The Atmosphere and Ionosphere

Physics of Earth and Space Environments

The series Physics of Earth and Space Environments is devoted to monograph texts
dealing with all aspects of atmospheric, hydrospheric and space science research
and advanced teaching. The presentations will be both qualitative as well as
quantitative, with strong emphasis on the underlying (geo)physical sciences.
Of particular interest are
l contributions which relate fundamental research in the aforementioned fields
to present and developing environmental issues viewed broadly
l concise accounts of newly emerging important topics that are embedded in a
broader framework in order to provide quick but readable access of new
material to a larger audience
The books forming this collection will be of importance for graduate students and
active researchers alike.

Series Editors:
Rodolfo Guzzi Louis J. Lanzerotti
Responsabile di Scienze della Terra Bell Laboratories, Lucent Technologies
Head of Earth Sciences 700 Mountain Avenue
Via di Villa Grazioli, 23 Murray Hill, NJ 07974, USA
00198 Roma, Italy
Ulrich Platt
Dieter Imboden
Ruprecht-Karls-Universität Heidelberg
ETH Zürich
Institut für Umweltphysik
ETH Zentrum
Im Neuenheimer Feld 366
8092 Zürich, Switzerland
69120 Heidelberg, Germany

For other titles published in the series, go to

www.springer.com/series/5117
Vladimir L. Bychkov l Gennady V. Golubkov l

Anatoly I. Nikitin
Editors

The Atmosphere and

Ionosphere
Dynamics, Processes and Monitoring

With 59 Figures and 8 Tables

Editors
Dr. Vladimir L. Bychkov Prof. Gennady V. Golubkov
Lomonosov Moscow State University Russian Academy of Sciences
Fac. Physics Semenov Inst. Chemical
Leninskie Gory 1 Physics
Moscow Kosygina St. 4
Russia 119991 Moscow
[email protected] Russia 117977
[email protected]

Dr. Anatoly I. Nikitin

Russian Academy of Sciences
Inst. Energy Problems of Chemical
Physics
Leninsky Prosp. 38, build. 2
Moscow, Russia 117829
[email protected]

ISSN 1610-1677 e-ISSN 1865-0678

ISBN 978-90-481-3211-9 e-ISBN 978-90-481-3212-6
DOI 10.1007/978-90-481-3212-6
Springer Dordrecht Heidelberg London New York
Library of Congress Control Number: 2010927479

# Springer Science+Business Media B.V. 2010

No part of this work may be reproduced, stored in a retrieval system, or transmitted in any form or by any
means, electronic, mechanical, photocopying, microfilming, recording or otherwise, without written
permission from the Publisher, with the exception of any material supplied specifically for the purpose
of being entered and executed on a computer system, for exclusive use by the purchaser of the work.

Cover design: eStudio Calamar S.L., F. Steinen-Broo, Pau/Girona, Spain

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Abstract

The first part of the book is devoted to the modern methods for calculating the
energy eigenvalues of Rydberg atoms A** and molecules XY** perturbed by
neutral particles of a medium and to the results of studying the interaction processes
with them. Interest in this study is caused by numerous applications in plasma
chemistry, aeronomy, and astrophysics.
The second part of the book is devoted to the atmospheric aerosol – one of the
most important factors affecting the Earth climatic and weather conditions. The
study of the mechanisms of formation and evolution of atmospheric aerosols is of
primary importance for predictions of the climatic changes on our planet. Special
attention is given to the last achievements in theory of particle formation and their
subsequent growth.
The third part of the book is devoted to numerous phenomena occurred in the
mesosphere, ionosphere and the magnetosphere of the Earth caused by the sources
located in the lower atmosphere and on the ground. Effects produced by lightning
activity and by ground-based transmitters operated in high frequency and very low
frequency ranges are described.
The fourth part of the book is devoted to modern methods of earthquake
prediction. First section contains first results of special satellite “COMPASS 2”
destined for detection of seism-electromagnetic effects. A whistler group in higher-
order guided mode was recorded. Probably it was propagating between two layers,
caused by onion-like structure of in homogeneities in the plasmasphere. Extremely
low and very low frequency effects observed over seism-active regions by the
satellite “INTERCOSMOS-24” are considered.
The achievements of the basic researches of the upper atmosphere and iono-
sphere processes with the mathematical modeling methods are briefly presented in
the fifth part of the book. The mathematical problem of the model atmosphere/
ionosphere description and existing global theoretical model of environment and
results of the investigations with their using are considered.

v
vi Abstract

The last part of the book is devoted to ball lightning investigations. Researches
historical review is presented. They consist of gathering and data processing of
observations, experiments on reproduction of long-lived shining formations in
electric discharges, and theoretical models. Detailed descriptions of three high-
energy ball lightning models are presented.
About Editors

Vladimir L. Bychkov is a leading researcher of the physical department of

Lomonosov Moscow state university; he is also head of a laboratory of Moscow
Radiotechnical Institute, Russian Academy of Sciences. He received MS degree in
theoretical physics from the University of Peoples Friendship, Moscow, Russia, a
Ph.D. in plasma physics and chemistry from Kurchatov Institute of Nuclear Energy,
and Doctor of Sciences degree from Moscow Regional State University. He has
35 years of experience in plasma physics researches, namely, in the physics of
elementary processes, gas discharges, electron-beam plasmas, plasma chemistry,
and ball lightning. He is head of the Russian Committee on ball lightning, vice
president of the International Committee on Ball Lightning, and a member of
Moscow Physical Society.
Anatoly I. Nikitin is a principal researcher of Institute for Energy Problems of
Chemical Physics, Russian Academy of Sciences, Moscow, Russia. He received a
MS degree in physics from the Lomonosov Moscow state university, a Ph.D. in
radiophysics from Lebedev Physical Institute, Russian Academy of Sciences,
Moscow and Doctor of Sciences degree from Karpov Institute of Physical Chemistry,
Moscow. He has 50 years of experience in quantum electronics (masers, lasers)
researches, chemical physics, selective by isotopes chemical reactions, plasma
physics, laser chemistry, and ball lightning. He is Secretary of the International
Committee on Ball Lightning and a member of the Moscow Physical Society.
Gennady V. Golubkov is a leading scientist of Semenov Institute of Chemical
Physics, Moscow, Russian Academy of Sciences. He received MS degree in theo-
retical and nuclear physics from Moscow Engineering Physical Institute, Moscow,
Russia, Ph.D. in chemical physics in Semenov Institute of Chemical Physics, and
Doctor of Sciences degree in chemical physics from the Semenov Institute of
Chemical Physics. He has 40 years of experience in the quantum scattering theory,
the theory of atom-molecular processes and elementary chemical reactions, chemi-
cal physics of atmosphere, and of low temperature plasma. He is the author of a
monograph “Rydberg states of atoms and molecules and the elementary processes

vii
viii About Editors

with their participation” (2001). Golubkov is a Member of the Bureau of Moscow

and National Russian Physical Society, Actual member of New York Academy of
Sciences, Member of International Committee on Theoretical Chemistry, and Chair
of International Advisory Committee of International Conference “Atmosphere,
ionosphere, Safety” (Kaliningrad, Russia, 2008).
Introduction

From July 7 to 12, 2008 in Zelenogradsk, a cosy resort on the bank of the Baltic
Sea near Kaliningrad in Russia, the 1st International Conference “Atmosphere,
Ionosphere, Safety (AIS-2008)” has been carried out. The State Russian University
of I. Kant, Semenov Institute of chemical physics of the Russian Academy of
Sciences, Pushkov Institute of terrestrial magnetism and radio-waves propagation
of the Russian Academy of Sciences, and Russian Committee on Ball Lightning
(BL) have acted as organizers of the conference. Financial support was made by
Russian Fund of Fundamental Research Project N. 08-03-06041 and European
Office of Aerospace Research and Development Grant award FA8655-08-1-5052.
The International conference “Atmosphere, Ionosphere, Safety” (AIS-2008) was
devoted to (i) the analysis of the atmosphere–ionosphere response on natural and
man-made processes, the reasons of occurrence of the various accompanying
geophysical phenomena, and an estimation of possible consequences of their
influence on the person and technological systems; (ii) the study of the monitoring
possibility and search of the ways for the risk level decrease. Discussion of the
physical and chemical processes accompanying the observable geophysical phe-
nomena was undertaken.
One can see from a list of the Conference sections that questions of safety took
only rather modest place, so main topics of the Conference became discussion of
processes taking place in the atmosphere, ionosphere and methods of monitoring
these processes.
At carrying out of the Conference besides plenary sessions, five sections worked
in parallel: (A) Dynamics of atmospheric aerosols; (D) dynamics of an ionosphere
and atmosphere – their communication through an ionosphere; (E) elementary
processes in the upper atmosphere and the ionosphere; (P) the electromagnetic
and optical phenomena in atmosphere, including long-lived and plasma objects and
ball lightning; and (S) information systems of environment monitoring and preven-
tion of incidents. At the Conference, nine plenary reports, 65 reports on sections and
40 poster reports have been presented.
The analysis of reactions in system “atmosphere–ionosphere” and influences
of natural and technogenic processes on them was the basic question brought for

ix
x Introduction

discussion of conference participants. In this connection, considerable attention has

been given to the study of reasons and cases of the various geophysical and
atmospheric phenomena display, an estimation of their influence on people and
technological systems, development of systems of monitoring, and decrease in risk
of negative influence of natural processes on mankind ability to live.
The physical and chemical phenomena proceeding in the upper atmosphere and
ionosphere occur in the conditions and the scales that are not available in usual
laboratories. Moreover, it is possible to create such nonequilibrium conditions to
study the response of an environment on the external perturbations, the realization
of which is difficult on the earth in general. All atmospheric layers interact among
themselves by means of various physical and chemical processes, forming the com-
plex system subject to influences of flashes on the Sun, earthquakes on the Earth, man-
caused catastrophes, etc. The primary goal of theoretical and experimental researches
consists in revealing interrelations of dynamics of various atmospheric layers, para-
meters of the atmosphere and ionosphere, an establishment of a role of various
physical factors, in studying, understanding, and, finally, forecasting of dynamics of
the environment in development of external perturbations. Research of these phenom-
ena is impossible without the deep analysis of features of interaction of participating
particles and also a detailed study of the elementary chemical processes occurring
here. The trustworthy information is necessary for their solution about reactionary
ability of the excited particles, about activation efficiency of various freedom degrees
of interacting reagents that requires development of absolutely new techniques of
measurements. In turn, it leads to the necessity of improving the existing theory that
would not only qualitatively but also quantitatively explain observable laws. The last
puts forward not trivial problems which at first sight seem in general insoluble for
theorists. Overcoming of difficulties arising here is probably possible only within the
limits of essentially new theoretical approaches different from traditional methods of
quantum chemistry, though substantially using its achievements. Now we have
extensive data on an electronic structure of atoms and molecules and about dynamics
of their interaction. Sometimes, it is reliable to calculate cross sections and rates of
elementary chemical processes because in most cases (and especially with participa-
tion of the electronically excited fragments), we do not have any trustworthy informa-
tion about features of these particles interaction acts.
One of the most reliable and effective enough tools for their studying is fast
molecular beams. This area of science has intensively developed since last 20 years,
thanks to its numerous practical applications – from space programs to problems of
ecology and chemical technology. Advantage of fast beams is connected first of all
with the possibility of carrying out the reagents relative speeds variation measure-
ments in a wide range – from thermal to high (/ 1010–1011 m/c). Thus, in crossed
(or combined) beams, the recording resolution of counter particles relative energy
can reach / 103 eV that opens possibility of a detailed study of elementary
chemical processes (including determination of endothermic reaction threshold, a
role of initial excitation for the reaction rate and course, etc.).
One of the major processes taking place in the Earth upper atmosphere is dissocia-
tive recombination (DR) of slow electrons and molecular ions. Researchers who study
Introduction xi

ionized states have to solve problems relating to various fields of physics. Many of
these problems are associated with the microscopic properties of plasma, that is, with
states of atomic and molecular species and with the elementary processes involving
them. These properties depend substantially on the presence of positively charged
molecular ions, since reactions involving these species even at low concentrations can
lead to a noticeable increase in the rate of the volume charge disappearance in
decaying plasma. The latter is accompanied, as a rule, by the formation of excited
atomic fragments followed by light emission. Thus, the recombination of electrons
and ions determining the ionization structure of the plasma and the recombination
spectrum, in turn, provides the necessary information on the physical conditions in the
medium in which the emitting species are located. However, the measurement of
partial cross-sections of recombination using the cross beam technique is complex due
to the fact that for beams, there are no reliable methods for detecting the initial and
final states of the recombining system. Therefore, the observed cross-sections are
averaged over the energy distribution in electron beams and over vibrational and
rotational states, which hampers the direct comparison of experimental and theoretical
results. Recently, results of new experiments where DR was studied in storage rings
were obtained. The measurements undertaken in the presence of external laser radia-
tion can also play an important role and allow control of a reaction course.
Atmospheric aerosols – particles suspended in air – play an extremely important
role in the “metabolism” of the atmosphere. Despite their very low mass concen-
trations and extremely small sizes they remain active agents in the atmospheric
chemical cycles and in the energy transfer in the atmosphere. The small sizes of the
aerosol particles (comparable to the molecular mean free path) make their physics
and chemistry quite unusual. To answer the questions, where are these particle
from? What are the mechanisms of their interaction with the atmospheric air and the
Sun radiation? How do they affect the weather condition and what is their role in the
climate changes?, etc., this is a far from complete list of the aerosol problems. Part
of this book devoted to aerosols intoduces the readers to the area of these problems.
As an essentially new method of elementary physical and chemical processes,
research can be done by carrying out laboratory measurements to circumterraneous
space with the use of techniques of active (radiating) influence and space com-
plexes. In these conditions (when the measuring device is placed in the reacting
environment), there are no difficulties with vacuum of the high resolution, no
foreign impurity, and so on.
The indicated problems are of interest for a wide range of the investigators
working in various areas of science and techniques. At the same time, it is necessary
to carry out the additional researches that are connected with the high human activity
in the atmosphere–ionosphere system, leading to occurrence of new risks. They
concern an active development of the manned and uninhabited orbital systems,
aircrafts (using height of an average atmosphere), new kinds of communication,
long-distance transmission circuits, etc. Non-stationary atmosphere–ionosphere
system is the subject of powerful natural affects. Its bottom level is disturbed by
earthquakes, volcanic eruptions, typhoons, thunderstorms, etc. From above, it is
influenced by the geomagnetic storms. As a result of these processes such disturbing
xii Introduction

factors, as powerful atmospheric perturbations, electric currents, electromagnetic

radiations in the various spectrum ranges, the plasma and optical disturbances, an
accelerated particles, the increased level of radioactivity, and changing of ionic and
molecular components are realized. Besides, the microwave radiation of highly
excited particles of the ionosphere accompanying the processes of solar activity
increase and the appearance of magnetic storms impacts mankind negatively. Its
spectrum, apparently, is completely defined by the neutral ionosphere components.
The knowledge of the influencing factor origination allows to use them for disaster
monitoring and to create the corresponding techniques on this base.
For low atmosphere, investigation of thunderstorm activity represents substantial
scientific and practical interest, in particular, such an uninvestigated phenomenon is
ball lightning. One can say that the nature of usual linear lightning is understood to
some extent and there are means of surface objects protection; however, the same
cannot be said about ball lightning. This state cannot be considered as acceptable since
in a number of cases destructions caused by the ball lightning are as serious as those
caused by the linear lightning and sometimes results of its impact are unpredictable
(ruination of aircraft, explosions in industrial objects). Besides, investigation of ball
lightning is interesting from the point of view of physics and power, nature at its
example demonstrates a possibility of high-energy density concentration and storing.
During AIS-2008 conference the tenth jubilee Symposium on Ball Lightning
was carried out at the electromagnetic and optical phenomena in the atmosphere
section. It was an occasion to make the review of Ball Lightning (BL) researches
history and, in particular, the work analysis for last 20 years. It can be seen that
these years were the time of active researches of a BL problem. Data banks of BL
observations collected and replenished; experiments on obtaining and research of
long-lived shining formations were carried out, works on creation and check of BL
models were conducted. Unfortunately, this activity has yet not brought notable
results. The reason for failure, apparently, is implied in the fact that we could not
choose “the main link” in properties BL and as a result have incorrectly chosen
ways of its experimental modeling. In the field of BL observation data collection
there was some saturation: new data practically add nothing to a “portrait” of an
average BL which for some reason yet has not helped to create adequate model of
this phenomenon. But there is a shortage of information on rare BL properties: its
high energy manifestation, capability to penetrate through subjects (glass, compos-
ite materials) connection with other geophysical events. Likely time has come to
pass to the publication of full descriptions of BL observations. Quite probably,
among them, there can be data that can become the “key” opening its secrets. Time
has come to seriously consider data file of UFO observations as among them more
than half of objects possess properties of BL. It is necessary to realize, that BL
science is the interdisciplinary science requiring participation of experts, working
in various areas of physics, chemistry, power, synergetic, biology and psychology.

V.L. Bychkov
G.V. Golubkov
A.I. Nikitin
Contents

1 Rydberg States of Atoms and Molecules in a Field

of Neutral Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
G.V. Golubkov, M.G. Golubkov, and G.K. Ivanov

2 Formation of Aerosols in the Atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

A.A. Lushnikov, V.A. Zagaynov, and Yu.S. Lyubovtseva

3 Atmosphere–Ionosphere Electrodynamic Coupling . . . . . . . . . . . . . . . . . . . . 97

V.M. Sorokin and V.M. Chmyrev

4 “COMPASS 2” Satellite and Ground-Based Experiments . . . . . . . . . . . 147

Yu.M. Mikhailov, V.D. Kuznetsov, C. Ferencz, L. Bodnar,
V.E. Korepanov, G.A. Mikhailova, L.P. Korsunova, V.V. Khegay,
S.E. Smirnov, and O.V. Kapustina

5 Mathematical Modeling of Dynamics Processes

in the Upper Atmosphere and Ionosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
I.V. Karpov and G.V. Golubkov

6 Ball Lightning Investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201

V.L. Bychkov, A.I. Nikitin, and G.C. Dijkhuis

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375

xiii
Contributors

L. Bodnar Bl-electronics, Solimar, Hungary, [email protected]

V.L. Bychkov M.V. Lomonosov Moscow State University, Moscow, Russia
V.M. Chmyrev Institute of Physics of the Earth, Russian Academy of Sciences,
10, Bolshaya Gruzinskaya Str., Moscow, 123995 Russia
G.C. Dijkhuis Department of Mathematics and Informatics, Eindhoven
University of Technology, P.O Box 513, 5600 MB Eindhoven, The Netherlands
C. Ferencz Space Research Groupe, Eotwös University, Budapest, Hungary,
[email protected]
G.V. Golubkov Semenov Institute of Chemical Physics, Russian Academy of
Sciences, Moscow, Russia
M.G. Golubkov Semenov Institute of Chemical Physics, Russian Academy of
Sciences, Moscow, Russia
G.K. Ivanov Semenov Institute of Chemical Physics, Russian Academy of
Sciences, Moscow, Russia
O.V. Kapustina Pushkov Institute of Terrestrial Magnetism, Ionosphere and
Radio Wave Propagation, Russian Academy of Sciences, IZMIRAN, Troitsk,
Moscow region, 142190 Russia, [email protected]
I.V. Karpov West Department IZMIRAN, Russian Academy of Sciences,
Kaliningrad, Russia
V.V. Khegay Pushkov Institute of Terrestrial Magnetism, Ionosphere and Radio
Wave Propagation, Russian Academy of Sciences, IZMIRAN, Troitsk, Moscow
region, 142190 Russia, [email protected]
V.E. Korepanov Center of Cosmic Investigations of NASU and NCAU,
Naukova 5, Lviv, 79000 Ukraine, [email protected]

xv
xvi Contributors

L.P. Korsunova Pushkov Institute of Terrestrial Magnetism, Ionosphere and

Radio Wave Propagation, Russian Academy of Sciences, IZMIRAN, Troitsk,
Moscow region, 142190 Russia, [email protected]
V.D. Kuznetsov Pushkov Institute of Terrestrial Magnetism, Ionosphere and
Radio Wave Propagation, Russian Academy of Sciences, IZMIRAN, Troitsk,
Moscow region, 142190 Russia, [email protected]
A.A. Lushnikov Karpov Institute of Physical Chemistry, Moscow, Russia
Yu.S. Lyubovtseva Geophysical Center of Russian Academy of Sciences,
Moscow, Russia
Yu.M. Mikhailov Pushkov Institute of Terrestrial Magnetism, Ionosphere and
Radio Wave Propagation, Russian Academy of Sciences, IZMIRAN, Troitsk,
Moscow region, 142190 Russia, [email protected]
G.A. Mikhailova Pushkov Institute of Terrestrial Magnetism, Ionosphere and
Radio Wave Propagation, Russian Academy of Sciences, IZMIRAN, Troitsk,
Moscow region, 142190 Russia, [email protected]
A.I. Nikitin Institute for Energy Problems of Chemical Physics, Russian
Academy of Sciences, Moscow, Russia
S.E. Smirnov Institute of Cosmophysical Research and Radio Wave Propagation
Far-East Branch of Russian Academy of Sciences, IKIR, Paratunka, Kamchatka
region, 68300 Russia, [email protected]
V.M. Sorokin Pushkov Institute of Terrestrial Magnetism, Ionosphere and Radio
Wave Propagation, Russian Academy of Sciences, IZMIRAN, Troitsk, Moscow
Region, 142190 Russia
V.A. Zagaynov Karpov Institute of Physical Chemistry, Moscow, Russia
Chapter 1
Rydberg States of Atoms and Molecules
in a Field of Neutral Particles

G.V. Golubkov, M.G. Golubkov, and G.K. Ivanov

Abstract This book is devoted to the modern methods of calculating the energy
eigenvalues of Rydberg atoms A

and molecules XY

perturbed by neutral parti-
cles of a medium and to the results of studying the interaction processes with them.
Numerous applications in plasma chemistry, aeronomy, and astrophysics have
contributed to conducting this study. These methods are based on the use of integral
variant of the theory utilizing Green’s function approach. Because of the closeness
to the continuum boundary, these energies cannot be properly described in terms of
the standard quantum chemistry. When the radius of electronic cloud of excited
states is large enough (i.e., Rc ¼ 2 n2  1, where n is the principal quantum number),
they cannot be regarded as isolated even in the case of a rarefied gas. The spectral
distortion is the strongest when the number N of perturbing neutral particles
falling into this region exceeds unity. This chapter is divided into four main parts.
In the first part, the generalized method of finite-radius potential (FRP) is
discussed. This method self-consistently takes into account the short- and long-
range interactions in the two-center system under consideration. It adequately
describes the scattering of a weakly bound electron by the ion core and a perturbing
atom with nonzero angular momenta l and L with respect to these centers, thereby
allowing the theory to be extended to the intermediate (on the order of and less than
electron wavelength l / n) interatomic distances R. As an application of the
theory, the detailed analysis is performed for the behavior of the potential energy
surfaces (PESs) of a system composed of a highly excited atom A

ðn; lÞ and a
neutral atom B with the filled electronic shell. It is demonstrated that the inclusion
of nonzero momentum L for the e  B scattering results in the additional splitting
of the PES into the separate groups of interacting terms classified by the projection
m of electronic angular momentum l on the quasimolecular axis. At distances
R  n, the FRP method exactly transforms to the zero-radius pseudopotential

G.V. Golubkov (*)

Semenov Institute of Chemical Physics, Moscow, Russia

V.L. Bychkov et al. (eds.), The Atmosphere and Ionosphere, 1

Physics of Earth and Space Environments,
DOI 10.1007/978-90-481-3212-6_1, # Springer ScienceþBusiness Media B.V. 2010
2 G.V. Golubkov et al.

(ZRP) model and, correspondingly, to the asymptotic theory in which the PESs
acquire a simple analytic form. It turns out that, at large values n  1, the ZRP
method is valid up to the distances R  n.
In the second part, the specific features of the diabatic and adiabatic PESs are
discussed by taking into account the dissociative, covalent, and ion configurations.
The potentialities and disadvantages of the existing ab initio approaches are ana-
lyzed. The matching method is suggested, which allows a unique self-consistent
picture devoid of the above-mentioned disadvantages to be obtained for the terms.
As an illustration, the potential curves are calculated for the nl ð2sþ1 LÞ states of the
Na

þ He quasimolecule (n; l , and L are the principal quantum number, angular
momentum, and its projection on the molecular axis, respectively, and S is the spin
of the system), and a detailed comparison with the computational results of other
authors is carried out.
In the third part, the possible applications of the theory to the shock ionization,
excitation, and quenching processes are discussed for the Rydberg states (RSs).
Among these are also the simplest dissociation, exchange, and charge exchange
reactions. They can be schematically represented as

XY

þ M ! XY

þ M; (1.1a)

XY

þ M ! X þ Y
þ M; (1.1b)

XY

þ M ! XM þ Y
; (1.1c)

XY

þ M ! XY þ þ e þ M; (1.1d)

XY

þ M ! XY þ þ M (1.1e)

The interaction of XY

with a neutral particle M includes the interactions with
both ion and a weakly bound electron. The former is characterized by small impact
parameters, whereas the latter has large impact parameters. As a result, the total
scattering cross-sections can appreciably exceed the gas-kinetic values. The mate-
rial is presented in terms of the PES of the XY

þ M system followed by the
description of the dynamics of processes (1.1a–e) within the framework of the
integral variant of the multichannel quantum defect (MQD) theory using the renor-
malized Lippmann–Shwinger equation technique. Such a formulation of the MQD
theory allows one to obtain a convenient representation for Green’s function of a
highly excited molecule, which opens up wide possibilities for various applications.
In the fourth part, the many-center perturbation of the atomic Rydberg states is
analyzed for the situation wherein two (or more) perturbing neutral centers fall
inside the electronic cloud. The behavior of Rydberg atom in a dense medium is
considered with allowance for the influence of finite number N of the neutral
particles chaotically distributed in its volume. The stochastic approach is proposed
for the solution to this problem.

Keywords Atommolecular processes  Elementary chemical reactions

1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 3

1.1 Introduction

The Rydberg states (RSs) are situated near the continuum boundary and are
characterized by the presence of a weakly bound electron with the orbit size on
the order of n2. These states are prone to the strong influence of, practically, any
external action: constant electric and magnetic fields [1], laser radiation [2], and the
surrounding medium [3] (including gas, liquid, or solid). This is important for the
study of the processes occurring in the upper atmosphere, in low-temperature plasma
and gaseous laser systems, for the study of the quantum size effects in nanostructures,
reactions involving excited particles at solid surfaces, etc. Researchers are interested
in these phenomena owing to the development of the fundamentally new experimen-
tal methods and their possible applications. Among these are the photoelectron
spectroscopy of superslow electrons [4–7], storage rings [8–11], and scanning tunnel
spectroscopy [12–14].
The spectral distortion in Rydberg atom A

by the ground-state neutral atom B
is the simplest situation modeling the presence of the surrounding medium. The
corresponding influence is assumed to be bordered by the sphere of radius R
,
which exceeds the classical radius Rcl ¼ 2n2 by several wavelengths l  n. The
region r > 2 n2 is classically forbidden for electron. Next, in the order of complex-
ity are quantum systems including N perturbing atoms in the sphere of radius R
,
with N  2.
Because of the difference in masses of electron and particles Aþ and B, the
energy-exchange processes in the A

 B collisions are unlikely. For this reason,
the investigation of the collisional dynamics, as a rule, is preceded by the cal-
culation of the potential energy surface (PES) of a united system with the aim
of revealing those regions of mutual particle arrangement, where the cor-
responding terms quasicross or draw close together. If the electron wavelength
l changes only slightly on the scale of the effective interaction region r0 with the
neutral particle,

dl r0
r0 l or l
dr ð kR Þ2

(k is the classical electron momentum), then the e  B interaction can be

described in terms of the free scattering approach.
The properties of PES and the dynamics of collisional processes in the simplest
(N ¼ 1) system are described in many publications, which are discussed in detail in
[15]. In most works, the approaches based on the information about the free-
electron scattering by the independent fragments are used, namely by the ion core
and atom B. Among the most elaborated approaches is the asymptotic treatment,
which is valid at sufficiently large interatomic distances R  n2 , where the main
theoretical results can be represented in a simple analytic form. An important fact
is that the wave functions and the corresponding Green’s functions describing
electron behavior in the field of ion Aþ are well known.
4 G.V. Golubkov et al.

On the other hand, this circumstance serves as a necessary base for developing
the theory of electronic structure for nanoclusters with ordered or disordered
structure, where the characteristic interparticle distances are of few angstroms.
This is also true for the electron-excited states of atoms and molecules adsorbed
at the solid surface. The use of the theory of multiple scattering for the solution of
the problem on perturbation of highly excited atomic states by the system of
chaotically or orderly arranged interaction centers within the framework of asymp-
totic approach was considered in [15, 16].
For the two-center axially symmetric system “Rydberg atom A

ðnlÞ – atom B”,
this problem was solved using the generalized zero-radius pseudopotential (ZRP)
model [17] for the electron S-scattering by the perturbing atom B at a sufficiently
large n. Taking into account the long-range interaction at “intermediate” distances,
R brings about the redetermination of the potential scattering length for e  B and
the removal of the degeneracy of Rydberg levels with respect to the projections m of
electronic angular momentum l on the quasimolecular axis. An important part is
played by the centrifugal potential, which reduces the number of degenerate
Coulomb states by forming covalent terms and violates the quasiclassical descrip-
tion used for the electron motion in the asymptotic theory [18–20].
At small values of the principal quantum number n, one should include higher
harmonics to describe electron scattering by the perturbing center. In [21], the
method of constructing one-center operators for the e  B scattering was devel-
oped, allowing the calculation of the PESs for the Rydberg atom A

ðnÞ – atom B
system at intermediate distances between atoms and for small values of n. The
method is based on the use of finite-radius potential (FRP) including nonzero
harmonics of the scattered electron and consistently takes into account the short-
range and long-range interactions of particles at these distances. At large values,
n  1, the ZRP results are reproduced in a natural way, and, at distances, R > n,
they exactly transform into the asymptotic theory. As an illustration, the terms were
calculated for the Na

ðnlmÞ þ He system, and comparison with the existing data
was carried out.
Thereupon, a systematic analysis of the strong nonadiabatic coupling between
the molecular Rydberg states in the collisional processes was carried out. We
discuss the main structural features of the vibronic PESs describing the interaction
of highly excited atoms and molecules with neutral particles. The asymptotic
method given here is an alternative to the traditional quantum chemical calculations.
It does not use the variational principle, while the PES of the system is determined
on the basis of the algebraic equations containing information on the free-electron
motion in the field of isolated fragments (of ion and neutral particle). The asymptotic
theory deals with the scattering T and K matrices, which are fundamental charac-
teristics in the quantum collision theory, even as the parameters appearing in it
can be directly associated with the experimentally observed electron-scattering
amplitudes (or phases) by ions, atoms, and molecules. The unique feature of the
asymptotic method is that it allows the multisheet PES to be represented in a simple
analytic form including the whole diversity of merging points and quasicrossing
regions responsible for the nonadiabatic transitions in the system.
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 5

The objects considered in our work also possess remarkable properties. The terms
split-off from the group of orbitally degenerate states of Rydberg atom X
or
molecule XY

demonstrate anomalous behavior. These terms are virtually indepen-
dent of the nature of Rydberg particle (if the perturbation is produced by molecule M,
the dependence on its orientation also vanishes). An important property of these
(valence) configurations is that they are capable of autoionization owing to the
escape to continuum as the interacting fragments approach each other. The weakly
bound states can also form in the valence configurations of Rydberg complexes. The
mechanism of their formation is caused by the Ramsauer effect in the elastic electron
scattering from particle M. They arise due to the presence of shallow, though rather
wide, hollows in the potential curve, where the weakly bound states can exist in the
Rydberg complex (of the X

B or XY

B type) at large distances from the ion core.
New features appear in the processes involving Rydberg molecules XY

. They
are caused by the strong nonadiabatic coupling between the states, i.e., by the fact
that each Rydberg nlNv – state (with small l) is a superposition of, at least, two (or
greater, if the rotation is taken into account) series closely spaced in v. This gives
rise to a sharp (resonance) n dependence of the cross-sections for the elementary
processes (even after averaging over the relative velocities of colliding particles),
which is particularly important in the analysis of kinetic phenomena under non-
equilibrium conditions.

1.2 Spectral Structure of the A

– B System

A highly excited atom A

(nlm) is electrically neutral and represents an Aþ sur-
rounded by the extended electronic cloud whose sizes R n2 appreciably exceed
the sizes of the unexcited atom. Here, l and m are, respectively, the angular
momentum of the Rydberg electron and the projection of angular momentum on
a certain axis, which can naturally be directed along the vector R connecting the
centers of gravity of particles A

and B in the A

– B system. The binding energy
of this electron is considerably lower than the atomic ionization potential I0 ,

1
En ¼  ; j E n j  I0 (1.2)
2 n2
(hereafter, atomic units are used, i.e.,

h ¼ e ¼ me ¼ 1). At large n2  1, a certain
electronic angular momentum l
exists, which, owing to the presence of centrifugal
barrier, separates the states strongly and weakly penetrating into the ion core (in
most atoms, l
¼ 3). The states with small angular momenta l < l
strongly interact
with the ion core and contain quantum defects ml that give rise to the shift Dnl of
Rydberg levels from the Coulomb levels En ; i.e.,

1 ml
Enl ¼  2
¼ En þ Dnl ; Dnl ffi : (1.3)
2 ð n  ml Þ n3
6 G.V. Golubkov et al.

The remaining levels l  l
are orbitally degenerate Coulomb states (with

ml ¼ 0), because quantum defects decrease as ml / l5 with increasing l [22].
The large sizes and multitude of energy-degenerate states are the characteristic
features of the system under consideration. Since the atom B occupies a small part
of volume in which electron moves, we will assume that the effective radius r 0 of
the e  B interaction is smaller than the electron wavelength,
pffiffiffi
r0 < l / R: (1.4)

In essence, condition (1.3) implies the transition to the intermediate distances,

1  R  n2 ; (1.5)

where (as also in the asymptotic region R / n2 ) electron moves with the total
energy j E n l j  1 in a combined field

1 lðl þ 1Þ
UðRÞ ¼  þ ;
r 2r 2

of the Coulomb and centrifugal potentials free from ion core.

Since the strong interaction Ve B of electron and atom B is concentrated near the
ðRÞ
point R inside the region of radius r 0 , the wave function F s ðrÞ of the Rydberg atom
A

changes only weakly on this scale. It is normalized to unity and has the form [3]

ðRÞ
FðRÞ
s ðrÞ ¼ Rnl ðrÞ Ylm ðr=rÞ; (1.6)
pffiffiffiffiffiffiffiffiffiffi
where r is the coordinate of the weakly bound electron, s ¼ fnlmg, n ¼ 1= 2E is
the effective principal quantum number, and Ylm ðr=rÞ is the spherical harmonic
[23]. The radial part is defined by

ðRÞ Wn; lþ1=2 ð2r=nÞ

Rnl ðrÞ ¼ (1.7)
rn ½Gðn  lÞ Gðn þ l þ 1Þ  1=2

(here, GðxÞ is the gamma-function and Wn; lþ1=2 ð2r=nÞ is the Whittaker function
[24]). Note that for the integer values n ¼ n, function (1.7) exactly coincides with
the Coulomb wave function.
The solution to the problem is based on the formalism of the reconstructed
Lippmann–Shwinger equations for the level-shift operator [15]

t ¼ Ke B ðGA

 G0 Þ t; (1.8)

where, after the separation of Green’s function of Rydberg atom A

into the
strongly and weakly energy-dependent parts,
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 7

X  ðRÞ ED
ðRÞ


GA

ðr; r0 ; EÞ ¼ pn 3 ðEÞ  Fn l m ðrÞ Fn l m ðr0 Þ cot p ðn ðEÞ þ m l Þ þ G0 ðr; r0 Þ
lm
(1.9)

the scattering matrix Ke B , which is defined for the kinetic energy of Rydberg
electron and exactly takes into account all the specific features of the e  B
interaction in the states entering into the function G0 , is introduced into the theory.
Formally, this operator satisfies the integral equation

K e B ¼ Ve B þ Ve B G 0 K e B ;

where Ve B is the operator for the e  B interaction. The determination of the Ke B
matrix is generally a challenge, because the action of operator Ve B can extend to a
rather wide region, where no unique analytic representation can be obtained for the
electron-perturbing particle potential. At the same time, the region of strong inter-
action, where the explicit representation is necessary for the function G0 , and the
region of a comparably weak interaction, where this function is not needed at all, can
always be found in the interaction potentials between an electron and the neutral
particles (atoms or molecules much smaller than the Rydberg atoms). Then, accord-
ing to the definition of Green’s function G0 , the Ke B -matrix elements must be
expressed through the characteristics of the electron scattering from atom B.
At small distances r  R from the perturbing atom, where the interaction Ve B
is the strongest, Green’s function G0 is chosen as follows [15]:
(
0 0 1 cos pe ðR Þ j r  r0 j; jEj < 1=R;
G0 ðr; r Þ ¼ g0 ðr; r Þ ¼ 
2pj r  r0 j exp½ae ðR Þ j r  r0 j; jEj > 1=R;
(1.10)

where pe ðRÞ ¼ ½2 ðE þ 1=RÞ 1=2 is the classical electron momentum at distance R

for energies jEj < 1=R, whereas the quantity ae ðRÞ is, correspondingly, determined
for j E j > 1=R and equals a e ðRÞ ¼ ½2 ð E þ 1=R Þ  1=2 .
Owing to the level degeneracy at l  l
(see comment to formula (1.3)), the pole
part of Green’s function (1.9) is separated into two parts. The first includes the
terms with ml 6¼ 0, which correspond to the nondegenerate states with l < l
. The
second part (with l  l
) contains the terms with ml ¼ 0. From the condition that
the determinant of the system of equations (1.8) must be zero and using Eqs. (1.9)
and (1.10) to write system (1.8) in the form
" #
X  ðRÞ E D ðRÞ  X  ðRÞ E D ðRÞ 
t ¼ K e B ’nlm ’nlm  cot p ðn þ ml Þ þ cot pn ’nlm ’nlm  t;
l<l
;m ll
;m
ðRÞ
pffiffiffiffiffiffiffiffiffi ðRÞ
’ nlm ¼ pn 3 F nlm; ð1:11Þ
8 G.V. Golubkov et al.

one obtains the PES of the combined system. The terms of the nondegenerate states
relate to the Rydberg configuration, while the remaining terms will be referred to as
covalent. When calculating the scattering Ke B matrix with allowance for the strong
interaction, one can use the fact that the properties of function (1.11) and Green’s
function of free electron are close and, hence, use the characteristics of the e  B
scattering observed in the beam experiments, at which point one can also include
perturbatively the long-range interaction. This greatly simplifies the procedure of
determining the corresponding matrix elements.

1.3 Finite-Radius Pseudopotential

The simplest method for the introduction of finite-radius pseudopotential into the
theory under condition (1.4) is proposed in [25]. The scattering amplitude is a
function of three variables: momenta pe ; p0 e , and the electron kinetic energy e. It is
ð0Þ
related to the Ke B -scattering operator by the expression

1 D ip0 e ðrRÞ  ð0Þ  ipe ðrRÞ E

Fðpe ; p0 e ; eÞ ¼  e Ke B e (1.12)
2p

or, in the representation of LM harmonics (where L is the electronic angular

momentum with respect to atom B, and M is its projection on the direction of
vector R),
X
Fðpe ; p0 e ; eÞ ¼ ð2l þ 1Þ FL ðpe ; p0 e ; eÞ PL ðcos YÞ
L
X ð0Þ
(1.13)


¼ 4p hLMj Ke B jLMi YLM ðpe =pe Þ YLM ðp0 e =p0 e Þ;
LM

where the wave functions jLMi are

pffiffiffi
jLMi ¼ 2 jL ðpe rÞ YLM ð r=rÞ (1.14)

jL ðxÞ is the L-order Bessel spherical function of the first kind, Y is the angle
between the vectors pe and p0 e , and r ¼ r  R. The observed scattering cross-
sections or amplitudes correspond to functions (1.12) and (1.13) determined at the
energy surface, i.e., under the condition

p2e p02e
e¼ ¼ > 0: (1.15)
2 2

Under condition (1.3), matrix elements (1.13) for slow electrons can be brought
to the separable form (k02 ¼ 2e),
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 9

 0 L  L
ð0Þ p ð0Þ pe
KL;L ¼ 2p e fL ðeÞ : (1.16)
k0 k0
ð0Þ
Here, fL ðeÞ is the free-electron elastic scattering amplitude, which is deter-
mined by [26]

ð0Þ k02L
fL ðeÞ ¼ ð1Þ ð2Þ
: (1.17)
kL þ cL e þ cL e2 þ    þ ik02Lþ1
ðsÞ
The expansion coefficients cL in Eq. (1.17) determine specific features of the
low-energy electron scattering and depend on the structure of potential created by
the perturbing particle. According to Eq. (1.17), an electron in the e  B system
can be bound at negative energies e in the states with L  1, whereas, at positive
energies, the quasistationary (resonance) levels can form.
ð0Þ
The Ke B operator can be represented in the following form [25]:

ð0Þ
X ð0Þ
K e B ¼ KL ;
L
ð0Þ ð0Þ
X (1.18)
KL ¼ 8p2 j2ejL fL ðeÞ dLM ðrÞ d
LM ðr0 Þ d ðr  RÞ ;
M

where the quantity

dLM ðrÞ ¼ YLM ðk=kÞ ði rr ÞL :

Then, the scattering-matrix elements in Eq. (1.18) are written as

X   
ð0Þ ð0Þ ðRÞ ðRÞ
Knlm; nl0 m ¼ 8p2 j2ejL fL ðeÞ dLM Fnlm d
LM F
nl0 m : (1.19)
LM

ðRÞ
The action of the operator dLM on the wave function Fnlm ðrÞ amounts to the
multiplication of the latter by ð4pÞ1=2 at L ¼ 0 and (at L  1) its differentiation at
ð0Þ
the point R. The scattering amplitude fL ðeÞ takes the form
8
< g1
L ðeÞ; r; r 0 <j2ej1 ;
ð0Þ
fL ðeÞ ¼ h pffiffiffiffiffiffiffiffi i (1.20)
: gL ðeÞ  2e 1 ; r; r 0 >j2ej1 ;

where gL ðeÞ is the integral function of e; its expansion in powers of e is

ð0Þ ð1Þ
ð2eÞL gL ðeÞ ¼ CL þ CL e þ    : (1.21)
10 G.V. Golubkov et al.

The inclusion of the states with L  1 perturbs the states with the nonzero
projections of electronic angular momentum m on the direction of vector R.
The effective e  B interaction can be represented as the sum of two terms
ð0Þ
Ve B ðr; RÞ ¼ V e B ðrÞ þ UAþ e B ðr; RÞ; (1.22)
ð0Þ
where the strong V e B interaction is concentrated inside a bounded region r  r 0 ,
where the wave functions of Rydberg electron change only weakly. Here, r and R
are the radius vectors of particles e and Aþ , measured from the center of gravity of
the perturbing atom B. The long-range portion of the force field UAþ e B contains all
polarization terms (including the polarization of atom B by electron) and can be taken
into account perturbatively. For the ground S state and the states of atom B with the
zero electronic orbital moment, the potential UAþ e B describes, in the case of LS-
coupling, the interaction with two oppositely charged particles. It has the form [27]
 
b r R 2 b b
UAþ e B ð r; RÞ ¼   ¼  4  4 þ DUAþ e B ð r; RÞ; (1.23)
2 r3 R 3 2r 2R

br R
DUAþ e B ð r; RÞ ¼ ;
ðrRÞ3

where b is the static polarizability of atom B.

The uncertainty in the separation of Ve B ð r; RÞ into the short-range and long-
range parts in Eq. (1.22) using the theory of perturbation for the interaction UAþ e B
can be eliminated under the following conditions:
 
ðb j E j; b R 2 Þ  1; b r 20 < 1: (1.24)

As a result, the scattering Ke B matrix takes the form

X ð0Þ
Ke B ð r; RÞ ¼ KL ð r; RÞ þ Uð r; RÞ; (1.25)
L

ð0Þ
where the partial operators
KL ð r; RÞ are
given by Eq. (1.18), and the long-range
part is Uð r; RÞ ¼ UAþ e B ð r; RÞ þ 2Rb 4  ðr  r0 Þ, where ðxÞ is the Heaviside
function. The corresponding matrix elements are written as
D X  E D   E
ðRÞ  ð0Þ  ðRÞ ðRÞ   ðRÞ
Knlm; n 0 l0 m ¼ F nlm  KL ð r; RÞ F n0 l0 m þ F nlm  U F n0 l0 m : (1.26)
L

(here, only elements with the specified value of projection m are nonzero). The first
term corresponding to L ¼ 0 in Eq. (1.26) is equal to
D   E
ðRÞ  ð0Þ  ðRÞ ðRÞ ðRÞ
F nlm  K0 ð r; RÞ F n0 l0 m ¼ 2pa0 F nlm ðRÞ F n0 l0 m ðRÞ; (1.27)
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 11

where a0 ¼ 1=k0 is the scattering length in the arbitrarily introduced short-range

ðRÞ
part of potential (1.22). Because the electron wave functions F nlm change
b
only weakly on the r0 scale, the integral for the polarization term  2r4 in potential
ðRÞ ðRÞ
(1.23) is equal to  2pðb=r0 Þ F nlm ðRÞ F nl0 m ðRÞ. This allows one to introduce the
quantity a ¼ a0  b=r0 into Eq. (1.26) (at small k0 , it exactly coincides with the
observed scattering length [27]) and recast matrix elements Knlm; n0 l0 m as
ðRÞ ðRÞ
Knlm; n0 l0 m ðRÞ ¼ 2pa F nlm ðRÞ F n0 l0 m ðRÞ dm0 þ D Knlm; n0 l0 m ðRÞ; (1.28)

where
D   E D X  E
ðRÞ   ðRÞ ðRÞ  ð0Þ  ðRÞ
DKn l m; n0 l0 m ðRÞ ¼ Fnlm  Uð r; RÞ Fn0 l0 m þ F nlm  KL ð r; RÞ F n0 l0 m :
L1
(1.29)

Note that, to an accuracy of terms / b=r20  a, this result does not depend on
r0 , so that the uncertainty in its choice vanishes [17]. In the quasiclassical condi-
tions of electron motion (at R / n2 ), expression (1.28) takes the form
D   E
ð0Þ ðRÞ
ðRÞ ðRÞ   ðRÞ
Knlm; n0 l0 m ðRÞ ¼ 2p f0 F nlm ðRÞ F n0 l0 m ðRÞdm0 þ F nlm  Uð r; RÞ F n0 l0 m ;
(1.30)
ð0Þ
where the amplitude f0 ðeÞ ffi ð a þ p bpe = 3Þ depends on energy and, for the
negative values of scattering length a, demonstrates the Ramsauer effect [28].
ð0Þ
ð0Þ d
Since, at small values of k0 , the amplitude is fL ðeÞ ffi kL0 / k02L [26], the
combination entering into Eq. (1.19) is written for L  1, according to Eqs. (1.20)
and (1.21), as
ð1Þ
1 CL
j2ejL fL ðeÞ ffi ð0Þ
 e; (1.31)
ð0Þ 2
CL ðCL Þ
ð0Þ ð0Þ
where the coefficient CL is related to the scattering phase dL by the expression
ð0Þ ð0Þ ð1Þ
dL ffi k02Lþ1 =CL . The constant CL is found from the energy dependence of phase
ð0Þ ð0Þ
dL ðeÞ. The corrections to the scattering amplitudes fL ðeÞ in the presence of long-
range interaction (1.23) are contained in the second term DKnlm; n0 l0 m in expression
(1.28). In the asymptotic region of distances, R / n2 , they are proportional to bk0 ,
i.e., e1=2 [27].
The mutual influence of the Rydberg and covalent states is significant only in the
vicinities of the points of possible mutual approach or term quasicrossings. According
to Eq. (1.11), the equation in this case takes the following form:
h ih i
ðRÞ ðcÞ
EU nlm ðRÞ E  U n0 Lm ðRÞ ¼ K 2nlm; n0 l0 m ðRÞ; (1.32)
12 G.V. Golubkov et al.

i.e., only states with the same projections can interact with each other. As a result,
ðRÞ ðcÞ
one has for the Rydberg U nlm and covalent U n0 Lm terms
h   i
ðRÞ 1  ðRÞ 2 b
U nlm ðRÞ ¼  2
þ 2 p a  F nlm ðRÞ  d m0 þ D Knlm; nl0 m ðRÞ  ;
2 ðn  ml Þ 2 R4
ðcÞ 1 b
U nlm ðRÞ ¼   ;
2 ½n  mnLm ðRÞ2 2 R4
1

mnLm ðRÞ ¼  arctan pn 3 KnLm; nLm ðRÞ ;

p
(1.33)

where the elements DKnlm;nlm are specified by expression (1.28), and the quantity
KnLm; nLm ðRÞ is
Xh ðRÞ
ðRÞ
i
KnLm; nLm ðRÞ ¼ 2pa F nlm ðRÞ F nl0 m ðRÞ dm0 þ D Knlm; nlm ðRÞ : (1.34)
l0 l

The interaction between these terms is determined by

 
D nlm; n0 Lm ¼ 2 K nlm; n0 Lm  cos 2 pn ðEÞ; (1.35)

where
Xh ðRÞ
ðRÞ
i
Knlm; n0 Lm ðRÞ ¼ 2pa F nlm ðRÞ F n0 l0 m ðRÞ dm0 þ D Knlm; n0 l0 m ðRÞ : (1.36)
l0 l

One can see that the quantum defect mnL ðRÞ induced in the covalent term by the
field of perturbing atom B is noticeably different from the corresponding expression
mL ¼ ape , which is obtained in the asymptotic theory at R / n2 [15].
The interaction between Rydberg terms is defined by
 
D nlm; n0 l0 m ¼ 2 K nlm; n0 l0 m  cos 2 pnðEÞ: (1.37)

Owing to the axial symmetry of the problem, only the states with different values
of angular momentum l and fixed projections m on the quasimolecular axis can
interact with each other. At some points (at energy E
c and distance R
c ), the Rydberg
terms can quasicross.
The covalent terms are found from the system of equations
X  ðRÞ E D ðRÞ 
t ¼ Ke B pn3 cot pn F nlm F nlm  t; (1.38)
l l

The unambiguous solution to this equation is found from

 
Det  d ll0 tan pnðEÞ  pn3 K nlm; nl0 m  ¼ 0: (1.39)
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 13

In the case of L ¼ 0 corresponding to the ZRP model, the nondiagonal elements

are separable
ð0Þ 2 ð0Þ ð0Þ
K nlm; nl0 m ¼ K nlm; nlm K nl0 m; nl0 m ; (1.40)

for which reason only one degenerate covalent term splits off from the Coulomb
levels [25].
According to Eq. (1.39), the inclusion of higher harmonics (L  1) in the ZRP
method should give rise to the branches containing N split-out covalent terms at
intermediate distances (R / n), each representing the l-mixed groups of states for a
given value m. This occurs because the separability
D  of nondiagonal elements
E
ðRÞ  P ð0Þ  ðRÞ
Knlm;nl0 m is broken because of the last term F nlm  KL ðr; RÞ F nl0 m in
L1
Eq. (1.28). Since all states, except m ¼ 0, are doubly degenerate with respect to
the m values, the total number of split-out covalent terms N ¼ 12 ðn  l
Þ
ðn þ 2l
 2Þ depends on the principal quantum number.
In the case of n ¼ 4, there are four noninteracting covalent terms with angular
momentum l
and projections m = 0–3. For the next value n ¼ 5, one has nine states.
They represent a group of four independent pairs (mixed with respect to l ¼ 3; 4 for
each of m = 0–3) of terms and one isolated term (with l ¼ m ¼ 4). In the case of
n ¼ 6, 15 terms appear, among which there are 4 independent triads (l ¼ 3–5; m ¼
0–3) of interacting terms, 1 interacting pair (l ¼ 4; 5 ; m ¼ 4), and 1 isolated
term with l ¼ m ¼ 5. For n ¼ 7, their number is 22, and they include four indepen-
dent groups of four interacting terms (l ¼ 3–6; m ¼ 4), one triad (l ¼ 4–6; m ¼ 4),
one pair (l ¼ 5 ; 6 ; m ¼ 5) of interacting terms, and one state with l ¼ m ¼ 6, etc.
Note also that, among each of the interacting groups containing ðn  l
Þ states
classified by the projection m, only the two first groups of states with m ¼ 0 and
m ¼ 1 are subjected to the strongest action. It is precisely these groups that are
of chief interest, because the remaining states are weakly perturbed and, hence,
are close to the Coulomb states. The shift of the terms with m ¼ 1 decreases
rapidly with increasing distance R between centers, because the corresponding
ð0Þ
matrix elements behave as hnl1j K1 jnl1i ¼/ R2nl ðRÞ R2 , whereas the elements
ð0Þ 
hnl0j K1 jnl0i ¼/ R2nl ðRÞ R decrease more slowly. For this reason, at distances
R  n, where a smooth transition to the ZRP model occurs, the S states are
perturbed most strongly.

1.4 PES of the A
– B System at Large Interatomic Distances

Taking large distances into consideration, the necessity arises of revealing the
correlation between the PES of the A

 B system at intermediate distances and
the corresponding terms in the asymptotic region R / n2 , where the role of centrif-
ugal barrier is insignificant and electron motion can be described using the quasi-
classical approximation in the plane-wave representation. For the Rydberg states
14 G.V. Golubkov et al.

corresponding to small values l < l
, no such problem exists. It arises for the PESs
of the covalent states (l  l
; m), because their number in Eq. (1.38) noticeably
increases with a decrease in R.
ðRÞ
The plane-wave expansion of the Rydberg wave functions F nlm at l < l
in the
asymptotic region of electron motion is written as [15]

ðRÞ
X pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðLÞ
jnlmi ¼ Fnlm ðrÞ ¼ 2p ð2L þ 1Þ Znl ðRÞ Ylm ðr=rÞ jLMi dM0 : (1.41)
L

Here, L and M are, respectively, the electron angular momentum relative to the
ðLÞ
atom B and its projection on the direction of vector R and Znl ðRÞ are the periodic
functions of the form
8 9
2 < sin SðLÞ ; ðL ¼ 2kÞ =
ðLÞ 3 1=2 nlðRÞ
Znl ðRÞ ¼ ð1Þ ð2ppe n Þ
k
; (1.42)
R : cos SðLÞ ; ðL ¼ 2k þ 1Þ ;
nlðRÞ

ðLÞ
where Snl ðRÞ is the corresponding quasiclassical phase.
Inasmuch as in the vicinity of perturbing atom at j r  R j; j r0  R j  R, the
Coulomb Green’s function for the degenerate states with l  l
has in the quasi-
classical approximation a simple form with respect to the angular variables [21],

X  ðRÞ E D ðRÞ  sin pe ðRÞ j r  r0 j

pn3 cot pn  Fnlm Fnlm  ffi cot pn ¼
l>l
; m
2 p j r  r0 j
X
pe ðRÞ cot pn j L M i h L M j; ð1:43Þ
LM

the matrix elements hLMj Ke B jnlmi and hnlmj Ke B jnlmi appear in the asymptotic
theory. They are expressed through the diagonal (in M) elements hLMj Ke B jL0 Mi
and are determined by
X pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
hLM j Ke B j nlmi ¼ 2p ð2L0 þ 1Þ
L0
ðL0 Þ
 Z nl ðRÞ Ylm ðR=RÞ hLMj Ke B jL0 Mi dM0 ;
X pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
h nlm jKe B j nlmi ¼ 2p ð2L þ 1Þ ð2L0 þ 1Þ
L;L0
ðLÞ ðL0 Þ
 Z nl ðRÞ Z nl ðRÞ j Ylm ðR=RÞ j2 hLMj Ke B jL0 Mi dM0 :
(1.44)

Since the functions Y lm ðR=RÞ are nonzero only for m ¼ 0, matrix elements
(1.44) are, in fact, the superposition of two waves propagating along the vector
R. Inasmuch as the electron scattering in this case is determined by the angular
momentum L relative to the perturbing atom B, the orbitally degenerate (in l  l
)
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 15

covalent term splits into the individual L components corresponding to m ¼ 0. The

brR
presence of the term ðrRÞ 3 in the potential D UAþ e B leads to an additional mixing of

the states with D L ¼ 1. Because of this, the summation over L0 in (1.44) includes
only three terms (L ; L 1).
In this representation, the matrix elements hLM j Ke B j L0 Mi can be conveniently
written as

ð0Þ
~  j LMi þ hLM j DU jL0 Mi;
hLM j Ke B j L0 Mi ¼ hLM j K (1.45)
e B

~ ð0Þ
where the operator K e B includes, in the outer region, the polarization interaction of
electron with atom B; i.e.,

~ ð0Þ ð0Þ b
K e B ðr; RÞ ¼ Ke B ðr; RÞ  ðr  r0 Þ;
2r4

whereas the long-range potential DU, by analogy with Eq. (1.25), is determined as

D Uðr; RÞ ¼ ½ D UAþ e B ðr; RÞ ðr  r0 Þ:

~ ð0Þ
In this case, the diagonal matrix elements hLM j K e B j LMi are independent of M
(because of the spherical symmetry of interaction) and can be expressed through the
observed characteristics of the potential e  B-scattering: scattering length a and
polarizability b [18]

ð0Þ pb pe 4
K~0;0 ¼ a þ þ ab p2e ln pe þ    ;
3 3
(1.46)
ð0Þ pb pe
K~L;L ¼ adL0  þ :
ð2L  1Þð2L þ 1Þ ð2L þ 3Þ

Since DU is the effective electric dipole interaction with the moment D ¼ b R2 ,
the nondiagonal (in L) elements hLM j DU j L0 Mi in Eq. (1.44) can be represented in
the form
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
0 2L0 þ 1
hLM j DU j L Mi ¼ ð 1L0 0 M j LM Þ ð 1L0 0 0 j L0 Þ PLL0 ðR; nÞ; (1.47)
2L þ 1

where
2 3
Z pe
2
2n
b 6 7
PLL0 ðR; nÞ ¼ 4 jL ðxÞjL0 ðxÞ dx 5:
p e R2
0
16 G.V. Golubkov et al.

For the sufficiently large values of n such that n2 pe  1, the upper limit can be
turned to infinity 1 to obtain the simple expression
8 " #
>
> 1 M 2
> ; L0 ¼ L þ 1;
< pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 1=2
>
b ð2L þ 1Þ ð2L þ 3Þ ðL þ 1Þ 2
hLMj DU jL0 Mi ¼ 2  
>
R pe > 1 M2 1=2
>
> p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi  ; L0 ¼ L  1:
: 1
ð2L  1Þ ð2L þ 1Þ L2
(1.48)
ð0Þ
In as much as the elements K~L;L rapidly decrease with increasing L, one can self-
restrict to the particular calculations of the two-channel situation L ¼ 0 and 1. Then,
the covalent terms are determined from the system of equations
X
tLM;l ¼ pe cot pn KLM;L0 M tL0 M; l ; (1.49)
L0 ¼0;1

whose solution describes two pairs of states classified by the effective angular
momentum L~ and j Mj ¼ 0; 1. It has the following form:

ð0Þ 1 b ð0Þ ð0Þ

~ ¼
ULM ð0Þ2
 ; ~ ðRÞ ¼ n  mLM
nLM ~ ðRÞ; (1.50)
2R4
~ ðRÞ
2nLM

where the quantum defects induced by the field of atom B for the levels with M ¼ 0
are determined by
(
ð0Þ 1 pe h ð0Þ ð0Þ
~ ðRÞ ¼
m LM arctan K 00;00 þ K 10;10 :
p 2
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi # ) (1.51)
ð0Þ ð0Þ 2 ð0Þ 2
ð K 00;00  K 10;10 Þ þ 4 ðK 10;00 Þ

(state with L~ ¼ 0 corresponds to the sign “þ”, and with L~ ¼ 1, to the sign “”). For
ð0Þ
the states with L ¼ 1 and jMj ¼ 1, the matrix elements K 1M;00 ¼ 0, so that

ð0Þ 1 ð0Þ
~ M¼1 ðRÞ ¼ arctan ð pe K 10;10 Þ:
m L¼1 (1.52)
p
ð0Þ ð0Þ ð0Þ
Since K00;00  K10;10 ; K10;00 at R  n2 , only one term with

ð0Þ 1 ð0Þ
~ M¼0 ðRÞ ¼
m L¼1 arctan ð pe K 00;00 Þ:
p

stands out, in fact, in the group of degenerate states. The other terms tend to the
unperturbed Coulomb levels.
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 17

1.5 Exchange Interaction Between the Rydberg

States in the A

– A System

We now turn to a quasimolecule consisting of the identical atoms A

and A. For
definiteness, we assume that the electronic shell of the ion residue is filled and the
atom A is in the ground state 2 S1=2 with one valence electron (as, e.g., in alkali metal
atoms). In this case, one can also use the approximation of LS coupling. Let us
denote the sets of quantum numbers for Rydberg atom A

(q1 ) and atom A(q2 ) in
the ground state by

q1 ¼ ðn1 l1 m1 ; s1 ¼ 1=2; s1 ¼ 1=2Þ and

q2 ¼ ðn2 ; l2 ¼ m2 ¼ 0 ; s2 ¼ 1=2; s2 ¼ 1=2Þ:

Here, an additional symmetry arises caused by the reflection about the plane
perpendicular to the molecular axis and passing through its midpoint. Since the
Hamiltonian of the system does not change as a result of electron reflection about
this plane, the quasimolecular states are divided into even ðgÞ and odd ðuÞ states,
and their wave functions do not change or change sign on electron reflection.
According to the general rules [26], the total wave functions of the odd and even
states with the given total spin S ¼ s1 þ s2 and its projection s ¼ s1 þ s2 on the
chosen direction are written (without the normalization factor) as
X
CðSsÞ
g;u ¼ ð s1 0s1 0 j s1 s1 Þ ð s2 0s2 0 j s2 s2 Þ ð s1 s2 s1 s2 j SsÞ
s1 þs2

 Cð1a

; 2bÞw s1 s1 ð1Þw s2 s2 ð2Þ  Cð2 a

; 1bÞw s1 s1 ð2Þw s2 s2 ð1Þ


Cð2a; 1b

Þw s1 s1 ð1Þw s2 s2 ð2Þ  Cð1a; 2b

Þw s2 s2 ð1Þw s1 s1 ð2Þ :

The indices 1 and 2 label the valence electrons, Cð1a

; 2bÞ and Cð2a

; 1bÞ
are the coordinate parts of the two-electron wave functions corresponding to
the location of these electrons near the corresponding atom (for simplicity,
atoms A

and A are denoted by a and b), wss is the electron spin function, and
ð j1 j2 m1 m2 j JMÞ are the vector composition coefficients [23]. Note that the expres-
sion in braces contains two terms enclosed in square brackets, of which the first
accounts for the electron exchange and the second additionally includes the electron
transfer between centers. After simple mathematics, this expression can be written
in the form

1 X
CðSsÞ
g;u ¼ ð s1 s2 s1 s2 j SsÞ
2 s1 þs2
nh i
 Cð1a

; 2bÞ þ ð1ÞS Cð2a; 1b

Þ w s1 s1 ð1Þ w s2 s2 ð2Þ
h i o
 Cð1a; 2b

Þ þ ð1ÞS Cð2 a

; 1bÞ w s1 s1 ð2Þ w s2 s2 ð1Þ :
18 G.V. Golubkov et al.

Owing to the electron identity and symmetry about the plane passing through the
midpoint of the quasimolecular axis, one has the following representation for the
total normalized wave functions CðSsÞ
g;u of the noninteracting system [29]:

ðS sÞ ðg; uÞ ðÞ
CgðuÞ ; n1 l1 m1 ; n2 ðr; rÞ ¼ F n1 l1 m1 ;n2 ðr; r Þ w S s
h i
ðRÞ ðRÞ ðÞ
¼ A g;u F n1 l1 m1 ð rÞ F n2 ð rÞ F n1 l1 m1 ð rÞ F n2 ð rÞ w S s ;
(1.53)
ðRÞ
where the even (g) and odd (u) functions have the signs ð Þ, Fn1 l1 m1 and Fn2 are the
wave functions of the excited atom A

and A, respectively, and r and r are the
electron coordinates measured from the centers of these atoms. The normalization
R h i2
ðg;uÞ
factors Ag;u are found from the condition drdr Fn1 l1 m1 ;n2 ð r; r Þ ¼ 1; they are
equal to

1
Ag;u ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi; (1.54)
2 ð 1 S212 Þ

where S12 is the overlap integral

Z
ðRÞ
S 12 ¼ dr F n1 l1 m1 F n2 : (1.55)

ð Þ
The normalized spin functions wSs are

ð Þ 1 X
wSs ¼ ð s1 s2 s1 s2 j SsÞ
2 s1 þs2

 w s1 s1 ð1Þw s2 s2 ð2Þ w s2 s2 ð1Þw s1 s1 ð2Þ : ð1:56Þ

The even (g) and odd (u) wave functions CðSsÞ g;u are antisymmetric about the
electron and spin permutation and correspond to S ¼ 0; 1 and s ¼ 0; 1. Since
spin functions (1.56) are mutually orthogonal, the elements of the e  A scattering
Ke A matrix determined, by analogy with Eq. (1.18), satisfy the relation
D   E
ðS sÞ   ðSsÞ ðS Þ
CgðuÞ  Ke A  CgðuÞ ¼ KgðuÞ;gðuÞ dSS0 d ss0 : (1.57)

They are nonzero only for the transitions between the states with the same parity
and independent of s. For this reason, to determine the PES of the system, one can
restrict oneself to the consideration of the situation with s ¼ 0. After insertion of
Eq. (1.57) into the initial equation for the level-shift operator, the perturbed diabatic
Rydberg terms n1 l1 m1 can be determined from the homogeneous system [29]

ðSÞ
KgðuÞ;gðuÞ
hgðuÞj tjgðuÞi ¼ hgðuÞj tjgðuÞi; (1.58)
E  En1 l1
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 19

whose nontrivial solution gives two (singlet g and triplet u) noninteracting per-
turbed Rydberg terms

ðSÞ ðSÞ b
Ug;u ðRÞ ¼ En1 l1 þ K gðuÞ;gðuÞ  ; (1.59)
2R4
split by

D ¼ K ð0Þ ð1Þ
g;g  K u; u : (1.60)

ðSÞ
The matrix elements KgðuÞ;gðuÞ entering in Eqs. (1.58) through (1.60) are
written as
ðSÞ
Kg ðuÞ; gðuÞ ðR Þ
D   E
ðRÞ ðÞ   ðÞ ðRÞ
¼ 2 A2g; u F n1 l1 m1 ð rÞ Fn2 ð rÞ w S 0  Ke A w S 0 F n1 l1 m1 ð rÞ Fn2 ð rÞ
D   E
ðRÞ ðÞ   ðÞ ðRÞ
2 A2g; u F n1 l1 m1 ð rÞ Fn2 ð rÞ w S 0  Ke A w S 0 F n1 l1 m1 ð rÞ Fn2 ð rÞ (1.61)
hD   E
ðRÞ ðÞ   ðÞ ðRÞ
¼ 2 A 2g; u F n1 l1 m1 w S 0  Ke A w S 0 F n1 l1 m1
D   Ei
ðÞ   ðÞ ðRÞ
S12 Fn2 w S 0  Ke A w S 0 F n1 l1 m1 :

The first term describes the level shift induced by the electron elastic scattering
from the perturbing center, and the second term corresponds to the excitation
transfer from one center to the other. According to Eq. (1.25), they take into account
the contributions from the short- and long-range interactions. In contrast to the
above-considered perturbation of atom A

by a foreign atom B, the contribution of
the short-range exchange interaction is determined through the singlet (aþ ) and
triplet (a ) electron-elastic-scattering lengths, whereas the contribution of the long-
range interaction depends on the total spin S. Note that, in the absence of the long-
range interaction, expression (1.60) takes the form
ðRÞ
D n1 l1 m1 ðRÞ
  2 
aþ  ðRÞ  ðRÞ
¼ 2p F
 n1 l1 m1 ðRÞ  þ S F
12 n1 l1 m1 ðRÞ F n2 ðRÞ
1 þ S 212
 2 
a  ðRÞ  ðRÞ
  Fn1 l1 m1 ðRÞ   S 12 Fn1 l1 m1 ðRÞ Fn2 ðRÞ dm 1 0 : ð1:62Þ
1  S 212
0 0
The interaction between the perturbed
D  n1 l1 m1 and En1 l1 m1 is
Rydberg states
ðg;uÞ ðÞ   ðÞ ðg;uÞ
determined by the matrix element Fn1 l1 m1 ;n2 wS0  K wS0 Fn0 1 l0 1 m1 ;n2 and is
given by the expression
20 G.V. Golubkov et al.

ðSÞ
Vn1 l1 m1 ; n0 1 l0 1 m1 ðRÞ

p a
ffi 2 FðRÞ
¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðRÞ
n1 l1 m1 ðRÞFn0 1 l0 1 m1 ðRÞ
ð 1 S 12 Þ ð 1 S 12 Þ
2 2
 
ðRÞ ðRÞ
Fn2 ðRÞ S 12 Fn0 1 l0 1 m1 ðRÞ þ S 10 2 Fn1 l1 m1 ðRÞ : ð1:63Þ

1.6 Potential Curves for the Na(nl)

+ He System

As a particular application of the theory, we analyze the PES of a quasimolecule

composed of highly excited atom A

and a neutral atom B with filled electronic
shell. It is natural to restrict oneself to the simplest and well-documented system
Na

ðnlÞ þ He, because it attracts increased interest in researchers today. This is
caused primarily by the discovery of brown dwarfs (BD) [30] and giant extrasolar
planets (GEPs) [31], which have become the subject of intensive study in the last
years [32, 33]. Brown dwarfs (BDs) are the substellar-mass objects in which the
thermonuclear fusion reaction of hydrogen transformation into helium does not
proceed. They substantially differ from Sun and planets of our solar system and
have effective atmospheric temperature as high as thousands of degrees. As distinct
from the main-sequence stars (which possess regions of radiation energy transfer),
the heat-transfer processes in BDs proceed only through convection, thereby
providing homogeneity of chemical composition in depth.1 Photon fluxes from
BDs are characterized by the optical wavelengths indicating the presence of alkali
metal atoms in their atmosphere [34, 35]. Analogous situation occurs in the giant
extrasolar planets (GEPs), although no radiation from them is observed directly.
Theoretical estimates also suggest the predominance of the sodium and potassium
resonance lines in the radiation spectra of GEPs [36, 37]. The atmospheres of these
substellar-mass objects almost totally consist of the hydrogen molecules and helium
atoms [32]. Owing to the collisions of alkali metal atoms with these particles, the
observed profiles of resonance lines are noticeably broadened (more than by 100
nm, depending on the temperature and density of the medium). In addition, these
profiles strongly deviate from the Lorentzian shape typical of an isolated atom.
Such broadening and line profiles are observed in the BD spectra and can be the
cause of the nontransparency of the GEP atmospheres [32, 36, 38–40].

1
It is generally believed that, for the thermonuclear reactions to proceed, the stellar mass must be,
at least, 80 times greater than the Jupiter mass (i.e., equal to about 0.08 of the Sun mass). The
hypothesis of BDs as star-like objects (with masses from five to 75 Jupiter mass) was put forward
in early 1960s of the twentieth century. It was assumed that the BD formation proceeds following
the scenario analogous to the formation of other stars. However, the detection of BD is hampered
by the fact that they virtually do not emit visible light. The most intense emission from them occurs
in the IR range.
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 21

The profiles of these lines are usually calculated on the basis of the quantum
mechanical description of collisional broadening and the available data on the
potential curves and dipole transition moments. The properties of “satellites” are
particularly informative, because they can arise at the line wings in the presence of
singularities in the differences between adiabatic potentials. The satellite shapes are
highly sensitive to the ambient temperature, whereas their positions and amplitudes
are sensitive to the details of interaction potentials. Moreover, an algorithm allowing
the temperature and pressure of the BD and GEP atmospheres to be reconstructed
from the line shape is formulated in [41, 42]. This explains the increased interest of
researchers in the correct theoretical calculations of the original potentials.
The shock broadening of the resonance 2 s ! 2p line of lithium atom was
studied in [42]. In [43], the collisional broadening of the sodium 3 s ! 3p and
potassium 4 s ! 4p lines by helium atoms was considered and the emission and
absorption coefficients on the red and blue wings were calculated. In addition to the
determination of the temperature and density, the obtained results allowed one to
develop the methods for the diagnostic of albedo and atmosphere composition of
the objects of interest [44, 45]. The potential curves of the low-lying electron-
excited states of the Na þ He system were calculated in [46–51] using different
numerical methods. The first one [46] is based on the semiempirical pseudopoten-
tial method that is widely used in atomic physics. The second [47] is the traditional
quantum chemical LCAO method. The calculations in [48–50] were made using the
standard MRD CI and MOLCAS programs. In [51], the terms of highly excited
states of the Na

ðnlmÞ þ He system were calculated and a detailed comparison
with the existing data was carried out.
In [52], the Hartree–Fock basis of cationic orbitals [53] was invoked to calculate
the potential curves and dipole moments for the 4 s ! 4p transition in the K þ He
system within the full ab initio approach and with the use of the multireference
configuration interaction (MRCI) method and the MOLPRO program [54, 55]. In
addition, the PESs of the K þ H2 system were constructed in [52] and the resonance
wavelengths and dipole transition moments were calculated as functions of the total
number of degrees of freedom, including the changes in the bond lengths and angles
to determine the position of potassium atom relative to the axis of molecular
hydrogen.2
Generally speaking, the medium-perturbed higher lying electronic states are also
involved in the formation of the frequency profile at its wings (under the condition
that Do  v=b0 , where v is the relative velocity and b0 is the Weisskopf radius
[57]). It then follows that the correct calculations require additional analysis of the
behavior of a set of potential curves, especially for the small impact parameters
where the standard quantum approaches are inapplicable [15].

2
Earlier, this system was studied in the work of Rossi and Pascale [56], where the PESs were
calculated with allowance for two orientations of the molecular axis of H2 about the atom K (the
dependence of the dipole transition moments on the geometry was disregarded).
22 G.V. Golubkov et al.

The perturbation of Rydberg terms is determined from the behavior of the

diagonal matrix elements Knlm; nlm . For instance, the l-dependent oscillatory struc-
ture of the terms in overbarrier region, according to [15], reproduces the specific
ðRÞ
features of electron wave function F nlm ðrÞ of Rydberg atom A

and dies out with
increasing R. With an increase in l, this structure becomes more pronounced. The
ðRÞ   ðRÞ 
effect of long-range interaction is determined by the element F nlm  U F nlm ,
the value of which is the most significant at the distances R  n. The azimuthal
angular dependence of long-range potential (1.23) is responsible for the P and
D components of Rydberg terms, whose shifts from the isolated unperturbed
levels decrease with  increasing
  l, because
 the corresponding matrix elements
satisfy condition  K nl2;nl2 < K nl1;nl1 < Knl0;nl0 .
One should also expect the appearance of quasicrossing between the terms with
different values of l and a fixed projection of angular momentum. This is caused by
the fact that the states with different quantum defects behave differently in the near-
barrier region.
Contrary to the interaction between the Rydberg terms (1.37), the interaction
parameters between the Rydberg terms and covalent term (1.34) can be signifi-
cantly greater. In the case of L ¼ 0, only one term corresponding to the covalent nL
state formed by the orbitally degenerate Coulomb states is split-off from the
Coulomb terms and characterized by the isotropic electron S scattering from the
perturbing atom. This state interacts only with those perturbed Rydberg states nlm,
which have zero projection m, in line with the ZRP model.
The aforesaid is clearly demonstrated in Figs. 1.1 through 1.3 where the potential
curves measured from the ground-state of the Na

ðnlÞ þ He system are presented.
The dependences on the interatomic distance R were calculated by formulas (1.28)
through (1.37) with account taken of the harmonics L ¼ 0 and L ¼ 1 in the matrix
elements Knlm; n0 l0 m ðRÞ with various n, l, and m. To this end, we used the data on
elastic e  He scattering [28] and the following expansion coefficients in (1.30)

ð0Þ ð1Þ
C 1 ¼ 1:631; C 1 ¼ 1:963: (1.64)

For the quantum defects in the sodium s; p, and d states, we used the values

m 0 ¼ 1:349; m 1 ¼ 0:857; m 2 ¼ 0:015; (1.65)

which were found using the data reported in [58]. These figures demonstrate the
adiabatic potential curves of the Rydberg nl2 Sþ states calculated for the quantum
numbers n ¼ 4–6 and l ¼ 0–3 [51]. The computational results of other authors
[46–50] are also presented. A comparison with them is of particular interest,
because the accuracy of quantum chemical computations for such values is not
too high [15]. One can see that our results [51] obtained for the component n ¼ 4 of
the term 4s2 Sþ (Fig. 1.1а) are in accordance with the data reported in [46, 48],
whereas the terms obtained in [35, 36] lie too low (this is typical of the MRD CI
program used in those works). The next Rydberg terms 4p2 Sþ and 4p2 P are shown
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 23

a b
31,200
26,200

26,000
30,800
25,800
Energy, cm–1

Energy, cm–1
Na(4s)+He
25,600
30,400
25,400 Na(4p)+He

25,200
30,000

25,000

24,800 29,600
4 6 8 10 12 14 16 R, a 4 6 8 10 12 14 16 R, a0
0

c d
34,700

35,200

34,600
Energy, cm–1

Na(4d)+He Energy, cm–1

34,800

34,500
Na(4f )+He

34,400

34,400

34,000
4 6 8 10 12 14 16 R, a 4 6 8 10 12 14 16 R, a
0 0

Fig. 1.1 Adiabatic terms of the Rydberg 4l 2 Sþ states in the Na + He quasimolecule for the
angular momenta l = 0–3. Curves (a) relating to the 4s2 Sþ state are the results of the following
calculations: □ [46], D [48], r [50], and ■ [51]. Curves (b) are for the 4p2 Sþ and 4p2 P states, first
corresponding to the ~ [46], r [48], ■ [50], and O [51] terms, and the second, to the D [46],
□ [50], and ● [51] terms. The potential curves (c) include the 4d 2 L states, which are denoted as
follows: ■ [47], □ [48], and ~ [51] for the 4d 2 Sþ state, and O and [51] for the 4d 2 P and 4d 2 D
states, respectively. The terms (d) are the results of the following calculations: □ [47], O [48], and
▾
▾ [51] for the 4f 2 Sþ state, and r [51] for the closely spaced 4f 2 P; D; F states. The horizontal
line ______ denotes the dissociation limit corresponding to the energy of the separated atoms in
the given states

in Fig. 1.1b. The term 4p2 Sþ best agrees with the results [48, 50], and the 4p2 P
term is in accordance with the calculations performed in [46, 48]. The potential
curves calculated by us for the 4d2 Sþ and 4f 2 Sþ terms (Figs. 1.1c, d) [51] agree
well with the data in [48] (the terms 4d; f 2 P ; 4d; f 2 D, and 4 f 2 F were calculated
for the first time). In the case of n ¼ 5, a good agreement for the term 5s2 Sþ at
intermediate distances (Fig. 1.2а) is achieved with the data in [46–48, 50]. At larger
distances, the term calculated in [50] lies too high. The potential curve calculated
for the 5p2 Sþ term [51] (Fig. 1.2b) well reproduces the result [48]. The data
reported in [47] are overstated, whereas the term 5p2 P was not calculated previ-
ously. The term 6s 2 Sþ is also in good agreement with [48] (Fig. 1.3).
It is worth noting that the results obtained in [51] are the most preferable,
because the accuracy of calculations performed by the ZRP method [27] are
controlled by the accuracy of determining scattering amplitudes (1.30), which are
usually calculated independently or can be determined from the experiment. With
an increase in n (due to the increase in the number of possible crossings), the
24 G.V. Golubkov et al.

a
33,400

33,300

33,200
Na(5s)+He
Energy, cm–1

33,100

33,000

32,900

32,800
4 6 8 10 12 14 16 R, a 0

36,000

35,600
Energy, cm–1

35,200
Na(5p)+He

34,800

34,400
4 6 8 10 12 14 16 R, a 0

Fig. 1.2 Adiabatic terms of the Rydberg (a) 5s2 Sþ and (b) 4p2 Sþ ; P states in the Na þ He
quasimolecule. Curves (a) are the results of the following calculations:▾ [46], □ [47], O [48], ●
[50], and D [51]. Curves (b) correspond to the □ [47], O [48], and ~ [51] calculations for the 2 Sþ
state and r [51] for the 2 P state

classification of the adiabatic terms without preliminary construction of the diabatic

picture becomes difficult. As to the quantum chemical programs used in [46–50],
they are adapted only to the calculation of adiabatic terms. The advantage of the
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 25

36,600

36,400
Energy, cm–1

Na(6s)+He

36,200

36,000
4 6 8 10 12 14 16 R, Å 0

Fig. 1.3 Adiabatic terms of the Rydberg 6s 2 Sþ state in the Na + He molecule, calculated in [47]
(□), [48] (●), and [51] (D)

method [27] is that it allows one to determine simultaneously the adiabatic and
diabatic potential curves and determine the exact structure of adiabatic terms with
all possible quasicrossings.
For the transitions to higher values, n  10, the adiabatic description of PESs
breaks down because of the appearance of many pseudocrossing points between the
Rydberg and covalent terms. For this reason, it is difficult to classify the
corresponding states, because, for each n, the degenerate in l covalent terms
ðcÞ
converge to the common limit U nlm ðR
Þ ¼  2 1n2 at the point R
¼ 2n2 .

1.7 Relation to Quantum Chemistry – Matching Method

The quantum chemically calculated terms possess some disadvantages that restrict
the accuracy of computations. First, they do not ensure proper asymptotics for the
calculated PES energies of the separated atoms and, thereby, understate (or over-
state) the potential energies at small interatomic distances. Second, these
approaches disregard the orbitally degenerate states, which form starting at n ¼ 5
and have quasicrossings with the Rydberg states. This hampers the correct determi-
nation of the adiabatic potential curves at intermediate distances (on the order of the
Rydberg electron wavelength). However, they give proper coordinate dependences
for the potential curves in the region of small and intermediate interatomic dis-
tances, though do not provide exact agreement with the energy scale. Hence, the
necessity arises of developing a fundamentally new computational procedure free
of the indicated disadvantages.
26 G.V. Golubkov et al.

The approach [27] gives the required solution, though it applies only in the
region of intermediate and asymptotically large distances. This is so because the
corresponding solutions are constructed on the basis of Rydberg functions centered
on the ion core. Therefore, the problem amounts to the matching of the quantum
chemical results with the results obtained within the framework of integral
approach and to the construction of general consistent picture of terms. To solve
this problem, the integral variant [27] needs modification. This amounts to the
transformation of PES through the inclusion of the additional dissociative valence
and ionic configurations into the theory. This procedure was partially accomplished
in [51]. The inclusion of the dissociative and valence configurations is usually done
by expanding the basis of eigenstates for the Green’s operator GA

of the Rydberg
atom A

[3] entering in Eq. (1.8) for the level shift operator. The ionic configura-
tions are introduced through the following renormalization of the scattering matrix:

j ’ r ih’ r j
K ¼ K e B þ V V: (1.66)
E  E r þ 1 =R

Here, E is the total energy of the system, ’r and Er are, respectively, the wave
functions and electron affinity in the ionic configuration Aþ B , R is the interatomic
distance, and V is the corresponding interaction. The poles of the second term
determine the positions of the ionic terms

U ðiÞ ðRÞ ¼ Er  1=R: (1.67)

Note that if the resonance occurs at distances Er >1=R, the potential curve
U ðiÞ ðRÞ passes into continuum and acquires the autoionization width. This result
follows from the eigenvalue solution to the equation for the level shift operator.
Then, the quantum chemical calculations are corrected through the direct coinci-
dence with our results at those intermediate distances where there is no quasicross-
ings of the Rydberg and orbitally degenerate covalent terms. As the result, we
obtain the required energy shift of the potential curve at small distances. Note that
the PESs determined in this manner must exactly transform, at small internuclear
distances, to the potential curves of the corresponding Rydberg molecules. We thus
obtain a consistent structure of the adiabatic PESs that can be used to reproduce the
full diabatic picture (including all quasicrossings) for the terms in the region where
the intercenter coordinates change. The use of diabatic terms is necessary for the
correct calculations of the quantum defects as functions of interatomic distances
that enter into the definition of electronic matrix elements.
An important role belongs also to the ionic configuration that interacts with the
Rydberg valent and dissociative states and gives rise to the additional quasicross-
ings between the diabatic terms. This result can be obtained by expanding the basis
of Green’s function entering into the main equation for the T matrix of multichan-
nel scattering [3]
X
~ 0 Þ T  i ~t
T ¼ ~t þ ~tðG  G jbihbjT; (1.68)
b
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 27

where the difference G  G ~ 0 is a portion of Green’s operator containing poles

corresponding to the discrete levels of the Coulomb center, and jb i are the wave
functions of the dissociative configurations. In contrast to the operator G0 , which
includes the Rydberg e þ XY þ states (without poles corresponding to the discrete
levels of the Coulomb center), dissociative X þ Y
, and discrete valence states, the
operator G~ 0 involves the states of ionic configuration, i.e.,

~ 0 ¼ G0 þ j ’r ih’r j
G : (1.69)
E  Er þ 1=R

After the renormalization, the reaction matrix ~t is written as

~
~t ¼ t þ V j ’ r ih ’ r j ~
V; (1.70)
E  Er þ 1=R

~ also includes the interaction

where the matrix t is defined in [3], and the operator V
with the ionic configuration. The dependences on the internuclear distance for the
electronic parts of the matrix elements corresponding to the configuration coupling
between the Rydberg, valence, dissociative, and ionic terms can be determined
from the difference between the adiabatic and diabatic potential curves in the
vicinity of quasicrossings.

1.8 Interaction of Rydberg Molecules with Atoms

and Molecules

The Rydberg states (RSs) of molecules (as well as atoms) can clearly be divided
into the two groups: strongly and weakly interacting with the core. They differ in
the electron angular momentum l, whose value characterizes the height of orbital
barrier. The number of the states in the first group is smaller than in the latter,
because the effective electron angular momentum l
dividing these groups is small.
For instance, in the two-atomic molecules (of the type H2 ; N2 ; NO, O2 , and CO)
l
¼ 2. In the case of weakly interacting states (with l > l
), the level positions can be
satisfactorily approximated by the expression

1
Enlq ¼ Eq  ;
2 n2
and the levels themselves can be assumed to be purely coulombic for every fixed
value of the index q characterizing the set of vibrational v and rotational N ionic
states; i.e., q ¼ fv; N g.
Let us consider the evolution of molecular spectrum with changing principal
quantum number n. For the fixed interatomic distance R, the positions of Rydberg
levels relative to the ionic term Ui ðRÞ is given by the formula
28 G.V. Golubkov et al.

1
Enl ðRÞ ¼  ;
2 ½n  m l ðRÞ 2

where ml ðRÞ is the R-dependent diabatic quantum defect of the level and l is the set
of quantum numbers determining this state with inclusion of the internal degrees of
freedom. Under the conditions of nonadiabatic motion, the set of levels is deter-
mined by the expression

1
Enlq ¼ E q  ;
2 ½ n  m l q ðnÞ  2

which describes the Rydberg series with different ionization limits that, similar to
atomic series, differ from each other by the energy Eq of ionic core excitation. In the
spectral regions where the period of electron rotation is comparable with the periods
of core vibration or rotation, the energy levels of Rydberg molecule (referred to as
rovibronic levels) demonstrate the irregular dependence on n [1, 3].
The next important feature of the spectrum of highly excited molecules is that,
along with the Rydberg states (i.e., one-electron configuration e–XY+), the states
mixed with other (non-Rydberg) valent configurations of molecule can be formed.
Among these are, first, the quasistationary states forming the additional set of
resonance levels in the XY þ þ e system and, second, the dissociative states having
a continuous spectrum of nuclear motion. The dissociative states lead to the decay
of a Rydberg molecule into atomic fragments.
The nonadiabatic coupling of the channels (in the excited states of the nuclear
subsystem) and the configuration interaction induce broadening that is determined
by the contribution from the autoionization states and, generally, is a nonmonotonic
function of energy. Under certain conditions, the autodecay widths of some levels
can turn to zero (for the interference-stabilized level). These states in continuum are
the most interesting in atomic and molecular physics and are the subject of wide
subject of wide interest in the literature (see. e.g., [59–63]).
The experimental investigations of the highly excited Rydberg molecules (with
the principal quantum numbers n  10) have been carried out over 20 years. The
results of these studies, including the developed and elaborated methods, are
reported in detail in the Freund’s review in [1], where the history of this problem
is given and extensive literature is presented.

1.8.1 General Equations of the Theory

The quasimolecules XY

þ A and XY

þ M, where atom A or molecule M are in

the ground electronic state, have the same PES configurations as in the above-
considered simpler systems X

þ A and XY

þ M. To obtain them, one should
determine explicitly the Green’s function for the e þ XY þ system. The inelastic
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 29

transitions induced in an ion XY þ by electron renders this problem multichannel,

because the states with total energy are formed owing to the interaction (and
interference mixing) of the states corresponding to the different channels q of
entrain excitation
E ¼ e q þ Eq ; (1.71)
where q is the set of vibrational v and rotational N quantum numbers q ¼ fv; N g of
the ion, Eq is the ion excitation energy, and eq is the electron energy measured from
the spectrum boundary in a given channel. The presence of a nonadiabatic coupling
between motion channels fundamentally distinguishes Rydberg molecules from
atoms and imparts to them a number of unique features. Among these are, e.g.,
irregular dependence of the quantum defects on the level number n [1, 3] and of
their near-threshold absorption spectra on the frequency of incident light [64], the
presence of bound states on the continuum background [61], the stroboscopic effect
[65–67], etc. The nonadiabatic interaction gives rise also to the interesting physical
phenomena consisting of a strong (resonance) dependence of the inelastic-transition
cross-sections, ionization cross-sections, and charge-transfer cross-sections for
highly excited molecules XY

on the principal quantum number [68, 69].
Let us consider the method of constructing Green’s function of a Rydberg
molecule in more detail. With this aim, it is convenient, following [70], to use the
Dyson equation

GV ¼ G þ GTG: (1.72)
The operator G (without interaction V between an electron and ion core)
represents a convolution of the Coulomb Green’s function GðcÞ
q and Green’s func-
tion of the nuclear subsystem,
X
GðEÞ ¼ j qi GðcÞ
q ðE  Eq Þ hqj; (1.73)
q

where jqi are the corresponding wave functions of the nuclear subsystem of XY þ ,
and GðcÞ
q is Green’s operator of electron in the Coulomb field. Inasmuch as the main
contribution to the matrix elements
  of the T operator comes from the region of
small r satisfying condition r eq   1, Green’s function can be expressed through
the matrix element of the operator T and represented in the following form [70]:

GðnÞ ðEÞ ¼ GðEÞ

X 0 pðvq vq0 Þ3=2  ðJÞ


þ ð1Þlþl  lq JMi Tlq; l0 q0 ðEÞ hl0 q0 JM  ð1:74Þ
ll0 ; qq0 ;JM
sin pvq sin pvq0

In this equation, the operator G is determined by the first term of expression

(1.72), and the basis functions are defined by

jlqJMi ¼ QlqE ðrÞ wJv ðxÞ fi ðxÞ F JM ^

lN ð^
r ; RÞ; (1.75)
30 G.V. Golubkov et al.

where QlqE coincides with radial function (1.7), ’i ðxÞ is the ionic electron wave
function, fxg is the set of coordinates of inner electrons, wvJ ðxÞ is the ionic
vibrational wave function, x is the internuclear distance, and nq is the effective
principal quantum number, nq ðeÞ ¼ ð2eq Þ1=2 , where eq ¼ Eq  E. The total
angular function F JM ^ of the system is determined in the representation with
lN ð^
r ; RÞ
the total angular momentum J, its projection M, nuclear rotational motion N, and, in
the case of LS coupling (e.g., for the state of XY þ ), has the form [71]
X
F JM ^ ¼ ^ N m M  m j JMÞ;
lN ð^
r ; RÞ Y lm ð^
r Þ Y N; Mm ðRÞðl (1.76)
m

r and R^ are the corresponding angular variables).

(^
For the subsequent consideration, it is important to separate the radial Green’s
ðcÞ
function G lq ðr; r 0 ; eÞ into the parts strongly and weakly depending on energy [72, 73]:
  
G lq ð r; r 0 ; eÞ ¼ cot pnq ðeÞ ’lqe ðrÞ ’lqe ðr 0 Þ  þ g lq ð r; r 0 ; eÞ:
ðcÞ ðcÞ
(1.77)

The first term in (1.77) reproduces positions nq ðeÞ ¼ ð2eq Þ1=2 of the Coulomb
levels at e < 0; it is expressed through the Coulomb wave functions regular at zero
point and is normalized as
  
f lq eðrÞ  flqe0 ðrÞ ¼ pdðe  e0 Þ: (1.78)

The elements of T matrix in (1.74) are determined from the integral equation [3]
X
T¼tþt jlqJMihlqJMj cot pnq T; (1.79)
q

where the reaction matrix t describes the electron scattering by the ion core (for all
open Rydberg channels the functions cot pnq ¼ i). Owing to the separable nuclear
structure (i.e., the factorable dependence on the variables r; R and r 0 ; R0 ), integral
equation (1.79) reduces to a system of linear algebraic equations for the elements of
collisional T-matrix. Under condition B n3  1 (where B is the ion rotational
constant), which encompasses a broad spectral range for the majority of molecules,
the Rydberg states can be described within the framework of adiabatic approxima-
tion. To transfer to the adiabatic basis, one should pass to the molecular-axis-fixed
coordinate system, in which the absolute value of the projection of electronic
angular momentum onto the molecular axis is fixed. On this basis, the states of a
two-atomic molecule are classified by the values of J and L; i.e., the electronic
angular momentum l is generally not conserved. The basis functions in the Rydberg
states (taking into account the vibrational and rotational motions) are [74]
X
jJMrLv; i ¼ Jl
arL jJMlLv; i; (1.80)
l
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 31

where r is the quantum number characterizing the effective electronic angu-

lar momentum in the Rydberg configuration with allowance for the mixing of
Jl
states l. The coefficients arL are normalized to unity and assumed to be known.
In the adiabatic basis, the elements of reaction matrix t are diagonal in
the indices r and L and expressed through the adiabatic quantum defects of the
levels

trL; r0 L0 ðxþ Þ ¼ tan pmrL ðxþ Þdr; r0 dL; L0 : (1.81)

Because the diabatic (1.80) and channel (1.76) bases are related by the unitary
J
transformation, the following relationship is true for the elements tlNv; l0 N 00 v0 inducing
vibrational–rotational transitions in XY þ [74]:
X  J   Jl0 Jl0
t JlNv; l0 N 0 v0 ¼   þ  J
r L U N L w v tan p mr L ðx Þ w v0 a r L U N L ;
a Jl Jl
(1.82)
rL

Jl
where ULN is the Fano rotation submatrix [71] with elements

Jl
UNL ¼ hlN j lLi ¼ ð1ÞJþLN ð2  dL0 Þ1=2 ðlJL  LjN0Þ: (1.83)

The nondiagonal matrix elements are generally smaller than the diagonal ele-
ments by the ratio of the amplitude of zero-point vibrations a0 to the interatomic
distance. For this reason, the strong vibronic coupling is significant, as a rule, only
for a pair of energetically close states.
Of the greatest interest for the problem under consideration is the coordinate
region ð j r  R j; j r0  R j Þ  R near the perturbing atom. For the states with
2=3
l  nq , the direction of electron momentum at the point R coincides with the
direction of vector R. Physically, this situation is the most important, because the
interaction with ion core noticeably distorts only the states with small l. Also, our
strongest assumption is that the perturbing particle is in the classical region of
electron motion.
In the quasiclassical conditions, it is convenient to use the rotationally adiabatic
approximation and consider it, similar to the translation, as the motion of a
representative point along the set of vibronic PESs corresponding to different
vibrational states of the molecule. In this case, expression (1.74) takes the form

GV ¼ g 0
X  D 
 þ 
þ  wv ð xþ Þ jL ð pv rÞ YLK ð^
rÞi wv ð x0 Þ jL0 ð pv r0 Þ YL0 K0 ð^
r0 Þ 
LK; L0 K 0 ; vv0
ðJÞ
 ð pv cot pnv dLK; L0 K0 dvv0 þ a Llv; L0 l0 v0 dK0 dK0 0 Þ;
(1.84)
32 G.V. Golubkov et al.

where r^ is the set of angles specifying the direction of vector r relative to R. The
ðJÞ
matrix a Llv; L0 l0 v0 is defined by

ðJÞ
X ðnv nv Þ3=2
a Llv; L0 l0 v0 ðRÞ ¼ 2p2 ½ð2L þ 1Þ ð2L0 þ 1Þ1=2
lL
sin pnv sin pnv0
ðJÞ ~ ðRÞ f ðlÞ ðl0 Þ
~ 0 0 ðRÞ YlL ðRÞ
^ Y
0 0 ðRÞ;
^
 TlLv; l0 L0 v0 fLvE LvE lL ð1:85Þ

The electron wave functions are

8
< QlqE ðR; nq Þ;
> L ¼ 2p;
~ ðlÞ
fLqE ðRÞ ¼ QlqE ðR; n  1=2Þ; L ¼ 2p þ 1 (1.86)
>
:
ðp ¼ 0; 1; 2; :::Þ

(for the even and odd values of L, they differ in the phase shift by p=2). Corre-
spondingly, the function g0 is
X cos pv j r  r0 j þ
g0 ¼  wv ðxþ Þ w ðx0 Þ: (1.87)
v
2p j r  r0 j v

According to Eq. (1.84), the function G  g0 is the sum of terms, each demon-
strating the factorable dependence on the variables r; r0 and x; x0 . Under these
conditions, the integral equations for the level-shift operator t describing the
eigenvalue spectrum of the system XY

þ M reduce to the algebraic equations.
The operator form of these equations is written as

~ v  g Þ t:
t ¼ KðG (1.88)
0

~ corresponds to energy ~eq ¼ p2 2 in the motion channel q.
The operator K q

1.8.2 Perturbation of the Rydberg Series in a Homoatomic

Molecule X2 by the Field of Neutral Particle

Let us consider a system X2

ðnlLÞ þ M with energy jEj  1, in which the total
energy E at a fixed distance between the particles X2þ and M is equal to the binding
energy of Rydberg electron in the ground vibrational state of the ion (here, the
electronic angular momentum l is a good quantum number). The ion core perturbs
the states with a small electronic angular momentum l relative to the Coulomb
center (l ¼ 0, 1, and 2). The field of perturbing particle, in turn, influences only those
superpositions of the Coulomb states that have small values of electronic angular
momentum L relative to the particle. These two groups of heterocenter terms that
are characteristic of a purely Coulomb center will be briefly called l and L terms.
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 33

The substitution of Eq. (1.84) into Eq. (1.88) reduces the homogeneous integral
equation with degenerate kernel to the following system of algebraic equations
[70]:
X
v; ss0 ¼
t LK K~Lv; ss00
s00
" #
X 0
 pvs00 cot pnvs00 t LK
v; s00 s0 þ d K0 a Lvs00 ; L0 v0 s00 t Lv0 ;Ks00 s0 :
L 0 v0
(1.89)

The index s labels the vibrational states of molecule M with the excitation energy
Es , and the quantities pvs and nvs are defined by
 
2 1 1=2
pvs ¼  ; nvs ¼ ½2ðEv þ Es  EÞ1=2 :
R n2vs

The system of equations (1.89) describes all the possible types of vibronic PESs
for the X2

ðnlLÞ þ M system with inclusion of the nonadiabatic coupling between
the electronic and vibrational motions. Note that the shift of Coulomb levels, as a
result of the joint effect of the perturbing particle M and the ion core, is the most
efficient for K ¼ 0. It is this case that holds the greatest interest. For the nonzero K,
the condition for unambiguous solution to the homogeneous system of equations
determines the set of Coulomb terms that are split by the field of particle M and do
not interact with the ion core.
There are two main mechanisms of the electron interaction with atoms and
molecules: direct and resonance. The direct mechanism (potential or background
scattering) is described in the adiabatic approximation

ð0Þ 1
K~Lv; ss0 ¼  hsj tan dvs ð~evs Þjs0 i; (1.90)
pvs

where dvs is the electron elastic-scattering phase determined for energy ~evs ¼ p2vs 2
at the point where particle M is situated. Since the vibrational coupling is weak, the
terms nondiagonal in s are small when compared with the diagonal terms, i.e., the
interaction between vibronic terms is appreciable only in the vicinity of their
quasicrossings.
For the resonance mechanism of interaction between electron and the perturbing
particle, the scattering operator has the form
 ED 
 ðrÞ ðrÞ 
X ’ v s ’ v s 
K ~ ð0Þ þ
~ ¼K V V: (1.91)
v s
E0  ðE r þ Es þ Ev Þ
34 G.V. Golubkov et al.

ðrÞ
Here, E0 ¼ E þ 1=R and j’ vs i are the total wave functions of the ðv; s Þ state
in the ionic configuration X2þ M , and Er is the resonance energy. The poles of the
second term in Eq. (1.91) define the positions

U ðiÞ
vs ðRÞ ¼ Er þ Es þ Ev  1=R; (1.92)

of different vibrational v and s states in terms of ionic configuration. Note that the
K~ operator is the same for both the positive and negative energies. The solution to
the above equations provides the full picture of the interaction between Rydberg
molecule and neutral particle.
By neglecting the anharmonic interaction in molecule M and taking into account
that the residues of the second term in Eq. (1.89) are appreciably smaller (by n2
times) than those of the first term, one obtains the following equation giving the
unambiguous solution of the energy eigenvalues:

1  p vs K~Lvs ctgpnvs  K~Lvs aLsv;Lvs ðRÞ ¼ 0; (1.93)

The solution to this equation points to the fact that there are two groups of
vibronic terms of the combined system XY

M. The first group is composed of the
Rydberg R-independent l terms

ðRÞ 1
U nv lLs ¼ Ev þ Es  ; (1.94)
2ðnv  mlLv Þ 2

(mlLv is the quantum defect of the lLv-series), whose positions are determined by
the poles of the electron–ion core scattering matrix T. The second group includes
covalent terms L that are split off from the Coulomb terms by the field of perturbing
particle.

1
UnðLÞ
vs
ðRÞ ¼ Ev þ Es  n o2 ; (1.95)
ð0Þ
2 nvs þ p1 arctan ½ pvs ðRÞ K~Lvs ðRÞ

where nvs corresponds to the principal quantum number of the nearest Coulomb
level at the classical turning point (R
vs 2n2vs ) determined from the condition
pvs ¼ 0.

1.8.3 Interaction with the Covalent Term

The collisions between particles X2

and A induce nonadiabatic transitions between
the Rydberg l terms and the covalent L terms in the vicinity of their quasicrossings.
The greatest contribution to the inelastic vibronic–transition cross-sections comes
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 35

ðcÞ
from the vicinities of the quasicrossing points Rnv0 between the Rydberg nv0 l and
covalent nv L terms corresponding to different vibrational states of the ion. It can be
shown that Eq. (1.93) in its vicinity takes the form
  
ðRÞ
E  UnðLÞ
v
E  Unv0 lL ¼ Vnv L; nv0 l :
2
(1.96)

The interaction between these states is given by the expression

BlLLv pv K~ 2Lv cos2 pnv ðlÞ

Vn2v L; nv0 l ¼ fnv nv0 ; (1.97)
pn3v

where the multiplier

 
ðLÞ
tlv; lv0 2  3
nv0 cos pnv
fnðlÞ
v nv0
¼    2 ; l¼ (1.98)
ðLÞ ðLÞ nv cos pnv0
l tan nv  tlv; lv 2 þ tlv; lv0

is defined as a nonadiabatic mixing factor.

On the background of continuum, the terms of quasimolecule X2

A are
described by Eq. (1.93) with the formal substitution tan pnv0 ¼ i corresponding to
the autoionization. The solution to this equation gives the following expression for
the covalent term:

1 pv ðRÞ K~Lv ðRÞ Gnv L ðRÞ

UnðLÞ ð RÞ ¼ Ev  2
þ i ; (1.99)
v
2nv pn3v 2

where Gnv L ðRÞ is the autoionization width whose explicit form is given in [64].

1.8.4 Interaction with the Ionic Term

Using expression (1.91), one can determine from Eq. (1.93) the interaction between
the Rydberg molecular X2

ðv0 Þ Mðs ¼ 0Þ and ionic X2þ ðvÞM ðs Þ configurations
[69],
  2  2
 ðs Þ   ^  f ðlÞ q0s :
V 2ivs ; nv0 lL ¼ 4p2  ALL   Fev lL ð RÞ YLL ðOÞ nv nv0 (1.100)

ðs Þ
Here, ALL is the coefficient characterizing the asymptotic expression

ðs Þ
ðÞ A 
fLL ¼ LL expðaðiÞ
s r Þ
r
36 G.V. Golubkov et al.

for the normalized electron radial wave function in the negative ion, L is the
projection L of the electronic angular momentum on the axis of molecule M,
ðiÞ
as ¼ ½2 ðEr þ Es Þ 1=2 and Fe v lL
ð RÞ are the functions normalized to unity
and coinciding, at energy e0 ¼ 1 2n2 , with the electron wave functions in the
field of Coulomb center, O ^ are the angular variables characterizing the orientation
ðlÞ
of molecule M relative to vector R, fnv nv0 is the nonadiabatic mixing factor (1.98),
and q 0 s ¼ h 0 js i is the Frank–Condon factor. The interaction of the covalent
2

nv L-term with the term X2þ ðvÞ M ðs Þ of ionic configuration on the continuum
background is expressed, according to Eq. (1.89), through the full autonization
width of the resonance level s [75]; i.e.,

1
V 2ivs ; nv0 L ¼ Gs q0s : (1.101)
2 pn3v0

1.9 Dynamical Models in the Theory of Collisional Transitions

Between Rydberg States

The dynamical models used in the description of the processes involving Rydberg
atoms are described in detail in the review of Hickman, Olson, and Pascale in [1].
For this reason, we will not discuss here all possible theoretical approaches and
concentrate only on those, which are used more frequently and efficiently. Among
these is primarily the impulse approximation [76, 77], in which the transition matrix
elements nlm ! n0 l0 m0 with a given momentum transfer Q are calculated by the
formula
Z
0 0 0
A ð k; nlm ! k  Q; n l m Þ ¼ F
n0 l0 m0 ð k  Q Þfe Fnlm ð k Þ dk: (1.102)

Here, the amplitude fe of the e  A scattering can be calculated using the low-
energy approximation suggested by O’Malley [78–80]. As a rule, this approxima-
tion is restricted to the first terms of expansion

4p
fe ¼ a  bQ: (1.103)
3

The formulas of Born approximation are described by the standard way using
the effective potential Vðr; RÞ chosen in such a way that it describes the real
characteristics of the e  B scattering [1].
Both approaches are, in essence, equivalent. Their limitation is associated with
the fact that they do not take into account a change in the mutual arrangement of the
terms between which transition occurs. However, the regions of mutual approach or
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 37

quasicrossing terms combined in the collisional transitions play the decisive role in
some cases.
The study performed above directly points to this fact. For example, when
studying the direct transitions between the nlm and nL0 groups of states, the
distortions in the group of nL0-terms should be taken into account explicitly.
Along with this, the interaction between these groups of states can be treated
perturbatively, because, when moving along a given trajectory, the transition
probability W is low enough (W  1).
In the approximation of linear trajectories, W is calculated by the following
formula [1]:
 1 2
Z 
 

W¼ 0 0 
dt V if ð R ðtÞÞ expðiD E nl;nL ðt Þ dt Þ ; (1.104)
 
1

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where R ¼ b2 þ Vc2 t 2 , b is the impact parameter, V i f ð R ðtÞÞ is the effective
interaction between the states of interest.
In the method of perturbed stationary states, the impulse and Born approxima-
tions correspond to the assumption that energies Ei and Ef are independent of time t.
The inclusion of the dependence D EðtÞ corresponds to the perturbation theory in
the distorted wave method. In addition, the fact is also taken into account that, as a
result of the interaction induced between the state l < l
and a large group of states
l > l
by the external particle A, its m components become mixed (in the case of
molecule XY

, the same is true for the M components of the total angular momen-
tum J).
It should be emphasized that the calculations utilizing information on the PES
reveal term quasicrossings and regions of the spatially localized transitions for
which
 1
Vc  Fi f   R 2 (1.105)

(Fif is the difference in the forces acting along the particle trajectory from the
crossing terms i and f ), so that the collisional processes can be described by the
Landau–Zener method. Under condition (1.105), the Landau–Zener formula is
obtained from Eq. (1.104) in the approximation of two coupled states. A more
complex situation arises for the case of multiple term crossings. However, the
exactly solvable models in the theory of nonadiabatic transitions [81–83] show
that, in many cases, a rather simple and reliable estimate of cross-sections can be
obtained in the approximation of independent Landau–Zener transitions at the term-
crossing points, which are characterized by the parameter
 2
2p  Vif 
 ¼   : (1.106)
Fif Vc
38 G.V. Golubkov et al.

Therefore, the use of the data on PES of the system X

þ M represents new step
in the study of collisional processes involving highly excited atoms.
Formula (1.104) (including the DEðtÞ dependence) and the model of independent
Landau–Zener transitions are used below in the calculations of the l-mixing, charge
transfer, and atomic ionization processes. The l-mixing process can be regarded as a
dynamic transition from the states with fixed value of l (for l < l
) to a group of
states with l > l
.
As an example, we refer to the lmixing process for the Rydberg states of atom
Na

ðnlÞ in slow collisions with atom Xe; i.e.,

Na

ðl  l
Þ þ Xe ! Na

ðl > l
Þ þ Xe; (1.107)

and compare the computational results [84] with the experiment and the data of
other approximate approaches [85]. The corresponding dependences are presented
in Fig. 1.4. The calculation was carried out for the relative velocity Vc ¼ 2  104
of the colliding particles (which corresponds to the temperature Tc / 300 K) and
parameters md ¼ 0:015, a ¼ 6:0, and b ¼ 27:06, taken from [86]. The black
circles on the right wing correspond to room temperature Tc ¼ 300 K [87]. On
the left wing of the curve, white circles are the results obtained in [88] for the
temperature Tc ¼ 430 K.

s, Å2

105

104

103
0 10 20 30 n

Fig. 1.4 l-Mixing cross-section for the Rydberg states of atom Na

in the collisions with atoms
Xe as a function of the principal quantum number n
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 39

The fact that the cross-section increases at small n is explained by the increase in
a scale factor of n4 , which is proportional to the square of a geometric size of the
Rydberg particle, whereas the decrease is explained by the fact that D E nl;nL weakly
depends on R at large n, although becomes small. In the limit n ! 1, the cross-
section of process (1.107) transforms to the known result [89]

a2
snlL ¼ 2 p ; (1.108)
n3 Vc2

which is well reproduced in the impulse and Born approximations [1].

As a next example, we consider the ionization process of Rydberg atom A

in
the collision with a diatomic molecule M possessing autodecaying state M . The
specific feature of the terms (denoted below by the index i) correlating with this
state at R ! 1 is that they intersect continuum boundary at the point Rr ¼ jEr j1 .
The ionization can be regarded as being completed if the characteristic residence
time of the system in continuum is much longer than the half-life of ion M ; i.e.,

Gi ð R 2i  b2 Þ1=2 Vc1  1: (1.109)

The transition to the ionic state can occur both at the approach stage and at
the stage of flying apart. One can show that the second stage is more efficient. The
characteristic ionization cross-sections at thermal velocity are on the order of the
gas-kinetic cross-sections. This is confirmed by the dependence of ionization cross-
section on the principal quantum number for the reaction (Fig. 1.5)

H

ðn0 Þ þ N2 ðs ¼ 0Þ ! Hþ N  þ 
2 ðs ¼ 1Þ ! H þ e þ N2 ðs ¼ 0Þ; (1.110)

At low electron energies, the e þ H2 system has a resonance corresponding to

the autodecaying Su -state (L ¼ 1; M ¼ 0). To describe the X
þ H2 processes
associated with the occupation of this H 2 state, one can set Er ¼ 1:7 eV and
G ¼ 0:5 eV (with these parameters, the theory reproduces well the experiment on
the ionization of atoms K from the ground state by molecules H2 [90]). Table 1.1

Fig. 1.5 Ionization cross-

section as a function of the
principal quantum number n,
as calculated by the harpoon
mechanism for atom H
40 G.V. Golubkov et al.

Table 1.1 Recommended parameters for the description of charge

transfer in the H

þ N2 system [75]a
s 0 1 2 3 4 5
G0s 1.1 2.3 2.4 1.7 0.9 0.4
G1s 2.3 1.3 0.01 0.7 1.6 1.5
RðsÞ 11.8 10.7 9.8 9.0 8.3 7.7
a
Partial widths are given in units of 103 au

reports the recommended parameters for process (1.110). These parameters were
calculated using the analytical model [91], which reproduces well the resonance
e þ N2 scattering (“shape” resonance in the 2 Pg -state with the resonance energy
Er ¼ 2:3 eV). The dependence of partial cross-section shown in Fig. 1.5 was
calculated in [75] using parameters taken from Table 1.1 for the relative velocity
of colliding particles Vc ¼ 1:32  103 au. This function has a typical bell-shaped
form with the maximal cross-section equal to 0:8  1017 cm2 at n0 ¼ 7. With
increasing n0 , the cross-section decreases as 1 n30 .

1.10 Ionization and Vibronic Transitions in Slow Collisions

of Highly Excited Molecules with Neutral Particles

The expressions presented above for the interaction of the terms belonging to
different states of the Rydberg atoms and molecules allow one to calculate the
processes specific to these systems. The modern optical methods of electronic
excitation of atoms and molecules, in conjunction with various methods of the
collision theory and ionization by an electric field [92, 93], as well as with the
methods of microwave multiphoton ionization allow the generation of atomic and
molecular beams in the state with close values of n, the same n, and even in the
individual Stark states [1]. In this connection, the question of how the cross-sections
of elementary processes (or reactions) depend on the energy of a high-resolution
initial excitation has great significance.

1.10.1 Probability and Cross-Section of the Collisional Ionization

When describing this process, it suffices to restrict oneself to the two-channel

vibronic basis and use (as in the analogous problem of Rydberg ionization due to
the perturbation introduced by the term intersecting continuum boundary [72]) the
model of independent Landau–Zener transitions in a system with many term cross-
ings [81]. In the approximation of linear trajectories, this problem can be solved
analytically.
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 41

For the given impact parameter b, the ionization probability is given by

WnðiÞ1 n0 ðbÞ ¼ pn1 n0 ðbÞqn1 n0 þ1 ðbÞ½ 1  S ðbÞ; (1.111)

where pn1 n0 ðbÞ is the probability of the first nonadiabatic transition from the
Rydberg n0 l to the covalent n1 L-term, and qn1 n0 þ1 ðbÞ is the probability to pass to
continuum at the point R
n1 along the diabatic n1 L term. The quantity S has a
meaning of the probability that the atom remains on the covalent term as the
representative point moves in the region of distances R  R
n1 , where the term is
autoionizing and determined by expression (1.99). The resulting ionization cross-
section takes the form

ðiÞ
sðiÞ
2
n0 ð Vc Þ ¼ p ð Rn1 Þ P10 ð Vc Þ; (1.112)

where

Z1   A 
ðiÞ 2 1=2
P10 ð Vc Þ ¼2 x 1  exp  ð1  x Þ
Vc
0
 
Q 2 1=2

 exp  ð1  x Þ ð1  SðxÞÞdx; x ¼ b RðcÞ
n0 : ð1:113Þ
Vc

The quantities A and Q are associated, respectively, with the probability of

the first Landau–Zener transition from the original term to the covalent n1 L-term
(A= Vc ) and the probability to remain on the original term (expðQ=Vc Þ). The
corresponding expressions for these quantities are given in [82].
In the case that the strong nonadiabatic mixing of the Rydberg n0 l- and
n1 l-states (i.e., the corresponding energy levels are close to each other) occurs
near some point Er , Aðn0 Þ and Qðn0 Þ depend essentially on n0 in the vicinity of
this point. With an increase in n0 , Aðn0 Þ increases sharply, whereas Qðn0 Þ
decreases. As a result, the cross-sectional peak becomes asymmetric with
respect to Er .
If the initial energy En0 of the Rydberg molecule is lower than Er , the efficient
vibrational excitation occurs owing to the transition from the covalent n1 L-term to
the Rydberg n1 l-term. This brings about sharp increase in the cross-section near the
point En0 , because the function Aðn0 Þ increases, thanks to the strong mixing of the
n0 l- and n1 l-states. In this case, along with the sharp peaks, the ionization cross-
section (1.112) contains oscillations caused by the fact that the quasiclassical phase
F depends on n0 . Such oscillations of the broadening cross-sections as a result of
quenching Rydberg states of alkali atoms in the rubidium atmosphere were studied
in [94] and reported in [75].
If an atom has the negative scattering length (a < 0), these two features are
supplemented by the appearance of a sharp dip in the cross-section, which is caused
42 G.V. Golubkov et al.

by the Ramsauer effect (i.e., turning to zero of the corresponding element in the
~
scattering K-matrix [68]).

1.10.2 Vibronic Transitions in the XY

+ A System

Let us now turn to the inelastic XY

 A collisions accompanied by the one-

quantum vibrational excitation. The vibronic transitions to the covalent n1 L-states
occurring with a change in the electronic angular momentum ðl0 6¼ l; l0  ðn1  1ÞÞ

XY

ðn0 ; l; v0 ¼ 0Þ þ A ! XY

ðn0 ; l0  1; v ¼ 1Þ þ A; (1.114)

and the transitions to the n1 l-states with the conservation of the angular momentum

XY

ðn0 ; l; v0 ¼ 0Þ þ A ! XY

ðn1 ; l; v ¼ 1Þ þ A: (1.115)

are of greatest interest.

In contrast to the ionization process, the inelastic vibronic transitions
ðn0 ; v ¼ 0Þ ! ðn1 ; v ¼ 1Þ proceed as a result of two passages of the representative
point through the regions where the Rydberg and covalent terms meet quasicross.
Since the corresponding transition probabilities are low, the total cross-sections of
the indicated processes are combined additively from the individual cross-sections,
each corresponding to the one-shot capture by the n1 L-state (accordingly, on the
approach or flying apart). Besides, in contrast to Eq. (1.114), one should include in
this case an additional factor describing the probability to escape from the n1 L-term
in the vicinity of E ¼ Er .
One should set off two regions corresponding to different values of n1 in the
energy dependence of cross-sections. In the first region, the n1 L-terms lie near the
continuum boundary and intersect it at the R
n1 points (Fig. 1.6a). For this reason,
one must take into account the possibility for autoionization in this region.
In the second region (Fig. 1.6b), the covalent term lies in the discrete spectrum
(E < 0). In this case, the cross-section of the inelastic transitions has the form [68]

 

0 Þ þ Aðn0 Þ 
Qðn0 Þ   Qðn
s01 ¼ 2pðRðcÞ
n0 Þ 2
E 3  E 3 : (1.116)
Vc Vc

(En ðxÞ is the integral exponential function). In this case, the ratio of the maximal
cross-section to the cross-section in the plateau region is

s01 ðE ¼ Er Þ=s01 ðE 6¼ Er Þ  1: (1.117)

1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 43

a
E, 10 –3 au
5
1 3 4
2

n1 = 7

–0.4

b
–3.2 3 4
1 5

–3.6

n1 = 6

–4.0

20 40 60 80 100
R , au

Fig. 1.6 The terms of the N

2 þ A fA ¼ He(1), Ne(2), Ar(3), Kr(4), Xe(5)g system classified for
n1 ¼ 7 and n1 ¼ 6 by the type of scattering (L ¼ 0) from the perturbing center. The order in which
the n1 L – terms are arranged is numbered for He through Xe [68]

Thus, the cross-sections of the inelastic vibronic transitions (by analogy with
ionization) must demonstrate “resonance” structure in the dependence of s01 ðn0 Þ
on the initial principal quantum number. Another feature is associated with
the fact that the cross-section increases in the vicinity of the En0 ¼ o  1 2n21
level (where the matrix element KL1 turns to zero for a < 0) owing to the Ramsauer
effect [68].
Note also that the cross-sections calculated for processes (1.114), (1.115) using
the impulse approximation with the hydrogen-like distribution of electron density
[95] coincide at l ¼ 0 with the well-known classical expression [96], which has
limited area of application Vc  1=n0 .
44 G.V. Golubkov et al.

1.10.3 Interaction of the N
2
Molecules with the Noble Gas Atoms

Let us use the collisional ionization and inelastic ðn0 ; v ¼ 0Þ ! ðn1 ; v ¼ 1Þ vibro-
nic transitions to illustrate the theory by the example of the system

N

þ 
2 ð n0 > 40; dsg Þ þ A ! N2 þ A þ e ; (1.118a)

N




2 ð n0 > 40; dsg Þ þ A ! N2 ðn1 ¼ 7; l  1Þ þ A; (1.118b)

N




2 ðn0 > 40; dsg Þ þ A ! N2 ðn1 ¼ 7; dsg Þ þ A; (1.118c)

(where A ¼ fHe, Ne, Ar, Kr, Xeg), whose covalent n1 L-terms intersect the contin-
uum boundary (Fig. 1.6a). For the energy Er ¼ 1:148  104 corresponding to
nr ¼ 66, the system is characterized by the strong nonadiabatic mixing of the
vibronic Rydberg dsg -states ðv ¼ 0Þ and ðv ¼ 1Þ. Figures 1.7 and 1.8 show the
dependence of the cross-sections for processes (1.118a–c) on the initial quantum
number n0 calculated in [68] by formulas (1.112, 1.113). The cross-sections are
characterized by the presence of sharp maxima covering a small group of Rydberg
states. As expected, the obtained curves are asymmetric about the position of level
nr ¼ 66.
Another common property of processes (1.118a–c) is the interference suppres-
sion of the cross-sections at the values of n0 corresponding approximately to the
condition sin Fðn0 Þ ¼ 0. For example, the first interference minima for He atom are
situated at the points n0 ¼ 80 and 147. For Ne, in which the covalent n1 L-term is the
smoothest, three oscillations are observed, respectively, at n0 ¼ 65, 77, and 114.
Since the first interference minimum is situated close to the level nr ¼ 66, processes
(1.118a and b) are only partially suppressed.
The characteristic feature of the interaction with the Ar, Kr, and Xe atoms
(possessing the negative scattering length) consists in the manifestation of the
Ramsauer effect and the resulting decrease in the cross-sections in the vicinity of
the n0 ¼ 59 level. Similar effect should also be observed in the inelastic collisions
of Rydberg atoms X

with atoms [1]. However, as distinct from the atomic
systems, where the Ramsauer minima are rather smooth, narrow localized dips
form in these systems (Fig. 1.8). This is caused by the fact that the cross-sections of
the processes reproduce the characteristic features in energy dependence of the K- ~
matrix, rather than its averaged characteristics (as in the case of collisions proceed-
ing by the impulse mechanism [1]).
For all inert gases (except Xе), ionization cross-section (1.118a) is smaller than
the inelastic-transition cross-section. For the Xе atom (where the ionization process
is rather efficient), they are comparable. In the region of n1 ¼ 6 (Fig. 1.6b), the
vibronic transitions to the covalent n1 L-state with the participation of the Rydberg
n0 ddg series should be the most efficient. In this case, the states with n0  ðnr ¼ 12Þ
should also be strongly nonadiabatically mixed and, as a result, demonstrate
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 45

s, cm2

10–13

He

10–15

2
3
10–17

10–19
3
2

1
10–21

10–23

10–13

Ne

10–15

10–17 3

10–19 1

10–21

10–23
50 75 100 125 150
n0

Fig. 1.7 Cross-sections of collisional ionization (1) and vibronic v ¼ 0 ! v ¼ 1 transitions (2)
with and (3) without a change of the electronic angular momentum in the system
N2

ðn 0 ; dg ; v ¼ 0Þ þ A ðA ¼ fHe and NegÞ. Calculations were carried out in [68] for the relative
velocity Vc ¼ 2  10 4 au
46 G.V. Golubkov et al.

Fig. 1.8 Cross-sections of s, cm

2

collisional ionization (1) and

vibronic v ¼ 0 ! v ¼ 1
10–13
transitions (2) with and (3) Ar
without a change in the 2
–15
electronic angular momentum 10
in the system 3
N


2 ðn 0 ; dg ; v ¼ 0Þ þ –17
10
AðA ¼ fHe and NegÞ. 1
Calculations were carried out
in [68] for the relative 10–19 3
velocity Vc ¼ 2  10 4 au 2

1
10–21

10–23 50 70 90 110 130 150

n0

10–12

Kr

10–14
2

3
10–16

1 3
2
10 –18 1

10–20
40 60 80 100 120 140

–12
10
Xe

10–14
2

10–16 1
3
2
1
10–18

10–20
40 60 80 100 120 140
n0
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 47

Table 1.2 Cross-sections of the inelastic ðn0 ; v ¼ 0Þ ! ðn1 ; v ¼ 1Þ transi-

tions with a change in the electronic angular momentum in the collisions of
N


2 ðn0 dsg ; v ¼ 0Þ molecules with noble gas atoms
n0 s01 ðn0 Þ; 1016 cm2
He Ne Ar Kr Xe
12 466.0 22.8 108.0 6.18 84.0
13 0.58 2.39 0.92 4.38 1.41
14 0.04 0.11 0.51 0.58 0.02
15 0.04 0.61 0.03 0.03 –
The cross-sections are calculated for the relative velocity Vc ¼ 103 аu

appreciably increased cross-sections. The results of calculation performed with the

adiabatic quantum defect [97, 98]

mdpg ð xþ þ
e Þ ¼ 0:11 þ 0:166ð x  xe Þ;

are presented in Table 1.2. Note that the cross-sections calculated near nr ¼ 12 are
two orders of magnitude larger than the gas-kinetic sections.

1.11 Dynamics of the Charge-Exchange Reaction in the X

2 þA
System

As in the case of ionization processes (including the inelastic vibronic transitions),

the admixed (v ¼ 1) state in the charge-exchange reaction can be more active than
the original (v ¼ 0) state. This is caused by the fact that the admixed state is
characterized by a higher transparency of the tunnel barrier. The charge-exchange
process induced by the movement of interacting particles with the relative velocity
Vc is a dynamic transition of the representative point from the Rydberg to the ionic
XY þ þ M term. This process can proceed both at the particle approach stage and at
ðsÞ
the stage of flying apart. The probability Rvv0 that the transition occurs in the vicinity
of the term-crossing point is described by the Landau–Zener formula [26]. In the
conditions under consideration, this probability is low, because the region in which
electron moves is wide and because it must overcome potential barrier (Fig. 1.9).
For this reason, the subsequent transitions from the ionic term to the covalent states,
as well as the transitions in continuum with the deactivation of vibrational excita-
tion in the XY þ ion can be ignored. Then the charge-exchange cross-section can be
calculated by formula [69]

ðss Þ ðsÞ V 2is v; nv0 lL

svv0 ¼ 4p2 ðR vv0 Þ4 ; (1.119)
Vc

here V is v; nv0 lL stands for the interaction of the Rydberg and ionic configurations
ðsÞ
(100) at the point Rvv0 . In the case that a group of states En1 l of the initial Rydberg
48 G.V. Golubkov et al.

E (s)
ion
+ω

R (s)
01
R (s)
00 R
0

E (s)
ion
ε0

Fig. 1.9 Schematics of the adiabatic terms of the system XY

þ M: Boundaries of the regions
classically allowed for the electron motion with v0 = 0 and 1 are shown by the dashed lines. The
subbarrier region transitions are dashed

series (v ¼ 0) is involved in the interaction with the level En0 l of the series (v ¼ 1),
the dependence of the nonadiabatic mixing factor on n0 assumes the resonance
character:

t210
fnðlÞ
1 n0
¼ 3 : (1.120)
ðn0 n1 Þ
p2 ð cos pn 2 ð En0 l  En1 l Þ2 þ t210
0 cos pn 1Þ

1.11.1 Charge Exchange in the Reaction of N

2 with the
Molecules NO and O2

In the case of charge exchange on the NO molecule, the ratio sðv ¼ 1Þ=sðv ¼ 0Þ is
much greater that unity starting at n0 >10. The NO molecule is known to possess a
low electron affinity Er ¼ 0:024 eV [99], whereas the molecular term of NO is
stable only in the state with s ¼ 0. The calculation of the reaction

N

0 þ  
2 ðv ¼ 0Þ þ NO ðs ¼ 0Þ ! N2 ðv ¼ 1Þ þ NO ðs ¼ 0Þ; (1.121)

(performed for the n0 ppu and n0 ddg -series using the data from [3]) clearly reflects
the irregular character of the dependence of the charge-exchange cross-sections
(1.121) from the n0 ppu and n0 ddg states of the nitrogen molecule on the level
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 49

s(i01), cm2

–15
10

–16
10

–17
10

–18
10

–19
10

–20
10

0 20 40 60 80
n0

þ 
Fig. 1.10 The plot charge exchange cross-section N

2 ðn0 lLÞ þ O2 ! N2 þ O2 as a function of
the initial principal quantum number n0 for the Rydberg series ppu (1) and ddg (2). Calculations
were carried out in [100] for the relative velocity Vc ¼ 4  10  4 au

number n0 , as is seen from the presence of sharp splashes whose positions depend
on the initial state of a highly excited molecule (Fig. 1.10). As a further application
of the theory, we present the computational results obtained for the reaction [100]

N

0 þ  
2 ðv ¼ 0Þ þ O2 ðs ¼ 0Þ ! N2 ðv ¼ 1Þ þ O2 ðs ¼ 3Þ; (1.122)

playing an important role in the upper and middle Earth atmosphere. The electron
affinity of oxygen molecule is Er ¼ 0:44eV [101]. Because of this, the negative ion
O 
2 ð pg Þ has four (s ¼ 03) stable vibrational states lying below the ground-state
2
3 
(X Sg ) energy of O2 . The computational results obtained for the charge-exchange
reaction cross-sections (1.122) are given in Figs. 1.11 and 1.12 for various Rydberg
lL series. These results clearly demonstrate that the dependence of cross-sections
on n0 has irregular character and demonstrates a noticeable increase for l  1 as L
increases. For n0 < 20, all the above-mentioned series contain the regions of a
strong nonadiabatic coupling. The resonance splashes in the region of large princi-
pal quantum numbers (n0 > 20) are the sharpest. This occurs for n0 ppu and n0 ddg
50 G.V. Golubkov et al.

a s, cm2
10–17

10–19

10–21

10–23

10–19

10–21

10–23

c
10–17

10–19

10–21

10–23
0 10 20 30 40 50 60 70 80
n

þ 
Fig. 1.11 The plot charge exchange cross-section N

2 ðn0 lLÞ þ O2 ! N2 þ O2 as a function of
the initial principal quantum number n0 for the Rydberg ssg ðaÞ, psu ðbÞ, and ppu ðcÞ series.
Calculations were carried out in [100] for the relative velocity Vc ¼ 1:2  10  3 au
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 51

a
s, cm2
10–15

10–17

10–19

10–21
b
10–20

10–22

10–24

c
10–17

10–19

10–21

0 10 20 30 40 50 60 70 80
n

þ 
Fig. 1.12 The plot charge exchange cross-section N

2 ðn0 lLÞ þ O2 ! N2 þ O2 as a function of
the initial principal quantum number n0 for the Rydberg dsg ðaÞ, dpg ðbÞ, and ddg ðcÞ series.
Calculations were carried out in [100] for the relative velocity Vc ¼ 1:2  10  3 au
52 G.V. Golubkov et al.

series, where the resonances are formed with the participation of the groups of
states and the strong vibronic-coupling effect is the most pronounced. For the
n0 ppu -series, the resonance occurs at n0 ¼ 26, and for the n0 ddg -series, at
n0 ¼ 66. The corresponding cross-sections at these points are equal to
4:0  10 18 and 4:8  10 18 cm2.

1.12 Nonadiabatic Effects in the l-Mixing Processes

The nonadiabatic electron–rotation coupling affects the PES structure of a quasi-

molecule consisting of the interacting two-atomic Rydberg molecule and a neutral
particle A (A is the noble gas atom), and to a great extent determines the character of
the processes proceeding in these systems. The role of rotational nonadiabatic
coupling in the collisional transitions is as yet poorly known. Here, we discuss it
in detail by the example of l-mixing process, i.e., of a change in the angular
momentum l of Rydberg electron in slow collision with the perturbing neutral
particle B [83]. This process constantly draws the attention of researchers, because
its cross-section reaches s  10 12 10 11 cm2 in the range n  1040 of the
principal quantum number.
In [84], the cross-sections of the elastic and inelastic transitions in a hydrogen
molecule excited to the Rydberg np0ð1 Sþ u Þ and np2ð Pu Þ series of the optical Rð0Þ-
1

branch characterized by the strong nonadiabatic coupling with rotation were calcu-
lated. According to the composition rules, the nuclear angular momentum N takes
in this case the values 0 and 2. It is known that, as regards the ion rotational motion,
the Rydberg spectrum of hydrogen molecule can be divided in the three character-
istic regions: adiabatic, nonadiabatic, and essentially nonadiabatic.
The first corresponds to the situation in which the period of nuclear rotation is
much longer than the characteristic rotation period of a weakly bound electron (i.e.,
Bn3  1), so that the projection L of electronic angular momentum l on the
molecular axis is conserved (Hund’s case “b” coupling [102]). As the principal
quantum number n increases, the regular level arrangement breaks down and the
spectrum structure becomes more complicated [103]. In the majority of states, the
electron–vibration coupling becomes much less pronounced; i.e., the spectrum
reproduces the v-repetitive system of terms shifted relative to each other by the
energy of vibrational quantum o. For this reason, the analysis of the nonadiabatic
effects can be divided in two steps. At the first step, one should consider the
nonadiabatic electron–rotation coupling and eliminate the vibrational motion (by
fixing the interatomic distance in the equilibrium internuclear position Rþ e ). At the
second step (if necessary), one can include electron–vibration interaction. Here, we
restrict ourselves only to the first step and study the processes in a fixed vibrational
state (e.g., v ¼ 0).
For the operator equation determining the PES of the XY

þ A system and the
interaction between the different groups of states, one has in this case [70]
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 53

 JM   JM 
X C ClN 
t¼ K~N lN
t
ll
; N
E  E JnlN
X
þ K~N pN cot pnN j YN ML i h YN ML j t ð1:123Þ
N; ML

(ML is the projection of the nuclear angular momentum N on the direction of radius
vector R). The quantity K~N ðRÞ is the diagonal element of the K-matrix describing
the free-electron elastic scattering from particle A in the state with orbital moment
L ¼ 0. This element coincides, to an accuracy of sign and the terms
 quadratic in pN ,
with the S-scattering amplitude determined at energy eN ¼ p2N 2. Formula (1.123)
contains the total wave functions (with regard to the influence of core and strong
nonadiabatic electron–rotation coupling effects) of the H2 molecule for the np0 and
np2 series, which represent the following superpositions [67]:
X
C JM ^
lN ð r; x; R Þ ¼ CNN 0 ðEnlN0 ÞQlN 0 ðrÞfi ðxÞF JM
lN 0 ð^
r ; RÞ; (1.124)
N0

where CNN 0 are the corresponding expansion coefficients, fi is the electron wave
function of the H+2 ion, and the functions QlN and F JM lN are given by expressions
(1.75) and (1.77).
The l-mixing process is defined as a dynamic transition from the m-mixed groups
of states with a fixed value of l (for l  l
) to the strongly m- and l-mixed groups of
states with l > l
. Note that the angular dependence of the wave function of Rydberg
electron is weak for the transitions from the states with l  l
and has no effect on the
dynamics of the process. The dependences of the quantum defects on the number n
and the quasiclassical oscillations of radial wave functions are more significant.
Analysis of the specific features of the l-mixing cross-sections can be carried out
using the quasi-elastic transitions as an example,

H




2 ð npNÞ þ Xe ! H2 ð nL L NL Þ þ Xe ; (1.125)

where NL takes the values 0 or 2. Indeed, for n  10, this process proceeds with
a small (when compared with the energy of rotational excitation) change DEnlN
/ mNL n3NL in the total energy of Rydberg molecule. At low relative energies of the
colliding particles, Ec  6B, only one component of the initial rovibronic state is
actually involved in this process. Recall also that the Rydberg npN-series of the
optical Rð0Þ-branch in the H

2 molecule are not autodecaying and possess, in the
spectral range of interest, radiative lifetimes (trad  106 s) considerably longer
than the characteristic collision times (t  1011 s).
One of the most characteristic features of the H

2 molecule is that it can
demonstrate the stroboscopic effect [66, 67] if the separate groups of states,
whose classification corresponds to the adiabatic representation, form in the spec-
trum. In the classical limit [65], it allows for a simple interpretation: the molecular
54 G.V. Golubkov et al.

axis assumes the same orientation in a sequence of interaction events between the
Rydberg electron and ion core. This effect appears if n0 and n2 satisfy the simple
relation: n 2 ¼ n 0  k ðk ¼ 1; 2; :::Þ. In this case, the adiabatic quantum defects
ms ¼ 0:191 and mp ¼ 0:078 are the roots of the secular equation

D ¼ ð tan pn0  t00 Þ ð tan pn2  t22 Þ  t202 ¼ 0; (1.126)

describing the spectrum of highly excited molecular states [67]. This has a simple
physical meaning and signifies that, when scattering from the ion core, the Rydberg
electron passes to new phase trajectory without changing its phase. Since the
quantities n0 and n2 satisfy the relation  21n2 ¼ 6B  21n2 , this situation, as noted
0 2
earlier, can be realized under the condition 6Bn2  1 and 6Bn3 k. In the hydro-
gen molecule, this spectral region is situated in the 10  n  30 interval.
The rotational nonadiabatic coupling in the npN ! nL LNL transitions should
manifest itself through the dependence of quantum defects mN ðnÞ on the level
number, because DEðtÞ in Eq. (1.73) strongly depends on n. As to the transitions
with a change in the ion rotational state in the npð0Þ and n0 pð2Þ series, the latter
should be markedly different if the principal quantum numbers n and n0 for the next
closely spaced pairs npð0Þ ! nL L; NL ¼ 2 or n0 p ð2Þ ! nL L; NL ¼ 0 also strongly
differ from each other. In the first case, the term L split out from the group of
orbitally degenerate states plays a significant part; for instance, the transitions
following the term-quasicrossing mechanism become possible at n0  n. In
contrast, the distortions in the term L can be disregarded at n0  n in the second
case. This is clearly demonstrated in Figs. 1.13 through 1.16, where the l-mixing

s,10–12 cm2
1.0

0.5

0
10 15 20 25 30
n

Fig. 1.13 Dependence of the H

2 ðnp; N ¼ 0Þ þ Xe ! H2 ðnL L ; NL ¼ 0Þ þ Xe cross-section on
the initial principal quantum number n. Dotted line (●) denotes the cross-section of quasi-elastic
nps ! nL L transition calculated in the rotation adiabatic approximation [84], solid line (~)
represents the calculation results with rotation coupling taken into account [70]
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 55

s,10–12 cm2

1.0

0.5

0
10 15 20 25 30
n

Fig. 1.14 Dependence of the H

2 ðnp; N ¼ 0Þ þ Xe ! H2 ðnL L ; NL ¼ 2Þ þ Xe cross-section on
the initial principal quantum number n. Dotted line (●) denotes the cross-section of quasi-elastic
nps ! nL L transition calculated in the rotation adiabatic approximation [84], solid line ()
represents the calculation results with rotation coupling taken into account [70]

cross-sections are shown as functions of the initial principal quantum number n for
the following quasi-elastic transitions [70]:
(
nL L; NL ¼ 0
npð0Þ ! : (1.127)
nL L; NL ¼ 2

They were calculated for the relative velocity Vc ¼ 9  10 4 with the para-
meters taken from [66]. As expected, the sharp nonmonotonic dependence on n is
the most characteristic feature of these cross-sections. The envelope of the cross-
sections has a typical bell-shaped form with a maximum of / 10 12 cm2. For
comparison, the curves calculated previously in the rotationally adiabatic approxi-
mation for the nps - and npp -states correlating at n  10 with the np0- and np2-
series are also shown in these figures [84].

1.13 Many-Center Perturbation of the Atomic Rydberg States

We now turn to the interaction of a Rydberg atom A

with the system of particles
situated near or inside this atom. To describe the spectrum in a rarefied gas, one can
restrict oneself to the perturbation introduced by only one particle [15, 27]. In a
denser medium, one must solve the problem of the influence of a force field
produced by the finite number N of neutral particles falling within the volume of
a highly excited atom. The dependence of N on the principal quantum number n and
56 G.V. Golubkov et al.

s,10–12 cm2

2.0

1.5

1.0

0.5

10 15 20 n

Fig. 1.15 Dependence of the H

2 ðnp; N ¼ 2Þ þ Xe ! H2 ðnL L ; NL ¼ 0Þ þ Xe cross-section on
the initial principal quantum number n. Dotted line (●) denotes the cross-section of quasi-elastic
nps ! nL L transition calculated in the rotation adiabatic approximation [84], solid line ()
represents the calculation results with rotation coupling taken into account [70]

concentration c of medium in unit volume is presented in Table 1.3 (a concentration

of 101 7 cm3 in the Earth atmosphere corresponds to the altitude 40 km, and
101 3 cm3 is attained at 100 km).
Since these particles are well separated in space, the electron scattering by each
s center does not correlate with the others. For this reason, the results obtained
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 57

s,10–12 cm2
2.0

1.5

1.0

0.5

10 15 20 n

Fig. 1.16 Dependence of the H

2 ðnp; N ¼ 2Þ þ Xe ! H2 ðnL L ; NL ¼ 2Þ þ Xe cross-section on
the initial principal quantum number n. Dotted line (●) denotes the cross-section of quasi-elastic
nps ! nL L transition calculated in the rotation adiabatic approximation [84], solid line (~)
represents the calculation results with rotation coupling taken into account [70]

above for the one-center electron scattering from one neutral particle B can easily
be extended to the many-center systems situated in the region of its action. Then,
assuming that the electron interaction with these N centers is additive, i.e.,

X
N
UðrÞ ¼ Uðr  Rs Þ;
s
58 G.V. Golubkov et al.

Table 1.3 N as a function of the principal quantum number n and concentration c

c (cm3) 10 1 3 10 1 4 10 1 5 10 1 6 10 1 7
n 45 60 30 40 20 27 15 18 10 13
N 1 5 1 5 1 5 1 5 1 5

one can get equations of the multiple-scattering theory, in which the solutions of
many-center problems reduce to the solution of the systems of algebraic equations
ð0Þ
containing the one-center scattering operators Ks ¼ Ke Bs :
In studying the general properties of such systems, it is convenient to restrict
oneself first to the study of a few-center situation, in which, on the one hand, the
interparticle correlations inherent in the complex systems are strictly taken into
account and, on the other hand, the characteristic features of energetic structure
containing a set of groups of levels are also allowed for. The presence of many
scattering centers placed outside or inside the atom A

breaks the axial and
azimuthal symmetry of interaction (when compared with the simplest two-center
system). The appearance of many l and m components of the Rydberg levels must
introduce certain changes also in the electronic structure of the system, as a result of
which the processes of electronic transitions between the Rydberg atom and neutral
cluster, as well as the autoionization processes of various kinds can become
efficient.
The corresponding system of equations for the level-shift operator takes the
form [27]

t ¼ TðGA

 G0 Þ t; (1.128)
P
T¼ Ts , where Ts is the operator for electron scattering from the s center. This
s
operator takes into account the rescattering from other centers and satisfies the
following equation:
X
T s ¼ ts þ ts G0 ðRs ; Rs0 Þ Ts0 : (1.129)
s0 6¼s

The function G0 in Eq. (1.129) is defined in Eq. (1.10) as a function of the

positions of vectors Rs ; Rs0 relative to the ion Aþ , with the momentum pe being
taken at the point R ¼ 12 jR1 þ R2 j. For instance, in a system of two scattering
centers,

1 h i
T1 þ T2 ¼ t1 þ t2 þ t 1 G0 ð R1 ; R2 Þ~
t2 þ t 2 G0 ð R2 ; R1 Þ~
t1 ; (1.130)
D

where

D ð R1 ; R2 Þ ¼ 1  t1 t2 G20 ð R1 ; R2 Þ (1.131)
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 59

is the determinant of the two-center system. The arrowed operators in Eq. (1.130)
are defined as

hFðrÞj t s ¼ ts FðRs Þ

and similarly for the operator acting toward the right. The contribution from the
combination t1 t2 is zero, because these operators act on different centers. For
definiteness, we will assume that the matrices ts of electron scattering from the
s center correspond to the S-scattering (L ¼ 0). Note that, according to Eq. (1.10),
the operator ts corresponds to the scattering K-matrix [3], which depends only on
ts ¼
the scattering length in the ZRP model; i.e., p 2p=ks ¼2pas . If Rs > 1=jEj; ts is
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
replaced by ts ¼ 2p=ðks  as Þ, where as ¼ 2 ðE þ 1=Rs Þ. Inasmuch as, under
certain conditions, determinant (1.131) can turn to zero, we consider this situation
in more detail.
If the centers 1 are identical and situated in the classically allowed region of
electron motion, one has for determinant (1.131) [27]
  
a cos pe ðRÞjR1  R2 j a cos pe ðRÞjR1  R2 j
DðR1 ; R2 Þ ¼ 1  1þ ; (1.132)
jR1  R2 j jR1  R2 j

where

1
R ¼ jR1 þ R2 j  1:
2

Note that near this pole, representation (1.130) is separable,

/ / ! !
1 t1 t2 ~
t1 ~
t2
T1 þ T2 ¼ pffiffiffiffiffiffi þ pffiffiffiffiffiffi pffiffiffiffiffiffi þ pffiffiffiffiffiffi ; (1.133)
D jt1 j jt2 j jt1 j jt2 j

which greatly simplifies further analysis. The roots of equation obtained from the
condition that the right-hand side of Eq. (1.129) turns to zero is denoted by Ur ðRÞ.
These roots depend parametrically on the ratio  ¼ d=a. If the centers are situated
on a sphere of radius R<1=jEj around the atom A

and the distance d between them
is small, i.e., 1 < d ¼ j R1  R2 j < jaj < R / n, the solution to this equation is
written as
  2
1 1 1 d
Ur ðRÞ ¼  þ Er ; Er ¼ arccos ; (1.134)
R 2 d jaj

where the position Er of the resonance level does not depend on the sign of
scattering length a. Consequently, the electron scattering from two centers with
momentum pe ¼ d1 arccos ðd=jaj Þ is accompanied by the formation of a quasi-
bound state near which the electron density is strongly redistributed. For small
60 G.V. Golubkov et al.

values of the argument, pe d  1, the position of this term in the region R<1 =jEj is
given by the expression
 
1 1 1
Er ¼  : (1.135)
d d j aj

Note also that, under the more rigid condition R ¼ 12 j R1 þ R2 j  jaj, this
situation appears also in the case where the scattering centers are arranged arbi-
trarily. Clearly, the energy spectrum is expected to be strongly rearranged near the
regions Ur ðRÞ  2n12 at a given n. Inasmuch as term (1.133) corresponds to the
electron localization at these two centers, there is full analogy with the ionic
configuration in the region E > 1=R, where, on passing to continuum with an
increase in R, it can become the autoionizing state of the system.
The terms of the ionic configuration Aþ  ð2BÞ can also form in the classically
forbidden region R > 1=jEj. The equation determining their position takes the form
  
eae d eae d
D¼ 1 1þ ¼0
ðk  ae Þ d ðk  ae Þ d
ð1Þ ð2Þ
and, in the general case, has two solutions Er and Er . It also follows from this
equation that its solutions must turn to zero at the turning point ae ¼ 0 where
ðRÞ
d ¼ j aj. Since the Rydberg wave functions Fnl0 ð RÞ decrease exponentially at
R > 1=jEj, it is natural to restrict oneself to the consideration of the small values
ð1Þ
of ae d. Under these conditions, the first solution Er coincides with expression
(1.135), and the second is
 
1 1 1
Eð2Þ
r ¼  þ 2
: (1.136)
8 d a
ð1Þ ð2Þ
The solutions Er and Er can exist simultaneously only in the case of 0 < a < d,
ð2Þ
whereas, at 1 < d < j aj, there is only one solution Er . Note that the case
j aj ¼ d, where ionic terms (1.134 through 1.136) merge with the Coulomb poten-
tial, corresponds to the formation of a virtual state with the zero binding energy.
One more feature arises under the condition k ¼ ae . At this point, D ! 1 and
scattering T operator (1.133) turns to zero; i.e., a “transparency window” appears
in the forbidden region and the initial state is not perturbed at the distances
l jaj
R ¼ pnffiffiffiffiffiffiffiffiffi
ffi >2n2 for a given energy Enl of the Rydberg level. Such a situation
a nl
2 2

must take place at nl / jaj, which corresponds to the small values of the principal
quantum number n. This result is not surprising and is well known in the quantum
mechanics [104]. In the subsequent analysis of the term behavior in the vicinities
of the corresponding quasicrossings, this particular case can be eliminated. It
is significant that, for a given arrangement of the scattering centers relative to
ion Aþ , the transition from the classically allowed region of electron motion
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 61

(corresponding to the large values of n) to the forbidden region (with small n)

should be accompanied by a change in the sign of electron affinity in the ionic
configurations. In addition, these and the nearest lying Rydberg terms must be
ðRÞ ðRÞ
strongly perturbed in the range U nlm ðRÞ < Ur ðRÞ < U nþ1;lm ðRÞ, resulting in a
change in their wave functions and in the corresponding electron densities inside
the volume of Rydberg atom. It is natural that these effects become more pro-
nounced with the increasing number of scattering centers.
Therefore, three main groups of states formed by the families of one-center
scattering harmonics with the nonzero angular momenta (L 6¼ 0) can simulta-
neously exist in the systems containing N  2 scattering centers. The stationary
states (with energies j E j < 1=R) relate to the first group and the ionic states (with
energies 0 < jUr j < 1=R) strongly interacting with the Rydberg configurations relate
to the second group. Finally, the third group includes the autoionizing states with
negative electron affinity, which should be observed at sufficiently large n.

1.14 Conclusions

In this book, the methods and results of systematic studies on the interaction of
Rydberg atoms and molecules with individual neutral particles (atoms or mole-
cules) are considered. To determine PES of the systems under consideration at
intermediate distances R / n, it is necessary to take into account the interaction
between a positively charged ion and atom B by introducing the operators for
scattering from each of them and then express their matrix elements through the
characteristics of electron elastic scattering without recourse to the quasiclassical
approximation.
Inasmuch as the electron momentum in the e  B interaction is not conserved
in the system of three particles, it is required to define the scattering K operator
outside of the energy surface. This problem can be solved on the following two
counts. First, the scattering amplitude at the short-range portion of the potential of
e  B interaction is factorized because of the behavior of the Bessel function at
small distances from the perturbing center. Second, there is a small parameter – the
ratio of electron mass to the reduced mass of the particles Aþ and B. Inclusion of the
long-range interaction at intermediate distances R gives rise to the azimuthal
dependence of the effective interaction. For this reason, the states of an A

B
quasimolecule become dependent on the projection of electronic angular momen-
tum m about Aþ on the axis connecting the interacting centers. Because of the axial
symmetry of interaction, these projections are integrals of motion, so that the states
with different values of m do not mix.
The study of a two-center system possessing axial symmetry shows that the
inclusion of long-range interaction (1.23) leads to the redefinition of the potential
e  B-scattering length and to the appearance of an additional term describing the
electrostatic interaction in Eq. (1.33). This term allows for the polarization of atom
B in the field of ion Aþ and for the effective electrodipolar interaction of atom with
62 G.V. Golubkov et al.

electron. At the intermediate distances R, the role of the latter proves to be greater
than that in the asymptotic region. At these distances, the centrifugal potential also
plays an important role. Its influence amounts to a drastic decrease in the number of
degenerate Coulomb states forming covalent term and to the breaking of the
quasiclassical description used in the asymptotic approach to the electron motion,
thereby bringing about a more severe distortion of the Rydberg states by virtue of
the increase in electron density at the intermediate distances. In the construction of
covalent terms, the asymptotic theory can be used only for R  n. At the interme-
diate distances R between centers, this inequality greatly restricts the area of its
applicability; i.e., this theory becomes valid only for small n. For such values of the
principal quantum number, the quasiclassical electron momentum pe is small and
the S-scattering approximation is valid. However, with a decrease in n, the momen-
tum increases, and one must also consider higher harmonics when describing the
electron scattering from atom B.
The additional features arise in PESs for the processes involving Rydberg
molecules XY

. They are caused by the strong nonadiabatic coupling between
the electron and nuclear motions, i.e., by the fact that each Rydberg nlNv-state (for
small l) is a superposition of at least two closely spaced v series (which can be more
in number if the rotation is taken into account). As a result, the processes become
dependent on the principal quantum number n (even after averaging over the
relative velocities of the colliding particles). This is particularly important in the
analysis of the kinetic phenomena under nonequilibrium conditions.
We have discussed the main structural features of the vibronic PESs describing
the interaction of highly excited atoms and molecules with neutral particles. The
asymptotic method presented in this review is an alternative to the traditional
quantum mechanical approaches. We do not use the variational principle and
determine PES of the system from the algebraic equations containing information
on the free-electron motion in the field of isolated fragments (of ion and neutral
particle). The theory operates with the scattering T and K matrices, which are the
fundamental characteristics in the quantum theory of scattering, whereas the para-
meters appearing in it can be directly related to the experimentally observed
amplitudes (or phases) of electron scattering from ions, atoms, and molecules.
The unique feature of the asymptotic method is that it allows many-sheet PES to
be represented in a simple analytic form with the whole diversity of the regions of
mutual approach and quasicrossings that are responsible for the nonadiabatic
transitions in the system.
The objects considered in this work possess remarkable properties. The terms
split out from the group of orbitally degenerate states of the Rydberg atom X
or
molecule XY

demonstrate anomalous behavior. In reality, these terms do not
depend on the nature of Rydberg particle (as well as on the orientation of the
perturbing molecule M). An important feature of such (valence) configurations is
that they are capable of autoionization after the term passes to continuum as a
result of the mutual approach of interacting fragments. Moreover, weakly bound
states caused by the Ramsauer effect in the elastic electron scattering from the
perturbing particle can form in the Rydberg complexes. They represent shallow,
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 63

though rather wide, potential wells that are responsible for the existence of weakly
bound states of the Rydberg complex (of the X

B or XY

B type) at large distances
from the ion core.
Next, the perturbations of the atomic Rydberg states in the field of chaotically
arranged neutral particles have been analyzed. The interaction with them splits
levels into a group of sublevels corresponding to the different values of the angular
momentum l of weakly bound electron and of its projection m onto the axis
connecting the positive ion and the center of gravity of the particles falling into
the electronic cloud of the Rydberg atom. This gives rise to a complex energy
structure containing many pseudocrossings. These electronic states can change in
the presence of a regular structure in the arrangement of the centers. Because of the
numerous mutual approaches and term quasicrossings of different groups of states,
the processes of electron transfer between different centers become rather active.
This opens up new spectroscopic possibilities in the study of physical properties of
the nanoclusters composed of atoms with a rather complex electronic structure.

Acknowledgment This work was supported by Russian Foundation for Basic Research, project
N. 10-03-00737.

References

1. Stebbings, R., Dunning, F. (eds.): Rydberg States of Atoms and Molecules. Cambridge
University Press, Cambridge (1983)
2. Golubkov, G.V., Ivanov, G.K.: Dynamic behavior of highly excited molecular states in the
strong monochromatic laser fields. Russ. Chem. Bull. 43(3), 327–345 (1994)
3. Golubkov, G.V., Ivanov, G.K.: Rydberg States of Atoms and Molecules and Elementary
Processes with Their Participation. URRS, Moscow (2001)
4. Fielding, H.H., Softley, T.P., Merkt, F.: Photoionisation and ZEKE photoelectron spectroscopy
of Ar, H2 and CO2 using a coherent XUV laser source. Chem. Phys. 155, 257–265 (1991)
5. Muller-Dethlefs, K., Schlag, E.W.: High-resolution zero kinetic energy (ZEKE) photoelec-
tron spectroscopy of molecular systems. Annu. Rev. Phys. Chem. 42, 109–136 (1991)
6. Merkt, F., Softley, T.P.: Final-state interactions in the zero-kinetic-energy-photoelectron
spectrum of H2. J. Chem. Phys. 96, 4149–4156 (1992)
7. Merkt, F., Fielding, H.H., Softley, T.P.: Electric field effects on zero-kinetic-energy photo-
electron spectra: an explanation of observed trends. Chem. Phys. Lett. 202, 153–160 (1993)
8. Larsson, M., Carlson, M., Danared, H., et al.: Vibrational cooling of D2+ in an ion storage
ring as revealed by dissociative recombination measurements. J. Phys. B27, 1397–1406
(1994)
9. Larsson, M., Broström, L., Carlson, M., et al.: Dissociative recombination of H2+ studied in
CRYRING. Phys. Scr. 51, 354–358 (1995)
10. Tanabe, T., Katayama, I., Kamegaya, H., et al.: Dissociative recombination of HD+ with an
ultracold electron beam in a cooler ring. Phys. Rev. Lett. 75, 1066–1069 (1995)
11. Schneider, I.F., Stromholm, C., Carata, L., et al.: Rotational effects in HD+ dissociative
recombination: theoretical study of resonant mechanisms and comparison with ion storage
ring experiments. J. Phys. B30, 2687–2706 (1997)
12. Balashov, E.M., Dalidchik, F.I.: Current peculiarities of scanning tunneling microscope
close by the collective state excitation thresholds of low-sized regular structures. Khim.
Fiz. 23(6), 38–42 (2004)
64 G.V. Golubkov et al.

13. Grishin, M.V., Dalidchik, F.I., Kovalevsky, S.A., Kovytin, A.V.: Atomic and electronic
structure of the surface nanoscaled graphite systems. Khim. Fiz. 23(7), 83–90 (2004)
14. Balashov, E.M., Dalidchik, F.I., Shub, B.R.: Structure and electronic characteristics of
imperfective oxides and nano-oxides. Khim. Fiz. 27(12), 10–27 (2008)
15. Golubkov, G.V., Ivanov, G.K.: Rydberg atoms and molecules in the medium of neutral
particles. Khim. Fiz. 22(10), 25–86 (2003)
16. Ivanov, G.K., Bodneva, V.L., Golubkov, G.V.: Perturbation of highly excited atomic states
by a system of chaotically and orderly arranged interaction centers. Chem. Phys. Rep. 19
(10), 1797–1809 (2001)
17. Golubkov, G.V., Golubkov, M.G., Ivanov, G.K.: Perturbation of Rydberg atom A

states by
neutral particles at intermediate distances. Khim. Fiz. 24(5), 3–13 (2005)
18. Ivanov, G.K.: Perturbation of the degenerate atom states by the neutral particles. Opt.
Spectrosc. 43, 1044–1049 (1977)
19. Ivanov, G.K.: Interaction between atoms in the electron-excited quasimolecules. Teor. Eksp.
Khim. 15(6), 644–650 (1979)
20. Golubkov, G.V., Ivanov, G.K.: Transitions between Rydberg states of diatomic molecules in
slow collisions with atoms. Z. Phys. A319, 17–23 (1984)
21. Ivanov, G.K.: Perturbation of the single-electron atom states by the field of diatomic
molecule. Teor. Eksp. Khim. 10, 303–309 (1974)
22. Freeman, R., Kleppner, D.: Core polarization and quantum defects in high-angular-momen-
tum states of alkali atoms. Phys. Rev. A14, 1614–1619 (1976)
23. Davydov, A.S.: Atomic Nucleus Theory. Fizmatgiz, Moscow (1958)
24. Bateman, H., Erdélyi, A.: Higher Transcendental Functions, vol. I. McGraw-Hill, New York
(1953)
25. Ivanov, G.K.: Finite-radius potential approximation for the atomic systems. Teor. Eksp.
Khim. 10, 450–455 (1974)
26. Landau, L.D., Lifshitz, E.M.: Quantum Mechanics. Nonrelativistic Theory. Pergamon,
Oxford (1978)
27. Golubkov, G.V., Golubkov, M.G., Ivanov, G.K.: Highly excited electronic states of nano-
scaled systems. Khim. Fiz. 24(9), 3–18 (2005)
28. Drukarev, G.F., Collisions of Electrons with Atoms and Molecules, Nauka, Moscow (1978);
Plenum, New York (1987)
29. Golubkov, G.V.: Exchange interaction of highly excited atom A

with non-excited atom A
at intermediate distances. Khim. Fiz. 25(2), 3–14 (2006)
30. Oppenheimer, B.R., Kulkarni, S.R., Matthews, K., et al.: Infrared spectrum of the cool
Brown Dwarf Gl 229B. Science 270, 1478–1479 (1995)
31. Mayor, M., Queloz, D.: A Jupiter-mass companion to a solar-type star. Nature 378, 355–359
(1995)
32. Burrows, A., Hubbard, W.B., Lunine, J.I., et al.: The theory of brown dwarfs and extrasolar
giant planets. Rev. Mod. Phys. 73, 719–765 (2001)
33. Christova, M., Allard, N.F., Kielkopf, J.F.: New line profiles of sodium and potassium
perturbed by helium for brown dwarf and very cool white dwarf stars. In: Gigosos, M.A.,
Gonzalez, M.A. (eds.) Spectral Line Shapes, vol. 15; XIX ICSLS. Valladolid. Spain,
pp. 308–310. Melville, New York (2008)
34. Burrows, A., Burgasser, A.J., Kirkpatrik, J.D., et al.: Theoretical spectral models of T dwarfs
at short wavelengths and their comparison with data. Astrophys. J. 573, 394–417 (2002)
35. Burgasser, A.J., Kirkpatrik, J.D., Liebert, J., et al.: The spectra of T dwarfs. II. Red optical
data. Astrophys. J. 594, 510–524 (2003)
36. Seager, S., Sasselov, D.D.: Theoretical transmission spectra during extrasolar giant planet
transits. Astrophys. J. 537, 916–921 (2000)
37. Brown, T.M.: Transmission spectra as diagnostics of extrasolar giant planet atmospheres.
Astrophys. J. 553, 1006–1026 (2001)
38. Jayawardhana, R.: Unraveling brown dwarf origins. Science 303, 322–323 (2004)
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 65

39. Lodders, K.: Brown dwarfs–faint at heart, rich in chemistry. Science 303, 323–324 (2004)
40. Burrows, A.: A theoretical look at the direct detection of giant planets outside the solar
system. Nature 433, 261–268 (2005)
41. Chung, H.-K., Kirby, K., Babb, J.F.: Theoretical study of the absorption spectra of the
sodium dimer. Phys. Rev. A63, 032516 (2001)
42. Zhu, C., Babb, J.F., Dalgarno, A.: Theoretical study of pressure broadening of lithium
resonance lines by helium atoms. Phys. Rev. A71, 052710 (2005)
43. Zhu, C., Babb, J.F., Dalgarno, A.: Theoretical study of sodium and potassium resonance lines
pressure broadened by helium atoms. Phys. Rev. A73, 012506 (2006)
44. Allard, N.F., Allard, F., Hauschildt, P.H., et al.: A new model for brown dwarf spectra
including accurate unified line shape theory for the Na I and K I resonance line profiles.
Astron. Astrophys. 411, L473–L476 (2003)
45. Burrows, A., Volobuyev, M.: Calculations of the far-wing line profiles of sodium and
potassium in the atmospheres of substellar-mass objects. Astrophys. J. 583, 985–995 (2003)
46. Pascale, J.: Use of l-dependent pseudopotentials in the study of alkali-metal-atom–He
systems. The adiabatic molecular potentials. Phys. Rev. A28, 632–644 (1983)
47. Kumar, A., Lane, N.F., Kimura, M.: Quenching of low-lying Rydberg states of Na colliding
with ground-state He: a semiclassical approach. Phys. Rev. A39, 1020–1028 (1989)
48. Geum, N., Jeung, G.H.: Undulating potential curves of the Rydberg states of NaHe. Chem.
Phys. Lett. 333(9), 314–318 (2001)
49. Theodorakopoulos, C., Petsalskis, I.D.: Potential energy curves and radiative lifetimes of
Rydberg states of NaHe. J. Phys. B26, 4367–4380 (1993)
50. Buenker, R.J., Liebermann, H.P.: Private communications (2008)
51. Golubkov, G.V., Golubkov, M.G., Ivanov G.K.: Rydberg atom A

in a field of neutral
atom B. In: Gigosos, M.A., Gonzalez, M.A. (eds.) Spectral Line Shapes, vol. 15; XIX ICSLS,
Valladolid, Spain, pp. 140–142. Melville, New York (2008)
52. Santra, R., Kirby, K.: Ab initio configuration-interaction investigation of optical transitions
in KþHe and KþH2. J. Chem. Phys. 123, 214309 (2005)
53. Szabo, A., Ostlund, N.S.: Modern Quantum Chemistry. Dover, Mineola, New York (1966)
54. Werner, H.-J., Knowles, P.J.: An efficient internally contracted multiconfiguration – refer-
ence configuration interaction method. J. Chem. Phys. 89, 5803–5814 (1988)
55. Knowles, P.J., Werner, H.-J.: An efficient method for the evaluation of coupling coefficients
in configuration interaction calculations. Chem. Phys. Lett. 145, 514–522 (1988)
56. Rossi, F., Pascale, J.: Pseudopotential molecular-structure calculations for alkali-metal-
atom–H2 systems. Phys. Rev. A32, 2657–2669 (1985)
57. Sobel’man, I.I.: Introduction to the Theory of Atomic Spectra. Fizmatgiz, Moscow (1963)
58. Golubkov, G.V., Golubkov, M.G.: Potential energies of the Na**-He quasi-molecule. Khim.
Fiz. 29(3), 1–13 (2010)
59. Aymar, M., Camus, P.: Multichannel quantum-defect analysis of the bound even-parity
spectrum of neutrale barium. Phys. Rev. A28, 850–857 (1983)
60. Aymar, M.: Stabilisation of 5dnd autoionising states of barium by destructive interference
effects between ionising channels. J. Phys. B18, L768–L770 (1985)
61. Frederich, H., Wintgen, P.: Interfering resonances and bound states in the continuum. Phys.
Rev. A32, 3231–3242 (1985)
62. Connerade, J.P., Lane, A.M.: Vanishing points of particle widths of resonances: an atomic
parallel of a well known nuclear phenomenon. J. Phys. B18, L605–L610 (1985)
63. Golubkov, G.V., Ivanov, G.K.: Interaction of auto-decayed states in molecule photodissoci-
ation processes. IL Nuovo Cimento 12D(1), 1–20 (1990)
64. Balashov, E.M., Golubkov, G.V., Ivanov, G.K.: Radiative transitions between Rydberg
states of molecules. Sov. Phys. JETP. 59(6), 1188–1194 (1984)
65. Labastie, P., Bardass, M.S., Tribollet, B., et al.: Stroboscopic effect between electronic and
nuclear motion in highly excited molecular Rydberg states. Phys. Rev. Lett. 52, 1681–1684
(1984)
66 G.V. Golubkov et al.

66. Bodrov, A.E., Dalidchik, F.I.: Stroboscopic effect in the spectra of highly excited states of
Rydberg molecules. Khim. Fiz. 8(7), 876–882 (1989)
67. Golubkov, G.V., Ivanov, G.K., Balashov, E.M.: Manifestation of nonadiabatic electron-
rotational coupling in radiative transitions between Rydberg states of the H2 molecule.
Opt. Spectrosc. 80(1), 33–40 (1996)
68. Balashov, E.M., Golubkov, G.V., Ivanov, G.K.: Strong irregular excitation-energy depen-
dence of the inelastic transitions and collisional ionization cross-sections of highly excited
molecules. JETP 76(2), 200–209 (1993)
69. Ivanov, G.K., Golubkov, G.V., Balashov, E.M.: Role of non-adiabatic mixing of the vibronic
states of highly excited molecules with neutral particles. Dokl. Akad. Nauk, SSSR. 323,
311–315 (1992)
70. Golubkov, G.K., Ivanov, G.K.: Nonadiabatic effects in Rydberg molecular states and their
manifestation in collision processes. Chem. Phys. Rep. 19(3), 549–571 (2001)
71. Fano, U.: Quantum defect theory of l-uncoupling in H2 as an example of channel-interaction
treatment. Phys. Rev. A2(2), 353–365 (1970)
72. Demkov, Yu.N., Komarov, I.V.: Ionization under slow collisions of two atoms. Sov. Phys.
JETP 23, 189–196 (1966)
73. Presnyakov, L.P., Urnov, A.M.: Excitation of the multicharged ions by electron impact. Sov.
Phys. JETP 41, 31–35 (1975)
74. Golubkov, M.G., Golubkov, G.V., Ivanov, G.K.: Low-temperature dissociative recombina-
tion of electrons with H2+, HD+ and D2+ molecular ions. J. Phys. B30, 5511–5534 (1997)
75. Golubkov, G.V., Ivanov, G.K.: Ionization and vibronic transitions under slow collisions of
highly excited atoms with molecules. Sov. Phys. Tech. Phys. 35, 401–410 (1990)
76. Demkov, Yu.N., Rudakov, V.S.: Method of the partial waves for a nonspherical scatterer.
Sov. Phys. JETP 32(60), 1103–109 (1971)
77. Beigman, I.L., Lebedev, V.S.: Collision theory of Rydberg atoms with neutral and charged
particles. Phys. Rep. 250, 95–328 (1995)
78. O’Malley, T.F., Spruch, L., Rosenberg, L.: Low-energy scattering charged particle by
polarizable system. Phys. Rev. 125, 1300–1306 (1962)
79. O’Malley, T.F.: Exstrapolation of electron-rare gas atom cross-sections to zero energy. Phys.
Rev. 130, 1020–1025 (1963)
80. O’Malley, T.F.: Low-energy expansion of the scattering amplitude for long-range quadru-
pole potentials. Phys. Rev. 134, 1188–1197 (1963)
81. Demkov, Yu.N., Osherov, V.I.: Stationary and time-dependent quantum problems solved by
the contour integral method. Sov. Phys. JETP 26, 916–923 (1968)
82. Nikitin, E.E., Umanskii, S.Ya.: Theory of Slow Atomic Collisions. Springer Verlag, Berlin,
Heidelberg/New York/Tokyo (1984)
83. Demkov, Yu.N., Golubkov, G.V.: Adiabatic approximation in time-dependent problems of
quantum mechanics. Khim. Fiz. 22(10), 7–24 (2003)
84. Golubkov, G.V., Ivanov, G.K., Balashov, E.M., Golubkov, M.G.: l mixing and dissociation
of Rydberg molecules accompanying slow collisions with inert-gas atoms, JETP 87(1),
56–63 (1998)
85. Petitjean, L., Gounand, F.: Simple analytical formulas for collisional l-mixing, n-changing
and ionization of Rydberg atoms with neutral particles at thermal energy. Phys. Rev. A30,
2946–2957 (1984)
86. Radtsig, A.A., Smirnov, B.M.: Handbook of Atomic and Molecular Physics. Atomizdat,
Moscow (1980)
87. Chapelet, M.J., Boulmer, J., Gauthier, J.C., et al.: Collisional angular momentum mixing of
very high (#) Rydberg d-states of Na by Ar, Kr and Xe. J. Phys. B15, 3455–3467 (1982)
88. Kachru, R., Gallagher, T.F., Gounand, F., et al.: Collisional angular-momentum mixing of
Na Rydberg states with Xe. Phys. Rev. A27, 795–804 (1983)
89. Omont, A.: Theory of collisions of atoms with neutral particles. J. Phys. 38, 1343–1359
(1977)
1 Rydberg States of Atoms and Molecules in a Field of Neutral Particles 67

90. Ivanov, G.K.: Transfer of electron in atomic collisions. Opt. Spectrosc. 37, 636–642 (1974)
91. Golubkov, G.V., Dalidchik, F.I., Ivanov, G.K.: Vibrational excitation of molecules under
resonant scattering by electrons. Sov. Phys. JETP 46(2), 230–238 (1977)
92. Merkt, F., Fielding, H.H., Softley, T.P.: Electric field effects on zero-kinetic-energy photo-
electron spectra an explanation of observed trends. Chem. Phys. Lett. 202, 153–160 (1993)
93. Merkt, F., Schmutz, H.: Very high resolution spectroscopy of high Rydberg states. J. Chem.
Phys. 108, 10033–10039 (1998)
94. Borodin, V.M., Kazansky, A.K.: Mechanism of the oscillation origin widths for the alkali-
metal atom Rydberg states. JETP 70, 252–262 (1990)
95. Bethe, H.A., Salpeter, E.E.: Quantum Mechanics of One and Two-electron Atoms. Springer-
Verlag, Berlin (1957)
96. Smirnov, B.M.: Physics of Weakly Ionized Gases. Mir, Moscow (1985)
97. Duzy, C., Berry, R.S.: Autoionization of N2. J. Chem. Phys. 64, 2431–2436 (1976)
98. Cremaschi, P.A.: A theoretical study of the Rydberg states of N2 obtained from excitations of
valence 3sg 1pu, 2su levels. Chem. Phys. 109, 117–124 (1976)
99. Schulz, G.J.: Resonanses in electron impacts on atoms. Rev. Mod. Phys. 45, 378–422 (1973)
100. Golubkov, G.K., Ivanov, G.K., Balashov, E.M.: Interactions of the highly excited atoms and
molecules with neutral particles. Features of the potential energy surfaces and the charge
exchange processes. Chem. Phys. Rep. 14(8), 1104–1127 (1995)
101. Parlant, G., Fiquet-Fayard, F.: The O2-P2g resonance. J. Phys. B9, 1617–1628 (1976)
102. Herzberg, G., Jungen, Ch: Rydberg series and ionization potential of the H2 molecule. J. Mol.
Spectrosc. 41, 425–486 (1972)
103. Golubkov, G.V., Ivanov, G.K.: Rydberg states of molecules. JETP 53, 674–683 (1981)
104. Galitskii, V.M., Karnakov, B.M., Kogan, V.I.: Problems on the Quantum Mechanics. URSS,
Moscow (2001)
Chapter 2
Formation of Aerosols in the Atmosphere

A.A. Lushnikov, V.A. Zagaynov, and Yu.S. Lyubovtseva

Abstract Atmospheric aerosol is one of the most important factors affecting the
Earth’s climatic and weather conditions. The study of the mechanisms of formation
and evolution of atmospheric aerosols is of primary importance for predictions of
the climatic changes on our planet. We hope that this short overview of the modern
state of art in aerosol science will be of use to all those who are involved to the study
of atmospheric processes that form the Earth’s climate. We introduce the readers to
the basics of physical chemistry of aerosols. Special attention is given to the latest
achievements in the theory of particle formation and their subsequent growth.

Keywords Aerosols
Nucleation
Condensation
Growth

2.1 Introduction

Atmospheric aerosols and trace gases affect considerably the global characteristics
of the Earth’s atmosphere [1–7]. The point is that these components play a decisive
direct and indirect role in the energy balance of the atmosphere. In particular, the
aerosols define the cloudiness of the sky: cloud formation is impossible without
cloud condensation nuclei [8]. Submicron atmospheric aerosol particles play the
role of such nuclei [4,6,7].
On the other hand, the aerosols are of primary importance in the atmospheric
chemical processes. It is enough to recall the ozone holes, not mentioning other
atmospheric chemical cycles, where aerosols participate either as catalysts or as an

A.A. Lushnikov (*)

Karpov Institute of Physical Chemistry, Moscow, Russia

V.L. Bychkov et al. (eds.), The Atmosphere and Ionosphere, 69

Physics of Earth and Space Environments,
DOI 10.1007/978-90-481-3212-6_2, # Springer ScienceþBusiness Media B.V. 2010
70 A.A. Lushnikov et al.

active reactant. Therefore, it is of interest to know the mechanisms of particle

formation in the atmosphere [1,3,6,7].
Attempts to answer questions such as, where are the aerosol particles from? what
structure do they have? and what is their chemical composition? appeared already
more than one and half century ago (see [9–11] ). Up to now no definite answers
have been found. Aerosol particles are even less studied objects than the elementary
particles and quarks. The reasons for such a situation are clear. The small sizes of
aerosol particles do not allow them to be watched through optical microscopes.
Normally, the sizes of particles that are especially interesting for investigations
do not exceed 100 nm, i.e., these particles are much smaller than the wavelength
of visible light. Next, they are transparent (and thus invisible) in UV light and
x-rays. The atmospheric aerosol particles immediately evaporate under the elec-
tronic beam in electron microscopes. The mass spectrometry is also practically
powerless in the study of tiniest atmospheric aerosols, because the samples contain
very small quantities of the aerosol substance, and it is impossible to come to some
definite conclusions on the chemical composition of the particles [12]. Next, the
atmospheric aerosols do appear under well-defined conditions. The point is that
the atmosphere is a stochastic medium and it is very difficult (better to say,
impossible) to control all the parameters governing intra-atmospheric physico-
chemical processes.
The submicron atmospheric aerosols were the most popular object for the study
since 3 decades ago. The reason for the interest to this item is clear. These very
particles are optically active. They play the crucial role in cloud formation. On the
other hand, these particles have their predecessors. At present the standard aerosol
instruments permit to detect the particles of nanometer sizes [12–15]. It occurs,
however, that this limit is not enough for answering all questions concerning the
mechanisms of gas-to-particle conversion.
The goal of this short overview is to introduce the outside reader to the circle of
problems related to the particle formation and growth in the atmosphere.

2.2 Classification of Aerosols

The aerosols are divided into two classes: primary aerosols and secondary aerosols,
according to the mechanisms of their origination. The primary aerosol particles
result, for example, from fragmentation processes or combustion and appear in
the carrier gas as already well-shaped objects. Of course, their shape can change
because of a number of physicochemical processes like humidification, gas–particle
reactions, coagulation, etc. Secondary aerosol particles appear in the carrier gas
from “nothing” as a result of gas-to-particle conversion. For example, such aerosols
regularly form in the Earth’s atmosphere and play a key role in the number of global
processes like formation of clouds. They serve as the centers for heterogeneous
nucleation of water vapor [8]. No aerosols – no clouds, so one can imagine how our
planet would look without the secondary aerosol particles.
2 Formation of Aerosols in the Atmosphere 71

Primary and secondary aerosols are characterized with the size, shape, and
chemical contents of aerosol particles. As for the shape, one normally assumes
that the particles are spheres. Of course, this assumption is an idealization necessary
for the simplification of mathematical problems related to the behavior of aerosol
particles. There are many aerosols comprising the irregularly shaped particles. The
nonsphericity of particles creates a heap of problems. There exist also agglomerates
of particles which in some cases reveal the fractal properties.
There are a number of classifications of particles with respect to their sizes.
For example, if the particles are much smaller than the molecular mean free path,
they are referred to as fine particles. This size ranges from 1 to 10 nm at normal
conditions. But from the point of view of aerosol optics these particles are not
small if the light wave length is comparable with their size. This is the reason
why such very convenient and commonly accepted classifications cannot compete
with natural classifications based on the comparison of the particle size with
a characteristic size that comes up each time one solves a concrete physical
problem.
The particle size distributions play the central role in physics and chemistry of
aerosol, although a direct observation of the distributions is possible only in
principle. Practically what we really observe is just a response of an instrument to
a given particle size distribution,

Z1
PðxÞ ¼ Rðx; aÞ f ðaÞda: (2.1)
0

Here f(a) is the particle size distribution (normally a is the particle radius), P(x) is
the reading of the instrument measuring the property x of the aerosol, and R(x,a) is
referred to as the linear response function of the instrument. For example, P(x) can
be the optical signal from an aerosol particle in the sensitive volume of an optical
particle counter, the penetration of the aerosol through the diffusion battery (in this
case x is the length of the battery), or something else. The function f(a) is normal-
ized to unity,

Z1
f ðaÞda ¼ 1: (2.2)
0

Although the particle size distribution is an elusive characteristic of the aerosol,

it is convenient to introduce it because it unifies all properties of aerosols.
In many cases the distribution function can be found on solving the general
dynamic equation governing the time evolution of the particle size distribution can
be found theoretically [1], but the methods for analyzing this equation are not yet
reliable, not mentioning the information on the coefficients entering this equation.
This is the reason why the phenomenological distributions are so widespread.
72 A.A. Lushnikov et al.

There is a commonly accepted collection of particle size distributions. It

includes:
The lognormal distribution [1,3,16],
 
1 1 2 a
fL ðaÞ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi exp  2 ln : (2.3)
2 p as ln s 2ln s as

Here a is the particle radius. This distribution depends on two parameters: as and s
where as is the characteristic particle radius and s is the width of the distribution.
Equation (2.3) is the famous lognormal distribution. It is important to emphasize
that it is not derived from a theoretical consideration. Rather, it is introduced “by
hands” for different set of the parameters.
Another size distribution (the generalized gamma-distribution) is given by the
formula:
 k "  #
1 a a j
fG ðaÞ ¼ exp  : (2.4)
Gðk=jÞ a s as

Here GðxÞ is the Euler gamma-function. The distribution fG ðaÞ depends on three
parameters, a s , k, and j [1,3,17].

2.3 Nucleation Scenarios

Nucleation is responsible for production of the tiniest particles due to gas-to-particle

conversion [18–21]. Condensable vapors cannot exist in the vapor phase under cer-
tain conditions; normally when the vapor pressure exceeds the saturation value that,
in turn, depends on the temperature. Statistical mechanics predicts the phase tran-
sition in these cases, but it does not answer the question, how does the transition go
on. For example, the vapor can condense either on the walls of the vessel containing
the gas–vapor mixture, or on foreign aerosol particles suspended in the carrier gas, or
form the particles itself without any help from other external factors. The latter case
is referred to as spontaneous nucleation. There are some principal difficulties in the
theoretical description of the nucleation process. We will try to elucidate the nature of
these difficulties and outline the ways for avoiding them.
Although everybody saw how a kettle boils producing visible vapor from its
nose, the quantitative measurements on the nucleation are very far from being
simple. The point is that respective nucleation theories are applicable for steady-
state and very clean (no foreign condensation nuclei) conditions. To provide such
conditions is not an easy task.
Already for many decades there exists an irreconcilable conflict between the
theoretical predictions and the experimental data on the spontaneous nucleation of
vapors. The reason for this is quite clear: the problem itself is far from being simple
2 Formation of Aerosols in the Atmosphere 73

either theoretically or experimentally. This opinion is commonly recognized and

shared by most of the researchers investigating the nucleation of vapors (and not
only vapors). After the brain storm of 60th and 70th (everybody remember the hot
battles around the Lothe–Pound correction) the development of our knowledge on
the nature of this remarkable phenomenon returned to a quiet stream. Although the
necessity in a correct and reliable Nucleation Theory had been emphasized many
times by very diverse leaders of Physics of Aerosols and Physics of Clouds, nobody
yet died because of the absence of such a theory. Decades elapsed, but as it was
many years ago the new generations of scientists continue to investigate the proper-
ties and the content of critical embrii, investigate the details of the particle forma-
tion under different specific conditions, and seek for the interpretable nucleation
events in the atmosphere. Of course, the situation with the experiment and espe-
cially with theory principally changed. First of all, it became possible to model the
nucleation processes on computers that allowed for performing very complicated
and quite informative numerical experiments. And still the nature of the discrepan-
cies remains unclear.
The diversity of the results of nucleation measurements makes us to think that
there could exist a number of principally different scenarios of the nucleation
processes. This section introduces the readers to some of them.

2.3.1 Thermodynamically Controlled Nucleation

Statistical physics predicts that after a long time under nucleation conditions a one
droplet þ nucleating gas will occur in the thermodynamic equilibrium [22–24]. The
mass of the droplet is / V (here V is the total volume of the system), i.e., it is
infinitely large. The starting point for the thermodynamic consideration is the Gibbs
distribution
 
Hðp; qÞ
W / exp ; (2.5)
kT

where p, q are generalized moments and coordinates respectively, k is the Boltzmann

constant, and T is the temperature in K. The Hamiltonian H possesses the property
of the translation invariance, which means that all correlation functions depend on
the differences of respective spatial variables. The density is thus independent of
the coordinates. This fact is somewhat strange in view of existence of the macro-
scopic droplet. The latter should possess a boundary (interface) that should locate
somewhere. On the other hand, the distribution (2.5) predicts the spatially uniform
density, i.e., it does not see the interface. Next, the pair correlation function
should distinguish the molecules in the gas and liquid phases. No conflict, how-
ever, arises, for the distribution function does not specify the location of the
interface. The droplet is simply spread over the coordinate space. If we fix the
74 A.A. Lushnikov et al.

location of the interface, then the translation invariance is broken (this step is
equivalent to adding an external force field depending on coordinates). Pay attention,
that in doing so we introduce (by hands) a new information in the system thus
changing its entropy.
The first attempts to describe the nucleation process based on the thermody-
namic theory of fluctuations (see [24] and references therein). The idea was (and is)
quite clear and transparent: the molecules of a condensing vapor should occur in a
volume, where the intermolecular interaction is sufficiently strong to keep them
together. In order to find the probability to form such a cluster (critical or supercrit-
ical embryo), it is necessary to find the free energy of the cluster and exponentiate it.
The problem is just how to find this free energy. It is quite natural to assume that
this free energy is identical to the free energy of a liquid (solid) droplet of the
condensing substance. This assumption is not so bad for large droplets, but if the
embryo is small, the doubts in the validity of such approach can arise. Of course, it
is possible to try to find the free energy starting with the microprinciples, that is,
to calculate the free energy of the cluster by solving the respective classical or
quantum mechanical problem.
It is commonly accepted to express the nucleation rate in terms of the probability
for the critical embryo to exist and a kinetic factor describing the flux of nucleating
vapor toward the growing embryo,
 
G
J ¼ Zc exp  : (2.6)
kT

Here Zc is the flux of vapor.

The values of central interest are the formation energy G and the size of the
critical embryo. There exist numerous models of critical embryo beginning with the
macroscopic (droplet) model describing the energetics of the critical embryo in
terms of macroscopic characteristics (surface tension, bulk density) and ending with
entirely microscopic models that use the pair intermolecular potential for evaluat-
ing the partition functions of the embryo.
This scenario will be referred to as “thermodynamically controlled nucleation”,
for it assumes setting a thermodynamic equilibrium inside the embrii. Many
nucleation processes (but not all) are thermodynamically controlled. The flux of
nucleating vapor to the growing embryo is reported in refs [25–30].

2.3.2 Kinetically Controlled Nucleation

It is very difficult to imagine that a very slow thermodynamically controlled

nucleation scenario can realize in the atmosphere. More likely is another scenario
that assumes the formation of the critical embryo after one successful collision of
two vapor molecules. In this case a dimer forms. This dimer can grow and cast to a
2 Formation of Aerosols in the Atmosphere 75

trimer etc., once the latter is stable. The formation of the dimer requires the
presence of a third body that takes away the excess of energy appearing after the
formation of the bound state of two molecules. The nucleation rate in this case is
proportional to the second power of the vapor concentration. This scenario has been
recently investigated in detail in ref. [25]. It has been shown that only the bound
states of the dimer contribute to the growth process, which means that a thermody-
namic equilibrium is never attained in this case. The states in the continuous
spectrum have a short lifetime and thus should be ignored in the kinetic consider-
ation of the process.
The kinetic approach allows one to assume that the supercritical embryo can
form via the formation of mixed clusters comprising the molecules of the conden-
sing vapor and the molecules of the carrier gas. When such a cluster grows and
reaches a thermodynamically controllable size, the molecules of the carrier gas are
lost (they evaporate back from the droplet). The mixed states of the growing
embryo are apparently non-equilibrium and thus cannot be predicted within the
scopes of the thermodynamically controlled scenario. As far as we know, nobody
yet tried to consider such a type of nucleation. Meanwhile, the existing experiments
sometimes display the dependence of the nucleation rate on the pressure of the
carrier gas.
The kinetics of the particle formation–growth process is described by the
Szillard–Farcas scheme which assumes that the vapor molecules can join to (or
escape from) by one until the growing particle reaches the critical mass. This
kinetic scenario produces a well-known chain of equations for the concentrations
of growing particles. This set of equations can be solved in the steady-state limit
and gives the expression for the nucleation rate in terms of the evaporation and
condensation efficiencies. If one assumes the Principle of Detailed Balance to be
applicable under nucleation conditions, then the evaporation efficiencies could be
expressed in terms of the equilibrium concentrations and condensation efficiencies.
Then the Szillard–Farcas scheme leads to the well-known expression for the
nucleation rate. The kinetic approach denies this step. Now it has become evident
that the secondary atmospheric aerosols in most cases form according to the kinetic
scenario.

2.3.3 Fluctuation-Controlled Nucleation

Even very strongly supersaturated vapors consist of independent (noninteracting)

molecules (the interaction time is much shorter than the free flight time). The
nucleation rate, however, in such vapors is extremely high. Moderately supersatu-
rated vapor, in principle, always contains very highly supersaturated areas that are
formed due to the density fluctuations. The nucleation process within these areas
goes very quickly, so the rate of nucleation is limited by the rate of formation of the
fluctuation areas. Therefore the details of the nucleation process inside these highly
supersaturated areas are not important, it goes instantly. Here we return to the
76 A.A. Lushnikov et al.

ancient idea on the role of fluctuations in the nucleation process. But in contrast to
the classical thermodynamic approach we consider the fluctuation areas wherein the
molecules do not (yet) interact. The vapor density is sufficient for creating high
supersaturation, but it consists of noninteracting molecules. It is clear that such a
scenario can be described by the scheme

1 þ 1 þ 1 þ


þ 1 ! g

and the rate of the process is proportional to the vapor concentration to the power
g  This very type of nucleation has been introduced and investigated in ref. [26]. It
is important to emphasize that the formation of particles by nucleation alone cannot
be observed directly. The point is that the nucleation process is accompanied by
coagulation, and the latter process is also very swift. So just formed (by nucleation)
particles coagulate, and we observe only the final result of this process. This means
that the observed nucleation rate is much slower than that predicted by the nuclea-
tion theory alone.

2.3.4 Ion-Induced Nucleation

Although the role of ions in the formation of aerosols has been understood long
ago, the ion-induced nucleation in the atmosphere has become popular only now
after the papers [27–29]. Cosmic rays are known to produce ions that serve as
active centers on which the aerosol particles can form. These charged clusters
then evolve forming neutral and charged aerosol particles. The rate of formation
of such particles has been shown to be enough for accounting for the rate of
aerosol particle formation observed in rural regions. In this presentation we
discuss the role of an additional mechanism of the particle charging, the aerosol
photo effect. It is commonly believed that the thresholds of photo effect are too
high for the photoelectrons from the aerosol particles to play any role in the
evolution of the atmospheric aerosols. Indeed, normally the photo effect thresh-
olds correspond to the wave lengths of the order of 250 nm. The ozone layer cuts
off these photons in the lower atmosphere. But there exists another mechanism of
photoionization: (the two-quantum photoeffect) that which assumes simultaneous
absorption of two photons. The wavelengths of these photons correspond to green
light, where the Sun’s radiation spectrum is a maximum. Of course, the effici-
ency of the two-quantum photo effect is by / 100 times smaller than the single-
quantum photo effect, but still it is enough to produce the photoelectrons in
noticeable quantity. In order to investigate the role of the photo effect we apply
a simple kinetic model that describes the formation of aerosols and includes all
charging mechanisms. The photoeffect is shown to produce the overcharging of
the smallest particles, that is, the total charge of the smallest fraction is always
negative.
2 Formation of Aerosols in the Atmosphere 77

2.4 Kinetics of Nucleation

Everyone knows what is going on when a vapor is supersaturated. Very many tiny
particles are formed. These particles then grow and are lost somewhere, let us say,
because of sedimentation or some other reasons. It is clear that this picture does not
bear on any thermodynamics and requires a kinetic description. The whole process
can be described by the scheme

Independent vapor molecules ! collective of tiny droplets

The droplets are, however, not quite independent. They grow by consuming the
vapor molecules and thus interact via vapor (condensation).
The Szillard–Farcas scheme assumes that nucleation goes along the route [1,3],

ðgÞ þ 1 $ ðg þ 1Þ: (2.7)

The rates of the forward and backward processes are assumed to be known.
In the case of kinetically controlled nucleation the chain of reactions remains
unchanged. The reaction rates, however, are controlled by the kinetic processes in
the carrier gas. For example, the dimerization requires triple collisions, where one
of the participants is a carrier gas molecule. Of course, a dimer decay never occurs
without a collision with a third body.
In principle, the mixed embrii can arise. For instance, the dimer can form as
follows:

1 þ 10 ! ð1 þ 10 Þ; ð1 þ 10 Þ þ 1 ! ð2Þ þ 10 (2.8)

The idea is clear. Very tiny embrii may consist of a mixture of the vapor and the
carrier gas molecules. At later stages of the nucleation process the carrier gas
molecules escape the growing embrii.

2.4.1 The Szillard–Farcas Scheme

The Szillard–Farcas scheme assumes the rates of the elementary processes

g  1 $ ðg  1Þ to be known. Although this scheme is widely known [1–3,30],
we refresh some central moments of the derivation of the expression for the
nucleation rate. Respective kinetics obeys the following set of equations

Jg1 ðCÞ ¼ Jg ðCÞ (2.9)

with Jg being the total rates of jumps from the state g  1 to g

Jg ¼ J ¼ ag1 Ccg1  bg cg : (2.10)

78 A.A. Lushnikov et al.

The notation is standard and is explained in detail elsewhere. The nucleation rate
J is independent of g if the steady state is assumed.
Equation (2.10) is a set of linear algebraic equations with respect to population
concentrations cg . The boundary condition (particles disappear on reaching a large
size) allows one to resolve the set and to find

1
JðCÞ ¼ J2 : (2.11)
1 þ x2 þ x2 x3 þ

Here J 2 ðCÞ is the rate of dimerization.

1
J2 ðCÞ ¼ a1 C2 : (2.12)
2

The notation C stands for the monomer concentration. This value is specially
introduced instead of the supersaturation

bg
xg ¼ : (2.13)
ag1 C

The nucleation rate J(C) is thus expressed in terms of C, ag , and bg . The

concentration C can be measured (in principle). The situation with ag and bg is
much less clear.

2.4.2 Condensation and Evaporation

From the first sight the evaporation is a very simple process and is as follows:

ag ¼ vT sc (2.14)

with vT being the thermal velocity (depends on g via the particle mass) and sc is
the capture cross-section. The condensation is always an inelastic process. The
dynamics of the collision process is far from simple. For example, the dimerization
requires the presence of the third body (usually a molecule of the carrier gas). For
larger particles the energy excess can be transferred to the internal degree of
freedom.
The force fields like van der Waals forces or particle charge can affect the value
of the capture cross section. In the simplest case cross-section equals sc ¼ p a2 .
The situation with b is more complex [31]. A molecule of vapor should escape from
the particle. The picture looks as if the molecule diffuses inside a self-consistent
field created by other interacting molecules. Its energy fluctuates and it can escape
reaching the interface (if it has an energy excess enough for over-jumping the
2 Formation of Aerosols in the Atmosphere 79

potential barrier). In principle, it is possible to find the evaporation rate by consid-

ering this or that model of interacting molecules.
However, very few researchers risked to apply this straightforward approach.
Most of them prefer to apply the principle of detailed balance. The idea is an
attempt to express the evaporation rate in terms of the condensation rate. It is
assumed that the conditions exist when a droplet of mass g is in thermodynamic
equilibrium with the surrounding vapor–gas mixture. If yes, then

ag1 Cc0g1 ¼ bg c0g : (2.15)

This condition means that we can fit the temperature and the vapor concentration in
such a way that this condition fulfills. Actually this condition is just the application
of the method of variable constants. Indeed, such choice of c0g makes J ¼ 0. This
step is typical for the method of variable constants. The solution to Eq. (2.10) is
sought as a sum of the solution to the homogeneous equation (J ¼ 0) þ particular
solution of inhomogeneous equation (J > 0). On the other hand, Eq. (2.15) reminds
the condition of the equilibrium which permits one statistical mechanics for calcu-
lating the concentration c0g . But the equilibrium under the nucleation conditions is
impossible. Here we wish to emphasize that the introduction of c0g is just the formal
trick for solving the set of equations (2.10).

2.4.3 Nucleation via Dimers

It is possible to imagine a different (and even opposite) scheme: very few vapor
molecules are grouped into a stable cluster whose life time is not related to the
contact with the carrier gas. In order to understand what we want to say, let us
consider a “dimer-controlled nucleation”. The dimers (serving as critical embrii)
form by three molecular reactions. Once occurred in the bound state the dimer
cannot leave it without colliding with a carrier gas molecule. The difference
between such microscopic critical embryo and the commonly accepted critical
embrii is seen from the first sight: nothing happens inside the embryo giving rise
to its decay. Of course, the dimer is distributed over its energy states, but this
distribution is controlled by the kinetics of the three body collisions between vapor
molecules and a molecule of the carrier gas [25]. Not all dimers can serve as the
critical embrii.
The rate of kinetically controlled nucleation is expressed as follows:

J ¼ Acg:

Here again g* is the mass of the critical embryo. Very close to this scheme is “quasi-
chemical nucleation” where a couple of substances can form a stable molecule. This
type of nucleation is known and quite widespread. For example, the sulfuric oxides
80 A.A. Lushnikov et al.

can form stable compounds with water molecules. These compounds have a tendency
to agglomerate and form rather stable clusters of several molecules. These clusters
can serve as critical embrii for a nucleation process. In this case the binding energy of
such embrii exceeds the thermal energy. The excess of energy should be somehow
taken away. The radiation processes or further multiple collisions can serve to this
end. The kinetics of such a nucleation process can be very complicated and related to
numerous specific factors.

2.5 Growth by Condensation

The role of particle condensational growth and evaporation is extremely important

for the particle size spectra formation. These very mechanisms are responsible for
appearance of climatically active aerosols (of sizes 100–1,000 nm) in the atmo-
sphere. It is commonly accepted to consider three regimes of the particle growth:
the free-molecule regime, the continuous regime, and the transition regime. Each
regime is characterized by the Knudsen number (Kn) – by the ratio of the molecular
mean free path to the particle size. Large Knudsen’s number corresponds to the
free-molecule regime, that is, the molecules of the carrier gas do not prevent the
condensable molecules to approach the target particle. In this case the efficiency a
of the molecule trapping is limited by the total particle surface area. The exact result
looks as follows [1,3]:

a ¼ p a2 v T ; (2.16)

where a is the particle radius and vT is its thermal velocity. The condensational
efficiency is introduced as follows

j ¼ a ðn 1  n e Þ; (2.17)

where n 1 andne are the concentrations of the condensable vapor far away from the
particle and over its surface, respectively. The notation j stands for the vapor flux
toward the particle. Most frequently ne is the equilibrium vapor concentration over
the particle surface. If, however, the processes of evaporation or condensation are
very fast then this concentration is different from its equilibrium value and should
be found after solving the diffusion-reaction equation inside the particle This
situation is typical for the fast evaporation.
The expression for the condensational efficiency in the continuous regime is also
very simple [1,3,9]

a ¼ 4 p D a: (2.18)

Here D is the diffusivity of the condensing molecules. In this case the condensa-
tional efficiency is proportional to the particle size.
2 Formation of Aerosols in the Atmosphere 81

In order to find the condensational efficiency in the transition regime a more

sophisticated approach should apply. In this case the full solution of the Boltzmann
kinetic equation should be found. It is impossible to do analytically. Therefore,
already long ago there appeared the semiempirical versions of the expressions for
a [1,2,13] that reproduced the respective limiting expressions and described the
experimental data on the transition regime.
The first simplest theories applied the continuous models of condensation (the
particle radius a much exceeds the condensing molecule mean free path l). Such
models were not able to describe very small particles with sizes less than the l [9].
It was quite natural therefore to try to attack the problem starting with the free-
molecule limit, that is, to consider a collisionless motion of condensing molecules.
Respective expressions for the condensational efficiencies had been derived and
can be found in refs. [1–3]. The important step directed to the reconciliation of these
two limiting cases was done by Fuchs [9] who invented the flux-matching theory.
The flux-matching theories are well adapted for studying the behavior of aerosol
particles in the transition regime. Although these theories mostly did not have a firm
theoretical basis, they successfully served for systematizing numerous experiments
on the growth of aerosol particles and until now these theories remain rather
effective and are very practical tools for studying the kinetics of aerosol particles
in the transition regime (see [9,32]). On the other hand, these theories are always
semi-empirical, that is, they contain a parameter that should be taken from some-
where else, not from the theory itself.
We introduce the reader to the ideology of the flux-matching theories by
considering the condensation of a nonvolatile vapor onto the surface of an aerosol
particle. The central idea of the flux-matching procedure is a hybridization of the
diffusion and the free-molecule approaches. The concentration profile of a con-
densing vapor far away from the particle is described by the diffusion equation.
This profile coincides with the real one down to the distances of the order of the
vapor molecule mean free path. A limiting sphere is then introduced inside with
which the free-molecule kinetics governs the vapor transport. The concentration
profile in the free-molecule zone is considered to be flat. The equality of the fluxes
in both the zones and the continuity of the concentration profile at the surface of the
limiting sphere define the flux and the reactant concentration at the particle surface.
The third parameter, the radius of the limiting sphere, cannot be found from such a
consideration.
Hence, we apply a more sophisticated scheme [33]. We also introduce a limiting
sphere outside of which the density profile of condensing vapor can be described by
the diffusion equation. But inside the limiting sphere we solve the collisionless
Boltzmann equation subject to a given boundary condition at the particle surface
and put an additional condition: the vapor concentration at the surface of the
limiting sphere coincides with that found from the solution of the diffusion equa-
tion. Even in the absence of any potential created by the particle the vapor profile in
the free-molecule zone depends on the radial coordinate, because the particle
surface adsorbs all incoming molecules. We thus gain the possibility to call for
the continuity of the first derivatives of the profile on both sides of the limiting
82 A.A. Lushnikov et al.

sphere. This additional condition defines the radius of the limiting sphere [13]. The
respective expression for the condensational efficiency has a serious advantage over
the formulas used before. The point is that the final result is expressed in terms of
the diffusivity and the thermal velocity of the condensable molecules, rather than in
terms of the mean free path. The final result is,

2 p a2 v T
a¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi : (2.19)
1 þ 1 þ ð a2vDT Þ2

The extension of this formula to the case of an arbitrary external potential is,

a ða; RÞ e UðRÞ=kT
a ðaÞ ¼ R1 : (2.20)
1 þ a ða; RÞeUðRÞ=kT e UðrÞ=kT 4pdrr2
R

This formula widely applies for calculation of the charging efficiency.

Charged particles play a notable role in the processes of tiny particles formation.
The nucleation process in the presence of ions goes much easier. The particle
growth also becomes faster because bipolar charges on particles accelerate the
coagulation. The study of atmospheric charged particles began a century ago
[35–48]. First simplest theories were grounded on the use of the diffusion equation
and applied to the particles of large sizes, much exceeding the ion mean free path in
air. Such models were not able to describe the charging of smaller particles, with
the sizes of order or less than the ion mean free path. Latest attempts to attack the
particle charging problem applied the free-molecule approximation, that is, the
collisionless motion of ion was considered as the charge transport mechanism.
Respective expressions for the charging efficiencies were reported in refs. [33,34].
However, as was shown in ref. [34] the free-molecule regime does not realize even
at normal conditions and arbitrary small particles. The point is that in the case of
charged particles another characteristic length comes up. It is the Coulomb length
l c ¼ q Q e2 =kT. Even at the particle charges equal to unity this length is equal to 60
nm, i.e., it is comparable with the mean free path of ions. This fact means that the
ion–molecular collisions are of importance at the distances where the Coulomb
energy of the charge particle is comparable with the thermal energy kT. Hence, the
criterion of validity of the free-molecule regime l >> lc fulfils only at very low air
pressure, i.e., in upper layers of the atmosphere. Next, the bound states of ions in
the Coulomb or the image potentials are also of importance for bipolar charging
processes. The point is that an ion can occupy a bound state only on colliding with
the carrier gas molecule. The diffusion model assumes that all bound states are
occupied. In the free-molecule model all these states are empty. On the other hand,
it is clear that both the models have nothing common with the reality. The ions
2 Formation of Aerosols in the Atmosphere 83

cannot lose a very large energy and occur very deep in the potential well. This fact
creates huge difficulties because it is necessary to solve the collision task in an
external field.

2.6 Uptake of Trace Gases by the Aerosol Particles

Trace gases are commonly recognized to react actively with the aerosol component
of the Earth’s atmosphere [3,49–59]. Substantial changes to the atmospheric chem-
ical cycles due to the presence of aerosol particles in the atmosphere make us look
more attentively at the nature of the processes stipulated by the activity of atmo-
spheric aerosols (see, e.g., [3]). The process of the gas–particle interaction is
usually a first-order chemical reaction going along the route:

X þ AP ! ðAPXÞ; (2.21)

where X, AP, and (APX) stand respectively for a reactant molecule, an aerosol
particle and the final product resulting from the reaction Eq. (2.21).
As an example we refer to ozone, which is a key substance for our atmospheric
system since it protects the living systems on our planet against the Sun’s UV
radiation. However, since the discovery of the ozone hole in the mid-1970s (details,
last achievements, and references see in [3] ) , it has been well established that
ozone is subject to periodical large depletion events at the poles and to continuous
decay in the global stratosphere. These trends are initiated by the presence of
halogenated radicals that are produced by a cycle of photochemical processes.
The amplitudes of ozone level variations are also driven by heterochemical reac-
tions occurring on the surfaces of polar stratosphere clouds which transform stable
reservoir molecules into radical precursors.
Yet some important aspects of aerosol heterochemistry are not so well studied
(see, however, [60–64]). Among them is the interconnection between uptake and
mass accommodation efficiencies [61]. Still there exist discrepancies between
the results because of different understanding of the meaning of uptake coefficient.
The whole issue of Journal of Aerosol Science 32(7) (2001) is entirely devoted
to the problems of gas–aerosol interaction in the atmosphere.
In the following paragraphs we wish to outline our point of view. Let a particle
of the radius a initially containing NB molecules of a substance B be embedded
to the atmosphere containing a reactant A. The reactant A is assumed to be able to
dissolve in the host particle material and to react with B. The particle will begin
to absorb A and will do this until the pressure of A over the particle surface will
be enough for blocking the diffusion process. Our task is to find the consumption
rate of the reactant A as a function of time. Next, we focus on sufficiently small
particles whose size is comparable to or less than the mean free path of the reactant
84 A.A. Lushnikov et al.

molecules in the carrier gas. The mass transfer of such particles is known to depend
strongly on the dynamics of the interaction between incident molecules and the
particle surface. In particular, the value of the probability b for a molecule to stick
to the particle surface is suspected to strongly affect the uptake kinetics. The
question “how?” has not yet found a full resolution.
Below we are trying to answer this question starting with a simple analysis of
the boundary condition to the kinetic equation for the molecules of a reactant A.
Let an aerosol particle be put in the atmosphere containing a reacting gas
admixture A. The molecules of A are assumed to react with a guest reactant B
dissolved in a host material of the particle (in principle, B. itself can be the host
material). The reactant A is assumed to react with the reactant A along the route:

AþB!C (2.22)

The particle initially containing no molecules of A begins to consume those

crossing the particle–carrier gas interface. Our goal is to investigate the kinetics of
this process.
Our basic integral principle asserts:

Flux of A from outside ¼ total consumption of A inside (2.23)

We consider four stages of the uptake process:

1. The diffusion of A toward the particle
2. Crossing the particle–air interface
3. Diffusion-reaction process inside the particle
4. Accumulation of nonreacted A-molecules in the particle
Balancing the fluxes gives the equation for the uptake rate.
In principle, the consideration of the uptake requires a solution of the time-
dependent transport problem. Here we give some order-of-magnitude estimations
allowing for a correct statement of the problem in realistic conditions of the Earth’s
atmosphere. Our idea is to get rid of the non-stationarity wherever it is possible.
Below we use the notation DX ðX ¼ A, B Þ; D g A for the diffusivity of the reactant
molecules inside the particle and in the gas phase respectively.
The characteristic time of the non-stationarity in the gas phase is estimated as
t / a2 =D: This time is extremely short. For D ¼ 0:1 cm2 /s and a ¼ 1,000 nm
t / 10 7 s. So the transport in the gas phase can be considered in the steady-
state limit. The diffusion process in the liquid phase is much slower. Its character-
istic time is t / 103 s for micron particles and DX  105 cm2 /s.
The time for the chemical reaction of A molecules in the liquid phase is
estimated as tA / k nB where k is the binary reaction rate constant for the reaction
given by Eq. (2.22). The maximal value of k / DX am / 1013 cm2/s for the mole-
cule radius am / 108 cm. The book [3] cites the values within the interval
k ¼ 1011  1018 cm3 /s.
2 Formation of Aerosols in the Atmosphere 85

The estimate of the characteristic of time tA depends on the value of nB . If the

gaseous reactant reacts with the host material, the characteristic reaction times are
very short t / 10 9 s for diffusion controlled reactions and much longer for other
types of chemical processes (up to seconds or even minutes).
The characteristic transport times should be compared to the characteristic times
of substantial chemical changes inside the particle that are stipulated by the flux
of A from outside. These times are of the order of tchanges / 1=jA , where jA is the
total flux of A-molecules trapped by the particle. This is the characteristic time for
one molecule of A to attach to the particle surface. Actually this time depends on
the sticking probability b and can reach 10 s.
These estimates show (see also [3]) that all the characteristic times for the tran-
sient processes inside and outside the particle are much shorter than the charac-
teristic time for the particle to change its chemical composition due to uptake. This
means that a quasi-steady-state approximation can be used for the description of
very slow changing parameters such as the total number of molecules inside the
particle or its size. Fast transport processes establish instantly the steady-state
concentration profiles.
Let us first consider condensation of A-molecules onto a spherical particle of
B-liquid. In the gas phase the distribution function fA of A-molecules over coordi-
nates and velocities satisfies the Maxwell boundary condition [11],

b
fA  ¼ ð1  bÞ fA þ þ nA e ; (2.24)
2p
where b is the sticking probability, fA  is the velocity distribution function of
molecules flying outward the particle, fA þ is the same for molecules flying towards
the particle surface, and nAe is the equilibrium concentration of A-molecules over
the particle surface. The first term on the right of this equation describes the mirror
rebound of A-molecules from the particle surface. The second term gives the
density of the accommodated and then emitted from the surface molecules of A.
The coefficient 1=2 p reflects the fact that the molecules fly only in the outward
direction. At f þ ¼ f  ¼ n e =2 p (full thermodynamic equilibrium) Eq. (2.24)
satisfies automatically. The total flux of A is expressed as

jA ¼ a ðaÞ ðnA 1  nA e Þ: (2.25)

Here nA 1 is the concentration of A far away from the particle.

The solution of the kinetic equation defines the concrete form of the dependence
of the condensational efficiency a ðaÞ on the particle radius and b. The form of the
second multiplier is universal and in neither way is related to any approximation.
Now we are returning to uptake. In this case the inward flux inside the particles
makes concentration nþ A of outflying particles lower than nA e . The boundary
condition Eq. (2.24) is replaced by

b
fA  ¼ ð1  bÞ fA þ þ n Aþ (2.26)
2p
86 A.A. Lushnikov et al.

with the value of the reactant concentration n A þ being determined from balancing
the fluxes. This is the principal point of our further consideration. This equation
differs from Eq. (2.24) by replacement nA e ! n A þ . Hence, the flux of nA e toward
the particle looks as follows:
 
n Aþ
jA ¼ aðaÞ nA e  : (2.27)
b

Instead of solving the kinetic equation, we apply any semi-empirical formula for
aðaÞ. For example, the Fuchs–Sutugin formula

af m ðaÞ
a ðaÞ ¼ ; (2.28)
1 þ b SðKnÞ

where
a f m ðaÞb ¼ p vT a2 (2.29)

is the condensational efficiency in the free-molecule regime and

3 0:466
SðxÞ ¼  ; (2.30)
4x x þ 1

n A ¼ H nA þ ; (2.31)

where H is the dimensionless Henri constant defined as

Hs RT
H¼ (2.32)
0:981  109

Here Hs is the Henri constant (in units mol/a, see the definition in [3]). At
T ¼ 300 K
H ¼ 26:4Hs : (2.33)

We demonstrate the application of the above scheme for the case of complete
mixing the reactants inside the particle, that is, we assume that the reactant
concentration profiles are independent of coordinates. Then the following set of
equations describes the evolution of the particle chemical composition,

dNA NA k
¼ aðaÞ ðnA 1  Þ  NA NB ; (2.34)
dt HV V
dNB k
¼  NA NB ; (2.35)
dt V

VðtÞ ¼ va NA þ vb NB þ vc NC ; (2.36)

where vx is the molecular volume of each component ðx ¼ a; b; cÞ:

2 Formation of Aerosols in the Atmosphere 87

2.7 Nucleation Bursts in the Atmosphere

Regular production of nonvolatile species of anthropogenic or natural origin in the

atmosphere eventually leads to their nucleation, formation of tiny aerosol particles,
and their subsequent growth. Thus formed aerosol is able to inhibit the nucleation
process because of condensation of nonvolatile substances onto the surfaces of
newly born particle surfaces. This process is referred to as the nucleation burst [1].
The dynamics of atmospheric nucleation bursts possesses its own specifics, in
particular, the particle production and growth is suppressed mainly by preexisting
aerosols rather than freshly formed particles of nucleation mode [12]. In many
cases the nucleation bursts have heterogeneous nature. The smallest (undetect-
able) particles accumulated during nighttime begin to grow at daytime because of
sunlight-driven photochemical cycles producing low volatile (but not nucleating)
substances that are able to activate the aerosol particles [65]. Stable sulfate
clusters [66–72] can serve as heterogeneous embryos provoking the nucleation
bursts.
The nucleation bursts were regularly observed in the atmospheric conditions and
were shown to serve as an essential source of cloud condensation nuclei [12,73–77].
Now it becomes more and more evident that the nucleation bursts in the
atmosphere can contribute substantially to the cloud condensation nuclei produc-
tion and can thus affect the climate and weather conditions on our planet (see e.g.,
[3,6,7] and references therein). Existing at present time opinion connects the
nucleation bursts with additional production of nonvolatile substances that can
then nucleate producing new aerosol particles, and/or condense onto the surfaces
of newly born particles, foreign aerosols, or on atmospheric ions. The production of
nonvolatile substances, in turn, demands some special conditions to be fulfilled
imposed on the emission rates of volatile organics from vegetation, current chemi-
cal content of the atmosphere, rates of stirring, and exchange processes between
lower and upper atmospheric layers, presence of foreign aerosols (accumulation
mode, first of all) serving as condensational sinks for trace gases and the coagula-
tion sinks for the particles of the nucleation mode, the interactions with air masses
from contaminated or clean regions [12,68,77]. Such a plethora of very diverse
factors most of which have a stochastic nature prevents direct attacks on this effect.
A huge amount of field measurements of nucleation bursts dynamics appeared
during the last decade (see [12,68,77] and extensive citation therein).
The attempts of modeling this important and still enigmatic process also appeared
rather long ago. All models (with no exception) start from a commonly accepted
point of view that the chemical reactions of trace gases are responsible for the
formation of nonvolatile precursors which then give life to subnano- and nano-
particles in the atmosphere. In turn, these particles are considered as active partici-
pants of the atmospheric chemical cycles leading to the particle formation. Hence,
any model of nucleation bursts included (and includes) coupled chemical and aerosol
blocks. This coupling leads to strong nonlinearities which means that all intra-
atmospheric chemical processes (not all of which are, in addition, firmly established)
88 A.A. Lushnikov et al.

are described by a set of nonlinear equations. In addition, there is no assurance that

we know all the participants of the chemical cycles leading to the production of low
volatile gas constituents that then convert to the tiniest aerosol particles.
Our main idea is to decouple the aerosol and chemical parts of the particle
formation process and to consider here only the aerosol part of the problem. We thus
introduce the concentrations of nonvolatile substances responsible for the particle
growth and the rate of embryo production as external parameters whose values can be
found either from measurements or calculated independently, once the input concen-
trations of reactants and the pathways leading to the formation of these nonvolatile
substances are known. Next, introducing the embryo production rate allows us to
avoid a rather slippery problem of the mechanisms responsible for embryos forma-
tion. Since neither the pathways nor the mechanisms of production of condensable
trace gases and the embryos of condense phase are well established so far, our semi-
empirical approach is well approved. Moreover, if we risk to begin from the basic
principles, we need to introduce too many empirical (fitting) parameters.
Aerosol particles throughout the entire size range beginning with the smallest
ones (with the sizes of order 1 nm in diameter) and ending with sufficiently large
particles (submicron and micron ones) are shaped by some well-established
mechanisms.
They are condensation and coagulation. Little is known, however, on atmo-
spheric nucleation. This is the reason why this very important process together
with self-coagulation is introduced here as an external source of the particles of
the smallest sizes. The final productivity of the source is introduced as a fitting
parameter whose value is controlled by these two processes simultaneously and is
thus always lower than the productivity of the nucleation mechanism alone. Next,
coagulation produces the particles distributed over a size interval, rather than
monodisperse ones of a critical size (like in the case of nucleation alone). The
productivity should be introduced as a function of the particle size and time,
respectively. In principle, the size dependence of the source can be found theoreti-
cally, but it is better to refuse this idea and to introduce it as the product of a
lognormal function and a time-dependent total production rate.
The condensational growth depends on the concentrations of condensable
vapors, with the condensational efficiencies being known functions of the particle
size. The concentrations of condensable trace gases are introduced as known
functions. They can also be calculated, once all reactions responsible for conversion
of volatile trace gases to low volatile ones and respective reaction rates are known
(þ stoichiometry of the reactions þ initial concentrations of all participants and
many other unpleasant things). Of course, nothing like this is known and there is no
chance to get this information in the near future.
The losses of particles are caused mainly by preexisting submicron and micron
particles. There are also other types of losses: deposition of particles onto leaves
of trees, soil losses, scavenging by deposits, and mists. Here the term ‘loss’ is
introduced as a sink of small particles on preexisting submicron and micron aerosol
particles.
2 Formation of Aerosols in the Atmosphere 89

Self-coagulation of particles with sizes exceeding 3 nm in diameter is entirely

ignored in the model. Many authors estimated the characteristic times of the self-
coagulation process and found them to exceed 104 s. In what follows we ignore this
process. On the contrary, the intermode coagulation (the deposition of newly born
particles onto preexisting aerosols) is of great importance and must be taken into
account.
Now it is easy to answer the question why the linear model is able to describe
the nucleation bursts in the atmosphere. The nucleation mode does not affect the
surrounding atmosphere whose chemical state is defined by other numerous exter-
nal factors. For example, the lifetimes of trace gases and the particles of nucleation
mode depend on the concentration and the size distribution of preexisting aerosol
particles.
The mechanisms of particle growth (condensation and coagulation) are well
established for the whole size interval of interest. This is the reason why the model
for the description of atmospheric particles is based on the continuity equation

@n @n
þ a_ þ ln ¼ J: (2.37)
@t @a

Here n ¼ n ða; tÞ is the particle size distribution, a_ is the rate of particle growth

vT V0 CðtÞ
a_ ¼ a CðtÞ ¼ ; (2.38)
4

where vT is the thermal velocity of condensing molecules, V0 is the volume of the

single molecule, and C ¼ CðtÞ is the number concentration of condensing mole-
cules in air. The value l is the coagulation sink (a fitting parameter), J ¼ Jða; tÞ ¼
jðtÞf ðaÞ is the productivity of the stable embrii source. The function f ðaÞ describes
the size distribution of the newly born particles.
At this stage it is especially important to emphasize that the coagulation sink
is created mainly by foreign aerosols with the sizes of order 1,000 nm. Then the con-
tinuity equation (2.20) becomes linear. We removed all the sources of nonlinearity
which arises from the nucleation term and the coagulation sink (the concentration
of condensing gases are considered to be known functions of time). Next, we ignore
the intra-mode coagulation. Equation (2.20) is the first-order partial differential
equation, whose solution can be presented in the form:

nða; tÞ ¼ n0 ða; tÞ þ nJ ða; tÞ; (2.39)

where n0 ða; tÞ is the general solution of the homogeneous equation and nJ ða; tÞ is a
particular solution of the inhomogeneous equation. The full spectrum in the form of
Eq. (2.39) allows one to classify the scenarios of nucleation bursts.
Scenario 1. At the initial moment of time (night time) the nucleation mode is
entirely absent. At sunrise the trace gases appear that are able to nucleate. They
90 A.A. Lushnikov et al.

0.25

GR = 1.5 nm/h Coag = 10–4s–1

squares 9h
circles 12h
0.20 crosses 15h
diamonds 18h
size spectrum (r.u.)

0.15

0.10

0.05

0.00
0 5 10 15 20
particle size (nm)

Fig. 2.1 The evolution of size distributions typical for summer periods is shown. Source
produces the particles of the nucleation mode. The sinks are seen not to prevent the particle
growth up to sizes 10 nm. This situation is typical for boreal regions of Finland and Siberia in
summer period

form the particles growing further by condensing low volatile trace gases that
cannot form the aerosol particles themselves. An example of such situation is
shown in Fig. 2.1. It is seen that if the coagulation sinks are sufficiently large the
particles larger 3 nm do not appear at all (see Fig. 2.2). It is important to stress that
there is no well-distinguishable peak in the particle size distribution.
Scenario 2. Nucleation does not occur at all, but the smallest particles formed
during nighttime are still present in the atmosphere. Then these particles begin to
grow as soon as condensable gases appear after the sunrise. These particles can pass
the threshold 3 nm in diameter and become detectable. In this case the peak in the
particle distribution is well expressed. The picture of the nucleation burst is shown
in Fig. 2.3.
Scenario 3. The running mode can appear after a new, more clean air mass
comes to the observation point.
2 Formation of Aerosols in the Atmosphere 91

7
10–3
GR = 1.5 nm/h Coag = 10–2s–1
6

size spectrum (r.u.) (9h)

5
(12h)
(15h)
4 (18h)

0
0 5 10
particle size (nm)

Fig. 2.2 In winter time the rate of particle formation is low. The sinks suppress the particle growth
and they do not reach up to detectable sizes

GR = 1nm/h Coag = 10–4s–1

squares 9h
circles 12h
crosses 15h
0.04
size spectrum (r.u.)

0.02

Fig. 2.3 If the particle

formation precedes the
condensational growth (e.g.,
small particles were born at
nighttime due to dark
reactions), then a well 0.00
0 5 10 15 20
distinguishable peak appears
at daytime particle size (nm)
92 A.A. Lushnikov et al.

2.8 Concluding Remarks

We have presented a short overview of existing data on the processes of aerosol

formation in the atmosphere. Our conclusions are as follows:
1. Although there is not full clarity in theory of gas-to-particle conversion, the
productivity of the nucleation process can be introduced as a fitting parameter
to the equation describing the atmospheric aerosol particle formation and
growth.
2. The mechanism of the condensational growth is clear and can be described
theoretically. Very often we do not know the participants of this process
therefore their concentrations should be introduced as parameters or calculated
from the respective chemical block.
3. Extremely important achievement is the theoretical description of the conden-
sation growth of mixed particles. Now we know how to include the accommo-
dation coefficient to the expression for the condensational efficiency of mixed
particles.
4. At sufficiently short times (of order of 24 h) the intra-mode coagulation pro-
cesses do not play an essential role. The inter-mode coagulation can be included
into the coagulation sinks and be considered as a fitting parameter of the
formation-growth models.

References

1. Friedlander, S.K.: Smokes, Haze, Mist. Wiley, New York/London (2000)

2. Hidy, J.M., Brock, J.R.: The dynamics of aerocolloidal systems. Pergamon, Oxford (1970)
3. Seinfeld, J.H., Pandis, S.N.: Atmospheric Chemistry and Physics. Wiley, New York (1998)
4. Taylor, F.W.: The greenhouse effect and climate change revisited. Rep. Prog. Phys. 65,
1 (2002)
5. Graedel, T.E., Crutzen, P.J.: The changing atmosphere. Sci. Am. 261, 58 (1989)
6. Charlson, R.J., Schwartz, S.E., Hales, J.M., Cess, R.D., Coslkey Jr., J.A., Hansen, J.E.,
Hofmann, D.J.: Climate forcing by anthropogenic aerosols. Science 255, 423 (1992)
7. Charlson, R.J., Heitzenberg, R.L.: Aerosol forcing of climate. Wiley, New York (1995)
8. Pruppacher, H.R., Klett, J.D.: Microphysics of clouds and precipitation. Reidel, Dordrecht,
The Netherlands (1991)
9. Fuchs, N.A.: Mechanics of aerosols. Pergamon, New York (2002)
10. Reist, P.C.: Introduction to aerosol science. Macmillan, New York (1984)
11. Williams, M.M.R., Loyalka, S.K.: Aerosol science, theory and practice. Pergamon, Oxford
(1991)
12. Kulmala, M., Vehkm€aki, H., Pet€aj€a, T., Dal Maso, M., Lauri, A., Kerminen, V.-M., Birmili,
W., McMurry, P.H.: Formation and growth rates of ultrafine atmospheric particles: a review of
observations. J. Aerosol Sci. 35, 143–176 (2004)
13. Alpin, K.L., Harrison, R.G.: A computer controlled Gerdien atmospheric ion counter. Rev.
Sci. Instrum. 71, 3037 (2000)
14. Eisele, F.L., Hanson, D.R.: First measurements of prenucleation molecular clusters. J. Phys.
Chem. 104, 830–836 (2000)
2 Formation of Aerosols in the Atmosphere 93

15. Gamera-Castano, M., de la Mora, F.: A condensation nucleus counter (CNC) sensitive to
singly charged sub-nanometer particles. J. Aerosol Sci 31, 757 (2000)
16. Aitchison, J., Brown, J.A.: The lognormal distribution function. Cambridge University Press,
Cambridge (1957)
17. Deirmenjian, D.: Electromagnetic scattering on spherical polidispersions. Elsevier, New York
(1969)
18. Abraham, F.F.: Homogeneous nucleation theory. Academic, New York (1974)
19. Kaschiev, D.: Nucleation: Basic principles and application. Butterworth, Heinemann (2000)
20. Laaksonen, A., Talanquer, V., Oxtoby, D.W.: Nucleation: measurements, theory, and atmo-
spheric applications. Annu. Rev. Phys. Chem. 46, 189 (1995)
21. Anisimov, M.A.: Nucleation: theory and experiment. Russ. Chem. Rev. 72, 591 (2003)
22. Hill, T.: Statistical mechanics. Principles and selected applications. McGraw-Hill, New York
(1956)
23. Gross, D.H.E.: Microcanonical thermodynamics. World Scientific, Singapore (2001)
24. Landau, L.D., Lifshits, E.M.: Statistical physics. Nauka, Moscow (1998)
25. Lushnikov, A.A., Kulmala, M.: Dimer in nucleating vapor. Phys. Rev. E58, 3157 (1998)
26. Lushnikov, A.A., Kulmala, M.: Nucleation controlled formation and growth of disperse
particles. Phys. Rev. Lett. 81, 5165 (1998)
27. Castleman Jr., A.W., Holland, P.M., Keesee, R.G.: The properties of ion clusters and their
relationship to heteromolecular nucleation. J. Chem. Phys. 68, 1760 (1978)
28. Yu, F., Turco, R.P.: From molecular clusters to nanoparticles: role of ambient ionization in
tropospheric aerosol formation. J. Geophys. Res. 106, 4797–4814 (2001)
29. Svensmark, H.: Influence of cosmic rays on Earth’s climate. Phys. Rev. Lett. 81, 5027 (1998)
30. Reiss, H., Tabahzade, A., Talbot, J.: Molecular theory of vapor phase nucleation: the physi-
cally consistent cluster. J. Chem. Phys. 92, 1266 (1990)
31. Ford, I.J., Harris, S.A.: Molecular cluster decay viewed as escape from a potential of mean
force. J. Chem. Phys. 120, 4428 (2004)
32. Fuchs, N.A., Sutugin, A.G.: High–dispersed aerosols. In: Hidy, G.M., Brock, J.R. (eds.)
Topics in current aerosol research, vol. 2, pp. 1–60, Pergamon, Oxford (1971)
33. Lushnikov, A.A., Kulmala, M.: Flux–matching theory of particle charging. Phys. Rev. E70,
046413 (2004)
34. Lushnikov, A.A., Kulmala, M.: Charging of aerosol particles in the near free-molecule
regime. Eur. Phys. J. D29, 345 (2004)
35. Arnold, F.: Multi-ion complexes in the stratosphere – implication for trace gases and aerosols.
Nature 284, 610 (1980)
36. Arnold, F., Stilp, T., Busen, R., Shauman, U.: Jet engine exhaust chemi-ion measurements:
implication for gaseous SO2 and H2SO4. Atmos. Environ. 32, 3073 (1998)
37. Castleman Jr., A.V.: Nucleation and molecular clustering about ions. Adv. Colloid Interface
Sci. 10, 73 (1979)
38. Horrak, U., Salm, J., Tammet, H.: Burst of intermediate ions in atmospheric air. J. Geophys.
Res. 103, 13909 (1998)
39. Karher, B., Yu, F., Schroeder, F.P., Turco, R.P.: Ultrafine aerosol particles in aircraft plumes:
analysis of growth mechanisms. Geophys. Res. Lett. 25, 2793 (1998)
40. Raes, F., Augustin, J., Vandingenen, R.: The role of ion induced aerosol formation in the lower
atmosphere. J. Aerosol Sci. 17, 466 (1986)
41. Turco, R.P., Zhao, J.-X., Yu, F.: A new source of tropospheric aerosols: ion–ion recombina-
tion. Geophys. Res. Lett. 25, 635 (1998)
42. Yu, F., Turco, R.P.: The role of ions in the formation and evolution of particles in aircraft
plumes. Geophys. Res. Lett. 24, 1927 (1997)
43. Yu, F., Turco, R.P.: Ultrafine aerosol formation via ion-mediated nucleation. Geophys. Res.
Lett. 27, 883 (2000)
44. Yu, F., Turco, R.P.: On the contribution of lightning to ultrafine aerosol formation. Geophys.
Res. Lett. 27, 1453 (2000)
94 A.A. Lushnikov et al.

45. Lushnikov, A.A., Kulmala, M.: A kinetic theory of particle charging in the free-molecule
regime. J. Aerosol Sci. 39, 1069 (2005)
46. Marlow, W.H.: Derivation of the aerosol collision rates for singular attractive contact poten-
tials. J. Chem. Phys. 73, 6284 (1980)
47. Pui, D.Y.H., Fruin, S., McMurry, P.H.: Unipolar charging of fine particles. Aerosol Sci
Technol. 8, 173 (1988)
48. Hoppel, W.A. Frick, G.M.: Ion-aerosol attachment coefficients and the steady-state charge
distribution on aerosols in a bipolar ion environment. Aerosol Sci. Technol. 5, 1 (1986)
49. Friedlander, S.K.: Dynamics of aerosol formation by chemical reactions. Ann. N Y Acad. Sci.
354, 1 (1983)
50. Prastinis, S.E., Friedlander, S.K., Pearlstein, A.J.: Aerosol reactor theory: stability and
dynamics of a continuous stirred tank aerosol reactor. AiChE J. 32, 177 (1986)
51. McGraw, R., Marlow, W.H.: The multistate kinetics of nucleation in the presence of an
aerosol. J. Chem. Phys. 78, 2542 (1983)
52. McMurry, P.H., Wilson, J.C.: Growth laws for formation of secondary ambient aerosols:
implication for chemical conversion mechanisms. Atmos. Environ. 16, 121 (1982)
53. Lushnikov, A.A., Kulmala, M.: Source enhanced condensation in monocomponent disperse
systems. Phys. Rev. E52, 1658 (1995)
54. Lushnikov, A.A., Kulmala, M., Arstila, H., Zapadinskii, E.L.: Source enhanced condensation
of a single component vapor in the transition regime. J. Aerosol Sci. 27, 853 (1996)
55. Pratsinis, S.E.: Simultaneous nucleation, condensation and coagulation in aerosol reactors. J.
Colloid Interface Sci. 124, 416 (1988)
56. Jacobson, M.Z., Turko, R.P.: Simulating condensational growth, evaporation and coagulation
of aerosols using a combined moving and stationary size grid. Aerosol Sci. Technol. 22, 73
(1995)
57. Tsang, T.H., Rao, A.: Comparison of different numerical schemes for condensational growth
of aerosols. Aerosol Sci. Technol. 9, 133 (1988)
58. Warren, D.R., Seinfeld, J.H.: Simulation of aerosol size distribution evolution in systems with
simultaneous nucleation, condensation and coagulation. Aerosol Sci. Technol. 4, 31 (1985)
59. Aikin, A.C., Pesnel, W.D.: Uptake coefficient of charged aerosols – implication for atmo-
spheric chemistry. Geophys. Res. Lett. 25, 1309 (1998)
60. Davis, E.J.: Transport phenomena with single aerosol particle. Aerosol Sci. Technol. 2, 121
(1983)
61. Davidovits, P., Hu, J.H., Worsnop, D.R., Zahnister, M.S., Kolb, C.E.: Entry of gas molecules
into liquids. Faraday Discuss. 100, 65 (1995)
62. Li, W., Davis, E.J.: Aerosol evaporation in the transition regime. Aerosol Sci. Technol. 25, 11
(1995)
63. Natanson, G.M., Davidovits, P., Worsnop, D.R., Kolb, C.E.: Dynamics and kinetics at the
gas–liquid interface. J. Phys. Chem. 100, 13007 (1996)
64. Shi, B., Seinfeld, J.H.: On mass transport limitation to the rate of reaction of gases in liquid
droplets. Atmos. Environ. 25A, 2371 (1991)
65. Aalto, P., Hammeri, K., Becker, S., Weber, R., Salm, J., M€akel€a, J.M., Hoell, C., O’Dowd, C.,
Karlsson, H., Hansson, H.-C., V€akev€a, M., Buzorius, G., Kulmala, M.: Physical haracteriza-
tion of aerosol particles during nucleation events. Tellus 53B, 344 (2001)
66. Arey, J., Atkinson, R., Aschmann, S.M.: Product study of gas-phase reactions of monoter-
penes with the OH radical in the presence of NOx. J. Geophys. Res. 96, 18539 (1990)
67. Juozaitis, A., Trakumas, S., Girgzdiene, D., Girgzdis, A., Sopauskiene, D., Ulevicius, V.:
Investigation of gas-to-particle conversion in the atmosphere. Atmos. Res. 41, 445 (1996)
68. Boy, M., Kulmala, M.: Nucleation events on the continental boundary layer: influence of
physical and meteorological parameters. Atmos. Chem. Phys. 2, 1 (2002)
69. Lushnikov, A.A., Kulmala, M.: Foreign aerosols in nucleating vapor. J. Aerosol Sci. 31, 651
(2000)
2 Formation of Aerosols in the Atmosphere 95

70. Lushnikov, A.A., Kulmala, M.: Nucleation burst in a coagulating system. Phys. Rev. E62,
4932 (2000)
71. Lushnikov, A.A., Kulmala, M.: Kinetics of nucleation controlled formation and condensa-
tional growth of disperse particles. Phys. Rev. E63, 061109 (2001)
72. Griffin, R., Cocker, D.R., Flagan, R., Seinfeld, J.H.: Organic aerosol formation from the
oxidation of biogenic hydrocarbons. J. Geophys. Res. 104, 3555 (1999)
73. Hoffmann, Th, Odum, J., Bowman, F., Collins, D., Klockow, D., Flagan, R., Seinfeld, J.:
Formation of organic aerosols from the oxidation of biogenic hydrocarbons. J. Atmos. Chem.
26, 189 (1997)
74. Kerminen, V.-M., Virkkula, A., Hillamo, R., Wexler, A.S., Kulmala, M.: Secondary organics
and atmospheric cloud condensation nuclei production. J. Geophys. Res. 105, 9255 (2000)
75. Korhonen, H., Lehtinen, K., Kulmala, M.: Multicomponent aerosol dynamic model UHMA:
model development and validation. Atmos. Chem. Phys. Discuss. 22, 471 (2004)
76. Kulmala, M., Pirjola, L., M€akel€a, J.M.: Stable sulfate clusters as a source of new atmospheric
particles. Nature 404, 66 (2000)
77. Lyubovtseva, YuS, Sogacheva, L., Dal-Maso, M., Bonn, B., Keronen, P., Kulmala, M.:
Seasonal variations of trace gases, meteorological parameters, and formation of aerosols in
boreal forests. Boreal Env. Res. 10, 493 (2005)
Chapter 3
Atmosphere–Ionosphere Electrodynamic
Coupling

V.M. Sorokin and V.M. Chmyrev

Abstract Numerous phenomena that occur in the mesosphere, ionosphere, and the
magnetosphere of the Earth are caused by the sources located in the lower atmo-
sphere and on the ground. We describe the effects produced by lightning activity
and by ground-based transmitters operated in high frequency (HF) and very low
frequency (VLF) ranges. Among these phenomena are the ionosphere heating and
the formation of plasma density inhomogeneities, the excitation of gamma ray
bursts and atmospheric emissions in different spectral bands, the generation of
ULF/ELF/VLF electromagnetic waves and plasma turbulence in the ionosphere,
the stimulation of radiation belt electron precipitations and the acceleration of ions
in the upper ionosphere. The most interesting results of experimental and theoreti-
cal studies of these phenomena are discussed below. The ionosphere is subject to
the action of the conductive electric current flowing in the atmosphere–ionosphere
circuit. We present a physical model of DC electric field and current formation
in this circuit. The key element of this model is an external current, which is formed
with the occurrence of convective upward transport of charged aerosols and
their gravitational sedimentation in the atmosphere. An increase in the level of
atmospheric radioactivity results in the appearance of additional ionization and
change of electrical conductivity. Variation of conductivity and external current in
the lower atmosphere leads to perturbation of the electric current flowing in the
global atmosphere–ionosphere circuit and to the associated DC electric field per-
turbation both on the Earth’s surface and in the ionosphere. Description of these
processes and some results of the electric field and current calculations are pre-
sented below. The seismic-induced electric field perturbations produce noticeable
effects in the ionosphere by generating the electromagnetic field and plasma dis-
turbances. We describe the generation mechanisms of such experimentally
observed effects as excitation of plasma density inhomogeneities, field-aligned

V.M. Sorokin (*)

Pushkov Institute of Terrestrial Magnetism, Ionosphere and Radio Wave Propagation, Russian
Academy of Sciences, IZMIRAN, Troitsk, Moscow Region 142190, Russia

V.L. Bychkov et al. (eds.), The Atmosphere and Ionosphere, 97

Physics of Earth and Space Environments,
DOI 10.1007/978-90-481-3212-6_3, # Springer ScienceþBusiness Media B.V. 2010
98 V.M. Sorokin and V.M. Chmyrev

currents, and ULF/ELF emissions and the modification of electron and ion altitude
profiles in the upper ionosphere. The electrodynamic model of the ionosphere
modification under the influence of some natural and man-made processes in the
atmosphere is also discussed. The model is based on the satellite and ground
measurements of electromagnetic field and plasma perturbations and on the data
on atmospheric radioactivity and soil gas injection into the atmosphere.

Keywords Electromagnetic field  Plasma disturbances  Upper ionosphere 

Atmosphere

3.1 Introduction

This chapter presents a review of the most interesting results of experimental and
theoretical studies of the electromagnetic field and plasma disturbances in the
ionosphere, which are initiated by different natural sources in the atmosphere and
on the ground, and of the technogenic disturbances connected with human activity.
A chain of interconnected processes in the lithosphere–atmosphere–ionosphere
interaction system causes the ionosphere to react on such phenomena as earth-
quakes, volcano eruptions, typhoons, lightning discharges, high power explosions,
functioning of powerful sources of electromagnetic radiation, etc. Due to such
coupling, the ionosphere appears to be a sensitive indicator of the many processes
occurring on the ground and in the near-Earth atmosphere. Apparently, this factor
defines growing interest to investigation of the atmosphere–ionosphere interaction.
A fundamental role in the energy transfer from the atmosphere to the ionosphere
belongs to thunderstorms. Large quasi-electrostatic fields in the mesosphere and the
lower ionosphere connected with cloud-to-ground lightning discharges and intense
electromagnetic pulses of 20 GW peak power generated by lightning current
cause significant ionospheric disturbances because of the heating and acceleration
of electrons, production of ionization, optical emissions, gamma ray bursts, etc.
Other powerful mechanism of lightning influence on the ionosphere is connected
with additional ionization and formation of ionospheric inhomogeneities that are
caused by the radiation belt electron precipitation due to the pitch-angle scattering
of trapped particles in the magnetosphere by whistler mode waves from lightning.
Technogenic effects in the ionosphere also became a subject of intense experi-
mental and theoretical studies. Significant attention is devoted to controlled influ-
ence on the ionosphere aimed at generating the desired ionospheric response
dependent on the parameters of influence. In this connection, one can mention the
generation of ultra low frequency (ULF)/extra low frequency (ELF) waves in a
process of ionosphere heating by the radiation of powerful high frequency (HF)
transmitters [1–4], the formation of artificial very low frequency (VLF) ducts and
other types of inhomogeneities [5–7], the acceleration of ions and the excitation of
atmospheric emissions in different spectral bands [8–11], etc. One can note the
3 Atmosphere–Ionosphere Electrodynamic Coupling 99

transition into an active phase of the investigation of VLF wave impact on the high-
energy particle distributions in the Earth’s radiation belt including the experiments
that use powerful VLF transmitter onboard spacecraft [12,13].
Since physical mechanisms for transfer of perturbations into the ionosphere from
the abovementioned natural and artificial sources in many cases are similar, we
thought it reasonable to discuss both types of influence (natural and man-made) in a
frame of one paper. Such phenomena as the effects of ballistic rocket launches and
of explosions of different kinds on the ionosphere stand by itself. These effects are
of a different physical nature, and, therefore, their consideration lies beyond the
scope of this chapter.
Among the numerous natural sources influencing the ionosphere, the most
important role is thought to be played by earthquakes. An importance is defined
by the potential of using the ionosphere as a sensitive indicator of earthquake
preparation processes and possible tool for the short-term earthquake forecasting
in future. Such potentiality is based on numerous ground and satellite observations
of seismic-related phenomena occurring in the atmosphere and the ionosphere of
the Earth several days or weeks before earthquake. Some results of these observa-
tions are reviewed in refs. [14–17]. Though the number of publications devoted to
seismic effects on the ionosphere is huge, there is no more or less complicated
understanding of cause–effect relations between the phenomena involved in the
ionosphere responding to the seismic processes. Because of the insufficient experi-
mental base and a lack of reliable statistical data on the ionospheric precursors to
earthquakes, currently, there is no tangible ground to use ionospheric phenomena
for sure earthquake forecasting. Nevertheless, a noticeable progress comes to pass
in accumulating and analyzing the data from satellite and ground-based observa-
tions and in theoretical modeling of interconnected processes in the lithosphere–
atmosphere–ionosphere system. Significant contribution is done from the Demeter
satellite research program (see, e.g. [18–20]). A comprehensive model of the lower
atmosphere and the ionosphere coupling was developed in ref. [21].
We can suppose that the earthquake effects in the ionosphere arise as a result of
simultaneous actions of various factors, such as acoustic waves, electric fields,
electromagnetic radiation, chemically active substances, etc. Since seismic activity
is accompanied by the enhanced injection of soil aerosols into the atmosphere, an
important role in the formation of these factors is played by aerosol fluxes, which
influence the electric conductivity and generate external electric currents in the
lower atmosphere [22]. Experimental evidences of earthquake-related enhancement
in the injection of soil aerosols into the atmosphere, the increase in concentration of
some gases (e.g., H2, CO2, and CH4) by several orders of magnitude in seismically
active zone, and the increase in atmospheric radioactivity associated with such
radioactive elements as radon, radium, uranium, thorium, and actinium and their
decay products on the eve of earthquake are presented in refs. [23–28]. Effects of
strong atmospheric processes associated with typhoons on the electric field and
plasma density fluctuations in the ionosphere were reported in refs. [29, 30].
Satellite measurements carried out during last 2 decades (see, e.g., [18, 20, 31, 32]
confirmed the existence of low frequency emissions associated with earthquakes,
100 V.M. Sorokin and V.M. Chmyrev

which was first reported by [33]. Comparison of the ground-based observations and
the satellite measurements of ELF waves performed during different seismic events
showed the similarity of the observed emissions; the only difference was in the wave
intensity in the ionosphere and on the ground [34]. The negative experimental
findings in [35] were connected with the use of incorrect method for data reduction:
short events ( 5 s), such as single or a small number of spherics or whistlers were
not classified as disturbances and were hence not considered. At the same time, the
ELF precursors to earthquakes could be associated with thunderstorm electromag-
netic pulses propagated in the Earth–ionosphere waveguide and scattered into the
ionosphere by plasma inhomogeneities formed in the lower ionosphere over the
seismic zone before an earthquake [36]. The duration of such pulses is much less
than 2 s, and the maximum power is concentrated at frequencies below 500 Hz.
DC electric field perturbations can play a key role in a coupling between the
lower atmosphere and the ionosphere [21]. Ground-based and satellite observations
confirm an enhancement of the electric field both in the ionosphere and on the
ground in a seismic zone before earthquakes [37–42].
Small-scale plasma irregularities and large-scale disturbances of electron and
ion density profiles in the upper ionosphere over seismically active region were
observed before earthquakes [28, 32, 43–46].
The formation of thermal anomalies in seismic zones several days before large
magnitude earthquakes presents another pronounced effect that is well observable
from satellites with the use of IR sensors [47–50]. In addition, we can mention the
observations of anomalous airglow at 557.7 and 630 nm correlated with the growth
of seismic activity [51].
Numerous theoretical models were suggested for explanation of separate phe-
nomena stimulated in the atmosphere and the ionosphere by seismic activity.
Reference [52] considered the generation and propagation of internal gravity
waves and their effects on the ionosphere. The generation of ULF emissions by
lithospheric sources and their possible penetration into the ionosphere was analyzed
in refs. [53–55]. The formation of electric currents in the lithosphere and the
propagation of the excited waves from the source region into the ionosphere were
considered in refs. [56,57]. Possible acoustic effects on the ionosphere and the
generation of geomagnetic pulsations were investigated in refs. [38, 58, 59] consid-
ered the conductivity and the electric field variations in the lower atmosphere that
can be caused by increase in radon concentration connected with the growth of
seismic activity. The enhanced injection of radon and metallic aerosols in the
epicenter region and its influence on the ionosphere were discussed in ref. [28].
The mechanisms for the electric field penetration from the source region in the
lithosphere into the ionosphere and related effects on the ionosphere were analyzed
in refs. [60,61].
Electrodynamic model of the ionosphere response to seismic-related lower
atmosphere disturbances is developed in ref. [22]. This model describes the com-
plete chain of processes in the lithosphere–ionosphere coupled system, which starts
from the injection of radioactive substances and charged aerosols into the atmo-
sphere and the formation of external electric current. This current is responsible for
3 Atmosphere–Ionosphere Electrodynamic Coupling 101

the enhancement of DC electric field and the subsequent development of acoustic-

gravity wave (AGW) instability in the ionosphere. It results in the generation of
conductivity inhomogeneities in the lower ionosphere, the formation of magnetic
field-aligned currents and plasma density disturbances in the upper ionosphere
stretched along geomagnetic field lines. Besides, the electric field increase leads
to an additional Joule heating of the lower ionosphere, which results in an elevation
of F-layer maximum, a decrease of electron density in the maximum of this layer,
and a growth of light ion density in the upper ionosphere. Thus, this model connects
disturbances of the electric field and the key ionosphere parameters with increase of
atmospheric radioactivity and injection of charged aerosols into the ionosphere.
Detailed consideration of all these processes is presented below.

3.2 Effects of Lightning and Powerful Ground-Based

Transmitters on the Near-Earth Space

3.2.1 Lightning Effects on the Ionosphere and the Magnetosphere

Thunderstorms play a fundamental role in the energy transfer from the atmosphere
to the ionosphere. At any given time, more than 2,000 thunderstorms are active over
the globe, and on average, lightning strikes the Earth about 100 times per second
[62]. Every cloud-to-ground lightning discharge transfers to the ground a charge of
the order of 300 C during several milliseconds, leading to the formation of large
quasi-electrostatic fields in the mesosphere and the lower ionosphere over millisec-
ond time scales [63]. These fields together with intense electromagnetic pulses of
20 GW peak power generated by lightning current [64] cause significant distur-
bances in the lower ionosphere due to heating of the ambient electrons and
acceleration of runaway electrons, producing ionization and optical emissions
[63, 65–67].
Plasma inhomogeneities caused by high-energy electron precipitation from the
Earth’s radiation belt in a process of pitch-angle scattering of trapped particles by
intense whistler mode waves from lightning discharges present other class of
lightning-induced disturbances in the ionosphere. Troposphere–magnetosphere
coupling of such kind has been many times confirmed in ground-based observations
[68–70], balloon measurements [71], rocket experiments [72, 73], and satellite
observations [74–76]. Theoretically, the mechanism of pitch-angle scattering and
particle precipitation from the radiation belt by ducted whistlers in the magneto-
sphere is also well investigated (see, e.g. [77, 78] and references therein).
Experimental indication of strong electrodynamic coupling of thunderstorms
to the mesosphere and the lower ionosphere includes such phenomena as fast
lightning-induced perturbations of subionospheric VLF transmitter signals propa-
gating over the thunderstorm areas [79–84], optical emissions associated with
sprites [85–91] and blue jets [92], and airglow enhancements [93] connected with
102 V.M. Sorokin and V.M. Chmyrev

short-term (<1 ms) “elves” emissions excited at 80–95 km altitudes before

sprites [65, 88, 94]. Sprites clearly associated with intense lightning discharges
[95] are correlated with fast perturbations of subionospherically propagating VLF
signals [82], providing an evidence of interconnection of these phenomena. An
important manifestation of strong upward energy pumping from thunderstorms
was obtained from the observations of gamma-ray flashes of atmospheric origin
[96] associated with lightning discharges [97] and of intense bursts of broadband
VHF emissions [98].
The existing generation models of the abovementioned phenomena are based
on heating of the ambient electrons by electromagnetic pulses from lightning
discharges [65, 99–102] or by quasielectrostatic thundercloud electric fields
[103–106], and on runaway electron acceleration processes [107–109].
First registration of optical emissions stimulated by lightning has been reported
by Franz et al. in ref. [110]. Later, numerous observations of similar flashes of
luminescence called sprites [111] have been performed from the ground [86, 90,
91], the aircraft [92, 111–113], and space-based platforms [89,114]. The main
characteristics of sprites obtained from these measurements could be summarized
as follows. This phenomenon is usually observed at 50–90 km altitudes over positive
cloud-to-ground lightning discharge with peak current of the order of 100 kA or
above. Transverse size of luminescence area is 5–50 km; sometimes containing
structural elements with a cross section of 2 km. Sprite arises 1.5–4 ms after
lightning discharge and lasts for several milliseconds. The brightness of sprites was
estimated to be 25–50 kR, for most intense events it was up to 100 kR. Maximum of
intensity ranges between spectral bands 650–680 nm and 750–780 nm.
Blue jets present other type of luminescence occurring at altitudes below 50 km
over thunderstorm clouds. They are observed as narrowly collimated light beams of
blue color propagating upward from thunderclouds with velocity 100 km/s up to
40–50 km altitudes, where they disappear. Blue jets in contrast to sprites are not
accompanied by strong electromagnetic radiation in VLF range. Observations did
not display direct connection with simultaneous cloud-to-ground lightning dis-
charges [115]. Blue jet generation theory based on accumulating charges in thun-
dercloud and subsequent breakdown ionization and upward propagation of
ionization wave has been developed by Pasko et al. in ref. [105].
Electron heating in the lower ionosphere by the thundercloud’s electric field can
lead to enhancement of infrared radiation, particularly to intensification of CO2
emission at 4.3 mm wave length [116]. Different electromagnetic emissions asso-
ciated with thunderstorm activity have been observed onboard the Demeter satellite
[117]. The effects of ion heating and generation of plasma turbulence in the
ionosphere under the influence of lightning discharges were reported by Berthelier
et al. in ref. [118] based on the Demeter data.
As mentioned above, one of important mechanisms of the troposphere–
ionosphere–magnetosphere coupling is lightning-induced precipitation of high
energy (>40 keV) electrons from the Earth’s radiation belt into the ionosphere
[76, 119, 120]. Comprehensive data of satellite observations of this phenomenon
were reported by Voss et al. [76], who used the low orbital (170–280 km) S81-1
3 Atmosphere–Ionosphere Electrodynamic Coupling 103

polar satellite measurements of electron and ion fluxes with energies 2–1,000 keV
and 10–1,500 keV, respectively. These observations were accompanied by the
ground wave measurements at Palmer station in Antarctica. Analysis of numerous
precipitation events associated with lightning has shown that the time delay
between causative spheric and appearance of first burst of precipitating particles
registered on satellite increased with L-shell according to prediction of theory [78].
The global distribution of intense precipitations of such kind correlated with
thunderstorm activity distribution and displayed maximum in the depletion zone
of the electron radiation belt at 2 < L < 3. The density of electron energy flux in
lightning-induced precipitations was about 103 ergcm2s1 (104 J/(m2 s)) and
particles with an energy of 100–200 keV prevailed. The most recent experimental
data on electron precipitations stimulated by lightning were obtained from the
Demeter satellite [121]. Bursts of 100–300 keV electrons registered within 1s
after an arrival of causative VLF pulses from lightning were observed.

3.2.2 Effects of the Ionosphere Heating by Powerful

HF Transmitters

The first successful observations of the lower ionosphere heating by electromag-

netic radiation from a powerful ground-based HF (3–30 MHz) transmitter were
carried out in ref. [122]. They used the experimental HF transmitter in Platteville,
Colorado, 50 MW effective isotropic-radiated power, as a heater and subiono-
spheric VLF (20 kHz) and LF (60 kHz) probe signals for detecting lower iono-
sphere heating by measurements of amplitude and phase perturbations in these
signals. Transmissions were made at 5.1 and 7.4 MHz to heat the midday D-region
and the diameter of the heated region at the reflection height of probe waves
(70 km) was 25 km. The heating effect in this region was obtained from the
amplitude perturbation 0.53% (0.046 dB) at 60 kHz and phase perturbation 0.3
at a frequency of 20 kHz.
Later, subionospheric VLF probe waves were used to probe D-region heating in
the experiments by Barr et al. in refs. [123, 124] with steerable heater in Tromso
(Norway) operated at the frequency of 2.759 MHz; the maximum effective radiated
power was 200 MW. VLF probe waves were from the 12.1 kHz Omega navigation
transmitter in Aldra, Norway, and the 16.4 kHz JXN transmitter nearby Helgeland.
The heated spot was located near the end of <500 km great circle paths from the
transmitters and had a form of a 27 by 34 km ellipse at 80 km height since the
heating wave beam was focused on the Omega probe wave path (38 from vertical).
Amplitude and phase perturbations of sub-ionospheric signals were correspond-
ingly 0.05 dB and 0.3 during the daytime heating and up to 6 dB and 50 at night.
The heating experiment with the use of 2.8 MHz HIPAS transmitter near
Fairbanks, Alaska, and 23.4 kHz probe waves from NPM transmitter in Oahu
received at Fort Yukon, Alaska has demonstrated the perturbations in NPM
amplitudes of 0.2 to þ 0.4 dB simultaneous with heater modulation patterns [125].
104 V.M. Sorokin and V.M. Chmyrev

Beginning practically from earlier experiments, the lower ionosphere heating by

intense modulated HF radiation in the presence of natural ionospheric electric
currents (auroral electrojet and middle latitude dynamo currents) was investigated
as a possible tool for the generation of electromagnetic ULF/ELF/VLF waves [1, 4,
126–131]. The generation mechanism is relatively simple. Periodic heating of the
ionosphere in the current region modulates the electrical conductivity in this region
and consequently causes the modulation of natural electric current. This process
forms giant radiating antenna at the altitudes of 60–100 km. The generation of ULF/
ELF/VLF waves by such antennas is extensively studied with the use of special
high-frequency (HF) heating facilities in Tromso (Norway), Arecibo (Puerto Rico),
HIPAS (High-Power Auroral Stimulation), and HAARP (High-frequency Active
Auroral Research Program) (Alaska) and SURA (Russia). The most interesting
results of these studies are presented below.
HF heater in Tromso operated at the frequencies from 2.7 to 8.0 MHz. A
successful generation of ELF/VLF waves was carried out with the use of amplitude
modulation by rectangular pulses in the frequency range 0.2–6.5 kHz. Maximum
amplitudes of generated waves 1, 0.1, and 0.03 pT were registered at the distances
of 20, 200, and 500 km from the HF transmitter correspondingly [124, 126].
Successful excitation of the Earth–ionosphere waveguide and reception of artifi-
cially generated ELF/VLF signals with amplitude 10 fT at the distance
2,000 km has been demonstrated in several transmissions [129].
Amplitude measurements at large distances were used to estimate the power of
ELF/VLF radiation from the ionosphere. Far-zone ground-based measurements
have shown the amplitudes corresponding to 1 W dipole radiation source at the
altitude of maximum modulation of Hall conductivity in the ionosphere [124].
However, the satellite measurements [2] over the same heating facility have
given an estimate of 30 W for the ionospheric ELF/VLF oscillator. Different
source characteristics related to simultaneous ground and satellite measurements
have been also obtained in ref. [132].
Several successful experiments on ELF/VLF wave generation were carried out
in Arecibo (Puerto Rico) where a powerful HF transmitter has modulated a current
of the equatorial electrojet. In these experiments, the heater frequency and the
radiated power were 3 MHz and 800 kW, respectively, and the modulation
frequencies were in the range from 500 to 5,000 Hz [133]. The efficiency of
ionospheric ELF/VLF wave generator depending on the polarization of HF heater
radiation and its daily work cycle was investigated by Ferrano in ref. [127].
High-frequency heating facility HIPAS located near Fairbanks (Alaska) was
successfully used for the modulation of auroral electrojet in the experiment [134],
in which the generation of ULF/ELF waves (11–76 Hz) with amplitudes 1 pT at the
phase modulation of a heater radiation has been observed. Villasenor et al. [135] have
investigated the dependence of ELF/VLF wave generation efficiency on the modu-
lation schemes of HF transmitter including the amplitude rectangular pulse modula-
tion, phase and two frequency modulations. It was shown that the amplitude pulse
modulation is most effective. ELF/VLF waves generated with the use of HIPAS
heating facility have been observed in space plasma in ref. [136]. In this paper, in
3 Atmosphere–Ionosphere Electrodynamic Coupling 105

contrast to similar measurements over the Tromso station [132], good correspon-
dence between the satellite and ground-based measurements was demonstrated.
Experiments on generation of ELF/VLF waves with HAARP heating facility
were carried out with use of different frequencies and polarizations of HF radiation
[137]. It was shown that the possibility existed to control the polarization of
artificially generated ELF waves by changing the frequency and the polarization
of the heater radiation [137]. Within the framework of this Program the
Stanford University has carried out a series of successful experiments on the
injection of artificially excited ELF/VLF waves into the upper ionosphere and
the Earth–ionosphere waveguide. Platino et al. [138] have reported the Demeter
satellite observations of ELF/VLF emissions generated by HAARP heater in the
ionosphere. For three of four modulation frequencies the satellite and the ground-
based measurements have shown the same results corresponding to radiated power
of ionospheric source 3–4 W. Multi-hop magnetospheric echo of ELF/VLF
signals induced by the HF transmitter have been observed by Inan et al. [3]. These
echoes displayed dispersion and amplification in equatorial region of the magneto-
sphere leading to generation of trigger emissions. ELF/VLF waves excited by the
HAARP heating facility were also injected into the Earth–ionosphere waveguide.
Moore et al. [4] have demonstrated unique results of observations of HAARP-
induced ELF emissions at the distance 4,400 km from the heater. Characteristics
of these emissions conformed to 4–32 W power of ionospheric oscillator at the
altitude 75–80 km [4]. Long-term program of the experiments with HAARP HF
transmitter performed by Stanford University have shown that heating of the lower
ionosphere by powerful amplitude modulated HF radiation could be an effective
tool for generation of ELF and VLF waves with given properties practically at any
geomagnetic conditions [139].
The formation of VLF ducts (magnetic field-aligned channels of enhanced
ionization) in the upper ionosphere recently detected by the Demeter satellite
presents other important effect of the ionosphere HF heating. Characteristics of
these ducts and other types of plasma disturbances and wave emissions observed
over the heating facilities were presented in [6, 7, 140, 141]. Typical size of ducts
across geomagnetic field was 40 km at the satellite altitude and the plasma density
excess within a duct over background level was 20–30%. Temperature and plasma
density fluctuations in the ionosphere over the heating area were about 10–15%.

3.2.3 Influence of Powerful VLF Transmitters on the

Ionosphere and the Magnetosphere

3.2.3.1 Artificially Stimulated Particle Precipitation into the Ionosphere

First experimental confirmation of artificial stimulation of radiation belt electron

precipitation was obtained in 1975 by Zhulin et al. [142] with use of powerful
106 V.M. Sorokin and V.M. Chmyrev

ground-based VLF transmitter located in Arkhangelsk region (Russia). Diagnostics

of electron precipitation was performed by special optical devices installed onboard
the aircraft-laboratory Yak-40. This experiment has shown significant enhancement
in intensity of atmospheric emission at 557.7 nm correlated with periodic VLF
transmitter pulses. Pulsations of the emission intensity were up to 40 R. An estimate
of precipitating electron energy flux with account of the experiment’s geometry
has given 0.13 erg/cm2s (1:3  104 J/m2s) that corresponded to flux of electrons
with energy 15 keV approximately 5  104 particles/cm2s and was a factor of 10
above the background level.
Direct measurements of artificially stimulated electron precipitations from the
radiation belt were carried out onboard the S81-1 satellite with use of USA VLF
transmitters NAA and NSS [143–145] and independently by the Aureol-3 satellite
[146, 147]. Bursts of precipitating electrons were observed by S81-1 satellite 1–2 s
after commencement of the transmitter pulses [143, 144]. Similar delay followed
from the theoretical estimates [145]. The most recent results on the electron
precipitations stimulated by ground VLF transmitter were reported by Sauvaud
et al. [148] and Gamble et al. [13], based on the Demeter satellite data.

3.2.3.2 Effects of the Ionosphere Heating by Powerful VLF Transmitters

Experiments on artificial stimulation of radiation belt electron precipitation led to

discovery of the effect of lower ionosphere heating by the radiation from ground-
based VLF transmitters with use of probe VLF signals [149]. First experiments were
performed with NAU transmitter in Puerto Rico. About 100 kW power at 28.5 kHz
were radiated by pulses of 3 s duration and 5 s repetition period. Search for lower
ionosphere perturbations was carried out using probe signals from several US VLF
transmitters at the receiving stations in Arecibo (Puerto Rico) and Palmer (the
Antarctic). Data analysis revealed the 5 s periodicity in a spectrum of amplitude
envelope of probe signals from the NAA transmitter. Perturbations of amplitude
with the modulation period of heating transmitter were in the range from 0.04 to
þ0.12 dB [149]. It is interesting to note that such disturbances caused by 100 kW
NAU VLF transmitter coincided in the order of magnitude with the heating effect of
the Tromso HF transmitter, the effective radiated power of which was 200 MW
[150]. This seeming contradiction was resolved by Taranenko et al. [151] who
showed that the amplitude disturbances of probe VLF wave in the Earth–ionosphere
waveguide in both cases should be comparable in magnitude since VLF heating is
maximum at the reflection heights of subionospheric signals in D-region, and
consequently the heating effect on the probe wave amplitude should be strongest
there.
A series of purposeful experiments on VLF heating of the lower ionosphere was
performed by Stanford University in the period from 1987 to 1992 [149]. In the
experiment of 1992, the amplitude and phase perturbations of probe signals at
21.4 kHz from NSS transmitter in Annapolis (Maryland) were registered in Gander
(Newfoundland). Investigated were the effects of ionosphere heating by the NAA
3 Atmosphere–Ionosphere Electrodynamic Coupling 107

transmitter in Cutler (Maine) operating at 24 kHz. For monitoring measurements,

the station in Gander received the NLK (24.8 kHz) and the NAU (28.5 kHz) signals
whose great circle paths did not pass through the ionosphic region heated by the
NAA transmitter. The ionosphere heating effect was found in 41 of 53 episodes of
NAA transmitter “ON/OFF”. The amplitude and phase perturbations of probe
signals lie in the ranges from 0.11 to þ0.84 dB and from 0 to 5.3 , respectively
[152, 153]. In a similar experiment with the NLK transmitter as a heater, the
observed perturbations of probe wave amplitude on the NPM (Hawaii) – Saskatoon
(Saskatchewan) propagation path ranged from 0.3 to 1.6 dB [152, 153].
Strong ionospheric disturbances in the zone of powerful VLF transmitter NWC
(Australia) were observed from the Demeter satellite [5]. Significant enhancement
of electron temperature and fluctuations of electron and ion densities were recorded
together with intensification of HF and ELF electrostatic waves. Dimension of the
perturbation zone in the ionosphere at the satellite altitude (700 km) was about
800 km along the orbit, the position of center of this zone corresponded to L-shell of
the NWC transmitter (L ¼ 1.41).
Thus numerous experiments show that intense VLF waves from different
sources influence on the lower ionosphere and therefore on the characteristics of
radio wave propagation sensible to changes in the ionosphere parameters.
Three-dimensional (3D) theoretical modeling of the lower ionosphere heating by
VLF transmitter radiation was carried out by Rodriguez and Inan [153] and
Rodriguez [152]. The heated area in the ionosphere has a form of a spot with a
radius of 150 km at the level half of the perturbation maximum. Position of
maximum is slightly shifted to pole from the position of heater. For transmitters
with radiated power 1,000 kW the electron temperature in maximum of heating
(87 km) exceeds the undisturbed level approximately threefold.

3.2.3.3 Acceleration of Ionospheric Ions

Effect of acceleration of ions in the ionosphere under influence of radiation from

ground-based VLF transmitter was first observed onboard the Aureol-3 satellite [8].
The observations were made in the zone of Arkhangelsk VLF transmitter (64 240 N,
41 320 E, L ¼ 4), which radiated 300 kW power at the carrier frequency 19.1 kHz
with amplitude modulation by periodic pulses (8 s ON–8 s OFF). The satellite
instrumentation set provided measurements of ion distributions for M ¼ 1–32
atomic mass units at energies from 0.01 to 3.5 keV/charge and pitch-angles 85
and 145 combined with analysis of electromagnetic fields in the frequency range
from 10 Hz to 20 kHz.
The observations showed that a mid-latitude increase in H+, He+ and O+ ion fluxes,
of about an order of magnitude above the background level, at the energies 250–330
eV and pitch-angles 85 and 145 occurred in the zone, in which the transmitter signals
were detected. Such quasi-transverse and upward ion fluxes with magnitude up to
106–107 cm2s1sr1keV1 were observed in five passages of the Aureol-3 satellite
108 V.M. Sorokin and V.M. Chmyrev

over the zone of Arkhangelsk VLF transmitter. It is significant that the upward (145 )
ion fluxes directed from the ionosphere into the magnetosphere were of the same
order of magnitude as the quasi-transverse (85 ) ion fluxes. Simultaneously with
anomalous ions the VLF emission at the frequencies 4.5 kHz well correlated with
the transmitter pulses have been observed. These frequencies were close to local low
hybrid resonance frequency. Let us note that the measurements were carried out in the
conditions of low geomagnetic activity (Dst ¼ 10), therefore the zone of natural
auroral precipitations lied in higher latitudes.
From the discussed Aureol-3 data Dzhordzhio et al. [8] have concluded that
intense electromagnetic waves radiated by ground VLF transmitter could stimulate
acceleration up to 300 eV energy and injection into the magnetosphere of iono-
spheric O+, He+ , H+ ions combined with excitation of ELF/VLF noises.
Due to high phase velocity of whistler mode waves, their interaction with supra
thermal (<1 keV) ions in the ionosphere is not effective. For explanation of these
experimental results, it seems reasonable to consider a two-step process – nonlinear
conversion of whistler waves from the transmitter into other plasma modes and the
subsequent interaction of these secondary waves with ionospheric ions. Mechan-
isms for transformation of waves and acceleration of ions are considered below.
Continuation of active experiments on the influence of VLF transmission on ion
distributions and ELF/VLF emissions in the ionosphere has been performed by
Chmyrev et al. [10]. These experiments were carried out onboard the COSMOS-
1809 satellite using a Arkhangelsk VLF transmitter operated at frequencies 15 and
19.1 kHz in the same regime as above (300 kW radiated power, pulse modulation:
8 s ON–8 s OFF). The satellite was operating on the near circular orbit with altitude
960 km and inclination 82.5 . Eight episodes of the transmitter operation were
carried out during the satellite passages over the transmitter zone. Distance in
longitude between the satellite trajectory and the transmitter location in these
episodes was 30 .
Figure 3.1 shows the distribution of low-energy ion fluxes in the ionosphere over
the transmitter zone obtained on the satellite orbit 1376, which was nearest to the
transmitter. Each of the four panels in Fig. 3.2, which corresponds to different energy
channels of electrostatic analyzer SF-3M, presents the ion fluxes for pitch-angles
a ¼ 5–40 , 50–85 , 95–130 , 140–175 . In this episode, the transmission started at
20.30 UT when COSMOS-1809 crossed the L-shell of the transmitter (L ¼ 4).
It is seen from Fig. 3.1 that effective heating of ions occurred in the ionosphere over
the transmitter zone, where the flux of ions with energy 28–74 eV increased from the
background level of 103 cm2s1sr1keV1 up to (2–4)104 cm2s1sr1keV1
during the time smaller or of the order of 1 min. An important peculiarity of the
particle distributions in Fig. 3.1 is that the downward ion fluxes were the strongest and
the particles with higher energy appeared later (at higher L-shells). For precipitation
and quasi-transverse fluxes, it means that particles accumulate energy at larger
length or at higher altitudes along the line of magnetic field above the satellite.
The characteristic transverse scale of the acceleration zone, as estimated from 28 eV
ion distribution, is 200 km at the satellite altitude (h  960 km). The analysis
3 Atmosphere–Ionosphere Electrodynamic Coupling 109

–1 –2 –1 –1
Ion flux (c cm str eV
Cosmos-1809 27 March 1987 Orbit 1376
4
210
9 eV
104

4104 28 eV

2104

2104
74 eV
104

2104
205 eV
104

UT 20.31.00 20.31.30
Latitude (deg) 64.8 66.4
Longitude (deg) 38.1 39.3
Altitude (Km) 961 962
L 4.6 5.1

Fig. 3.1 Distribution of low energy ion fluxes along the orbit 1376 of COSMOS-1809 satellite
over the Arkhangelsk VLF transmitter zone for energies  9, 28, 74, and 205 eV and pitch-angles
5–40 (–––), 50–85 (– – –) , 95–130 (–  –), and 140–175 (  )

of energy spectrum of ions for all pitch-angles showed that the enhancement of
ion fluxes in the transmitter zone took place in the range from 10 to 200 eV with
maximum at 20–70 eV.

3.2.3.4 Parametric Excitation of ELF Waves in the Upper Ionosphere

Enhancement of intensity (stimulation) of ELF noises correlated with VLF trans-

missions was observed on three passages of COSMOS-1809 over the transmitter
zone where the discussed above ion acceleration occurred [10]. Stimulated
ELF emissions were found very near the transmitter (within 5 in longitudes
and 3 in latitudes) in the region of maximum intensity of the transmitter signals.
The spectral density of noises at the lower part ( f < 400 Hz) of the device
frequency range increased during the transmitter pulses 2–3 times above the
undisturbed level observed in pauses between the pulses. A maximum of induced
ELF wave intensity appeared to be at 140–180 Hz.
110 V.M. Sorokin and V.M. Chmyrev

Spectra of VLF transmitter signals during the observation of stimulated ELF

emissions at all three orbits was found to be broadened; the characteristic broaden-
ing was D f  (200–400 Hz). This value corresponds to the frequency band of
ELF emissions correlated with VLF transmitter pulses.
Let us note that the effect of spectral broadening of VLF signals in the iono-
sphere and the magnetosphere of the Earth were numerously observed from satel-
lites and on the ground [154–157]. Titova et al. [156] interpreted this effect in terms
of VLF signal scattering by low-frequency plasma turbulence (in particular the ion-
cyclotron one) in the ionosphere below and/or near the satellite. At such scattering
process the observed electrostatic ELF emissions were not correlated with the
transmitter signals and the broadening effect took place for all transmitters inde-
pendently on their power [156].
In the experiment [10] such scattering of VLF signals was also observed; ELF
emissions in these cases were not correlated with transmitter pulses and the
spectrum-broadening effect was registered for all VLF transmitters (independently
on power) observed on satellite at the frequencies below 20 kHz. However, besides
the scattering by ionospheric turbulence, the nonlinear VLF wave interaction
process occurred in the experiment [10], which resulted in the generation of ELF
waves and depended on the intensity of pump wave. Such conclusion was made
based on the correlation between the observed ELF emissions and the transmitter
signals and also from the fact that spectrum broadening in such events has been
observed only for powerful signals. The latter followed from the comparison of
simultaneously observed signals of the Arkhangelsk VLF transmitter (radiated
power 300 kW) at 19.1 kHz with signals from the OMEGA transmitter in Norway
(radiated power 10 kW) at 10.2 kHz.
The measurements of wave intensity in separate spectral channels showed that the
propagation of a strong electromagnetic wave from a VLF transmitter through the
ionosphere was accompanied by the generation of rather intense (8 mV/mHz1/2)
secondary emissions in the electric field at frequencies 140 Hz. This emission
was well correlated with the transmitter signals, while other spectral channels did
not indicate any reaction on the transmitter operation. However, magnetic component
of the emission at 140 Hz did not exceed the level of noise (0.3 pT/Hz1/2). The filter
bandwidth in a channel 140 Hz of the spectrum analyzer was 20 Hz. The ratio
of magnetic-to-electric field components for the transmitter signals was N ¼ B/E  18
that corresponded to the refraction index of the whistler mode wave.
Thus, the presented experimental results confirm the generation of ELF waves
and acceleration of ionospheric ions under the influence of strong VLF transmis-
sions. Before considering physical mechanisms for such influence, let us resume the
main characteristics of the observed phenomena:
l The excitation of ELF emissions correlating with Arkhangelsk VLF transmitter
signals was observed in the frequency range 70–400 Hz with maximum intensity
8 mV/mHz1/2 at frequencies 140–180 Hz.
l Simultaneously, with the generation of ELF emissions, the spectral broadening
D f ¼ (200–400 Hz) was observed in Arkhangelsk transmitter signals while
3 Atmosphere–Ionosphere Electrodynamic Coupling 111

broadening of weaker signals from the OMEGA transmitter in the same zone
was not found.
l The enhancement of ion flux in the transmitter zone took place in the range of
energies from 10 to 200 eV; maximum of flux intensity (2–4)104 cm2
s1sr1eV1 was observed at 20–70 eV.
l Accelerated ion fluxes were anisotropic; the highest anisotropy was observed at
28 eV.
As we mentioned the interaction of low energy (1 keV) ions with whistler
waves at the transmitter frequencies and in ELF range is not effective due to high
phase velocities of these waves in the ionosphere. Therefore, it was reasonable to
suppose that suprathermal ions were accelerated by one of the ion branches of
plasma waves, which could be excited due to parametric instability of intense
whistler mode waves. In the frequency range 70–400 Hz, it could be electrostatic
ion-cyclotron waves, ion-acoustic, and electromagnetic ion-cyclotron waves, the
generation of which has been considered in many papers (see, e.g. [158, 159]). The
highest growth rates are realized in a process of decay interaction of whistler with
low frequency plasma wave and ion-cyclotron wave. Detailed analysis of the ion
wave modes generation by decay instability of quasimonochromatic whistler waves
with account of several types of ions and variations of ion composition with altitude
in the ionosphere has been performed by Chmyrev et al. [160]. Effectiveness of
decay process with participation of electromagnetic ion-cyclotron waves in the
ionosphere is substantially higher than with electrostatic waves due to strong linear
attenuation of latter [159]. Electromagnetic ion-cyclotron waves present one of the
most effective sources of ion acceleration in the upper ionosphere [161]. With
account of these circumstances Taranenko and Chmyrev [159] and Chmyrev et al.
[160] suggested the following scheme for explanation of the above-discussed
experimental results. Intense whistler mode wave from the ground VLF transmitter
decays in the upper ionosphere on other whistler propagating near the resonance
cone and highly oblique left-polarized electromagnetic ion-cyclotron wave. The ion
waves propagating upward in the region of decreasing geomagnetic field will be
effectively absorbed by ions providing their acceleration. This mechanism is
similar to [161] with the difference that in [161] the electromagnetic ion-cyclotron
waves are excited due to instability of precipitating auroral electrons while in [159,
160] – due to parametric instability of intense whistler waves. Analysis of growth
rates g1 and frequencies o3 of electromagnetic ion-cyclotron waves for this process
showed that excitation was possible only for waves with large transverse compo-
nents of wave vectors [159]. The corresponding frequency range o3  100–350 Hz
calculated for altitudes below 2,000 km was in agreement with the data on genera-
tion of ELF emissions and VLF wave spectra broadening. Maximum growth rate
was g1  2–3 s1 at the amplitude of pump wave 10 mV/m [159].
Let us consider now the interaction of parametrically excited electromagnetic
ion-cyclotron waves with ions [10]. Dispersion equation for these waves as known
can be written as follows:

o3 ¼ kjj3 vA ð1  o3 2 =oBi 2 Þð1 þ k2 Þ1=2 ; k ¼ ck?3 =ope ; (3.1)

112 V.M. Sorokin and V.M. Chmyrev

where k||3, k⊥3 are longitudinal and transverse components of wave vector, oBi is
cyclotron frequency of ions, vA is Alfven velocity, c and ope are light velocity and
electron plasma frequency. Resonance energy of ions interacting with wave (3.1) at
first cyclotron resonance is defined as


2
Mv2==res Mv2A o2 oBi
W==res ¼ ¼ 1  23  1 ð1 þ k2 Þ1 ; (3.2)
2 2 oBi o3

where v//res ¼ (oBi  o3)/k//3, M is ion mass, v//res and k//3 are oppositely directed.
Let us estimate W//res for H+ ions in the ionosphere at altitudes 2,000 km, where vA 
108 cm/s. At k ¼ 3, which corresponds to maximum growth rate we find W//res ¼ 1.3
eV for frequency o3/oBi ¼ 0.9 and W//res ¼ 12.5eV for frequency o3/oBi ¼ 0.8.
Following ref. [161], let us estimate an effectiveness of resonance heating of
ions in the conditions of the above-considered experiment [10]. We assume that
some part of the geomagnetic field tube is filled by electromagnetic ion-cyclotron
waves, born in the decay process, with a sufficient wide-frequency band so that in
every part of the trajectory in an inhomogeneous magnetic field, the particle could
find the wave, which satisfies the condition of local cyclotron resonance and
therefore experiences continuous acceleration along the trajectory. The transverse
energy increment of resonance ion can be written as

MðDV? Þ2 e2 E2?3
DW? ¼ ¼ ðDtÞ2 ;
2 2M

where e and M are particle charge and mass, E?3 is transverse component of wave
electric field, Dt is a time of particle being in resonance with wave. In the vicinity of
resonance frequency, one may assume E⊥32 ¼ P( f )D f, where P( f ) is a spectral
density of the emission and D f is the frequency bandwidth. For the wide band wave
packet, it is possible to believe the resonance time Dt to be equal to the characteris-
tic time of phase correlation in the packet, that is, Dt(D f )1. Then, for the energy
growth rate, we obtain [161]

dW? e2 Pð f Þ
¼¼ : (3.3)
dt 2M

The movement of a particle in an inhomogeneous magnetic field B0(s) in

approximation of a leading center with an account of Eq. (3.3), neglecting the
geomagnetic field-aligned electric field, is described by the following equations
[161]:

dW? dv==
¼ W? v== B1
0 dB0 =ds þ e Pðf ; sÞ=2M; M
2
¼ W? B1
0 dB0 =ds; (3.4)
dt dt
3 Atmosphere–Ionosphere Electrodynamic Coupling 113

where v// is longitudinal velocity of particle, W⊥ is transverse energy, and ds is an

element of length. For dipole magnetic field, Eq. (3.4) has a form

2dW? 3 sin ’ð3 þ 5 sin2 ’Þ e2 Pð f ; ’Þ

v== ¼ v== W? þ LR0 cos ’ð1 þ 3 sin2 ’Þ1=2 ;
2d’ cos ’ð1 þ 3 sin2 ’Þ 2M

2dv== 3 sin ’ð3 þ 5 sin2 ’Þ

Mv== ¼ W? ð3:5Þ
2d’ cos ’ð1 þ 3 sin2 ’Þ

where ’ is geomagnetic latitude, L is McIlwain parameter, and R0 is the Earth’s

radius. When obtaining Eq. (3.5) we assumed d=dt ¼ v== ð@’=@sÞd=d’:
Numerical integration of Eq. (3.5) along the part of magnetic field line at L ¼
4.64, where ion acceleration took place was performed in [10]. Integration was
carried out in the range of D’ from initial latitude ’0 corresponding to altitude h at
a given magnetic field line downward to final latitude ’1 ¼ 61.3 corresponding
to altitude 450 km. While calculating, it was assumed that electromagnetic ion-
cyclotron waves were generated along a geomagnetic field line and that their
intensity was constant (dP/d’ ¼ 0) within D’. A resonant particle with initial
parallel energy W//0 in the point ’ ¼ ’o was moving downward. Calculations were
made for H+ ions with initial transverse energy W⊥0 ¼ 0 and initial parallel energies
W//0 ¼ 12.5 and 6.25 eV. Spectral density of electromagnetic ion-cyclotron waves
was taken according to experimentally observed value P ¼ 8  1011 V2 m2Hz1.
It was found that at this wave intensity, the particle energy increase could reach
DW  20–30 eV in the interval D’ ¼ 3–5 , which corresponds to interval of heights
Dh  1,400–2,300 km on a given magnetic field line. Pitch-angle, with which the
particle comes to satellite at the altitude 1,000 km, depends on the altitude (or
latitude D’) and on initial parallel energy, from which this particle starts accelera-
tion. The particles accelerated at altitudes below satellite and the accelerated
particles came from above and turned back in mirror points below 1,000 km provide
some upward flux corresponding to pitch-angles a > p/2. All these characteristics
are in agreement with the experiments.
So this modeling shows that the mechanism of ion acceleration by electromag-
netic ion-cyclotron waves generated by parametric instability of strong VLF waves
in the ionosphere is sufficiently effective to explain existing experimental data.
The VLF transmitter described in ref. [8] started to work several hours before
satellite arrival in the transmitter zone, and there was enough time for pumping the
accelerated particles into the wide range of geomagnetic field tubes (life time of
such accelerated particles is 104 s). Experimental data from COSMOS-1809
collected at orbit 1376 allow to analyze the development of ion acceleration process
in time and in altitudes along geomagnetic field lines since the transmitter located at
L ¼ 4 was switched on at 20.30 UT, when the satellite was practically on the same
L-shell (L ¼ 3.9). Thus, the accelerated ions whose velocities exceed the velocity of
satellite could reach the satellite only at higher L-shells moving to it from the
acceleration region, while the magnetic field tube is not completely filled by
114 V.M. Sorokin and V.M. Chmyrev

accelerated particles. Note that the lower energy (10–200 eV) accelerated ions
measured by COSMOS-1809 could be observed in a wider range of latitudes than
250–330 eV analyzed in ref. [8]. On the whole, the presented results of ion
measurements onboard the COSMOS-1809 satellite do not contradict the AUR-
EOL-3 observations [8].
Though the scientific instruments onboard the COSMOS – 1809 did not analyze
mass distribution of ions, we can assume that anomalous ion fluxes observed at
orbit 1376 have been accelerated in the upper ionosphere at altitudes below several
thousand kilometers. At the same effectiveness and location of acceleration mech-
anism, the O+ ion needs by a factor of 4 longer time to reach the satellite altitude
than the H+ ion. With account of the discussed data this means that the oxygen ions
could be observed by satellite only from narrower range of altitudes near the
satellite trajectory (or significantly later when satellite was far away from the
transmitter). Moving downward from the acceleration region at the sufficient
parallel velocity the H+ ions started at several thousand kilometers and the O+
ions from the lower altitudes could be able to “run down” the satellite on ascending
orbit at L > LP, where LP is the transmitter L-shell.
So the presented materials can be considered as experimental evidence of the
acceleration processes in the upper ionosphere initiated by electromagnetic radia-
tion from the powerful VLF transmitter. As a confirmation of this thesis, we can
mention the observation of 630 nm [OI] and Hb atmospheric emissions from the
“Intercosmos-Bulgaria 1300” satellite, which demonstrated the possibility of stim-
ulation (or intensification) of these emissions due to forming the fluxes of acceler-
ated ions and suprathermal electrons under the influence of intense VLF waves [9].

3.2.3.5 Artificial Stimulation of Geomagnetic Pulsations

In 1975 Frazer-Smith and Cole from Stanford University first paid attention on the
fact that powerful ground-based VLF transmitters influenced on the generation
regime of geomagnetic pulsations [162]. They have shown that the occurrence
rate and the intensity of Pc-1 oscillations increased when the VLF transmitter
located near the registration site operated.
Beginning from 1975 the Institute of Terrestrial Magnetism, Ionosphere and
Radio Wave Propagation of Russian Academy of Sciences (IZMIRAN) has carried
out a series of active experiments on the ionosphere modification by electromag-
netic radiation from VLF transmitters. Below we consider the results of these
experiments related to artificial stimulation of geomagnetic pulsations.
First of these experiments called “Juliana” has been carried out in February 1975
with use of Arkhangelsk VLF transmitter operated for this experiment at the fre-
quency 12.5 kHz in the mode of amplitude pulse modulation, 15 s ON–15 s OFF.
Module magnetic measurements were carried out at IZMIRAN expedition base 200
km away from the transmitter. Data reduction of the magnetic records with use of the
epoch superposition method revealed an excitation of the magnetic field oscillations
with the repetition period (30 s) of the transmitter signals and with amplitude 16 nT
3 Atmosphere–Ionosphere Electrodynamic Coupling 115

[163]. In pauses between the transmissions, no regular 30 s oscillations were observed.

This observation was the first direct experimental evidence of artificial generation of
geomagnetic pulsations under influence of VLF transmitter on the ionosphere and the
magnetosphere.
The experiments of 1987 were also carried out with use of Arkhangelsk VLF
transmitter. The radiated power was 300 kW, the carrier frequencies were 15 and
19.1 kHz, and the modulation regimes were 2 s ON–2 s OFF, 8 s ON–8 s OFF, and
50 s ON–10 s OFF. Measurements of geomagnetic pulsations were performed at
Lekhta station in Karelia (Russia) and at Oulu, Kevo, and Sodankyla stations in
Finland. In these experiments the ground-based geomagnetic observations were
accompanied by the measurements of ELF/VLF waves and particle fluxes onboard
the COSMOS-1809 satellite.
Figure 3.2 shows one of the episodes of artificial stimulation of geomagnetic
oscillations [164]. In this episode the transmitter was operating in the modulation
regime 8/8 during the night time in quiet geomagnetic conditions. Presented are the
wave forms of D-component of geomagnetic pulsations recorded in Lekhta in two
frequency channels (LF: 0.005–1.0 Hz and HF: 0.1–1.0 Hz) and the spectrogram of
H-component of geomagnetic pulsations observed in Kevo. The bar near 21.00 UT
in the lower part of Fig. 3.2 indicates the time interval when transmitter was active.
It is seen from Fig. 3.2 that the pulsation bursts of Pi-1 and Pi-2 types have
appeared 10–15 min after beginning of transmitter operation and terminated
approximately 15 min after transmitter switching off. Maximum amplitudes were
0.1 nT (HF channel) and 4 nT (LF channel) [164].
We already discussed the COSMOS-1809 satellite data on 9–205 eV ions observed
over the transmitter zone during this episode (see Fig. 3.1). When the ground
27 March 1987

5 nT Lekhta
D (0.005 – 1.0 Hz)

0.1 nT Lekhta D (0.1 – 1.0 Hz)

Kevo
Hz

1.0

UT 20.00 21.00 22.00

Fig. 3.2 Wave forms of the magnetic field oscillations (D – component) at Lekhta station and the
spectrogram of geomagnetic oscillations (H – component) observed at Kevo station 27 March 1987
116 V.M. Sorokin and V.M. Chmyrev

multistation measurements indicated the generation of geomagnetic oscillations, the

satellite observed significant enhancement of ion fluxes with energies 9–205 eV in
narrow zone  3 above transmitter, the maximum  4  104 cm2s1sr1eV1 was
at 20–70 eV [10]. This combination of simultaneous ground and satellite observations
has given a base for assumption that the generation of Pi-1 and Pi-2 geomagnetic
pulsations in this case was stimulated by the VLF transmitter [164].
The characteristic time delay between the commencement of VLF transmission
and the appearance of stimulated magnetic oscillations was about 10–15 min.
Owing to such conjunction in observations of geomagnetic pulsations and acceler-
ated ions it was natural to conclude that transmitter influence on the generation of
magnetic oscillations was performed by means of artificially accelerated iono-
spheric ions carried into the magnetosphere.
VLF transmitter-induced formation of conic ion distributions and transfer of
accelerated ionospheric ions (O+, He+, and H+) into the magnetosphere was
observed from the AUREOL-3 satellite [8] as we already discussed.
The considered experimental data have been used for the development of
theoretical model of artificially stimulated geomagnetic oscillations [160, 164].
According to this model transverse acceleration of ions takes place in the upper
ionosphere below 2,000 km, where the intensities of injected VLF waves and of
parametrically excited ELF emissions are sufficiently high, while the generation
(amplification) of geomagnetic oscillations by accelerated ions occurs in the mag-
netosphere (mainly in the near equator region), where the magnetic field and plasma
spatial gradients are small. The time delay between the commencement of VLF
wave injection and the appearance of geomagnetic pulsations is about the time
needed for accelerated ions to fill up the magnetic field tube in the acceleration
zone. The estimates made in refs. [160, 164] have shown that the ion fluxes
observed in the discussed experiments could provide the amplification of oblique
Alfven waves in the magnetosphere near L  4 with exponential growth rate 0.2
s1 for frequencies 5 Hz and 0.05 s1 for frequencies 0.5 Hz. At such growth
rates the stimulation of geomagnetic pulsations could be provided at the time scale
of the order of several minutes.

3.3 Electric Current and Field Perturbation

in the Atmosphere–Ionosphere Circuit

3.3.1 Equations Describing the Electric Potential

in the Ionosphere

The below-considered electrodynamic model of electromagnetic and plasma pre-

cursors to earthquake was formulated in refs. [21, 165] to explain some of the
observed phenomena by the electric field enhancement in the ionosphere over a
seismically active zone. The idea of such explanations arose from the observations
3 Atmosphere–Ionosphere Electrodynamic Coupling 117

of anomalous DC electric fields 10 mV/m in the low-latitude ionosphere over a

region shortly before an earthquake [45]. An existence of seismic-related electric
fields and their possible role in the lithosphere–ionosphere coupling were discussed
in refs. [38, 43, 166, 167]. The main conclusion made in these papers was that to
excite more or less significant (
1 mV/m) electric field in the ionosphere, the
vertical component of long-term DC electric field on the Earth’s surface should be
1–10 kV/m in the area 100 km. Such fields at the time scale about several days
have never been observed. Vershinin et al. [42] reported the observation of short-
time (1–2 h) bursts of the localized vertical electric field 500 V/m 8–10 h before
an earthquake. According to Buchachenko et al. [15], the meaningful ionospheric
precursors should exist at least for several days. The long-time electric field
variations observed before earthquakes have a magnitude 100 V/m or less [42].
So for proper modelling, we should find the mechanism providing the generation of
10 mV/m electric field in the ionosphere at slow variations of the vertical electric
field near the Earth’s surface with magnitude 100 V/m or below.
Such mechanism based on the electric field perturbation by the conductivity
currents in the atmosphere and the ionosphere was suggested in refs. [30, 168].
Their source was an external electric current generated by the injection of charged
aerosols into the atmosphere with soil gases, the upward transfer of these aerosols,
and their gravitational sedimentation with charge relaxation. Evidently the time
scales of the external current generation and the injection of soil gases into the
atmosphere should be similar.
Let us consider the generation of the electric field perturbation by an external
current je in the Earth–ionosphere layer. We will find a system of equations that
determine the potential ’ of the electric field perturbation E ¼ r  ’. Let us
introduce Cartesian coordinates ðx; y; zÞwith the z-axis directed vertically upward.
The homogeneous magnetic field B is directed at epy angle of a with respect to the
x-axis. The z ¼ 0 plane coincides with the ideally conducting Earth’s surface. We
assume that the electric field potential is zero on this plane, fjz¼0 ¼ 0. The layer
0 < z < z1 is the atmosphere, whose conductivity sðzÞ depends on the altitude z.
The z ¼ z1 plane coincides with the thin conducting ionosphere characterized by
the integral conductivity tensor with the components SP ; SH (the Pedersen and
Hall conductivities, respectively). According to refs. [30, 168], for slow processes
with characteristic times t  1=s, the distribution of the horizontal electric field
components in the ionosphere is derived by the equations:

Z1 Z1
1
Ex ðx; yÞ ¼ Kx ðx  x0 ; y  y0 Þ j1 ðx0 ; y0 Þ dx0 dy0;
4pSP
1 1
Z1 Z1
1 (3.6)
Ey ðx; yÞ ¼ Ky ðx  x0 ; y  y0 Þ j1 ðx0 ; y0 Þ dx0 dy0;
4pSP
1 1
x sin3 a y sin a
Kx ðx; yÞ ¼ 2 ; Ky ðx; yÞ ¼ :
x sin2 a þ y2 x2 sin2 a þ y2
118 V.M. Sorokin and V.M. Chmyrev

Equations (3.6) are used to calculate the spatial electric field distribution in the
ionosphere.

3.3.2 Electric Field Limitation on the Earth’s Surface

As mentioned above, the long-time ( several days) variations of the electric

field on the ground do not exceed 100 V/m. Such limitation can be connected
with the feedback between the vertical electric field perturbation and the causa-
tive external electric current near the Earth’s surface [30]. The feedback is
provided by the potential barrier at the Earth–atmosphere boundary that controls
the upward moving of charged aerosol particles through this boundary. The
movement occurs owing to viscosity of soil gases injected into the atmosphere.
For instance, if a positively charged particle emerges from the Earth into the
atmosphere, the Earth’s surface gets charged negatively, and the electric field that
appears, which is directed downward, impedes particle emergence to the surface.
At the same time, this field stimulates the emergence of negatively charged
particles onto the surface.
The vertical electric field component in the atmosphere Ez ¼ @’=@z is defined
by the following equation [30]:

Zz1 ,Zz1
1 h i je ðr; zÞ dz
Ez ðr; zÞ ¼ j1 ðrÞ  _je ðr; zÞ ; j1 ðrÞ ¼ dz : (3.7)
sðzÞ sðzÞ sðzÞ
0 0

We assume that an external current is formed as a result of the injection of

positively ( jp ) and negatively ( jn ) charged soil aerosols into the atmosphere,

je ðx; y; zÞ ¼ jp ðx; yÞsp ðzÞ  jn ðx; yÞsn ðzÞ; sp ðz ¼ 0Þ ¼ sn ðz ¼ 0Þ ¼ 1:

The sp ðzÞ and sn ðzÞ functions describe the altitude distribution of external
currents. Substituting the above equality into Eq. (3.7) yields an equation for the
vertical electric field component on the Earth’s surface, namely,

1 
Ez0 ðx; yÞ ¼ j1 ðx; yÞ  jp ðx; yÞ þ jn ðx; yÞ ;
s0
1 
j1 ðx; yÞ ¼ jp ðx; yÞkp  jn ðx; yÞkn :
r
(3.8)
Ez0 ðx; yÞ ¼ Ez ðx; y; z ¼ 0Þ; s0 ¼ sðz ¼ 0Þ;
Zz1 Z z1
sp;n ðzÞ dz
kp;n ¼ dz ; r¼
sðzÞ 0 sðzÞ
0
3 Atmosphere–Ionosphere Electrodynamic Coupling 119

External currents of positively and negatively charged aerosols depend on the

vertical electric field component on the Earth’s surface as required by the feedback
mechanism,
  
jp ðx; yÞ ¼ jp0 ðx; yÞf Ez0 ðx; yÞ Ecp ; jn ðx; yÞ ¼ jn0 ðx; yÞ f ðEz0 ðx; yÞ=Ecn Þ;

where jp0 ðx; yÞ; jn0 ðx; yÞ are determined by the intensity of charged aerosols injec-
tion in the absence of an electric field. When the negative field reaches some critical
value Ecp , it blocks the flow of positively charged particles. Accordingly, a positive
field blocks the flow of negatively charged particles. The critical field can be
estimated in the order of magnitude from the balance equation for viscous, gravity,
and electrostatic forces,

Ecp ¼ ð6 pRp V  mp gÞ=eZp ; Ecn ¼ ð6 pRn V  mn gÞ=eZn ;

where  is the viscosity of air, V is the velocity of upward movement of soil gases in
earth, Rp;n is the aerosol particle radius, mp;n ¼ ð4=3Þ pR3p;n m is the particle mass,
and m is the particle density. The viscous force of soil gases that rise in the Earth
acts on a particle upward. The gravity force is directed downward. The electrostatic
force, which appears as a result of the emergence of a positively charged particle
onto the surface, is directed downward. For simplicity, we assume that positively
and negatively charged aerosols have equal sizes and masses, Ecp ¼ Ecn ¼ Ec . To
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi let us specify the functional dependence f on the electric field
performpcalculations,
as f ¼ 1 þ Ez0 =Ec . Using this dependence in Eq. (3.8) yields

"
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi#

1 kp Ez0 ðx; yÞ kn Ez0 ðx; yÞ
Ez0 ðx; yÞ ¼ jp0 ðx;yÞ 1 1þ  jn0 ðx; yÞ 1 1 :
s0 r Ec r Ec
(3.9)

Given jp0; jn0 , this equation allows us to calculate the vertical electric field
component on the Earth’s surface. After solving Eq. (3.9), we can determine the
horizontal distribution of the conductivity current flowing from the atmosphere into
the ionosphere. It follows from Eqs. (3.8) and (3.9) that

" sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi #

1 Ez0 ðx; yÞ Ez0 ðx; yÞ
j1 ðx; yÞ ¼ jp0 ðx; yÞ 1 þ kp  jn0 ðx; yÞ 1  kn : (3.10)
r Ec Ec

Let us assume  ¼ 1.72  107 kg/(cms), V ¼ 104 m/s, R ¼ 5  107 m, m ¼

1.5  103 kg/m3, and Z ¼ 100. Then the estimates of the critical field give Ec ¼
450 V/m and s0 Ec ¼ 10 pA/m2. It means that the vertical electric field on the
Earth’s surface cannot exceed the value Ezm ¼ 90 V/m.
120 V.M. Sorokin and V.M. Chmyrev

3.3.3 Mechanisms of External Electric Current Generation

in the Lower Atmosphere

Various mechanisms can be responsible for the formation of external currents in the
near-ground atmospheric layer [169, 170]. One of the most effective mechanisms is
associated with intensification of charged soil aerosol injection into the atmosphere
or with changes in meteorological conditions at a stable altitude distribution of
aerosols. It was experimentally shown [23–27] that, several days before an earth-
quake, the concentration of metal ion-containing soil aerosols in the atmosphere
increases by one or two orders of magnitude. The quasi-stationary aerosol distribu-
tion is formed as a result of turbulent upward transfer and gravitational sedimenta-
tion. Turbulent transfer occurs due to two main reasons. First is connected with
vertical gradient of horizontal wind velocity and transformation of wind kinetic
energy into the energy of turbulent pulsations. Second is caused by the thermal
instability of the atmosphere arising when negative temperature gradient exceeds its
adiabatic gradient. Turbulent vortices transfer aerosols from the altitudes, where
their concentration N is high to the altitudes of a lower concentration. Equilibrium
is attained when the vertical flux of aerosols is balanced by their gravitational
sedimentation at the rate w.
For describing the dynamics of particles determined by stochastic differential
equations we will use the probability distribution function f ðq; z; tÞ of aerosols,
which has the meaning of a probability that a particle has the charge q at the time
t and the altitude z [21]. If transport coefficient K weakly depends on the altitude,
the kinetic equation for the f ðq; z; tÞ distribution function takes the form [21]:

@f @f @ @2f
 w  4 psðzÞ ðqf Þ ¼ K 2 :
@t @z @q @z

The space–time distributions of aerosol concentration, Nðz; tÞ, their electric

charge density, re ðz; tÞ, and the density of external current connected with their
motion, je ðz; tÞ, can be written in terms of the moments of the f ðq; z; tÞ distribution
function. The equation for the moments has a form

@N @N @2N @re @je

w ¼K 2 ; þ 4 psðzÞre ¼  :
@t @z @z @t @z

The last equality describes the density of external charge and current in a
conducting medium. Changes in the number of external charges in a distinguished
volume are determined by two processes. First, this is electromotive force-induced
transfer through the surface that bounds this volume. Secondly, there is a decrease
in the external charge caused by its relaxation in the environment with conductivity
s. For instance, if the total flux of external charges through the surface binding the
volume becomes zero r  je ¼ 0, the number of these charges in the volume
decreases according to the law re  exp(4pst). The relaxation time is 1/4ps.
3 Atmosphere–Ionosphere Electrodynamic Coupling 121

If the process is fairly fast (t  1=4ps), charges fail to relax, and the continuity
condition takes the form @re =@t þ r  je ¼ 0. Otherwise, if external charges are
formed at a low rate (t  1=4ps), the external charge density is connected with the
external current as 4psre þ r  je ¼ 0. This equality means that, in the stationary
state, the disappearance of external charges in the volume due to relaxation is
balanced by their transfer through the boundary surface. We assume that the
characteristic time of the processes under consideration is longer than the relaxation
time 1=4 ps. Equation for the altitude distribution of external current in the quasi-
stationary approximation can be found from the above consideration in the follow-
ing form:


@ 1 @je ðz; tÞ w @je ðz; tÞ je ðz; tÞ
þ  ¼ 0:
@z 4psðzÞ @z 4psðzÞK @z K

An influence of variations in conductivity and radioactivity in the lower atmo-

sphere on DC electric field over a seismic region was investigated by Sorokin
et al. [171]. This effect is associated with the occurrence of ionization source due
to seismic-related emanation of radon and other radioactive elements into the atmo-
sphere. The natural radioactivity of the lower atmosphere is mainly associated with
such elements as radon, radium, thorium, actinium, and their decay products. Radio-
active elements enter the atmosphere together with soil gas, and then they are
transferred by the air streams upwards up to the altitude of a few kilometers.
An increase in the level of atmospheric radioactivity, for example, prior to an
earthquake, leads to an increase in the ion production rate q. The number density of
light ions and the mean charge of aerosols are determined by recombination of ions
and their adhesion to aerosols at given ionization sources. The vertical distribution of
ion production rate is formed as a result of atmospheric absorption of gamma
radiation and alpha particles from the decay of radioactive elements in the atmo-
sphere. The total ion production rate in the lower atmosphere is formed by two
sources – the cosmic rays and the atmospheric radioactivity. To calculate the external
current and the atmosphere conductivity, it is necessary to find an equilibrium ion
number density depending on the ion formation rate. The equilibrium value of ion
number density is defined by their recombination in the air and adhesion to aerosols.
The perturbations of conductivity and external electric current as a function of
altitude have been calculated from the following equations [171]:


d 1 djp;n 1 djp;n jp;n
Hp;n þ  ¼ 0; s ¼ 2emn;
dz s dz s dz 2e0 vp;n
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 rffiffiffi
em Np Nn q
n ¼ N þ n0  N; N ¼
2 2 þ ; n0 ¼
4e0 a lp ln a

where s ¼ 2emn is the atmospheric conductivity, n0 is the ion number density in the
clean atmosphere without aerosols, Np;n are the positive and negative aerosol number
densities, a is the recombination constant, and m is the light ion mobility in the
atmosphere. Rapid growth of the conductivity occurs in the near surface layer.
122 V.M. Sorokin and V.M. Chmyrev

At altitudes up to 6 km, the conductivity increases with the growth of the radioactivity
level. The aerosol number density increase leads to decrease of conductivity due to
loss of the light ions caused by their adhesion to aerosols. The calculations showed
significant current densities at the altitude up to 12–14 km with maximum at 1–2 km.
External current decreases with growth of atmospheric radioactivity.

3.3.4 Results of the Electric Field Calculation

Let us consider the electric field generation by external currents associated with the
charged aerosol dynamics in the lower atmosphere. We accept the large-scale
axially symmetrical external current distribution. Such distribution corresponds to
vertical transport of aqueous aerosols in typhoon regions [30]. The horizontal
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
electric field component in the ionosphere Er ðr; ’Þ ¼ E2x þ E2y is calculated for
various ’ ¼ arctanðy=xÞ angles, ’ ¼ 0 and corresponds to the magnetic meridian
plane. The external current is induced by vertical atmospheric convection,
which acts as an electrostatic generator. Air moving upward transfers small
positively charged particles, whereas gravitational sedimentation transfers negative
charges downward. The index of charge separation in unit cloud volume is

dQ dt  1 C/km3 min  1011 C/m3 s. Perhaps, the vertical convective movements
in typhoon regions are characterized by smaller indices; their values are not exactly
known. We assume that, at altitudes of z0 ¼ 10 km, the mean charge separation
index is of 4  1012 C/m3 s. The estimates then give the current density je0 
(dQ/dt)z0  4  106 A/m2. It was found that the electric field component in the
magnetic meridian plane is much smaller than the component in the perpendicular
plane. The distribution substantially depends on the field tilt angle a. The spatial
field structure has two maxima with a very small component in the magnetic
meridian plane (in the center of a typhoon) for small (<20 ) field tilt angle [30].
Seismic and volcanic activities initiate the enhanced injection of charged
aerosols with soil gases into the atmosphere. These aerosol fluxes as we discussed
above generate the external electric currents and related electric field perturba-
tions. Sorokin et al. [168] have calculated the structure of such perturbations for
the seismic event with axially symmetric horizontal distribution of external
currents 100 km in radius. Figure 3.3 presents the distributions of a vertical
component of the electric field on the Earth’s surface and of the horizontal
electric field component in the ionosphere. It is seen from this figure that the
ionospheric field can reach the values up to 10 mV/m, whereas the vertical field
component on the Earth’s surface does not exceed 100 V/m. This is good
illustration of the feedback mechanism for limitation of the ground electric field
considered in Section 3.3.2.
The field in the ionosphere reaches a maximum at the edge of an area covered by an
external current. The vertical electric field enhancement on the Earth’s surface occurs
in the region, which is approximately three times larger than the horizontal scale of the
external current. Within this region, the field virtually does not vary with distance.
3 Atmosphere–Ionosphere Electrodynamic Coupling 123

a
Er , mV/m
10

500
300
100
–100
500
x, km 300
–300 100
–100
b –500 –300 y, km
–500
Ez, V/m

40

20

500

300

100
x, km
–100
500
300
–300 100
–100
–300 y, km
–500
–500

Fig. 3.3 Spatial distributions of the (a) horizontal electric field component in the ionosphere and
(b) vertical field component on the Earth’s surface

3.4 Plasma and Electromagnetic Effects of Seismic-Related

Electric Field in the Ionosphere

3.4.1 Instability of Acoustic-Gravity Waves and the

Formation of Horizontal Conductivity Inhomogeneities
in the Lower Ionosphere

Growth of the electric field leads to the instability of acoustic-gravity waves in the
ionosphere [172, 173]. The instability is connected with the transformation of Joule
heat of the ionosphere currents into the wave energy. According to [174], the
124 V.M. Sorokin and V.M. Chmyrev

ionosphere can be treated as a continuous medium with tensor conductivity in

the low-frequency approximation. The propagation of acoustic-gravity waves in
this medium is accompanied by the perturbation of conductivity and, therefore, of
currents. Under certain conditions, these perturbations are sufficiently strong for
Joule heat of the perturbed currents to increase an amplitude of acoustic-gravity
waves. The source of energy for this instability is the electromotive force of
external electric field. Field energy transforms into the energy of waves without
disturbing the heat balance of the medium. Let us analyze the stability of acoustic-
gravity waves in the presence of an external electric field on the assumption that the
magnetic field B is directed along the z-axis, and the electric field E, along the
x-axis. We shall use the equations of motion, continuity, conservation, and ideal gas
state in the gravitational field g. The gas is characterized by the velocity v, density
r, pressure p, and temperature T. We shall also use the Ohm’s law for the current
density j in the ionosphere.
Let us consider the horizontal propagation of a plane wave along the x-axis in a
homogeneous medium, when the free fall acceleration can be neglected. We
assume that the unknown values depend on the coordinates and time as
expðiot þ ikxÞ. The dispersion equation then takes the form [172]:

oðo þ iom Þ½o þ iðom  o2 Þ

k2 ¼ ;
a2 ½o þ ðom þ o1 Þ
o1 ¼ ð2a þ 1Þðg  1ÞsP0 E2 =2a2 r0 ; o2 ¼ gðg  1ÞsP0 E2 =2a2 r0 ;
om ¼ sP0 B2 =c2 r0 ; a2 ¼ gRT0

where g is the ratio between the specific heat capacities, R is the universal gas
constant, and sP0 is the Pedersen conductivity of the unperturbed ionosphere. The a
coefficient characterizes the ratio between the relative changes in density of ions
and gas density in the wave. The time dependence of perturbation is defined by
complex frequencies. If o ¼ o0 þ iG, where G  o0 , one obtains o0 ¼ ak and
G ¼ ðom  o1  o2 Þ=2. If om > o1 þ o2 , then G < 0, and the wave turns
damp. At o1 ¼ o2 ¼ 0, the damping of the wave is determined by the parameter
om ¼ sP0 B2 =c2 r0 , which characterizes induction deceleration. If om < o1 þ o2 ,
we have G > 0, that corresponds to instability. The equality om ¼ o1 þ o2 defines
the critical field value,

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
aB 2
Ek ¼ :
c ðg  1Þð2a þ g þ 1Þ

If the field is lower than critical, the initial perturbation fades out, and if it is
higher, the wave amplification occurs. Estimation gives Ek ¼ð7  8ÞmV/m. For the
estimates we used a ¼ 3  102 m/s; B ¼ 0.3Oe; c ¼ 3  108 m/s; g ¼ 1:4; a ¼ 2.
3 Atmosphere–Ionosphere Electrodynamic Coupling 125

Let us consider the horizontal propagation of an acoustic-gravity wave (AGW)

along the x-axis on the assumption that @=@z ¼ 0. The dispersion equation then
takes the form [172]:

2 
ak ðo þ iom Þ o2 ½o þ iðom  o2 Þ  ðo þ iom Þo2a þ io2 o23
¼  ;
o o o2 ½o þ iðom þ o1 Þ  ðo þ iom Þo2g

where o2a ¼ gg=4H is the boundary acoustic frequency,

 o2g ¼ ðg  1Þg=gH is the
Brunt–Vaisala frequency, and o23 ¼ o2a ð2a þ 3Þ g. Let us introduce the complex
refractive index n þ ik ¼ ka=o. The n ¼ nðoÞ and k ¼ kðoÞ dependences have
been calculated from this equation in the vicinity of og . It appeared that the absorp-
tion coefficient is negative and has a maximum at the frequencies o  og. This means
the instability regime, which provides the exponential growth of acoustic-gravity
waves at o  og and produces the periodic plasma structure. Along with density and
pressure oscillations, the conductivity oscillations occur in the wave. Thus the AGW
instability leads to formation of periodic horizontal inhomogeneities of the iono-
sphere conductivity with the characteristic scale l  l/2, where l is the wavelength at
o  og, where the  AGW
 growth rate is maximum. At these frequencies, the
refractive index n og reaches a maximum
  and the phase velocity becomes less
than the velocity of sound, vg ¼ a n og < a. The horizontal inhomogeneity scale is

l ¼ l=2 ¼ pvg =og ¼ pa=og nðog Þ: (3.11)

Thus the enhancement of DC electric field causes the instability of acoustic-

gravity waves and generation of horizontal conductivity inhomogenities in the
lower ionosphere.

3.4.2 The Formation of Longitudinal Currents and Plasma

Inhomogeneities in the Upper Ionosphere
and the Magnetosphere

Horizontal conductivity inhomogeneities in the lower ionosphere change the struc-

ture of the electric fields and form the magnetic field-aligned plasma layers in the
upper ionosphere [172, 173, 175]. A high conductivity along magnetic field lines
results in electric field propagation into the upper ionosphere and the magneto-
sphere. This causes the formation of an electric circuit, which includes the longitu-
dinal currents in the magnetosphere that transfer the electric field along
geomagnetic field lines and the transverse closure currents provided by Pedersen
conductivity in the ionosphere. Note that the longitudinal currents are carried by
electrons, whereas ions are transverse current carriers. Therefore, the electric field
propagation along geomagnetic field lines and the formation of closure currents
should be accompanied by local changes in plasma density. Thus, the alternation of
ionospheric E-layer conductivity in the presence of an external electric field results
126 V.M. Sorokin and V.M. Chmyrev

in the appearance of a polarization electric field. The upward propagation of this

field causes the changes in plasma density structure in the upper ionosphere. Let us
estimate this effect.
Although an acoustic-gravity wave propagates along the x-axis, the conductivity
inhomogeneities associated with AGW instability are extended along the y-axis.
We assume that the electric field E0 lies in the (x, y) plane and the magnetic field B
is directed along the z-axis. When a conductivity inhomogeneity with the amplitude
DsP;H moves in the E-layer of the ionosphere as an isolated band of width l  l/2
extending along the y-axis, the plasma inhomogeneity extending along the mag-
 [176]. The band moves along the x-axis
netic field is formed in the upper ionosphere

with the AGW velocity vg ¼ a n og ; o1 . The magnetosphere  is characterized by
the Alfven velocity va or wave integral conductivity Sw ¼ c2 4pva . The formation
of the conducting band in E-layer results in the occurrence of a polarization electric
field DE, which is transferred along geomagnetic field lines into the upper iono-
sphere and changes the plasma density in this region. The relative magnitude of this
change can be found from the equation [172]

N 1 þ D1  vg vex
¼  ;
N0 1 þ D2  vg vex
ni Ex0 ni ðSP0 þ Sw ÞEx0 þ ðSH  SH0 ÞEy0
D1 ¼ ; D2 ¼ ;
oi Ey0 oi ðSP þ Sw ÞEy0

where N andN0 are the equilibrium densities of ions inside and outside the band,
vex ¼ cEy 0 B is the particle drift velocity, vg is the horizontal velocity of band
movement along the x-axis, SP and SH are the integral conductivities of the lower
ionosphere, and ni is the ion collisions frequency. We assume that the electric field
is directed along the x-axis (Ey0 ¼ 0). The equality SP ¼ Sw holds  to a fairly high
accuracy in the ionosphere. If DSP =SP0 ¼ DsP =sP0 and ni cEx0 oi vg B  1, the
change in plasma density DN ¼ N  N0 can be estimated from the equation

DN DsP ni cEx0 ni cnðog ÞEx0

¼  :
N0 sP0 oi vg Bð2 þ DsP =sP0 Þ 2oi aB

The altitude dependence of DN=N0 is described by the function ni ¼ ni ðzÞ. Thus

the appearance of horizontal inhomogeneities in the lower ionosphere conductivity
results in the formation of plasma layers extended along the geomagnetic field. The
transverse size of these layers coincides with the scale of horizontal spatial structure
of conductivity.

3.4.3 Electromagnetic Perturbations in ULF/ELF Ranges

As mentioned above, satellite data on intensification of extremely low-frequency

(ELF) radiation over earthquake region were reported repeatedly. Several
3 Atmosphere–Ionosphere Electrodynamic Coupling 127

mechanisms of the generation of such radiation have been discussed during the past
years (e.g., see [177]). Calculations showed that these mechanisms produced much
weaker effects than those observed experimentally over the spectral range covering
several hundred hertz. A new mechanism based on the transformation of atmo-
spheric ELF noise on the small-scale inhomogeneities in the ionosphere was
proposed by [36]. This pulse electromagnetic noise is generated by lightning and
always exists in the Earth–ionosphere waveguide. The excitation of small-scale
plasma inhomogeneities in the ionosphere before earthquake was experimentally
confirmed [32] and explained by AGW instability in the presence of an overcritical
DC electric field [172, 173].
Let us consider the mechanism suggested by Borisov et al. [36]. The lowest
eigenmode of subionosphere waveguide (the TM mode) has the weakest attenuation
at frequencies below 1 kHz and can therefore propagate through large distances.
Because of the high conductivity of the Earth near the surface, the electric field of this
mode is directed vertically. A horizontal electric field component appears as the
altitude increases. Its value approaches the amplitude of the vertical component over
the spectral range 100–1,000 Hz at altitudes of 115–120 km, at which the conductiv-
ity of the ionosphere is maximum. Horizontal components of the electric field pulses
from lightning discharges induce polarization currents on the conductivity inhomo-
geneities. The radiation from these currents, which depends on frequency, propagates
in a whistler mode upward into the upper ionosphere and the magnetosphere. Satellite
should observe this radiation at the same geomagnetic field lines where plasma
density inhomogenities are observed. The spectral characteristics and the intensity
of this radiation were found to be near the experimentally observed values [32].
Generation of small-scale plasma density irregularities in the ionosphere over
seismic zone, and the effects of these irregularities upon characteristics of very low-
frequency transmitter signals propagated through these disturbances and then
registered onboard a satellite were analyzed in [178]. The main effect is in observ-
able spectral broadening of signals. The calculations showed two characteristic
spatial scales of plasma density irregularities across the magnetic field. The first is
4–40 km, which has been confirmed by satellite observations, and the second is of
the order or less than 100 m [18]. These smaller-size irregularities produce notice-
able effect in very low-frequency signal spectral broadening, which is most pro-
nounced when the transmitter frequency is above, but close to the local low-hybrid
resonance frequency in the region where the small scale irregularities are present,
which in turn sets the requirement that the transmitter frequency be in the range
from 10 to 20 kHz. This corresponds to operational band of most VLF transmitters.
For the 100 m irregularities, we get the spectral broadening 100 Hz that can easily
be registered by simple very low-frequency receiver onboard a satellite, provided
the transmitter power is high enough. This effect together with the direct satellite
measurements of plasma density variations can be used as an effective tool for
diagnostics of seismic-related ionospheric disturbances and therefore considered as
a possible ionospheric precursor to earthquake.
The observations of ultra low-frequency electromagnetic noises in the
vicinity of a forthcoming earthquake area were reported in several papers
128 V.M. Sorokin and V.M. Chmyrev

(e.g., see [177, 179–181]). In particular, Frazer-Smith in ref. [179] have observed
ULF oscillations in the frequency band 0.01–5 Hz approximately 10 days before
a strong earthquake at a distance 50 km from the epicentre. Fitterman [56],
Molchanov et al. [55], and Pilipenko et al. [57] considered a source of this
phenomenon to be located in the lithosphere. An alternative ionospheric mecha-
nism for the generation of ULF magnetic field oscillations based on excitation of
gyrotropic waves (GW) was presented in refs. [182, 183].
Gyrotropic waves earlier observed by Sorokin and Fedorovich [59] propagate
within a thin layer of the lower ionosphere in low and medium latitudes with weak
attenuation at phase velocities from tens to hundreds of kilometers per second.
Some geophysical effects related to the generation and propagation of GW in the
horizontally homogeneous ionosphere were considered in refs. [184, 185]. A
comprehensive theory of these waves in the mid-latitude ionosphere was developed
by Sorokin and Pokhotelov [186].
Let us consider the generation mechanism for gyrotropic waves suggested in ref.
[183] where the key role, as for ELF radiation, belongs to ionospheric inhomogene-
ities. This mechanism involves the generation of GW in the lower ionosphere by the
interaction of electromagnetic noises with periodic horizontal inhomogeneities of
electrical conductivity excited by the acoustic-gravity wave instability in the
presence of an overcritical DC electric field. Polarization current excited on such
periodic structure forms the distributed source of gyrotropic waves with horizontal
spatial scale 10 km. These waves propagated along the ionospheric E-layer
produce the magnetic field oscillations that can be observed on the ground in the
frequency band 1–10 Hz.
The ground-based measurements yielded the detection of discrete narrow-band
spectra of ULF magnetic field oscillations during seismic enhancements, volcanic
eruptions, and space shuttle launches and landing [187]. It was found that
the spectrum maxima in these oscillations are located at separate frequencies of
2, 6, 11, and 17 Hz. An attempt to explain such discrete structure in terms of
gyrotropic waves was undertaken by Sorokin and Hayakawa [188] who have
analyzed the generation and propagation of GW in the conducting layer of a finite
thickness in the lower ionosphere in the presence of conductivity inhomogeneities.
Discrete spectrum containing several lines was obtained; the position of lines
depended on the thickness of the layer with Hall conductivity. The width of those
spectral lines was defined by width of a spatial spectrum of ionosphere irregularities
and the ratio between Pedersen and Hall conductivities. The attenuation was
determined by Pedersen conductivity and a number of frequency lines depended
on the dimension of source region. For instance, if conductivity irregularities with a
size of 80 km in horizontal direction are distributed within the ionospheric region
over seismic area, then the spectrum of excited pulsations consists of six spectral
lines in the frequency range 1–30 Hz. Amplitude of magnetic oscillations is defined
by the intensity of ionosphere irregularities, their spatial structure, and by the wave
attenuation.
Irregularities of the ionosphere conductivity can influence on the propagation of
ULF waves excited in the magnetosphere through the ionosphere. In ref. [189],
3 Atmosphere–Ionosphere Electrodynamic Coupling 129

Sorokin et al. have shown that such irregularities could produce the screening effect
on the geomagnetic pulsations penetrating through the ionosphere and decrease the
amplitude of the oscillations observed on the Earth’s surface. This effect should
take place during night time and be more significant for higher frequencies [189].

3.4.4 Perturbation of the Ionosphere

The heat flux emitted by a thin conducting layer of the ionosphere in the horizontal
electric field is (104  103) W/m2. One of the main sources of the ionosphere
heating is short-wave solar radiation ðl < 102:6 nmÞ. The supply of heat caused by
the absorption of this radiation at altitudes above 100 km is approximately several,

103 W m2 . Depending on the solar cycle, it changes several times in either
direction. An estimate made in ref. [190] shows that the Joule heat of ionospheric
currents over the earthquake preparation region constitutes a substantial fraction of
the total heat balance of the ionosphere. Evidently this source of heat has determin-
ing action on the state of the ionosphere. Heating the ionosphere by currents
increases the scale of altitude distribution of ionospheric components and, there-
fore, the altitude profile of F2 layer. Apart from the other possible mechanisms, this
heating mechanism should contribute to the observed ionosphere response on the
earthquake preparation processes [190]. The ionosphere is isothermal at altitudes
exceeding 200 km, and there is a positive temperature gradient over the altitude
range 100–200 km. Owing to heat conductivity, the presence of the temperature
gradient generates a heat flux directed downward. The Joule heat source is localized
in the lower ionosphere at 120–150 km. The upper ionosphere heating can then
occur only if gases move in the vertical direction. Changes in the altitude distribu-
tion of ionization in the ionosphere at a given spatial inhomogeneity of the electric
field on the ground were considered in ref. [191]. They assumed that the ionosphere
modification was caused by plasma drift.
An increase of the electric field and current in the lower ionosphere over a
seismic region leads to additional release of Joule heat that produces additional
vertical flux of neutral particles and changes the ionosphere temperature [190].
Such mechanism of effecting the formation of F2 layer is realized through colli-
sions between ions and neutral particles and an increase in the vertical scale of their
altitude distribution. The modelling of seismic-related modification of the iono-
sphere F region performed by Sorokin and Chmyrev [190] with use of some
simplifying assumptions [192] showed that heat release in the lower ionosphere
increases the height of F layer maximum and decreases the concentration of
electrons at this height. The altitude profiles of ion densities are also modified
towards the increase in density above F layer maximum.
Along with the rearrangement of the altitude profile of plasma density in the
upper ionosphere above a seismically active region, the formation of sporadic
layers in the lower ionosphere has been observed [193, 194]. The critical frequency
130 V.M. Sorokin and V.M. Chmyrev

of the sporadic E layer, f0 Es , reached 8  9 MHz in daytime. This value corre-

sponds to a number density of electrons 106 cm3 . High-altitude rocket measure-
ments in the medium-latitude ionosphere showed that electron number density
in the sporadic layer was 2  105 cm3 , and the electric field in this layer reached
10 mV/m [195]. A model of the formation of an anomalous sporadic E layer based
on the effect of atmospheric electric current flowing into the ionosphere was
considered in ref. [196]. This current initiates the occurrence of a longitudinal
current, which flows from the ionosphere into the magnetosphere, and transverse
closure current, which flows in the conducting ionosphere layer. The atmospheric
current transfers ions into the ionosphere, and the longitudinal current, which
balances electric charge, transfers electrons into the ionosphere. As a result, plasma
density in the lower ionosphere increases [196]. The presence of metal ions in the
lower ionosphere causes the appearance of a thin (3 km) layer of enhanced
electron concentration when electric current flows into the ionosphere. This layer
could be recorded as a sporadic E layer.

3.4.5 Electrodynamic Model of the Atmosphere

and the Ionosphere Coupling

According to the electrodynamic model [21], an enhancement of a DC electric

field in the ionosphere can be caused by an increase in conductivity of the lower
atmosphere due to the intense outbursts of radioactive substance from a seismic
source at the preparatory stage of earthquake. Such outbursts were repeatedly
observed several days or weeks before earthquake. The time dependence of radon
concentration in soil gases and in natural water was investigated by Virk and
Singh [25]. An intense radon burst was registered approximately a week before
the earthquake. The concentration of radon then increased by factors of 2.5 and
above 1.5 in gases and water, respectively. Measurements were taken at a distance
of 300 km from the epicenter. According to ref. [26], the concentration of radon
increased approximately fourfold 5 days prior to an earthquake. The statistical
analysis of the data obtained during 3 years of observations for about 300 micro
earthquakes (M < 4) in south-east Germany shows that 75% of the earthquakes
were preceded by a substantial increase in the concentration of radon. It can be
suggested that, simultaneously with radon, other radioactive elements are injected
into the atmosphere. In addition, such injections should not necessarily be related
to radon. The seismic-related increase in the concentration of soil aerosols
containing metal ions in the atmosphere was reported in several papers (e.g., see
[23, 43].
According to the considered model [21] these injection processes are responsible
for the formation of external currents and enhancement of electrical conductivity in
the lower atmosphere that causes an increase in the electric field in the ionosphere,
as it is shown in Fig. 3.4.
3 Atmosphere–Ionosphere Electrodynamic Coupling 131

Such approach finds support in the data obtained in [197] after the Chernobyl
accident. It was found that intense injections of radioactive substances into the
atmosphere were accompanied by changes in the phase and amplitude of VLF
signals along the propagation path crossing the accident area. The VLF signal
alternations were connected with the localized lower ionosphere perturbation,
which could be caused by the electric field growth in this region [198].
According to modeling [21], changes in the electrophysical parameters of the
lower atmosphere lead to amplification of the electric field in the ionosphere and
stimulate several effects observed experimentally. As an example, we can mention
the satellite observations reported in refs. [32, 40]. The data on ULF magnetic field
oscillations over the frequency range (1–8) Hz and on the vertical DC electric field
component obtained onboard the “Intercosmos-Bulgaria 1300” satellite 15 min
before the earthquake of 01.12.1982 were discussed in ref. [40]. An increase in
the electric field  (3–7) mV/m and the appearance of ULF magnetic field oscilla-
tions with an amplitude of 3 nT were observed in two regions: above the epicenter
and in the magnetically conjugated zone. COSMOS-1809 satellite measurements
over Spitak earthquake region showed that intense ELF radiation was generated
over the longitude range 6 and latitude range 24 about the epicenter [32]. The
intensity of radiation was about 10 pT at the frequencies 140 Hz (in the band

Z
B

4 2

Z1

5
3

6 7 8 1

Fig. 3.4 Scheme of the model used for calculations of electric field in the atmosphere–ionosphere
circuit. 1 – Earth’s surface, 2 – ionosphere, 3 – conducting current in the atmosphere and
the ionosphere, 4 – field aligned current, 5 – zone of upward convection of charged aerosols
and external electric current formation, 6 – zone of perturbation of atmospheric conductivity
induced by radioactive elements emanation, 7 – charged aerosols moving upward with soil gases,
8 – radioactive elements emanation
132 V.M. Sorokin and V.M. Chmyrev

25 Hz) and about 3 pT at 450 Hz(in the band 75 Hz). In the same region, small-
scale (4–10 km along the orbit) plasma density inhomogeneities with the relative
amplitude (37)% were observed. An existence of plasma density perturbations
above seismically active regions was confirmed in ref. [45].
As discussed, an increase in the ionosphere electric field above the earthquake
preparation region results in the instability of acoustic-gravity waves [172, 173].
The development of this instability is accompanied by the generation of horizontal
inhomogeneities in the ionosphere conductivity, whose interaction with the electric
field in the ionosphere works as a source for the radiation of guided Alfven waves
into the magnetosphere. These waves form the field-aligned electric currents and
plasma layers extended along the geomagnetic field. The transverse spatial size of
these layers coincides with the scale of the horizontal spatial structure of conduc-
tivity. Let us consider numerical estimates. At an altitude of 1,000 km, the
summed frequency of ion collisions with ions and molecules is ni  0.5 s1
[199], and the gyrofrequency of ions is oi  30 s1. For the velocity of sound
a ¼ 3 102 m=c, perturbed electric field value E  9mV=m, refractive index n 
(1  10 and og  2  102 s1 , we obtain:

l ¼ pa/ogn(og)  (4  40) km; DN/N  (1.6  16)%.

When a satellite moving at the velocity vs  8 km=s crosses plasma inhomo-

geneities of horizontal scale l  (4  40) km, plasma density fluctuations with the
period Dt ¼ l/vs ¼ pa/vs ogn(og)  (0.4  4) s should be observed. Since these
inhomogeneities are formed by field-aligned currents, geomagnetic oscillations
with the same period should be recorded when satellite crosses these currents.
Their amplitude is found to be b 5 nT. This estimate is in agreement with the
satellite measurements. A scheme of such measurements is illustrated by Fig. 3.5.
Geomagnetic oscillations observed on the Earth’s surface before earthquakes are
interpreted in frame of the discussed model [21] in terms of gyrotropic waves
(GW). These waves are connected with polarization currents associated with the
horizontal inhomogeneities of conductivity and their interaction with the back-
ground electric field.
This interaction generates GW, which propagate in the horizontal direction and
form geomagnetic oscillations in ULF range on the Earth’s surface [182, 183]. The
normalized spectrum of geomagnetic oscillations has the intensity maximum
40% of the unperturbed (background) value and lies in ULF range. The spectral
maximum frequency decreases as the horizontal scale of inhomogeneities increase.
The amplitude of perturbation decreases with a distance from epicenter due to
attenuation of gyrotropic waves.
According to a mechanism of electromagnetic ELF precursors to earthquakes [36],
the enhanced ELF emission in the upper ionosphere arises when pulsed electromag-
netic radiation from lightning discharges propagating in the Earth–ionosphere
waveguide is scattered by lower ionosphere conductivity inhomogeneities over
earthquake region. Intensities of lightning-induced whistler mode waves in the
upper ionosphere vary substantially depending on lightning activity. The modeling
3 Atmosphere–Ionosphere Electrodynamic Coupling 133

a
By, nT
10

0
3
–10
4

b 2
DN/N

10%
1

60 c

Fig. 3.5 Scheme of satellite observations of (a) ultralow-frequency geomagnetic field oscillations
and (b) plasma density inhomogeneities. 1 – Horizontal ionosphere conductivity inhomogeneities,
2 – plasma density inhomogeneities stretched along the geomagnetic field, 3 – field aligned electric
currents, and 4 – satellite

yields the maximum of magnetic field amplitude of ELF emission borne in

this scattering process of the order of b  (1  2) pT over the frequency range
(200  500) Hz. These characteristics are in agreement with the experimental data [31,
32, 200].
The estimates of the characteristic transverse size (10 km) and the magnitude
of plasma layer density inhomogeneities obtained earlier have allowed to suppose
that small-scale plasma structures excited this way could play a role of ducts or
waveguides that canalized whistler mode waves along the geomagnetic field [201].
An appearance of such structures in the ionosphere above the epicenter region could
increase the probability of the emergence of magnetospheric whistlers into the
Earth–ionosphere waveguide and change the dispersion characteristics of signals
recorded on the Earth’s surface. An influence of seismic activity on the parameters
of magnetospheric whistlers was reported in ref. [202]. The statistical analysis
carried out demonstrated the unusual changes in mid-latitude whistler character-
istics, such as an increase in the dispersion and the occurrence rate, associated with
enhancement of seismic activity. Hayakawa et al. [202] discussed the possible
mechanisms of such influence. In particular, they considered the seismic effect on
the wave trapping in duct (or excitation of duct) at its entrance and the enhanced
occurrence of ducts in the ionosphere disturbed by the acoustic or internal gravity
134 V.M. Sorokin and V.M. Chmyrev

waves generated by earthquakes. This idea was experimentally confirmed by

satellite observations of small-scale plasma density irregularities and correlated
ELF emissions over the seismic zone [32].
At the preparatory stage of earthquake, large-scale changes in the ionosphere
parameters were observed several days before the main seismic shock using the
topside ionosphere sounding from the INTERCOSMOS-19 satellite [43]. It was
shown that the altitude of F2 layer maximum increased from approximately 280 to
360 km, and the maximum electron density decreased from 3  105 to 105 cm3
1 day before the earthquake.
It was shown in Section 3.4.4 that an enhancement of the electric field and
related release of Joule heat in the lower ionosphere increased the temperature and
the upward mass flux in the upper ionosphere. This modifies the diffusion process
that forms ionospheric F layer and changes its characteristics [190]. Let us estimate
the energy flux and the temperature variation in the ionosphere caused by electric
field enhancement. Assuming that the heat flux produced by solar radiation is

q0 ¼ 103 W m2 , and the integral Pedersen conductivity and the ionosphere elec-
1
tric field value to be equal to S  P 2 10 O and E0  3mV=m, respectively, we
3
obtain the heat flux q ¼ 10 W m . Such a flux will heat the ionosphere up to the
temperature T  1; 400 K. If above the region of preparing earthquake the electric
field was increased by the quantity  6 mV/m, as observed in ref. [40], it results in
the thermal flux q ¼ 2  103 W m2 and in the ionosphere temperature growth up
to T  2,100 K. This in turn leads to an elevation of the F-layer and a decrease of the
maximum electron and ion densities in the layer [190] as it was observed from the
INTERCOSMOS-19 satellite [43].
The other observable characteristics of the seismic-related ionosphere distur-
bances are ion composition and the density of the light ions. It was shown that in a
near-earthquake zone during the earthquake preparation processes the H+ and He+
ion densities in the ionosphere increased by about a factor of 3 [44]. The modeling
[190] has shown that the relative increase of the ion density (including the light
ions) at altitudes 700–1,000 km could be six- to eightfold at the electric field
enhancement in the ionosphere by the quantity 6 mV/m.
In conclusion, we note that the electrodynamic model of the atmosphere and the
ionosphere coupling [21] is applicable to the interpretation of some effects pro-
duced in the ionosphere by tropical storms and typhoons [29, 30, 203, 204].

3.5 Conclusions

Experimental studies and theoretical modeling show that the ionosphere is influ-
enced by many processes, which are developed on the ground, in the lower atmo-
sphere and in the stratosphere. This influence provides the ionosphere response to
such phenomena as earthquakes, volcano eruptions, typhoons, some types of techno-
genic disasters, etc. The mechanisms of influencing include a set of rather compli-
cated chemical, electrophysical, hydrodynamic and electromagnetic processes in the
3 Atmosphere–Ionosphere Electrodynamic Coupling 135

atmosphere that accompany the development of these phenomena. The most com-
prehensive model of the atmosphere and the ionosphere coupling [21], which is
presented above describes the complete causal-sequence chain of seismic-induced
processes beginning from modification of the lower atmosphere state to excitation of
the plasma density variations, the geomagnetic field-aligned currents and ULF/ELF
emissions in the ionosphere, an elevation of the F-layer maximum, a decrease of
electron density in the maximum of this layer, and a growth of light ion density in the
upper ionosphere. The lower atmosphere modification is produced by the injection
and turbulent transfer of charged aerosols and radioactive substances, which lead to
the formation of external electric current in the atmosphere, changes in the Earth–-
ionosphere electric circuit, an enhancement of DC electric field in the lower iono-
sphere, and related development of acoustic-gravity wave instability responsible for
generation of the plasma density and electric conductivity inhomogeneities. The
scattering of electromagnetic radiation from lightning by these inhomogeneities
produces additional ELF wave flux observed over an earthquake area in the upper
ionosphere together with plasma density irregularities.
Two other noticeable effects of the lower ionosphere modification by DC
electric field enhancement are the amplitude and phase perturbations of subiono-
spheric VLF signals and the generation of narrow band gyrotropic waves that can be
observed on the Earth surface as ULF geomagnetic oscillations.
An intensified inflow of atmospheric conductivity current into the ionosphere
changes the plasma density distribution in the lower ionosphere and causes the
formation of an anomalous sporadic E-layer.
Let us note that the considered model is applicable to interpretation of some
effects produced on the ionosphere by intense atmospheric disturbances connected
with tropical storms, typhoons, etc.
In spite of definite progress in experimental studies of man-made and natural
effects on the ionosphere, there is significant gap in the understanding of an
origin and interconnection of many processes involved in the lithosphere–
atmosphere–ionosphere coupling. The way of filing this gap is the development
and implementation of long-term international research program based on coordi-
nated observations from the ground networks, aviation, and satellite platforms of all
the processes occurred in different layers of near-Earth space in association
with catastrophic natural phenomena such as earthquakes, volcano eruptions,
typhoons, etc.

References

1. Getmantsev, C.G., Zuikov, N.A., Kotik, D.S., Mironenko, L.F., Mityakov, N.A., Rapoport,
V.O., Sazonov, Y.A., Trakhtengerts, V.Y., Eidman, V.Y.: Combination frequencies in the
interaction between high-power short-wave radiation and ionospheric plasma. JETP Lett. 20,
101–112 (1974)
2. James, H., Inan, U.S., Rietveld, M.T.: Observation on the DE-1 spacecraft of ELF/VLF
waves generated by an ionospheric heater. J. Geophys. Res. 95, 12187 (1990)
136 V.M. Sorokin and V.M. Chmyrev

3. Inan, U.S., Golkowski, M., Carpenter, D.L., Reddell, N., Moore, R.C., Bell, T.F., Paschal,
E., Kossey, P., Kennedy, E., Meth, S.Z.: Multi-hop whistler-mode ELF/VLF signals and
triggered emissions excited by the HAARP HF heater. Geophys. Res. Lett. 31, L24805
(2004)
4. Moore, R.C., Inan, U.S., Bell, T.F., Kennedy, E.J.: ELF waves generated by modulated HF
heating of the auroral electrojet and observed at a ground distance of 4400 km. J. Geophys.
Res. 112, A05309 (2007). doi:10.1029/2006JA012063
5. Parrot, M., Sauvaud, J.A., Berthelier, J.J., Lebreton, J.P.: First in-situ observations of strong
ionospheric perturbations generated by powerful VLF transmitter. Geophys. Res. Lett. 34,
L11111 (2007). doi:10.1029/2007GL029368
6. Frolov, V., Rapoport, V., Komrakov, G., Belov, A., Markov, G., Parrot, M., Rauch, J.,
Mishin, E.: Density ducts formed by heating the Earth’s ionosphere with high-power HF
transmitter. JETP Lett. 88, 790–794 (2008)
7. Milikh, G.M., Papadopoulos, K., Shroff, H., Chang, C.L., Wallace, T., Mishin, E.V., Parrot,
M., Berthelier, J.J.: Formation of artificial ionospheric ducts. Geophys. Res. Lett. 35, L17104
(2008). doi:10.1029/2008GL034630
8. Dzhordzhio, N.V., Mogilevskii, M.M., Chmyrev, V.M., Kovrazhkin, R.A., Molchanov,
O.A., Galperin, YuI, Boske, J.M., Roche, J.L.: Acceleration of ions in the plasma environ-
ment of the Earth by the radiation from a low-frequency transmitter on the ground. JETP
Lett. 46, 405–409 (1987)
9. Chmyrev, V.M., Kuzmin, A.K., Lazarev, V.I., Isaev, N.V., Bilichenko, S.V., Taranenko,
YuN, Teltsov, M.V., Teodosiev, D.K.: Correlation of stable red arcs and Hb emissions with
ion fluxes, electric fields and VLF radiation. Geom. Aeron. 28, 813–819 (1988)
10. Chmyrev, V.M., Mogilevsky, M.M., Molchanov, O.A., Sobolev, Ya.P., Titova, E.E.,
Yakhnina, T.A., Suncheleev, R.N., Gladyshev, V.A., Baranets, N.V., Dzhordzhio, N.V.,
Galperin, Yu.I., Streltsov, A.V.: Parametric excitation of ELF waves and acceleration of
ions at the injection of strong VLF waves into the ionosphere. Kosmich. Issled. 27, 248–257
(1989)
11. Bernhardt, P.A., Scales, W.A., Grach, S.M., Keroshtin, A.N., Kotik, D.S., Polyakov, S.M.:
Excitation of artificial airglow by high-power radio waves from “SURA“ ionospheric heating
facility. Geophys. Res. Lett. 18(8), 1477–1480 (1991)
12. Cohen, D., Weiber, J., King, J., Kemper, S., Stephens, S., Davis, L., Spanjer, G., Winter, J.,
Adler, A., Easley, S., Tolliver, M., Guarnieri, J.: The SSTE-4: DSX flight experiment: design
of a low-cost, R&D space mission with responsive enabling technologies. Paper N.2005-
3004, AIAA 3rd Responsive Space Conference 2005, Los Angeles, CA (2005)
13. Gamble, R.J., Rodger, C.J., Clivlend, M.A., Sauvaud, J.-A., Thomson, N.R., Stewart, S.L.,
McCornick, R.J., Parrot, M., Berthelier, L.-J.: Radiation belt precipitation by manmade VLF
transmission. J. Geophys. Res. (2008). doi:10.1029/2008JA013369
14. Molchanov, O.A.: Wave and plasma phenomena inside the ionosphere and the magneto-
sphere associated with earthquakes. In: Stone, W.R. (ed.) Review of Radio Science
1990–1992, pp. 591–600. Oxford University Press, New York (1993)
15. Buchachenko, A.L., Oraevskii, V.N., Pokhotelov, O.A., Sorokin, V.M., Strakhov, V.N.,
Chmyrev, V.M.: Ionospheric precursors to earthquakes. Phys.-Usp. 39, 959–965 (1996)
16. Varotsos, P.: A review and analysis of electromagnetic precursory phenomena. Acta
Geophys. Pol. 49, 1–42 (2001)
17. Hayakawa, M., Molchanov, O.: Seismo-Electromagnetics (Lithosphere–Atmosphere–Iono-
sphere Coupling), pp. 1–477. Terrapub, Tokyo (2002)
18. Parrot, M., Berthelier, J.J., Lebreton, J.P., Sauvaud, J.A., Santolik, O., Blecki, J.: Examples
of unusual ionospheric observations made by the DEMETER satellite over seismic regions.
Phys. Chem. Earth 31, 486–495 (2006)
19. Parrot, M., Berthelier, J.J., Lebreton, J.P., Treumann, R., Rauch, J.L.: DEMETER observa-
tion of EM emissions related to thunderstorms. Space Sci. Rev. 137, 511519 (2008). doi:
10.1007/s11214-008-9347-y
3 Atmosphere–Ionosphere Electrodynamic Coupling 137

20. Bhattacharya, S., Sarkar, S., Gwal, A.K., Parrot, M.: Satellite and ground-based ULF/ELF
emissions observed before Gujarat earthquake in March 2006. Curr. Sci. 93, 41–46 (2006)
21. Sorokin, V.M., Chmyrev, V.M., Yaschenko, A.K.: Electrodynamic model of the lower
atmosphere and the ionosphere coupling. J. Atmos. Solar-Terr. Phys. 63, 1681–1691 (2001)
22. Sorokin, V.M.: Plasma and electromagnetic effects in the ionosphere related to the dynamic
of charged aerosols in the lower atmosphere. Russ. J. Phys. Chem. 1, 138–170 (2007)
23. Alekseev, V.A., Alekseeva, N.G.: Investigation of metal transfer in the biosphere during
gaseous emission in zones of tectonic activity using methods of nuclear physics. Nucl.
Geophys. 6, 99–105 (1992)
24. Voitov, G.I., Dobrovolsky, I.P.: Chemical and isotope – carbonic instability of the soil gases
in the seismic regions. Izvestiya AN SSSR. Fizika Zemli 3, 20–27 (1994)
25. Virk, H.S., Singh, B.: Radon recording of Uttarkashi earthquake. Geophys. Res. Lett. 21,
737–741 (1994)
26. Heincke, J., Koch, U., Martinelli, G.: CO2 and Radon measurements in the Vogtland area
(Germany) – a contribution to earthquake prediction research. Geophys. Res. Lett. 22,
774–779 (1995)
27. Igarashi, G., Saeki, T., Takahata, N., Sano, Y., Sumikawa, K., Tasaka, S., Sasaki, Y.,
Takahashi, M.: Groundwater radon anomaly before the Kobe earthquake. Science 269,
60–61 (1995)
28. Pulinets, S.A., Alekseev, V.A., Legenka, A.D., Khegai, V.V.: Radon and metallic aerosols
emanation before strong earthquakes and their role in atmosphere and ionosphere modifica-
tion. Adv. Space Res. 20, 2173–2176 (1997)
29. Isaev, N.V., Sorokin, V.M., Chmyrev, V.M., Serebryakova, O.N., Ovcharenko, O.Ya.:
Electric field enhancement in the ionosphere above tropical storm region. In.: Hayakawa,
M., Molchanov, O.A. (eds.) Seismo electromagnetics: litosphere–atmosphere–ionosphere
coupling, pp. 313–315. Terrapub, Tokyo (2002)
30. Sorokin, V.M., Isaev, N.V., Yaschenko, A.K, Chmyrev, V.M., Hayakawa, M.: Strong DC
electric field formation in the low latitude ionosphere over typhoons. J. Atmos. Solar-Terr.
Phys. 67, 1269–1279 (2005)
31. Serebryakova, O.N., Bilichenko, S.V., Chmyrev, V.M., Parrot, M., Rauch, J.L., Lefeuvre, F.,
Pokhotelov, O.A.: Electromagnetic ELF radiation from earthquake regions as observed by
low-altitude satellites. Geophys. Res. Lett. 19, 91–94 (1992)
32. Chmyrev, V.M., Isaev, N.V., Serebryakova, O.N., Sorokin, V.M., Sobolev, Ya.P.: Small-
scale plasma inhomogeneities and correlated ELF emissions in the ionosphere over an
earthquake region. J. Atmos. Solar-Terr. Phys. 59, 967–973 (1997)
33. Gokhberg, M.B., Morgunov, V.A., Yoshino, T., Tomizawa, I.: Experimental measurements
of electromagnetic emissions possibly related to earthquake in Japan. J. Geophys. Res. 87,
7824–7828 (1982)
34. Koons, H.C., Roeder, J.L.: A comparison of ULF/ELF measurements associated with earth-
quakes. In: Hayakawa, M. (ed.) Atmospheric and ionospheric electromagnetic phenomena
associated with earthquakes, pp. 171–175. Terrapub, Tokyo (1999)
35. Henderson, T.R., Sonwalkar, V.S., Helliwell, R.A., Inan, U.S., Fraser-Smith, A.C.: A Search
for ELF/VLF emissions induced by earthquakes as observed in the ionosphere by the DE-2
satellite. J. Geophys. Res. 98, 9503–9511 (1993)
36. Borisov, N., Chmyrev, V., Rybachek, S.: A new ionospheric mechanism of electromagnetic
ELF precursors to earthquakes. J. Atmos. Solar-Terr. Phys. 63, 3–10 (2001)
37. Kondo, G.: The variation of the atmospheric electric field at the time of earthquake. Memoirs
of the Kakioka Magnetic Observatory, Kakioka. Japan 13, 17–23 (1968)
38. Pierce, E.T.: Atmospheric electricity and earthquake prediction. Geophys. Res. Lett. 3,
185–188 (1976)
39. Hao, J.: The anomalous of atmospheric electric field at the ground level and earthquakes.
Acta Seismol. Sinica. 10, 207–211 (1988)
138 V.M. Sorokin and V.M. Chmyrev

40. Chmyrev, V.M., Isaev, N.V., Bilichenko, S.V., Stanev, G.A.: Observation by space-borne
detectors of electric fields and hydromagnetic waves in the ionosphere over on earthquake
center. Phys. Earth Planet. Inter. 57, 110–114 (1989)
41. Tate, J., Daily, W.: Evidence of electro-seismic phenomena. Phys. Earth Planet. Inter. 57,
1–9 (1989)
42. Vershinin, E.F., Buzevich, A.V., Yumoto, K., Saita, K., Tanaka, Y.: Correlations of seismic
activity with electromagnetic emissions and variations in Kamchatka region. In: Hayakawa,
M. (ed.) Atmospheric and ionospheric electromagnetic phenomena associated with earth-
quakes, pp. 513–518. Terrapub, Tokyo (1999)
43. Pulinets, S.A., Legenka, A.D. Alekseev, V.A.: Pre-earthquakes effects and their possible
mechanisms. Dusty and dirty plasmas, noise and chaos in space and in the laboratory,
pp. 545–557. Plenum Publishing, New York (1994)
44. Boskova, J., Smilauer, I., Triska, P., Kudela, K.: Anomalous behaviour of plasma parameters
as observed by the Intercosmos-24 satellite prior to the Iranian earthquake of 20 June 1990.
Studia Geophys. Geodet. 8, 213–220 (1994)
45. Afonin, V.V., Molchanov, O.A., Kodama, T., Hayakawa, M., Akentieva, O.A.: Statistical
study of ionospheric plasma response to seismic activity: search for reliable result from
satellite observations. In: Hayakawa, M. (ed.) Atmospheric and ionospheric electromagnetic
phenomena associated with earthquakes, pp. 597–617. Terra Scientific Publishing Company
(TERRAPUB), Tokyo (1999)
46. Pulinets, S.A., Legenka, A.D.: Spatial-temporal characteristics of the large scale distur-
bances of electron concentration observed in the F-region of the ionosphere before strong
earthquake. Cosmic Res. 41, 221–229 (2003)
47. Tronin, A.A.: Satellite thermal survey application for earthquake prediction. In.: Hayakawa,
M. (ed.) Atmospheric and ionospheric electromagnetic phenomena associated with earth-
quakes, pp. 717–723. Terrapub, Tokyo (1999)
48. Tronin, A.A., Hayakawa, M., Molchanov, O.A.: Thermal IR satellite data application for
earthquake research in Japan and China. J. Geodyn. 33, 519–534 (2002)
49. Qiang, Z.J., Dian, C.G., Li, L.Z.: Satellite thermal infrared precursors of two moderate-
strong earthquakes in Japan and impending earthquake prediction. In: Hayakawa, M. (ed.)
Atmospheric and ionospheric electromagnetic phenomena associated with earthquakes,
pp. 747–745. Terrapub, Tokyo (1999)
50. Tramutoli, V., Di Bello, G., Pergova, N., Piscitalli, S.: Robust satellite techniques for remote
sensing of seismically active areas. Ann. Geofis. 44, 295–312 (2001)
51. Toroshelidze, T.I., Fishkova, L.M.: Analyzes of the middle and upper atmosphere lumines-
cence before earthquakes. DAN SSSR, Fizika Zemli. 302, 313–319 (1986)
52. Gokhberg, M.B., Nekrasov, A.K., Shalimov, S.L.: On influence of the unstable injection of
green gases to the ionosphere in seismic region. Izvestiya AN SSSR, Fizika Zemli. 8, 52–60
(1996)
53. Draganov, A.B., Inan, U.S., Taranenko, YuN: ULF magnetic signatures at the earth surface
due to ground water flow: a possible precursor to earthquakes. Geophys. Res. Lett. 18,
1127–1131 (1991)
54. Surkov, V., Pilipenko, V.: The physics of pre-seismic electromagnetic ULF signals. In:
Hayakawa, M. (ed.) Atmospheric and ionospheric electromagnetic phenomena associated
with earthquakes, pp. 357–363. Terrapub, Tokyo (1999)
55. Molchanov, O.A., Hayakawa, M., Rafalsky, V.A.: Penetration characteristics of electromag-
netic emissions from an underground seismic source into the atmosphere, ionosphere, and
magnetosphere. J. Geophys. Res. 100, 1691–1712 (1995)
56. Fitterman, D.V.: Theory of electrokinetic–magnetic anomalies in a faulted half-space.
J. Geophys. Res. 84, 6031–6040 (1979)
57. Pilipenko, V.A., Fedorov, E.N., Yagova, N.V., Yumoto, K.: Attempt to detect ULF electro-
magnetic activity preceding earthquake. In: Hayakawa, M. (ed.) Atmospheric and
3 Atmosphere–Ionosphere Electrodynamic Coupling 139

ionospheric electromagnetic phenomena associated with earthquakes, pp. 203–214. Terra-

pub, Tokyo (1999)
58. Alperovich, L.S., Gokhberg, M.B., Sorokin, V.M., Fedorovich, G.V.: On generation of the
geomagnetic variations by acoustic oscillation occurring at the time of earthquakes. Izvestiya
AN SSSR, Fizika Zemli 3, 58–68 (1979)
59. Sorokin, V.M., Fedorovich, G.V.: Propagation of the short periodic waves in the ionosphere.
Izvestiya VUZov, Radiofizika. 25, 495–507 (1982)
60. Grimalsky, V.V., Hayakawa, M., Ivchenko, V.N., Rapoport, YuG, ZAdoroznii, V.I.: Pene-
tration of an electrostatic field from the lithosphere into the ionosphere and its effect on the
D-region before earthquakes. J. Atmos. Solar-Terr. Phys. 65, 391–407 (2003)
61. Rapoport, Y., Grimalsky, V., Hayakawa, M., Ivchenko, V., Juarez, R.D., Koshevaya, S.,
Gotynyan, O.: Change of ionospheric plasma parameters under the influence of electric field
which has lithospheric origin and due to radon emanation. Phys. Chem. Earth 29, 579–587
(2004)
62. Volland, H.: Atmospheric electrodynamics. Springer, New York (1984)
63. Pasko, V.P.: Dynamic coupling of quasi-electrostatic thundercloud fields to the mesosphere
and lower ionosphere: sprites and jets. Ph.D. thesis, Stanford University, Stanford, CA
(1996)
64. Uman, M.A.: The lightning discharge. Academic, Orlando, FL (1987)
65. Inan, U.S., Sampson, W.A., Taranenko, Y.N.: Space-time structure of lower ionospheric
optical flashes and ionization changes produced by lightning EMP. Geophys. Res. Lett. 23,
133–138 (1996)
66. Roussel-Dupre, R.A., Gurevich, A.V., Tunnell, T., Milikh, G.M.: Kinetic theory of runaway
air breakdown. Phys. Rev. E 49, 2257–2269 (1994)
67. Bell, T.F., Pasko, V.P., Inan, U.S.: Runaway electrons as a source of Red Sprites in the
mesosphere. Geophys. Res. Lett. 22, 2127–2135 (1995)
68. Helliwell, R.A., Katsufrakis, J.P., Trimpi, M.L.: Whistler-induced amplitude perturbation in
VLF propagation. J. Geophys. Res. 78, 4679–4688 (1973)
69. Carpenter, D.L., Inan, U.S., Trimpi, M.L., Helliwell, R.A., Katsufrakis, J.P.: Perturbations
of subionospheric LF and MF signals due to whistler-induced electron precipitation bursts.
J. Geophys. Res. 89, 9857–9867 (1984)
70. Burgess, W.C., Inan, U.S.: The role of ducted whistlers in the precipitation loss and
equilibrium flux of radiation belt electrons. J. Geophys. Res. 98, 15643–15650 (1993)
71. Rosenberg, T.J., Helliwell, R.A., Katsufrakis, J.P.: Electron precipitation associated with
discrete, very low frequency emissions. J. Geophys. Res. 76, 8445–8456 (1971)
72. Rycroft, M.J.: Enhanced energetic electron intensities at 100 km altitude and a whistler
propagation through the plasmasphere. Planet. Space Sci. 21, 239–247 (1973)
73. Goldberg, R.J., Curtis, S.A., Barcus, J.R.: Detailed spectral structure of magnetospheric
electron bursts precipitated by lightning. J. Geophys. Res. 92, 2505–2512 (1987)
74. Voss, H.D., Imhof, W.L., Mobila, J., Gaines, E.E., Walt, M., Inan, U.S., Helliwell, R.A.,
Carpenter, D.L., Katsufrakis, J.P., Chang, H.C.: Lightning-induced electron precipitation.
Nature 312, 740–749 (1984)
75. Imhof, W.L., Voss, H.D., Walt, M., Gaines, E.E., Mobila, J., Datlove, D.W., Reagan, J.B.:
Slot region electron precipitation by lightning. J. Geophys. Res. 91, 8883–8892 (1986)
76. Voss, H.D., Walt, M., Imhof, W.L., Mobila, J., Inan, U.S.: Satellite observation of lightning-
induced electron precipitation. J. Geophys. Res. 103, 11725–11732 (1998)
77. Inan, U.S., Bell, T.F., Helliwell, R.A.: Nonlinear pitch angle scattering of energetic electrons
by coherent VLF waves in the magnetosphere. J. Geophys. Res. 83, 3235–3246 (1978)
78. Chang, H.C., Inan, U.S.: Lightning-induce energetic electron precipitation from the magne-
tosphere. J. Geophys. Res. 90, 4531–4539 (1985)
79. Armstrong, W.C.: Recent advances from studies of the Trimpi effect. Antarctic J. 18,
281–286 (1983)
140 V.M. Sorokin and V.M. Chmyrev

80. Inan, U.S., Shafer, D.C., Yip, W.Y., Orville, R.E.: Subionospheric VLF signatures of
nighttime D-region perturbations in the vicinity of lightning discharges. J. Geophys. Res.
93, 11455–11467 (1988)
81. Inan, U.S., Rodriguez, J.V., Idone, V.P.: VLF signatures of lightning-induced heating and
ionization of the nighttime D-region. Geophys. Res. Lett. 20, 2355–2360 (1993)
82. Inan, U.S., Bell, T.F., Pasko, V.P., Sentman, D.D., Wescott, E.M., Lyons, W.A.: VLF
signatures of ionospheric disturbances associated with sprites. Geophys. Res. Lett. 22,
3461–3466 (1995)
83. Inan, U.S., Slingeland, A., Pasko, V.P., Rodriguez, J.: VLF signatures of mesospheric/lower
ionospheric response to lightning discharges. J. Geophys. Res. 101, 5219–5228 (1996)
84. Dowden, R.L., Adams, C.D.C., Brundell, J.B., Dowden, P.E.: Rapid onset, rapid decay
(RORD), phase and amplitude perturbations of VLF subionospheric transmissions.
J. Atmos. Terr. Phys. 56, 1513–1521 (1994)
85. Sentman, D.D., Wescott, E.M.: Red sprites and blue jets: thunderstorm-excited optical
emissions in the stratosphere, mesosphere and ionosphere. Phys. Plasmas 2, 2514–2522
(1995)
86. Lyons, W.A.: Characteristics of luminous structures in the stratosphere above thunderstorms
as imaged by low-light video. Geophys. Res. Lett. 21, 875–881 (1994)
87. Lyons, W.A.: Low-light video observations of frequent luminous structures in the strato-
sphere above thunderstorms. Mon. Weather Rev. 122, 1940–1950 (1995)
88. Lyons, W.A.: Sprite observations above the U.S. high plains in relation to their parent
thunderstorm systems. J. Geophys. Res. 101, 29641–29652 (1996)
89. Boeck, W.L., Vaughan, O.H., Blakeslee, R.J., Vonnegut, B., Brook, M., McKune, J.:
Observation of lightning in the stratosphere. J. Geophys. Res. 100, 1465–1472 (1995)
90. Rairden, R.L., Mende, S.B.: Time resolved sprite imagery. Geophys. Res. Lett. 22,
3465–3469 (1995)
91. Winckler, J.R., Lyons, W.A., Nelson, T., Nemzek, R.J.: New high-resolution ground-based
studies of cloud-ionosphere discharges over thunderstorms (CI or Sprites). J. Geophys. Res.
101, 6997–7012 (1996)
92. Wescott, E.M., Sentman, D., Osborne, D., Hampton, D., Heavner, M.: Preliminary results
from the Sprites94 aircraft campaign: 2. Blue jets. Geophys. Res. Lett. 22, 1209–1213 (1995)
93. Boeck, W.L., Vaughan, O.H., Blakeslee, R.J., Vonnegut, B., Brook, M.: Lightning-induced
brightening in the airglow layer. Geophys. Res. Lett. 19, 99–103 (1992)
94. Fukunishi, H., Takahashi, Y., Kubota, M., Sakanoi, K., Inan, U.S., Lyons, W.A.: Lightning-
induced transient luminous events in the lower ionosphere. Geophys. Res. Lett. 23,
2157–2163 (1996)
95. Boccippio, D.J., Williams, E.R., Heckman, S.J., Lyons, W.A., Baker, I.T., Boldi, R.: Sprites,
ELF transients, and positive ground strokes. Science 269, 1088–1093 (1995)
96. Fishman, G.J., Bhat, P.N., Mallozzi, R., Horack, J.M., Koshut, T., Kouveliotou, C., Pendleton,
G.N., Meegan, C.A., Wilson, R.B., Paciesas, W.S., Goodman, S.J., Christian, H.J.: Discovery
of intense gamma-ray flashes of atmospheric origin. Science 264, 1313–1319 (1994)
97. Inan, U.S., Reising, S.C., Fishman, G.J., Horack, J.M.: On the association of terrestrial
gamma-ray bursts with lightning discharges and sprites. Geophys. Res. Lett. 23,
1017–1022 (1996)
98. Holden, D.N., Munson, C.P., Devenport, J.C.: Satellite observation of transionospheric pulse
pairs. Geophys. Res. Lett. 22, 889–893 (1995)
99. Taranenko, Y.N., Inan, U.S., Bell, T.F.: Interaction with the lower ionosphere of electro-
magnetic pulses from lightning: heating, attachment, and ionization. Geophys. Res. Lett. 20,
1539–1545 (1993)
100. Taranenko, Y.N., Inan, U.S., Bell, T.F.: The interaction with the lower ionosphere of
electromagnetic pulses from lightning: excitation of optical emissions. Geophys. Res. Lett.
20, 2675–2680 (1993)
3 Atmosphere–Ionosphere Electrodynamic Coupling 141

101. Milikh, G.M., Papadopoulos, K., Chang, C.L.: On the physics of high altitude lightning.
Geophys. Res. Lett. 22, 85–91 (1995)
102. Rowland, H.L., Fernsler, R.F., Huba, J.D., Bernhardt, P.A.: Lightning driven EMP in the
upper atmosphere. Geophys. Res. Lett. 22, 361–367 (1995)
103. Pasko, V.P., Inan, U.S., Taranenko, Y.N., Bell, T.F.: Heating, ionization and upward
discharges in the mesosphere due to intense quasi-electrostatic thunderstorm fields. Geo-
phys. Res. Lett. 22, 365–370 (1995)
104. Pasko, V.P., Inan, U.S., Bell, T.F.: Sprites as luminous columns of ionization produced by
quasi-electrostatic thunderstorm fields. Geophys. Res. Lett. 23, 649–655 (1996)
105. Pasko, V.P., Inan, U.S., Bell, T.F.: Blue jets produced by quasi-electrostatic pre-discharge
thunderstorm fields. Geophys. Res. Lett. 23, 301–307 (1996)
106. Pasko, V.P., Inan, U.S., Bell, T.F., Taranenko, Y.N.: Sprites produced by quasi-electrostatic
heating and ionization in the lower ionosphere. J. Geophys. Res. 102, 4529–4539 (1997)
107. Roussel-Dupre, R.A., Gurevich, A.V.: On runaway breakdown and upward propagating
discharges. J. Geophys. Res. 101, 2297–2310 (1996)
108. Taranenko, Y.N., Roussel-Dupre, R.A.: High altitude discharges and gamma-ray flashes: a
manifestation of runaway air breakdown. Geophys. Res. Lett. 23, 571–576 (1996)
109. Lehtinen, N.G., Walt, M., Inan, U.S., Bell, T.F., Pasko, V.P.: X-ray emission produced by a
relativistic beam of runaway electrons accelerated by quasi-electrostatic thundercloud fields.
Geophys. Res. Lett. 23, 2645–2652 (1996)
110. Franz, R.C., Nemzek, R.J., Winckler, J.R.: Television image of a large upward electric
discharge above a thunderstorm system. Science 249, 48–54 (1990)
111. Sentman, D.D., Wescott, E.M., Osborne, D.L., Hampton, D.L., Heavner, M.J.: Preliminary
results from the Sprites94 campaign: red sprites. Geophys. Res. Lett. 22, 1205–1211 (1995)
112. Sentman, D.D., Wescot, E.M.: Observation of upper atmosphere optical flashes recorded
from an aircraft. Geophys. Res. Lett. 20, 2857–2864 (1993)
113. Sentman, D.D., Wescott, E.M.: Red sprites and blue jets. Geophysical Institute Video
Production, University of Alaska, Fairbanks (1994)
114. Vaughan, O.H., Blakeslee, R.J., Boeck, W.L., Vonnegut, B., Brook, M., McKune, J.:
A cloud-to-space lightning as recorded by the space shuttle payload-bay TV cameras.
Mon. Weather Rev. 120, 1459–1465 (1992)
115. Wescott, E.M., Sentman, D.D, Heavner, M.J, Hampton, D.L.: Blue starters, discharges above
an intense thunderstorm over Arkansas, July 1, 1994. In: Proceedings of the EOS Transac-
tions. AGU, 1995 Fall Meeting, 76, F104 (1995)
116. Picard, R.H., Inan, U.S., Pasko, V.P., Winick, J.R., Wintersteiner, P.P.: Infrared glow above
thunderstorms. Geophys. Res. Lett. 24, 2635–2643 (1997)
117. Parrot, M., Berthelier, J.J., Lebreton, J.P., Treumann, R., Rauch, J.L.: DEMETER observa-
tion of EM emissions related to thunderstorms. Space Sci. Rev. 137, 511–519 (2008).
doi:10.1007/s11214-008-9347-y
118. Berthelier, J.J., Malingre, M., Pfaff, R., Seran, E., Pottelette, R., Lebreton, J.P., Parrot, M.:
Lightning – induced plasma turbulence and ion heating in equatorial ionospheric depletion.
Nat. Geosci. (2008). doi:10.1038/ngeo109
119. Inan, U.S., Carpenter, D.L.: Lightning-induced electron precipitation events observed at L
2.4 as phase and amplitude perturbations on subionospheric VLF signals. J. Geophys. Res.
92, 3293–3299 (1987)
120. Clilverd, M.A., Rodger, C.J., Nunn, D.: Radiation belt electron precipitation fluxes asso-
ciated with lightning. J. Geophys. Res. 109, A12208 (2004). doi:10.1029/2004JA010644
121. Inan, U.S., Piddyachiy, D., Peter, W.B., Sauvaud, J.A., Parrot, M.: DEMETER satellite
observation of lightning-induced electron precipitation. Geophys. Res. Lett. 34, L07103
(2007). doi:10.1029/2006GL029238
122. Jones, T.B., Davis, K., Wieder, B.: Observation of D-region modifications at low and very
low frequencies. Nature 238, 33–37 (1972)
142 V.M. Sorokin and V.M. Chmyrev

123. Barr, R., Rietveld, M.T., Kopka, H., Stubbe, P.: Effects of heated patch of auroral ionosphere
on VLF radio wave propagation. Nature 309, 534–538 (1984)
124. Barr, R., Rietveld, M.T., Kopka, H., Stubbe, P., Nielsen, E.: Extra-low-frequency radiation
from the polar electrojet antenna. Nature 317, 155160 (1985)
125. Bell, T.F., Inan, U.S., Danielson, M., Cummer, S.: VLF signatures of ionospheric heating by
HIPAS. In: Goodmaned, J.M. (ed.) Proceedings of the 1993 Ionospheric Effects Symposium,
pp. 622–628. SRI International, Arlington, VA (1993)
126. Stubbe, P., Kopka, H., Rietveld, M.T., Dowden, R.L.: ELF and VLF generation by modu-
lated heating of the current carrying lower ionosphere. J. Atmos. Terr. Phys. 44, 1123–1128
(1982)
127. Ferrano, A.J., Lee, H.S., Allshouse, R., Carroll, K., Lunnen, R., Collins, T.: Characteristics of
ionospheric ELF radiation generated by HF heating. J. Atmos. Terr. Phys. 46, 855–863
(1984)
128. Barr, R., Stubbe, P., Rietveld, M.T., Kopka, H.: ELF and VLF signals radiated by the “polar
electrojet antenna”: experimental results. J. Geophys. Res. 91, 4451–4462 (1986)
129. Barr, R., Stubbe, P., Kopka, H.: Long-range detection of VLF radiation produced by heating
the auroral electrojet. Radio Sci. 26, 871–989 (1991)
130. Papadopoulos, K., Wallace, T., McCarrick, M., Milikh, G.M., Yang, X.: On the efficiency of
ELF/VLF generation using HF heating of the auroral electrojet. Plasma Phys. Rep. 29,
561–567 (2003)
131. Pashin, A.B., Mochalov, A.A., Bosinger, T., Rietveld, M.T.: Physics of auroral phenomena.
In: Proceedings of the XXVI Annual Seminar, Apatity, pp. 111–114. Kola Science Centre,
Russian Academy of Science (2003)
132. Kimura, I., Stubbe, P., Rietveld, M.T., Barr, R., Ishida, K., Kasahara, Y., Yagitani, S.,
Nagano, I.: Collaborative experimentsby Akebono satellite, Tromso ionospheric heater,
and European incoherent scatter radar. Radio Sci. 29, 23–29 (1994)
133. Ferrano, A.J., Lee, H.S., Allshouse, R., Carroll, K., Tomko, A.A., Kelly, F.J., Joiner, R.G.:
VLF/ELF radiation from the ionospheric dynamo current system modulated by powerful HF
signals. J. Atmos. Terr. Phys. 44, 1113–1119 (1982)
134. McCarrick, M.J., Sentman, D.D., Wong, A.Y., Wuerker, R.F., Chouinard, B.: Excitation of
ELF waves in the Schumann resonance range by modulated HF heating of the polar
electrojet. Radio Sci. 25, 1291–1298 (1990)
135. Villasenor, J., Wong, A.Y., Song, B., Pau, J., McCarrick, M., Sentman, D.: Comparison of
ELF/VLF generation modes in the ionosphere by the HIPAS heater array. Radio Sci. 31,
211–217 (1996)
136. Kimura, I., Wong, A., Chouinard, B., McCarrick, M., Okada, T.: Satellite and ground
observations of HIPAS VLF modulation. Geophys. Res. Lett. 18, 309–314 (1991)
137. Milikh, G.M., Papadopoulos, K., McCarrick, M., Preston, J.: ELF emissions generated by the
HAARP HF-heater using varying frequencies and polarization. Izvestiya VUZov, Radio-
fizika 42, 728–733 (1999)
138. Platino, M., Inan, U.S., Bell, T.F., Parrot, M., Kennedy, E.J.: DEMETER observations of
ELF waves injected with the HAARP HF transmitter. Geophys. Res. Lett. 33, L16101 (2006)
139. Moore, R.C.: ELF/VLF wave generation by modulated HF heating of the auroral electrojet.
Ph.D. thesis, Stanford University, Stanford, CA (2007)
140. Rapoport, V.O., Frolov, V.L., Komrakov, G.P. Markov, G.A., Belov, A.S., Parrot, M.,
Rauch, J.L.: Some results of measuring the characteristics of electromagnetic and plasma
disturbances stimulated in the outer ionosphere by high-power high-frequency radio emis-
sion from the “Sura” facility. Radiophys. Quantum Electron. 50, 645–651 (2007)
141. Frolov, V., Komrakov, G., Rapoport, V., Ryzhov, N., Belov, A., Markov, G., Parrot, M.,
Rauch, J., Reitveld, M.: Phenomena observed by HF heating of middle – and high-latitude
ionosphere and registered with DEMETER instruments. Geophys. Res. Abstracts 10, EGU
2008-A-03872 (2008a)
3 Atmosphere–Ionosphere Electrodynamic Coupling 143

142. Zhulin, I.A., Lyakhov, S.B., Majorov, A.D., Managadze, G.G., Mogilevsky, M.M.,
Chmyrev, V.M.: Artificially stimulated electron precipitation from the Earth’s magneto-
sphere. Dokl. Akad. Nauk SSSR 230, 1073–1077 (1976)
143. Imhof, W.L., Reagan, J.B., Voss, H.D., Gaines, E.E., Datlowe, D.W., Mobilia, J., Helliwell,
R.A., Inan, U.S., Ratsufrakis, J.P.: Direct observation of radiation belt electrons precipitated
by the controlled injection of VLF signals from a ground-based transmitter. Geophys. Res.
Lett. 10, 361–366 (1983)
144. Imhof, W.L., Reagan, J.B., Voss, H.D., Gaines, E.E., Datlowe, D.W., Mobilia, J., Helliwell,
R.A., Inan, U.S., Ratsufrakis, J.P.: The modulated precipitation of radiation belt electrons by
controlled signals from VLF transmitter. Geophys. Res. Lett. 10, 615–620 (1983)
145. Inan, U.S., Chang, H.C., Helliwell, R.A., Imhof, W.L., Reagan, J.B., Walt, M.: Precipitation
of radiation belt electrons by man-made waves: a comparison between theory and measure-
ment. J. Geophys. Res. 90, 359–370 (1985)
146. Kovrazhkin, R.A., Mogilevsky, M.M., Bosqued, J.-M., Galperin, Y.I., Dzhordzhio, N.V.,
Lisakov, Y.V., Molchanov, O.A., Reme, A.: Observation of particle precipitation from the
ring-current zone stimulated by powerful ground-based VLF transmitter. JETP Lett. 38,
397–402 (1983)
147. Kovrazhkin, R.A., Mogilevsky, M.M., Molchanov, O.A., Galperin, Y.I., Dzhordzhio, N.V.,
Lisakov, Y.V., Bosqued, J.-M., Reme, A.: Direct detection of the precipitation of ring current
electrons and protons stimulated by artificial VLF emission. Geophys. Res. Lett. 11, 705–709
(1984)
148. Sauvaud, J.-A., Maggiolo, R., Jacquey, C., Parrot, M., Berthelier, J.-J., Gamble, R.J.:
Radiation belt electron precipitation due to VLF transmission. Satellite observations. Geo-
phys. Res. Lett. 35, L09101 (2008). doi:10.1029/2008GL033194
149. Inan, U.S., Rodriguez, J.V., Lev-Tov, S., Oh, J.: Ionospheric modification with a VLF
transmitter. Geophys. Res. Lett. 19, 2071–2077 (1992)
150. Barr, R., Rietveld, M.T., Stubbe, P., Kopka, H.: The diffraction of VLF radio waves by a
patch of ionosphere illuminated by a powerful HF transmitter. J. Geophys. Res. 90,
2861–2869 (1985)
151. Taranenko, Y.N., Inan, U.S., Bell, T.F.: VLF-HF heating of the lower ionosphere and ELF
wave generation. Geophys. Res. Lett. 19, 61–66 (1992)
152. Rodriguez, J.V.: Modification of the Earth’s ionosphere by very-low-frequency transmitter.
Ph.D. thesis, Stanford University, Stanford, CA (1994)
153. Rodriguez, J.V., Inan, U.S.: Electron density changes in the nighttime D region due to
heating by very-low-frequency transmitter. Geophys. Res. Lett. 21, 93–98 (1994)
154. Likhter, Ya.I., Molchanov, O.A., Chmyrev, V.M.: Modulation of spectrum and amplitude of
low-frequency signal in the magnetosphere. JETP Lett. 14, 325–327 (1971)
155. Bell, T.F., James, H.G., Inan, U.S., Katsufrakis, J.P.: The apparent spectral broadening of
VLF transmitter signals during trans-ionospheric propagation. J. Geophys. Res. 88,
4813–4818 (1983)
156. Titova, E.E., Di, V.I., Yurov, V.E., Raspopov, O.M., Trakhtengerts, V.Y., Jiricek, F., Triska,
P.: Interaction between VLF waves and turbulent ionosphere. Geophys. Res. Lett. 11,
323–327 (1984)
157. Bell, T.F., Inan, U.S., Lauben, D., Sonwalkar, V.S., Helliwell, R.A., Sobolev, YaP,
Chmyrev, V.M., Gonzalez, S.: DE-1 and COSMOS-1809 observations of lower hybrid
waves excited by VLF whistler mode waves. Geophys. Res. Lett. 21, 653–656 (1994)
158. Taranenko, YuN, Chmyrev, V.M.: Interaction between whistler waves and ion-cyclotron
waves in magnetospheric plasma. Radiophys. Quant. Electron. 29, 373–376 (1986)
159. Taranenko, YuN, Chmyrev, V.M.: Parametric interaction of whistler and electromagnetic
ion-cyclotron waves in ionospheric plasma. Geom. Aeron. 29, 459–464 (1989)
160. Chmyrev, V.M., Draganov, A.B., Taranenko, YuN, Teodosiev, D.: Acceleration of particles
in the upper ionosphere and the magnetosphere due to decay interactions of whistlers, part 1.
Phys. Scr. 43, 495–502 (1991)
144 V.M. Sorokin and V.M. Chmyrev

161. Chang, T., Crew, G.B., Hershkowtz, N., Jasper, J.R., Retterer, J.M., Winningham, J.D.:
Transverse acceleration of oxygen ions by electromagnetic ion cyclotron resonance with
broad band left hand polarized waves. Geophys. Res. Lett. 13, 636–640 (1986)
162. Frazer-Smith, A.C., Cole, C.A.: Initial observations of the artificial stimulation of ULF
pulsations by pulsed VLF transmissions. Geophys. Res. Lett. 2, 146–149 (1975)
163. Chmyrev, V.M., Roldugin, V.K., Zhulin, I.A., Mogilevsky, M.M., Di, V.I., Koshelevsky,
V.K., Bushmarin, V.A., Raspopov, O.M.: Artificial injection of very low-frequency (VLF)
waves into the ionosphere and the magnetosphere of the Earth. JETP Lett. 23, 409–412
(1976)
164. Chmyrev, V.M., Taranenko, Yu.N, Kopytenko, Yu.A., Voronov, P.V., Draganov, A.B.:
Observation of ULF pulsations correlated with transmission of VLF waves and amplification
of ULF waves by O+ ion conics in the equatorial magnetosphere. Paper presented at AGU
Chapman Conference on auroral plasma dynamics, Minneapolis, MN, 21–25 Oct 1991
165. Sorokin, V.M., Chmyrev, V.M.: Electrodynamic model of ionospheric precursors of earth-
quakes and certain types of disasters. Geom. Aeron. 42, 821–830 (2002)
166. Molchanov, O.A., Hayakawa, M.: VLF transmitter earthquake precursors influenced by a
change in atmospheric electric field. In: Proceedings of 10th International Conference on
Atmospheric Electricity, pp. 428–431. Osaka, Japan, 10–14 June 1996
167. Boyarchuk, K.A., Lomonosov, A.M, Pulinets, S.A., Hegai, V.V.: Variability of the Earth’s
atmospheric electric field and ion-aerosols kinetics in the troposphere. Studia Geophys.
Geod. 42, 197–206 (1998)
168. Sorokin, V.M., Chmyrev, V.M., Yaschenko, A.K.: Theoretical model of DC electric field
formation in the ionosphere stimulated by seismic activity. J. Atmos. Solar-Terr. Phys. 67,
1259–1268 (2005)
169. Sorokin, V.M., Yaschenko, A.K.: Perturbation of the conductivity and electric field in the
Earth–ionosphere layer over preparing earthquake. Geom. Aeron. 39, 100–106 (1999)
170. Sorokin, V., Yaschenko, A.: Electric field disturbance in the Earth–ionosphere layer. Adv.
Space Res. 26, 1219–1225 (2000)
171. Sorokin, V.M., Yaschenko, A.K., Hayakawa, M.: A perturbation of DC electric field caused
by light ion adhesion to aerosols during the growth in seismic-related atmospheric radioac-
tivity. Nat. Hazards Earth Syst. Sci. 7, 155–163 (2007)
172. Sorokin, V.M., Chmyrev, V.M., Isaev, N.V.: A generation model of mall-scale geomagnetic
field-aligned plasma inhomogeneities in the ionosphere. J. Atmos. Solar-Terr. Phys. 60,
1331–1342 (1998)
173. Sorokin, V.M., Chmyrev, V.M.: On acoustic gravity waves instability by electric field in the
ionosphere. Geom. Aeron. 39, 38–45 (1999)
174. Piddington, J.H.: The transmission of geomagnetic disturbances through the atmosphere and
interplanetary space. Geophys. J. 2, 173–189 (1959)
175. Chmyrev, V.M., Sorokin, V.M., Pokhotelov, O.A.: Theory of small scale plasma density
inhomogeneities and ULF/ELF magnetic field oscillations excited in the ionosphere prior
to earthquakes. In: Hayakawa, M. (ed.) Atmospheric and Ionospheric Electromagnetic
Phenomena Associated with Earthquakes, pp. 759–776. Terrapub, Tokyo (1999)
176. Lyatsky, V.B., Maltsev, YuP: Magnetosphere–Ionosphere Coupling. Nauka, Moscow (1983)
177. Molchanov, O.A., Mazhaeva, O.A., Golyavin, A.N., Hayakawa, M.: Observation by Inter-
cosmos-24 satellite of ELF–VLF electromagnetic emissions associated with earthquakes.
Ann. Geophys. 11, 431–440 (1993)
178. Chmyrev, V.M., Sorokin, V.M., Shklyar, D.R.: VLF transmitter signals as a possible tool for
detection of seismic effects on the ionosphere. J. Atmos. Solar-Terr. Phys. 70, 2053–2060
(2008)
179. Frazer-Smith, A.C., Bernardi, A., McGill, P.R., Ladd, M.E., Helliwell, R.A., Villard Jr., O.G.:
Low-frequency magnetic field measurements near epicentre of the MS 7.1 Loma Prieta
earthquake. Geophys. Res. Lett. 17, 1465–1468 (1990)
3 Atmosphere–Ionosphere Electrodynamic Coupling 145

180. Kopytenko, Y.A., Matiashvili, T.G., Voronov, P.M., Kopytenko, E.A., Molchanov, O.A.:
Detection of ultra-low-frequency emissions connected with the Spitak earthquake and its
aftershock activity, based on geomagnetic pulsations data at Dusheti and Vardzia observa-
tories. Phys. Earth Planet. Inter. 77, 85–89 (1993)
181. Hayakawa, M., Kawate, R., Molchanov, O.A., Yumoto, K.: Results of ultra-low-frequency
magnetic field measurements during the Guam earthquake of August 1993. Geophys. Res.
Lett. 23, 241–250 (1996)
182. Sorokin, V.M., Chmyrev, V.M., Yaschenko, A.K.: Ultra low frequency oscillations of
magnetic field on the Earth’s surface generated by irregularities of the ionosphere conduc-
tivity. Geom. Aeron. 41, 327–331 (2001)
183. Sorokin, V.M., Chmyrev, V.M., Yaschenko, A.K.: Ionospheric generation mechanism of
geomagnetic pulsations observed on the Earth’s surface before earthquake. J. Atmos. Solar-
Terr. Phys. 64, 21–29 (2003)
184. Sorokin, V.M.: Wave processes in the ionosphere related to geomagnetic field. Izvestiya
VUZov, Radiofizika. 31, 1169–1179 (1988)
185. Sorokin, V.M., Yaschenko, A.K.: Propagation of the Pi2 pulsations in the low ionosphere.
Geom. Aeron. 28, 655–660 (1988)
186. Sorokin, V.M., Pokhotelov, O.A.: Gyrotropic waves in the mid-latitude ionosphere.
J. Atmos. Solar-Terr. Phys. 67, 921–930 (2005)
187. Rauscher, E.A., Van Bise, W.I.: The relationship of extremely low frequency electromag-
netic and magnetic fields associated with seismic and volcanic natural activity and artificial
ionospheric disturbances. In: Hayakawa, M. (ed.) Atmospheric and Ionospheric Electromag-
netic Phenomena Associated with Earthquakes, pp. 459–487. Terrapub, Tokyo (1999)
188. Sorokin V. M., Hayakawa, M.: On the generation of narrow-banded ULF/ELF pulsations in
the lower ionospheric conducting layer. J. Geophys. Res. 113, A06306 (2008). doi:10.1029/
2008JA013094
189. Sorokin, V.M., Fedorov, E.N., Schekotov, AYu, Molchanov, O.A., Hayakawa, M.: Depres-
sion of the ULF geomagnetic pulsation related to ionospheric irregularities. Ann. Geophys.
47, 191–198 (2004)
190. Sorokin, V.M., Chmyrev, V.M.: Modification of the Ionosphere by Seismic Related Electric
Field. In: Hayakawa, M. (ed.) Atmospheric and ionospheric electromagnetic phenomena
associated with earthquakes, pp. 805–818. Terrapub, Tokyo (1999)
191. Kim, V.P., Khegay, V.V., Nikiforova, V.V.: On possible perturbation of the night ionosphere
E region over large scale tectonic fault. Izvestiya RAN, Fizika Zemli. 7, 35–39 (1995)
192. Bowhill, S.A.: The formation of the daytime peak of the ionospheric F2-layer. J. Atmos.
Terr. Phys. 24, 503–520 (1962)
193. Ondoh, T., Hayakawa, M.: Seismo discharge model of anomalous sporadic E ionization
before great earthquakes. In.: Hayakawa, M., Molchanov, O.A. (eds.) Seismo electromag-
netics: litosphere–atmosphere–ionosphere coupling, pp. 385-390. Terrapub, Tokyo (2002)
194. Ondoh, T.: Anomalous sporadic-E layers observed before M7.2 Hyogo-ken Nanbu earth-
quake; Terrestrial gas emanation model. Adv. Polar Upper Atmos. Res. 17, 96–108 (2003)
195. Yokoyama, T., Yamamoto, M., Pfaff, R.F., Fukao, S., Iwagami, N.: SEEK-2 campaign
measurement of the electric field in the E-region and its association with the QP echoes.
In: Abstracts for the 112th SGEPSS Fall Meeting, pp. 12–13. University of Electro-
Communications, Tokyo (2002)
196. Sorokin, V.M., Yaschenko, A.K., Hayakawa, M.: Formation mechanism of the lower
ionosphere disturbances by the atmosphere electric current over a seismic region.
J. Atmos. Solar-Terr. Phys. 68, 1260–1268 (2006)
197. Fuks, I.M., Shubova, R.S.: Anomaly in ELF signals as response to the low atmosphere
processes. Geom. Aeron. 34, 130–134 (1995)
198. Martynenko, S.I., Fuks, I.M., Shubova, R.S.: Ionospheric electric-field influence on the
parameters of VLF signals connected with nuclear accidents and earthquakes. J. Atmos.
Electr. 15, 259–269 (1996)
146 V.M. Sorokin and V.M. Chmyrev

199. Schunk, R.W., Nagy, A.F.: Ionospheres of terrestrial planets. Rev. Geophys. Space Phys. 18,
813–852 (1980)
200. Parrot, M.: Statistical study of ELF/VLF emissions recorded by a low-altitude satellite
during seismic events. J. Geophys. Res. 99, 23339–23347 (1994)
201. Sorokin, V.M., Chmyrev, V.M., Hayakawa, M.: The formation of ionosphere–magneto-
sphere ducts over the seismic zone. Planet. Space Sci. 48, 175–182 (2000)
202. Hayakawa, M.T., Yoshino, T., Morgounov, V.A.: On the possible influence of seismic
activity on the propagation of magnetospheric whistlers at low latitudes. Phys. Earth Planet.
Inter. 77, 97–102 (1993)
203. Sorokin, V.M., Cherny, G.P.: It is quite possible to monitor typhoons from outer space.
Aerospace Courier. N. 3, pp. 84–87 (1999)
204. Isaev, N.V., Sorokin, V.M., Chmyrev, V.M., Serebryakova, O.N.: DC electric fields in the
ionosphere related to sea storms and typhoons. Geom. Aeron. 42, 670–676 (2002)
Chapter 4
“COMPASS 2” Satellite and Ground-Based
Experiments

Yu.M. Mikhailov, V.D. Kuznetsov, C. Ferencz, L. Bodnar,

V.E. Korepanov, G.A. Mikhailova, L.P. Korsunova, V.V. Khegay,
S.E. Smirnov, and O.V. Kapustina

Abstract This review is devoted to modern methods of earthquake (EQ) pre-

diction. Section 4.2 contains the first results of special satellite “COMPASS 2”
destined for the detection of seism-electromagnetic (EM) effects. A whistler group
in the higher-order guided mode was recorded. Probably it was propagating
between two layers, caused by onion-like structure of inhomogeneities in the
plasma sphere. In Section 4.3, ELF-VLF effects observed over seism-active regions
by the satellite “INTERCOSMOS-24” are considered. It was revealed that the
D-region conductivity decreases during the earthquake preparation at Kp < 3 and
increases during geomagnetic disturbances. The seismic and geomagnetic effects
are assumed to be attributable, respectively, to the increase in near-ground atmo-
spheric conductivity caused by radon emanation during fissuring and to the precipi-
tation of high-energy particles from the inner radiation belt. In Section 4.4, results
of the earthquake precursor occurrence time analysis in quasi-static electric field of
the surface atmosphere on the Kamchatka Peninsula are presented. The propagation
velocity of the precursors and their occurrence time are estimated for different
geophysical conditions. ULF-effects in magnetic fields as a result of aftershocks are
considered. In Section 4.5, relation between precursors in quasistatic electric fields
and in ionosphere parameters is analyzed. This relation reflects the processes of
lithosphere–ionosphere interaction.

Keywords Satellite experiments
Ionosphere
Precursors
Earthquake prediction

Yu.M. Mikhailov (*)

Pushkov Institute of Terrestrial Magnetism, Ionosphere and Radio Wave Propagation, Russian
Academy of Science, IZMIRAN, Troitsk, Moscow region 142190, Russia
e-mail: [email protected]

V.L. Bychkov et al. (eds.), The Atmosphere and Ionosphere, 147

Physics of Earth and Space Environments,
DOI 10.1007/978-90-481-3212-6_4, # Springer ScienceþBusiness Media B.V. 2010
148 Yu.M. Mikhailov et al.

4.1 Introduction

The basic principles of the search methodology of earthquake (EQ) predictions

have been already formulated by Gufeld [1] by next propositions:
1. Proper choice of observation point and fixed parameters
2. Several extraction methods of key parameters
3. Some sign criterion as a combination of time sequences of key parameters
4. Cases of an extra low level of electromagnetic (EM) effects connected
with EQ
At present, it is interesting to deal with special satellites, the main goal of which
is to look for EQ-precursors (COMPASS-2, Demeter). They allow us to carry out a
global control over the seismic situation under consideration. But some difficulty
arises because of slow time-varying effects when preparing EQ. Further observa-
tion of these effects in the case of a quick satellite flight is rather complicated. To
register precursors, it is necessary to understand their nature.
For this reason, the main attention in the present issue is paid to a velocity of EQ
precursors. It has been shown that the same values of velocities (near 1 m/s) appear
in different processes. But this immediately means that precursor registrations
confirm real process under consideration.

4.2 Satellite Experiment “COMPASS 2”

4.2.1 Satellite Investigations

The “COMPASS 2” S/C (Fig. 4.1a) is a space platform consisting of a set of

housekeeping systems including the science instrumentation. It was launched 24
May 2006 on orbit with parameters: apogee 534 km, perigee 402 km, inclination
78,89 degree.
The science instrumentation includes the following units:
l RFA (radio-frequency analyzer)
l Radiation and ultraviolet detector (“TATYANA”) – (high energy particles)
l Total electron content (TEC) detector – GPS receivers for the ionosphere
tomography
l NVK (low-frequency wave complex)
l Dual Frequency Radio Transmitter RBE-150/400 (“MAYAK”).
Main characteristics of the satellite are

Average orbit power 25 W

Mass of the S/C 80 kg
Mass of the science payload 20 kg
4 “COMPASS 2” Satellite and Ground-Based Experiments 149

Fig. 4.1 (a) Satellite COMPASS 2; (b) Receiver SAS2 (NVK-complex); (c) magnetic sensor
LEMI106 HS; (d) electric sensor LEMI502; (e) total electron content detector; (f) RFA-
equipment; (g) block (TATYANA); (h) ultra violet emission detector
150 Yu.M. Mikhailov et al.

4.2.1.1 The Scientific Instrumentation

1. Total Electron Content Detector (IZMIRAN)

Total Electron Content (TEC) detector is designed for high-precision electron
concentration distribution measurements (from the lower ionosphere to the satellite
altitude) obtained by trans-ionospheric sounding using global navigation system
(GPS) spacecraft (Fig. 4.1e).
Experiment aims are
Study of vertical structure of the ionosphere global distribution
Diagnostics of the active experiments in the ionospheric plasma, anthropogenic
factors in the ionosphere, and effects caused by natural phenomena in the Earth’s
atmosphere, including seism-active and magnetic conjugation regions [8]
Main characteristics of the TEC device are
Registration of relative phase shifts of GPS signals at frequencies F1 ¼ 1,250
MHz and F2 ¼ 1,600 MHz
Registration of absolute code delay of the C/A signal, transmitted by GPS
satellites at the frequency F1 ¼ 1,250 MHz
Calculation of the location vector of the TEC device in PZ-90 or WGS-84
reference system
2. Dual Frequency Radio Transmitter RBE-150/400 (“MAYAK”)
The onboard RBE-150/400 (“MAYAK”) transmitter is designed for high-
precision measurements of space–time distribution of the electron concentration
structures above the sensitive regions, as predicted by the orbital transionosferic
sounding (radio beacon method). It consists of two coherent transmitters at 150 and
400 MHz.
3. RFA (Radio Frequency Analyzer) IZMIRAN (Russia), CBK (Poland), IRF
(Sweden)
The scientific objectives assigned to this device are the search for ionospheric
precursors of the earthquakes as well as ionospheric phenomena caused by other
natural and technogenic catastrophes. Its main functions are registering the electric
component of plasma oscillation spectra in the 50 KHz–17.9 MHz frequency band
and obtaining the electron density profiles along the orbit, including regions where
strong earthquakes are expected (Fig. 4.1f).
4. Radiation and Ultraviolet Detector (“TATYANA”)
TATYANA device was designed and manufactured by the Nuclear Physics
institute (NIIYaF) of the Moscow State University (MSU). TATYANA’s main
purpose is to register radiation, cosmic rays, and ultraviolet emission coming from
the Earth’s atmosphere (Fig. 4.1g – radiation detector block, 4.1h – ultraviolet
emission detector).
4 “COMPASS 2” Satellite and Ground-Based Experiments 151

5. NVK (Low Frequency Wave complex)

The general scientific goals of the complete COMPASS mission are the study of
the electromagnetic (EM) activity in the ionosphere concerning the seismic events,
especially the existence of EM precursors of the earthquakes, the detailed investi-
gation and monitoring of the electromagnetic environment of the Earth in the
ULF–VLF bands, as well as the detailed characteristics of the lightning activity
and whistler propagation and space weather-related tasks.
One of the main scientific experiments was the electromagnetic one performed
with the help of advanced Signal Analyzer and Sampler (SAS2-K2), which used
Ukrainian electromagnetic sensors: two spherical electric probes spaced at the
distance of about 4 m producing one (differential) electric signal measured between
these probes and one search coil magnetometer producing one magnetic signal, both
operated in the ULF–VLF bands. The SAS2-K2 and the sensors worked perfectly,
and a good balance of the efficient sensitivities of both electric and magnetic channels
was revealed. It was also noted in the data that a very low noise of both sensors was
obtained, according to the original specifications [2], which confirms good EM
cleanliness of the satellite itself. The general view of SAS2-K2 for COMPASS-2
instrument is given in Fig. 4.1b. Inside the box of SAS2-K2 two identical SAS2 units
were integrated operating in cold reserve mode to increase the reliability of the whole
mission, as it is possible to see in the block diagram of the system.
The main characteristics of the SAS2 wave analyzer are the following:
1. Frequency range (search coil and the electric probes) from 1 Hz to 20 kHz
2. Search coil transfer function from 1 Hz to 1 kHz linear and from 1 kHz to
20 kHz flat
3. Electric sensor transfer function from 1 Hz to 20 kHz nearly flat (Fig. 4.1d)
4. Noise bands
5. Magnetic sensor of 10 Hz–2 pT/Hz1/2, 100 Hz–0.2 pT/Hz1/2, 1 kHz–0.03
pT/Hz1/2, and 10 kHz–0.05 pT/Hz1/2 (Fig. 4.1c)
6. Two electric probes of 10 Hz–40 nV/Hz1/2 and 10 kHz–20 nV/Hz1/2. The weight
of the SAS2-K2 electronic unit is 470 g. The total weight of the system with
sensors etc. is 1,260 g. The size of the electronic unit is 150  70  110 mm3.
The power consumption is 3 W, including the sensors, by this, one electronic
unit (the SAS2-A or the SAS2-B inside the SAS2-K2 box) is in a cold
backup mode.
During the operation of the system it was possible using telemetry com-
mands to select which of A or B unit is active inside the SAS2, the input gain
of the input analog (ULF–VLF) amplifiers and the main operation mode and the
parameters of the selected operation mode including the sampling rate of the
registered signals. The maximum sampling rate was 43.2 kHz for each channel
and the gain of input amplifiers was possible to change from 18 to þ 20 dB in
three steps.
The SAS2 has three different memory modules beside the internal memory of
the DSP. The first is the boot memory (128 KB EPROM) from which the operating
152 Yu.M. Mikhailov et al.

program is loaded after switch-on or power-on reset. The second is a 4 MB SRAM

used for measurement data storage organized as circular buffer and telemetry
buffer. The third is a 64 KB EPROM to set and store the actual parameters and
reference spectrum of the measuring (operation) modes. Using this hardware and
software possibilities the following SAS2 main operation modes are realized:
1. Monitoring of the average EM noise spectra of both channels. The averaging
time can be set by commands from 1 s high-speed monitoring to 10-min long
time averaging. This operation mode runs simultaneously with possibly two or
three operation modes.
2. Event detection. After the processing of the signals coming simultaneously to
both channels the processed signal (spectrum) of one selected channel is com-
pared to a stored (and selected or modified by ground commands according
to the actual onboard EM situation, etc.) reference signal (spectrum). If the
recorded spectrum is higher than the reference one by one or more spectral lines
(see more about these criteria given by Lichtenberger et al. [3]), it is accepted as
a real event and valid trigger. Then, the system reads the recorded signal values
stored in a circular buffer memory before a predefined time of the trigger signal
and follows the trigger signal to another predefined time (e.g., 0.5 s before the
trigger and 1 or 2 s after the trigger). If the registration of the detected electric
and magnetic signals during this time period is finished, the whole record is
rewritten to the telemetry buffer.
3. Periodical, time-controlled data collection. In this mode, the SAS2 detects and
stores the incoming data of both channels with predefined sampling rate (normal
“burst” mode) without triggering an event detection using a command controlled
(predefined) registration time schedule list.

4.2.2 Experimental Results

Despite these difficulties related with submarine launch and short time of work, the
satellite allows to receive interesting scientific results. We present a few results
received by the instruments of the scientific complex and then in detail VLF-data.
Two-frequency receiver GPS – Glonass showed high quality of the definition
of space–time parameters of the satellite. Continuing processing of information
connected with reconstruction of the ionosphere profiles.
High-frequency receiver RFA registered dynamical spectrum in the frequency
range from 0.3 to 18 MHz. In Fig. 4.2a local plasma resonances in frequency range
from 12 to 14 MHz in conditions of high and low impedance of antennae are
presented.
Radiation Detector “TATYANA” measured proton flows with energies 7–15
arrhythmic pulse number of the counter. Plato on curves shows penetration
of energetic particles in the inner magnetosphere on open magnetic force lines
up to altitudes near 400 km. The Earth in proton flows from South was shown
in Fig. 4.2d. Pulses of ultraviolet emission 02.12.06 at 21.04.41 UT were also
4 “COMPASS 2” Satellite and Ground-Based Experiments 153

Fig. 4.2 (a) Local plasma Resonances over WS-America (RFA); (b) SpW (SAS2); (c) FDW
(SAS2); (d) sight of Earth in proton flows; (e) proton flows 7–10 (red) and 15–40 MeV (yellow)
over South Atlantic (altitude 400 km); (f) map of Kamchatka with EQ 28.02.07; (g) FDW and
whistlers 27.02.07 at 21.35 UT (SAS2)
154 Yu.M. Mikhailov et al.

registered. This event is connected with the storm phenomena over Kenya
at nighttime. Double pulses are connected with near-storm discharges (device
DUF).
Low-Frequency Wave complex (NVK) registered electric and magnetic com-
ponents of electromagnetic field in the range of frequencies from 1 Hz to 20 kHz.
Frequently dispersed whistlers have been registered over Indonesia (Fig. 4.2b).
The duration of signals equal to 200–400 ms, D  5, frequency of maximum
 2,500 Hz. Spiky Whistlers (SpW) phenomena was recorded on Demeter and
COMPASS 2 the SAS2, 16.03.2007 (Fig. 4.2c). At 18.03.07 in 17.09.17 UT was
detected a broadening of VLF-transmitter on frequency 12.0 kHz. The band of
signals equals 200–400 Hz. The fluctuations of frequency have a time character-
istic of about 0.05–0.13 that accords with dimensions of uncertainties about
300–500 m. Now there is an actual problem  penetration of strong atmospheric
on satellite altitudes. In times of flight over Central Europe, 11.02.07 10:02 UT
satellite “COMPASS 2” registered strong noise burst in the frequency range from
500 to 2,500 Hz on the latitude of 450 km [4]. An attempt was made to register a
penetration of strong atmospheric signals, which can be considered as precursors
of earthquakes. In Fig. 4.2h a dynamic spectrum of such a signal is represented;
it was registered 27.02.07 at 21:35:49 UT over Kamchatka (Fig. 4.2g) before
the earthquake. It is a frequently dispersed whistler with the dispersion rate of
5.1/s2. Later, a group of multipath whistlers with the dispersion rate of about
75.1/s2 were noticed.
In the paper [5], analysis of an isolated whistler group with very low back-
ground received on “COMPASS 2” was presented. It confirms that every whistler
has a special structure, that is, the spectrogram of these signals has no asymptote
and zero asymptote; however, these signals were predicted by the theory of
UWB propagation in waveguides filled with magnetized plasmas [6, 7] Using the
theoretical model cited above it is possible to make a detailed comparison of
computed and measured data. The form, spectrogram of computed (simulated)
UWB signal and measured signals recorded on board of COMPASS 2 by SAS2
were nearly identical if simulated UWB signal propagates not in a tube, but
between two layers (inside an onionskin-like structure) and it propagates in the
third-order guided form (Fig. 4.3a). The correlation and similarity of these signals
were very high.
Therefore, it is possibly to state, that SAS2 has recorded a whistler group, which
propagated in the guided mode, probably between two layers of plasma inhomo-
geneities in the form of higher (m ¼ 3) order guided mode. So, the new advanced
EM analyzers on board electromagnetically very low-noise satellite has the capa-
bility to detect and identify VLF signals (whistlers) propagating in free space mode
without guiding and in guided modes in waveguiding structures. The general shape
of this recorded waveform (Fig. 4.3a, b) identifies its presence (tube-like or surface,
i.e. onion-like etc.). In our case, there is evidence to confirm the existence of
surface-like (onionskin-like) waveguide structure (Fig. 4.3c) in the magnetosphere;
however, there was no evidence of a tube-like (“spaghetti-like”) structure that
appeared with rather high probability.
4 “COMPASS 2” Satellite and Ground-Based Experiments 155

Fig. 4.3 (a) Sonogram of whistlers, electric component (SAS 2); (b) sonogram of whistlers
magnetic component. At low frequencies curve f(t) reaches zero in finite time; (c) onion-type
waveguide model; (d) frequency histograms of maxima in the whistler spectrum, (e) its relative
intensity at quiet conditions at Kp < 3 (solid line); at Kp < 3 but during the seismically active
period (dash-dotted line); and at Kp >3 in the absence of earthquakes (dashed line)
156 Yu.M. Mikhailov et al.

4.2.3 Conclusions

Data of devices RFA and “TATYANA” are under process. The satellite has high
cleanliness that allows registering of very thin electromagnetic VLF-effects. Wave
analyzer SAS2 has recorded a whistler group in the higher-order guided mode. The
recorded whistlers propagated probably between two layers confirming the exis-
tence of an onionskin-like structure of inhomogeneities in the plasmasphere. These
whistlers propagated probably in the form of a third-order guided mode between
the boundary layers. These ideas, developed in ref. [5] have been confirmed by
“COMPASS 2” data.

4.3 ELF and VLF Electromagnetic Background in Topside

Ionosphere over Seismically Active Areas
(Intercosmos-24 Satellite)

4.3.1 Introduction

This section is devoted to the continuation of our studies of the ionospheric

D-region properties at the earthquake preparation phase on the basis of the observed
“INTERCOSMOS-24” whistler spectra of broadband data, but using a large vol-
ume of experimental data. For comparison, we consider the amplitude spectra of
whistlers during magnetically active and quiet periods in the absence of seismic
activity.

4.3.2 Input Data and Processing Methods

We used broadband records of the electric component of the electromagnetic field

received at the station of the Institute of Terrestrial Magnetism, Ionosphere, and
Radio Wave Propagation (’ ffi 55.5 N, l ffi 37.3 E) from “INTERCOSMOS-24”
to study the response of the lower ionosphere to earthquakes. For this receiving
station, the radio-visibility zone was extended from 25 N to 80 N in latitude and
from 340 W to 80 E in longitudes; it included the two seismically most active
areas; the Mediterranean Sea and the Caucasus with Iran. According to the catalog
(NEIC), the epicenters of the earthquakes that occurred in 1990–1992 had the
following limiting geographical coordinates: ’max ¼ 45 N, ’min ¼ 34 N and l ¼
10 E–31 E (L ¼ 1.35–1.85) for the first area; and ’max ¼ 43 N, ’min ¼ 27.5 N
and l ¼ 43.5 E–58 E (L ¼ 1.3–1.85) for the second area. For our data processing,
we selected records taken in orbits whose latitudinal projection extended from
25 to 45 N, i.e., in the range of L shells 1.3–1.9. The data processing consisted of
4 “COMPASS 2” Satellite and Ground-Based Experiments 157

the following steps. The signal tape records were first examined with an analog
spectrum analyzer. Of the 250 satellite passes from October 1989 through June 1992,
75 append in the longitude sectors 10–30 and 45–60 E. For the subsequent digital
data processing, we selected the most characteristic records on the following three
conditions: in the absence of seismic and geomagnetic activity, at enhanced geo-
magnetic activity but in the absence of earthquakes, and during seismic activity but
for Kp < 3 at daytime. By the analogy with ground-based observations of earth-
quake effects using whistlers, we choose the signal rate and the frequency and
intensity of maxima in the spectra as the quantitative whistler parameters. We used
the spectral density of a calibration sinusoidal signal of frequency 4 kHz and
duration 2 s to quantitatively estimate the whistler intensity. This signal was fed
to the input of the onboard broadband receiver every 8 min simultaneously with
useful signals from the output of the antenna system. Next, we estimated the
whistler’s spectral density at a maximum of the spectrum (Smax).
To carry out a detailed digital data processing, the analog signals were read
from the magnetic tape into a computer. A spectral-time analysis of the read
signals was performed by the method of a sliding time “window”  25 ms in
duration in the frequency range 70 Hz–4 kHz, which yielded a digital spectro-
gram, that is, with one dependence of f(t). Concurrently, we estimated the spectral
density of the maximum Smax2 ( fmax, t). Next, using the f(t) curve near fmax, we
estimated the dispersion D from two points on both sides of this frequency by
assuming that the law D ¼ t/f was valid. For each whistler, we thus had t, fmax,
Smax, and D, which, together with the satellite orbital data, give full informa-
tion about the characteristics of the whistler and its position in space (’, l, h, L,
and LT). We determined the signal rate from these data and the propagation
path of the signal  whether it follows along a short path from the Earth–iono-
sphere waveguide or from the southern hemisphere  from the dispersion. The
next processing step was to construct a statistical frequency distribution of
maxima and the relative maximum spectral density in the whistler spectra for
each satellite pass.
In addition, we calculated the terminator line at ground level and at heights of
60 and 100 km for all passes. The coordinates ’* and l* of the conjugate regions
[9] were estimated from ’ and l at the ends of the orbit projections in the
latitudinal belt under consideration. The terminator curve at 60 and 100 km and
the (’, l and ’*, l* data were used to determine the sunrise and sunset times
in the northern and southern hemispheres to estimate the illumination conditions
for the whistler propagation path from the southern-hemisphere sources. For
each satellite pass, there are also data on seismic and geomagnetic (Kp indices)
activity.
Since the strongest seismic event, both in intensity and in duration, during the
period 1990–1992 under consideration occurred in Iran from June 20 to 30, the
greatest attention in our data analysis is given to this period. According to the data
from seismic Catalog, there were six repeated shocks during this period with Mp in
the range 4.8–6.1 and about 60 with Mp < 4.8.
158 Yu.M. Mikhailov et al.

4.3.3 Main Results

To study the behavior of the D region of the ionosphere during earthquake prepara-
tion and high magnetic activity, we selected 53 turns when the satellite passed a
wide range of longitudes from 10 to 65 E, but mostly during the period that
included the Iranian-earthquake events of 1990. Of them, ten turns were at Kp < 3
during the active seismic period, ten passes were at Kp > 3, and 33 passes were at
Kp < 3 but in the absence of earthquakes. Since the damping coefficients of VLF
waves for extraordinary polarization as they pass into the topside ionosphere from
the Earth–ionosphere waveguide are at a maximum at h  60–90 km and decrease
by three orders of magnitude at h  1,000 km [10], all the whistlers observed at
h  700–2,400 km were included in the data processing.
Quiet background, Kp < 3, no earthquakes. Fractional hop whistlers observed
at morning and day local time, when the diurnal variation of thunderstorm activity
and atmospheric interference near the ground exhibits a minimum, which is
deeper than that in the morning, are of greatest interest in the analysis. These
patterns of behavior showed up during satellite observations, when at Kp < 3
under seismically quiet conditions at heights up to 2,400 km fractional hop
whistlers, irrespective of the longitude, were either completely unobservable or
were observed very rarely (up to 4/min) and were less intense; only ELF compo-
nents in the frequency band from 200 Hz to 2 kHz were present in their spectra. In
the morning, ELF hisses were observed in all orbits in the frequency band
400–500 Hz. In addition, a secondary maximum was also observed in the spec-
trum of fractional hop whistlers over the Mediterranean Sea at maximum seasonal
thunderstorm activity (May), but was of lower intensity than that for the first
ELF. The whistler rate and intensity before the sunset increase compared to the
morning and daytime.
Geomagnetic effects in the D region. To estimate the geomagnetic effects in
the D region of the ionosphere in such low L shells (1.4 < L < 2), we considered
two periods of high geomagnetic activity: May 19–21, 1990, and August 22–23,
1990, at Kp  5–6. A data analysis for these periods revealed that exclusively
fractional hop whistlers, whose amplitude spectra exhibited only ELF compo-
nents with an upper cutoff frequency 0.7–1.5 kHz, that is, lower than in the signals
under quiet-time conditions were observed. These signals had strongly diffuse
traces on the f(t) spectrograms. As the Kp index decreased, the whistler rate
( 6–8/min) was essentially constant, but the ELF spectrum extended to higher
frequencies (up to 2 kHz). For all passes, ELF hisses were observed in the band
300–400 Hz at Kp > 3 [16].
Seismic effects in the D region. We analyzed the whistler spectra for seven
satellite passes over Western Iran from June 21 through 29, 1990; one pass over
Southern Iran on November 6, 1990; one pass over the Mediterranean Sea on
January 18, 1990; and one pass over Romania on May 29, 1990, during earthquake
preparation (Mp  5–6.1, h  3 and 30 km). Our analysis of the whistler amplitude
spectra revealed the following: first, the extension of the ELF spectra to higher
4 “COMPASS 2” Satellite and Ground-Based Experiments 159

frequencies (up to 3 kHz) compared to the quiet background; second, the appear-
ance of VLF spectral components; third, the observation of isolated fractional
hop whistlers of anomalously high intensity (Smax)2 with a broad spectrum DF ¼
100–10,000 Hz without the minimum at frequencies of 2–3 kHz which is typical
for the spectra of whistlers observed in the Earth–ionosphere waveguide; in the
evening, these whistlers were accompanied by echo signals reflected in the southern
hemisphere, suggesting that magnetospheric ducts are formed over seismically
active areas at the earthquake preparation phase.
Figure 4.3d shows the frequency distributions of maxima in the spectra of
fractional hop whistlers with a dispersion D < 20 normalized to the total number
of maxima for three different geophysical conditions: quiet background (S ¼ 69),
geomagnetic activity (S ¼ 54), and seismic activity (S ¼ 300). Figure 4.3e shows
the frequency distribution of the relative intensity of maxima for the same geophys-
ical conditions. The following general trend clearly shows up both in the sample
passes and in the histograms: the decrease in the upper cutoff frequency of the ELF
spectra at Kp> 3 and its increase during the seismically active period compared to
the quiet background values; the appearance of additional maxima at frequencies
above 3 kHz during the seismically active periods. At the same time, there is no
strong dependence of the whistler intensity on D-region conditions, and the number
of whistlers during the seismically active period increases at the expense of weak
signals (we excluded anomalously strong isolated whistlers when constructing the
histograms).

4.3.4 Discussion

Our results show that fractional hop whistlers with dominating ELF components in
their spectra are observed at daytime at heights 700–2,400 km in the absence of
seismic and geomagnetic disturbances. VLF components are also observed in the
spectra of daytime whistlers over the Mediterranean Sea at maximum thunderstorm
activity, but they are weaker than the ELF components. This implies that at daytime,
VLF waves undergo a stronger damping as they pass into the topside ionosphere
than do ELF waves. This conclusion is consistent with the theoretical calculations
[11], according to which the transmission coefficients of electromagnetic waves in
the frequency range 1–30 kHz have their maximum at f ¼ 5 kHz and change only
slightly on both sides of this frequency. At the same time, the transmission coeffi-
cient at daytime is, first, an order of magnitude smaller than that at nighttime, and,
second, being at a maximum at  1 kHz (for a definite ionospheric model), it steadily
decreases with frequency, for example, by a factor of 5 at f  15 kHz. As the
damping in the D region increases, this curve shifts to lower frequencies.
Based on the latter theoretical conclusion, we can qualitatively explain the
amplitude spectra during seismically active and magnetically active periods. The
shift of the upper cutoff frequency of the ELF components in the spectra of
fractional hop whistlers to higher frequencies at the earthquake preparation
160 Yu.M. Mikhailov et al.

phase and the appearance of VLF components in their spectra suggest that the
damping of these waves decreases as they pass into the topside ionosphere. With
increasing geomagnetic activity, the upper cutoff frequency in the whistler ELF
spectra steadily decreases and becomes lower than 1 kHz at Kp  6. This suggests
that the damping of ELF waves is enhanced as they pass into the topside iono-
sphere. It thus follows that the D-region of the ionosphere is modified by different
agents during seismic and geomagnetic activity. During geomagnetic disturbances,
high-energy electron beams, which produce additional ionization in the lower
ionosphere, are the main agents responsible for the increase in the D-region
conductivity. During the preparation of earthquakes, several hours before their
active phase, an abrupt increase in the count rate of precipitating electrons from
the inner radiation belt with energies 0.7–25 MeV [12] and 40–250 keV [13] was
actually detected. Moreover, parameters of the seismogenic variations over the area
of the Iranian earthquake in 1990 were estimated by using the Meteror-3 satellite
counters of electrons with energy 0.7–25 MeV [14]. It turned out that the repeated
shock of June 21 with Mp  6.1 was preceded by an anomalous burst of particles
13 h before, while the repeated shocks with a smaller magnitude (June 21 and 22)
were preceded by such a burst 2–4 h before. The appearance of these bursts did not
coincide in time with the anomalous increase in intensity of the fractional hop
whistlers observed from INTERCOSMOS-24, particularly during pass 3309 on
June 21 [10]. These experimental facts suggest that the changes in D-region proper-
ties during the earthquake preparation are caused by an agent other than particles.
An attempt to explain the increase in the signal intensity of very-long-wave
ground-based transmitters over the area of the preparing earthquake by the change
in near-ground atmospheric conductivity due to the radioactive radiation associated
with radon emanation during fissuring was first made by Martynenko [15]. Accord-
ing to their estimate, this causes the D-region electron density to increase only
slightly and the frequency of their collisions with neutral particles to decrease
appreciably. As a result, the lower-ionosphere conductivity decreases, as does the
damping of VL waves as they propagate in the Earth–ionosphere waveguide and as
they pass from the waveguide into the topside ionosphere.

4.3.5 Conclusion

An analysis of the amplitude spectra for the fractional hop whistlers observed from
INTERCOSMOS-24 at daytime under various geophysical conditions has revealed
the following [16]:
During a quiet period, in the absence of seismic and geomagnetic disturbances,
the signal spectra exhibit mostly ELF components ( fc < 3 kHz) with its upper
cutoff frequency fc  1.5 kHz. This result is consistent with theoretical views of
the properties of ELF–VLF-wave passage into the topside ionosphere from the
Earth–ionosphere waveguide.
4 “COMPASS 2” Satellite and Ground-Based Experiments 161

With growing geomagnetic activity, the upper frequency of the ELF components
shifts to lower frequencies (700–1,000 Hz), which, according to theoretical views,
suggests that the damping of ELF waves enhances as they pass through the
D-region of the ionosphere.
By contrast, during a seismically active period, in the earthquake preparation
phase, the upper cutoff frequency of the ELF components in the whistler spectra
shifts to higher frequencies (2.5–3 kHz), and VLF components emerge, that is, the
damping of ELF and VLF decreases as they pass through the D-region.
Our analysis of the whistler amplitude spectra has allowed the seismic and geo-
magnetic effects in the region of the ionosphere to be clearly separated. They are
related differently with wave damping, other words, D-region conductivity, they are
attributing different physical agents which modify this on. The additional ionization in
the D-region provided by the precipitation of high-energy particles from the inner
radiation belt during geomagnetic disturbances does cause the damping of ELF and
VLF waves to be enhanced, which is confirmed by experiment.
Other phenomena should be invoked, which allow a mechanism of the lithosphere–
ionosphere interaction in its lowest part to be loped. There is no such mechanism
so far. In this section, there was a hypothesis about the change of whistler character-
istic as a result of lower ionosphere (D-layer) conductivity variations. But it is
necessary to prove that these changes are related to lithosphere processes. It is a
complicated problem. In Section 4.5, we will try to solve this problem experimen-
tally with the help of ground-based measurements.

4.4 Quasi-static Electric and ULF Magnetic Fields

on Kamchatka

4.4.1 Experimental Results

We analyzed earthquake precursors observed in various periods in the quasi-static

electric field in the surface atmosphere on the Kamchatka Peninsula at the Institute
of Cosmophysical Research and Radio Wave Propagation, RAS Fur East Division
(Paratunka settlement; 52 58.30 N;158 14.90 E). Figure 4.4a, b illustrate two cases of
the precursors occurrence before an earthquake.
A bay-like drop in the quasi-static field down to 300 V/m was observed before
the earthquake of March 6, 1992 (M ¼ 6.1) at the epicentral distance of 130 km,
with the epicenter in the Avachinskii Bay and a source depth of 30 km (Fig. 4.4b)
[17]. A similar decrease in the value of the quasi-static electric field was also
observed before the earthquake of September 18, 1999 (M ¼ 6.0; a source depth
of 60 km) at the epicentral distance of 190 km (Fig. 4.4a) [18]. It is noteworthy that
the precursors had similar shapes, but their premonitory times were different: 10 h in
the first case and 29 h in the second. Assuming that the time difference depends
on the precursor propagation velocity, we estimate the latter as V ¼ 0.87 m/s [19].
162 Yu.M. Mikhailov et al.

Fig. 4.4 Variations in the quasi-static electric field Ez from Kamchatca measurements in periods
of nucleation and developments of the earthquakes 19.09.99 (a) and 06.03.92 (b). A and A0 are
tentative earthquake precursors, and the interval B is the pulsation time of Ez in the neighborhood
of an earthquake; (c) daily power spectra of horizontal component H for 01–10.03.92 in range of
periods 6–40 min. Harmonics with periods 20 and 35 min are seen
4 “COMPASS 2” Satellite and Ground-Based Experiments 163

In 1983, Zubkov, in his paper [20] informed that at earth current observations, the
velocity of earthquake precursors is found to be about 0.3 mc1.
Korsunova and Khegai [21] analyzed occurrence times of earthquake precursors
in the variations of critical frequencies of the ionospheric F2 layer in the seismically
active Pacific Ocean region. For strong crustal earthquakes with sources shallower
than 50 km, the precursor propagation velocity was found to be of the order of
1 m/s. It is unclear as to why these values coincide since the precursory effects were
discovered in different parameters and different media. These questions will be
considered in detail in Section 4.5.
Now we consider Fig. 4.4 in greater detail and analyze the seismic situation on
the Kamchatka Peninsula at the beginning of March 1992. Two series of earth-
quakes occurred in this period: on March 2 and 5 (according the NEIC catalog).
Sequence of 2 March includes three shocks (12.29.39, 12.49.18, 14.07.59 UT)
with deeps 38, 33, 34 km and magnitudes (6.9, 4.6, 6.0). Interval between shocks is
near 20 min. Sequence of 5 March includes three such shocks (14.39, 15.14, 16.01
UT) with deeps (45, 33, 39 km) and magnitudes (6.4, 4.7, 4.4). The interval between
shocks is near 40 min.
We analyzed by spectral method [18] a horizontal component of magnetic field.
The results of data analysis obtained during 1–10 March (1992) are presented in
Fig. 4.4c, where on the right-hand side of the vertical the time scale for this interval is
given from the bottom upwards. Horizontal lines are concerned with EQ, whose
magnitude exceeds 5. In doing so, the corresponding scale is at the bottom. Special
attention here should be paid to some sequences consisting of two or three above-
mentioned shocks for EQ 3.03.92 and 6.03.92. On horizontal lines one can see the
daily spectra of the H-component of the magnetic field occurred at Petropalovsk-
Kamchatsky’s station in the case when the time period was varying from 6 to 40 min.
It is interesting to note that the 35 min-harmonics could be roughly put in correspon-
dence with the time interval between the shocks of EQ. These harmonics either
appears most brightly or responds with some delay to the occurrence of EQ. Unfor-
tunately it is difficult to evaluate correctly this delay in light of the small time interval
of available data. Nevertheless from preliminary estimates one can conclude that this
delay corresponds to the precursor propagation velocity of 1 m/s order.

4.4.2 Conclusions

The analysis of the properties of earthquake precursors in the quasi-static electric

field of the surface atmosphere has led to the following conclusions:
– The precursor propagation velocities vary within the range of 0.87–1.8 m/s and
depend on the focal depth.
– In electromagnetic signals, harmonics with equal periods of interval between
foreshocks were found.
164 Yu.M. Mikhailov et al.

4.5 An Experimental Confirmation of the Relation Between

Precursors of Earthquakes in Near-Earth Electric Fields
and in Ionosphere Parameters

4.5.1 Introduction

Appearance of seism-ionosphere disturbances before/at earthquakes inspires a con-

siderable interest for understanding of litho-, atmo-, ionosphere coupling, so for
possibility of earthquake prediction. Two mechanisms of such coupling are dis-
cussed: electrodynamics and acoustic gravity wave (AGW) influence [22]. Thereby
precursor effects are very different, having different duration and deviation from the
background level; therefore, it is necessary in each occasion to choose available
coupling mechanism. So, for interpretation of disturbances with periods T1.5–2.5 h
in daily variations of F2 and Es-layers frequency parameters, there is often a chosen
hypothesis about the influence of those scales on ionosphere by AGW, increasing of
activity which was observed during the seism-active periods [23].
However, in ref. [24], it was obtained that propagation of the ionospheric
disturbances with t ¼ 2–3 h, identified as medium-term precursors of earthquake,
is realized with low velocities (4–8 km/h), compared with the earth core seism-
genic disturbance velocities described in ref. [25]. Such velocities by two orders of
magnitude are smaller than AGW velocities and therefore this hypothesis cannot
serve as the satisfactory explanation of the observed ionospheric disturbances. On
the other hand, the appearance of anomalously high Es at h ¼ 120–140 km in the
period of the earthquakes preparation, which is the basic morphological sign during
the identification of ionospheric precursors [24], corresponds to the model calcula-
tions, that consider the action of electric fields on the ionosphere [26, 27].
Simultaneous measurements of the electric field and ionosphere parameters can
play an important role in the understanding of such coupling nature in the seism-
active region. Such measurements are realized on geodynamic area in Paratunka
village, Kamchatka region, at the Institute of Cosmophysical Research and Radio
Wave Propagation of Far East Branch of Russian Academy of Science. The first
results of simultaneous measurements of the vertical electric field component near
the Earth’s atmosphere and ionosphere parameters at EQ-preparation were pub-
lished in ref. [28], where coupling between F2-variations and the electric field has
been considered. However, in this paper only one ionosphere parameter  the
critical frequency of F2-layer – was considered, and clear correlation between Ez-
and F2-variations has not been obtained in the result of mask influence of earth
magnetic field disturbances. Anomalies of electric field in the atmosphere, which
may be considered as the precursors, and the velocity of their propagation
have been discussed in ref. [29]. It was stated that virtual velocities of theirpro-
pagation are equal to 1–3 km/h, that is, close to virtual velocities of ionosphere
precursors propagation, according to ref. [24]. This fact correlates with this paper’s
purpose to find the coupling between the atmospheric electric fields and ionospheric
4 “COMPASS 2” Satellite and Ground-Based Experiments 165

parameters in the periods of earthquake preparation. To find this coupling for search
of ionosphere precursors, a new method was applied, which took into account
the simultaneous measurements of Es and F2 parameters. It used Paratunka ionos-
phere data (’ ¼ 52 58.30 N; l ¼ 158 14.90 E) during February– March 1992,
September–October 1999, and September–October 2002. During September–October
1999, full processing of ionosphere observations with sampling time of 30 min was
made; for other months, there has been formed an hourly data of this station, obtained
from archive of NOOAA’s National Geophysical Data Center [31]. Processing the
ionosphere data in more detail is necessary since the appearance of Es in considered
time intervals (equinox) is a rare event, and hence the existence of two sporadic
formations is possible, which do not reveal in hourly data of the archive.

4.5.2 Method of the Analysis

In ref. [24], the ionospheric earthquake (IPE) precursors identification method

proposed earlier in ref. [30] is tested. The main criterion for detecting precursors
is to serve the appearance of an abnormally high-layer Es with a duration (t) of
about 2 h in excess of 10 km or more background values of h0 Es.
Besides, only those events are taken into account, when increasing h’Es are
accompanied by bursts in frequencies parameters of Es and F2 of the same duration.
This method allowed the detection of IPE for 33 strong EQ with M 6 in Japan in
the period from 1985 to 2001.
For Kamchatka weaker EQ are typical; therefore, weaker ionosphere effects
of EQ premonitory are expected. It is connected with the fact that values of
seism-ionosphere disturbances depend on a magnitude (M) of future EQ [30].
Nevertheless, the criteria of precursors determining proposed above were used for
Kamchatka also. But in contrast to the cited work, average values on ten magneto-
quiet days, which preceded earthquake were used as the background values of the
ionospheric parameters. It is connected with the fact that at considered latitudes
high influence of the magnetic disturbances on parameters of F2 layer is observed.
For the identification of the supposed IPE the instantaneous deviations values
of the h0 Es, fbEs, foEs, and foF2 from background values for each hour (and for
half-an-hour interval for the selected periods of 1999) were calculated: Dh0 Es ¼
h0 Es – (h0 Es)mean, DfbEs ¼ fbEs – (fbEs)mean, DfoEs ¼ foEs – (foEs)mean, DfoF2 ¼ foF2
– (foF2)mean. For the frequency parameters, we calculated the deviation relative
values: DfbEs/(fbEs)mean, DfoEs/(foEs)mean, and DfoF2/(foF2)mean. The analysis was
fulfilled from the moment of the earthquake to the time of the precursor appearance.
All events, which coincided with higher mentioned signs, has been marked [24].
In Fig. 4.5a there are changes of base ionosphere parameters with sampling time
30 min and Ez-component of quasistatic electric field in preparing period of
EQ 18.09.99 with M ¼ 6.0. A distance from the epicenter to the observation point
(R) equals to 190 km. The moment of earthquake is indicated by a dashed line arrow.
166 Yu.M. Mikhailov et al.

Fig. 4.5 (a) Changes of vertical electrical potential gradient (panel 1), ionosphere parameters
(panels 2–4) and Kp-index (panel 5) in preparing period of 18.09.1999 EQ; (b) variations of
ionosphere parameters (panels 1–3) an Kp-index (panel 4) in the period of preparing of
EQ 05.10.1999. Different groups of supposed IPE have been marked by I and II; (c) the logarithm
of the product of the lead-time (DT) up to the epicentral distance (R) in the dependence on the
magnitude of earthquake (M) for two groups of assumed precursors of EQ and approximating lines
(solid line – IPE, dashed line – EPE)

The time of electrical precursor appearance was defined on time registration of

negative electric field bay shape curve at time Ez-variations, but IPE- on registration
anomaly high Es (type h – thick dotted line on Fig. 4.5a). Blackened bursts,
corresponding to directed higher criteria, are presented as precursors. On lower
panel there have been carried the variation of Kp-index at same time interval. It is
4 “COMPASS 2” Satellite and Ground-Based Experiments 167

evident that the preparing of EQ 18.09.99 is accompanied by magnetic distur-

bances. That is why the positive deviation in the frequency parameters of Es and
F2 layers are very small, but their appearance in the same period of time as the
anomalies in the Ez and h0 Es, and violation of the observed trend toward negative
deviations in foF2 show their seismogenic nature.
However, when the data on Es-layer are absent, the identification of IPE only
with foF2 is very complicated; it is indicated in ref. [28].
Comparison of ionospheric parameters variations and Ez-component of electric
field shows that precursor in electric field appeared few hours earlier than in
ionosphere. Besides, a few hours earlier EQ are noticed bursts in Es-parameters,
which coincide with increasing of Ez.
Analysis of time ionosphere parameters variations in the absence of noticeable
geomagnetic disturbances (when the average daily magnetic index Ap  17 nT)
shows that it was possible to select two groups (I, II) of Es-parameters essential
deviations, which satisfied to morphological signs of IPE-identification, defined in
[24], distinguish on lead-time of EQ-moment, IPE of first and second orders on
lead-time. It is illustrated by Fig. 4.5b, in which are given one half hours data, and
two groups of IPE. It is evident that these groups differ not only in the earthquake
lead-time (DTIPE), but also the deviations values within groups and their duration.
The most deviations in the ionospheric parameters are noted in the second group
for the Es-layer. The duration of bursts and their numbers in the second group are
greater than in group I, for which individual bursts are typical.
Quasistatic electric field variations at September–October 2002, presented in ref.
[29] were compared with corresponding variations of Es and F-layer calculated on
hourly data. The onset of precursors and time interval from the moment of their
appearance to the main shock (the lead-time for EPE, DTEPE) was determined.
Table 4.1 contains results of such a comparison and so EQ-parameters. Unfortu-
nately, for March 1992 electric field data are absent, but for EQ 5.10.1999 pre-
cursors of II group were not identified (see dashes in Table 4.1).
By stars there are marked precursors, identified before in ref. [14] only by
electric field measurements. For other events, there are identified bay-shape dis-
turbances of electric field during day interval centered IPE relatively. Moment of
appearance of such disturbance with maximum amplitude and DT are indicated in
Table 4.1. It is seen that Ez anomalies appear before the appearance of ionospheric
disturbances. Therefore one can suppose that appearance of two groups of different
precursors corresponds to different phases of EQ preparing.

4.5.3 Results and Discussion

Results of analysis of the ionospheric parameters variations for before EQ periods

(for 6 EQ) show that it is possible to select two groups of bursts with t  1–2
h duration, which correspond to precursor signs, and, as a rule, are accompanied with
specific “bay-shapes” in Ez electric component curves. Appearance of such bursts
168 Yu.M. Mikhailov et al.

Table 4.1 The characteristics of ionospheric and electric EQ-precursors

Date and time M R, km IPE EPE
of EQ, UT Date and DT, days Date and DT, days
time, UT time, UT
Group 1
08.10.02 5.0 120 06.10.02 1.5 06.10.02 1.4
09 h 19 min 22 h 23 h
03.10.02 5.2 280 02.10.02 0.8 02.10.02 0.8*
15 h 57 min 20 h 21 h
20.10.02 5.3 110 16.10.02 3.2 16.10.02 3.4
01 h 35 min 22 h 15 h
05.10.99 5.6 190 02.10.99 2.6 02.10.99 2.7
05 h 02 min 15 h 30 min 12 h
02.03.92 6.0 160 25.02.92 5.8 – –
14 h 08 min 19 h
18.09.99 6.0 190 14.09.99 4.6 14.09.99 4.7
21 h 29 min 07 h 30 min 06 h
05.03.92 6.1 130 21.02.92 13.2 – –
14 h 39 min 10 h
16.10.02 6.2 160 04.10.02 11.6 04.10.02 11.8
10 h 12 min 20 h 14 h
Group 2
08.10.02 5.0 120 07.10.02 0.5 07.10.02 0.7*
09h 19m 20 h 17 h
03.10.02 5.2 280 03.10.02 0.4 03.10.02 0.5
15 h 57 min 06 h 04 h
20.10.02 5.3 110 18.10.02 1.1 18.10.02 1.1*
01 h 35 min 23 h 23 h
05.10.99 5.6 190 04.10.99 1.2 – –
05 h 02 min 00 h.
02.03.92 6.0 160 29.02.92 1.9 – –
14 h 08 min 15 h
18.09.99 6.0 190 17.09.99 1.1 17.09.99 1.2*
21 h 29 min 19 h 16 h 30 min
05.03.92 6.1 130 02.03.92 2.7 – –
14 h 39 min 22 h
16.10.02 6.2 160 14.10.02 1.8 14.10.02 1.5*
10 h 12 min 14 h 22 h

took place during daytime. The duration of the bursts into Es for the second group, as
a rule, has a duration t  1.5–2 h, which is somewhat greater than that for the first
group. The most considerable deviations in ionospheric parameters have groups with
the short lead-time, which we signed as IPE for second group (II). As it follows from
Table 4.1, for the same epicentral distances, the precursors in both groups appear
earlier, if the value M of the subsequent earthquake is greater. This property is
characteristic for those investigated by us before IPE [30]. However, the existence
of two groups of IPE, which are distinguished not only by DT but also by the value of
deflection, is detected by us for the first time. For Ez-component of electric field
two groups of specific disturbances before EQ are selected, which differ by the
value and lead-times. These times are given in Table 4.1, from which it follows that
4 “COMPASS 2” Satellite and Ground-Based Experiments 169

Ez-disturbances were observed in the same days, when the IPE were noticed, although
times of the Ez-bursts appearance were different from IPE. Depending on the time
of appearance they are referred in the Table 4.1 as groups I or II, respectively.
The analysis of values of Ez-disturbances shows that in group II, there are
predominated disturbances of Ez with negative swings and amplitudes ranging
from 200 to 1,500 V/m. In group I deviations of Ez are small with alternate
signs
(150–300 V/m) bays shape disturbances with Dt  1–1.5 h. These differ-
ences in amplitudes and their durations are analogous to the earlier mentioned
peculiarities of IPE for the same group. Therefore, one can suppose that appearance
of two groups of different precursors (in EPE and IPE) corresponds to different
phases of EQ preparation.
It is necessary to note that precursors of EQ in electric fields with long lead-time
(3–26 days) for EQ with M ffi 5.0–6.4 have been observed in China [32] and also
identified by us as precursors (IPE) of the first order. For Kamchatka EQ in
Ez-variation, reliable precursors in the time interval of 1 day before EQ are selected
[33, 34], which correspond to characteristics of IEP of the second order according to
our classification. It is possible that this is connected with a circumstance that
values of seismogenic anomalies in Ez of the second group usually by several times
exceeds changes of Ez of the first group, so that they are undistinguished on the
background of daily variations. The longest lead-times of EQ (2–3 and 6 months)
were exposed in gradients and phase velocities ULF ( f ¼ 0.03–0.1) of geomagnetic
disturbances before EQ with M ¼ 5.8 and 6.4 in Japan for epicenter distances
R < 150 km [35]. Though the nature of these anomalies may be other than those
detected in the ionosphere and vertical gradient electrical potential, their appear-
ance shows strong evidence for longer time of EQ preparation, including a few
stages. Trends in the logarithm of the product of earthquake lead-time (DT) up to
epicentral distance (R) depending on the earthquake magnitude (M) for both groups
of the assumed precursors are shown in Fig. 4.5c (points – IPE, squares – EPE).
The approximating lines are carried out on the basis of the least squares method.
For precursors of the first group next expressions are obtained:

lgðDT
RÞIPE ¼ 0:81M  1:83; lgðDT
RÞEPE ¼ 0:81M  1:81 (4.1)

For precursors of the second group, which are nearer to the moment of EQ we
have:

lgðDT
RÞIPE ¼ 0:54M  0:79; lgðDT
RÞEPE ¼ 0:35M þ 0:25 (4.2)

It is seen from Fig. 4.5c that obtained empiric dependences are close especially
for the first group. This confirms the relationship of Ez-anomalies seismogenic
origin to the appearance of disturbances in the ionospheric parameters, which can
be attributed to ionospheric precursors of earthquakes.
Besides that, formulae (4.1) are close to that, obtained in ref. [25]:

lgðDT
RÞ ¼ 0:72M  0:72 (4.3)

170 Yu.M. Mikhailov et al.

for precursors of EQ, propagating in the earth crust. This allows to establish a
connection of them with the same phase of preparing EQ. The differences in the
constant term in formulas (4.1)–(4.3) reflect, apparently, the structural special
features of the earth’s crust in the earthquake preparation zone.
As it was shown in ref. [24], on base (4.1), (4.2) it is possible to estimate the
apparent velocities of disturbances from the seismic source in the atmosphere and
the ionosphere. At propagation of disturbance from 100 to 200 km for M ¼ 6.0 we
obtained a velocity V ¼ (0.7–0.8) km/h for the first group: for the second group we
obtained VIPE ¼ 3 km/h; VEPE ¼ 3.7 km/h. For the same magnitude of EQ and
directed epicenter distances we have V ¼ 0.3 km/h in accordance with ref. [24].
Comparison of obtained velocities shows that disturbances in ionosphere and Ez in
the first group follow the border of preparing EQ region. The apparent velocities of
the seism-genic disturbances propagation in group II practically coincide with the
velocities of disturbances in the ionosphere according to the data of the stations of
vertical sounding in Japan (4.4
3.3) km/h [24] and with the obtained earlier
velocities of EPE in ref. [29]. Close times of IPE and EPE appearance for simulta-
neous observations in the same point of observations, the correspondence of
disturbance structure for two groups of precursors, and propagation’s velocities of
disturbances in the atmospheric electric fields and in the ionosphere are the experi-
mental confirmation of the physical mechanism of seism–ionospheric interaction,
based on the modification of near Earth surface quasi-electrostatic fields and their
further action on the ionosphere [36]. As it follows from this model, the emanation
of radioactive gases from the ground water and the earth’s surface in the period of
the earthquakes preparation is one of the important factors of impact on the surface
quasi-electrostatic field.
It is possible to suppose that the region of exit of high radioactive gases follows
the broadening border of EQ preparing zone on the earth’s surface, formed by
water arrival in microtransform fraction zone of the stress part of the earth’s core.
Indeed in many occasions few days before EQ an increase of radon concentration
in waters and surface layers of earth at far distances from the epicenter (up to
450 km) was observed [37,38]. Next, tests of physical mechanism of lithosphere–
ionosphere, proposed in ref. [36], require organization of complex experiments on
simultaneous measurements in several points of radon and aerosols concentra-
tions, atmospheric electrostatic fields, and ionosphere parameters. Geodynamic of
Kamchatka region where seismic activity is high is in this respect most promising
in Russia.

4.5.4 Conclusions

The fulfilled investigation allows us to make the following conclusions:

1. The empirical dependences which reflect the connection between the lead–time
of EQ, the distance to the epicenter from the observation point and the
4 “COMPASS 2” Satellite and Ground-Based Experiments 171

magnitude of earthquake are obtained. The similarity of these dependences

for the IPE and EPE confirms their interrelation in the process of lithospheric–
ionospheric interaction during the preparation of the earthquakes. As a rule EPE
appear earlier than IPE.
2. In the Kamchatka region two groups of earthquake precursors in the ionosphere
and in the near Earth surface electric fields with different lead–time of EQ (DT)
from several hours to 2 weeks are selected. The amplitudes of precursors in the
group, which is nearer up to the moment of the earthquake, are substantially
greater.

Acknowledgments The authors thank Yu.Ya. Ruzhin, V.S. Dokukin, L.N. Leshenko, G.I.
Druzhin, V.V. Bogdanov, O.E. Ferencz, D. Hamar, Ja. Lichtenberger, and P. Steibach for helpful
discussions.

References

1. Gufeld, L.I.: Seismic process physical-chemistry aspects RSA, Academia of Cosmonavtics,

IPE. Moscow, pp. 153 (2007)
2. Korepanov, V.E.: The modem trends in space electromagnetic instrumentation. Adv. Space
Res. 32(3), 401406 (2003). doi:10.l016/S0273-–1177(03)9020-8
3. Lichtenberger, J., Ferencz, O.E., Bodnar, L., Hamar, D., Steinbach, P.: Automatic whistler
detector and analyzer system: Automatic whistler detector. J. Geoph. Res., 113, A12201
(2008). doi:10.1029/2008JA013467
4. Mikhailov, Yu.M., Mikhailova, G.A., Kapustina, O.V., Druzhin, G.I., Smirnov, S.S., Ferencz,
C.H., Lichtenberger, Ja., Bodnar, L., Korepanov, V.: E.: COMPASS 2 Satellite and ground
base VLF-experiments. In: Abstracts. AIS-2008 “Atmosphere, Ionosphere, Safety”, July
7–12, Kaliningrad, pp. 31 (2008)
5. Ferencz, O.E., Bodnar, L., Ferencz, C., Hamar, D., Lichtenberger, J., Steinbach, P.,
Korepanov, V., Mikhailova, G., Mikhailov, Yu, Kuznetsov, V.: Ducted whistlers propagating
in higher order guided mode and recorded on board of COMPASS-2 satellite by the advanced
signal Analyzer and Sampler SAS2. J. Geoph. Res. 114, A003213 (2008). doi:1029/2008.
JA013542, (2009)
6. Ferencz, O. E., Steinbach, P., Ferencz, C., Lichtenbcrger, J., Hamar, D., Berthelier, J.J.,
Lefeuvre, F., Parrot, M.: Full-wave modeling of long subionospheric propagation and frac-
tional hop whistlers on electric field data of the DEMETFR satellite, paper presented at
International Symposium on DEMETER, Cent. Nat. d’Etud, Spalialcs, Toulouse, France,
14–16 June (2006a)
7. Ferencz, O. E., Steinbach, P, Ferencz, C., Lichtenberger, J., Parrot, M., Lefeuvre, F.: UWB
modeling of guided waves in anisotropic plasmas, paper presented at 2nd VERSIM Workshop
2006, Sodankyla Geophys. Obs., Sodunkyla, Finland (2006b)
8. Ruzhin, Yu.Ya., Zakharova, O.K.: Magnetic conjugation as a subject of investigation by the
method of active experiments. Preprint N.23 (970), IZMIRAN, Moscow (1991)
9. Al’pert, Ya.L.: Radio-wave propagation and the ionosphere. Akad. Nauk SSSR, Moscow
(1960)
10. Mikhailova, G.A., Golyavin, A.M., Mikhailov, Yu.M.: Dynamic spectra of VLF emissions in
the topside ionosphere connected with Iranian earthquake of 21 June 1990 (The Intercosmos-
24 Satellite). Geomag. Aeron. 31(5), 801 (1991)
11. Aksenov, V.I.: On the origin of the electromagnetic waves of extremely low frequency
through the ionospheric plasma. Radiotekhn. Elektron. 11(6), 1030 (1966)
172 Yu.M. Mikhailov et al.

12. Aleshina, M.E., Voronov, S.A., Gal’per, A.M., et al.: On the relationship between the
positions of seismic centers and regions of high-energy particle precipitations from the
radiation belt. Kosmich. Issled. 30(1), 79 (1992)
13. Gal’perin, Yu.L, Gladyshev, V.A., Georgio, N.V., et al.: Precipitation of energetic trapped
particles in the magnetosphere over the epicenter of a preparing earthquake. Kosmich. Issled.
30(1), 89 (1992)
14. Ginzburg, E.A., Malyshev, A.V., Proshkin, I.P., Pustovetov, V.P.: Correlation of strong
earthquakes with variations of the radiation-belt particle flux. Geomag. Aeron. 34(3), 60
(1994)
15. Martynenko, S.I., Fuks, I.M., Shubova, R.S.: Response of the lower ionosphere to variations in
near-ground atmospheric conductivity. Geomagn. Aeron. 34(2), 121 (1994)
16. Mikhailov, Yu.M., Mikhailova, G.A., Kapustina, O.V.: ELF and VLF Electromagnetic back-
ground in the topic ionosphere over seismically active areas. Geomag. Aeronom. 37(4),
450–455 (1997)
17. Rulenko O.P., Ivanov, A.V., Shumeiko, A.V. :A short-term atmospheric electric precursor of
the Kamchatka earthquake of March 6, 1992 (M ¼ 6.1). Dokl. Ross. Akad. Nauk. 326(6),
980–982 (1992)
18. Mikhailov, Yu. M.: Seismoelectromagnetic signals before earthquakes on Kamchatka, solar-
terrestrial bonds and electromagnetic precursors of earthquakes. In: Rep.4th International
Meeting, Paratunka, Kamchatka region, pp. 398–402, Aug 2007
19. Mikhailov, Yu. M.: On the possible role of atmospheric waves in the formation of electro-
magnetic precursors of earthquakes. In: 2nd All-Russia Workshop on Electromagnetic
Sounding of the Earth. November 28–30, 2005, Moscow. Tsentr Geoelektromagnit. Issled,
pp. 91. IFZ RAN, Moscow (2005)
20. Zubkov, S.I.: About dependence time and radius of appearance of electrotelluric precursors
from energy of earthquake. Phys. Earth. 5, 101–106 (1987) (in Russian)
21. Korsunova, L.P., Khegai, V.V.: Seismoionospheric effects of strong crustal earthquakes in the
Pacific Region. Geomagn. Aeron. 45(5), 706–711 (2005)
22. Sorokin, V.M., Chmyrev, V.M., Pohotelov, O.A., Liperovski, V.A.: Review of lithosphere-
ionosphere relations at preparing of earthquake periods. In: Strachov, V.N., Liperovsky, V.A.
(eds.) Short time prediction of catastrophe earthquakes by help of radiophysics ground-based-
kosmic methods, pp.64–83. OIFZ RAS, Moscow (1998)
23. Husamiddinov, S.S.: Ionosphere Research Electrical and Magnetic Precursors of Earthquakes.
In: Golovkov, V.P. (ed.) Tashkent FAS Uz SSR, pp. 90–111 (1983)
24. Korsunova, L.P., Hegai, V.V.: Analysis of seism-ionosphere disturbances on chain of Japan
vertical sounding ionosphere station. Geom. Aeron. 48(3), 407–415 (2008) (in Russian)
25. Sidorin, A.Ya.: Dependence of appearance time of earthquakes precursors of epicenter
distance. Reports of RAS. 245(4), 825–828 (1979)
26. Kim, V.P., Hegai, V.V.: About possibility of formation of metal ion layer in E-region of
average-latitude of ionosphere before strong. Geom. Aeron. 33(5), 114–119 (1993)
27. Sorokin, V.M., Yaschenko, A.K., Hayakawa, M.: Formation mechanism of the lower iono-
spheric disturbances by the atmospheric electric current over a seismic region. J. Atmos. Sol.-
Terr. Phys. 68(11), 1260–1268 (2006)
28. Mikhailov, Yu.M., Mikhailova, G.A., Kapustina, O.V., Depueva, A.X., Buzevich, A.V.,
Druzhin, G.I., Smirnov, S.E., Firstov, P.P.: Variation of different atmosphere and ionosphere
parameters in periods of earthquake preparing on Kamchatka. Preliminary results. Geom.
Aeron. 42(6), 769–776 (2002)
29. Mikhailov, Yu.M.: On the properties of earthquake precursors in the electrostatic field of the
surface atmosphere. Phys. Solid Earth. 43(4), 336–339 (2007)
30. Korsunova, L.P., Khegai, V.V.: Medium-term ionospheric precursors to strong earthquakes
Int. J. Geom. Aeron. 6, GI3005 (2006). doi:10.1029/2005 GI000122
31. Solar-Geophysical Data. Coffey, H.E., Kroehl H. W. (eds.) Solar –Terrestrial Physics
Divisions, NOAA, Boulder, CO (1999)
4 “COMPASS 2” Satellite and Ground-Based Experiments 173

32. Hao, J., Tang, T., Li, D.: Progress in the research on atmospheric electric field anomaly as an
index for short-impending prediction of earthquakes. J. Earthquake Prediction Res. 8(3),
241–255 (2000)
33. Rulenko, O.P.: Short time precursor in near Earth atmosphere electricity. Vulcanol. Seismol.
4, 57–68 (2000)
34. Smirnov, S.E.: Peculiarities of negative anomalies of electrical field at near Earth at
Kamchatka. Geom. Aeron. 45(2), 282–287 (2005)
35. Ismagilov, V.S., Kopytenko, Yu.A., Hattori, K., Hayakawa, M.: Use of gradients and phase
velocities of ULF geomagnetic disturbances for definition point of source of future strong
Earthquake. Geom. Aeron. 46(3), 423–430 (2006)
36. Pulinets, S.A., Boyarchuk, K.A., Hegai, V.V., Kim, V.P., Lomonosov, A.M.: Quasielectro-
static model of atmosphere-thermosphere-ionosphere coupling. Adv. Space Res. 26(8),
1209–1218 (2000)
37. Virk, H.S., Singh, B.: Radon recording of Uttarkashi earthquake. Geoph. Res. Lett. 21(8),
737–740 (1994)
38. Steinitz, G., Vulkan, U., Lang, D.: Monitoring of the tectonically related radon flux in Israel.
Isr. Geol. Surv. Curr. Res. 10, 148–153 (1996)
Chapter 5
Mathematical Modeling of Dynamics Processes
in the Upper Atmosphere and Ionosphere

I.V. Karpov and G.V. Golubkov

Abstract In this chapter, the achievements of the basic researches on the upper
atmosphere and ionosphere processes with the mathematical modeling methods are
briefly presented. The mathematical problem of the model atmosphere/ionosphere
description, the existing global theoretical model of environment, and the results of
the investigations as well as their applications have been considered. The numerical
experiments using the theoretical models show that the results obtained through
model calculations have a good relation with the experimental data on the quiet
geophysical conditions. However, under perturbed conditions, the difference in the
model results and observations can be more evident. The reasons for such differ-
ence can be attributed to the defects of the experimental data on the dynamics of the
atmospheric perturbations sources, as well as the imperfection of the theoretical
models. In particular, the experimental studies have observed an increase in the
microwave radiation (MWR) flows from the upper atmosphere in the geomagnetic
and solar-disturbed conditions. This work has focused on the increase in the MWR
flows, which can be associated with the appearance of high-disturbed particles (n > 10)
in the atmosphere. The correlation between the efficiencies of the chemical reac-
tions and the participation of such particles is one of the directions of improvement
of the theoretical models. Furthermore, the approaches to describe the chemistries
of high-disturbed particles within the framework of theoretical atmosphere/iono-
sphere models also discussed.

Keywords Global theoretical models
Ionosphere perturbations
Precursors

Upper atmosphere

I.V. Karpov (*)

West Department IZMIRAN, Kaliningrad, Russia

V.L. Bychkov et al. (eds.), The Atmosphere and Ionosphere, 175

Physics of Earth and Space Environments,
DOI 10.1007/978-90-481-3212-6_5, # Springer ScienceþBusiness Media B.V. 2010
176 I.V. Karpov and G.V. Golubkov

5.1 Introduction

Currently, the application of mathematical modeling methods is the standard

practice in the theoretical researches on the atmosphere and ionosphere structure
and dynamics. Numerical models have been increasingly applied owing to the
complexity of the physical and chemical processes determining the dynamics of
the neutral and charged components of the upper atmosphere. The primary goals of
theoretical researches with the application of numerical models include the creation
of physical interpretation of the geophysical phenomena proceeding in a particular
region of the atmosphere, research on the role of separate physical processes, study
of the dynamic connections of the processes in the upper atmosphere and iono-
sphere along with the processes in the underlying atmospheric layers, as well as
examination of the processes in the magnetosphere. In the review of the upper
atmosphere and ionosphere, theoretical global models, systems of the equations
describing their dynamics, some data on the achievements in modeling researches,
and ways to achieve further perfection of the models are briefly discussed.

5.2 Global Numerical Models of the Upper Atmosphere

and Ionosphere

The application of the three-dimensional models based on the hydrodynamic

equations for neutral and charged components of the environment is necessary for
the theoretical research on large-scale dynamic processes (scale of more than 1,000
km) in the upper atmosphere and ionosphere (heights of more than 100 km). A brief
chronology of the global upper atmosphere and ionosphere modeling is given in [1].
Initially, the first global numerical models of the neutral components of the upper
atmosphere (thermosphere) dynamics were elaborated in the 1980s [2–4]. The
charged components’ influence on thermosphere was taken into account by includ-
ing the necessary data from the empirical ionosphere models or the simple theoreti-
cal ionosphere models. Further development of the upper atmosphere global models
resulted in the creation of self-consistent models of the dynamics of the neutral and
charged particles as well as the electric fields. The first numerical global model,
GSM TIP, achieved such a self-consistent description [5]. Furthermore, the blocks
of the ionosphere-parameters calculation [6–8] were included in the thermosphere
models [2–4]. Subsequently, the blocks of calculation of electric fields were
included in these advanced models, and were termed as TIME-GCM [9–11] and
CTIP [7] models. The advanced version of model GSM TIP [5] – model UAM [12, 13] –
was developed for high-latitude ionosphere modeling investigations as well as for a
wide spectrum of researches comprising the empirical model data. All the above-
mentioned models provided similar ionosphere–thermosphere description based on
the equations for charged and neutral components of the upper atmosphere. Here,
we consider the system of the equations as the model [5].
5 Mathematical Modeling of Dynamics Processes 177

The system of the equations in a hydrostatics approximation for the basic neutral
gas component of the thermosphere, namely, N2 ;O2 ,O, includes the following:

@ nn
þ r½nn ðV þ Vd n Þ ¼ Q n  L n ; (5.1)
@t

 XX
@V
r þ ðV
rÞV þ 2OV ¼ rP  m ni n ni n n ðV  Vi Þ þ n ðr2
VÞ;
@t i n

   
@T g
r Cv
þ ðV
rÞ T þ P r V ¼ rðl n rT Þ þ r D T r T þ
@t Cp

 X X
4 2 2
þ n ½r  V  ðr VÞ þ nn m n i n n i ðVi  Vn Þ2 þ P un v
3 n i
þ P cn  P rL þ P Tn ;

@r X X @P
þ rðr VÞ ¼ 0; P ¼ n n k T; r ¼ m n n n; ¼ r g;
@t n n
@r

X    
@ ni ni ni @ T @ ni ni ni @ T
Vd n ¼ Din þ þ
 DT þ þ

i
@ r Hi T @ r @r H T @r

In these equations, r, P, T are the density, pressure, and temperature of the

neutral gas, respectively; n n is the concentration of the gas components; V is the
mean velocity of the neutral gas; Vd n is the diffusion velocity of the nth gas
component; O is a vector of angular speed of rotation of the Earth; g is the
acceleration of a gravity; m n i is the given weight for the nth neutral and ith ion
component, equal to m n
m i =ðm n þ m i Þ; n n i is the ion–neutral collision frequency;
V i is a ith ion component velocity;  n ; l n are the viscosity and heat conductivity
coefficients of the neutral gas, respectively; D T ; D i j are the coefficients of turbu-
lent and molecular diffusion, respectively; C v ; Cp are the thermal capacities of
neutral gas at constant volume and pressure, respectively; P un v ; P cn ; P Tn are the
heating of neutral gas from UV radiation of the Sun, corpuscular flaxes, and
turbulent energy dissipation, respectively; Pri  is the losses of energy owing to
radiation; Q n ; L n are the sources and losses of the gaseous component owing to
photochemical reactions, respectively.
Equation (5.1) takes into account an exchange of the movement and energy in
collisions with the charged particles. During the calculation of the characteristics,
the charged components of the upper atmosphere are taken into account, to distin-
guish the dynamics of the field-aligned (atomic ions Oþ ; Hþ ; Heþ in the F-region
of the ionosphere and in the plasmosphere) and non-field-aligned particles (molecular
ions – Oþ þ þ
2 ; N2 ; NO in the lower ionosphere).
178 I.V. Karpov and G.V. Golubkov

The following are the equations for the molecular ions of the lower ionosphere:

@ n ðM þ Þ
¼ Q ðM þ Þ  L ðM þ Þ  r ð n ðM þ Þ V i ðM þ ÞÞ; (5.2)
@t

@ Ti X X
ð 3n i k=2Þ ¼ nn m n i nn i ðV i  V n Þ 2 þ P ei T þ P ni T ;
@t n i

@ Te
ð 3n e k=2Þ ¼ P e q þ P ie T þ P ne T ;
@t
X
n i m i g  r ðn i k T i Þ  m n i n n i ðV i  V n Þ þ en i f E þ ½V i  Bg ¼ 0;
n

Here, n e ¼ n i ¼ nðM þ Þ, and E; B are the electric and magnetic fields, respectively.
The following are the equations for the description of the ionosphere parameters
of F2-region and upper ionosphere:

D ni k
þ r k ð n i V i Þ ¼ Q i  L i  n i r? V ?
i ; (5.3)
Dt

k ni
 2 m i n i ½O  V i  i ¼ n i m i g k  r k pi  r k Pe
ne
X k k
X k
 m i jn i j n i ð V i  V j Þ  m i n n i n nn ð V i  V kn Þ
j n
X
 rðn i k T i Þ  m n i nn i ðV i  V n Þ þ e n i f E þ ½V i  B g;
n

 
3 D T i; e
n i; e k þ V i; e r k T i; e þ P i; e r k V i; e  r k ð l i; e r
k
T i; e Þ
2 Dt
¼ P i; e; Q  P i; e; L ;
k
X
V?
i ¼ V?
e ¼ E  B=B ; V i ¼
2
nj Vj =ne :
j

Here, || and ? are the components of the vectors along the direction and in the
normal direction of the geomagnetic field, respectively (including the vector
operatorr). The operator DDt ¼ @@t þ ð V ? ; r Þ represents a full derivative on time
along a trajectory of the electromagnetic drift, and Qi ; Li are the gains and losses of
ions, Oþ ; Hþ , respectively, taking into account the photos and corpuscular ioniza-
tion, ion–molecular reactions, and charge-exchange processes. The heating of
the electronic gas PeQ is considered as the nonlocal heating by the escaping
photoelectrons.
5 Mathematical Modeling of Dynamics Processes 179

The following is the equation for the calculation of the potential of the electric
field:

r ½^
s
ðr’  V  B Þ  J m  ¼ 0; (5.4)

where ^ s is the ionosphere conductivity tensor, f is the potential of a large-scale

electric field, and J m is the current along the geomagnetic field lines. Furthermore,
as the sources of the electric field of the magnetosphere, the system of currents at
the geomagnetic latitudes of 75 and 65 is set accordingly.
The equations (5.1–5.4) describe the dynamical interaction between the neutral
and charged components of the upper atmosphere and ionosphere.
The solar radiations and corpuscular flaxes determining the chemical and ioni-
zation processes in the upper atmosphere can be chosen from the empirical models
of the appropriate parameters. For example, the boundary conditions for the
solution of equation (5.1) for the parameters of neutral particles can be set at
the lower boundary thermosphere (in the model [5] at a height of 80 km) from
the empirical models of the upper atmosphere [14–16].
At the upper boundary of the thermosphere (in the model [5] at a height of 526
km), the components of wind and temperature can be considered, with their vertical
gradients assumed to be zero owing to dissipation, and by presuming that the gases
are in diffusion balance.
In the model [5] equation for the atomic ions and the electric field integrated
along the magnetic field lines and used as boundary conditions (at the base of the
field lines) the results of integrating the equations for the molecular ions (5.2). In
other models [6–8], the equation for ions are solved in a spherical system of
coordinates, which assumes the choice of the upper boundary conditions at the
fixed height of the top level of the integration region of the equations.
The equations for the neutral and charged particles of an atmosphere (5.1–5.4)
are solved using finite-difference numerical methods on global spatial grids. The
spatial grid scales in such models usually comprise some degrees of latitudes and
longitudes. For example, in one model [6], the steps of integration have been
considered as 5 along the latitudes and longitudes, and in another model [5], 5
and 15 were considered accordingly. It can be noted that in various versions of
these models, the problems of numerical calculations in the grids with higher spatial
resolution were solved [12, 13, 17]. The reduction in the steps of integration on
spatial coordinates allows to precisely describe the changes in the sources of
perturbations of the atmosphere and reaction of the atmospheric parameters in
disturbed conditions.
Verification of such complex theoretical models was carried out by comparing
the data on the parameters of global empirical models of the upper atmosphere and
ionosphere [15–18]. The results of numerical calculations of global distributions of
atmospheric and ionospheric parameters, calculated with the application of such
theoretical models, have shown good consent with the data on empirical models, for
the quiet and moderately disturbed geomagnetic conditions with average levels of
solar activity. The results of electronic-concentration global distribution model
180 I.V. Karpov and G.V. Golubkov

calculations for various seasons and levels of solar activity, calculated on model
UAM [12], are shown in Figs. 5.1 and 5.2, and for comparison, the ionosphere data
from the empirical model IRI [15] is presented.

Fig. 5.1 Electron concentration in a maximum of F2-layer of ionosphere for equinox conditions
and moments UT ¼ 12 h (left) and UT ¼ 18 h (right), obtained in empirical model IRI-2001 (top)
and calculated in model UAM [12] (below)

Fig. 5.2 Electron concentration in a maximum of F2-layer of ionosphere under conditions of a

solstice and various levels of solar activity, obtained in empirical model IRI-2001 (top) and
calculated in model UAM [12] (below)
5 Mathematical Modeling of Dynamics Processes 181

Apparently, the theoretical model was found to reproduce all the basic morpho-
logical features of planetary distributions of the atmospheric and ionospheric
parameters. Similar calculations of the parameters of the upper atmosphere and
ionosphere were also carried out in researches with the application of the models
TIME GCM, GSM TIP, and CTIP. These results show that complex global theo-
retical models adequately describe large-scale structure of the upper atmosphere
and ionosphere, and can be successfully applied in studies on dynamic processes in
a wide range of geophysical conditions.

5.3 Modeling Atmosphere and Ionosphere Perturbations

All atmospheric layers interact among themselves by means of various physical and
chemical processes, forming the complex system subject to the influences of flashes
on the Sun, earthquakes on the Earth, human activity, etc. A change in the
environment’s physical characteristics as a result of these influences is termed as
perturbation. The primary goal of theoretical researches includes revealing the
interrelations of various atmospheric layers’ dynamics, parameters of the atmo-
sphere and ionosphere, validation of various role factors, and studying, understand-
ing, and finally, forecasting the dynamics of environment in the development of
perturbations.
The complexity of modeling of the perturbation situations is determined by
several factors. First, the usual system of upper atmospheric and ionospheric
parameter observations does not allow for obtaining experimental data with the
necessary spatial and temporal resolution. Besides, in the observations, separate
parameters (e.g., either electron concentration in the wind or temperature is
measured, etc.) and measurements indirectly associated with the characteristics of
the parameters of the environment (e.g., intensity of radiation in various ranges of
frequencies) are frequently carried out. However, the accuracy of empirical models
in the description of the disturbed situations in an atmosphere and ionosphere is
insufficient and does not allow for making exhaustive comparison with the theoretical
researches.
Second, theoretical models have some limitations owing to the low spatial
resolution of the equations’ integration grids, as well as insufficient completeness
of the energy transfer channel’s perturbation description in the environment.
Besides, boundary conditions in the numerical models and sources of perturbations
are determined by the results of experimental researches. Thus, owing to the
uncertainty of the experimental data, there exists an uncertainty as a result of
theoretical modeling.
The general problems associated with developing models of the upper atmo-
sphere and the accuracy of their comparison with experimental results have been
formulated in an earlier work [19]. Nevertheless, mathematical modeling as the
universal method of research on processes, even under conditions of insufficiency
182 I.V. Karpov and G.V. Golubkov

of the experimental data for comparative analysis, allows for developing adequate
interpretations of the observed physical phenomena.

5.4 Modeling Ionospheric Disturbances on the Geomagnetic

Storms

The most complex physical phenomena of the upper atmosphere occur at high
latitudes where the direct connection of the ionosphere and neutral atmosphere is
possible through magnetosphere with the solar wind. The disturbances on the Sun
and the changes in the solar-wind parameters as well as their associated interplane-
tary magnetic field result in significant perturbations in the upper-atmospheric
condition. Development of atmospheric and ionospheric perturbations during such
geophysical events influences various fields of activity of people and functioning of
engineering. These circumstances determine the intensity of research on the influ-
ence of solar and geomagnetic activity on the characteristics of the parameters of
environment. The complexity of the theoretical studies of such situations is deter-
mined by the diversity and dynamics of disturbing factors – change over time of the
magnetic field components, perturbation of spectrum corpuscular precipitations
and, therefore, complex and spatially heterogeneous reaction of atmospheric and
ionospheric parameters [20, 21]. More accurate description of the structure of
perturbation factors’ dynamics requires higher accuracy of modeling, i.e., higher
spatial and time resolution of the numerical model (e.g., the characteristic scale on
latitudes is not 1,000 km as in middle latitudes, but 100 km or even less). The
versions of models of the upper atmosphere in which calculations with high spatial
resolution [12, 17] are realized were developed for this purpose. However, nowa-
days, the problems associated with reduction in the steps of integration of the
equations of numerical models are not essential owing to the application of super-
computers in modeling researches. On the whole, the theoretical researches on
disturbed geomagnetic situations in the upper atmosphere and ionosphere have
allowed ofthe development of physically proven interpretations of the observed
phenomena in experiments.
At the same time, numerical experiments have shown that an important setback
for the adequate description of parameters of the upper atmosphere and ionosphere
in theoretical models is the description of spatial and time variability of sources of
perturbations [22–24]. Hence, in [23–26], it was shown that the account of dynamics
of sources of geomagnetic perturbations, nonstationary behavior magnetospheric
electric fields, and changes in the spectra and zones corpuscular precipitations allow
for the more adequate description of the behavior of modeled parameters, with high
conformity with the experimental data. Naturally, obtaining such observation data
on the behavior of electric fields and corpuscular flaxes, necessary for inclusion in
global numerical models, is impossible. In most cases, variability of these para-
meters is defined based on the observation of separate ground stations or satellites.
5 Mathematical Modeling of Dynamics Processes 183

Hence, preparation of the data on perturbation factors characteristics for global

models assumes development of methods of approximation of available given mea-
surements. Owing to the insufficiency and separation of the experimental data, the
spatial in such methods as well as the spatial and time characteristics of the sources of
perturbations have essential uncertainty. Thus, further development of modeling
researches assumes perfection of theoretical models by inclusion of the equations
for magnetosphere dynamics in the modeling description. In this case, sources of
perturbations of the magnetosphere–ionosphere–atmosphere system will be para-
meters of a solar wind and interplanetary magnetic field. Theoretical models in
which blocks of calculation of magnetosphere parameters are included and executed
calculations of distributions of parameters of the atmosphere and ionosphere are
presented in [12, 13, 27–29]. It is obvious that such models are the most complete
models describing large-scale structure and dynamics of a near-earth space.
Application of theoretical models to examine the atmospheric and ionospheric
perturbations dynamics initiated by geomagnetic events has allowed of the deve-
lopment of the proven physical interpretations of the observable phenomena.
Thus, it is necessary to note that the basic aim of these researches was to under-
stand the physical processes that develop in the environment during geomagnetic
perturbations.
However, at the height of the upper atmosphere and ionosphere, only a part of
the solar radiation is absorbed. A significant part of this energy is absorbed in the
underlying atmospheric layers. Absorption of solar radiation at the height of the
middle atmosphere is an important energy source of the upper atmosphere, and it is
taken into account in the theoretical models in the description of variations in the
atmospheric parameters in the lower boundary of the modeling region, at the height
of the upper mesosphere–lower thermosphere (70–90 km). Experimental researches
have shown that the periods of geomagnetic perturbations are accompanied by
perturbations of the parameters of the atmosphere at the height of lower thermo-
sphere (70–110 km) [30–32], and the consequences of geomagnetic storms are
revealed in the ionosphere perturbations for several days after the termination of
geomagnetic perturbation [33, 34]. Thus, existential scales of these perturbations
surpass those of the perturbations of the upper atmosphere and ionosphere initiated
by geomagnetic storms. In lower thermosphere, perturbations of parameters define
the changes in the characteristics of global tidal variations and changes in the daily
average condition of the environment, while in the pretreated ionosphere, perturba-
tions of the parameters are shown in the amplification of longitudinal and UT
variations of the parameters.
Thus, changes in the characteristics of atmospheric parameters owing to geo-
magnetic perturbations can be observed in the upper atmosphere and ionosphere, as
well as in the underlying atmospheric layers. From the perspective of the modeling
researches, this indicates that in the field of the lower boundary conditions of the
theoretical models [5, 7, 9, 13], the perturbations of the atmospheric parameters
initiated by geomagnetic perturbations can be observed. Hence, to improve the
accuracy of the description of the perturbation atmosphere during geomagnetic
storm and the subsequent relaxation to quiet conditions, it is necessary to take into
184 I.V. Karpov and G.V. Golubkov

account the changes in the atmospheric parameters on the lower boundary of

numerical models. Till now, in modeling researches, the influence of geomagnetic
perturbations on the boundary conditions in numerical models was not considered.
Thus, the essential influence of variations in the atmospheric parameters on the
lower boundary of the numerical models on large-scale dynamics of the upper
atmosphere and ionosphere has been confirmed with detailed researches with the
application of all the above-mentioned global models [32–35].
Obviously, the natural development of modeling researches on storm situations
in the upper atmosphere should be accompanied by perfection of the theoretical
models, by assuming the expansion of region of integration of the equations not
only upward in the magnetosphere but also downward, including the underlying
layers of the atmosphere.

5.5 Modeling of Interaction of Various Layers

of the Atmosphere

The principal causes determining the necessity of expansion of area of integration

of the equations in models of the upper atmosphere and ionosphere are connected as
follows:
1. In experimental researches, it has been established that owing to the perturba-
tions of solar and geomagnetic activity, there are perturbations of atmospheric
parameters at the height of mesosphere, as well as perturbations of the upper
atmosphere and ionosphere, which can proceed for a long time even after the
geomagnetic storms have ceased. Besides, experimental researches have revealed
wave variations of the parameters of ionosphere and upper atmosphere, with
periods of more than a day [36, 37]. The occurrence of such variations in the
upper atmosphere is thought to be due to the penetration of planetary waves from
the lower and middle atmosphere in the upper atmosphere. Also, it is necessary to
note that at the height of the upper atmosphere and ionosphere, the ionospheric and
atmospheric parameters’ perturbations are caused by the development of meteoro-
logical processes in troposphere and stratosphere [37].
2. Most part of the solar radiation is absorbed at the height of mesosphere and
lower atmosphere. Photochemical and dynamic processes at these heights of atmo-
sphere cause existential distributions of atmospheric parameters in the height of the
lower boundaries of the existing theoretical models. It is natural to assume that the
boundary conditions in the models on the heights of the lower atmosphere, where
the characteristic times of processes is much higher than the corresponding magni-
tude in the upper atmosphere, thus eliminating the problem of choosing the bound-
ary conditions in the lower thermosphere and rightly raise the possibility of
integrating the equations for long periods. In such a statement of modeling research,
sources of perturbations of environment will operate within the modeled region.
The increase in the duration of integration of the equations of models is necessary
5 Mathematical Modeling of Dynamics Processes 185

for studying the problem of penetration of perturbations from the lower atmo-
spheric layers to the height of the upper atmosphere and ionosphere.
In fact, the requirement for the expansion of area of modeling researches is
determined by an attempt to consider the most essential chemical and dynamic
mechanisms of the atmosphere and ionosphere perturbations caused by absorption
of solar energy within the framework of the theoretical model. Experiments with
such a development of theoretical model were executed on the basis of models
UAM and TIME-GSM [10, 13]. In [33], the results of the research on the processes
of generation and distribution of planetary waves in thermosphere with the period
of 2 days have been presented. As shown in the calculations, for the distribution of a
planetary wave at the height of thermosphere, a very long duration (month) was
found to be necessary. Experimental researches have shown that the time delay of
occurrence of perturbations at the height of thermosphere, when compared with that
of mesosphere, is much less. In this case, numerical experiment demonstrated an
inefficiency of direct wave distribution of a planetary wave from the lower atmo-
sphere up to the height of the thermosphere and ionosphere. In [38], modeling
research with the application of model GSM TIP [5] was presented, in which the
opportunity to generate large-scale perturbations in thermosphere owing to the
vertical propagation of the short-scale (the period of waves studied was close to
Brent–Vaisala period) internal gravitational waves (IGW) were located along a line
of solar terminator, from mesosphere–lower thermosphere. These numerical
experiments have shown efficiency of such channel of penetration of perturbations
from the lower atmosphere to the upper atmosphere, and in another study [38], a
variant, without taking into account the perturbations from short-scale IGW
(Fig. 5.3a), has been shown along with the indignations from IGW in the field of
solar terminator (Fig. 5.3b). As can be observed from Fig. 5.3, the short-scale IGW
influences the formation of circumpolar vortex at the height of 140–160 km under
polar night conditions. Similar modeling researches will help to perform complex
physical processes, providing penetration of perturbations from lower to upper
atmosphere.
For a long time, the modeling of the dynamics of the lower and middle atmo-
sphere was an independent scientific problem. The important purpose of such
researches was to study a difficult and complex photochemical and dynamic
processes initiated by the absorption of solar radiation. Currently, numerical global
mesosphere models that take into account the chemistry and dynamics [39,40] have
been realized. One of the limitations of the mesosphere–lower thermosphere mod-
eling is the description of the upper boundary conditions, which is the height of
150 km; in such region, the account of ionospheric influence on the dynamics of
environment is necessary. As a rule, this problem was solved by including the data
of empirical models. However, insufficiency of experimental researches and expe-
rience of modeling of ionospheric parameters’ distributions demonstrate the major
limitation of the application of such approach. Obviously, a more natural step in the
development of models would be the association of models of the upper atmosphere
and ionosphere with those of the middle atmosphere.
186 I.V. Karpov and G.V. Golubkov

V, m/s geograph. 0h 0n UT 11.8.1990 h=144km

a
north 30° south 30° 84 m/s

12h 12h
180° 180°

270° 06 –270°

0° 0°
h
24 24h

V, m/s geograph. 0h 0n UT 11.8.1990 h=144 km

b
north 30° south 30° 77 m/s

12h 12h
180° 180°

270° 06h –270°

0° 0°

24h 24h

Fig. 5.3 (a) Circulation of wind over the polar caps, designed on model [5] with the absence of
short-scale IGW in the solar terminator region; (b) the same, as on (a), with the account of short-
scale IGW perturbations in the solar terminator region
5 Mathematical Modeling of Dynamics Processes 187

5.6 Perfection of the Description of Photochemical Reactions

The trends in the development of the experimental and basic researches are finally
directed towards the prognostic system of the condition of the upper atmosphere
and ionosphere on the meteorological prognostic systems. From the perspective of
the theoretical researches, this may result in the improvement of the numerical
models owing to the increase in the technical features of the models (the reduction
in the integration equation’s time, increase in the spatial and temporal resolution),
and thus, may provide more complete description of the sources of the perturbations
and channel of the energy of the perturbations transformation in the model equations.
An important aspect in the problem of creation of realistic model of upper atmo-
sphere and ionosphere is the insufficient certainty in describing the complexity of the
chemical processes taking place at the height of upper atmosphere and ionosphere.
Traditionally, in the model of upper atmosphere, including the most developed model
[3, 5, 7, 12, 27], neutral components comprise the basic gas of atmosphere – O, O2 ; N2 ,
and charged components include molecular ions – Oþ þ þ
2 ; NO ; N2 and atomic ions –
þ þ þ
O ; H ; He . The description of the photochemical processes in the theoretical
model, with the account of chemical composition of the neutral particles and ions
considered, includes the processes of photoionization, photodissociation, dissocia-
tive recombination, radiative recombination, etc. The contribution of these chemi-
cal processes in mass, ionization, and heat balance of the ambience is known as
reaction velocity. It is necessary to note that the reaction velocities, along with the
sections of absorption and efficiency of heating from chemical reaction, possess the
significant uncertainty. In the model study using the theoretical models, detailed
researched processes of the formation of the global spatial structure of the neutral
gas components and their influence on heat balance thermosphere and ionosphere
have been presented [41–44]. On the whole, under quiet geophysical conditions, the
applicable schemes including chemical reactions and their influences on heat
balance allow for reproducing the particulars of the morphologies of planetary
structures’ atmospheric parameters, including spatial distribution of neutral and
ion components. In disturbed condition, the difference in the model result and
experimental data can be more significant. The important reasons for these diver-
gences include the description of the defects of the atmosphere’s chemistry. The
theoretical researches show that under perturbed condition, the role of small gas
components of atmosphere increases, influencing the heat balance of the upper
atmosphere and ionization processes [41–47]. The numerical experiments show the
possibility of the reproduction of the experimental results by special choice, for
instance, the efficiency of heating [41]. On the whole, such situation is indicative of
the defect of the description of the chemical processes with the participation of
atmospheric gas and uncertainty in the knowledge of the chemical-reaction velo-
cities. Obviously, the basic researches point to the need to include small gas
components and account of the chemical-reaction efficiencies with their participa-
tion in the schemes of the chemical processes, in the model. Such efficiencies of
chemical reactions in the atmosphere can be a change in the chemical-reaction
188 I.V. Karpov and G.V. Golubkov

velocities, as a consequence of turning the particles and participating in the reaction

under highly disturbed conditions with the main quantum number n of > 10. The
analysis of the experimental data on the observations of electromagnetic radiations
from atmosphere [48] points out the need to prove such suggestions and the
necessity for theoretical researches to examine processes that influence the upper
atmosphere and ionosphere.
Microwave radiations (MWR) are caused by the optical transitions between
Rydberg states of the atoms and molecules in the upper atmosphere. For example,
the Rydberg-state radiation of atomic oxygen (the basic gas component of upper
atmosphere) with a change in the principal quantum number Dnon the unit for n ¼
10–20 lies in a centimetric range, and in millimetric range for larger changes D n.
Decimeter radiation appears in case of transitions without change in the principal
quantum number for n ¼ 20–40.
Satellite registration that allows for MWR in the set spectral range along with the
measurement of intensity can serve one of the essentially new methods of monitor-
ing and provide a frequency profile of radiation to define the distribution of
differential luminosity at different heights. On the basis of the results of the
theoretical analysis of shock and radiative quenching of Rydberg states and those
of the subsequent satellite measurements, it is possible to almost replicate the
research on the density distribution of the atmosphere from 40 to 100 km, for
which, currently, the standard information is practically absent. Owing to the spatial
extent of the area of an electronic cloud, Rydberg states are subject to the influence
of neutral particles of the environment, which are not weak and should be taken into
consideration during the calculation of the corresponding optical characteristics.
For their definition, it is necessary to solve a quantum problem on radiation of
Rydberg atoms or the molecules that are chaotically located neutral particles in the
field. The finding of the resultant intensity and a frequency profile of radiation is
impossible without preliminary definition of characteristics of shock suppression
that, in turn, demands the information on the potential energy surfaces (PES) for all
the fragments of a mix of the gases that make atmosphere (the well-known structure
and relative concentration within the specified layer do not depend on height). The
method of construction of PES in the case of one revolting center (atom or a
molecule) has been developed in detail [49]. To account for the several revolting
centers in the general formalism, the theory of repeated dispersion is employed, i.e.,
the possibility of the analysis of a power spectrum of compound system (quasimo-
lecule) should not cause doubts. The description of shock suppression quasimole-
cule conditions is usually based on the calculated PES within the limits of the
existing dynamic approaches [50]. Calculation of an optical spectrum demands
construction of electronic wave function of a system, which is a key problem,
because on the whole, the regular method of the decision of this problem is not yet
developed. Currently, there is only a scheme of construction of wave function for a
case of one revolting center [51].
During the periods of increase in the solar activity, along with ionospheric
magnetic storms at the top atmosphere (which stretches from 40 to 100 km and
above the surface of the Earth), MWR is formed, whose intensity on some usages
5 Mathematical Modeling of Dynamics Processes 189

exceeds the background. Over the last 10 years, it has been established that this
radiation is caused by the radiating transitions between Rydberg states of atoms and
molecules, which are formed by electron flux and thrown out of the ionosphere [52].
Superfluous MWR, practically lost-free, passes through the atmosphere of the Earth
and produces an undesirable influence on the biosphere. In fact, its resonant
influence on various living organisms, including human beings, has already been
directly acknowledged experimentally [53].
Currently, it is not clear whether the maximum intensity of the ultrabackground
MWR at greater heights (considerably exceeding 100 km) is small owing to the
rarefaction of the atmospheric gas where it is exactly formed, and whether small
(lower or about 40 km) superfluous excitation is effectively extinguished at the
expense of interaction with the neutral particles of the environment. On the one
hand, this fact denotes the absence of reliable information on the character of
distribution of density of the atmospheric gas in a strip from 40 to 100 km over
the surface of the Earth. The density of the environment, as it is known, essentially
influences the speeds of the elementary chemical reactions (dissociation of mole-
cules and recombination of electrons and ions, hemoionization and Penning ioniza-
tion, shock and radiating suppression of the raised conditions, etc.). On the other
hand, there are no uniform mechanisms to form the frequency profile of MWR with
respect to density. Construction of an overall picture of the phenomenon demands
definition of the dependence of dynamics of shock and radiative quenching of
highly excited states of atoms and molecules from the density of the revolting gas.
In the rarefied gas, for the description of a spectrum of radiation, it is possible to be
limited to the account of the indignation created by one neutral particle. In denser
environment, it is necessary to consider the influence of the final number N of
chaotically moving particles in the volume of the raised atom or a molecule, on the
optical characteristics of the system (positions and shifts of the levels, wave
functions, and corresponding dipole moments). The dependence of N on the
principal quantum number n and concentration c of the medium in unit volume is
presented in Table 5.1 (a concentration of 10 17 cm3 in the Earth’s atmosphere
corresponds to the altitude of 40 km, and 10 13 cm3 is attained at 100 km).
For the low-lying electron-excited states, the existing quantum-chemical methods
(e.g., see [54, 55]) allow for calculating the demanded characteristics. In a case with
conditions, these methods are hardly suitable, and it is necessary to use the
alternative approach [56–58]. The greatest difficulties in describing the shock
suppression of the raised conditions include the necessity for correct decision of a
dynamic problem of dispersion. The basic processes leading to a fall in the excita-
tion of reagents are the direct exchange of energy (the momentum mechanism) [59]
and non-adiabatic transitions (through intermediate valence or ionic configurations)
[60, 61]. Their efficiency is defined by the extent of the temperatures of the free

Table 5.1 N as a function of c (cm3) 1013 1014 1015 1016 1017

the principal quantum number
n and concentration c N 45 60 30 40 20 27 15 18 10 13
N 1 5 1 5 1 5 1 5 1 5
190 I.V. Karpov and G.V. Golubkov

electrons and neutral medium components. For fast collisions (when the relative
speed of the facing particles considerably exceeds the root-mean-square speed of
the loosely-coupled electron), the momentum mechanism dominates. At lower
environmental temperatures (usually found in the upper atmosphere), the second
mechanism appears to be more preferable. In this case, consideration demands
preliminary construction of PES particles interacting with all the areas of merge and
the quasicrossings responsible for non-adiabatic transitions. The following stage is
reduced to the calculation of differential and full sections of reaction, which is
carried out within the limits of the semiclassical approach using representation of
the movement of the representing points on various branches of PES [62].
Development of a method of construction of electronic wave functions of the
interacting system for the calculation of the radiation dipole transition moments is
necessary. In a case of highly excited states, the integrated device of the theory using
a method of functions of Green [49] is considered to be the most convenient.
Furthermore, on the basis of the calculated PES and dipole matrix elements, the
dynamic problem defining a frequency profile (the line form) radiation spectrum
needs to be addressed. For simple “atom-nuclear” of the systems (taking into
account the radiating transitions between two low-lying electron states), such
procedure has been well developed [63]. Thus, it appears that the presence of
extremes in the differences of the considered potential curves leads to the occurrence
of “satellites” on the wings of a line of the radiation, which possess high sensitivity
to environmental temperature, as well as positions and amplitudes to details of
potentials of interaction; that is, they depend on its density and structure [64].
The problem of radiative decaying of the raised conditions at the lower part in
the field of several neutral particles is solved in the frameworks of quasi-static
approach [65]. During transition to highly excited states, there is a necessity for a
set of pseudocrossing PES and the corresponding transformation of the surfaces of
the dipole moment optical transitions. This problem has been formulated for the
first time.
The fundamental problems of quenching and radiation of highly excited atoms
and molecules in the rarefied gas have a direct influence on the applied studies on
the structure and dynamics of the upper atmosphere and ionosphere. A number of
experimental facts that have not received satisfactory physical interpretation (in
particular, sporadic splashes in radio emissions from the ionosphere and their high
correlation with the levels of solar and geomagnetic activity, presence of the highly
excited atomic oxygen at the upper atmosphere, etc.) till date, can explain the
excitation Rydberg states of atmospheric gases.
It is necessary to note that researches on the spectra of electromagnetic radia-
tions from an ionosphere are the basis for the most informative methods of research
on environment; for example, a method of non-coherent dispersion. The solution to
the problems associated with examining the influence of highly excited particles in
the atmosphere on the form of a spectrum provides more exact diagnostics of an
environmental condition. The account of such raised conditions of particles should
be considered important in the model description of the upper atmosphere and
ionosphere. Till date, the most developed theoretical models use rather simplified
5 Mathematical Modeling of Dynamics Processes 191

description of the photochemical processes taking place in a mix of atmospheric

gases [3, 5, 66–68]. The structural blocks of these models describing the chemistry
of the upper atmosphere and ionosphere include a complex of chemical reactions
with the participation of the basic gas components of the atmosphere:O, O2 ; N2 .
They also contain the ionized components: molecular ions (Oþ þ þ
2 , NO , N2 ) at the
þ þ
height of the lower ionosphere and atomic ions (He ; H ) at the height F – areas of
the ionosphere and above. The influence of small gas components (O3 , NOx ) is
considered in a parametrical manner in the model of atmosphere and ionosphere
under the indignant solar and geomagnetic conditions. In all the above-mentioned
models, the speed of the chemical reactions has been observed to depend on the
representations of the elastic impacts of the cooperating particles.
The influence of the raised atoms and molecules on the condition of the upper
atmosphere and ionosphere can be investigated using methods of mathematical
modeling with the application of the global self-consistent model of thermosphere,
ionosphere, and plasmosphere (GSM TIP). It is one of the most developed models
of the upper atmosphere applied till date for research on the structure and dynamics
of the environment. Furthermore, an improved description of the chemical reactions
in the framework of the GSM TIP model has also been attempted. The new
description of the dissociative-recombination complex of the atmospheric pro-
cesses can be applied to theoretical investigations of the upper atmosphere and
ionosphere, which may expand the possibilities to receive a more complete physical
interpretation of the studied phenomena.
Development of the intricate theoretical models of the upper atmosphere is
influenced by the necessity to study the dynamic reaction of the atmosphere on
external influences. Recently, owing to the fast development of experimental
researches, the extent of the problems of modeling of the phenomena and processes
in the atmosphere and ionosphere has considerably increased, and thus, the require-
ment for an accurate physical interpretation has essentially increased. The need for
detailed conformity of the theoretical researches to modern requirements demands
further perfection of the model description of the environment. Consecutive com-
plication of the schemes of chemical processes and an increase in the accuracy of
the description of the elementary chemical processes represent one of the most
important directions of development of theoretical models of the upper atmosphere
and ionosphere.

5.7 Modeling Ionosphere Precursors of Earthquakes

One of the most important problems of the theoretical researches on upper atmo-
sphere and ionosphere is the development of physically motivated interpretation of
the observed phenomena. Obviously, model studies should allow for evaluatin the
joint action with many factors simultaneously influencing the condition of the
system, executing the main role in the decision of this problem.
192 I.V. Karpov and G.V. Golubkov

The volume of the experimental data on the observations of the atmospheric and
ionospheric parameters for the past decennial event has vastly increased. This may
be associated with the quick development of the methods of the satellite studies of
the atmosphere. Relatively recently, the possibility of the analysis of the planetary
distribution of the observed atmospheric parameter appeared. For instance, the
incessant observations of the ionosphere parameter TEC (total electron content)
allowed for obtaining the planetary distribution with a time resolution of 2 h [69].
Such data helped to carry out the planetary perturbation analysis and increase the
physical representations of the mechanism defining the planetary structure varia-
bility of the ionosphere. For instance, analysis of the spatial indignations TEC at the
geomagnetic storm periods allowed for motivating the suggestion on the generation
of high-frequency (a period of less than 12 h) planetary Poincare waves in the upper
atmosphere [70], executing important role in the relaxation of atmosphere to quiet
condition after perturbations.
The developed systems of the experimental study on the global structure of the
atmosphere and ionosphere, in particular, the TEC observations, have allowed for
providing exact definition of spatial and temporary efficiencies of the modification
of the ionosphere before earthquakes as an ionosphere precursor. Obviously, these
problems regarding the explanation of the mechanism shaping these phenomena are
very important, and will ultimately help to create the system of the seismic-event
forecast. The efficiencies of the ionosphere perturbations during the periods of
preparing for the earthquakes have been explained in detail [71–73]. The perturba-
tions are shown in Figs. 5.4–5.6. In Fig. 5.4, the deflection of the critical frequency
F-layer of the ionosphere from the average level based on satellite IK-19 is
presented. The negative variations of the electronic density existed for 2 days and
1 day before the advent of earthquake on 16.07.1980 in the region of Australia and
New Zealand [71].
More extensive morphology of the ionosphere perturbation has been obtained
during the last years of TEC observations. The TEC diurnal variations on the
middle latitude stations are shown in Fig. 5.5, which were observed in the field of
epicenter of the earthquake that occurred on 8.01.2006 [73]. As can be seen from
Fig. 5.5, well-marked TEC perturbations, relative to the median values, existed a
day before the seismic event.
The spatial maps of the ionosphere perturbations over the epicenter of the
earthquake, allowing the establishment of the spatial sizes of the area of perturba-
tions and efficiencies of the manifestation of the ionosphere precursors in different
latitude regions are presented in Fig. 5.6. The analysis of the experimental studies
has shown that in most cases, the amplitudes of the seismic indignations’ iono-
spheric parameters form 20% of the median values. The perturbations with
such amplitude can be associated with different factors, determined by instability
of the geophysical situation. An important problem of the ionosphere’s seismic
precursors obtained from the ionosphere variations of the other nature is the
development of physical interpretation of ionosphere perturbations, determined
by seismic activity.
5 Mathematical Modeling of Dynamics Processes 193

Lat 14.07.80
20
8
–0.8 –0.
– 1 .3
10
–1.8 –2

.3
.3

–1

–0
–1

.3
.8
.3

.8
–2
–1

0 .3 –0.8
–1.3
–2
.8

–1.8
–1

–2.

–2
–10

.3
8
.8
–1 dfoF2 MHz
–20
110 120 130 140 150 160 Long
Lat 15.07.80 –0.30
20
–0.3 –0.80
–1.8
–2.3

–3.3
–2.8
–1.3
–0.8

–0
10

.8
–1 –1.30
.3
–1
.8
–3.3

–2 –1.80
–2.

0 –0.3 .3
8

8
–0. –1.3 1.8 .3 –2.30
– –2
–2
–10 .8
–2.80
–2.3
–2.8
–20 –3.30
110 120 130 140 150 160 Long
–3.80
Lat 17.07.80
20
–4.30
–
10 –1
.3 –0.8 0.3
–0 –0.3
–1.3 .8 –0.3

0 –0.8
–1. –1.3
8 –1.3
–2.8
–2.3 –1.8
–10 –2
.8 –2.3
–2.3

–20
110 120 130 140 150 160 Long

Fig. 5.4 Spatial distribution on the area of earthquake in New Zealand on 16.07.1980 (M7.3)
according to satellite IK-19

The basic approaches to interpret ionosphere perturbations are based on the

possible increase in the electric fields. Based on such suggestions, it may be
possible to try to determine the answer to two important questions:
1. Do electric fields get into the ionosphere?
2. Do the values of these electric fields cause the observed effects in the ionosphere?
194 I.V. Karpov and G.V. Golubkov

ORID (41.07° N, 20.47° E) MATE (40.39° N, 16.42° E)

20 20
16 16

TECU
TECU

12 12
8 8
4 4
0 0
3 4 5 6 7 8 9 3 4 5 6 7 8 9
TUBI (40.47° N, 29.27° E) NOT1 (36.52° N, 14.59° E)
20 20
16 16

TECU
TECU

12 12
8 8
4 4
0 0
3 4 5 6 7 8 9 3 4 5 6 7 8 9

Fig. 5.5 The TEC diurnal variations for station ORID, TUBI, MATE, and NOT1 during January
3–9, 2006 (1 TECU ¼ 1016 el/m2). The fine line – current TEC variation, solid line – a median.
Arrow denotes moment earthquakes

Fig. 5.6 The differential percent TEC maps for European region on January 7, 2006. The points
are the specified epicenter of the earthquake

In [72, 73], the qualitative schemes of the electric-field perturbation have been
considered, owing to the changes in the atmospheric conductivity over the area of
the epicenter. In [74], the results of the study on the influence of the localized
perturbations of the electric field on the structure of the ionosphere have been
presented. The additional local potential of the electric field has been considered as
the source of such field. In this work, it has been demonstrated that the development
of such source electric fields with an amplitude of 3–10 mV/m causes perturbations
in the ionosphere, which are very cognate with the ionosphere precursors of earth-
quakes, which appear in TEC observations. Here, a similar space configuration of
5 Mathematical Modeling of Dynamics Processes 195

the additional potential allows for the reproduction of the known observed particu-
larities of the ionosphere precursors in the middle and low latitudes. In [74], the
thrash-out particularities of the change in ionosphere drift cause corresponding
perturbations to electronic concentration, as well as demonstrate the importance of
the amplitudes of the electric field, which are in agreement with the findings of the
other authors [75]. Further studies on the ionosphere precursors of earthquakes need
to examine the physical processes and associate local perturbation of electric field
with the amplitude observed in [74, 75]. The model realization of any scheme of the
penetration seismic-electric field in the upper atmosphere requires the expansion of
the area of modeling from the surface of the land up to the upper atmosphere and
ionosphere, and data corresponding to the outraging factor. However, there exists
an uncertainty in the determination of these outraging factors and the absence
corresponding to the experimental data, allowing their quantitative estimation. In
such situations, searching for possible mechanism of the propagation of perturba-
tions from landing layers in the upper atmosphere and ionosphere is carried out in
the model study. Such studies may be useful to obtain experimental proofs regard-
ing the connections between ionosphere and seismic perturbations. Thus, for
instance, in [76], model GSM TIP [5] was applied to examine the possibility of
generation of large-scale non-uniformities at the height of the ionosphere, owing to
the propagation of IGW with small periods (close to Brent–Vaisala period), loca-
lized in the epicenter of the earthquake. The results of such model calculations [76]
are presented in Fig. 5.7, which demonstrate the possibilities of such mechanism of
ionosphere perturbations excitation and their similarity with the earlier results
obtained using more simple model situation [74]. It can be noted that in this
study, the source of the generation of the short-scale IGW is not discussed, but
the role of such waves in the realization of dynamic connections of atmospheric
layers is presented. The possibility of realization of such mechanism of the atmo-
sphere and ionosphere perturbation over the earthquake epicenter can be achieved
only by undertaking experimental studies.

5.8 Conclusion

Existing theoretical global models of the upper atmosphere and ionosphere are the
powerful tool for studying the physical processes developing in the environment.
Prospects of further development of theoretical models depend on the expansion of
area of modeling from the surface of the Earth to the magnetosphere. Thus, it is
obviously necessary to improve the description of various chemical processes,
paying particular attention to the processes including small gas components of
the atmosphere and the exited particles. It is also necessary to develop theoretical
methods to describe the influence of the final number of neutral particles in the
environment (Rydberg atoms and molecules chaotically moving in volume) on the
dynamics of the elementary processes with the participation of the basic atmo-
spheric particles. Development of theoretical models in these directions will allow
196 I.V. Karpov and G.V. Golubkov

100 100
(TECd - TECq)/TECq, % 02:00 UT (TECd - TECq)/TECq, % 02:00 UT
90 90
90
80 80

60 70 60 70
Latitude (deg)

Latitude (deg)
60 60
30 50 30 50
40 40
0 0
30 30
20 20
–30 –30
10 10

–60 0 –60 0
–10 –10
–90 –20 –20
–180–150–120 –90 –60 –30 0 30 60 90 120 150 180 –180–150–120 –90 –60 –30 0 30 60 90 120 150 180
–30 –30
Longitude (deg) Longitude (deg)
100 100
(TECd - TECq)/TECq, % 10:00 UT (TECd - TECq)/TECq, % 10:00 UT
90 90
90
80 80

60 70 60 70
60 60
Latitude (deg)

Latitude (deg)
30 50 30 50
40 40
0 0
30 30
20 20
–30 –30
10 10
–60 0 –60 0
–10 –10
–90 –20 –20
–180–150–120 –90 –60 –30 0 30 60 90 120 150 180 –180–150–120 –90 –60 –30 0 30 60 90 120 150 180
–30 –30
Longitude (deg) Longitude (deg)
100 100
(TECd - TECq)/TECq, % 14:00 UT (TECd - TECq)/TECq, % 14:00 UT
90 90
90
80 80

60 70 60 70
60 60
Latitude (deg)

Latitude (deg)

30 50 30 50
40 40
0 0
30 30
20 20
–30 –30
10 10
–60 0 –60 0
–10 –10
–90 –20 –20
–180–150–120 –90 –60 –30 0 30 60 90 120 150 180 –180–150–120 –90 –60 –30 0 30 60 90 120 150 180
–30 –30
Longitude (deg) Longitude (deg)
100 100
(TECd - TECq)/TECq, % 24:00 UT (TECd - TECq)/TECq, % 24:00 UT
90 90
90
80 80

60 70 60 70
60 60
Latitude (deg)

Latitude (deg)

30 50 30 50
40 40
0 0
30 30
20 20
–30 –30
10 10
–60 0 –60 0
–10 –10
–90 –20 –20
–180–150–120 –90 –60 –30 0 30 60 90 120 150 180 –180–150–120 –90 –60 –30 0 30 60 90 120 150 180
–30 –30
Longitude (deg) Longitude (deg)

Fig. 5.7 Global maps of deviations of the total electron content in geographic Cartesian coordinate
system for the various moments of time, obtained from calculations carried out with and without
the IGW parameter generated in the epicenter region (the left panel) and calculated according to
GPS TEC, obtained 1 day prior to earthquake in Greece, and median values of GPS TEC (the right
panel). Epicenter of earthquake is marked by asterisk

for the improvement of the quality of researches on the physical processes defining
the structure and dynamics of the environment. This, in turn, will promote the
understanding of the features of the dynamics of the atmosphere and ionosphere
5 Mathematical Modeling of Dynamics Processes 197

owing to various perturbations, both of geophysical and technologic nature. Fur-

thermore, the obtained results may be helpful to experts in the area of physics and
atmosphere chemistry.

References

1. Namgaladze, A.A., Yurik, R.Yu.: Mathematical modeling of Earth upper atmosphere pertur-
bations, RFBR, Earth science, 1998
2. Dickinson, R.E., Ridley, E.C., Roble, R.G.: A three dimensional general circulation model of
the thermosphere. J. Geophys. Res. 86, 1499–1512 (1981)
3. Dickinson, R.E., Ridley, E.C., Roble, R.G.: Thermospheric general circulation with coupled
dynamics and composition. J. Atmos. Sci. 41, 205–219 (1984)
4. Fuller-Rowell, T.J., Rees, D.: A three-dimensional, time dependent global model of the
thermosphere. J. Atmos. Sci. 37, 2545–2567 (1980)
5. Namgaladze, A.A., Korenkov, Yu.N., Klimenko, et.al.: Global model of the thermosphere–
ionosphere–protonosphere system. Pure Appl. Geophys. 127(2/3), 219–254 (1988)
6. Roble, R.G., Ridley, E.C., Richmond, A.D., et al.: A coupled thermosphere and ionosphere
general circulation model. Geophys. Res. Lett. 15, 1325–1328 (1988)
7. Fuller-Rowell, T.J., Rees, D., Moffett, R.J., et.al.: A coupled thermosphere ionosphere model,
CTIM. In: Schunk, R.W. (ed.) Paeronomical Models of the Ionosphere STEP Handbook,
pp. 217–242 (1996)
8. Fuller Rowell, T.J., Rees, D., Quegan, S., et.al.: Simulations of the seasonal and universal time
variations of the high latitude thermosphere and ionosphere using a coupled, three-dimen-
sional model. Pure Appl. Geophys. 127(2/3), 189–217 (1988)
9. Richmond, A.D., Ridley, E.C., Roble, R.G.: A thermosphere ionosphere general circulation
model with coupled electrodynamics. Geophys. Res. Lett. 19, 601–604 (1992)
10. Roble, R.G., Ridley, E.C.: A thermosphere–ionosphere–mesosphere–electrodynamics general
circulation model (TIME-GCM): Equinox solar cycle minimum simulations (30–500 km).
Geophys. Res. Lett. 21, 417–420 (1994)
11. Roble, R.G.: On the feasibility of developing a global atmospheric model extending from the
ground to the exosphere. In: “Atmospheric Science Across the Stratopause,” Geophysical
Monograph 123, American Geophysical Union, pp. 53–67 (2000)
12. Namgaladze, A.A., Martynenko, O.V., Volkov, M.A., et.al.: Highlatitude version of the global
numerical model of the Earth’s upper atmosphere. Proceedings of the MSTU 1(2), 23–84
(1998)
13. Namgaladze, A.A., Martynenko, O.V., Volkov, M.A., et.al.: Mathematical modeling of large-
scale upper atmosphere perturbations, Modelling of upper polar atmosphere processes. PGI
KSC RAS, pp. 167–249 (1998)
14. Fuller-Rowell, T.J.: A two-dimensional, high-resolution nested-grid model of thermosphere
2: response of the thermosphere to narrow and broed electrodynamic features. J. Geophys.
Res. 90(A6), 6567–6568 (1985)
15. Bilitza, D.: International reference ionosphere 2000. Radio Sci. 36, 261–275 (2001)
16. Hedin, A.E.: MSIS-86 thermospheric model. J. Geophys. Res. 92, 4649–4973 (1987)
17. Hedin, A. E., Fleming, E.L., Manson, A.H., et.al.: Vial empirical wind model for the upper,
middle and lower atmosphere. J. Atmos. Terr. Phys. 58(13), 1421–1447 (1996)
18. Klimenko, V.V., Korenkov, Yu.N., Bessarab, F.S., et.al.: Model/data comparison of the F2-
region parameters for August 11, 1999 Solar eclipse. Adv. Space Res. 31(4), 995–1000 (2003)
19. Richmond, A.M., Wiltberger, M., Hanli, L.: The upper atmosphere problems in developing
realistic models. Geophysical Models at NCAR. A Scoping and Synthesis Workshop, NCAR.
Boulder, CO, 13–14 Nov pp. 1–3 (2006)
198 I.V. Karpov and G.V. Golubkov

20. Fuller-Rowell, T.J., Codrescu, M.V., Roble, R.G.: How do the thermosphere and ionosphere
react to a geomagnetic storm? Magnetic storm. Geophys. Monogr. AGU 98, 203–225 (1998)
21. Prölss, G. W.: Ionospheric F-region storms. In: Volland, H. (ed.) Handbook of Atmospheric
Electrodynamics, 2, pp. 195–248. CRC Press, Boca Raton, FL (1995)
22. Kamide, Y.: Interplanetary and magnetospheric electric fields during geomagnetic storms:
what is more important – steady-state fields or fluctuating fields? J. Atmos. Sol.-Terr. Phys.
63(5), 413–420 (2001)
23. Rees, D.: Observations and modeling of ionosphere and thermospheric disturbances during
geomagnetic storm: a review. J. Atmos. Terr. Phys. 57(12), 1433–1457 (1995)
24. Korenkov, Yu.N., Bessarab, F.S., Klimenko, V.V., et.al.: Numerical modeling of the thermo-
sphere–ionosphere coupling during substorm. Adv. Space Res. 18(1), 41–44 (1996)
25. Fuller-Rowell, T.J., Codrescu, V.C., Wilkinson, P.: Quantitative modeling of the ionospheric
response to geomagnetic activity. Ann. Geophys. 18(7), 766–781 (2000)
26. Codrescu, M.V., Fuller-Rowell, T.J., Foster, J.S.: On the importance of E-field variability for
Joule heating in the high-latitude thermosphere. Geophys. Res. Lett. 22(17), 2393–2396
(1995)
27. Wang, W., Wiltberger, M., Burns, A.G., et.al.: Initial results from the coupled magneto-
sphere–ionosphere and -thermosphere model: thermosphere-ionosphere responses. J. Atmos.
Solar-Terr. Phys. 66, 1425–1438 (2004). doi:10.1016/j.jastp. 2004. 04. 008
28. Wiltberger, M., Wang, W., Burns, A.G., et.al.: Initial results from the coupled magneto-
sphere–ionosphere–thermosphere model: magnetospheric and ionospheric responses. J.
Atmos. Solar-Terr. Phys. 66, 1411–1424 (2004). doi:10.1016/j.jastp. 2004. 04. 026
29. Moffett, R.J., Millward, G.H., Quegan, S., et.al.: Results from a coupled model of the
thermosphere, ionosphere, and plasmosphere (CTIPM). Adv. Space Res. 18(3), 33–39 (1996)
30. Salah, J.E., Goncharenko, L.P.: Search for geomagnetic storm effects on lower thermospheric
winds at midlatitudes. J. Atmos. Sol.-Terr. Phys. 63(9), 951–963 (2001)
31. Rees, D., Fuller-Rowell, T.J.: Comparison of empirical and theoretical models of species in
the lower thermosphere. Adv. Space Res. 13(1), 107–129 (1993)
32. Karpov, I.V., Korablinova, N.A.: The migrating tide influence on lower thermosphere circu-
lation at middle latitudes. Geom. Aeron. 38(2), 112–119 (1998)
33. Fessen, C.G., Roble, R.G., Ridley, E.C.: Thermospheric tides at equinox: Simulation with
coupled composition and auroral forcing. 1. Diurnal component. J. Geophys. Res. 96(A3),
3647–3661 (1991)
34. Fessen, C.G., Roble, R.G., Ridley, E.C.: Thermospheric tides at equinox: Simulation with
coupled composition and auroral forcing. 2. Semidiurnal component. J. Geophys. Res.
96(A3), 3663–3677 (1991)
35. Korenkov, Yu.N., Klimenko, V.V., Bessarab, N.S.: The investigation of boundary conditions
influence on thermodynamical balance OH thermosphere at winter solstice. Geom. Aeron.
43(2), 119–128 (2003)
36. Lashtovicka, J.: Forcing of the ionosphere by waves from below. J. Atmos. Sol.-Terr. Phys.
68, 479–497 (2006)
37. Altadill, D., Apostolov, E.M.: Time and scale size of planetary wave signatures in the
ionosphere F-region. Role of the geomagnetic activity and mesosphere/lower thermosphere
winds. J. Geophys. Res. 108(A11), 1403–1429 (2003)
38. Karpov, I.V., Bessarab, F.S.: Model investigation of solar terminator influence on thermo-
spheric parameters. Geom. Aeron. 48(2), 217–227 (2008)
39. Frohlich, K., Pogoreltsev, A., Jacobi, Ch: The 48-layer COMMA-LIM model. Rep. Inst.
Meteorol. Univ. Leipzig. 30, 157–185 (2003)
40. Miyahara, S., Miyoshi, Y.: Migrating and nonmigrating atmospheric tides simulated by a
middle atmosphere general circulation model. Adv. Space Res. 20, 1201–1205 (1997)
41. Wells, G.D., Rodger, A.S., Moffett, R.G., et al.: The effects of nitric oxide cooling and
photodissociation of molecular oxygen in the thermosphere/ionosphere system over the
Argentine islands. Ann. Geophys. 15, 355–365 (1997)
5 Mathematical Modeling of Dynamics Processes 199

42. Killeen, T.L., Burns, A.G., Azeem, I.: A theoretical analysis of the energy budget in the lower
thermosphere. J. Atmos. Sol-Terr. Phys. 59, 675–689 (1997)
43. Bessarab, F.S., Korenkov, Yu.N.: The dynamical processes influence on heat balance of upper
thermosphere. Geom. Aeron. 33(5), 120–126 (1993)
44. Bessarab, F.S., Korenkov, Yu.N.: The nitric oxide influence on global distributions of
thermosphere and ionosphere parameters. Geom. Aeron. 38(5), 107–116 (1998)
45. Fuller-Rowell, T.J.: Modeling of solar cycle change in nitric oxide in the thermosphere and
upper mesosphere. J. Geophys. Res. 98, 1559–1570 (1993)
46. Fuller-Rowell, T.J., Rees, D., Rishbeth, H., et al.: Modelling of composition changes during
F-region storms – a reassessment. J. Atmos. Terr. Phys. 53, 541–550 (1991)
47. Fuller-Rowell, T.J., Rees, D., Tinsley, B.A., et.al.: Modeling the response of the thermosphere
and ionosphere to geomagnetic storms: Effects of a mid-latitude heat source. Adv. Space Res.
10(6), 215–222 (1990)
48. Avakyan, S.V.: Physics of solar-terrestrial connections: some results and new approaches.
Geom. Aeron. 48(4), 435–443 (2008)
49. Golubkov, G.V., Ivanov, G.K.: Rydberg atoms and molecules in the medium of neutral
particles. Khim. Fiz. 22(10), 25–86 (2003)
50. Golubkov, G.V., Ivanov, G.K.: Rydberg states of atoms and molecules and elementary
processes with their participation. URRS, Moscow (2001)
51. Golubkov, G.V., Golubkov, M.G., Ivanov G.K.: Rydberg atom A** in a field of neutral atom
B. In: Gigosos, M.A., Gonzalez, M.A. (eds.) Spectral Line Shapes, Vol. 15; XIX ICSLS,
Valladolid, Spain, pp. 140–142. Melville, New York (2008)
52. Golubkov, G.V., Ivanov G.K., Golubkov, M.G., Karpov, I.V.: Rydberg atom microwave
radiation the field of the chaotically located neutral particles. Abstracts of 1st International
conference “Atmosphere, Ionosphere, Safety” (AIS 2008), pp. 99–101. Kaliningrad (2008)
53. Avakyan, S.V., Voronin, N.A., Serova,. A.E.: The role Rydberg atoms and molecules in the
upper atmosphere. Geom. Aeron. 37(3), 99–106 (1984)
54. Avakyan, S.V.: The microwave radiation from ionosphere as factor of solar flashes and
geomagnetic storms influence on biosystems. Opt. J. 72(8), 41–48 (2005)
55. Szabo, A., Ostlund, N.S.: Modern Quantum Chemistry. Dover, Mineola, New York (1966)
56. Werner, H.-J., Knowles, P.J.: An efficient internally contracted multiconfiguration – reference
configuration interaction method. J. Chem. Phys. 89, 5803–5811 (1988)
57. Golubkov, G.K., Ivanov, G.K., Balashov, E.M.: Interactions of the highly excited atoms and
molecules with neutral particles. Features of the potential energy surfaces and the charge
exchange processes. Chem. Phys. Rep. 14(8), 1104–1127 (1995)
58. Golubkov, G.V., Ivanov, G.K., Balashov, E.M., and Golubkov, M.G.: l mixing and dissocia-
tion of Rydberg molecules accompanying slow collisions with inert-gas atoms, JETP. 87(1),
56–63 (1998)
59. Matsuzava, M.: Theoretical investigations of Rydberg atom collisions with molecules. In:
Stebbings, R., Dunning, F. (eds.) Rydberg States of Atoms and Molecules. Cambridge
University Press, Cambridge (1983)
60. Ivanov, G.K., Golubkov, G.V., Balashov, E.M.: Role of non-adiabatic mixing of the vibronic
states of highly excited molecules with neutral particles. Dokl. Akad. Nauk SSSR. 323,
311–315 (1992)
61. Balashov, E.M., Golubkov, G.V., Ivanov, G.K.: Strong irregular excitation-energy depen-
dence of the inelastic transitions and collisional ionization cross-sections of highly excited
molecules. JETP 76(2), 200–209 (1993)
62. Nikitin E.E., Umanskii S.Ya. Theory of Slow Atomic Collisions. Springer-Verlag, Berlin,
Heidelberg/New York/Tokyo (1981)
63. Devdariani, A.Z.: Radiation of quasimolecules, Opt. Spectrosc. 86(6), 954–959 (1999)
64. Zhu, C., Babb, J.F., Dalgarno, A.: Theoretical study of sodium and potassium resonance lines
pressure broadened by helium atoms. Phys. Rev. A. 73, 012506 (2006)
65. Lehner, M., Xu, R., Jungen, M.: The emission spectrum of the Li(2p)He2. J. Phys. B 38,
1235–1247 (2005)
200 I.V. Karpov and G.V. Golubkov

66. Namgaladze, A.A., Korenkov, Yu.N., Klimenko, V.V., et.al.: Global model of the Earth
thermosphere–ionosphere–protonosphere. Geomagn. Aeron. 30(4), 612–619 (1990)
67. Namgaladze, A.A., Korenkov Ju.N., Klimenko V.V. et al.: Numerical modelling of the
thermosphere-ionosphere-protonosphere system. J. Atmos. Terr. Phys. 53(11, 12),
1113–1124 (1991)
68. Bessarab, F.S., Korenkov, Yu.N., Karpov, I.V.: Sensitivity of GSM TIP model to the varia-
bility of the UV and EUV solar radiation. Phys. Chem. Earth. 25(5, 6), 543–546 (2000)
69. Afraimovich, E.L., Perevalova, N.P.: GPS-Monitoring of the Earth Upper Atmosphere. SC
RRS SB RAMS, Irkutsk, Russia (2006)
70. Karpov, I.V. Shagimuratov, I.I., Yakimova., G.A.: Ionospheric Effects of Poincare Waves. In:
Proceedings of International Conference “Fundamental Space Research”, pp. 75–77, Sunny
Beach, Bulgaria (2008)
71. Liperovskyi, V.A., Pohotelov, O.A., Shalimov, S.A.: The Ionosphere Precursors of Earth-
quakes. Nauka, Moscow (1992)
72. Pulinets S.A., Boyarchuk K.: Ionospheric Precursors of Earthquakes. Springer, Berlin, Ger-
many 315 p (2004)
73. Zakharenkova, I.E.: The using of GPS signals measurements for finding of earthquake
ionosphere precursors.// Cand. Diss., RSU, Kaliningrad 134 p (2007)
74. Namgaladze, A.A., Klimenko, M.V., Klimenko, V.V., and Zakharenkova, I.E.: Physical
mechanism and mathematical modeling of earthquake ionospheric precursors registered in
total electron content. Geom. Aeron. 49(2), 252–262 (2009)
75. Sorokin, V. M., Chmyrev, V. M., Yaschenko, A. K.: Theoretical model of DC electric field
formation in the ionosphere stimulated by seismic activity. J. Atmos. Sol.-Terr. Phys. 67,
1259–1268 (2005)
76. Namgaladze, A.A., Klimenko, M.V., Klimenko, V.V., Zakharenkova, I.E.: Physical mecha-
nism and mathematical modeling of earthquake ionosphere precursors registered in total
electron content. Geom. Aeron. 49(2), 267–277 (2009)
Chapter 6
Ball Lightning Investigations

V.L. Bychkov, A.I. Nikitin, and G.C. Dijkhuis

Abstract Ball lightning (BL) researches’ review and theoretical models of three
different authors are presented. The general review covers investigations from
1838 until the present day, and includes a discussion on observation data, experi-
mental modeling, and theoretical approaches. Section 6.1 is written by Bychkov
and Nikitin; authors of the sections 6.2, 6.3 and 6.4 are, respectively, Bychkov,
Nikitin and Dijkhuis.

Keywords Ball lightning
Charged eigenstate
Combustion
Concentric gear

shells
Condensed melted hot material
Discharges
Dynamic electric capacitor

Electrically charged object
Electron ring
Experiments
Explosion
High energy

Levitation
Observations
Protons
Spherical vortex crystals
Theoretical models

Vacuum cavity

6.1 Modern State on Ball-Lightning Investigations

The historical review of researches on ball lightning (BL) is presented. It is

conditionally separated into two stages: first – till 1988, the moment of commence-
ment of the International Committee on BL work, and second – since 1988 till
today. Furthermore, researches on gathering and data processing on BL observa-
tions, experiments on the reproduction of long-lived shining formations in electric
discharges, and BL models, are described. The conclusion that the work on BL
observations’ data collection has allowed to reliably determine its basic properties
has been drawn. Now, there is a necessity for an analysis of its rare properties.
Theoretical models need to provide an explanation for BL mechanisms of

V.L. Bychkov (*)

Physical Electronics Chair, Physical Department, M.V. Lomonosov Moscow State University,
Moscow 119992, Russia
e-mail: [email protected]

V.L. Bychkov et al. (eds.), The Atmosphere and Ionosphere, 201

Physics of Earth and Space Environments,
DOI 10.1007/978-90-481-3212-6_6, # Springer ScienceþBusiness Media B.V. 2010
202 V.L. Bychkov et al.

high-specific energy content and holding of non-compensated electric charge. In

the last part of the chapter, detailed descriptions of the three high-energy BL models
developed by Bychkov, Nikitin and Dijkhuis, are presented.

6.1.1 Introduction

The Tenth Jubilee International Symposium on Ball Lightning was held in July
2008 in Zelenogradsk city of Kaliningrad region, Russia, on the Baltic seashore.
The first symposium on this topic took place exactly 20 years before, from 4th to 6th
July 1988, at Waseda University in Tokyo, Japan [1]. The organizers of the first
symposium were Professor Dr. Y.-H. Ohtsuki, Professor Dr. N. Kitagawa, and
Professor Dr. H. Kikuchi. It was attended by 54 scientists from 8 countries. The
first symposium was biannually followed by the second one in Budapest, Hungary
(1990), the third in Los Angeles, USA (1992), the fourth in Canterbury, Great
Britain (1995), the fifth in Tsugawa-Town, Niigata, Japan (1997), the sixth in
Antwerp, Belgium (1999), the seventh in St. Louis, USA (2001), the eighth in
Chung-li, Taiwan (2004), and the ninth in Eindhoven, the Netherlands (2006).
Usually, the results of scientific activity of the regular symposia were analyzed in
reviews of the International Committee on Ball Lightning by President Dr. Stanley
Singer [2–5]. Unfortunately, in 2008, Dr. Singer passed away, and hence, the tenth
symposium was dedicated to his memory.
Usually, the results of the BL investigators’ 2 years’ activities were summed up
in analytic reports and review articles describing the symposia work [6, 7]. In the
present review, we consider reasonable to expand the time in a timetable and
describe the most interesting works made before the start of International Symposia
constraints and after it, that is, to write something similar to a history of BL
investigations.
Traditionally, the following subjects come under discussions at the symposia:
(1) data on BL observations; (2) laboratory experiments on BL realization; (3) BL
theory; and (4) BL observations video data. Items 1 and 4 are, in essence, devoted to
the same subject. In the following review, we will dispose our material in the
sequence of subjects 1–3. We will describe the collection and analysis of the
observation data, we will consider most interesting works on experimental inves-
tigations of laboratory BL analogs, and we will try to present an image of activity in
the area of BL theoretical models. Besides works discussed and published in BL
Symposia and Conferences transactions, as well as in peer-reviewed journals,
Internet, and popular editions, regular information about BL observations and
some ideas concerning its physical nature are presented. We have also tried to
include such data in our consideration, trying to separate sensation-aimed canards
and really reliable reports. Obviously, we might have missed something interesting
in this “information sea” distributed over nonspecialized information sources.
Therefore, we express our apologies to those authors whose works were left out
of our zone of attention.
6 Ball Lightning Investigations 203

6.1.2 Collection and Analysis of Ball Lightning Observation Data

6.1.2.1 Data Collected Before 1988

The first meeting of scientists investigating atmospheric electricity with BL was in 1753.
Professor George William Richmann was killed by its stroke during a thunderstorm. He
was measuring a potential of a metallic bar placed on a roof of his house with a help of an
electrometer – a metallic rod that was placed inside the house. The rod was isolated from
the ground by a crystal glass and a silk thread was attached to the rod through one of its
ends. He judged the value of the rod potential by an angle of a thread deflection. The rod
and the electrometer were connected by an iron wire going through an inner porch and a
corridor [8]. The event was investigated by Mikhailo Vasil’evich Lomonosov. He
established that a fire ball was formed outside the house, which penetrated into a
room through either a door or a window. The report of Lomonosov can be considered
as the first qualified description of traces left by BL. Besides this event, Lomonosov also
knew other cases of BL observations, which have been presented in his article “The
word about air phenomena resulting from the electric force” [9]. In this work, he had
proposed his explanation of this natural phenomenon as: “combustion of fats, gathered
together in air.” In fact, this was the first in the world model of BL with a chemical
source of energy. Unfortunately, Richmann’s death had, for a long time, retarded studies
on the atmospheric electricity (including investigations of BL). Leonard Euler stated:
“This case has taken away courage from the local naturalists, who were studying the
thunderstorm phenomena, and they have interrupted their investigations” [8, P. 7].
We can consider the study by Francois Arago as the next surge of interest in BL.
He had collected and published 30 evidences of BL observations and explained this
as “lightning energy condensation” [10]. The later BL observation cases and its
models were discussed at Paris Academy of Sciences sessions.
An important landmark in BL problem investigation is Walter Brand’s book
“Ball Lightning” publication in 1923 [11]. This book publication in Germany at that
time under conditions of unprecedented inflation was a heroic deed of the scientist.
Brand, in his book, had represented 215 BL observation descriptions from 1665 to
1919 as described by eyewitnesses, and indicated 14 main features of BL. As a result
of his observation and analysis, nowadays, when we know many other observations,
one can make the following conclusions that are based on his first analysis:
1. The thunderstorm in the majority of cases is the cause for BL appearance;
however, BL can be observed in fair weather.
2. One can most often observe a single BL; however, there are cases when two or
more number of balls were observed simultaneously.
3. BL was observed on a plain, in mountains, and near seas.
4. BL is formed not only during thunderstorms, but also during hurricanes and
tornado.
5. The maximum number of BL observations in Central Europe occurs during
the summer months – from May to August; this correlates with thunderstorm-
appearance frequency. BL is most often observed from 14 to 20 h during day
204 V.L. Bychkov et al.

time, but a period of most thunderstorm activity falls during 12–18 h. It means
that BL more often appears at the end of the thunderstorm.
6. In a majority of cases, the form of BL is spherical; sometimes, it is a hollow
sphere; and sometimes, it is compared with a soap bubble. Sometimes, one
observes deviations from the spherical form: BL has the appearance of an egg,
an ellipsoid, a flame tip, a bar, or a pole.
7. The BL contour is usually clearly outlined, but sometimes, it seems to be
blurred, surrounded by a glowing gas. During its motion, one can sometimes
observe projections on its surface that protrude in different sides. Sometimes,
one can see a halo around BL. In a majority of cases, its color is different from
those of the ball.
8. Sometimes, one can observe an internal structure of a ball: a decrease in the
light intensity from the center to the edge; the ball is dark red in the center and
brighter at the edge; or a transparent ball, with a blue flame in the center.
9. Sometimes, sparks, flame bunches, and beams of approximately 50 cm in
length fly from the surface of the ball.
10. BL sometimes changes its form: it elongates going through a hole, it is
transformed from a “fire snake” to a ball and vice versa; the ball slowly
elongates, transforming to thin weakly glowing strip.
11. The average diameter of BL is 20 cm, and its size range is from 1–2 cm to 27 m.
During its lifetime, BL size can be constant, increasing, or decreasing.
12. In a number of cases, BL is rotating or rolled over a ground.
13. The major colors of BL are: white, white-blue, dark-blue, yellow-green,
yellow-red, and red (the last one appears more often than others). A change in
the color of the ball usually takes place before an explosion of the ball: it becomes
white from red-yellow; the ball being yellow before an explosion becomes
red-blue; the white central part of a ball becomes red before an explosion.
14. There are reports of observations of a black BL, as if consisting of a dense
smoke.
15. The light irradiated by BL is characterized in a range from a dim to a dazzling
bright light. Observers often state that it was bright but not blinding. The
brightness of BL often increases before its disappearance. There are cases
when the luminescence brightness of different BL parts was different.
16. Sometimes, BL leaves a smoky trace behind itself or is covered by a smoke or
vapor. Observers usually compare the smell of the smoke with those of burning
sulfur or burning powder. They generally describe the vapor as the orange gas
that can be identified as nitrous oxide (N2O). In other cases, the smell was
identified as the smell of ozone, especially when BL was observed by people
belonging to science.
17. In most cases, there are no material traces after a BL explosion.
18. In a vast majority of cases where observers were at a close distance to BL or
have touched it, they point out the absence of the feeling of heat from the heat
emitted from it. Sometimes, on the contrary, observers report about heat
irradiation and even about cases of singing or inflammations of objects by
BL. The effects with heat release most often take place during BL explosions. It
6 Ball Lightning Investigations 205

is usually difficult to separate the consequences of BL explosions and the

effects caused by a stroke of usual linear lightning.
19. An encounter of BL usually leaves an indelible impression in an observer’s
memory. Most often, it causes a feeling of danger and fear.
20. BL at close contact with an observer’s body can cause temporary loss of
consciousness, paralysis, or even death. There are many cases of animal deaths
during a BL explosion. A “mechanism” of breather affection is similar to the
impact of usual linear lightning.
21. There are data about the mechanical work executed by BL: stone spreading,
transportation of people and animals in space, and making holes in thick walls,
wooden doors, or window glasses. BL is capable of making a deep pit or a
trench in the ground, throwing a tile from the roof, or swinging a church bell. It
can evaporate wine, oil, and water in earthenware, leaving the vessel safe; tear
a shirt to pieces; instantly melt down a piece of ore or gold; and tear away a ring
from a finger.
22. BL can impact electrical devices: it can make electrical bells to ring and cause
glowing of bulbs; its explosion is accompanied by a melting down of safety
fuses and burnout of light bulbs.
23. Sometimes, BL is a sound source. This sound is compared with a whistle
of a rising rocket or a shell. Sometimes, a crackling noise accompanies BL
motion. St. Elms fires produce a similar set of sounds, but the sound gener-
ated by BL is much louder. Probably, one can connect BL sound with electric
charge loss.
24. In a large number of cases, the ball dropped out of the sky is observed.
Sometimes, BL reaches the earth’s surface, but sometimes, BL’s fall is stopped
at a height of 1–15 m, and BL starts moving horizontally. There are cases when
BL, on the contrary, rose up to the cloud from the earth’s surface. Sometimes,
BL stops and stays motionless in air. BL’s path is never straight, but sometimes
it falls down from a cloud to earth along the vertical trajectory.
25. BL moves along a complex trajectory in closed premises or near buildings. It can
jump aside as an elastic object when it comes into contact with walls or ground.
26. In premises, it often moves round a circle approaching walls and floor not
closer than 1 m. BL flying in a room through a window or a door usually moves
to an opposite wall. Sometimes, it leaves the room along the same way through
which it flew in. This can be caused by electric-field variation in the room or a
complicated airflow gas-dynamics in the premises [12] with respect to the
location of intake and air vent holes.
27. A sharp change in BL motion direction can be caused by air flows or an impact
of electric conductors. However, sometimes, air motion does not visibly impact
it, and it may move upwind. The features of BL motion can be explained by its
attraction to metallic objects, and its tendency to move through chimneys can
be associated with the electrical conductivity of smoke.
28. The path of BL motion, in some cases, remains along the electric conductors: it
moves along the telephone, telegraph, and electric cables, wires of lightning
rod groundings, drainpipes, and tree trunks.
206 V.L. Bychkov et al.

29. There were marked cases of BL moving along the surfaces of people and
animals, leaving skin burns. However, more often it demonstrates a tendency
to evade contacts with people, animals, and good conductors.
30. BL can penetrate through keyholes, slits, and cracks. In rare cases, it can
penetrate through glasses, leaving it intact; sometimes, it makes a fist-size
orifice.
31. During its lifetime, BL can cover a substantial distance of several hundreds of
meters.
32. BL moves with a constant velocity only in rare cases; usually, it accelerates and
decelerates. BL can be separated into two groups with respect to its motion
velocity. The first one consists of BL moving from clouds to earth or between
the clouds. On an average, its velocity is of several tens of meters per second.
The second group consists of BL observed in premises or outdoors near the
earth. Its average velocity is 2 m/s.
33. The duration of existence of BL is from 1 s to 5 min and more. In a majority of
cases, one can observe BL with the observation time below 5 s. Short-lived BL
is observed inside premises and in the open air, but the long-lived ones were
observed only outdoors.
34. A stroke of usual linear lightning often precedes the BL. There are observations
of BL originating in the linear lightning channel. However, observers often
report nothing about BL appearance associated with usual lightning. In most of
the cases, BL had not appeared directly after the linear lightning stroke, but
several minutes after and in the absence of linear lightning.
35. BL often disappears silently and does not leave any traces. Its disappearance
can take place in the air and also when it comes into contact with some objects,
ground, or water. In some cases, BL’s disappearance is accompanied by a
sizzle and burst of flame tips or sparks. In many cases, BL’s life ends with an
explosion. Sometimes, an explosive sound is described as a gunshot or rifle
shot. The explosion is often accompanied by a bright light flash, lightning, and
pieces of matter and sparks flying out of BL.
36. Sometimes, BL divides into several separate balls that either continue their
existence or go out. BL division can take place in the air and at a collision with
some objects. Balls of a smaller size practically do not differ from the initial
ball.
37. There is a description of the case when BL, on the contrary, was composed of
smaller-sized balls, which when moving towards a cable or a wire, came nearer
and formed a luminescent mass.
38. There are rare cases of BL pairs located one over the other and connected by a
“thread” consisting of small luminescent beads.
Besides the conclusions made on the basis of Brand’s descriptions, one can
formulate additional BL features:
39. BL can lose a part of its matter after an impact with an obstacle, but it does not
interrupt its existence as a single whole (Br.14, Br.36, Br.92) (the mark (Br.)
6 Ball Lightning Investigations 207

indicates the book [11], and a numeral signifies the number of the report from
this book).
40. Natural tissues (cotton, cloth) become fragile or tear to pieces after coming into
contact with BL. However, traces of “burns” left on a tissue disappear after
washing (Br.3, Br.95).
41. Tinted tissue becomes colorless after coming into contact with BL (Br.95).
42. As a linear lightning, BL can strike a tree and split the tree trunk, and people
under the tree can get killed (Br.33, Br.146, Br.149).
43. According to a report, St. Elms fires had appeared at mast tops, 2 min after BL
explosion in water at a distance of 2.5 m from the sailing-ship side (Br.47).
44. There is a detailed description of circular holes in a glass produced by BL
(Br.95). “Their diameter as a rule 8 cm, the fracture was even, not the star-
shaped, smooth on the inner surface and exhibited a small sharp edge on the
outer surface. In certain cases, there was a series of concentric wave patterns
around the circular hole.”
45. An electric discharge had occurred between BL and the grounded electric wire
when BL came into contact with it. This caused melting of the wires (Br.136).
46. Sometimes, the disappearance of BL is accompanied by a loud noise; however,
sparks do not fly from it (Br.171). This resembles a collapse of evacuated
sphere.
Brand finished his review with the summary of 14 items, in which he presented
conclusions in a concerted form. We have enumerated them in Items 1–38 men-
tioned earlier. At the same time, some conclusions made by Brand were represented,
which could lead (and unfortunately sometimes lead) to their erroneous interpreta-
tion. In this way, in Item 1, he presented BL definition as “rare forms of spherically
shaped electrical discharges.” Usually, the term, an “electrical discharge,” is under-
stood as the process with the electric current participation and energy delivery to the
discharge zone. The presence of such processes does not follow from the observa-
tions collected by Brand. In Item 8, he describes a possibility of BL “shoot down to
earth” from the lower part of a cloud and its transition to usual streak lightning.
However, there is no such case in his list of observations. Hence, it is difficult to
determine this phenomenon owing to the incommensurability of linear and BL
lightning lifetimes. In Item 10, he describes the existence of two BL types – the
freely floating and the stationary ones, which have different characteristics and can
transform into one another. However, there are no clear evidences in favor of such
BL types among the observation descriptions collected by Brand.
Nevertheless, the indicated inaccuracies do not disparage the great scientific
value of the materials collected and analyzed by Brand. It will be shown in the
following sections that the main conclusions made by him remain practically
unchanged even 85 years after the publication of his book. The massive data
collected by Brand increased with the results of new observations, and only a few
of them gave “new colors” to a complicated BL picture.
It must be noted that Brand did not make any qualitative estimates of energy
contained in BL (or brought by it from external sources) when analyzing cases
208 V.L. Bychkov et al.

executing the mechanical work of BL. This soon became connected with the fact
that there were no detailed numerical data about destructions left by BL in the
sources used by him. However, 14 years after the publication of the book of Brand,
an important event took place, and it forced to fundamentally change the opinion
regarding BL as an insignificant formation with small energy.
47. A large, orange-fruit-sized, red-colored BL had fallen in a tub with 18 L of
water in Dorstone, a London suburb. The water began to boil and it was boiling
for several minutes [13]. A part of the water had evaporated. The BL energy
had been evaluated as 10 MJ based on this “natural” experiment and its energy
density had been found to be 3  1010 J/m3 [14]. Let us consider that there was
no BL explosion in this case, and it had “lived” in the tub gradually giving
energy to water (the number at the beginning of this paragraph indicates that we
have added a new BL characteristic to the list of features taken from the book
of Brand).
In 1965, in the USSR, a popular scientific book of R.A. Leonov, titled “Ball
Lightning Enigma,” was published and 53,000 printed copies were circulated [8]. In
that book, the author described in detail the circumstances of Professor Richmann’s
death, investigated by Lomonosov. Lomonosov thought that BL could come into
the house through an open window of a room adjoining the room where Richmann
was. A clock stopped in the room with the opened window and the sand flew from
the stove. Further, Leonov described several BL observation cases borrowed from
different magazines. These were reports about: a spark that grew to a size of a
10-cm ball in front of the observer [15]; a “golden” ball of 10-cm diameter, which
exploded and caused fire on a fur coat and a shed [15]; BL that flew twice around
people sitting near a table in a hut, which then flew out of a window. There was also
a report about the appearance of a “walnut”-sized bright-red ball from the heated
Russian stove. The ball had flown over a bag with surgeon tools, and subsequently
flew out of the hut. In a yard, it had transformed into white-colored spindle form and
had exploded near a tree, which caught fire. It was determined that metallic tools in
the bag became magnetized; however, “their magnetization had disappeared after a
day.” On the basis of this observation, one can make a supposition about one more
feature of BL:
48. BL can magnetize steel objects close to them.
Apparently, this BL feature is presumed to have caused the stopping of the clock,
as observed by Lomonosov.
A review of J. Rand McNally appeared in May 1966, which was devoted to the
analysis of BL features [16]. The author had distributed a questionnaire about the
details of lightning observations among 1,962 OAK Ridge National Laboratory
employees. One hundred and ten of the questioned people (5.6%) reported that they
saw BL. Subsequently, McNally undertook a wider questioning of all Oak Ridge
Union Carbide Nuclear Company employees (15,923) and received 513 positive
answers (3.1%). The questions in the questionnaire were about BL’s association
with a stroke of usual lightning, its life span, its disappearing process, size, form,
6 Ball Lightning Investigations 209

sound, traces left by it, its motion type, color, and heat irradiated by it. An analysis
of the questionnaire results demonstrated that BL, in majority of the cases, appears
after a linear lightning discharge, and the place of its origin can be metallic
constructions, transmission facilities, nonmetallic constructions, telephone lines,
and free space. It can move along wires, float in air, jump back from earth, rotate
and roll, move straightly, or stay motionless. A spherical form is the dominate one.
Its sizes can range from 2 cm to 2 m, and the most probable size is 24 cm. The total
BL volume shines uniformly in a majority of cases, and rarely, the BL surface
shines. The predominant color group is red–orange–yellow (57.5%), with white
color observed in 28.8% of the cases, and 13.7% of the cases falls under the color
range green to violet. The duration of existence of BL ranges from parts of the
second to 2 min; the average time is about 3 s. It usually disappears quickly (often
with an explosion).
One can observe that the review of McNally practically add nothing to a list of
BL features discovered by Brand. We will also not determine something new in the
list of selected witnesses’ observations. There are reports about BL appearance over
wires of high voltage transmission line at 6 m from a pole and its motion along a
wire (Mn.12, Mn.17, Mn.57, Mn.69, Mn.184); vertical descending of BL on a side
of a thunderstorm cloud; its jump back when it hits the earth (Mn.49) or water
(Mn.64); and BL origination as linear lightning that strikes a tree (Mn.113,
Mn.129), electric fence for cattle (Mn.169), or a telephone line (Mn.184) (here,
Mn. indicates the reference of [16], and the numeral – the number of the observa-
tion). There is an interesting report about BL penetration through a steel mosquito
net (Mn.49) and about the production of hissing sound similar to that produced by a
red-hot iron when BL came into contact with water (Mn.403). Thus, one can add
one more feature to the list of BL features:
49. BL can penetrate metallic net.
A report by a NASA employee, Warren D. Rayle, had been published in 1966, in
which the author questioned the NASA Lewis Research Center employees [17]. The
questioning consisted of two stages. First, a questionnaire was sent to 4,400 employ-
ees with questions about: how often they saw a linear lightning striking a place; bead
and BL; do they like to be outdoors during thunderstorms; and do they like an
observation of lightning discharges. The author received 1,764 answers, revealing
that: 409 people saw a nearby lightning discharge, 180 people saw BL, and 112
people saw a bead lightning. He made a conclusion on the basis of these results that
BL is not a rare phenomenon: the ratio of the number of observed BL and the nearby
linear lightning discharges was 0.44 (and the ratio of bead-lightning number and the
number of nearby linear lightning discharges was also high – 0.27). The second
questionnaire was sent to the people who reported that they saw BL, after the
analysis of the first questionnaire’s results. It comprised the following questions:
did the observer see the moment of BL origination, what was its diameter, how often
did he come across a BL, in which year it happened, in which month, in which time
of the day, where did it happen, was it associated with a thunderstorm, wind, where
was BL at the beginning of the observation, what was its distance, what was its form,
210 V.L. Bychkov et al.

diameter, brightness, and color, for how long did the observation last, did its size and
brightness change, did the observer hear a sound and smell an odor, did he feel the
heat, what was its velocity, did the ball guiding take place, did the ball rotate, how
did the ball disappear, and were there any traces after BL disappearance. One can
note that the number of these questions is much larger than those in the McNally
questionnaire. Rayle received 112 BL-observation descriptions. The main purpose
of the investigation was an attempt to find a correlation between different BL
characteristics. The analysis of the obtained data demonstrated that the average
BL life span was 6 s (whereas McNally obtained 4 s), and a proportion of the balls
that remained for more than 30 s was 12% (McNally obtained 8%). The different-
diameter BL probability observation lognormal distribution graphs were developed.
It was found that the points laid on a straight line (also similar to McNally results)
and the slope of the BL square diameter lognormal distribution lines (i.e., the ball
surface area) coincided with the slope of the charge (transported by the linear
lightning stroke) value frequency lognormal distribution line. The average diameter
was found to be 35 cm (McNally obtained a value of 25 cm). The value of the
distance from which BL was observed sharply decreased after 30 m, and this was
apparently associated with the worsening of its observation conditions. The maxi-
mum velocity of BL motion was about 30 m/s, and the minimum velocity was about
2 m/s. BL most often moved horizontally (54% of the cases), and the vertical motion
was marked only in 19% of the cases. The brightness was most often characterized
as “sufficient to see a ball during the day light,” its form was mainly the spherical
one, and the sphere shined uniformly. The predominant colors were orange and
yellow (while the red color was predominant in McNally analysis). Thirty six per
cent of the observers reported rotation of the ball. It was reported that the brightness
and sizes of BL did not change during the observation in 85% of the cases, and did
not change up to a moment of BL disappearance. A few evidences were obtained in
favor of BL’s ability to contain high amount of energy, and information about bright
flashes and destructions was very rare.
The correlation analysis of the described events’ different parameters did not
allow the determination of their visible connections. The size, shining brightness,
and existence duration proved to be unconnected. The observed weak correlation
between BL diameter and its distance to the observer may possibly be associated
with the errors in defining the indicated parameters and the difficulty in observing
small objects from large distances. Among the 112 observations, one could separate
two groups with 12 cases each. In the first group, BL observation after a stroke
of linear lightning to the earth was reported. The second group consisted of
BL observations first noticed in the open air and remaining there until their
disappearance.
Thus, on the basis of Rayle’s analysis, one can establish new features of BL:
50. Processes responsible for the size, luminescence brightness, color, and exis-
tence duration of BL are practically independent.
51. The value of the BL surface area correlates with a charge value transported by a
linear lightning to the earth.
6 Ball Lightning Investigations 211

On the basis of BL-characteristics constancy during its lifetime (size and lumines-
cence brightness), Rayle made a conclusion that appears skeptical. For convenience,
we call it “Myth No 1” (M1):
M1. As it is difficult to propose a mechanism of uniform stored energy
expenditure, one has to support the idea according to which BL gets energy
from the external source.
The argumentation is rather strange, because in nature and technique, there are
numerous systems where the stored energy expends uniformly in the long duration
(e.g., energy of water from a water reservoir with a dam; benzene energy in an
automobile; and energy of a compressed gas released to the atmosphere through a
throttle).
A new symbolic event took place in 1971 in USA – the book of Stanley Singer,
titled “The Nature of Ball Lightning” [18] was published. After 2 years, its
translation into Russian was published in the USSR. The number of copies of the
Russian edition was about 10,000, it was typical for that time, and the book price
was about a dollar. Owing to this book, the material collected and analyzed by
Brand became available to the readers: the book of Brand became the rarest book by
that time. Singer added to the data of Brand’s new data that appeared during the
30th–60th years of the twentieth century. Thus, the book included a short mention
about the 10-MJ BL-energy case in 1936 [13, 14] and BL-observation description
near Onega river, published by M.T. Dmitriev in 1967 [19]. For a minute, Dmitriev
was observing a flight of a bright yellow-white ball of 8-cm diameter. The ball was
surrounded by two shells; the internal – the violet one was of 1–2-cm thickness, and
the external – the blue one was of 2-cm thickness. The ball moved with a velocity of
1.5 m/s along the rafts in the river. At the bank, its velocity dropped down to 0.4 m/s,
and it hovered above for 30 s. The ball left behind a trace of bluish acrid smoke.
Then, the ball began to move to a side of a forest and ejected sparks at each collision
with the trees. At the end of the observation, the ball was transformed from white to
bright red and disappeared. The noise increased in the transistor receiver during the
approach of BL, which transformed into a continuous roar. Dmitriev took four air
samples after the ball flight from its trail. In the first and the fourth samples taken at
the 55th and 70th seconds after the BL flight, the ozone (O3) and nitrogen dioxide
(NO2) concentrations proved to be at the level of 0.1 mg/m3, but in the second and
third samples taken during the 60th and 65th seconds, the concentration of these
gases proved to be ten times greater. One can formulate two more features of BL on
the basis of Dmitriev’s observations:
52. BL is a source of radio-radiation at frequencies of 0.3–3 MHz.
53. Ozone and nitrogen dioxide are formed in the BL trail with a delay of about 1 min.
Singer described several cases of BL’s encounter with aircrafts. It can fly into a
cabin through an open window and fly to the plane’s tail. It can explode outside near
the aircraft’s fuselage or propeller, and can melt a metal and surround the molten
place with soot. Furthermore, BL can get into a hermetic cabin of an all-metal
212 V.L. Bychkov et al.

aircraft. It often moves by a straight line along a passage to the tail part of the saloon
at a height of about 75 cm over a floor. From this feature of BL behavior in the
aircrafts, one can formulate the following:
54. BL can penetrate inside an all-metal aircraft, and usually moves in the center of
the saloon at a constant height and velocity of 1–2 m/s in the direction from the
cabin to the tail.
Singer devoted a special chapter of his book to the analysis of BL’s typical
features derived from the observations. He considered that the whole set of BL
features contains obvious contradictions: BL can appear both during a thunderstorm
and in fair weather; its color can be arbitrary; sometimes it can move and sometimes
it can remain motionless; sometimes it can move down a wind and sometimes it can
move towards a wind; it can disappear silently and sometimes with an explosion;
and sometimes it can move over objects and sometimes without any support.
Subsequently, he enumerated 14 BL features defined by Brand and the results of
the statistical processing of the questionnaires created by McNally and Rayle. With
regard to BL luminescence, Singer compared it with a black body and a plasma-
generator torch, whose temperatures remained in the range of 2,000–14,000 K. In
addition, in conception of a chemical reaction whose products’ outcome depends on
the temperature, he considered the ratio of ozone and nitrogen dioxide ([O3]/[NO2]
ratio) in the air samples found by Dmitriev. The lowest value of [O3]/[NO2] ratio
proved to be close to 0.8:1, which, according to Singer, indicates that the BL
temperature is 4,000 K. Such reasoning has quite a right to life, similar to a
hypothesis. However, it does not prove that thermal radiation is a source of BL
radiation and that there are some different processes involved. Unfortunately, such
an ungrounded opinion of BL radiation cause subsequently became another myth:
M2. BL irradiates energy as heated black body.
Speaking about BL release of extremely high energy, Singer mentioned its falling
into a tub with water (its energy was estimated as 6–11 MJ) [13, 14] and a destruction
of a cob house at the explosion of BL (with an estimated energy of 400 MJ) [20].
Singer devoted one of the chapters of the book to the analysis of BL photos. A
number of these photos are very small owing to the unpredictability of BL and its
short life span. BL most often appears in a vision field of a photo camera in an
amateur recording of linear lightning discharges. A photographer directs a camera to
some section of a sky where thunderstorm discharges occur, and opens an objective
for several minutes. Sometimes, in developed films, one can find twisting traces on a
background of a landscape, which could be left by moving BL. For example, Singer
showed two photos obtained by this method [18, Figs. 14 and 15]. At the same time,
he warned against false photos in which a continuous trace could appear as a result
of camera motion in the presence of a bright light source in its field of vision; for
example, a street light. In the book, one can also find two BL photos made with short
exposure. In one of them [18, Fig. 17] (see Fig. 6.1), one can see a big oval form
whose large axis is oriented vertically, whose upper and lower parts are somewhat
smeared, and the axes lengths’ ratio is 1.28:1. It is reported that the ball diameter was
6 Ball Lightning Investigations 213

Fig. 6.1 Still photograph taken

during a thunderstorm in 1935
by H. Schneidermann [48]

35 cm and the ball during recording was descending to the earth. Let us make our
own analysis of this photo. One can find that the length of the large axis of the oval is
35  1.28 ¼ 44.8 cm, and the difference of the axes lengths is 9.8 cm. This distortion
of a ball image could occur if the ball, during the exposition, moved for a distance of
4.9 cm. As the exposure duration is not given in the book, let us consider it to be
equal to the average exposure time of a photo camera, that is, 1/30 s. The ball motion
velocity in this case is vb ¼ 4.9  102  30 ¼ 1.47 m/s, and this is a typical velocity
of motion for BL. The smaller level of the negative blackening in the upper and the
lower parts of the oval form can be easily explained by the vertical motion of the
ball. For some reasons, all researchers, including Singer, described an oval but not a
ball. In the second photo represented by Singer [18, Fig. 18], one can see a photo of
BL with a diameter of 50 cm, which is accompanied by ejection of sparks (this case
resembles several descriptions from Brand’s book).
A review by W.N. Charman about the BL features was published in 1979 [21].
Charman described the results of statistical investigations of McNally and Rayle,
and also presented 14 observation examples collected by him. However, there are
no new details with respect to what was known before. Among the new data that
appeared in the literature, he mentioned Jennison’s evidence when BL flew past
him at a distance of 50 cm. Jennison discovered that steel objects in his pockets did
not react to BL’s presence. From this finding, one can conclude that:
55. BL does not reveal a visible magnetic impact at a distance greater than 0.5 m.
214 V.L. Bychkov et al.

Charman gave special attention to the cases of BL release of extremely large

values of energy. He added several cases to those described by Singer with regard to
the tub [13, 14] and house destruction [20]. Among them, there was a report about
BL, whose size was at 60 cm at the beginning and 8 cm at the end of the observation.
It made a trench of 100-m length and 1.2-m depth [22]. A work executed by BL
proved to be equal to 107 J according to the estimates. In the case when BL
descended from a cloud and split a pile of moorage, its minimal energy was
estimated as 105 J [23, 24]. Thus, according to Charman, BL energy can be in the
range from 103 to 107 J. With regard to the BL photos, Charman described the photos
published in Singer’s book [18], and concluded that they practically do not add new
information to that which follows from visual BL observations. Nevertheless, they
serve to confirm this phenomenon’s existence, as until now, some scientists doubt
about the reality of BL and explain its observation using the physiological effects of
the bright object’s image on the retina of the eye.
Besides BL photos obtained occasionally, there are rare cases when they were
obtained in a course of systematical observations. In the 1960s, a system called
Prairie Meteorite Network was effective in the USA for observing meteorites [25].
Sixteen photo cameras were continuously photographing the night sky through
objectives that were opened for 25 ms with a period of 75 ms. When a discharge
of a linear lightning was in the objective’s field of vision, it was recorded during one
exposure time (a duration of a stepped leader in an average was about 20 ms, and a
duration of a lightning return stroke was about 1 ms). Investigators found that in
2 out of 1.2  105 photos of a linear lightning, there were recorded events of some
objects going out of the linear lightning trace, which remained for about 3 s. A trace
of these objects resembled a dashed line consisting of 40 bright sections. It is quite
probable that this trace was left by a moving BL. In another case, when a ball was
recorded near a channel of a linear lightning, it remained for about 180 ms. It was
observed during the recording of a linear lightning discharge by a high-speed
camera [26]. Charman also discussed the question of the possibility of residual
radioactivity in the objects that came into contact with BL. It was shown by the
thermal luminescence method that there were no traces of radioactivity in stones of
the church wall that was hit by BL in 1846 [27, 28].
A book by Igor Pavlovich Stakhanov, titled “Physical Nature of Ball Lightning”
[29] was published in 1979 in the USSR, and the number of copies sold was 23,000.
After 6 years, in 1985, its second edition was published with a title “On Physical
Nature of Ball Lightning”, and the number of copies sold was 50,000 [30]. In 1996,
9 years after the death of Stakhanov, the third edition of his book appeared (in this
case, only 1,000 copies were sold) [31]. The book was written based on the BL-
observation analysis, and the data was collected by I.P. Stakhanov and
S.L. Lopatnikov by questioning the readers of the popular-scientific magazine,
“Nauka i Zhizn” (“Science and Life”). The readers answered the questions in
the questionnaire that was published in the magazine in December 1975. Besides
the descriptions collected personally by Stakhanov in these books, he also used the
materials of observation analysis published by McNally. A main feature of these
books was that they were written by a man who, in the first place, deeply understood
6 Ball Lightning Investigations 215

physics, and in the second place, created his own model of BL. On the one
hand, this plays a positive role, as the author, besides a statement of one or another
BL-feature observation, tried to provide some physical explanation. On the other
hand, he started to excessively strictly consider the facts that did not match his own
theory. He announced them doubtful, and often “spills out a baby with water.”
Examples of his selective attitude towards facts can be found in the preface [31],
where he wrote: “There a ball lightning is chasing a man mad with fear and plays
with him as a cat with a mouse. Here appeared a ball lightning with a diameter of 5,
10, not, even 20 m. There BL destroys a house that is at 50 m from it, another one
breaches a huge tunnel in a ground, the third evaporates water in a tub. This is a
letter in which a man states that his teeth and hair has gone out and signs of the
radiation sickness have appeared after a ball lightning moved by him . . . It is not a
ball lightning. It is a myth of a ball lightning.” One can expect that the author who
follows such reasons can throw away letters describing large-sized BL, BL with
large energy, or how BL impacts people’s psyche or health. It is doubtful whether
impartial investigators, such as Brand, Singer, Charman, and Rayle could venture to
take such a step. However, there is a small probability that such “doubtful” data did
not disappear, because according to a report by I.G. Stakhanova [31, p. 228], there
are about 1,000 unanalyzed observation reports in Stakhanov’s data bank from
which only 126 descriptions were published in [31]. Expecially, in [29–31] it has
no luck the case in which BL got into a tub with water and heated it up to boiling
point [13, 14]. Stakhanov considered that “Goodlet’s estimate is erroneous. Really
there is a possibility that water ‘boiling’ in the tub was connected not with high
temperature of a whole water mass but with gas bubbles formation in places of local
heating (i.e., in the places where BL substance has appeared) at low water temper-
ature . . .. But the witness said that water was hot by touch even after several
minutes. But these feelings could be exaggerated, and real water temperature was
not measured.” Ignoring such evident facts seems to be more than strange, as the
author of [30] was aware of the book by Imyanitov and Tikhiy [32] (it was marked
by number 23 in the literature list [30]). An even more striking case of BL heating
100 L of water to boiling point was described. Furthermore, in this book, attempting
to prove that a mistake was made in BL-energy estimation when BL split the pile of
moorage [23, 24], Stakhanov executed his own calculations and found that a spent
energy of about 100 kJ was involved in the splitting of wood. However, during
calculation, he, for some reasons, considered that the pile length was only 0.1 m.
Nevertheless, by assuming that the pile was longer, for example, 2 m, one can
obtain the energy estimate of 2 MJ.
Disputable from our point of view is the BL problem solution method proposed
by Stakhanov. He presumed “that if we try to explain all facts reported about ball
lightning then we pose an unsolvable problem for us from very beginning. It is not
only the unsolvable but, very likely, a senseless; because at that we will try to
explain also mistakes inevitably included to actual facts. Therefore it is quite
natural to refuse from an explanation of a number of exotic facts at first steps (for
example, as appearance of black ball lightning, ball lightning with a huge energy
release of 10 MJ order of magnitude and higher, or ball lightning with a diameter of
216 V.L. Bychkov et al.

several meters) and to direct towards explanation of main observation mass”

[29, P. 44]. One could accept the proposed principle of “thoughts economy” as
quite reasonable if it was known beforehand that parameters excluded from a
consideration (e.g., those enumerated by Stakhanov) were inessential. As this is
not evident, the opposite principle can be considered to be more accurate: for the
construction of BL theory, one has to use all its features, even those rare ones that
are “omitted” during statistical analysis of the observation results.
Description of the massive data collected by Stakhanov confirms the main
observation results published by Brand, Singer, and McNally. Again, we have
reports of BL of different colors [30, P. 10; 27], absence of heat at close contact
with BL [30, P. 10; 22; 72], BL explosions and sooty spots left by it, separation of
object parts [30, P. 10; 22; 91; 124; 125], melting and evaporation of metals at BL
explosion [30, P. 22; 76; 85], and BL formation near a linear lightning [30, P. 23; 27]
and at a lightning stroke to the earth [30, P. 107; 108]. He also discussed the ejection
of sparks from BL [30, P. 27; 45; 88; 91], BL reddening before its disappearance
[30, P. 27, 127], a form change – transformation of an arrow to a ball [30, P. 75; 121],
appearance of holes in glasses or walls [29, P. 79; 90], and kickbacks at hitting
objects [30, P. 121]. There are reports about BL propagation through small holes
and slits [30, P. 93; 121; 123], barking of trees at BL strokes [30, P. 97; 112], BL’s
appearance from sockets [30, P. 110; 111], and a case when pellets propagated
through BL [30, P. 62].
This last observation, if, of course, a hunter could hit a ball from a distance of
15 m, is a new one and quite interesting. It was reported that the ball after the shot
only shook but left safely. Let us consider a pellet radius rp equal to 1 mm; thus, its
mass is mpl ¼ 4prp3 rPb =3 ¼ 0.047 g (here, rPb is lead density). A pellet momentum
moving with a velocity of 100 m/s is pp ¼ 4.7  103 kg m/s. If ten pellets got into
the ball and each of them gave it 1/10 of its momentum and the ball obtained a
velocity of vBL ¼ 10 cm/s, then the BL mass would be MBL ¼ pp =vBL ¼ 47 g. On
the basis of this “experiment,” one can conclude that:
56. BL is a material body with a mass of 10–40 g. Propagation of small-sized
metallic ungrounded particles through it does not lead to its ruination.
In a similar case, another hunter [33] hit a ball using buckshot from a closer
distance. The ball size increased by 1.5 times; after that, it decayed with a bang, and
its asymmetric pieces of torn form went out falling. This observation also demon-
strates that BL is rather a heavy material body, whose parts are heavier than air, and
that the type of its interaction with metallic particles can be changed with respect to
their energy and size.
Stakhanov, while making a statistical analysis of 884 cases of BL observations,
found that 83% of the observation cases fell on three summer months, and 93% of
the observation cases took place during the period from May to September. This
coincides with the results of Rayle, indicating that BL appearance correlates with
the thunderstorm activity. Stakhanov, by analyzing BL with different lifetimes’
observation probability distribution, discovered that there are two groups of BL:
“short-lived” (57%) with the average lifetime t1 ¼ 11 s and “long-lived” (43%)
6 Ball Lightning Investigations 217

with the average lifetime t2 ¼ 54 s. Furthermore, he discovered two similar groups

by analyzing McNally’s data. The BL-diameter probability density distribution in
the case of Stakhanov’s data practically coincided with the analogous McNally’s
and Rayle’s data distribution. The most probable diameter value proved to be 10–20
cm. It was revealed that about of 50% BL had time to cover a distance of about 50 m
during an observation time. About 75% of them moved horizontally and smoothly
(about 83%), and the most probable velocity was 0.5–2 m/s. “Guiding” – BL
motion along the lengthy conductors – was observed approximately in 20% of the
cases. Color in 60% of the observation cases could be attributed to the red edge of
the spectrum (red–yellow), and 15–20% belonged to the short-wave edge of the
spectrum (green–violet). Stakhanov presumed that “a color of ball lightning is not
its typical feature, and in particular say nothing about its temperature and com-
position.” We also agree with this conclusion of Stakhanov. A light flux from BL,
on an average, can be compared with that from an incandescent lamp of 100 W.
Stakhanov conceded that “ball lightning can emit soft ultra-violet radiation, by
absorption of which in surrounding air one can explain the observed blue halo.”
After making photometric measurements of BL trace in a photographic film,
Stakhanov concluded that light freely comes out of the internal part of a ball. It
means that BL is transparent for visible light. In the data collected by Stakhanov, an
amazing BL feature was confirmed: by radiating light, it radiates almost no heat.
Only 8.5% of the witnesses at a distance less than 1 m from the ball reported about
feeling the heat (but hardly noticed). This indicates that BL temperature cannot be
equal to a thousand or several thousands of degrees. BL, according to the observa-
tions, is a source of radiation in radio-range. Stakhanov presumed that the radio-
radiation “has essentially non-equilibrium character” (i.e., heated body cannot be
its source). A valuable conclusion about BL features was one that “ball lightning
substance possesses surface tension.” However, he could not indicate a reason for a
surface tension existence and proposed to accept it as the fact [30, P. 118]. The
presence of surface tension with a value higher than s ¼ 106 J/cm2 allows to
explain the absence of BL substance diffusion to air during its motion (absence of
Rayleigh–Helmholtz instability). Besides, the surface tension also allows saving
BL from the convective instability (Rayleigh–Taylor instability). The presence of
the surface tension enables “ball lightning substance to create a separate phase in
air, different from a gas surrounding it. . . . Most time of its life a ball lightning has
sufficiently clearly defined boundary.”
Along with the deep conclusions set forth, Stakhanov had great doubts regarding
the assumption that BL contains energy greater than 100 kJ. With the same persis-
tence, he did not accept any presumptions that BL can carry high uncompensated
charge. However, by trying to explain BL motion, he supposed its presence, but in
“homeopathic” doses. Thus, for an explanation of BL rising to a thunderstorm cloud,
he considered that it “has to carry a positive charge similar by the sign to those of
earth surface” [30, P. 63]. Assuming BL mass to be equal to 2 g and thunderstorm
electric-field strength of 105–106 V/m, with respect to Stakhanov’s estimates, one
can observe that BL charge should not be lesser than Q ¼ 2  108  2  107 C for
the compensation of gravitation force. The electric capacity C of a ball with a radius
218 V.L. Bychkov et al.

R ¼ 10 cm equals to C ¼ 4pe0R ¼ 1.1  1011 F, and the electric energy stored in the
capacitor carrying this charge Q is We ¼ Q2/2C ¼ 2  105  2  103 J. BL – a
capacitor with such insignificant energy – cannot express visible electric effects: to
melt metallic substance, electric current should strike a breather.
As these effects undoubtedly take place, Stakhanov had to find some explana-
tions for them. The following is the essence of his explanations. He stated that “ball
lightning possesses a feature to take off charges from conductors. At that, it has to
be a conductor itself and the conductor with low work function of charges” [30,
P. 96]. For example, “a tree splitting can take place under an influence of a current
pulse that appears at the contact with ball lightning. Energy releasing at ‘explosion’
in this case does not relate to the energy stored in ball lightning itself. Releasing
energy is stored in charged conductors, and the ball lightning is a trigger for release
of this energy . . ., and the energy can be released also at large distance from the
place of ball lightning contact with the conductor.” Let us try to understand where
the charge can appear from, which BL discharges through itself. Let the lower part
of a thunderstorm cloud carry the negative charge of several tens of Coulomb; then,
the positive charge of the same value will be gathered on the earth surface under the
cloud. An area of such a charged “spot” is approximately equal to the area of the
lower part of the cloud (S ¼ 1,000  1,000 m2 ¼ 106 m2), and a distance between
the center of the spot and the uncharged area of earth may be about 1 km.
Stakhanov, for the estimation of the maximal electric energy store in the “spot,”
considered its radius as 10 m, and a value of a charge induced in it as 103–102 C
[30, P. 104]. Thus, he seemingly considered that an area at 5 m from the “spot”
center stays uncharged. Later, he considered the following: BL whose size is 20/0.2
m ¼ 100 times smaller than the “spot” diameter gathers charges by some incom-
prehensible means from the “spot” and carries them to the uncharged part of the
earth. Those of us who “short-circuited” charged capacitors using a piece of wire
can clearly understand that if a distance between the condenser clips is 100 times
greater than the length of the wire, then it is impossible to discharge the capacitor
through this way. Hence, one can consider the process, according to which

M3. “Ball lightning consists of a conducting substance with low working

function which possesses a feature to easily scatter charges at the contact with
charged conductors”, and “powerful current pulses appearing at that . . . are
capable to scatter considerable charges induced during a thunderstorm over a
large territory – of area . . . of tens . . . and hundreds of square meters,”
as a type of myth about BL.
Also, another Stakhanov’s conclusion about BL is still unproven, which one can
consider as another myth.

M4. “A density of ball lightning is approximately equal to air density”.

(See the remark to the feature 56).
A book by Imyanotov and Tikhiy, titled “Beyond Boundary of Science Laws”
was published in 1980 in Moscow (the number of copies sold was 50,000) [32].
6 Ball Lightning Investigations 219

In that book, a form of a conversation between a scientist and a journalist is

presented, along with the analysis results of 500 BL reports collected by Imyanotov.
As in fiction, the book starts with a story about the disappearance of a golden
bracelet from the hand of a woman during her sleep, the bracelet which she could
not remove for several years from her grown plump hand. She was awaked by
“a sound of a thunderstorm discharge similar to a shot.” When she came to a
window, she noticed a hole of about 3-cm diameter in a glass and “suddenly noticed
that the bracelet had disappeared from her hand. Only a dark wale was on her wrist.
A circular glass with fritted edges was found on a floor.” Imyanotov made an
estimate of energy value necessary to heat, melt, and evaporate the golden bracelet
of 50-g mass, and found that the bracelet heating to a temperature of melting point
of gold (1,336 K) needs to spend 6.3 kJ, its melting requires 3.3 kJ, temperature
rising to the boiling point (3,239 K) requires 10.5 kJ, and vaporization requires 87
kJ. Finally, the energy spent for bracelet vaporization was found to be 107 kJ [32, P.
94]. As there was no burning of the hand, the process of bracelet disappearance was
very quick. If it lasted no longer than 102 s, then the power realized by BL was not
lesser than 10 MW. Brand, in his book, also discussed cases of traceless “vaporiza-
tion” of golden chains, rings, and other round objects [11, P. 8]. Apparently, these
adornments were actually “vaporized”, and were not stolen.
Thus, one can formulate one more feature of BL:
57. When passing over ring-type metallic objects with diameters from some milli-
meters to several centimeters or when exploding near them, BL can instantly
vaporize these objects.
Imyanitov’s review added some new information to BL maximal energy esti-
mate. It was reported that BL at an explosion separated into two balls, and “the
explosion was so strong that it slightly raised a ceiling and slid walls of the house
apart” [32, P. 31]. The more interesting was the description of an event that was
very much similar to the case in Dorstone [13], which occurred near Perechin city in
Western Ukraine. S.S. Mach, a reporter of Imyanitov, wrote the following: “In
August of 1962 near 11–12 o’clock p.m. a ball lightning with a size of a tennis ball
had fallen into a trough with water for cattle; it was shining by rainbow colors
during 10 s. Water from the trough was vaporized almost completely and boiled
frogs were on a bottom. Sizes of the trough were 0.3  2.5 m. Depth of a water
layer was 15 cm. Boiled frogs were also found in two other troughs.” A calculation
of energy spent by BL for heating and vaporization of water led to a result of
270 MJ, and the energy density (if we accept the radius of the ball as 3 cm) was
2.4  1012 J/m3. Based on this “natural experiment,” one can formulate two more
features of BL:
58. BL can release an energy value of 270 MJ, and that its average energy density
can be 2.4  1012 J/m3.
59. BL energy absorbed by water can be transported through a dielectric wall
(in this case, through wood).
220 V.L. Bychkov et al.

In this case, it must be noted that there were no doubts about water heating to
boiling temperature, and it did not lead to bubbling. The frogs that gave their lives
for the science were the indicators of temperature rise.
In the book by Imyanitov, there is one more report exhibiting the magnetic
features of BL [32, P. 72]: “It was physics exam for 7-th grade. A thunderstorm was
approaching. Magnets of different forms and other devices were on a table.
Suddenly . . . a small cloud flew in through the open ventilator window . . . its
color was bluish–violet, it approached the table where aids were lying, it immedi-
ately rose and again flew out through the ventilator window without breaking the
window. When the cloud rose from the table, all of us saw that the magnets rose and
flew through the ventilator window as if they had life in them. One horseshoe-
shaped magnet made a hole in the iron tank and the other one fell down . . . and
deeply went into earth.” On the basis of this description, one can define three
additional features of BL:
60. In the presence of magnetized objects, BL can also become a magnet and be
attracted to the magnet.
61. The magnetic features of BL can unexpectedly disappear even in the presence
of the magnetized object.
62. BL is capable to hold objects with a mass of about 1 kg near itself and raise
them in air.
One can estimate a height H, to which BL could raise two magnets. Let
us suppose that a projectile of a mass of ms ¼ 50 g moving with a velocity of
vs ¼ 100 m/s can make a hole in a wall of the iron tank. Let us consider that a kinetic
energy of the falling magnet of a mass mm ¼ 0.5 kg is equal to the energy of
the projectile. Then, the height from which the magnet could fall down is H ¼
msvs2/2mmg ¼ 50 m (here, g is gravitational acceleration).
In other cases, BL features derived from the observations collected by Imyanitov
agree well with the data observed by other investigators. Also, cases of roar at BL
explosion, production of the smell of sulfur, noises in electric net at BL moving by,
its appearance from radio-sockets, creation of round holes in windows, motion of
BL along the wires, melting of metallic objects at BL explosion, and the mass
deaths of four-legged animals when BL gets discharged to the earth, have been
reported. Surprisingly, during the two centuries of BL observation, nobody reported
about the death of poultry, that is, biped beings. This may be owing to the pace
voltage striking factor when BL strikes the earth; it releases strong currents that
spread in the earth. However, Imyanitov, similar to other investigators of his time,
accepted the obvious for the real: he considered that a fact of free soaring of a ball
could be explained only by the equality of the Archimedean raising force and the
ball weight [31, P. 96].
In the Journal of Technical Physics in 1981, an article in which a case when BL
of D ¼ 150 cm hit the earth during a thunderstorm observed by several people was
described, which then raised up [34]. A layer of caked earth with a volume of 0.4 m3
was formed in the soil in the place of BL touching the earth. The authors of [34]
modeled the formation of an earth layer impacting the soil by different means, and
6 Ball Lightning Investigations 221

came to the conclusion that most close structures realized in nature appear at a high
frequency radiation zone of the earth. From their estimate for caking of a piece of
earth of indicated volume, it was found that one requires an energy of W ¼ 109 J.
Let us consider that BL spent only a part of its energy to the indicated action, and in
this case, a density of spent energy rW ¼ 6W/pD3 ¼ 5.7  108 J/m3 proved to be
high. Certainly, there are no direct evidences to show that BL impacted earth
through electromagnetic radiation, and thus, other mechanisms of energy transfer
are possible. However, in any case, one can consider that a level of energy value
was estimated correctly.
A book by James Dale Barry, titled, “Ball Lightning and Bead Lightning” was
published in 1980 in the USA [35], and in 1983, its translation into Russian was
published (with 6,000 copies being sold). Unlike previous monographs on BL,
where the authors used the data that they have personally collected, Barry did not
present a review of reports about BL observations and information from a vast
literature list consisting of 1,979 references compiled by him. Instead, he dedi-
cated a substantial part of the book to BL photos and processes that accompany it,
and to the analysis of the experimental attempts to realize BL in a laboratory.
Based on the works of previous investigators, he gave a short review of BL
characteristics: its form and size, color, structure, motion, sound, and smell. He
discussed the traces of destruction and heating left by BL, its lifetime, processes
of decay and connection with linear lightning, and its tendency to get into the
closed premises. With regard to BL density, he again repeated an old idea that it
is close to those of air. Barry represented data of different authors on the
estimation of energy density of BL that got into the tub water [13, 14], which
proved to be in the range from 2.5  109 to 1.9  1010 J/m3. With respect to the
case of splitting of the pile [23, 24], he found the energy value of 1.5  105 J and
the energy density of 8.5  107 J/m3. He also analyzed other cases where it was
possible to determine BL energy. Their BL-energy density proved to be in the
range from 4  105 to 1.7  109 J/m3. With the help of his energy-estimates
results, Barry built a graph of BL-observation probability lognormal distribution
with definite values of energy density. The experimental points were located on a
straight line. From this fact, he concluded that there is only one type of BL in
nature. Unfortunately, this conclusion has not yet been comprehended by BL
investigators. For example, O.A. Sinkevich again wrote about BL as a physical
object “which realization mechanisms . . . can have different nature” in the
afterword to the third edition the book by Stakhanov [31], published 16 years
after the printing of Barry’s book [35].
In contrast to Stakhanov’s conclusion, Barry considered that BL color “can be
connected with the radiation characteristics of the absolute black body.” Thus, the
red BL should have a temperature of 4,600 K and the white one – higher than
10,000 K. However, Barry also allowed another possibility: “that application of
color temperature to a phenomenon of ball lightning . . . might be unacceptable”
[35, P. 78]. With regard to the reason for the church bell swinging deceleration, he
made a supposition that one had to blame BL magnetic field with the induction of
1.5  102 T [35, P. 81].
222 V.L. Bychkov et al.

Barry represented a large number of BL traces photos, many of which could

be faulty owing to a camera motion during a photographic survey. Unfortunately,
with rare exception, they do not bring something essential for the formation of BL
database.

6.1.2.2 Discussion of Ball Lightning Observation Results During the First

Symposium on Ball Lightning in 1988

The above-mentioned 62 features of BL can be considered as “a scientific store”

that was available to investigators gathered at the First International Symposium on
BL in Japan, in July 1988 [1].
Stanley Singer made a report “Ball lightning – The continuing challenge” at the
Symposium [1, P. 1–18]. He described the BL phenomenon investigation history,
starting from the first publication of F. Arago. The BL problem has passed a period
of negation, reducing it to optic illusions, and a skepticism period rose on behalf of
an outstanding scientist, M. Faraday. “In view of the skepticism which has dimin-
ished but not disappeared in recent consideration, it is necessary to assess the
significance of the phenomenon. Certainly, the repeated observation in nature of
a striking occurrence which has not been clearly explained demands a serious
interest.”
Unfortunately, this phenomenon, owing to its unpredictability, is impossible to
study with the help of devices. The only data left was the information from
unprepared observers, who could report nothing but the size, velocity, color,
and the life span of an object. Observations indicate a connection of BL with
thunderstorms and a very small probability of their observation, which was about
3  1011 km2s1 (for usual lightning, it was 3  108 km2s1). A ball usually
has a red or orange color, its diameter is 30 cm, and its average velocity is 2 m/s.
The estimates of its energy density gave values from 5  105 to 105 J/cm3.
In spite of the appearance of about five articles a year devoted to the new
theoretical models, neither of them could explain all the features of natural BL.
Successful experimental attempts to reproduce BL were also quite limited. Plasma
formations (plasmoids) were obtained in electrical discharges, including a magnetic
plasma confinement in installations of spheromak types. Fire balls were observed at
ignition of lean mixtures of flammable gases with air, and at explosion of metallic
wires during the transmission of high currents. “Our understanding of the ball
lightning phenomenon has progressed significantly since the . . . investigations in
the early 1800s, yet we lack knowledge of the mechanism of formation and the
structure.”
George Egely (Hungary) made two reports: “Hungarian Ball Lightning Obser-
vations in 1987” and “Physical Problems and Physical Properties of Ball Light-
ning”. Egely, in the first report [1, P. 19–30], described nine cases of BL
observations; its energy in three of them was determined as exceeding 10 kJ.
Grey BL in the first of them had a diameter of 2 m, which exploded in an attic
floor of a house. The explosion slightly raised the roof with a 35  15 m2 and threw
6 Ball Lightning Investigations 223

it down to a field of a neighbor (man). After that, according to the observation of the
neighbor (woman), a reddish ball had flown from the destroyed house and began
moving horizontally over the field. The mass of the 1 m2 of the roof was 20 kg. If
one considers that the roof was raised for about 1 m, then a work executed by BL is
about 100 kJ. It must be noted that BL continued its existence after that. In the
second case, an explosion of a yellow ball of 20-cm diameter was reported. As a
result of the explosion, aluminum wires of a concealed wiring in the walls melted,
their total length was estimated to be about 100 m, and the diameter was 2.5 mm.
Starting from this, a minimal BL-energy density was estimated as 350 J/cm3. Other
reports described similar destructions occurring at BL explosions in closed pre-
mises: wires melting, electrical bulbs exploding, brick walls moving, and sooty
spots appearing. There were also reported cases of spontaneous BL formation
without a stroke of usual lightning. BL in one of them unexpectedly turned up
from a lake; in another, five balls of diameter of 10 cm each had unexpectedly
appeared on a bank, and then gathered to form a ball of 25–30-cm diameter. In the
third case, glowing balls of 20-cm diameter appeared one after another in a garden
near the earth, and they unexpectedly disappeared and again silently appeared at a
height of 20 m.
Egely, in the second report [1, P. 81–87], grouped the main BL features and
demonstrated the difficulties in their theoretical explanation. Thus, according to
him, models that considered the presence of surface tension cannot explain the
existence of BL in a form of a pear or ellipsoid [36]. Models in which BL
represented ring currents [37] or rotating dipoles [38] cannot explain the reason
for the decrease in BL size when it penetrates through small holes. The color of BL
does not change even at its interaction with metallic objects or at substantial energy
losses. This lowers the belief on models in which BL color is connected with its high
temperature as well as the chemical and plasma processes. BL-energy density can be
of 0.6 kJ/cm3 and even 34 kJ/cm3, when the maximal energy density for chemical
and plasma models is only 0.2 kJ/cm3. Even holes in glasses and walls left by BL
cannot be reproduced in laboratory conditions. We cannot understand the reason for
the electric effects realized by BL. If one considers that electric-field strength on
the surface of BL of 20-cm diameter does not exceed the breakdown value of 2.5 
106 V/m, then the BL charge can be considered to be equal to 3  106 C, and its
electric energy may only be 0.8 J. At the same time, it is known that a current
conditioned by BL charge can cause the melting of 100 m of aluminum wire of
1.5-mm diameter. If this effect was caused by the current with a duration of about
1 s, then the charge of about of 80 C would have been transported through the wire.
Ohtsuki and Ofuruton (Japan) [1, P. 31–57] reported about BL observations in
Japan. The Japanese Information Center of the Ball Lightning (Fire Ball) started its
work in July 1986, and has collected 2,060 evidences of witnesses from all over the
islands of Japan. It turned out that 89.7% of BL in Japan were observed in fair or
cloudy weather without rain, in contrast to those of Europe and America where the
substantial BL part appeared during thunderstorms. Nevertheless, thunderstorms’
probability and BL observations over a month’s distribution strictly correlate, and
their maximum falls on July–August as in Europe. The authors found that high air
224 V.L. Bychkov et al.

humidity in Japan – the humidity in summer rises to 80% even in fair weather may
be the reason; however, the higher humidity in the Continent occurs only during
thunderstorms. Other BL characteristics – size, color, form, and lifetime – turned
out to be the same as in Europe and America.
Alexander Keul (Austria) made a report, “Phenomenological and Physiological
Analysis of 150 Austrian Ball Lightning Reports” [1, P. 58–80]. Austrian BL,
similar to others, demonstrated some unusual features along with typical, often
repeated ones: there are cases of one ball division into several ones; BL motion as
“a rubber ball”; destruction of walls, and creation of holes in them. Keul undertook
a computer analysis of 150 selected reports using nine parameters concerning
observation conditions and the main BL features. Subsequently, he developed the
following conditions: (1) If the diameter of the observed BL turned to be larger, the
distance to it was longer and (2) If the feeling of fear when seeing a BL was lesser,
then the BL was farther. Both these effects have a physiological basis under them:
the larger object – the more attention it attracts at large distances, and small objects
can leave unnoticed. BL observations in Austria correlate with the frequency of
lightning and the density of population. BL was observed both in the plains and on
high land in the country. BL investigations are based on the analysis of observations
and not on laboratory experiments; thus, a specialist in plasma physics may be as
helpless as a specialist in psychology and vice-versa. Besides, there is an additional
difficulty: there is no clear definition regarding what is BL.
An article by A.I. Grigorjev, I.D. Grigorjeva, and S.O. Shiryaeva, titled,
”Statistical Analysis of the Ball Lightning Properties” [1, P. 88–134], was pre-
sented at the Symposium. The authors of the article have collected 2,082 descrip-
tions of BL in Russia; they have executed a statistical analysis of materials by the
method proposed by Stakhanov. They had an opportunity to make a sampling of
observations where BL was observed: (1) from a moment of its origination till its
death, (2) sometime after linear lightning discharge, and (3) when it went out of a
field of vision. They found that one can represent a decay probability dependence
in a chosen group of BL as a combination of three exponents (but not two, as in
Stakhanov’s analysis) with typical times t1 ¼ 3 s, t2 ¼ 29 s, and t3 ¼ 213 s. A
predominant BL form was a sphere (91.2% of the cases out of 2,013). There were
cases of a ball transformation to a tape (18.66% of the cases out of 134) and vice
versa, from the tape to the ball (11.19%). A total of 119 witnesses reported that
BL had a tail. In 266 reports, information that it had a coat (shell) of 3-cm
thickness was observed. It was reported in 143 cases (6.9%) that BL had an
internal structure: bright, luminescent points chaotically moving inside the coat,
or beams of bright filaments (threads). A form of diameters probability distribution
dependences coincided with those found by McNally, Rayle, and Stakhanov.
However, the distribution maximum moved from 20 to 7.5 cm. There were cases
of BL separation into several parts (3.5%), and that it bounced back from the
obstacles like a rubber ball. BL in 83% of the 319 cases did not emit heat, and in
other (17%) cases, witnesses indicated a feeling of heat. In three cases, BL caused
burns, similar to those of ultraviolet radiation. The authors came to a conclusion
that BL appearance is closely connected with thunderstorm activity. Cases of
6 Ball Lightning Investigations 225

BL observations in fair weather can include errors: (1) Observers considered a

weather to be “fair” always when the sun shines; (2) Thunder can be heard from a
distance up to 10 km; however, overvoltage waves can be propagated over high-
voltage (power) lines (BL origination on a conductor took place in 68.53% of the
286 cases); and (3) Witnesses can forget weather conditions because of event
remoteness.
We found that by 1988, the investigators’ team from different countries has
made a large work on collection and analysis of BL observation data. It was
revealed that the data from different investigation groups practically completely
coincided. This favorably testifies that a subject of investigations was a unique
natural phenomenon. However, a search of correlations between different charac-
teristics of BL gave an ambiguous result: only evident connections of BL with other
events were determined, such as with thunderstorms, connection of an observed
object with a distance to it, and a feeling of fear with respect to the distance to BL.
At the same time, there are no clear correlations between properties such as size,
life span, energy, color, and radiation intensity. Besides, there is no clear definition
of an object that is understood as “ball lightning,” as it was truly noted by Keul in
his report. In a majority of cases, we could distinguish BL from St. Elmo’s fires,
corona discharge, meteorites, electric arc, swamp lights, etc. However, in our
opinion, undefined boundaries are left, in which an object can be attributed to
BL. For example, if the average diameter of BL is 20 cm, then what one can say
about a spherical object with a diameter of 20 or 100 m? Is it a BL or something
else? or How to call an object that is not observed during 10 s but during 1 h?
Here, we encroach upon an area of unknown flying objects (UFO) observations.
An important part of UFO is the optic illusion, as it was shown by Menzel [39]
(remember that BL was also for a long time considered as optic illusion by
scientists). However, some of the UFO observations are reliable, and it is striking
that the main UFO characteristics coincide with those of BL. Thus, the following
dilemma appears: either to include observations of UFO into the data bank on BL or
to ignore these results. The first step in the positive side has been made during the
First Symposium on BL – participants listened to the report by Y.S. Zou on UFO
observation in China [1, P. 273–287].
On March 18, 1988, at 21 h 35 min, the Beijing pilots and passengers of aircraft
TU-154 during the night flight at height of 11,000 m over a mountain region saw a
strange fire ball to the right of the aircraft and somewhat higher than it. The ball
emitted a bright beam of white color resembling a beam of a searchlight. People
considered that UFO was flying towards the aircraft, which could be seen better and
better. Till that time, all aircraft electronic devices and radio were working normally.
The flying object turned to the left after 3 min and began to move from the aircraft.
After several minutes, it separated into two parts; a ball was created from above, and
under it, there was a part in the form of a kidney bean. One of the objects was
rotating with a high speed and was surrounded by a green ring. After that, both the
bodies associated into one, and the green ring surrounding them left safe. Then, the
flying object began to become smaller and smaller. At 21.48, it disappeared among
the stars. Thus, the object was observed for 13 min. The air velocity at a height of
226 V.L. Bychkov et al.

11,000 m was 27.8 m/s, and it blew towards the aircraft and approximately perpen-
dicularly to the flight direction of the flying object. The air temperature was 50 C.
P.A.M. Koloc, in his report [1, P. 289–309], devoted to explain his own BL
model as a magneto-plasma configuration in which there are relativistic electrons,
had cited a report on a case that took place in Venezuela in 1886 [40]. He stated that
in that case, the symptoms of radiation sickness were observed in a group of
inhabitants after their contact with BL: vomiting occurred and lips swelled. “The
next morning the swellings had subsided, leaving upon the face and body large
blotches. No special pain was felt until the ninth day, when the skin peeled off, and
the blotches were transformed into virulent raw sores. The hair of the head fell off
upon a side which happened to be underneath when the phenomenon occurred, the
same side of the body being, in all nine cases, the more seriously injured.”
On the basis of this description, one can suppose that
63. BL can cause radiation-sickness symptoms in humans.

6.1.2.3 Results of Ball Lightning Observations Published After 1988

Now, let us consider the new traits in the BL “portrait,” which brought observations
made after the First Symposium.
In Yaroslavl city (USSR) in 1990, Professor A.I. Grigorjev published, on his own
account, a small book, titled, “Fiery Killers” (of 7,000 copies) [41]. Having a
databank consisting of 6,000 cases of BL observations, the author represented
descriptions of frequently repeated and rarely encountered features of BL. The
book included information about the death of people and animals caused by the
electric current (Gr.1; 3; 5; 10) (here Gr. means the book [41], and the numeral is
the number of the report from this book). There is written information about the
appearance of burns after contacts with BL (Gr.7; 8; 9; 35) and about the severe
destructions at BL explosions (Gr.2; 3; 6; 8; 13; 60). In Grigorjev’s collection, there
are about 50 reports mentioning that BL penetrated window glasses without causing
any damage (Gr.30; 31; 32; 33; 35). He represented many descriptions on how BL
went out of electric sockets or electrical devices (Gr.11; 20; 24; 25; 26; 30; 43; 47; 49;
50; 54), or rolled up as a ball after initially being a tape (Gr.19; 21; 22; 61). This
character of rolling is complicated: BL thread rolls up the coil or hanks; hence, the
character of the tape transformation in the ball is different from the opposite process.
There are reports that BL ejected sparks or small BL (Gr.17; 46; 53; 54; 57); however,
there is no case in the collection with regard to the confluencing of BL pieces. All this
information has allowed Grigorjev to add new features to the list developed by Brand:
64. BL can possess high electric potential and electric charge.
65. At the explosion, it can cause the appearance of a very large induction electro-
motive force in the surrounding wires.
66. During the process of its existence, BL can emit electromagnetic waves in the
radio range.
6 Ball Lightning Investigations 227

67. It can be a source of hard ultraviolet and soft X-ray radiations.

68. BL can appear on wires at electrical overloads.
In Grigorjev’s opinion, BL “represents . . . a charged plasma – a highly ionized gas
with domination of unipolarly charged ions” [41, P. 27]. Grigorjev considered that
BL’s ability to store high energy is its very important feature: “in the football-sized
BL can be stored the amount of energy that is released at the explosion of some ten
kilograms of trinitrotoluene.”
At the beginning of the 1990s, Grigorjev wrote a new book – the monograph
“Ball Lightning.” However, because of events that took place in Russia during
those years, it was published only in 2006, with a modest copy number of 200 [42].
Concerning the creation of a theoretical model, Grigorjev considered that it is
unproductive to create “averaged BL portraits.” In his opinion, “the ideal situation
for the investigator’s entry to the BL problem is such when he makes himself a list
of its features basing on a sufficiently large data of detailed descriptions of ball
lightning’s behavior in natural conditions, the features that later will be lain down in
the basis of the future model.” Following this, Grigorjev represented in his book
“a set of ball lightning behavior descriptions in natural conditions concentrating
attention on its features that are not in a scope of standard averaged portraits.”
In the first chapter of the book, Grigorjev described the general, often repeating
BL features. He considered that the BL portrait created by Brand [11] is “the most
compact, complete, detailed list of ball lightning properties that is not distorted by
false constructions of the author.” However, Grigorjev stated that the list of Brand
is not precise. “Thus, apparently BL that is attached to or is sitting on a conductor
does not exist”; sooner, this was found to be St. Elmo’s fires. Furthermore, BL’s
size and form are not always constant (Grg.1; 2; 5; 9–11), and its color can be
changed (Grg.2) (here Grg. means the book [42], and the numeral is the observation
number from this book).
An analysis of BL origination and disappearance as well as the description of
facts represented in the second chapter can give very important information about
BL’s nature. Among 5,315 descriptions collected by Grigorjev, witnesses in 1,138
cases saw the birth of BL lightning. In 8% of the cases, it appeared from a channel
of linear lightning, and in 8% of the cases, it appeared in the place where the linear
lightning hit the earth, while in 4% of the cases – in clouds and in 66% of the cases –
on metallic conductors. Grigorjev considered that BL can originate at the linear
lightning channel decay through an intermediate state – bead lightning type, or can
throw itself out of the incomplete branch of the lightning channel (Grg.21; 23;
24–27). However, in the predominant number of cases, BL origination was marked
on electrical wires, radiator’s tubes, loudspeakers, radio sockets, and electric sock-
ets (Grg.44–67). “This can occur at sharp rise of the conductor’s potential when the
overvoltage wave that has appeared in it (e.g., at nearby discharge of a linear
lightning) comes to an end of a long wire.” BL at explosion can scatter sparks
(Grg.74–80), can soundlessly thaw in air (Grg.81–91), or go into earth or a
conductor (Grg.88–91). BL carries some uncompensated electrical charge and is
a “material object, sooner consisting of some variety of plasmas.” Therefore “an
228 V.L. Bychkov et al.

ability to pass through windowpanes not leaving holes in them” is very intriguing
attributed to it. This BL feature was for the first time marked by Brand, and among
the 5,315 descriptions collected by Grigorjev, there are 42 such cases
(Grg.100–115). The only material objects capable of penetrating through a glass
are electric and magnetic fields, radiowaves, and visible light. This, by some means,
can decrease a number of BL-energy carriers. In the list of observations collected by
Grigorjev, there are 26 events reporting that BL left holes in glasses (Grg.117–132).
Disks that fell out of glasses have “sharp edges and do not carry any traces of
melting.” There is still no answer to the question “about a physical mechanism of
ball lightning penetration glasses without their damage . . . a possibility of the
answer to it is closely connected with a clarification of a substance nature of
which it is composed.” The data obtained from witnesses who saw BL nearby are
very valuable. In the descriptions collected by Grigorjev, there is a report about a
“ball with a core of 30 cm diameter . . . Something was rotating inside the core
leaving streaks of white threads after itself ” (Grg.133). A ball in another case has
penetrated a metallic net against mosquitoes: “it resembled a tangle of golden-
orange threads or dots, to be more exact” (Grg.134). There was an observation of
two balls with clear boundaries, “a flame was boiling filling a whole volume behind
their transparent covers. . . . The whole mass was in continuous motion” (Grg.135).
There is a report of an observer, who at a distance of 5–7 m “at a height of two
meters over earth” saw “something resembling a sheaf consisting of a multitude of
balls as if of melted metal. Each yellow with redness ball was separated from
another one, and slightly trembling and moving in the ‘sheaf’ it did not connect to
others, but to the contrary it, as if, slightly pushed them away. ‘The sheaf’ had a
form of irregular oval of two-three meters size. . . . All the balls were absolutely
identical” (Grg.136). There were balls resembling “a tangle of bright threads”
(Grg.138) or “consisting of stirring small white-red sparks“ (Grg.139). Grigorjev
stated that BL can be separated into sparks or balls at collisions with obstacles; this
makes one to suppose that they are “elementary bricks of which a ball lightning is
composed . . .. A small ball can be considered as a small ball lightning.”
With regard to BL radiation detection, Grigorjev informed that only 1.5% of the
witnesses who saw BL from the distance lesser than a meter (35%) reported about
the presence of a thermal flow from it. Fires and burns to people owing to BL
occurred only at its explosion. A more interesting BL feature is “its ability to emit
electromagnetic waves in a microwave and radio-wave ranges.” Thus, a ball
moving by a room antenna of a TV set “has caused strong interferences: streaks
began moving over a screen, an image was distorted, and one could hear crackling”
(Grg.141). It was told that the hairs of girls rose when the ball of 8–14 cm diameter
was passing over their heads. Their hairs dropped down when the ball flew away.
“An incandescent hair in a bulb began to glow reddish color when a ball was
passing by a lamp, though a switch was shut downed . . . When the ball has flown
away then the incandescent filament went out” (Grg.143). There is an interesting
report about BL formation process. “A trolleybus’s bow has come off. It several
times hit wires causing short-time shorting, crackles have sounded, and sparks
began to fall. Suddenly a bright, round ball of 12–20 cm in a diameter has separated
6 Ball Lightning Investigations 229

from a total mass of sparks. . . . The ball crossed a roadway . . . and began to move
along the trolleybus wires. . . . The wires were fixed to poles with lighting lamps.
Lamps in turn flashed as the ball moved under the lamps. The ball in this moved by
3–4 lamps” (Grg.145). There are evidences in favor of BL having a hard component
in its radiation: injured persons got skin burns typical for the burns under action of
ultraviolet radiation (Grg.148–152). Grigorjev presented cases of injury of persons
and animal by BL (Grg.153–167): on touching it answers “either by an electric
discharge or by an explosion”. The lifetime of BL explosion by Grigorjev’s
estimates does not exceed a millisecond, and hence, a shock wave cannot extraor-
dinarily harm a human. “Apparently a man loses consciousness only for a time of
shock wave propagation.” “Release of a large charge at ball lightning disappear-
ance” can cause more serious effects. There is a report about BL explosion near an
electrical supply distributional box in Kiev studio of feature films (Grg.168). “All
the lamps began to shine” at the moment of the explosion. Then it “turned to be that
all the incandescent lamps of different power were burned out, independently of the
fact were they switched on or no.”
Among Grigorjev’s observations, there are evidences that BL changes its size
during an observation (4%) or changes its form (6.5% of the cases). In one of them,
a transformation “of a burning needle of about ten centimeter length and about of a
half of a centimeter thickness into a ball of 5–6 cm in a diameter” after hitting a
door (Grg.176) is described. “Burning yellow-red color tape” in another case “has
been assembled into a football sized ball near a threshold of 1–2 cm height”
(Grg.177). In the third case: “A thin yellow-white line-thread has flown through a
glass. . . . It hit a brick and has transformed into a ball of about ten centimeters
diameter” (Grg.178). “A yellow-red tape of a meter length with a diameter of 3 cm
stopped having flown to a threshold, then it started to rotate on the spot winding into
a football sized ball” (Grg.180). A transformation of a ball into a tape is also
possible: the ball slowly during 5 s “has elongated into a long tape, and the whole
tape immediately went through a ventilator window to a street when one of its ends
has reached the ventilator window, then it leaped up” (Grg.184). BL can move over
the water, leaping back from its surface (Grg.193–198), but can descend into the
water. It was reported that “there was no explosion, a hissing has sounded, and
ripples began to move as from a thrown stone from the place where the ball has hid
in the water. For several seconds there was nothing, but then bubbles have appeared
in the place of the ball lightning fall and a water column of two-three meters height
has gushed”(Grg.199).
During an analysis of his data, Grigorjev tried to find correlation dependences
between different characteristics of BL, but similar to other investigators, he found
practically a complete absence of connection between BL diameter and its lumi-
nescence brightness. There were weak signs that large BL is more often red; and
red BL has a somewhat shorter life than the yellow ones. The only parameter that
one succeeded to connect is BL lifetime tL and its diameter D, which can be
described by a relation tL ¼ t0 exp(D0/D) (t0 ¼ 75 s, D0 ¼ 7 cm). One can see
that tL limits to t0 ¼ 75 s at D  D0, but more long-lived BL were observed in
nature.
230 V.L. Bychkov et al.

As investigator, Grigorjev was confident regarding the opinion that for the crea-
tion of BL theory, it is necessary to know all its features in detail. He thought that
even 6,000 descriptions collected by him were still “insufficient for serious investi-
gation of ball lightning.” New information is necessary about rare and poorly
investigated BL features; for example, how it propagates through glasses [41, P. 32].
In 1989, in “Pravda (Truth)” (Moscow) newspaper, a news story under the title:
“Visit of Fiery Lady” appeared. It was reported in it that a “curious case” has
occurred in Khabarovsk city. “There ball lightning has flown into a cauldron with
about of 7,000 L of water. The water has started to boil immediately. Lightning was
bathing in water for about of ten minutes, and then it went out” [43]. To heat 7 t of
water from 20 to 100 C, one needs an energy of 2.35  109 J. There was no
indication about BL size in the news story, and hence, one can only expect that it
was smaller than the sizes of the cauldron. If one considers that its form was close to
a spherical one, then its diameter would be 2.4 m. Let us assume that BL diameter
was by two times smaller, that is, 1.2 m. Subsequently, its volume would be 0.9 m3
and the energy density contained in it would be 2.6  109 J/m3. Based on this
observation, let us add one more feature to BL properties:
69. BL can release an energy of 2.3  109 J.
In 1993, Egely published BL observations data bank in Hungary [44]. He
published in four Hungarian newspapers articles about BL and asked the readers
who saw BL to send him a detailed description of their observations. Egely’s
method is comparably more informative, as it allows getting data about a physical
nature of a phenomenon without the loss of important details, which can disappear
at statistical or questionnaire analysis. Data collection was started in April 1986 and
was finished in September of the same year. At the end, Egely collected 300
BL-observation descriptions.
According to the received reports, “Hungarian” BL did not differ from BL
observed in other countries. In some of them, an internal structure (luminescent
grains) and a cover (sharp projections were sometimes observed on it) were noticed
(Eg. 54; 63; 64; 112; 125; 183) (here Eg. means Collection [42], and a numeral – a
serial number of an observation from this Collection). There were marked cases
when BL took the form of a long tape, which then coiled into a ring or a ball (Eg. 54;
57; 76; 177; 248), went out of sockets or keyholes, and then returned to them
(Eg. 12; 26b; 31). They often moved over wires, cables, or railroad rails (Eg. 32;
36; 67; 68; 69; 136). A case when the ball was accompanying a carriage moving
along a fence (seemingly metallic) and left behind it, when the fence ended (Eg.
56a), or “felt” a motion of a man in a room (Eg. 56c) was also reported. There are
cases when BL, for a long time, was flying nearby an automobile or a running man,
keeping up with him (Eg. 101; 223). The most striking case of such an accompany is
described in (Eg. 178a, b). A driver and doctors of an ambulance carrying a patient at
night saw a spherical luminescent object of about 4-m diameter. It was surrounded
by a halo, and three “flame tips” were gushing from it. This ball was flying nearby
the automobile moving with a velocity of 40 km/h during 15 min. After that, it
descended to the earth and disappeared. On the way back, they met a second BL.
6 Ball Lightning Investigations 231

It was of the same size as the first one, but already had 6–7 “flame tips.” Moving at a
height of 10–15 m, it moved parallel to the automobile with a velocity of 60 km/h.
The velocity of the ball decreased to 5–10 km/h when the car stopped. The
automobile again went with the velocity of 70 km/h and BL again flew besides it.
After this, it disappeared in the same place where the first BL disappeared. The
weather on that day was cloudy, but without rain and thunderstorm. Both BLs were
moving with constant velocities; however, their motion was visibly accelerated
when they moved near villages. The halo thickness around the ball was approximately
equal to its diameter and the “flame tips” or “tails” went out, and then flashed again.
Electric bells and electronic alarm clocks began to ring, and electric lamps start
to flash when BL appeared in premises (Eg. 146; 197; 247; 248). There are reports
about cases of injuries and death of people when BL touched them (Eg. 39–41; 56b;
100). On a victim’s body, burns in the form of lines branching as roots of trees were
found. However, in one case, BL exploded near a patient lying in a resuscitation
department with electrocardiograph suction caps on his chest, and it had completely
cured him from arrhythmia (Eg. 179). There are marked multiple cases of animal
deaths caused by BL (Eg. 55; 97). There is also a case when cows escaped from BL
when they put their front legs in a trough. However, they did not give milk for
several days after this (Eg. 166). There has been a report of BL flying into a room,
which then jumped from nail to nail, throwing down pictures hanging on them, and
then flew out after a silver spoon hit a window (Eg. 99b). In another case, it moved
along a railroad platform passing round the benches with steel feet (Eg. 102). One
witness noticed that a ball jumped up and went down at the moment when a linear
lightning struck (Eg. 13). In Egely’s opinion, BL jump had occurred owing to the
action of the external electromagnetic field.
There has been a report where a witness described in details the whole process
starting with the formation and ending with the death of BL (Eg. 69). First, he, for
3 s, saw weakly glowing ball of 4-m diameter with diffuse edges. Then, gradually
growing bright, “small islands” appeared on its surface (this lasted for about 3 s). In
the center of the ball, there was an area with a diameter of 0.8 m, free of “islands.”
Suddenly, a ball of 50-cm diameter flew out of this area with a high velocity in a
direction of a high-voltage line. The color of the ball was orange-red, and it had a
clear black contour to the left. The ball existed for 1 s. The ball’s diameter
decreased to 10 cm during the next moment, the luminosity of its surface sharply
increased, and its color became blue. The ball in this state remained for 5–8 s. On
approaching high-voltage line, the ball “floated” over the wires and exploded after
covering 10 m. It produced a dazzlingly shining white ball of 8-cm diameter. Red
jets were ejected from the ball surface, and their length was 30 cm. This lasted for
0.5–1.5 s, and everything disappeared.
BL sometimes appeared in groups (Eg. 184). There is a report that first in a
witness field of vision, BL of 3-cm diameter appeared flying over the earth at a
height of 2.5–3 m. Then, after sometime, two balls of the same size appeared
approximately at the same height, which were moving parallel to the earth at a
distance of 10 cm from each other in the same direction as of the first BL. Several
seconds after this, three small balls appeared at the same height, which were located
232 V.L. Bychkov et al.

in the corners of an isosceles triangle: two of them were at a distance of 15–20 cm

from each other and the third one was at a distance of 30 cm from them.
There are case descriptions when BL demonstrated its ability to carry heavy
objects. Thus, it can attract a fork (Eg. 53) or through from itself, a black stone with
a smooth surface of 10-cm diameter (Eg. 60).
There is a report that a red-brown spot of burned ground appeared after an
explosion of BL of 100-cm diameter (Eg. 220b). The witness discovered that a
magnet was attracted to this absolutely smooth section of earth. At the same time,
the magnet was not attracted to the earth around the spot that the BL did not touch.
The witness supposed that the red-brown color of earth could be explained by the
presence of iron in it.
Egely represented a description of an “immersion” case of a witness inside a
glowing sphere (this case resembled a similar event described by Singer [18, Case 12]).
A young man with a bicycle before a thunderstorm was inside a yellow cloud for
1–2 min, feeling himself being enclosed in a yellow glass sphere. After that, the
spherical cover was torn and disappeared without traces. Being inside the ball, the
young man did not see anything outside, and it was also invisible for an observer
outside.
In Egely’s data bank, there are several cases that allow estimating the energy
store of BL. He reported about melting of wires at BL explosion (Eg. 10; 173; 174),
and the formation of holes in glasses, walls, metallic plates, and stones (Eg. 4; 56b;
58; 71). There is a case (Eg. 272) about BL that made a hole of 30-cm diameter in an
adobe wall of a house, and went away from a room through a window without
damaging the glass. However, a metallic handle of a frame disappeared somewhere.
There is a report that a potato-shaped BL flew over a ditch and left a brown mass of
fluidized sand of 70-cm length, 10-cm width, and of several centimeters thickness
(the case resembled the one described in [34], where it was possible to estimate BL-
energy store). The volume of the fluidized mass of sand was Vpt ¼ 0.7  0.1 
(0.02–0.04) m ¼ (1.4–2.8)  103 m3. By assuming the density of the sand to be
rs ¼ 2  103 kg/m3, heat capacity Cs ¼ 0.84 kJ/kg K, and a melting temperature of
the silica gel T ¼ 1,570 K, one can find that the energy that was spent for melting
was W ¼ CsrsVptDT ¼ 3–6 MJ. Note that BL did not perish from the indicated
impact, but continued its existence (as in the case [34]).
We would finish Egely’s book description with the case (Eg. 222) of its energy
measurement, with a help of “water calorimeter,” which is very much similar to the
case of BL falling into a tub [13] or trough with water [32]. On a hot July day, a
yellow BL of 25-cm diameter rolled over a glass light of a workshop’s roof and then
with a deafening crash, fell down into a pit near the corner of a building. This was
observed by a multiple group of workers who went into a yard during lunch hour.
The sizes of the pit were 0.6  0.6  0.5 m, and 2/3 of it was filled with water. All
120 L of water evaporated; the vapor from the pit was appearing for half an hour
after the ball fell. An examination of the event place showed that water was
evaporated, but was not spilled out. Burnt leaves and grass stems were lying around
the hole. Egely estimated the energy of this BL as 2.8  108 J, and its energy density
as 3.4  1010 J/m3.
6 Ball Lightning Investigations 233

Summing up, from the data analysis collected and analyzed by Egely and
Grigorjev, one can say that their method of complete events description texts
representation turned out to be very useful.
Along with the observations of BL “standard” properties that can be statistically
analyzed, rare cases gradually came out (that allow them not to be neglected at
statistical analysis), which were also considered to demonstrate important BL
features: its large energy store, the respond to variations of external electric field,
its internal charge, capability to transmit energy “at distance,” and features of its
motion. Their rare events were summed up and allowed to formulate new features
of BL:
70. BL has a cover inside which one can observe a motion of smaller parts.
71. BL can change its form: it can transform its form from a tape to a ball and from
a ball to a tape.
72. BL “feels” variations of external electric field; it is sensitive to motions of the
objects (people and automobiles).
73. One can observe an interaction of separate elements in a group of BL: two BLs
move at a definite distance from each other, and three BLs create an isosceles
triangle.
74. BL can carry a heavy object, for example, a stone or a steel fork.
75. At explosion, BL can reduce iron oxides to metal.
76. BL can transmit energy at a distance, most probably in a form of electromag-
netic radiation. This causes electric devices (bells and lamps) reaction or can
lead to melting of wires, sand, or glasses.
77. BL is capable of making holes or deepenings in walls, metallic plates, and
glasses.
78. BL can penetrate a glass and leave it unharmed.
Unfortunately, we are not aware of any investigations of glasses through which
BL has penetrated, and hence, we do not know if BL could change their properties.
79. By getting into a tank with water, BL can heat water to boiling point
and evaporate it. BL-energy density estimated in this case equals to 3.4 
1010 J/m3.
In 1994, in Moscow, a Collection of works, titled, “Ball Lightning in a Labora-
tory” [45] was published. In the introduction, signed by all editors of the Collection,
there are words that “the most important feature of natural ball lightning is
its capability to carry high uncompensated electric charge as large as some
Coulombs.” Stating this, the authors have addressed Stakhanov’s works or talks,
which, in our view, has been misunderstood. We have mentioned earlier that
Stakhanov gave quite a decorative role to BL charge, and considered that a reason
for strong electric effects was some hypothetical BL capability to transport charges
accumulated in the conductors. It is interesting that at that time, there was no
uncompensated charge in the BL models developed by the editors and authors of
the collection.
234 V.L. Bychkov et al.

Some cases in [45] confirmed that BL can have a substantial uncompensated

electric charge. During the Second World War, a girl who was on a wooden
watchtower of 20-m height was killed by BL [45, P. 222]. There was “a halo of
white fluff” around her body, which “turned out to be of underwear and soldier’s
puttees comminuted in the fluff. At that a soldier’s blouse and a skirt were safe, and
there were no traces of burns on the body. Only a round burn was noticed on a place
where there was a ring before, but the ring itself has disappeared.” There is also a
description of another case when BL came at 1 m from a man and a woman sitting
on the hay [45, P. 222]. The woman’s hair “stood on end (like a fan), and hairs of the
man began to stir under a service cap.” When they came back to their senses after
BL went away, the observers discovered that “rings have disappeared from
women’s fingers,” and from the man, “the penknife chain, a metal tip from a
mechanical pen and metal rims from the boots, fixing edges of holes for laces
have disappeared.”
On the basis of these reports, one can consider that
80. BL can possess a substantial uncompensated electric charge.
81. BL can be a source of powerful electromagnetic radiation causing resonance of
currents in the ring conductors with diameters from 1 to 2 cm.
The case that occurred on July 10, 2006 in J€ urmala (Latvia) can be a confirma-
tion of this conclusion. When the resident of Yelgava city, Oleg Andrejev, went
from the sea to the coast, “the yellow-orange bright ball about a tennis ball in size
has flown” nearby him and “the strong roar has appeared. Oleg has fallen face
downwards the sand. He did not lose consciousness, but he was very sluggish with a
disseminated consciousness. When Oleg was turned, his face was in a smoke black
(salts of gold?). People bathing at this time in a gulf have felt an appreciable current
blow . . . The golden chain on Oleg’s neck has evaporated (see Fig. 6.2), having left
an appreciable burn round his neck in the form of a dotted line. On his stomach and
on his heels also there was a trace of a lightning (the current channel)” [46].
Apparently, the BL has blown up at the level of Oleg’s waist and the current of

Fig. 6.2 Still photograph of a

man’s neck hit by ball lightning
(BL) [46]
6 Ball Lightning Investigations 235

its discharge has gone to earth through the lower part of the man’s body, not having
touched the heart area. At the moment of explosion, there was a generation of high-
frequency radiation pulse with a basic frequency, resonant with the absorption
frequency of golden chain links (with the size of 3–4 mm). Having accepted a
chain weight of m ¼ 17 g, we can find that the energy Q1 ¼ c1m (Tm  T0) that has
been spent for its heating to a fusion temperature Tm ¼ 1,063 C was Q1 ¼ 2.46 kJ.
Energy Q2 ¼ c2m (Tb  Tm) ¼ 4.96 kJ has been spent on the heating of a melt to
boiling temperature Tb ¼ 2,887 C. And, finally, the energy spent on chain evapo-
ration was Q3 ¼ c3m ¼ 28.56 kJ (here, T0 ¼ 30 C is air temperature, c1 is the
specific thermal capacity of solid gold, c1 ¼ 140 kJ/(kg·K), c2 is the specific heat
capacity of liquid gold, c2 ¼ 160 kJ/(kg·K), and c3 is the specific heat of gold
evaporation, c3 ¼ 1,680 kJ/(kg·K) [47]). Thus, the total energy spent for the chain
evaporation is Q ¼ Q1 + Q2 + Q3 ¼ 36 kJ. If the explosion took place during 102 s,
then a power of BL radiowave radiation would be about 3.6 MW.
There is an interesting description in [45, P. 223] about BL movement over a
narrow-gauge railway rails at a height of 3 m. When it came nearer to the observers,
they “were shined with a narrow bright beam, outside of which it was dark, and
behind them the beam sharply broke that does not happen at the illumination by
searchlights or headlights.” Therefore, it is possible to consider that in rare cases
82. BL visible radiation can have a form of a narrow beam.
In 1999, Mark Stenhoff’s monograph, “Ball Lightning” [48] was published.
Unlike the books about BL written by the experts having their own ideas on how
it is arranged, and consequently and subconsciously choosing the “most authentic”
among the facts, the book by Stenhoff was written by the impartial researcher.
However, another extreme is peculiar to Stenhoff’s approach: its excessive skepti-
cism is shown in aspiration to reduce many power effects caused by BL on the action
of usual linear lightning discharge. The temptation of it is great, as the energy of an
ordinary linear lightning (2  108 J) greatly exceeds the energy of an “English” BL
(107 J), known to Stenhoff [13, 14]. However, here, he again questioned the
reliability of the classic example of energy estimate. Hence, by reproducing the
calculations of a sphere energy content made by Goodlet [48, P. 194], he considered
only the lower estimation of energy (3.1 MJ) as correct, when water heated up only
to 60 C. With the assumption that water was heated up to 100 C and its part had
boiled away, he speculated that “it was not in event descriptions of Mr. Morris and in
additional inquiries of Sir Charles Boys.” He was more inclined to consider that
water heating could be caused by a stroke of a linear lightning, and as a proof,
presented an opinion of V.M. Uman who suggested that “lightning caused the arc
from nearby electric power lines.” Such an attitude to the event description appears
strange enough, as Mr. Morris “in the clear” stated that “the water boiled for some
minutes,” and near a tub with water, there was no power electric line, but “telephone
wires” [48, P. 7]. According to Stenhoff, the reason of observable destructions
comprised “close association between CG (cloud – ground) lightning and a ball
lightning. Ball often precedes a CG lightning flash to the immediate vicinity, which
coincides with the apparent decay of the ball” [48, P. 196] (note that, here, Stenhoff
236 V.L. Bychkov et al.

departs from the principle of impartiality as among the observations collected by

him, and in the observations collected by other researchers, the specified time
correlation cannot be found). Stenhoff also cautiously considered the assumption
that BL can bear in itself a considerable electric charge. For “it seems quite
implausible that ball lightning could contain within its small volume sufficient
charge per unit volume or sufficient electrical potential energy to deliver such
large currents. It is much more credible that an ordinary CG flash was responsible.”
Thus, in the approach regarding BL-energy store and its charge, Stenhoff appeared
to stick to a myth that
M5. BL is some sort of a “trigger” switching on a linear lightning. “. . . ball
lightning may not in itself represent a hazard, . . . it may be a precursor of an
impending lightning flash” [48, P. 196].
Basically, Stenhoff “has walked over already plough up field” of data about BL
observations, and his book practically does not contain new data that could add to
the full list of BL properties made by us. An exception is the analysis of BL behavior
in the planes, undertaken by Stenhoff. There is a story as once “a dim yellowish light
appeared ahead of the aircraft’s nose . . . This light expanded in circumference . . .
When the light was close to the windscreen, it was possible to hear a swishing noise.
At the same time the light seemed to be rotating rapidly . . . There was a loud report
followed by a strong smell of ozone . . . But it was accompanied by . . . a blinding
flash . . . They saw a ball of fire come out of the flight deck and roll down the aisle. It
was about the sizes of a football . . . It continued to the rear of the aircraft and
disappeared . . . On arrival . . . we examined the airframe . . . and found a burn mark
on the radome about the size of a half-crown” [48, P. 112–113].
In another case, “the aircraft was evidently struck by lightning (observers saw a
bright flash of light and heard a loud bang) and some seconds later, a perfectly sym-
metrical glowing sphere of diameter 22  2 cm emerged from the pilot’s cabin and
traveled at constant height and speed (75 cm above the floor at 1.5  0.5 m/s relative
to the aircraft) and in an undeviating path down the central aisle of the aircraft
. . . The blue-white sphere had no structure, and was somewhat limb-darkened
and optically, hence appearing almost solid. It did not seem to radiate heat, and
appeared to have an optical power of about 5 to 10 W . . . It disappeared into
the toilet compartment at the rear of the aircraft” [48, P. 114].
“A DC6 aircraft was flying north over Alps in heavy turbulence when a luminous
ball with fuzzy edges, about 10 in. (25 cm) in diameter, was seen traveling rapidly
along the floor on the central aisle. There was no integral damage, but on landing
blackness holes were found in the nose and tail of the aircraft, 3 to 4 in. and 10 to 12
in. (8–10 and 25–30 cm) in diameter, respectively” [48, P. 115].
During the flight of Ilyushin-18 aircraft “at an altitude of 1,200 m, a ball of fire
10 cm in diameter appeared on the fuselage in front of the cockpit. It disappeared
with a deafening noise, but re-emerged several seconds later in the passenger’s
lounge, after piercing in an uncanny way through the air-tight metal wall. The
fireball slowly flew above the heads of the stunned passengers. In the tail section of
6 Ball Lightning Investigations 237

the airliner it divided into two glowing crescents which then joined together again
and left the plane almost noiselessly . . . The radar and several other instruments
were damaged, so the aircraft returned to the airport. Two holes were found in the
fuselage – one in the nose section and one in the tail plane. There were no traces
inside the aircraft, and none of the passengers was hurt” [48, P. 115].
Thus, one can assert that
83. BL can get in the all-metal plane through cabin glass, aerial fairings, or through
apertures in a fuselage. Having got into the plane salon, it, as a rule, slowly
moves between the rows of seats in a direction from a plane nose to a tail. It can
leave a plane tail silently or with an explosion, leaving a hole.
Besides BL considered by us, objects that have obtained the name of “the
Unknown Flying Objects” (UFO) are often observed in the nature. They, along
with BL, represent shining spheres, ellipsoids, disks, or cylinders, moving in air
sometimes smoothly and sometimes very quickly. They can throw out sparks and
substance streams from themselves, and can leave circles of the burnt earth.
Therefore, in our opinion, the distinct border that could be placed between BL
and UFO does not exist. If the distinction of these objects is made, it is possible to
say that it does not have a basic, but “ideological” character. The BL’s nature is not
yet well understood, and we fairly declare it, whereas with regard to UFO, there
appears a steady myth that they are the ships of newcomers from other planets.
The complete report about UFO observations can be found in Chernobrov’s
book [49]. After reading this book, it is difficult to get off the thought that in an
overwhelming number of cases reported, there is a description about BL that is
familiar to us. The specificity of UFO observation is revealed only in the fact that all
messages concern objects out of premises, and great attention is focused on the
emotional condition of the observer – to a stream of his/her imagination caused by
the influence of a shining object. As it was mentioned earlier, an encounter with BL
for any strange reason “is imprinted” forever in the observer’s memory; that you
will not tell about other, not smaller impressions of dramatic events of a life.
Another feature of UFO observation reports comprised the fact that sometimes
objects of large sizes (up to 100 m), existing for hundreds of minutes, have been
reported. However, this can hardly serve as a criterion for the differentiation of BL
and UFO. If one looks at the dependence graphs of BL observations probability of
the size and its lifespan (e.g., see [48, P. 14; 15]), it is possible to notice that at the
growth of the size and time, they come near an abscises axis smoothly, thereby
supposing the possibility of BL existence with large sizes (>10 m) and large life
span (>10 min). Hence, Stenhoff, without any doubts, considered BL as a sphere of
6 m in diameter [48, P. 173]. On the other hand, Chernobrov named all shining
yellow-red spheres from 30 cm to 4 m in diameter as UFO [49, P. 26; 33; 38; 85; 89;
92; 95; 126; 252; 391; 414]. Thus, having added the data from the data banks of
UFO to BL data banks, and supposing that they describe objects of similar nature,
we can make those BL properties more authentic, which were considered earlier as
rare. Thus, it is possible to assert that in an upper atmosphere, the existence of
238 V.L. Bychkov et al.

spherical, disk, and cylindrical form of BL with sizes up to 100 m is probable. These
objects may exist for several hours.
The most interesting feature of these objects (of both small and large sizes) is
that they can emit narrow beams. Sometimes, it is simply an emission of “flame
tips,” and sometimes, narrow rays of light similar to the laser beams are emitted [49,
P. 20; 44; 83; 88; 97; 100; 102; 109; 113; 114–117; 122; 152; 187; 404]. It must be
noted that among the above-mentioned BL observations, there was only one case
when light emitted by it looked like a projector beam [45, P. 218]. In one of the
descriptions [49, P. 83], a light going from a sphere, which decayed to ring zones of
violet, yellow, and pale yellow color, was described. This report wonderfully
agrees with the observation description of shining object from a plane board in an
article that appeared in 1985 in the Soviet newspaper “Trud (Labor)” [50]. In this
article, a long-time observation of an object of 20 m in size at a height of 10,000 m
was described. It emitted narrow light beams. When the beam slipped on a plane
cabin, the pilots saw “the dazzling white point surrounded with concentric color
circles.”
There is a report about an orange-red disk of 20 m in diameter, which dived on
the parking of tourists [49, P. 42]. The fire occurred because of it, the tents were
broken, the frozen water became warm, and all clocks stopped. This, as we
remember, is very similar to BL action which can heat up water and even bake a
soil section. UFO can move with a plane speed [49, P. 88; 113] and make sharp
turns, by not reducing the speed [49, P. 153]. We found similar properties in the
descriptions of BL behavior [48, P. 110; 111]. The UFO as well as BL generates
noise in radio receivers: when the object flew by a car at a height of 100 m over it,
“the receiver as if has blown up” produced a strong whistle [49, P. 164].
If one recognizes an identity (or likeness) of UFO and BL, then it is possible to
consider a typical specificity of pairs, and especially, three fireballs’ interaction. As
discussed earlier, in one of Egely’s reports, the property of BL groups to create a
structure in the form of an isosceles triangle, was described (Eg. 184). According to
the book by Chernobrov, such a tendency to self-organize is one of the important
properties of UFO: observers everywhere saw “three-star structures” in the sky [49,
P. 23; 39; 43; 200; 238; 246; 248; 269; 297; 382]. It is marked that close location of
UFO leads to “a paralyzing action” on a person; it causes spasms, catalepsy,
hallucinations, and loss of consciousness [49, P. 21; 44; 111; 290]. It is quite
probably that these objects (including BL) impact not only the basic senses of the
person (sight, hearing and sense of smell), but also the brain.
So far, we have considered the information on BL properties, “concentrated” on
monographs and reviews, covering an observation material from the beginning of a
“history” till the end of the first decade after the International Symposia on Ball
Lightning. Now, we will shortly analyze what has changed in this area during the
last 10 years. During this period, results of the researches devoted to BL were
published in journal articles as well as in the proceedings of the International
Symposia on Ball Lightning and the Russian Conferences on Ball Lightning.
At the Fifth International Symposium on Ball Lightning in Japan, from August
26 to 29, 1997, statistical data about BL properties observed in Japan [51] have been
6 Ball Lightning Investigations 239

presented. Again, as in the report presented 10 years ago, 89% of BL in Japan was
observed in fair weather, and other characteristics of the Japanese BL did not differ
from the worldwide.
In the report by Keul [52], a collision case of a BL of 50 cm in diameter with a
car was reported. At the moment of the collision, the driver – the school teacher
(female) – felt as if the car has run into an obstacle. Everything that was to the right
of her became bright green – the cowl, windshield, instrument panel, and even
upholstery. Having flown around the car from above, the sphere had landed on a
highway. Thus, the Faraday cage – the metal car – could protect the driver.
Stakhanova (Russia) [53] presented some BL photos from the archive of I.P.
Stakhanov at the Symposium. The pictures were taken in September 1985 near
Undugun lake 100 km from the city of Chita. Amirov and Bychkov (Russia) [54]
reported about the results of the statistical analysis of SKB data bank: 1,503 cases of
the observations collected by Stakhanov, 150 supervision collected by Keul, and
100 events collected by Bychkov. The increase in BL lifetime with the growth of its
size was revealed, and the most probable diameter of BL was defined as 10–20 cm.
On August 23–25, 1999, in Antwerp (Belgium), the Sixth International Sympo-
sium on Ball Lightning took place.
In the report by Stakhanova [55], a connection between BL observable char-
acteristics and weather conditions has been described. An analysis of 1,086 cases of
observations where the weather condition was specified, demonstrated that BL
characteristics, such as form, diameter, brightness, and color, appeared to be the
same as in the cases where information on weather conditions was not provided.
However, BL of larger sizes, greater brightness, and longer lifetime were more
often observed in fine weather, which can be connected with improvement in the
observation conditions.
The Japanese researchers, H. Ofuruton and M. Kamogava, informed about the
continuation of BL researches in Japan [56]. Two years after the fifth symposium,
they had collected 500 additional observations. Their analysis demonstrated that
unlike Europe where the overwhelming number of BLs is observed in summer, in
Japan, there are more chances to observe them during other seasons. With regard to
the color of the BL, they assumed a possibility of radiation of infrared and ultra-
violet light by it. For BLs of the small size (<30 cm), two peaks of average lifetime
were found: 10 and 60 s. The first peak can be explained by the fact that people do
not always see BL from the moment of its formation. Therefore, the actual lifetime
of BL may be close to 60 s.
Keul (Austria) [57] reported about the attempt of BL detection by means of
linear lightning detecting system. In Austria, since 1992, ALDIS – Austrian Light-
ning Detection and Information System of eight field stations has been operating,
with the help of which, the place of lightning stroke is determined with an error of
less than 1 km, and polarity of discharge and amplitude of a current are defined. The
system can distinguish intercloud discharges from discharges between clouds and
earth. Keul chose nine cases of BL observations after May 1992 from his data bank,
and compared them with the results of the linear lightning strokes registration
obtained by means of the ALDIS system. Coincidence between the data when the
240 V.L. Bychkov et al.

linear lightning has struck at the same time and in the same area and those of
observed BL has occurred only in two cases.
Bychkov [58] stated that for all the cases of BL observations, the increase in
lifetime with the diameter was evident. Definition of this dependence is important
for the development of BL theory. Hence, the lifetime in plasma models should
not depend on diameter, and in theories based on vortical structures, the lifetime
should be proportional to the square of the diameter. Bychkov found that BL’s life
span increases approximately linearly with its diameter in the case of small-sized
(<10 cm) BL. In [59], Amirov and Bychkov analyzed the interaction of luminous
objects with airplanes. Having compared the statistics of intercloud discharges and
BL occurrence, they came to a conclusion that only in 18 cases of the 73 observa-
tions known to them (from and in the planes) with BL sizes smaller than 5 m, BL’s
occurrence was evident. In one case of plane crash, BL’s impact could be observed.
It occurred in February when the plane flew over a powerful radio-transmitting
station. The BL has appeared on a control panel, and then struck the pilot’s chest.
He was injured and catapulted, and traces typical for electric current blow were
found on his body.
Toselli and Fedele [60] reported about the project of the Ball Lightning Italian
Observations Database (BLITA). The beginning of this database took place with
the analysis of 12,000 UFO observations collected by the Italian Center for UFO
Studies. From this database, 40 cases similar to BL observations were selected.
In July 2001, in the St. Louis (USA), the Seventh International Symposium on
Ball Lightning took place. Unfortunately, the proceedings of this symposium were
not published. Three reports devoted to BL observations and analysis were pre-
sented. In the report by I.G. Stakhanova, read by D. Turner, the data about the
electric effects made by BL was presented. P. Handel read the report of A.G. Keul
and O. Stummer about the gathering of the new BL observations in Austria. The
statistics of the results of BL observations appeared invariable on time and space of
all Central Europe. According to Bychkov, the old question of “Whether ball
lightning exists?” needs to be replaced with the new one “Whether highly energetic
ball lightning exists?” As a positive answer to the second question, he presented
examples of 17 observations known to him, comprising reports on BL with huge
energy content. Moreover, he asserted that the analysis of descriptions leads to a
conclusion that BL’s surface is hot in 25–30% of cases [33].
Subsequently, the Eighth International Symposium on Ball Lightning took place
on August 3–6, 2004, in Chung-li, Taiwan, and four reports were devoted to the BL
observation analysis. The first report [61] described how the sparks fell down to the
earth from a support of a high-voltage line after a lightning stroke. One of them
changed the flight direction horizontally at 1 m from the earth. Its diameter
was equal to 2 cm, and it was lightly yellow and fluffy. The length of the “hairs”
was 2–10 cm and on the tip of each of them, small droplets were visible. In another
case, the witness (a female) within 1 s could observe a shining sphere with a boiling
surface. “It consisted of randomly located ‘hillocks’ with strongly pronounced
peaks and deepening between them.” The streams were ejected in all directions
from the sphere. The diameter of the sphere was 35 cm and the distance between the
6 Ball Lightning Investigations 241

peaks was 3 cm. The streams had the form of cylinders of 5 mm in diameter and
15–20 cm in length These hillocks were similar to the fountain streams in a
gravitational field: the long ones, having reached the maximum height, “abruptly
changed a direction and in the form of drops fell downwards, on a sphere surface.”
Earlier, we had presented Egely’s conclusion of the presence of a shell over BL.
However, the case analysis required further investigation [62], and it was found that
the BL that blow on the earth could break up to pieces of a shell – a form that is
small and could melt (going out) – which are similar to pomegranate grains. Thus,
the destruction is viscous as rolled clod of snow. The observation [63] also affirmed
that the BL has a cellular, granular structure. Along with the observations of
streams beating from a surface of a BL and falling back to the form of drops
[33, 61], such BL descriptions lead to the conclusion that
84. BL cover represents a complicated system comprising solid internal layer and
liquid outside layer.
There is a report about the explosion of BL (BL hit it) nearby an electric lamp.
The lamp glass appeared with holes of small apertures of 0.1–0.2 mm in diameter.
At explosion of BL near to an iron oven in a tent of a Nenets family the aluminum
table ware standing about the stove has fused and has turned to a metal ingot, and
grains of small shot in cartridges was alloyed into a uniform piece of lead. Other
cases reported about how BL went out of radio transmitter plugs, the electric socket,
a phone tube, or power cable; in the latter case, during the strong thunderstorm.
Arora (India) [64] reported about BL mass occurrence case in India in the dry
long summer of 2002 when the day temperature of air was +42 C and the night
temperature was +28 C. There was no thunderstorm activity during that time, and
BL could appear because of the currents flowing in the earth generated by power
stations. In the report by Keul [65], the analysis of shining object video-recording
was reported. The shining object came out of the next house roof, had risen
upwards, and through 2.24 s, disappeared into the clouds. Its color was yellow-
orange, and at flashes, the object became white. Kavano (Japan) [66], in a substan-
tiation of his version, stated that BL is an undeveloped channel of a linear lightning,
had demonstrated a drawing of BL with a tail on an ancient Japanese engraving.
The Ninth International Symposium on Ball Lightning took place on August
16–18, 2006, in Eindhoven, the Netherlands. In the report “About observation
properties of a ball lightning” [67], Bychkov brought the result of 10 years’ work
on the statistical analysis of BL properties. The definition of the average characteri-
stics of this phenomenon was the result of this work. It was found that the diameter,
life span, and radiating ability of BL depend on where it is observed – indoors or in
the open air. The distribution of colors in the BLs does not depend on their site. The
statistical analysis demonstrated that the increase in the humidity of air and the
intensity of the atmospheric electric field led to a reduction of BL’s life span. The
estimated levels of BL-energy density based on observations are in the range from
102 to 3  106 MJ/m3. According to Bychkov, cases of occurrence of burns on the
bodies of people injured by a BL show that “the part of BLs possesses a very hot
surface.”
242 V.L. Bychkov et al.

Fig. 6.3 Still photograph taken

during a thunderstorm in1960
by I. Shagin [48]

Keul [68] presented the results of BL properties that were compared with the two
German data banks: Hentschel Data Bank (200 events) and Central European Data
Bank Germany (CEBLD) (132 events) with data from CEBLD Austria (150
messages). The key parameters on which comparison was made were: time of
observation, weather condition, observation time duration, distance to the object,
its diameter, surface kind, color, brightness, features of movement, and event
details. A comparison of data demonstrated that the properties of BL are common
across Central Europe.
In the report by Nikitin, Velichko, Vnukov, and Nikitina [69], an analysis of the
photo taken in 1960 in Sochi city in the USSR (see Fig. 6.3) was carried out. One
can observe the trace of a bright object, which fell down, with short-term stops. In
the comment to the first publication of a picture in the USSR magazine “Ogonyok
(Small light),” this trace was identified with BL trace, and image doubling was
explained using light reflection from the lenses of a camera objective. However,
Stenhoff was skeptical about this photo [48, P. 137] because of the plurality of the
traces. Modeling of the photographing process of bright light sources demonstrated
that the cause of double images is the reflection of light from lenses, and not the
camera movement at shooting. It has been defined that the BL descended with a
constant speed of about 25 m/s, and its stops were caused by the action of linear
lightning discharging an oscillating electric field. Thus, it was concluded that the
6 Ball Lightning Investigations 243

BL had a positive electric charge of nearly 104 C, a weight of nearly 1 g, and a

diameter of nearly 3 cm.
In the illustrations in the report by Papaelias [70], in which he considered that BL
is an antimatter display, he presented an interesting photo, which can be interpreted
as a trace of BL leaving a bend of linear lightning channel. Tar [71] told about his
observation how BL was formed after a linear lightning struck the earth. Within 2 s
after the blow, at the distance of 2.5 m from the blow place at a height of 1.2 m over
the earth, there was a bright sphere of 30–40 cm in diameter, rotating counter-
clockwise round a horizontal axis. To the right of the blow place at the same
distance (2.5 m), there was a bush whose branches were bent under the pressure
of the wind which started from the place of a lightning stroke. The fireball had one
or two reddish tails. After the third of the second tail was drawn into the sphere, it
continued its slow movement to the left. In 3 s, the sphere instantly disappeared.
Based on his observation, the author proposed a model of BL formation: (1) After a
lightning struck, the vortical toroidal ring was formed. It was cold and consequently
remained invisible. (2) Having come across a bush, a ring was broken off, and the
figure resembling a sickle of moon was formed. (3) The broken-off ring was pulled
together in the sphere. (4) Because of strong electrification, the gas molecules were
excited and the sphere became visible.
On 7–12 July 2008, in Zelenogradsk, the Kaliningrad area of Russia, the Tenth
International Symposium on Ball Lightning took place. New reports on BL obser-
vation were discussed during this event.
In the report by [72] Bychkov and Nikitin, the likeness of BL and UFO
phenomena was underlined and it was concluded that the part of these objects
can represent the result of geological and geophysical processes, during the shift of
the Earth and rock layers.
Keul [73] compared the results collected from six BL European national data
banks: 200 observations of Hentschel in Germany, 132 observations of Keul in
Austria, and 249 reports from France, collected by Hubert. From two Italian
databases (Toselli, Carbognani), 201 reports were collected. Egely provided
descriptions of 281 BL-observation cases in Hungary. Two Russian data banks
contained 3,104 documents. Europe is an ideal place for doing research on light-
ning, including BL, unlike America with its “hot” storm areas like Florida and
Paraguay, where moderate storm activity with an annual frequency of lightning
strokes from 8–10 km2 (Italy, Spain) to 1–2 km2 (Southern Scandinavia) could
be observed. A comparison of the European cards of storm activity and the
“density” of BL observations showed the absence of correlation among these
parameters. There is no wonder that lightning is recorded immediately by means
of tools, and that data on BL observations pass through some social filters such as
population density type, association between public and thunderstorms observation,
educational level, and critical attitude to reports. The observation period for all
banks was 1900–2000, and a maximum number of observations were recorded
during the 1950s and 1960s. Furthermore, the maximum number of BL observa-
tions were reported during summer months. The more observations occurred during
the afternoon hours (from 12 to 18 h); in Italy, the peak of observation shifted to
244 V.L. Bychkov et al.

18–21 h. In 80% of the cases, BL observation was connected with thunderstorms. In

the rest of the data banks, one can trace the usual characteristics of BL observed in
former researches. A comparison of the European databases with those of the
American and Japanese ones was carried out. A distinction was observed only
with regard to Japanese data, where 2.6% of the observation cases were connected
only with a thunderstorm; however, BL-observation frequency distribution during
the months of the year (a maximum of 26% in August, and 16% in July) in Japan
was similar to those observed all over the world.
VanDevender (USA) [74], in his report, described the results of field researches
on the event that occurred on August 6, 1868 in County Donegal, Ireland [75]. The
witness, Michael Fitzgerald, during a walk, noticed a red fiery sphere of about
60 cm in diameter, slowly moving in air. “After passing the crown of the ridge . . . it
descended gradually into the valley, keeping all the way about the same distance
from the surface of the land, until it reached the stream” (coast) in 300 yards (90 m)
from a place where an observer was standing. “It then struck the land and reap-
peared in about a minute, drifted along the surface for about 200 yards (60 m) and
again disappeared into the boggy soil, reappearing about 20 perches (100 m) further
down the stream; again it moved along the surface, and again sunk, this time into
the brow of the stream, which it flew across and finally lodged in the opposite
brow, leaving a hole in the peat bank, where it buried itself . . . I . . . found a hole
about 20 feet square (6 m), where it first touched the land, with the only pure peat
turned out on the lea as if it had been cut out with a huge knife. This was one minute
work . . . It next made a drain about 20 perches (100 m) in length and 4 feet (1.2 m)
deep, afterwards ploughing up the surface about 1 foot (30 cm) deep, and again
tearing away the bank of the stream about 5 perches (25 m) in length and 5 feet
(1.5 m) deep, and then hurling the immense mass into the bed of the stream, it flew
into the opposite peaty brink. From its first appearance till it buried itself could not
have been more than 20 minutes . . . It appeared at first to be a bright red globular
ball of fire, about 2 feet in diameter (60 cm), but its bulk became rapidly less,
particularly after each dip in the soil, so that it appeared not more than 3 inches
(8 cm) in diameter when it finally disappeared.” A brigade of researchers found out
that even 137 years after the incident on earth, traces of the ditches made by
BL could still be observed. In [74], the authors estimated the work spent by BL
as 4  107 J and considered that the ditch has been pressed through the earth by a
microscopic black hole with a weight from 2  104 to 106 kg, which “magnetically
levitated by a large electromagnetic field.”
In another report [76], studies on glass-disk characteristics were presented. The
glass disk was of 8 cm in diameter, cut out of a window by a BL. The disk had a
cone edge with expansion in the direction of the room where it dropped out of a hole
in a windowpane. The analysis demonstrated that the hole formation most likely
occurred because of local heating of glass, and glass heating occurred from the
room side that is from the side opposite to those to which the BL touched. It
specified that BL energy passed in any form through the glass. If it was electro-
magnetic radiation, then its frequency should be no more than 1010 Hz [77]. The
disk from the inside of the glass had a sharp edge that torn off from the edge of an
6 Ball Lightning Investigations 245

aperture; its thickness was about 50 mm. By means of X-ray spectrometer, it was
revealed that in the direction from the center of the disk to the edge, the element
structure of glass on an edge surface varied: its silicon content (from 26% to 18%)
decreased and the oxygen content (from 59% to 69%) increased.
Apart from the above-stated BL-observation reports that have entered into the
books and in the proceedings of the symposia, articles reporting new BL features
continue to appear in the current literature. Earlier, we mentioned about cases when
the BL had lifted air magnets weighing about 1 kg [32, P. 72] or stones weighing
about 1.3 kg (Eg. 60). Some researchers doubt the validity of these observations as
these actions by light BL violate the law of momentum conservation. However, if
we consider that the BL has a charge and acts as a hook on a rope of the elevating
crane, “it is suspended on a power electric line” to a cloud, then this contradiction
disappears. In [78], even more surprising BL properties are described – its ability to
lift objects weighing up to 100 kg. It is reported that “flying by over a tractor station,
BL has crushed a framework welded of metal rods, and other framework . . . at first
it has dragged over the ground. When BL has overtaken it, the framework has been
lifted, carried for 300 m and softly descended. The framework weight was not less
than 100 kg.” In the same book, it is reported that “flying by over an apartment
house covered with slate, BL has torn off nails together with the slate from a house
roof (from a surface of some hundreds square meters order) and, having raised all
this slate in air, has entailed it behind itself, scattering over all village.” Further-
more, photos of the carried steel framework and the house with the damaged roof
have also been presented.
We have already discussed the case of BL birth by means of “technical mean” –
trolley-bus bars (Grg.145). It appears that this is not the only case. A witness of the
same event in 1997 was Professor R.N. Kuzmin [33]. On a road hollow, the trolley-
bus bar came off the wires. “The spark discharge between a boot of a bar of a trolley
bus and current carrying wire has generated a white ball which has come off the
birthplace, and has grown up a little.” Subsequently, the sphere of blue color with a
yellowish shade, 10–12 cm in diameter, smoothly passed through a handrail of the
bridge and “has blown up practically without noise in one and a half or two meters
from an earth surface, scattering a sheaf of sparks. Before the explosion BL has
decreased in size, has reddened, and there was a blackout inside it.”

Prospects of Future Work on Collection of Ball Lightning Observation Data

Now, let us draw a result. In the present review, we have tried to make a represen-
tation of the ways of accumulation and processing of data on properties of an
unusual short-term natural phenomenon – BL. It seems that the hope that the
statistical processing of data files can help to get into its physical nature was not
successful. It did not reveal any, except trivial, correlations between its key para-
meters: size, life span, brightness, and color. This may be owing to the fact that they
really do not depend on each other. Work on data gathering and their averaging, in
our opinion, “came to saturation” and does not give practically any new results
246 V.L. Bychkov et al.

useful for the construction of BL theory. In this connection, the change in the
“paradigm” is more reasonable: a transition from operating by the processed data to
the analysis of detailed descriptions of real events. The supporters of such approach
are G. Egely and A.I. Grigorjev. Certainly, it will require granting open access to a
full set of data on the occurred events of observations to the researchers. However,
whether it will be possible to achieve this during our period of the Internet and
permanent world crisis remains an open question. Nevertheless, it is possible to
hope that, as in the past, enthusiasts will continue collecting BL data with their own
expenses, time, and health. Perhaps, a time will come when people will start
showing interest in this topic.
At data gathering, however, it is not necessary to forget the limitation of means
by which BL data are extracted: about the limitation regarding the range of human
senses. In fact, through vision, we cannot define whether the fireball radiates in
infrared and ultraviolet areas of a spectrum, whether its light is polarized, whether it
consists of narrow spectral lines or has a continuous spectrum, and is radiated in the
form of short impulses or is continuous. Unfortunately, cameras and video-cameras
cannot help in the search of answering these questions, as the range of their
perception coincides with the human vision possibilities’ range. We do not feel
the presence of weak and average magnetic and electric fields, radiowaves, we
cannot define all the shades of a strong smell, and we do not hear ultrasonic and
infrasonic sounds that BL can generate. Certainly, all these can be defined by means
of devices, but where is this universal portable device that could be supplied to all
potential observers of BL?

6.1.3 Experimental Modeling of Ball Lightning

For many years, BL studies included repeated attempts to reproduce this phenome-
non in the laboratory. These experiments can be divided into two groups. The first
one is an experimental verification of the definite conclusions of BL models. The
second one comprises attempts to “invent” the new form of the electric discharge in
which conditions of “a fiery sphere”, similar to that realized in the nature would be
generated automatically. This division is conditional to a large extent, as in both the
approaches, there is the general aspiration to imitate nature and in any way to
deduce a “technology” that can be used for the creation of BL.
Long before it was confirmed through statistical analysis of observations,
researchers intuitively felt that BL’s birth is somehow connected with the discharge
of a linear lightning. Therefore, it was natural to consider that it can be obtained by
means of the electric discharge in the air or in other natural environment. According
to Ford [79, P. 211], BL’s occurrence in the discharge can promote: the presence of
water vapor in air, dust, or soot; presence of nonhomogeneities in current conduc-
tors of bends and waists type; very slow increase in electric-field intensity to
breakdown potential; and the presence of a shock or sound wave, for example, a
thunder.
6 Ball Lightning Investigations 247

Fig. 6.4 A scheme of G. Plante’s experiment [35]

6.1.3.1 Experiments on Ball Lightning Modeling Made Before 1988

Experiences on the generation of shining spheres during electric discharge began

from the time of creation of the first electric devices (influence machine), first
electric capacitors (Leyden jar), and first chemical sources of current (galvanic
batteries and accumulators). The description of the experiments on the generation
of shining balls (see Fig. 6.4) at the application of the direct current (Plante [80],
N.A. Gezekus) or alternating current can be found in some earlier books [8, 18, 32,
35, 48]; therefore, it is senseless to repeat them here. Instead, we will mention some
experiments that have left an appreciable trace in the history of BL experimental
researches.
The very first experiment, in which the ball with the properties of a BL was
casually formed, was carried out in 1757 by the natural philosopher, Arden, in the
presence of Wm. Constable, Esq. [79, P. 216–221]. In the experiments, he used the
Leyden jar, made from crown glass, of 4 gal (18 L) in volume, which was connected
to the high-voltage generator through the brass core located in the glass isolating
tube. After the researcher made “nearby 100 or 150 turns” by rotating the handle of
the electric machine’s wheel, in the bottom part of the jar, he “perceived a ball of
fire, much resembling a red-hot iron bullet, and full three quarters of an inch in
diameter, turning round upon its axis, and ascending up the glass tube that
contained the brass wire.” The researcher continued to rotate the wheel handle.
The fiery ball “continued turning upon its axis, and ascending up the glass tube till it
got quite upon the top of the prime conductor. There it turned upon its axis some
little time, and then gradually descended, turning upon its axis as it had done in its
ascent,” on the bottom of the jar. “But soon after this, a very bright flash was seen; a
large explosion was heard, and a strong smell of sulfur was perceived all over the
room; a round aperture was cut through the side of the jar, as fine as it had been cut
with a diamond, rather more than three quarters of an inch (2 cm) in diameter, and
248 V.L. Bychkov et al.

between two and three inches (5–7.6 cm) below the top of the coating.” Though
researchers have replaced the punched Leyden jar, the effect was not reproduced.
A very large resonance was caused by cases of fiery sphere formations in the
American submarines [81]. Spheres appeared at disconnection of the relay silver
contacts at the passage of a strong current through it. The green fiery spheres
separated from the contacts, which “floated” indoors for some seconds. “In 1947
in Philadelphia experiments onboard a submarine ‘Catles’ have been made. The
sphere has been created in an engine room at voltage of a direct current 260 V and a
current strength of 156 kA, the peak value of power was thus equal 40 MW. As time
of disconnection for breakers of mechanical type is equal to 0.1–0.01 s then energy
of plasmoid formation is of an order 0.4–4.0 MJ. The diameter of the sphere has
appeared equal to 10–15 cm.” In the 1980s, G.C. Dijkhuis undertook experiments
on the reproduction of the processes described by Silberg [82]. For this purpose, he
used the complete set of submarine accumulators with the power of 30 MW
developed at short circuit. At a current higher than 150 kA, orange fiery spheres
of 10 cm in diameter with a lifetime of 1 s were obtained, which scattered from the
place of the relay contacts’ disconnection over the parabolic trajectories.
The approach to BL modeling goes in two directions: development of concep-
tions about it as an independent object and as an electric discharge in air. The most
impressing experiences on the generation of high-frequency discharges in air were
realized in 1941 by G.I. Babat in blockade Leningrad [8]. Babat used the generator
with a frequency of 60 MHz and a power of 20 kW. The energy of the generator was
delivered to a glass cylinder of 5 L in capacity through its walls in the form of an
alternate magnetic field. At first, the cylinder was pumped out to a pressure of 0.1–5
Torr. After switching on the high-frequency generator in a cylinder, a shining red
ring appeared. After that, the cock was opened and air filled the cylinder. The ring
began to pull together in a lump and turned to a fiery sphere. In the process of
pressure increase, its color at first became violet, then green, and at last, became
dazzling white. Fiery tips were pulled out from the sphere’s surface, and it was
possible to distinguish concentric zones of different colors and different intensities
of luminescence. At the shutdown of the generator, the sphere went out, and dull,
green shining fog remained in the cylinder.

6.1.3.2 Experiments on Ball Lightning Modeling Discussed During the First

Symposium on Ball Lightning

During the symposium, two works devoted to experimental modeling of BL were

presented. Ofuruton and Ohtsuki (Japan) [1, P. 310–324] repeated the experiments
of Barry [35] on the inflammation of strongly diluted mixtures of flammable gases
with air. They used methane, ethane, and an aerosol in the form of cotton fibers as
fuels. The electric spark was applied to initiate the reaction in a gap with a length of
3 mm between the copper electrodes. The mixture of gases filled a glass vessel of 73
 37  43 cm in size. During the inflammation of 2% methane–air mixture, a flame
6 Ball Lightning Investigations 249

of the spherical form of 3–6 cm in diameter appeared. At first, it had green color
because of the presence of copper atoms (from electrodes), and then it took the
color of a usual flame. In the experiments with the mixtures containing 2% ethane,
spheres of about 4 cm in diameter, in the beginning were greenish, and then turned
red. The lifetime of fiery spheres was nearly 0.3 s. Also, experiments were carried
out when in a vessel that contained about 100 cm3 of cotton fibers that were broken
into small bits. The fibers were distributed over the volume of the vessel by a stream
of compressed air. During the inflammation of this mixture, a shining white sphere
of 3 cm in diameter was formed. After that, its size decreased and it disappeared
0.8 s after the discharge initiation. In the experiments, two shining objects appeared
after the discharge pulse near electrodes, in the presence of a mixture of 2.7%
ethane, 100 cm3 of cotton fibers, and air in the vessel. The first of them (5 cm in
diameter) disappeared after 2 s, and the second (1 cm in diameter) moved upwards
and downwards and vanished after 1 s.
G.C. Dijkhuis (Netherlands) [1, P. 325–337] reported about the creation of
installation to verify his BL model. According to this model, in the nature, long
life span and huge energy of BL are the result of superconductivity and nuclear
reactions as well as the presence of fusion reactions in strongly turbulent plasma of
BL. He considered that BL is formed in the atmosphere by means of free-electron
gas-phase transition. Electrons, owing to exchange interaction, unite in Cooper
pairs, and collapse in a collective basic state in particles satisfying Bose–Einstein
statistics. The “bosonizing” threshold exceeds the value of electric-field strength
arising before a linear lightning breakdown development only by four times. On the
basis of the nonrelativistic theory of the charged superconductors, microscopic
bosonic vortex is represented as a core of self-captured Cooper electron pairs,
around which positive ions rotate on circular orbits owing to Coulomb’s force. If
the radius of their orbit is close to Bohr radius, then the speed of the ions reaches a
level at which reactions of nuclear fusion are possible.
An experimental installation for research on the possibility of “bosonic” BL
existence has been created. Laval nozzles were used for cooling a gas mixture to the
temperatures necessary to start the process of bosonization. A small amount of
deuterium was introduced into a nitrogen stream. High-voltage electric capacitors
were used for the charge accumulation, whose value can be compared with the
charge of the thunderstorm clouds. In the space between the electrodes, the dis-
charges were initiated, whose currents were comparable with those of usual light-
ning. Two capacitor batteries in the capacity of some microfarads were used, and
were charged up to a voltage of 240 kV. Thus, a full charge of the battery was nearly
1 C. The voltage of each capacitor’s battery was put to opposite located electrodes.
Full inductance of the batteries was equal to 124 mH, and the period of current
oscillations was about 50 ms. The experiment was undertaken as follows. Batteries
of capacitors were charged, a gas was delivered to a pipeline with a supersonic
nozzle, and two dischargers were simultaneously short-circuited. The voltage on
the electrodes on the nozzle side increased, and at last, breakdown took place. After
the breakdown, the voltage on the electrodes decreased and experienced strong
oscillations. The highest value of the voltage appeared to be equal to 94 kV.
250 V.L. Bychkov et al.

This value was smaller than those necessary according to the calculations to start
the process of electron bosonization (143–157 kV). The number of neutrons
measured with the help of the activation analysis did not exceed 3  1011, which
was smaller than expected. By means of a high-speed video camera, the place of an
exit of the gas from a nozzle into the air was recorded. No long-lived fiery spheres
were discovered; however, the gas luminescence proceeded approximately ten
times longer than the duration of the electric discharge.

6.1.3.3 Experimental Investigation of Ball Lightning After 1988

As stated earlier, there are simultaneously two approaches to the conception of BL

entity. BL, according to the first, is considered as a special discharge form in air,
which gets energy from an external source. According to the second, BL is an
independent body with an internal energy source. There are two branches of
experimental investigations with respect to these conceptions.

High-Frequency Discharge in Air

We have described the experiments of Babat with high-frequency discharge in air.

However, he did not connect them with BL. A hypothesis according to which BL is
fed by radio-frequency radiation energy of thunderstorms was proposed by Kapitsa
[83] in 1955. He was also one of the few scientists who have verified their ideas in
practice. Kapitsa, along with other research workers, managed to realize the high-
frequency constricted discharge in the atmosphere of helium, argon, carbon diox-
ide, and air at a pressure from 10 Torr to several atmospheres [84]. The discharge
was lighted in the spherical cavity resonator working on H011 wave type, the
wavelength of radio emission was equal to 19 cm, and power introduced into the
resonator ranged from 0.6 to 6.4 kW. At low powers, the discharge had an oval form
with dim borders; however, with an increase in power, it took the form of a cord,
came up to the top wall of the resonator, and soared at some distance from it. The
discharge became steadier when there was a gas rotation, and the temperature of its
central part was about 6,000 K. In spite of such a high temperature, heat exchange
of hot area with a wall appeared insignificant because of the occurrence of a double
layer of charges on its border. The plasma formations obtained by Kapitsa had a
faint resemblance to BL. However, the discharge in the air was supported only in
the presence of a continuous supply of a considerable radio-radiation power to it.
With the shutdown of the radio-frequency generator, the discharge died away.
The shining spheres, similar to plasmoids investigated by Kapitsa, was also
obtained by the Japanese scientists [85]. Spheres in air at atmospheric pressure
appeared inside the cylindrical cavity resonator of 16.1 cm in diameter and 37 cm in
length, with a supply of a continuous flux of radiation power of 5 kW at a frequency
of 2.45 GHz (l ¼ 12.24 cm). Four types of plasma spheres were observed. Spheres
of the first type were steady and only sometimes showed the tendency to be divided
6 Ball Lightning Investigations 251

into two spheres. Formation of plasma spheres of the second type began with bright
flash with 0.2-s duration. After that, the sphere was divided into two spheres, and
these spheres started to move in the resonator in different directions. The discharge of
the third type had the form of a flame and was located in a place of the connection
with the waveguide bringing the radiation to the resonator. Sometimes, this flame was
drawn into the waveguide. If the ceramic plate of 3-mm thickness was located
between a waveguide and the resonator, then the plasma formation passed through
a plate, neither destroying nor melting it. The fourth type of plasma formations had
the form of the flame flowing from the resonator in free space through a hole in an
aluminum foil, which was burnt by the flame, closing the resonator’s end face. This
“plasma fire” continued to exist during 1–2 s after the shutdown of the magnetron.
The shining plasma formations, similar to the plasmoids obtained by Kapitsa and
plasma spheres obtained by Ofuruton and Ohtsuki, have been observed by the authors
of another study [86].
Later, Ofuruton along with his coworkers, received plasma fiery spheres not in a
waveguide but in free space [87, 88]. In their experiments, a magnetron continu-
ously radiating power up to 5 kW at a frequency of 2.45 GHz was used. The radio-
frequency radiation went to air through an open waveguide. Two copper electrodes
of 3 mm in diameter were placed near the waveguide; the gap between them was
3 mm. The distance from a spark gap to a waveguide cut varied from 0 to 9 cm. The
capacitor with the capacity from 0.2 to 10 mF, charged to 8–10 kV, was discharged
through the electrodes. When the magnetron was switched on, fiery balls appeared
in air after the discharge; they were of spherical or ellipsoidal forms. At distances
from a spark gap to the waveguide, lesser than 3 cm, the plasma sphere usually
moved horizontally and was drawn in the waveguide. At the distances greater than 5
cm, it moved upwards. The color of the sphere changed from red to white. After the
discharge initiation, the sphere, gradually decreasing in sizes, disappeared after
0.5 s. The spheres appeared with 100% probability at the large discharge power
and/or the large power of the microwave radiation.
As can be seen, if it was possible to find the processes providing continuous
delivery of some kilowatt of high-frequency power in some allocated area of space,
then “Kapitsa’s BL” would become the real fact in nature. Regarding the continuity
factor, it seems to be an intractable problem. However, the second part of a
problem – concentration of energy in a definite place – can be solved with an
interference of radiowaves. In this connection, “Anderson localization” is an
intensively investigated problem, which is a possibility of local increase in high-
frequency field intensity owing to the summation of amplitudes of the waves
reflected from subjects located in a random way [89].
Tanaka and Tanaka [90] presumed that “Anderson localization” in air conditions
of a fireball formation in the form of the plasma spheres obtained by Ohtsuki and
Ofuruton can be created. The authors [91] made numerical calculations of electric-
field strength with a frequency of 2.45 GHz in a rectangular waveguide for a case in
which a row of ceramic plates of 2.25 mm in thickness was located in it. Two kinds
of one-dimensional disordered systems of an arrangement of plates were investi-
gated: a quasi-periodic system of some Fibonacci series type and a fractal system
252 V.L. Bychkov et al.

of the Cantor lattice type. In both the cases, from the calculations of the authors,
the intensity of a high-frequency field was found to increase more than 100 times in
the limited areas of waveguide. Further work of the Japanese scientists was devoted
to experimental verification of the obtained results [92–96]. When ceramic plates
of 2.25 mm in thickness were used, whose sizes were equal to those of a waveguide
(109  55 mm), the passing and reflected power was measured at a power of
1–2 kW of the microwave radiation. The plates were placed according to Fibonacci
series from first to sixth generation (22 plates in the latter case were used). At
transition from fourth to fifth and from fifth to sixth generation, it was possible to
hear a sound of breakdown of air in the waveguide; however, the plasma discharge
did not appear.
As it is improbable that in the nature there can be conditions when a high-
frequency field strength of external sources exceeds the breakdown value of air (>3
 106 V/m) or there may be a local-pressure decrease promoting ignition of the
discharge, researches on the discharge ignition processes by means of passive
initiators has gained interest. Metal needles, multi-needle metal whisks, metallic-
dielectric plates, flame, and aerosol can serve as initiators. Using the initiator made
of a wire of 1 mm in diameter, the authors [45, P. 119–136] obtained the low-
threshold spherical high-frequency discharge at the frequency of 7 GHz, which
existed at energy fluxes of 10–100 mW/cm2 in an air pressure range from 3 to
110 Torr. The maximum diameter of a sphere was equal to a wavelength of
radiation, and the minimum was equal to the quarter of the wavelength. With the
increase in pressure, the discharge took the form of a torch, and the intensity of its
luminescence sharply increased. At atmospheric pressure, the discharge was sup-
ported by the power of some hundreds watt. In [45, P. 136–159], for discharge
ignition, the short pulse of powerful high-frequency radiation was used, and plasma
maintenance was carried out by the second pulse of small amplitude, which was
incapable to independently cause a breakdown of the gas.
Chukanov [97] obtained shining spheres in a quartz flask of 10 cm in diameter,
placed inside an industrial microwave of 2.2 kW. The flask could be pumped out by
the forevacuum pump and could be filled with various gases. It had two tungsten
electrodes, brought to its internal wall. The alternating voltage (60 Hz) of 1,500 V at
a current of 10 A could be delivered to the electrodes. Besides, the electron beam
could be injected into the flask in a radial direction (from the periphery to the
center). At first, the flask was pumped out. When the microwave was switched on,
the discharge was lighted. A gas was delivered to the flask, and brightness of the
luminescence increased. The limiting pressure of the gas was 1 atm. When
the electron gun was switched on, the gas luminescence became very bright. The
alternating current was given to the electrodes – the plasma started to shine, similar
to a lamp of 12-kW power. The discharge burned easily in air, oxygen, nitrogen,
and carbon dioxide. In hydrogen, helium, and water vapor, the discharge did not
burn.
The special group of experiments comprised tests that investigated the burning
processes in a high-frequency field. Golka obtained plasmoids by placing a 5-cm
burning candle in a household microwave. During the time while it was switched
6 Ball Lightning Investigations 253

on, fiery spheres [98] were flying. In [45, P. 154–155], the interaction of the high-
frequency discharge with a candle flame was investigated. It was found that a
diffusion-type electrodeless discharge appeared in a zone of a burning candle
because of the formation of a cavity with the lowered density of the gas.
In [99], the technique of realization of fiery spheres from hot points of the fused
substance was described. The rectangular waveguide with a length of 20 cm and
section of 8.6  4.3 cm2 was used, whose open end was blocked by a mirror made
of vanes under cutoff, and in the other end, power of 0.6 kW was delivered from a
magnetron working at a frequency of 2.45 GHz. In the center of the wide side of a
waveguide at a distance of 5 cm from its open edge, the hole, into which the metal
drill reaching an internal surface of the bottom side of a waveguide could be
inserted, was drilled. This configuration was termed “a microwave drill” [99]. In
the bottom part of the waveguide, the substrate of silicon, glasses, germanium,
alumina, or basalt, in the form of plate, bulk, or powder could be placed. At first, the
drill resulted in contact with a substrate, thereby concentrating the release of
microwave power in a contact point. There appeared a hot point, and approximately
after 1 s, there was substance melting. “After that the microwave drill bit was pulled
out and detached from the molten spot while dragging a condensed, nearly vapor-
ized drop out of the substrate. A stream of fire is ejected then from the molten hot
spot forming a fire column . . . Shortly after, this fire column rises, detaches from the
hot spot, and forms a fireball in the air. The confined fireball flies up to the metallic
ceiling, where it is slightly squeezed to a form of a Gaussian-like shape. The
floating fireball looks like a dense heavy vapor, almost a liquid, glowing in yellow
red . . . It becomes buoyant in the air and quivers elastically as a jellyfish. The
floating fireball ( 15 cm3 in volume) may remain stable as long as microwave
power is on, and it continues to glow for another 30–40 ms after the microwave
energy is turned off.” It was possible to observe kickbacks of a fiery sphere at
collisions with walls, rotation of a fiery column, and sphere movement in the
waveguide in a direction towards the power source. At the contact of a teflon
partition, the sphere left burn traces on it, and its contact to a glass plate of 2-mm
thickness resulted in its melting, cracking, and destruction of the fiery sphere.
In [100], to obtain shining spheres, streamer discharges created by high-
frequency Tesla generator at a frequency of 67 kHz, with an average power of
3.2 kW submitted to a high-voltage electrode were used. The authors stated that
they used reconstructed N. Tesla installation with reduced size. It has been experi-
mentally established that during the development of discharges, a large number of
carbon particles and particles of the evaporated metal from the electrode in small
near-electrode area were formed. In the same area, there may be strong electric
fields (with a strength of up to 10–20 kV/cm), and thus, the formed particles were
quickly heated up. The spheres appeared near the high-voltage electrode as if they
appeared “from nowhere,” because they were not observed on the previous frame
that was video-recorded, i.e., their time of formation was tcr < 40 ms. The spheres
had different colors, including red, yellow, blue, and white, their lifetime
reached 2 s, and their diameter was 1–5 cm. Often, at the end of existence, they
blew up with a loud bang that indicated the energy store E 5/2
kT
N > 0.2 J/cm3
254 V.L. Bychkov et al.

(where 5/2
kT
N – equilibrium energy content of air at room temperature). The best
results were obtained in those experiments in which the high-voltage electrode was
covered by wax or charred wood.

Plasmoids

In 1999, at the Sixth Symposium on Ball Lightning, P. Koloc reported about the
experiments regarding the installation of “Plasmak”; he had stated its theory at the
First Symposium in 1988 [101, 102]. The scheme of its installation in [101] is not
represented; however, through indirect signs, it is possible to conclude that it is the
coil of a thin copper wire, which blows up when a strong current pulse passes
through it. Energy from 20 J to 5 kJ was put into the discharge. Koloc managed to
realize plasma formations of approximately spherical form with smeared
borders. The plasma spheres had the sizes from 2 to 18 cm, radiated white light
during 5–50 ms, and usually emerged upwards. Emission of a substance from
sphere poles (along an axis of a plasma torus) was sometimes observed. Energy
measurements of plasmoids were not undertaken. The small lifetime of the plasma
formations testifies to fast attenuation of currents in a “mantle” of the plasmoid,
which indicates the absence of superconductivity in it.
Since 1971, in the Vladimir State University (Russia), experiments on obtaining
electric discharges with possibly large current strength have been undertaken [103].
The last variant of power installation is the system with the inductive storage of
energy [104]. The electromagnet with the opened U-shaped core of a transformer
steel of 22 t in weight serves as energy accumulator. Fifteen coils of several parallel
bundles of a copper wire are reeled up on the core with the total section of 4,000
mm2. The inductance of the storage is 6.5  104 H. At a feeding current of 30 kA,
an energy of about 500 kJ is stored in it. The current pulse duration at removal of
energy of the storage can be changed from 70 to 120 ms. This installation works as
follows. The storage inductance is connected with the block of rectifiers through the
switch. After accumulation of magnetic energy, the storage by means of the second
switch is connected with a load, and 70 ms after that, the first switch is discon-
nected. Extra current of disconnection is generated in a working chain. This current
passes through a round metal diaphragm of 56 mm in thickness. One electrode is
connected with a peripheral part of the diaphragm, and the second, with its center.
As a result, the current flows over the disk radii. At the moment of diaphragm
explosion, the currents flowing in it and lead wires create a toroidal magnetic field.
Ions appear to be “frozen in” this field. Convection electric currents appear in
electrically neutral plasma. These currents take the form of a torus, and then the
form of a sphere [105–107]. A transformation of a torus in a sphere occurs at 0.25 s,
whereas the magnetic field fades at 0.4 s. The generated shining object has the form
of a spheroid with a sharp border. After that, diffuse decay of electromagnetic
energy occurs from thermal to optical ones, and the object slowly dies away.
Sometimes, at local destruction of a sphere surface, “energy flash decay, including
a beam with characteristics close to those of a laser beam” occurs. It is revealed that
6 Ball Lightning Investigations 255

radiation is polarized by 1.5–5.5% and the polarization plane is parallel to the plane
of a diaphragm [108]. The authors [109] considered that it is possible to explain
these phenomena by accumulation of electromagnetic radiation energy in the
internal area of plasma formation, which occurs owing to light reflection from
internal walls of the toroid caused by increase in free-electron concentration in the
direction of a vortex radius.
Some authors [110, 111] obtained plasmoids in the form of heated air spheres
kept from expansion by atmospheric air pressure. They came to a conclusion that
heating of a gas in a volume of some cubic centimeters by the powerful spark
discharge leads to pressure increase in it to several thousand atmospheres, which
causes fast destruction of the sphere by a supersonic shock wave. As a result, the
lifetime of the heated sphere does not exceed 1 ms. In this connection, they have
chosen a mode of slow heating of a gas volume. They used six high-voltage
transformers (60 Hz, 120 V/15 kV, 60 mA), whose secondary windings
were switched in parallel through the spark gaps exciting small volume of air.
The transformers were periodically switched on for 0.5 s, and then were switched
off for 0.5 s. The shining formations rose upwards from the discharge zone existing
from 0.25 to 0.5 s.

Erosive Discharge

In the early 1990s, the pulse electric discharge in the channel with evaporating
walls (the erosive discharge) attracted the interest of researchers. The plasma
stream that is ejected out of the channel has the density close to air density, and
at low temperature, the intensity of radiation is small, the density of energy is high,
and the concentration of the charged particles is high. This stream burns holes in the
metal plates, but leaves a paper intact, and seeks to keep the integrity at an
encounter with an obstacle. According to some authors [45, P. 15–66; 112], this
set of properties is typical for BL. Owing to this fact an erosive plasma stream was
called a laboratory analog of BL. This analogy looks adequately artificial (and
should cause the active protest in those who saw a real BL), as a stream of flying
plasma, neither by the form nor by the motion character, remind of BL in any way.
Nevertheless, analogy declared in [112] has been unconditionally accepted by other
researchers.
In typical installation for generation of erosive plasma streams, capacitors from
630 mF to 2.5 mF charged to a voltage of 450 V are used. Capacitors are discharged
through the channel of 1 mm in diameter and 5 mm in length drilled into a plate of
an insulator (organic glass is used more often). For the maintenance of the laminar
flow of the evaporated substance from the channel, it is necessary that the increase
in the speed of the current does not exceed 105 A/s; at higher speeds, the stream
becomes turbulent. For the creation of the required form of a current pulse, an
inductance is included into a discharge circuit. The typical length of the stream at
atmospheric pressure is about 10–40 cm, and speed of its motion is 60–100 m/s. The
stream consists of a thin central part (core) of 1–2 mm in diameter and a cover of
256 V.L. Bychkov et al.

6–10 mm in diameter. The continuous spectrum of radiation of the stream is caused

by a luminescence of soot particles, and lines of atomic radiation of hydrogen,
carbon, copper, sodium and calcium, and molecular bands of C2 and CN are also
observed. On the whole, the plasma formation turns out to be electrically neutral;
the excess negative charge in the core does not exceed 103 from the charge value
that has passed through the discharge channel. The maximum floating potential of
plasma is 12 V. The radiation power of the plasma formation in wavelengths
ranging from 0.2 to 10 mm appeared to be about of 30 W, and thus, no more than
1% of the energy stored in it (50 J) leaves during its existence. The temperature of a
neutral gas in the stream appeared to be equal to 1,000–6,000 K [113], and the
temperature of ions was defined as 6.5  103 K. The plasma stream is capable to
pass through a hole in a textolite plate. If the diameter of the hole is small, then only
the core passes through it, and the cover “is taken off.” It was revealed that at the
direction of the plasma stream in a ditch with liquid nitrogen in it, bright spherical
plasmoids up to 5 cm in diameter, whose color during a lifetime (20–40 s)
changed from white to violet, and dark spheres from 1 to 2 cm in diameter existing
for 1–2 s were formed. The energy stored in bright plasmoids reached 100–200 J.
Some authors [45, P. 72–78] considered that plasmoids consist of Rydberg sub-
stance – an ensemble of highly excited atoms.
Investigation of the material objects that are taken out by a stream from the
discharge channel also attracted considerable interest. To increase the weight of the
evaporated substance, the wall of the channel was covered with wax, and the length
of the channel was increased to 5 cm [45, P. 66–72; 114]. In the head part of the
stream, an ellipsoidal “shaggy” plasmoid consisting of a set of branching shining
filaments from 2 to 6 cm in the length, created by such a generator was formed.
After plasmoid extinction, the polymeric threads of 1–10 mm in diameter and 1 cm
in length stayed in its place. According to some authors [114], in the course of
plasmoid material condensation, a formation of mosaic, such as a charged fractal
polymeric structure, capable of saving electric energy up to 100 J, was observed.
For obtaining the fractal charged objects, the authors [45, P. 87–95] placed the
activated coal or lumps of cotton wool in the discharge chamber. At the discharge
from the channel, parts of 5 mm in size were flown out, which were shining for
about 6 s with red light. Under a microscope, it was possible to see white porous
structure with short fibers on the surface of the parts; their density was about 5 
102 g/cm3. The authors of [115] placed metal particles of 50–100 mm in size in the
channel of the plasma generator made of organic glass. Thus, on an exit from the
plasma generator at a distance of 30–100 cm from an injector, there was an
explosive formation of bright areas of 10–30 cm in size, shining for about 0.1 s.
At the interaction of a plasma stream with a tin wire, there were yellow-red balls of
3–5 mm in diameter. Falling on a plastic surface, they jumped aside, leaving traces
of burns on it. The cooled down surface of the balls appeared covered with threads
of tin oxide of 50 mm in length. Besides the balls, there were fractal structures up to
3 cm in length and of some millimeters in width, with fractal dimension D > 1.8 
0.2. Within several minutes, they became highly elastic, and then became fragile
and quickly got destroyed.
6 Ball Lightning Investigations 257

The authors of [115] carried out an experiment whose conditions were close to
those which can be realized when a lightning strikes the earth. The electric
discharge occurred in a thin tube of an insulator with thick walls. The products of
the discharge consisting of steams of a material of a tube and electrodes, and also of
ionized air, having broken through a tube wall, went outside. The lifetime of similar
formations reached 5–7 s. Having hit an obstacle, they blew up. Their remains
reminded shells of the oxidized metal. It was one of the few experiments in which
an explosion of luminescent formation was observed. In [116, 117], the stream of
erosive plasma with energy of about 200 J passed over a dish with melted paraffin or
wax. After the stream passage through a cloud of steams over the dish, there were
emerging formations of spherical or mushroom forms of 10–20 cm in size.
They were shining for 0.5 s. With the use of the plasma generator with a capillary
of 1–2 mm in diameter, drilled in a composite material of a mixture of
paraffin, rosin and wood sawdust of size 1  0.3  0.3 mm shining objects of
1–2 cm in size with a lifetime of 1–2 s were obtained. They left the channel
approximately 30 ms after the start of the plasma stream. Their residuals repre-
sented particles of black color of 1–2 mm in size with polymeric fractal-porous
structure of the surface.
In [118], the interaction of the erosive discharge stream tip with an oncoming
supersonic stream of air was investigated. It was revealed that the shock wave is not
formed in front of the plasmoid owing to the fact that a drag force falls by 20%. The
plasmoid is a source of radiowaves with a wavelength of 0.5–2 cm and it amplifies
laser radiation at wavelengths of 0.63 and 10.6 mm. In [119–121], it is shown that a
cluster plasmoid obtained in the erosive discharge bears a positive charge on itself.
The floating potential of the plasmoid was equal to 100 V. This value is 100 times
greater than the expected potential of decaying plasma with an electron temperature
of 104 K.
In the nature, the configuration similar to the device of laboratory plasma
generator can appear when a lightning strikes the earth. As is known, fulgurites
thus appear, which are tree-like formations from caked sand and soil. Abrahamson
and Dinnis [122] assumed that at high temperature, sand (SiO2) and organic
substances of soil can react resulting in the formation of silicon:

SiO2 þ 2 C ! Si þ 2 CO:

Silicon vapors are thrown out to the atmosphere, where their condensation takes
place and nano-size silicon balls are formed. Under the influence of electric fields,
balls gather in aggregates that start to shine because of warming up owing to the
reaction of silicon oxidation, Si þ O2 ! SiO2 : Abrahamson undertook experiments
to imitate conditions when lightning strikes the earth [123]. He passed a discharge
current of capacity of 204 mF, charged to a voltage of 20 kV, through a layer of
humidified soil of 3-mm thickness. The soil filled the bottom electrode – a graphite
substrate; as the top electrode, the graphite rod of 15-mm diameter was used,
located at a distance of 22–36 mm from the surface of the earth layer. The energy
released at the discharge was 110 kJ, and a charge of 1.3–3.4 C was transferred
258 V.L. Bychkov et al.

through the soil. In experiments with the highest power, the negative potential was
applied to the top electrode. The reaction products were sucked away by the pump
and deposited on the filter of quartz fibers and the lattice of nickel. Fiber threads of
100 nm in diameter and 7 mm in length were found on the filter, by means of a
scanning electron microscope. Sometimes, spheres of 2 mm in diameter were
deposited on the fibers. Chains from 25 to 120 nm in length consisting of balls
with an average size of 25 nm were found on the surface of the nickel lattice. It was
not possible to define the elemental structure of the threads and chains. Spheres of
2 mm in diameter basically consisted of silicon, and impurities in them included Al,
Mg, Ca, Na, and Fe (components of minerals of soil). The author of [123] did not
manage to observe the formation of aerogel-isolated spheres. He explained that it
was because of a strong shock wave, and accordingly, a strong gas disturbance near
the discharge area. Therefore, he considered that in the nature, vapors from a cavity
made by a lightning come out after the discharge is terminated. The ring whirlwind
is formed of charged nanoparticles.
The authors of [124] repeated the experiments of Abrahamson, by replacing a
layer of soil with a silicon plate of 5 cm in diameter and a thickness of 350  50 mm.
This plate was placed on the steel base of 10  10 cm in size and 5 mm in thickness,
serving as an earthed electrode. The rod of tungsten or graphite of 4 mm in diameter
and 30 cm in length was used as the second electrode whose movement could be
operated manually. The alternating voltage of 20–25 V (50 Hz) was applied to the
electrodes, and the current could be varied from 100 to 140 A. The rod was lowered,
and when it came into contact with the silicon plate, current appeared. After that,
the electrode was raised to 1–2 mm, and an electric arc appeared between the
electrode and the plate. At this time, shining fragments extensively flew out
from the arc. At the further lifting of the electrode, the arc went out. This cycle
took 1–2 s. Shining formations appeared, on an average, once in 30 attempts.
However, with the replacement of the silicon plate with an aluminum or copper
foil, addition of some salt water, or dry or humid microspheres of SiO2, or humid
wood, it was impossible to get shining spheres. The spheres that appeared in the
experiments with a silicon plate had a diameter from 1 to 4 cm and existed for 8 s,
whereas their fragments went out at 1 s. When hitting a floor, they jumped aside and
were scattered into small parts. There were observed cases when the shining sphere
passed through a crack under a plank or left behind a spiral smoky trace. Examina-
tion of the surface of a silicon plate using a scanning electron microscope demon-
strated that after the discharge, there were deepenings of about 5 mm in size and
aggregates of particle chains of micron sizes on the surface of the originally smooth
plate. The obtained shining formations had some properties coinciding with the
properties of natural BL; however, freely levitating sphere was never observed.

Electric Discharge in Water

For the purpose of reproducing the conditions in submarines, which produced fiery
spheres [81], Golka carried out experiments using a transformer of 150 kW power,
6 Ball Lightning Investigations 259

with 60-Hz frequency, a current of 10 kA, and voltage of 15 V [98]. At short circuit
of a copper cable of 2.5 cm in thickness and an aluminum plate of 3–6 mm in
thickness, under water, fiery balls of 6 mm in diameter appeared on the water
surface. Hissing, they slid on a surface. Sometimes, fiery spheres of the irregular
form rose in the air, leaving spiral smoky traces. Such fiery spheres could also be
obtained without water; however, in this case, they scattered with much larger
speeds, and it was difficult to watch them. These fiery spheres were shining with a
bright white light, which may be due to the slow burning of aluminum. At contact
with iron electrodes, yellow fiery spheres appeared. Their luminescence lasted for
2–5 s, and at the end of their life, their size decreased. The cooled-down spheres
represented metal spheres of about 2 mm in diameter covered with oxides in the
form of small hairs.
In 2000, in Gatchina near St. Petersburg (Russia), the new type of electric
discharge named “the Gatchina discharge” was achieved by its inventors
[125–135]. They consider the erosive discharge (“Avramenko stream”) as a proto-
type of this discharge, but unlike the last one, they could increase the size and life
span of the shining formation by two orders of magnitude [126, 127]. The basis of
installation for discharge realization is the condensers battery with a capacity of 0.6
mF, charged to a voltage of 5.5–6.0 kV. The plastic cylindrical vessel with water
was used. The earthened anode represented a ring made of a wire, lying at the
bottom of the vessel. The thickness of the water layer was usually equal to 15 cm.
As the cathode, the graphite electrode of 6 mm in diameter protruding over a water
surface at a height of 2–3 mm was used. The lateral surface of an electrode was
isolated from water by a quartz tube. By means of a manual discharger, the
capacity’s voltage was given to the cathode. The sliding discharge was developed
on a water surface, and from the electrode protruding over water, the stream was
thrown out upwards. The discharger was disconnected at the 80th ms (at that time,
the residual voltage of the battery was 3 kV), and the stream, separating from the
electrode was transformed into shining formation. From 60th to 100th ms, it had the
form of jelly-like body, and sometimes, took the form of an ideal sphere (see
Fig. 6.5). The object rose upwards and decayed. The decay started at 200th ms,
and the decay itself lasted about 400 ms.
By means of electric probes, the authors of [127–130, 132, 134, 135] found
that the object bore non-compensated negative charge on itself in the size of 2 
107 C on its cover. The cover potential exceeded 10 kV. The plasmoid did not
influence dielectrics, but melted off thin wires. Hence, when entering inside the
plasmoid with a nichrome wire of 0.08 mm in diameter on its tip, the ball of 0.2
mm in diameter [128] was formed. In [135], it was reported that at melting of a
wire of 0.1 mm in thickness, the ball of 0.5 mm in diameter was formed. The
authors considered that “the basic energy of ball lightning is stored in the form of
the electric field created by electron sphere, and this energy has the potential
character” [131]. We can estimate the energy of “Gatchina Plasmoid.” Let the
diameter of the charged sphere be D ¼ 13 cm, and its charge be Q ¼ 2  107 C.
The potential of the sphere is U ¼ Q/2pe0D, and energy of electric field is
We ¼ Q2/4pe0D. For accepted values of D and Q, one can obtain U ¼ 27.7 kV
260 V.L. Bychkov et al.

Fig. 6.5 A photo of Gatchina

discharge [135]

and We ¼ 2.8  103 J. Now, we will define the energy necessary for the melting
of the nichrome wire. A specific heat capacity of the nichrome is Cp ¼ 0.46 kJ/
kg·K, its density is rN ¼ 8.4  103 kg/m3, and temperature of melting is Tm
¼ 1,390 C [47]. For the formation of the melted ball with a diameter d, energy
Wm ¼ pd3rN CpDT/6 is required (here, DT is the difference between the melting
and room temperature). For the formation of a ball of 0.2 mm in diameter, energy
of 2.2  102 J is required, and the ball of 0.5 mm in diameter can be formed at
an energy expense of 3.5  101 J. This energy is 10–100 times greater than
those of plasmoid electric energy. However, in [132], the authors reported that
right after the discharge termination, the plasmoid charge was Q0 ¼ 4  106 C,
and accordingly, we found that the energy of its electric field was equal to 1.1 J.
At the same time, according to the authors’ observation, in the initial stage of
existence (during t ¼ 100–200 ms), the plasmoid intensively lost its charge
because of the corona discharge. Apparently, these can explain that at the
moment of its lifting to height where the probe is located (when a wire melts),
the plasmoid charge decreases by 99%. From here, an average current of the
corona discharge is Ic ¼ Q0/t ¼ 20–40 mA. For a sphere having a charge Q, a
current defined by air conductivity la is Ia ¼ laQ/e0 [136]. Hence, Ia ¼ 1.6 
102 mA at la ¼ 3.5  1014 O/m (a conductivity of dry air) [137] and Q ¼ 4 
106 C. For a “Gatchina” BL, current Ic exceeds Ia by 2,000 times. Apparently,
this can be explained by high conductivity of the channel (“leader”) con-
necting the emerging shining formation with the cathode of the installation.
6 Ball Lightning Investigations 261

Inconsistency of the results of the measurement of the laboratory BL energy

obtained from the results of melting conductors and its electric energy estimation
testify that there should be other energy sources in it.
The authors of [127, 138] considered that “Gatchina” shining formations appear
“because of structurally-energy self-organizing of a metastable substance on a basis
of coulomb and polarizing nonideality of a plasma containing the smallest particles
of metal and dielectric. The existence of this nonideality is maintained by catalytic
energy liberation in the electric discharge at a decay of chemically active plasma.
The majority of large particles concentrates near the external border, forming
‘a cold cover’, and small particles form ‘a hot core.’” “The object luminescence
has the chemical nature and, besides elementary reactions of chemiluminescence, is
supported by the energy accumulated by particles of an aerosol in the form of
vibrational and electronic excitation. It occurs thanks to burning and catalytic
reactions on surfaces of particles. To one of channels of liberation of excess
electronic excitation serves ejection as dust particles of electrons, ions, excited
molecules and atoms by dusty particles.”
In [125–135], to obtain shining formations, the electric charge was located on a
cover surface. Therefore, it cannot cause the force creating a surface tension. This
force could appear, if the charge has been located in the volume limited by the cover
[136, 139, 140]. Therefore, the sharp plasmoid’s border, most likely, appears
because of the vortical movement of heated gas volume rising upwards.
Creators of a new kind of discharge [130] reported that in case the capacity
discharge is broken off before the 80th ms when there is a voltage greater than 3 kV
on the battery of the capacitors, the second sphere takes off from a copper electrode
of the bar closing and disconnecting a circuit of the discharge. The color of both the
spheres is approximately identical. On the basis of it, the authors [130] concluded
that water in installation only plays a role of variable resistance. This conclusion
has found experimental confirmation [141]. It has been shown that the current falls
down faster than the voltage of the discharge and the voltage on the water surface at
a point located 1 cm from the cathode starts to decrease 50 ms earlier than at a point
at 7 cm from the cathode . Besides, it has been found that the shining formations,
similar to those observed at a standard “water” configuration of the discharge,
appear at short-term short circuit and disconnection of electrodes made of coal,
steel, copper, and aluminum or at passing of a current pulse through a glass with
metallic sputtering [142, 143].
In 2007, “the Gatchina discharge” was investigated by the German scientists
[144, 145]. Their installation was made under the scheme of [125]. The cylindrical
vessel of glass or plastic of 25 cm in diameter and height, filled with tap water
or distilled water with the salt additive (NaCl or CaCl2), was used. At the bottom of
a vessel, the copper ring (anode) was placed, and the central electrode (cathode)
acted at some millimeters over the water surface. The battery of the capacitors of
1-mF capacity charged to a voltage of 4.8 kV was connected to the anode through a
discharger that produced a current of the discharge from 10 to 100 A. After
100–150 ms, the current was broken off. At the 30th ms after the current interrup-
tion, the spherical object with accurate border was formed. In a photograph taken
262 V.L. Bychkov et al.

with large exposure, it looked as a bright homogeneously shining sphere; however,

at small exposure in a picture, drop-shaped bright core whose size was two times
smaller than the diameter of the sphere, surrounded by a weakly luminescent cover,
was observed. Usually, energy about 8 kJ was put in the discharge. Calorimetry
demonstrated that the energy spent for heating and water evaporation was about
5 kJ, and the other 3 kJ was spent for the formation of a plasma sphere. About 0.7 kJ
(23%) of plasmoid energy went to radiation. The temperature at t ¼ 235 ms from
the beginning of the discharge appeared equal to 900 K, and at t ¼ 260 ms,
T ¼ 600 K. At t ¼ 200 ms, the thermocouple detected a temperature greater than
1,300 K. By means of a probe, the floating potential of a rising plasma sphere was
measured, which appeared positive. From the 200th to 300th ms, it was equal to
about 0.5 V, and at the 650th ms, it decreased to 0.05 V. This indicates the presence
of positive charges in the sphere, also proving the presence of radiation lines of Ca
(II) and Mg(II) ions in the spectrum. The radiation spectrum consisted of lines of
elements with a small energy of excitation, whereas the lines with energy greater
than 5 eV were absent. Therefore, the electronic temperature in a sphere hardly
exceeded 1 eV. On the basis of Stark broadening value measurement of hydrogen,
Li(I), and Cu(I) lines, it was found that the electron density was about 1020–1022
m3 in the initial stage (from t ¼ 0 to t ¼ 75 ms); however, in the generated
plasmoid, it has decreased to 1014–1015 m3. The electron temperature after initial
fast falling from 5,000 to 4,000 K smoothly decreased to 2,000 K during the
subsequent plasmoid’s life. The molecular band of CaOH was a prevailing source
of visible radiation. Thus, it is possible to consider that the energy reservoir of
radiation is the chemical energy stored in water-dissociation products. Measure-
ments of the laser beam deviation angles at the passage through the sphere volume
showed that the average density of the particles in it is approximately eight times
smaller than the air density, which follows that the temperature of the sphere
substance is T ¼ 8  273 K ¼ 2,200 K [145]. It can be observed that the conclu-
sions of [144, 145] are somewhat differ from those of “the Gatchina discharge”
study: the sphere temperature appeared greater than that assumed earlier; the sphere
charge appeared smaller, and the chemical energy seemed to be the energy source
and not the electric one. In this connection, the burning of the sphere with a
combustible substance was already considered in [146].

Explosion of Metallic Wires

The authors of [147, 148] got shining formations at the explosion of wires of 0.5
mm in diameter of copper, tungsten, iron, titan, and nickel in small plastic ditch that
was filled with mixture of organic material with water. In the ditch, there was a hole
through which explosion products could go out to the atmosphere. At passing of a
current of 5–20 A through the wire, the wire evaporated, and from the gas that
moved through the hole, the shining sphere of 5–6 cm in diameter was formed. The
sphere was shining for 2–3 s. Then, the sphere was flattened and gradually turned to
a nonshining toroidal ring. It rose upwards – at first quickly, and then slowly – and
6 Ball Lightning Investigations 263

its size increased from 5 to 50 cm. According to the authors, the sphere lumines-
cence could not be supported by charges recombination, as the characteristic time
of this process was very short. Most likely, the reason for luminescence was the
burning of the disperse wire material in the air.
A.N. Vlasov considered that BL represents toroidal current layer induced inside
a vortical gaseous ring [149–152]. The toroidal vortex can be formed at the blow of
a linear lightning in the earth. For modeling this process, Vlasov undertook experi-
ments on the induction of a strong toroidal magnetic field at the passing of a current
pulse through the long wire coil rolled in the torus [153–155]. The winded coil was
made of a copper wire of 0.5 mm in diameter. Through it, by means of an electronic
key, the capacitors’ battery with a capacity of 45 mF charged up to a voltage of
300 V, was discharged. Subsequently, an electric explosion of the coil took place.
From the explosion place, the shapeless or pear-shaped body of 15–20 cm in size
rose up quickly, and was shining for about 0.3 s. Before decay, the plasmoids took
the toroidal form. At explosion of the spiral wire that had not been rolled in torus,
long-lived plasmoids with large sizes did not appear.

Microscopic Ball Lightning

In [156], the results of a relaxation of the expanding plasma formed at the high-
voltage discharge in water were described. For obtaining the discharge with a
duration of 2 ms, Marx’s generator with a voltage of 60 kV and an energy of 30 J
that has been accumulated in the capacitors was used. The discharge chamber, filled
with water, had a volume of 1.5 dm3. Observation of the discharge was carried out
through the windows located on an axis of the chamber. Under the impact of the
discharge in water, the pulsating cavity up to 2.5 cm in diameter was formed. It was
revealed that sometimes the plasma in the cavity did not disappear, and decayed to
separate shining objects from 0.1 to 1 mm in the size. Their number in different
experiments ranged from several units to several tens. Some objects went through
the cavity walls into a liquid and existed in it for about of 0.5 ms. Research on the
formations’ luminescence spectrum showed that at the change of wavelength from
550 to 750 nm, the increase in the intensity of luminescence was observed, and the
dependence course approximately coincided with a kind of a short-wave wing of a
black body spectrum at T ¼ 2,850 K. However, in the area of wavelengths from
750 to 900 nm, the intensity of radiation sharply fell. In the radiation spectrum,
there were no lines and bands of atoms and molecules. The authors considered that
the luminescence cannot be caused by the radiation of heated solid particles, as
photometric measurements of images demonstrated that shining objects are opti-
cally transparent. Subsequently, researchers included a device into the installation,
which allowed to throw out products of water plasma decomposition into the air
[157, 158]. At the discharge of objects from water, their size increased by three to
five times (from 1 to 5 mm), and the luminescence time reached 0.1 s. Sometimes, it
was possible to observe how spherical objects are formed during approximately
300 ms from the shining cloud, consisting of water plasma decomposition products.
264 V.L. Bychkov et al.

These objects were destroyed at passing through a zone of hot air (T ¼ 800–1,000

C), and on the contrary, condensation of objects on the surface of the plate cooled
to a temperature of liquid nitrogen was observed. Sometimes, objects blew up, and
the scattering parts’ speed was greater than 50 km/s. On the basis of the observation
over the behavior of these objects, the authors of [158] assumed that they can be the
porous friable bodies whose “skeleton” consists of unusual metastable composi-
tions of hydrogen and oxygen. At a transition from metastable to basic state, energy
release takes place.
T. Matsumoto [159–161] carried out researches on the spark discharge in tap
water and solutions of electrolytes. In one modification of his installation, the wire
was used as an electrode; it was made of lead, cadmium, iron, or gold, with a
diameter of 1 mm. Its one end was dipped into a solution of KOH (1.5 mol/L) or a
carbonate of cesium, Cs2CO3 (0.6 mol/L), to a depth of 3 mm. As the other
electrode, a box of 25  25  25 mm in size, made of polished copper plates,
was used. The box with the wire into it stood at the bottom of the Petri cup, in which
there was a layer of electrolyte of 8-mm thickness. Between the electrodes, pulses
of an alternating current with a voltage of 120 V (50 Hz) and a duration of 20 ms
were introduced, and the time between the pulses was 20–30 s. Traces of the round
form from 20 to 200 mm in diameter were found on the copper plates. The same
traces appeared in films with nuclear photoemulsion, which were located around the
discharge zone. Matsumoto considered that microscopic BL left these traces. In his
opinion, they represent clusters of strongly compressed hydrogen atoms. Binding of
atoms in such a cluster is ensured by exchange forces of degenerate electrons’
attraction. These clusters are capable to pass through plates made of dielectric, over
thin copper wires, jump aside from obstacles, and blow up at the end of its life.
In [162], the results of investigations of electric explosion of metallic foils of
50-mm thickness and 1  5 cm size located in the hermetic tank filled with water
were presented. Eight tanks were used, in which the discharge energy of 50 kJ at a
pulse duration of 60–150 ms was maintained. The spheres of 10–15 cm in diameter
appeared over the tanks at the discharge, and they were lighted for 5 ms. In the
photos of frame-to-frame shooting of the process, it was possible to see that the
originally diffuse luminescence of spheres changed to radiation of the bright
particles filling their volume. This can explain the process of condensation of
metal vapors’ plasma, which penetrated into the atmosphere through the destroyed
vacuum sealing of the tank.
In [163], “strange tracks” of the shining particles created by the pulsed erosive
plasma generator in external electric field were observed. The size of the particles
was 10–100 mm, they moved with a speed of 20–200 km/s, possessed an electric
charge, and at impact with a target, emitted soft X-ray radiation directed forward.
The characteristic time of these particles’ luminescence ( 1 s) considerably
exceeded the luminescence time of the cluster plasmas when it was discharged
from the plasma generator ( 10 ms). The shining particles flew off the capillary
dielectric channel of 1–2 mm in diameter with the walls made of organic glass,
paraffin, or wax. They could be collected in a ditch with liquid nitrogen. The
analysis demonstrated that they consist of pyrolitic graphite. In another experiment,
6 Ball Lightning Investigations 265

in a stream leaving the channel, the aluminum powder with the grains of size
10–100 mm was admixed. When they were discharged from the nozzle, these
particles burned. They appeared negatively charged and sometimes blew up
because of Coulomb’s forces. In a spectrum of burning aluminum particles along
with the lines of Al and AlO, the lines of Na, Li, and K were observed, and the last
ones appeared with a delay greater than 10 ms. Radiation in quasi continuum in the
spectra amplified approximately during the same time, corresponded to the equilib-
rium radiation of heated solid-state particles.

Experiments on Modeling of Ball Lightning Features

Along with the above-described attempts to obtain an artificial BL in laboratory,

experiments on modeling its separate properties by means of known scientific and
technical processes have been increasing. Among these properties accessible to
experimental verification, there were cases of its interaction with windowpanes and
features of its movement over the conductors.
There are many reports on the ability of BL to cut out round holes of 5–8 cm in
diameter in glasses, and sometimes, there are no cracks both in the cut-out disk and
in the glass that remained in a frame [11, 18, 32, 42, 76, 164, 165].
In [166], the case when the BL, in the presence of a group of schoolboys during a
lesson, heated up a window glass, forming hole was presented. On the basis of this
observation, the author of [166] undertook experiments to obtain holes in the
glasses through their local heating with the help of a beam of laser on carbon
dioxide, with a power of 500 W, or with a flame of a gas torch. The glass was heated
up to the fusion of an aperture in it. After that, the heating source was switched off,
glass was cooled, and a round crack covering a zone of fused glass in it appeared.
The diameter of the crack exceeded the diameter of a heated-up zone approximately
by 1.4 times. To heat the glass to fusion, it was necessary to spend energy of about
20 kJ. In these experiments, it was important to heat the glass area to a softening
that removed the stress of compression from other parts of the glass. At smaller
power of heating, the glass sprung. The results obtained in [166] do not absolutely
correspond to those of BL’s influence on glass, because in the latter case, burning-
off of a disk is not observed and there is no aperture in it.
This finding stimulated the authors of [167–169] to carry out experiments to
study the heating influence on windowpanes. Glass of 3.7-mm thickness was heated
up by a flat spiral of a nichrome wire of 0.7-mm thickness. The external diameter of
a spiral was equal to 40 mm, which contained 3–5 coils. The spiral was heated up
using an alternating current (50 Hz) to red or yellow heat. At values of the energy
from 12 to 24 kJ spent in heating the spiral, the glass got destroyed with the
formation of radial cracks, and at the heated place, there was a round or oval
piece of glass. It had the form of the truncated cone, and the cone base was turned
towards the surface that was heated up by the spiral. However, in one of the
experiments, both a round disk and other part of the glass practically remained
intact. In this case, on the brink of the glass, there was occasional crack nucleus.
266 V.L. Bychkov et al.

After 11 s following glass heating by a spiral fed with power of 260 W, the crack
from edge extended to a heating zone. After 5 min of heating, the current was
switched off, and the glass started to cool. During the course of cooling of the glass,
there appeared a ring crack whose diameter (50 mm) exceeded the diameter of an
external coil of the spiral by 1.25 times. Simultaneously, in an external part of the
glass, the second crack with its edge going tangentially towards a ring crack was
formed. Similar kind of cracks could be observed in some pictures of the glasses
that were affected by BL [32, 42, 164]. However, full identity of results when
without cracks remained both a disk, and the glass with a hole, it was not possible to
achieve in experiments.
The important property of BL, which can serve as a criterion for an artificial and
natural BL identity is its ability to move in a horizontal direction over a conductor
surface. In [136, 139, 140], the hypothesis that it is possible to explain such
BL behavior through the interaction of alternating magnetic fields generated by
currents flowing in a substance of BL with magnetic fields of eddy electric currents
induced in conductors, was stated. The results of the experimental verification of
this hypothesis were described in [170, 171]. The value of the force that pushed
away the coil fed by an alternating current with a frequency of 440 kHz from a
copper plate was measured. The coil from three densely laid loops of a copper tube
of 3 mm in diameter with an external diameter of D ¼ 52 mm was used. The water
stream that passed in the tube cooled it. The coil was connected to the high-
frequency generator with a limiting power of 10 kW. The tubes connecting the
coil with the generator simultaneously played the role of a spring, which at
displacement of the coil along an axis to 10 mm, had no residual deformation.
The coil was located at some distance l from the copper plate of 24  14 cm size
and 2-mm thickness. The plane of loops of the coil was parallel to the plate’s plane.
It was found that the force F that pushes away the coil from the plate appeared
proportional to the square of the value of the current flowing through the coil
(F I2), and was inversely proportional to the seventh degree of distance l (mm)
from the center of the coil to the plate’s plane (F ¼ K/(b + l)7, where the current
was I ¼ 26 A and the coefficient was K ¼ 3.7  107 mN mm7, and b ¼ 4.5 mm).
Because of the sharp dependence of the force F on the distance l, it appears to be
appreciable only at l
D; however, it sharply decreased at l ¼ D. Therefore, the
mechanism of the interaction of currents cannot explain the cases of BL levitation
over a conductor at the distances equal to its several diameters.
In this connection, in [170, 171], the assumption has been made that the presence
of the corona discharge on BL surface might be the reason for its levitation. This
discharge appears because of continuous leakage of a part of BL charge to the
atmosphere. At the corona discharge in the restricted area of space near an electrode
(in an ionization zone), there exists neutral plasma in which ionization and excita-
tion of a gas by electron avalanches occur. From the ionization zone, the flux of
charge carriers whose sign coincides with a charge sign of the corona-forming
electrode moves to an external zone. The space charge formed as a result of it
decelerates the further development of the ionization processes, weakening the field
near the ionizing electrode. It leads to localization of the ionization zone near the
6 Ball Lightning Investigations 267

electrode. Thus, the corona-forming sphere appears surrounded by a “fur coat” of

charges whose sign coincides with the sphere sign. If the sphere is at a large
distance from the conductors, then this “fur coat” envelops it uniformly. However,
when the sphere approaches a conductor, the cloud of the space charge moves to the
area between the sphere and the conductor. As a result, the “pillow” weakening
the attraction of the sphere charge to a charge of an opposite sign, induced in a
conductor, is created.
The experiment of George S. Piggott confirmed the subjects’ levitation possibil-
ity owing to corona discharge, as described in [79, P. 192–195]. Piggott observed a
levitation of a silver ball weighing 1.3 g in an electric field of a sphere of 10 cm in
diameter, charged up to a voltage of 500 kV. The distance from the center of the
sphere to the floor was about 50 cm, and the ball was hanging half of this height. In
[172], experiments on levitation observation of corona-forming spheres in the field
of the flat condenser were described. The sphere of 12-cm diameter was made of a
steel microwire of 20-mm thickness, the weight of the sphere was 0.03 g. Vertically
directed electric field was created between two parallel metal disks of 120-cm
diameter, being at the distance of 100 cm from each other. Negative voltage of
20–160 kV was put to the top electrode, and the bottom electrode was grounded. At
voltages above 60 kV, the corona appeared on the sphere surface and it made
fluctuations in a vertical direction, not touching the electrodes, or hung at some
height between the electrodes. At distortion of the field uniformity, one could
observe the sphere aspiration to be drawn into an area of stronger field. On entering
several (two or three) spheres into the condenser, they were lined along one vertical
axis, and connected through the common ion current, made group flight in a
horizontal plane.
In [173], measurements of the weakening force of the flat capacitor plates’
attraction in the presence of the corona discharge in space between them were
carried out. The electric capacitor consisting of two duralumin disks of 12 cm in
diameter was used. The top plate was hung on the shoulder of the scale’s balancing
arm. To other shoulder, the steel rod, which could be drawn in the solenoid was
hung. By measuring the current at which the rod was drawn in the solenoid, it was
possible to define an attraction force of plates of the capacitor. In the top plate, there
were compressed sewing needles (90 or 420 units) coming out of it at a height of 5 or
8 mm. The surface of the plate with needles was faced inside the capacitor. The
distance between the plates varied from 4 to 7 cm, and a voltage on the top plate
changed from 0 to 30 kV. At values of U < 15 kV, the attraction force F of the
capacitor plates increased proportionally to a square of the voltage U2. However,
after reaching the corona discharge ignition conditions (at U ¼ 15 kV), this force
became smaller than those defined by the law F U2. Also, it was revealed that the
attraction force of plates started to increase quickly at U ¼ 25 kV. The ion wind
creating rarefaction under the top plate of the condenser might be the reason for this.
However, after 88 holes were drilled in the plate, this effect was gone (Fig. 6.6). On
the basis of these experiments, the force DF decreasing the attraction of the plates
owing to the presence of the corona discharge DF ¼ zU2 (z ¼ 4  1012 N/V2) was
found. It was assumed that if this dependence remained at potentials U  30 kV;
268 V.L. Bychkov et al.

20

15
F , mN

10

0
0 200 400 600 800
U 2, (kV)2

Fig. 6.6 The dependence of the attraction force F (mN) on a voltage square U2 (kV2) for the plate
with 90 needles and 88 orifices [173]

the results of the experiments allowed to provide an explanation for BL levitation

over a conductor surface at the heights exceeding a value of its several diameters.
In the first part of the review, we discussed a case when BL with a “boiling”
surface from which substance fountains were thrown out was observed [63, 174].
Thus, it could be assumed that the BL surface can possess properties of a liquid, and
dynamics of its surface can be defined by the electric field of a unipolar charge of
BL. For studying the behavior of the surface of the liquids in the electric field, the
authors of [175] have carried out experiments on corona-discharge influence on tap
water, double distilled water, ethyl alcohol, glycerin, kerosene, and a mixture of
alcohol with water and glycerin. The liquid was poured into the flat vessels of the
rectangular or cylindrical form made of a dielectric or steel. More often, the positive
voltage from the rectifier (3–50 kV) was brought through the resistor of 100–300
MO to a liquid through an electrode sealed in a wall of a vessel, and the negative
electrode (cathode) was placed over a liquid surface at a height of 5–15 mm and was
inclined to it by 40 . The diameter of the cathode was 3 mm, which sharpened under
a corner of 30 , and the radius of its rounding tip was 0.4 mm. For a more exact
definition of the moment of breakdown between the electrodes, the capacitor with a
capacity of 500 pF was included. The breakdown moment was detected by the
occurrence of a spark between the cathode and liquid surface, as well as by a sharp
reduction of voltage on the capacitor.
At the ignition of the corona discharge on the surface, there was a funnel of a
rotating liquid, which may be due to the ion wind. When the negative corona
voltage increased, streams started to be thrown out from the funnel edges (see
Fig. 6.7). These streams sometimes turned to the liquid columns whose tips were
shining with a violet light. At the voltage above 15–25 kV, the liquid surface started
6 Ball Lightning Investigations 269

Fig. 6.7 Appearance of bursting streams over a surface of alcohol under a corona discharge [175]

to shake, and there was a breakdown in the formation of the plasma channel
(Fig. 6.7). The authors of [175] explained the occurrence of streams and distur-
bances of a liquid by developing Tonks–Frenkel instability of the charged surface.
A field strength, at which there is instability, is greater for liquids with greater
densities and forces of surface tension; therefore, instability easily develops on an
alcohol surface than on a water and glycerin surface. The bursting of streams
observed in the experiments is connected with the development of charged drops’
instability. Thus, if the BL surface represents the unipolarly charged liquid, then
one can observe streams, columns, and bursting of streams because of the develop-
ment of hydrodynamic instabilities on it.

Intermediate Conclusions and Perspectives of Future Experiments

Apparently, within the last decades, many qualified researchers worked hard to
create laboratory analogs of BL. It has led to the finding of new properties at known
electric discharges and creation of new types of discharges. However, in no case it
was possible to obtain shining object with the properties of natural BL. We presume
that in future work, it is necessary to consider two circumstances. The first – it is
270 V.L. Bychkov et al.

necessary to search and experimentally model only systems that basically consider
the possibility of considerable quantities of energy concentration in small volumes.
It will allow to considerably narrow a circle of searches and to concentrate efforts to
studying and creation of systems with high density of energy. The second – it is
necessary to shift from creation of equilibrium systems (or close to equilibrium) to
essentially nonequilibrium systems capable of self-organizing. Therefore, with the
generation of objects with properties of BL, it is necessary to not only provide
sufficient energy store in them, but also conditions of their ordering (entropy
decrease).
It is necessary to thoroughly analyze cases when objects with properties of BL
appeared as a result of “extraordinary” work of electric installations with certain
parameters. In the first part of our review, we have mentioned two cases of formation
of BL at short-term distortion of contact of a pantograph bar of a trolley-bus with a
contact wire [42, 176]. Trolley-buses use a direct current of about 200 A at a voltage
from 500 to 750 V, and a current collection is carried out by carbonic-graphite
brushes [177]. Reproduction of discharges with the specified range of voltage and
currents is not difficult in laboratories. Moreover, the specified cases indicate that
the threshold value of the current necessary for the creation of BL is much lower
than 156 kA, required to start the process of “bosonization” of electrons [82]. It is
interesting to carry out the analysis, and probably, reproduce the experimental
conditions when the plasma bunch that has got into a vacuumed plasma gun,
initiated an origination of a fireball [178]. It is also interesting to define the physical
nature of “microscopic fireballs,” appearing in the discharges in water [156–161]. It
is necessary to continue research and modeling of traces of BL’s impact on material
objects, in particular, on glasses. It is interesting to understand how holes are formed
with the separation of thin edge on the edge of the hole, and what is the reason for the
change of element structure of the glass coming into contact with BL [76].

6.1.4 Development of Ball Lightning Theory

Usually, development of scientific idea goes under the following scheme: an idea
! a theoretical model ! an experimental check of the idea, etc. With respect to
BL, the first part of this chain  formation of an idea, which in this definite case,
was shown as data gathering about the characteristics of the observable phenome-
non  has developed more or less safely (see Section 6.1.2). Numerous researchers
have worked in this area and are still working. The data obtained by them have been
rechecked, and most importantly, are not lost, but are constantly replenished. In
some measures, it concerns the last link: experiments on realization of “fire balls”
(which is unrelated to BL) are repeated regularly and checked. The middle link of
this chain drops out of an overall picture and stands independently. In most cases,
each new model of BL does not arise as criticism or development of existing
models, but as though from nothing. This, in turn, remains almost unnoticed
(neither critics nor potential followers) and dies along with its author. Besides,
6 Ball Lightning Investigations 271

each model is capable to explain only a limited number of BL properties. Authors,

as a rule, also do not try to cover all the BL properties, leaving the final “finishing”
of the model to the descendants. And, at last, there is another extreme: aspiration to
reduce BL problem to even more difficult unresolved problems (to dark energy,
black holes, an antimatter, etc.), transforming it into a certain place where all
unclear concepts are “merged.”
We will try as much as possible to understand the reasons for such a state, and to
realize whether there is “light in the end of a tunnel,” leading to the creation of BL
theory. Following our scheme, along with other two directions – data gathering of
observations and experiments – let us take a look at the history. We should
remember that because of the above-mentioned continuity absence, the concept
of historicity concerning BL theory works is relative to a considerable extent.
Following our scheme, we will first speak about the theoretical works executed
before 1988 (year of the First Symposium on Ball Lightning). Then, we will get
acquainted with the reports on the theory made on this symposium. After that, we
will look at what has occurred in this area during the next 20 years. Basically, the
time boundary dividing works on the BL theory might not be 1988, but 52 years
earlier, in 1936, when for the first time, an extremely large energy content of this
natural formation was found out experimentally. However, up to 1998, this obser-
vation either remained unnoticed or was announced doubtful [29, 30, 48].

6.1.4.1 Ball Lightning Models Created Before 1988

The review of early theoretical approaches to the BL problem solution can be found
in monographs of Brand [11], Singer [18], and Stenhoff [48]. It is natural that till
“fatal” 1936, BL model suited everything that could explain the existence of the
spherical object radiating light during tens of seconds. Stenhoff [48, P. 179–180]
gave a rather uncomplimentary appraisal to research approaches to create the BL
theory. He considered that prior to the beginning of the work of BL Symposia,
“theoretical proposals were initiated by publication of quite brief and tentative
papers stimulating an occasional letter by way of debate. Some of these papers
could better be described as light-hearted musings rather than disciplined contribu-
tions to discussion.” The majority of the theories contain essentially weak positions.
“The most serious . . . is that many theories do not correlate with observational
evidence. Some theories predict overt behavior by ball lightning that is simply not
reported by eyewitnesses. Equally other theories fail to predict frequently reported
observational characteristics of the natural phenomenon.” Despite the demonstrated
severity, Stenhoff [48], for the first time in BL studies’ history, gave the fullest
review of works on search of approaches for the creation of its theory. Owing to
space constraint, we will present only a short abstract of his analysis.
Similar to Brand, Stenhoff divided BL theories based on the characteristic of
whether they have an internal energy source or get energy from an external source.
The simplest BL model with an internal energy store is a sphere of heated air. For
the sake of justice, it is necessary to remind that comparison of a fireball with
272 V.L. Bychkov et al.

club of air heated by the energy of nuclear explosion has been made by Kapitsa [83]
in 1955. He came to a conclusion that a sphere of air of 20 cm in diameter,
originally heated to a temperature of 10,000 K, should be cooled to a temperature
below 6,000 K during the time of about 102 s. On the basis of this, he concluded
that the existence of BL with an internal energy source is impossible, and it can live
only owing to an energy supply from the outside. From the logic point of view, this
conclusion is incorrect: there was a thesis substitution in it, and a particular aspect
was replaced by the general one. Kapitsa only proved that independent existence of
BL in the form of heated air sphere is impossible.
Uman and Lowke returned to the idea of BL in the form of heated air area, in
1968 [179], and calculated that cooling speed of heated spheres decreases during
the process of their temperature fall. Thus, cooling of a sphere of 20-cm diameter,
originally heated to T ¼ 10,000 K, in a range of temperatures from 3,000 to 2,000
K should occur in 4 s. Thus, the radiating ability of a hot body is proportional to T4.
However, the considered model cannot be connected with BL, not only due to a fast
reduction of its luminosity, but because the cloud of hot gas should necessarily float
in air and cannot have a distinct border.
Bostick [180, 181] experimentally obtained plasma structures (plasmoids) that
were kept by their own magnetic field for a short time (some ms). It served the
beginning of BL plasma model creation, which was appreciably supported by the
development of works on magnetic confinement of hot plasma for thermonuclear
fusion of light elements. V.D. Shafranov theoretically considered the balance
conditions of magneto-plasma structures and demonstrated that spherical plasmoid,
defined as the self-limited clot of electromagnetic energy, cannot exist without the
compressing force directed to its center, for example, gravitation [182, 183]. If the
atmospheric pressure is such a force, then the energy of the plasmoid of 10-cm
diameter cannot exceed several hundreds of joules [184, 185], and according to
the virial theorem [186], this means limiting of the plasmoid energy density by
rE ¼ 2.5  105 J/m3.
Attempts to find a force, which along with atmospheric pressure would promote
confinement of the plasmoid in the limited area of space, have been undertaken.
Wooding [187] assumed that BL can be a vortical plasma ring in which centrifugal
force is counterbalanced by pressure force of the captured magnetic field.
Bergström [188] stated that plasmoids from expansion, except magnetic forces,
can keep short-range forces of dielectrics’ attraction to diamagnetics.
Another mechanism that can promote the creation of forces keeping from the
scattering system of charges is the quantum mechanical exchange forces. Neugebauer
[189, 190] demonstrated that such forces can arise in neutral plasma if the density of
the electrons in it is high (not less than ne ¼ 2.7  1025 m3), and its temperature Tex is
low (not above 600 K). Considering that for electron separation from an atom, it is
necessary to spend about of 14 eV energy, Neugebauer found that in such a disordered
system, the density of the internal energy will be equal to rE ¼ 6.7  107 J/m3.
Neugebauer’s model was developed by G.C. Dijkhuis, who replaced it with the
ordered system of vortical tubes of quantum mechanically connected electron pairs,
6 Ball Lightning Investigations 273

around which positive ions rotate [82, 191, 192] (we will discuss the Dijkhuis model in
more detail later).
The search for the conditions promoting the preservation of the plasma forma-
tion’s spherical form and restraint of its expansion, as well as the fact that BL was
often seen rotating, has guided some researchers to a thought that it can be a
whirlwind of the spherical form (a whirlwind of Hill [193]). This whirlwind of
ionized air can arise because of a gas stream in the channel of a linear lightning if
there are opposite-directed channels [37, 194] or the channels have a sharp bend
[195]. Some models where the fireball was considered as “a bubble filled with
radiation” – a cover of highly ionized gas, in which strong superficial current flows
were proposed. This cover keeps in itself the powerful resonant high-frequency
radiation reflected from walls of a cover [196]. The pressure of radiation upon a
cover wall is counterbalanced by the atmospheric pressure; thus, the energy of
radiation for a sphere of 10 cm in radius does not exceed 400 J (rE ¼ 105 J/m3).
Endean [38], independent of Dawson’s finding [196], proposed a model according
to which BL consists of electromagnetic field enclosed in a vacuum spherical
cavity, separated from the atmosphere by a cover of the ionized gas. According
to him, the stability of the cover is provided by the quickly rotating magnetic field.
Unlike the above-mentioned models, where it is assumed that BL in the course
of its life only spends the energy obtained by it at the moment of formation, in other
models, it is supposed that it obtains the energy continuously from chemical
reaction. As mentioned earlier, the first who admitted that such energy source
existence is possible in BL, was Lomonosov [9]. In 1838, Arago published an
article [10] in which the BL-energy source was considered to be the burning of the
substances that appeared in air under the influence of the linear lightning discharge.
Later, some power-intensive compounds were proposed: nitrogen oxide (NO)
[197], ozone (O3) [198], and a mixture of hydrogen with oxygen [199–202] or
with nitrogen oxides [203]. It was also considered that organic substances and
gases, which are constantly present in air – as the products of rotting of the
vegetative remnants and methane coming out of the earth [35, 204, 205] – could
be the fuel delivering BL energy. It was supposed that condensation of the charged
particles of dust, droplets, and aerosols can result in BL formation [206–209].
Estimations show that none of the above-mentioned models could explain the
origin of high specific energy content of “Dorstone” BL [13, 14] (rE ¼ 1010 J/m3).
Alternatively, models with a nuclear energy source, which could explain such high
density of energy were proposed. An assumption was made that under the influence
of strong electric fields in the channel of a linear lightning, there can be an accelera-
tion of the charged particles and formation of neutrons. Neutrons, reacting with
nitrogen nuclei 14N, can form nuclei of radioactive carbon 14C with a half-life decay
period of T1/2 ¼ 5,570 years [210]. Fast protons (1 MeV), facing oxygen nuclei 16O
or nitrogen 14N, can form radioactive isotopes 17F (T1/2 ¼ 66 s) and oxygen 15O
(T1/2 ¼ 124 s), which decay with the emission of positrons [22]. The model of BL
with a nuclear energy source, basically, is capable to explain its high energy content
and comprehensible lifetime, but is ineffective to explain something about the
274 V.L. Bychkov et al.

electric effects caused by it and about the features of its movement. Besides, such
fireball should be a source of high radiation that took place only in several cases [40].
The authors of other models (also low energy ones) considered that BL energy is
reserved in electric field of the separated charges. For the first time, an idea that BL
is the natural electric capacitor (Leyden jar) was stated by Tessan [211]. Other
researchers considered BL as the sphere filled with positively and negatively
charged droplets of water, separated by an insulator – water steam [212]. In some
models, molecular ions were considered as carriers of charges, and BL was consid-
ered to be similar to “a tiny storm cloud” [213, 214]. Stakhanov assumed that
recombination of ions of a different sign can be suppressed, if ions are surrounded
by covers of water molecules [29–31, 215].
For models with an external energy source, there are no difficulties connected
with the virial theorem (which are caused by necessity to constantly store the large
amount of energy in BL). Such BL should have the source located outside, from
which they obtain energy for the maintenance of their luminescence. However,
these models experience difficulties in explaining the cases of allocation of consid-
erable quantities of energy at BL explosions, the electric effects, and trivial proper-
ties such as fireball movement (especially – in a vertical direction) and its
occurrence in the closed premises.
The energy supply to BL from an external source can occur in the form of
constant or alternating current. In the model developed by Finkelstein and Rubinstein
[184], it was supposed that BL is the localized glow discharge fed by the currents
flowing in the atmosphere. According to this model [185], BL appears as the rest of a
linear lightning discharge or St. Elmo’s fire. If the intensity of the electric field near the
channel after the termination of the lightning discharge exceeds the value of 100 kV/m
in the area of ionized gas, then the Tawnsend process of reproduction of charges of a
different sign will occur, which, under the influence of a field, will flow down from a
sphere in the opposite direction. Because of the electrons’ higher mobility, the sphere
will be charged positively. After that, the cloud of the positive space charge, which
flows down from the sphere, will start to contain electrons in it from the space area,
which is much larger than the sizes of the cloud. This model describes the fireball
passage through the glass as follows. As electric field lines of a cloud charge penetrate
the glass, the charges of a different sign gather on the opposite side of the glass. In
the BL approach, the glass is heated up, becomes a conductor, and is broken down.
The current passes through the glass, and a sphere, having separated off the glass, is
formed indoors.
Kapitsa [83] proposed the BL concept as high-frequency electric discharge in
air, which arises in the field of a standing electromagnetic wave. Based on Kapitsa’s
estimations, BL diameter D and a wavelength l of radio-frequency radiation
feeding it are connected by the relation l ¼ 3.65·D, and the discharge appears
in a loop of a standing wave. Starting from a range of a typical BL size of
(D ¼ 10–50 cm), an area of radio frequencies f can be defined as f ¼ 160–820
MHz. Later, it was shown that BL, having arisen in a loop of wave, should move to
wave knot under the action of pondermotive forces, where it is maintained owing to
the balance of radiating and atmospheric pressures [216, 217]. The basic difficulty
6 Ball Lightning Investigations 275

for a substantiation of Kapitsa’s theory is the absence of any appreciable continuous

radio-frequency radiation in the indicated wavelength’s range (from 37 to 183 cm),
and investigations have given a negative result.
All the above-mentioned models has appeared that scientific reserve in develop-
ment of BL theory by the beginning of the First International Symposium on a Ball
Lightning in 1988. Now, we will describe the works on BL theory discussed at this
symposium.

6.1.4.2 Works on Ball Lightning Theory Represented at the First

Symposium on Ball Lightning

Earlier, we presented reports in which the results of BL observation were discussed

and we observed large consistency and reproducibility of the obtained data. On this
background, works on the theoretical research on a BL problem looked inadequate
at the symposium. Apparently, as stated in the previous section, instead of the
allocation of key BL properties and work on “quality” increase (deep elaboration of
models), the activity, basically, has poured out in “quantitative” increase in a
number of models describing only separate features of the phenomenon. Some
researchers employed a combination of different models. A typical example of such
association is the report by K. Nickel (Germany) at the First Symposium [1,
P. 177–191]. He considered that BL represents a sphere of rotating air, gas, or
plasma, which is kept by the hydrodynamic forces operating in Hill’s whirlwind or
in a smoke ring of the smoker. Its luminescence and heat allocation is made by
burning chemical process of the gases captured by a sphere. BL creation is initiated
by the blow of a usual lightning, which transfers the moment, kinetic energy, and
electric charge to BL. Nickel’s model is presented in traditional for BL theory
developers manner – at the level of declarations that have not been confirmed even
with simple estimations after completely ignoring the properties of the real BL
obtained by the analysis of thousands of observations. Hence, with regard to BL
charge, he estimated its potential U as 106 V and more, without explaining how this
potential was retained by this system. It is known that the electric capacity C of a
sphere D ¼ 0.2 m in diameter is C ¼ 2 pe0D ¼ 1011 F, and its charge is Q ¼
C
U ¼ 105 C. The energy of the sphere’s electric field is WE ¼ CU2/2 ¼ 5 J,
which is obviously insufficient to cause appreciable actions – fusion of wires or
other destructions. An extremely unpersuasive reasoning is that BL, for mainte-
nance of burning in itself, can collect fuel from car exhaust gases.
B.M. Smirnov (USSR) [1, P. 192–228] considered that the BL problem “cannot
be solved on the basis of successful idea” . . . “Because of complexity of this
multilateral phenomenon in order to become successful, it is necessary to divide a
problem into separate elements and to investigate each of these elements.” The main
BL problems, in his opinion, are the structure of BL substance; processes of storage
and energy transformation; chemical, thermal, gas-dynamic, electric processes, and
radiation mechanism. As an example of such an approach to a problem, he consid-
ered BL phenomenological model with substance comprising separate threads or,
276 V.L. Bychkov et al.

more precisely, fractal clusters. The waves of the chemical reaction propagate along
the threads, accompanied by a luminescence of hot zones. These waves can be
transferred to the subsequent threads. This chemically active substance is similar to
a pyrotechnic mixture. BL possesses the firm framework similar to the aerogel; the
specific density of this framework is approximately equal to the specific density of
air. The active combustible substance is located in the pores of this framework.
Owing to the active substance’s chemical reaction, there is a release of heat and
heating of air in BL to a temperature of some thousands of degrees. As a result, the
BL shines as a black body, and its color is defined by impurities, giving color to the
burning of pyrotechnic mixture. The framework bears an electric charge on itself,
which is about 106 C, which provides its strength, and vibrations of the ends of the
charged threads are a radio-radiation source.
There is nothing reprehensible in the represented approach: everyone
has the right to propose an idea and to build the BL model on the basis of this
idea. However, it is necessary that conclusions of this model coincide with the
properties of a real BL. Unfortunately, it is impossible to imply this on fractal BL
model. Its charge is very small to be dangerous to the person, whereas there are
many cases when the opposite was observed. The energy stored in such BL is
chemical energy of several grams of solid fuel; therefore, it cannot cause the
destructions left by BL and is substantially less than 20 kJ, which are allowed for
an “average” BL.
A detailed description of Smirnov’s idea has been represented in the book
[146]. In an explanation of the nature of BL’s electric action, he is in agreement
with Stakhanov’s explanation. He considered that “in these cases electric energy
is taken from an external source (a cloud or a dust cloud), and BL is a conducting
body which creates initial ionization of air and transfers energy of the external
source” [218]. From our point of view, it is impossible to imagine that the
ephemeral object with a charge of 8  107 C and energy which “is possible
to provide approximately by 10 matches” is capable to impact the linear
lightning, transferring a charge of tens of Coulomb and energy of thousands of
megajoules.
H. Jamamoto’s report [1, P. 229–241] at the symposium is a fresh sight at BL
problem of the person working in another area of physics which has gotten
acquainted with the discussed problem only recently. He had an impression that
based on the observation, it is possible to draw a conclusion that there are many
types of fireball. Therefore, in his opinion, “it is impossible to explain all proper-
ties of BL within the limits of one model.” He also considered that a statistical
analysis of observation results could lead to an exclusion of the important infor-
mation of consideration. Therefore, he considered that it is much more useful to
publish original reports of witnesses, and leave the right to make conclusions to
the readers. Personally, he proposed a BL model in the form of a combination of a
ring electron current and electrons, moving on a surface of the torus whose axis is
the above-named ring. He considered that the full charge of the system is equal to
zero, indicating that there are also positive charges besides electrons. However,
how they move, how collisions of particles with air are eliminated, and how
6 Ball Lightning Investigations 277

recombination charges of a different sign is prevented, have not been discussed in

the article.
G.P. Gladyshev (USSR) [1, P. 242–253] considered that high-temperature
physical and chemical processes serve as BL-energy source in thunderstorm atmo-
sphere. Under the influence of lightning discharges in it, combustible components,
such as nitrogen oxide and ozone, are formed, which can react with the oxygen of
air with heat release. BL is formed as follows. Under the influence of the lightning
discharge in some part of space, the concentration of ions increases and the
temperature raises, indicating that conditions for a flame ignition are created.
This flame takes the form of a sphere and continues to be supported by the direct
currents flowing in the atmospheric air. Necessity of their presence is caused by the
fact that heat of oxidation reaction of strongly diluted combustible gases does not
suffice for the maintenance of temperature of a zone of reaction even at a level of
600–700 K. Thus, it appears impossible to create a purely “chemical” model of the
phenomenon without the presence of an external energy source. External currents
heat up a sphere to temperatures at which burning and synthesis of nitrogen oxide
(NO) begins. Heat of the reaction is transferred to an environment, and inside BL,
a quasi-stationary distribution of temperatures is established. The author considered
that “reduction of atmospheric currents force leads to going out of BL, and, on the
contrary, at critical values of the discharge (during a thunderstorm), there can be an
explosion or the discharge of a linear lightning (spark, arc) to huge release of energy.
It is obvious that though the BL has a rather small store of internal energy, it can
initiate high-energy discharges in atmosphere.” Again, for some reason, without any
substantiation, properties are imposed to the nature, which it never shows! The
author recognized that the part of the concepts, which have been employed in its
model, was proposed earlier by other researchers. However, nobody had proposed
any physical and chemical mechanism of ion formation and maintenance of the
level of charge concentration in BL.
In 1955, Kapitsa proposed a model of BL fed by radio-frequency radiation of a
linear lightning discharge [83]. As discussed earlier, the basic problem of this
model is the search of the energy source supporting BL during the time between
the two lightning strokes. However, P. Handel (USA) [1, P. 254–259] found this
source. In his opinion, under the influence of the strong electric-field short pulse
generated by the linear lightning discharge, a population inversion between the
pairs of rotation energy levels of water molecules takes place. It leads to the
formation of an atmospheric maser with a huge volume of active medium –
102–109 m3. Owing to the maser radiation under a cloud near the earth’s surface,
the standing electromagnetic wave is formed. If at some point, there is a coinci-
dence of local plasma frequency of ions and the maser frequency, then it will cause
local increase in the intensity of electric field. It, in turn, will create pondermotive
forces, which will “pump out” ions from the area with a strong electromagnetic
field, thus creating a cavity in the plasma. It will result in the formation of the so-
called “caviton” – a practically empty sphere containing only a resonant electro-
magnetic field and ions. Thus, “the caviton is a stable quasi-spherical configuration
of the captured electromagnetic field surrounded with plasma. It moves horizontally
278 V.L. Bychkov et al.

at a loop level of a standing wave and manifests, as BL.” Handel represented the
equations describing the process of population inversion creation in the two-level
system under the influence of resonant high-frequency radiation pulse; however, for
some reason, there are no terms in them, which are considered to be responsible for
the relaxation of excited water molecules at collisions with the molecules of air.
The equations describing the caviton formation in the atmosphere were also
represented.
According to H. Kikuchi (Japan): [1, P. 288], now the theory of a lightning and
electric discharges is at an initial stage and requires development. For its develop-
ment, the following are necessary: (1) Definition of more exact laws of electrostatics
and electromagnetism; (2) Working out of the new theory of plasma in which the
dominating role is played by collisions; and (3) Creation of a new scientific area – of
electro-magneto-hydrodynamics (EMHD), which combines the physics of plasma
and meteorology. According to this approach, for the solution of a BL problem, it
meaningful sense to bring three new concepts into the theory of the electric dis-
charge: (1) process of reconnection of the electric field lines, similar to the phenom-
enon of reconnection of the magnetic field lines, leading to solar flashes; (2) the
theory of critical speed for gases, in which the dominating role is played by the
processes of collisions, similar to Alfven critical speed; and (3) the account for
pondermotive forces.
P.M. Koloc (USA) [1, P. 289–309] considered that BL represents the steady
plasma configuration, which he termed as the “Plasmak”. “Plasmak” grows out
of evolutionary development of hot plasma magnetic confinement systems, and
Tokamak was the first to be developed [219, 220]. In the Tokamak, the toroidal
chamber, filled with plasma, represents a secondary winding of the transformer.
Under the influence of a vortical electric field on an axis of the toroid chamber, the
current flows through the chamber, which creates a “poloidal” magnetic field (in
which field line flows are similar to that wound on the torus generatrix). For plasma
confinement on a toroid axis in Tokamak, the “toroidal”magnetic field created by a
current through a wire wound on the toroidal vacuum chamber is used. Lines of this
magnetic field go on an axis of the torus that remains along the current lines in
plasma. As a result, the configuration of “magnetic fields, free from forces” appears.
It is possible to replace the toroidal metal chamber with the spherical one for the
creation of the magnetic field, and by keeping the charges on an axis of the toroidal
area of plasma, it is possible to use the vortical currents flowing in the walls of the
chamber. Such a configuration has been named as “Spheromak” [221]. In the
“Plasmak” of Koloc, the metal spherical chamber is replaced with a cover of plasma
[222]. According to Koloc, BL consists of a core (Kernel) – a toroidal ring of hot
magnetized plasma and a cover (Mantle) – a shell in the form of a plasma layer with
high conductivity. It is considered that the current in a shell is supported by
relativistic electrons. Magnetic lines of the field, created by the current flowing
on a ring axis, go to infinity on a ring axis. Therefore, on the shell’s “poles,” hot
electrons can leave the system, and it should have ledges on the poles. To confirm-
ing his conclusion, Koloc used a photo of the elliptic BL (see Fig. 6.1). As
mentioned earlier, elongation of the image in a photo, most likely, has been caused
6 Ball Lightning Investigations 279

by BL movement at shooting. The typical form of BL, according to a overwhelming

number of supervisions, is the ideal sphere. Estimating the energy that BL can
contain in the form of “Plasmak”, Koloc addressed the case of “lightning in a tub”
[13] and came to the conclusion that his model can explain it if its energy release
did not exceed 1 MJ and its diameter was about 80 cm (however, it is not clear
how BL of 80 cm in diameter and volume of 270 dm3 could get into the tub of
nearby 20 dm3 in volume).

6.1.4.3 Ball Lightning Theory Development After 1988

Now, we will discuss what has changed in the field of work on the creation of the
BL theory in 20 years, after the first symposium. For the purpose of analysis, we
will use the material of the theoretical articles presented at the subsequent symposia
and also the reviews on BL models [223–226].

Ball Lightning Models with External Source of Energy

During the considered 20 years, Kapitsa’s idea was developed according to which
BL is the discharge in air, which obtains energy from an external source. Lowke
[227] assumed that the electric field of the currents spreading in the earth from a
place where a linear lightning struck can be this source. It can explain the observa-
tion cases of horizontal BL movement along the earth’s surface, but, unfortunately,
can explain nothing about the reasons for the other types of BL motion.
Handel [228–230] worked on the development of the concept of atmospheric
maser on water molecules (see [1, P. 254–259]). The existence of maser radiation in
the nature, in principle, could solve the problem of maintenance of discharge burning
in the air during the interval between two lightning strokes. According to Handel,
water molecules serve as the keepers of energy during the period between lightning
discharges. In the system of close rotation levels of water molecules, the inversion of
population occurs under the influence of the electric-field pulse. As the radiative
transition between these levels is forbidden, a lifetime of a molecule at the excited
level can reach several tens of seconds. “In absence of the high sharpened conductors
equipotential surfaces are flat and the population inversion can be formed in a volume
of many cubic kilometers. When the electric field increases to a value of 104 V/m,
some of the forbidden transitions become weakly allowed and start to feed a huge
atmospheric maser which volume exceeds the critical. The volume is considered big
enough if the number of photons, created in unit of time due to the induced radiation,
is greater than a speed of their losses which is mainly caused by the absence of the
resonator [1, P. 254–259].” According to Handel, the population inversion in a
system of two singled-out rotation levels of the water molecule W1 and W2 is realized
owing to the action of electromagnetic radiation pulse at the transition frequency
o12 ¼ ðW2  W1 Þ= h: This method of pulse pumping is usually used in quantum
280 V.L. Bychkov et al.

paramagnetic amplifiers [231, 232]. Under the influence of a high-frequency field,

the population of rotation levels periodically change with a frequency [233, 234]:

2O ¼ ½ðo  o21 Þ2 þ ðm21 E0 =hÞ2 1=2 : (6.1)

Here, o is a circular frequency of a high-frequency field, E0 is the amplitude of

electric-field strength, and m12 is a matrix element of a transition between the levels
W1 and W2. At o ¼ o21, population densities of the levels oscillate with the Rabi
frequency:

O0 ¼ m21 E0 =ð2
hÞ: (6.2)

If one impacts a medium by a pulse of resonant high-frequency radiation of

duration Dtp ¼ p/O0 (by so-called p-pulse), the population densities of the upper
and lower energy levels exchange places. As in the conditions of thermal balance, the
population of the lower level is greater than those of the upper level by
expðho21 =kB TÞ times, and in the system of levels, the population inversion arises,
that is, the number of molecules capable to radiate energy will be greater than the
number of molecules capable to absorb it (here, kB is Boltzmann constant). However,
this process is possible only in the case when the pulse action time Dtp is much
smaller than the relaxation time of the upper level (or, in other words, the time of
thermal balance establishment in the system of considered levels). Under the condi-
tions of terrestrial atmosphere, a main factor of such balance establishment is the
collisions of water molecules with air molecules. It is known that the rotational
relaxation occurs quickly enough, with the probability close to unity during one to ten
collisions [235]. At atmospheric pressure and room temperature, the average free
path of water molecule is 6  106 cm, and the average speed is v ¼ 6  104 cm/s.
Thus, the molecule will make 1010 collisions during a second. If 100 collisions are
necessary for the rotational relaxation realization, then the time of balance establish-
ment will be teq ¼ 108 s. This time is much smaller than those of a linear lightning
discharge of tltn ¼ 104 s [236], during which radiation, resonant with the frequency
of the transition o21 between the rotational levels, can be generated. Moreover, even
if it is possible to create the population inversion by a large duration pulse (Dtnp ¼
np/O0), it will be destroyed in time teq [231]. Thus, in terrestrial atmosphere, the idea
of energy accumulation in the form of molecules’ rotation energy cannot be realized
in spite of even large rotational transitions’ radiating lifetime.
As demonstrated by Wieder [237], maser existence on water molecules at
atmospheric pressure and BL feeding by it are also impossible owing to other
reasons. Among the rotational transitions in water molecule, transitions with the
frequencies being in the area of “Kapitsa’s radiation” from 429 to 857 MHz have
not been found. The longest wave transition found in the radiation of space masers
on water molecules was the rotational transition 616 ! 523 with a wavelength of
lM ¼ 1.35 cm (f ¼ 22.235 GHz) [238]. According to Kapitsa’s criterion, powerful
high-frequency radiation with such a wavelength can form a plasma bunch in the
6 Ball Lightning Investigations 281

size D ¼ lM/3.65 ¼ 3.7 mm. However, this turned out to be much smaller than the
observed BL. Nevertheless, though there are lines with wavelengths from 35 to 70
cm in a rotational spectrum of water molecule, the width of these lines will be about
7,000 MHz, because of collisions of water molecules with air molecules, that is, it
will be greater than the distance between the hypothetical energy levels (400–900
MHz). In this case, both levels become indistinguishable.
Maser of Handel, has no a resonator with the walls effectively reflecting a radio-
wave radiation. Hence, it had to radiate isotropically [239]. Therefore, it is not clear
what could have caused the radiation focusing in a small area of space. The
statement of Handel that the functioning of atmospheric maser and, hence, BL
formation is impossible in the conditions of a mountainous terrain, is also not clear
[229]. Actually, in the literature, there are reports that fireballs were repeatedly
observed in mountains [18, P. 108].
Despite criticism, Handel continued to work to provide an explanation to BL
properties on the basis of his model. In the reports presented at the Eighth Interna-
tional Symposium on BL, Handel gave an explanation of two features of BL
movement based on his model. In [240], he tried to find the reason for why a sphere
of air heated by the discharge did not rise upwards owing to convection and
remained at place. Handel explained this stating that maser does not continuously
radiate energy, but in the form of often repeating pulses (spikes). Because of this,
the discharge often dies away and again lights up within a second at the same place.
Handel explained the movement of BL along the earth surface using the difference
in the frequencies of two running waves feeding BL [241]. As a result, the focus
place of a resultant standing wave slowly moves, which is perceived by the observer
as BL movement. At the Ninth Symposium on BL, Handel proposed another reason
for BL movement: phase shift of the waves feeding the discharge [242]. This shift
can arise because of an asymmetrical distribution in the space of inversion density
of water molecules population levels. As the inversion density experiences chaotic
variations, BL movement can also be chaotic.
It can be seen that for Handel’s model, a major difficulty not only represents the
explanation of the reasons for the continuous feeding of the discharge, but also a
trivial BL property such as its movement. Nevertheless, his idea of “caviton” (an
area with the lowered pressure of the gas, surrounded by plasma cover) formation in
the air under the influence of high-frequency radiation can appear useful in explain-
ing “material” BL passage processes through the glass.
Stakhanov considered the models of BL in the form of electric discharge
obtaining feeding from an external source, as contradicting the reality [29–31].
He absolutely and fairly wrote that Kapitsa’s and Finkelstein’s models [83, 184]
cannot provide answer to the simple questions connected with BL behavior. Hence,
the mechanism of BL moving in space (more precisely, places where there is an
allocation of energy of an external source) is not clear. Furthermore, it is not clear,
why BL appears single, instead of many plasma bunches in the neighboring loops of
standing electromagnetic wave. The intensity of the radio-radiation generated at the
linear lightning discharge is, by many orders of magnitude, smaller than those
which are required for discharge maintenance in the air, and the basic frequency of
282 V.L. Bychkov et al.

this radio-radiation ( 50 kHz) is by 104 times smaller than the frequency of

radiation that presumably feeds BL (500 MHz). It is not clear how it manages to
remain in the conditions of strong convection, which takes heat away from it
released at the high-frequency discharge.
Nevertheless, researches on BL model with the external energy supply continue,
and an authority of P.L. Kapitsa, the Nobel Prize winner, has played a great role in
this field.

Ball Lightning Models Without Feeding from External Energy Source

Certain complicated questions stand in front of models with an internal energy store.
Hence, according to Stakhanov, BL cannot be a plasmoid, that is, the high-tempera-
ture plasma held by its own magnetic field [31, P. 175]. In fact, the systems held by
their own magnetic field can be steady only in the presence of additional compressing
forces (external pressure or gravitation) [182, 183]. If the atmospheric pressure plays
a role of constraining force, then according to the virial theorem [184], energy of the
plasmoid Wpl cannot exceed a value 3PaV (where Pa – atmospheric pressure,
Pa ¼ 105 Pa, and V – a volume of the plasmoid). For V ¼ 103 m3, energy is Wpl
 300 J, which is much smaller than the values of BL energy found on the basis of its
action on various objects (water heating, pile splitting, etc.). Hot plasma will quickly
lose its energy because of electrons braking radiation. Somehow, to reduce the speed
of this loss, the plasma temperature should be below 107 K [184]. BL cannot exist in
the form of a low-temperature plasma volume at atmospheric pressure, as the atomic
ions transform to molecular ions during 108 s, which, in turn, recombine for a
thousand part of a second [31, P. 177; 146, P. 47].
The properties of BL cannot be explained by the model based on the assumption
that it represents a mixture of neutral molecules capable to react with each other
chemically [31, P. 181]. First, it is impossible to explain the maintenance of the
spherical form of BL, if the presence of large uncompensated charge of BL
substance should be excluded. Second, BL with a chemical store of energy cannot
provide energy storage of some tens kilojoules in a liter. And, third, the chemical
model of BL (even supposing the presence of a small charge of 107 C in it) is
incapable to explain the electric effects connected with it.
In addition, the system consisting of charged dust particles [31, P. 184] cannot
provide a comprehensible energy store. According to estimations, the energy
density, stored by such system, cannot exceed 1 J/m3. The substance of BL cannot
contain free electrons. Owing to small mass, the acceleration of an electron is 105
times greater than those of an ion, and the radiated power by it is greater by 1010
times. This has led to the conclusion that the radiation power of BL of 20 cm in
diameter will be equal to 10 MW, which will release its energy for 103 s [31,
P. 188]. BL cannot exist during adequately large time, in which energy is stored in
atoms or molecules excited to metastable electronic levels. In spite of the fact that
the lifetime with respect to radiation of such atoms and molecules can reach
6 Ball Lightning Investigations 283

hundreds of seconds, the time of their relaxation in collisions with air molecules is
only 108–102 s [31, P. 49].
However, the energy of BL cannot be stored for a long time in the form of the
ions surrounded with covers of polarized water molecules (this assumption con-
siders the basic idea of Stakhanov model [31]). As shown in [146, P. 47], for
recombination processes of unlike charged ions in air, the parameter et, represent-
ing the product of density of the stored energy and the characteristic time of a
process, shifts from 1012 to 1011 (J/cm3) s. If we assume that charges are located
on aerosol particles, it is possible to admit that the speed of the aerosol particles
approaching, caused by their Coulomb attraction, will be slowed down by forces of
friction in the gas. Therefore, the parameter et 107 ro/q2 for aerosol plasma,
where ro is the average radius of an aerosol particle (in a mm) and q is its charge in
terms of an electron charge. Having put ro ¼ 1 m and q ¼ 1, we can observe that et
107 (J/cm3) s [146, P. 47]. Thus, the time of aerosol plasma existence with the
energy density of 5 J/cm3 will be only t 2  108 s.
On the basis of this not very comforting analysis made by Stakhanov and
Smirnov, we can conclude that the overwhelming majority of hypotheses of BL
structure proposed in 90th years of 20th century cannot accurately describe its
properties. In this connection, within the last 20 years, attempts to eliminate the
contradiction in the “traditional” models of BL, were undertaken.

Models with Chemical Source of Internal Energy

Experiments with the burning of methane–air mixtures have shown that at a
concentration of methane of about 1.5%, sometimes balls with the diffuse borders,
bearing a faint resemblance to the BL form, were observed [35, P. 166]. In the
nature, such processes, in principle, could take place with the release of significant
amounts of methane from bogs or earth-crust breaks. However, the natural mainte-
nance of methane in the air does not exceed 104%, which explains that burning of
BL by methane is improbable.
The assumption that energy-containing reagents are labile substances, created in
the course of storm activity (ozone, nitrogen oxides), is more natural [243, 244].
According to Smirnov [244], the concentration of ozone in BL volume can be about
of 1%; this can provide it with an energy of 100 J. This energy can support its
luminescence within 5 s. However, in [243], an attempt was made to estimate the
value of the force connecting BL together, which showed that the interaction of
particles inside it (dipoles and charges) cannot create surface tension, sufficient for
the maintenance of BL spherical form.
Eventually, Smirnov’s model evolved, and BL was considered to be the light
aerosol framework containing a fuel (coal, stearin, and ozone) and a small amount
of charges of one polarity in its pores [146, 218]. The structure described in [218] is
similar to the aerosol model of BL proposed in [208], according to which, it consists
of “the volumetric threadlike structures suspended in air”. The authors of [208]
considered that the particles with which BL is formed are gathered by the electric
field from a large volume in the small one. “If . . . particles possess excess internal
284 V.L. Bychkov et al.

energy it means . . . an increase of such energy density in the volume.” Nevertheless,

the density of chemical energy in an aerosol BL appears very small (about 1 J/cm3).
This allowed Smirnov to figuratively compare its energy with the energy of ten
matches [218].
Absence of electric charge presence necessity in chemical BL model (or its
auxiliary role – only for stabilization of its skeleton [218]) reduces a merit of this
model. In fact, electric manifestations of BL interaction with various objects are
one of its basic features. Therefore, one can understand the aspirations of some
authors to connect together chemistry and its electric properties.
For example, B.B. Kadomtsev considered BL as the self-organized mixture of
chemically active dust particles with plasma ions – a certain isotropic accumulator
of electric energy [245]. The authors of [246, 247] considered that in the volume of
BL consisting of an aerosol and dust, chemical reactions take place in which the
charged particles (electrons and ions) are formed. If there is an electric field, then
the originated electrons will move in it, colliding with the polarized extended
particles of the dust. These dusty particles will try to be oriented so that their dipole
moment is directed along the lines of a field. At collision of an electron with the end
of the particle charged negatively, it will bend around a particle, and, on the
contrary, if the particle’s end is charged positively, it will collide with it. Therefore,
the impact cross-section of an electron with the particle is different in these cases. In
the space between the cathode and anode, the particles try to be guided, so that their
negatively charged end is turned to the anode. Therefore, the electrons, moving
from the anode to the cathode (in the direction opposite to the direction of a
conductivity current), will less often encounter with the dusty particles, than the
electrons moving from the cathode to the anode. Owing to this, in space between
the electrodes, the current from the cathode to the anode will flow, charging the
electrodes and increasing the field strength between them. The energy for current
maintenance is drawn from the kinetic energy of the electrons, which, in turn,
obtain it from the energy of the chemical reaction. Owing to the described
process of “negative conductivity,” electric-field strength can increase to
Ech ¼ 6  105 V/m. However, the electric energy density in the volume of such
“chemical” BL appears equal to rch ¼ eoEch2/2 ¼ 1.6 J/m3, and it is obviously not
enough for an explanation of conductors’ fusion processes and other manifestations
of electric charges present in BL.
Summing up, it is necessary to agree with Stakhanov’s opinion that BL with the
chemical energy source, having a density equal to the air density, can neither
explain its source of high-specific energy content nor the reasons for the electric
phenomena caused by it [31].

Cluster Model of Ball Lightning

This model was proposed by Stakhanov [29, 31, 248]. According to this model, the
substance of BL represents the nonideal plasma, consisting of positive and negative
ions with the big covers of polar neutral molecules. In hydrated ions or hydrates, a
cluster cover, from three to five molecules of water, has a binding energy of about
6 Ball Lightning Investigations 285

4 eV, and it should not be destroyed by collisions at temperatures below 1,000 K.

Such a cover can prevent recombination. However, cases when hydrated ions’
recombination speed decreases to the values necessary for the explanation of
BL’s long existence, have not been revealed.
In the model developed by Stakhanov [29, 31], the question on the mechanism of
the origination of the surface tension interfering with the mixing of a cluster ion
cloud with air was left open. Turner [249] tried to find the answer this question. He
considered that BL consists of the central area filled with plasma, which is
surrounded by three concentric spherical zones. In the hot central area, there is a
formation of ions and metastable molecules, which, leaving this zone, go to the
external layers saturated with cold humid air. Here, they become covered with a fur
coat of water molecules, and heat release takes place. Strongly hydrated ions move
outside and a cluster recombination reaction takes place in an external layer. The
most probable cluster types are the clusters formed by the hydrated proton H3 Oþ
and ion NO 2 : Nitric acid is formed at their recombination:

H3 Oþ
nH2 O þ NO
2
nH2 O ¼ HNO2 þ ð2n þ 1ÞH2 O: (6.3)

Reaction (6.3) is endothermic and should lead to cooling of an adjoining layer of

air that according to [249], will cause the appearance of an air stream directed inside
the BL. This stream will create the pressure compressing an area where there are
chemical reactions, and simultaneously deliver reagents for their feeding, that is,
BL will work as the thermo-chemical thermal pump.
Turner’s model, to some extent, is a development of Stakhanov’s idea: it allows
finding the reasons of the occurrence of surface tension and spatial separation of
areas of clusters formation and recombination. However, together with this, it
shows a number of difficult compatible requirements for the considered system
(a chemical reactor). First, the model assumes the presence of the substance’s
constant movement from the center to the periphery. Then, to fill in the substance
losses in the center, it is necessary that there are gas streams moving in an opposite
direction  from the periphery to the center. It should break the central symmetry of
the system. Second, the generation and maintenance of high-temperature plasma
mechanism in the central zone is not clear. Turner assumed that the formation of
metastable molecules with the subsequent spontaneous ionization of the latter can
be the underlying mechanism that, however, does not remove the question on an
energy source necessary for this process. Third, for the creation of the force
compressing Turner’s chemical reactor, it is necessary that not only temperature
but also pressure has to be decreased in the external layer. As to the latter, the
answer remains undetermined, because in the course of the chemical reaction (6.3),
there is a growth in particle number, and hence, pressure. And, fourth, for an
explanation of BL passage through the intact glass cases, Turner supposed that
cluster molecules and ions can generate electromagnetic radiation, which can be
absorbed by similar types of molecules on the glass’ opposite side; however, he did
not describe the mechanism of this radiation occurrence.
286 V.L. Bychkov et al.

In [250], an attempt to create a BL model by combining the models of various

authors, was undertaken. This model known as condensational, includes Stakhanov’s
ideas about the cluster nature of BL [29], Smirnov’s idea that these clusters form
fractal structures [146], Bychkov’s idea that BL structure includes polymeric threads
[251], and Stepanov’s, who proposed the mechanism of electric-field generation in
chemical reactions [246, 247]. According to the condensational model, BL is formed
owing to a nonequilibrium condensation of the overcooled steam leaving a super-
sonic nozzle. The basic energy source of BL is the energy of the clusters, excited to
metastable high-lying energy levels. Besides, the authors of the model considered an
oxidation process of BL substance as an additional energy source.
Thus, the hypothesis that BL consists of charged clusters has aroused interest
among researchers in this poorly investigated area of electrochemistry. However,
restriction of limiting energy density, which the system of these clusters is capable
to store with a value of 5  105 J/m3 [31] forces to recognize that the specified
processes have nothing to do with BL.

Ball Lightning Model on the Basis of “Rydberg” Substance

It is possible to avoid the difficulties connected with fast recombination of charges
if BL substance consists of neutral particles. In this case, energy can be stored, for
example, in the form of electronic excitation of atoms and molecules. At excitation
of atoms to states with high principal quantum number n (n ¼ 10), the system is
capable to get new properties. We will consider, for example, the gas consisting of
hydrogen atoms. The radius of a Bohr orbit of hydrogen atom at n ¼ 1 is equal to
a0 ¼ 0.529  108 cm. At the growth of n, the orbit radius increases as rn ¼ a0n2;
thus, for n ¼ 10, it is 0.529  106 cm. The average distance between the atoms is
latm ¼ 0.33  106 cm at atmospheric pressure, which is comparable with the size
of highly excited “Rydberg” atom. As a result, there can be an overlapping of
electron shells of the neighboring excited atoms, which can result in “collectiviza-
tion” of electrons. Exchange and correlation effects lead to electron attraction
among each other. The authors of [252–255] considered that BL consists of “Ryd-
berg” substance with the density close to that of air. The energy density stored in BL
substance lies from 1 to 10 J/cm3, and the surface tension, caused by the electron
attraction, is close to 106 J/cm2 (for comparison – the surface tension coefficient of
water at 20 C is equal to 7.3  106 J/cm2). As the density of electrons is small, the
plasma frequency appears to be much smaller than the frequency of visible radia-
tion. In this connection, BL, consisting of “Rydberg” substance, should seem
transparent. This can be observed due to the radiation caused by the recombination
of the charges. It can conduct an electric current and possess a small electric charge.
The idea that BL substance is a nonideal plasma in which the potential energy of
particles’ interaction is comparable with their kinetic energy [256], has also been
expressed. In [257–259], it was shown that a hypothesis according to which the
substance of BL represents the overcooled metastable non-ideal plasma, and a
hypothesis about “Rydberg” BL substance, supplement each other and can be
combined into a general hypothesis.
6 Ball Lightning Investigations 287

Unfortunately, the model on the basis of the “Rydberg” substance cannot explain
the reason for the high energy content of BL. The limiting values of energy density,
which can store such substance lie in the range from 106 to 107 J/m3 [254].

Ball Lightning Models with Energy Accumulation in Electromagnetic Field

For the first time, the idea that BL energy can be stored in the form of the electric
capacity energy was stated by de Tessan in 1859 [211]. In the process of perfection
of work techniques with electric fields, the authors of new BL models repeatedly
addressed this issue. Hence, according to Mesenjashin, BL is the charged bubble
whose shell consists of water molecules oriented in the electric field [260, 261]. The
thickness of the bubble shell ranges from 0.01 to 100 mm, and charges of an
opposite sign settle down on the bubble poles. The bubble formed at inflating of a
water drop with an impurity of dust is charged in the electric field of a thunderstorm
cloud, thus, the dipole moments of water molecules are guided along the force lines
of the electric field. Owing to this, the bubble gets a mechanical durability. The
main component of free energy of a bubble is the energy of the molecules’
orientation. For a bubble of 10-cm radius with a shell with a thickness of 10 mm,
this energy is 11 kJ (thus, the energy density in regard to the bubble volume is 3 
106 J/m3). Mesenjashin’s bubbles shine by reflected light, and their color can vary
because of an interference of beams in a thin film of a bubble.
The model of unipolarly charged BL, representing a vacuum bubble in atmo-
sphere, was proposed in 1975 [36]. The balance in this system was provided with
the equality of atmospheric pressure force and the force of Coulomb charges’
repulsion at the border of the vacuum cavity. The authors found that the limiting
value of energy density was 4  105 J/m3 in their model, which is in full agreement
with an estimation that can be obtained on the basis of the virial theorem.
The model of BL in the form of the charged water bubble was also described
[262]. Furthermore, in [36], the bubble with a shell bearing charges of one polarity
was considered. In this model, it was assumed that the pressure of electric field
eoE2/2 was counterbalanced by the pressure of the bubble shell, proportional to a
gradient of electric-field strength. For a bubble of 7 cm in radius, this pressure was
2.3  108 N/m2 at a field strength of E ¼ 5.5  109 V/m on its surface. The basic
contribution to energy was given by the energy of the external electric field, which
for the specified conditions, was equal to 5.8  105 J. Having divided this value of
energy by the volume of a sphere of 7 cm in radius, we found an energy density of
rE ¼ 4  108 J/m3.
R. Fedele stated that the behavior of electromagnetic radiation bundles and
bunches of the charged particles can be described by the equations, which resemble
Schrödinger’s equation. The solution of these equations leads to the description of
the processes of self-modulation, self-capture, and self-channeling of bunches with
the formation of solitons [263]. He considered that these processes can be related to
a BL. It can be formed in the atmosphere as follows: (1) under the influence of
intensive electromagnetic radiation in the field of space with the raised level of
ionization, radiation self-capture takes place; (2) owing to the pondermotive force,
288 V.L. Bychkov et al.

electrons are pushed outside and form a spherical layer, and heavier ions remain in
the sphere; (3) the spherical capacity is formed, in which electric-field energy is
stored; and (4) as the formed system is unstable, an ionic core starts to extend and
eventually reaches an electronic shell, returning the stored energy. The time during
which the proposed system exists depends on the competition of the processes of
core inertial compression and Coulomb ion repulsion.
As it can be seen, in Fedele’s model, BL energy is stored in the form of electric-
field energy We ¼ CU2/2 (where C is the condenser capacity and U is the potential
difference between an ionic kernel and electronic sphere). This energy is not large
owing to the small capacity value.
A serious obstacle on the way of accumulation of energy in the electric field is a
rather small value of breakdown electric-field strength in the air, Ebra ¼ 3  106 V/m.
Therefore, it seems that the problem can be facilitated if energy is stored in a
magnetic field. The energy density of a magnetic field in vacuum is rM ¼ B2/2m0
(where B is a field induction and m0 is a magnetic constant). For obtaining “Goodlet”
energy density rBL ¼ 1010 J/m3, it is necessary for BL to have a magnetic field with
an induction of B ¼ 158 T (we noticed that in the case of accumulation of such
energy density in electric field, its intensity should be 4.7  1010 V/m, which is 104
times greater than the breakdown strength of air). Magnetic fields with energy
density up to 1012 J/m3 could be obtained at a conductor with a current produced
by the compression energy of explosion, and the pressure of the explosive counter-
acting the pressure of a magnetic field exceeded the atmospheric pressure by 107
times [264].
The holding requirements of an area with a strong magnetic field somewhat
softened when the magnetic-field configurations free from forces were considered
(Force Free Field) [265]. The basic idea of such magnetic-field configuration is to
direct a current j, creating the second magnetic field along the force lines of the first
magnetic field H. This condition is described by the equations:

j ¼ rot H ¼ aH;
(6.4)
div H ¼ 0:

Having taken advantage of this idea, the authors of [266, 267] proposed a BL
model in the form of a free from forces bundle of magnetic field lines. According to
their model, in BL of 30 cm in diameter, only a small area of 2  102 cm3 in
volume heated up to a temperature of 18,000 K, emitting light. This area repre-
sented a streamer with fast movement consisting of several knots. The streamer had
a diameter of about 100 mm and a length of some meters. At the magnetic-field
induction of B ¼ 3 T, the energy stored in the field was 24 kJ. The described kind of
BL in the form of a ball of magnetic force lines faced sharp criticism from Jennison
who had seen the real BL [268]. In his opinion, the “magnetic” BL has nothing in
common with the natural one. Besides, the question connected with fast light
emission of such high-temperature BL remains unexplained [31].
6 Ball Lightning Investigations 289

One should also remember that the constant magnetic field can be supported
only by currents. In usual conductors, these currents tend to fade because of ohmic
losses, and in this connection, the time of magnetic-field existence is limited.
Callebaut [269] demonstrated that a characteristic time of attenuation for magnetic
field, free from forces, is

1
tdp ¼ a2  : (6.5)

Here, a is the proportionality factor in the formula (6.4), which has a dimension
of m1, and  is the specific resistance of a medium in which the current flows,
 ¼ (ms)1 (m is the magnetic permeability and s is the specific conductivity). For
plasma, it is possible to accept  ¼ 1 m2/s. Starting from the BL size of dBL ¼ 0.1
m, the parameter a can be estimated as a ¼ dBL1 ¼ 10 m1. By substituting these
figures in the formula (6.5), we can obtain the magnetic-field existence time tdp
¼ 102 s.
If superconductivity conditions are somehow provided, the time tdp can be
increased. Superconductivity can create runaway electrons. They can appear in
special conditions, for example, owing to acceleration in an initial phase of the
strong solitary wave initiated by the discharge of a linear lightning [270].
Many researchers proposed a BL model in the form of the cavity resonator for
high-frequency radiation with walls of plasma. One of the early examples of such
model is “a bubble filled with radiation” [196]. It is considered that in a dielectric
plasma sphere, there is the high-frequency radiation whose resonant frequency
considerably exceeds the frequency of electron collisions with the molecules. The
existence of plasma in a sphere is supported by high-frequency radiation, and
conducting plasma walls of the cavity resonator are formed owing to the currents
flowing on an internal surface of its walls. The radiation pressure upon the resonator
walls is counterbalanced by the atmospheric pressure; therefore, the energy density
of a high-frequency field appears to be low – approximately 105 J/m3. The authors
considered collisions of electrons with molecules as the main channel of energy
losses; however, they did not consider the attenuation of radiation because of ohmic
losses in the resonator walls. In fact, even at a very high quality factor of the
resonator (e.g., Q ¼ 106), oscillations with a frequency of n ¼ 2.6  109 Hz will
attenuate during

Q=n ¼ 4  104 s:

The authors of [271–273] believed that it is possible to find conditions when

energy of the radiation, “locked” in a cavity with reflecting walls, whose separation
displacement is restrained by atmospheric pressure, can considerably exceed the
limit allowed by the virial theorem (rE ¼ 3  105 J/m3). However, the analysis of
such configuration [225] has shown that the virial theorem acts in this case with all
its restrictions.
290 V.L. Bychkov et al.

The authors of [105, 106] found out that at an explosion of round metal
diaphragms, at the impact of a powerful current pulse, shining spheres that existed
for some seconds, were formed. They considered that it is the analog of BL
representing a potential hole for captured electromagnetic radiation.

Models Capable to Explain Specific Ball Lightning Energy Content Greater

than 109 J/m3
As observed earlier, the force creating a surface tension in the ensemble of
“Rydberg” atoms arises owing to the exchange interaction of electrons. The idea
about the important role of this interaction with reference to BL was proposed for the
first time by Neugebauer in 1937 [189]. In a case when plasma is almost completely
ionized and the density of the charges is high enough, force of their mutual attraction
can appear, because of the exchange interaction of electrons exceeding their
Coulomb repulsion force. At electron density ne ¼ 2.7  1019 cm3, this condition
is reached if the temperature of the ionized gas is not higher than 600 K.
Neugebauer’s idea was developed in Dijkhuis’s works. Unlike Neugebauer, who
considered a homogeneous sphere of plasma, and electrons as fermions, Dijkhuis
considered that at the large density of electrons, formation of vortical tubes of
coupled electrons, satisfying the Boze–Einstein statistics, is possible [191]. The
“Bozonized” electrons in the tubes are attracted to each other not owing to the
exchange forces, as considered by Neugebauer. These forces are too weak for
maintenance of high-energy BL existence. Thus, the necessary binding energy is
provided by the nonuniformity of “Bozon” charge-density distribution.
Condensation of electrons and formation of vortices can occur at the lightning
discharge at critical values of temperature of 1,844 K, pressure of 6.04 MPa, and an
electron density of 1.28  1027 m3 [82, 274]. Nonuniform distribution of elec-
tronic vortices density r over the sphere volume (that occurs because of the
temperature difference between the central and peripheral areas of a sphere at
constant pressure) should lead to the occurrence of the force directed to the center
of a sphere whose value is proportional to r2 r1=2 =r1=2 [275]. Dijkhuis considered
that the virial theorem in the traditional form is inapplicable to two-component
plasma considered by him [276].
According to Dijkhuis, the radius of vortical electronic tubes is equal to 53 
1012 m (Bohr radius), and around these tubes, positive ions rotate at a distance of
about 107 m. At the extension of coaxial electron-ion tubes, there is an accelera-
tion of ions to a limiting energy of 511 keV (energy of electron–positron pairs
production). Dijkhuis considered that this restriction is artificially limited to non-
relativistic consideration of processes. At the extension of tubes, new bends appear
that allows them pack up in limited volume. BL formation occurs at displacement
of the channel of a linear lightning. If BL can accumulate in itself a full charge of
typical linear lightning Q ¼ 6 C (in the form of the sum of positive and negative
charges, equal in value) the full kinetic energy of ions will be about 3 MJ, which is
comparable with the energy value of “Goodlet” BL that entered a tub with water
(10 MJ) [13, 14]. In Dijkhuis’s model, new lines of the approach to the BL problem
6 Ball Lightning Investigations 291

solution appeared. The BL substance can be conditionally called highly ordered

plasma. Along with the usual plasma, it consists of unlike charges; however, they
do not recombine as they are spaced. Another important property of Dijkhuis’s
model is that the main BL energy is stored in the form of kinetic energy of ions.
Basically, the capacity of this reservoir is boundless. According to the virial
theorem, movement of particles demands the presence of the force compensating
the centrifugal force of ions. According to Dijkhuis, the nature of this force is the
quantum-mechanical mutual attraction of electrons, and the electrostatic force
keeps the ions on an orbit near the electronic vortical tube.
Dijkhuis considered the reaction of nuclear fusion of deuterium nuclei in water
vapor in atmospheric air as the energy source of BL. The kinetic energy of the ions
rotating around the vortical electronic tubes is quite sufficient for the realization of
such reaction [192].
The problem of dense packing of rotating tubes, such that they do not brake each
other in places of contacts, has spilled over into an interesting mathematical
problem. To obtain a solution to this problem, Dijkhuis devoted a number of studies
[277–282].
Thermonuclear reaction of deuterium nuclei fusion also serves as BL-energy
source in the model proposed by A.N. Vlasov. He considered that BL looks like a
toroidal vortical air ring [149–152]. In this ring, the toroidal current circulates,
creates a magnetic field, and is directed along a toroid axis. The toroid layer is
formed by a cover of positive ions, around which relativistic electrons move, and
are collected in a layer of one electron in thickness. The vortex with a system of
moving charges is formed at a stroke of a linear lightning in earth. Thus, the mode
of continuous electron acceleration is formed and they get the speed close to a
velocity of light. Owing to an exchange of photons and interaction with positive
ions, a Bose-condensation of electrons takes place, and the electronic monolayer
becomes superconducting. Electrons in the monolayer remain practically motion-
less with respect to each other; in this connection, their temperature is near to
absolute zero. Vlasov assumed that ions, for example, deuterium nuclei in a
monolayer of electrons, are screened by the last ones. Approaching of the nuclei
at distances of about 1012 cm, may result in a possible nuclear synthesis reaction.
The analog of this process is muonic catalysis. According to Vlasov’s estimations,
in a volume of BL of 30 cm in diameter, a power of 1.7 kW owing to nuclear
reaction can be realized. As neutrons are formed in the course of nuclear synthesis,
such BL will represent danger to the observer.
It is known that with the reaction of one deuterium nucleus, energy of 1.6
MeV ¼ 2.9  1013 J is released [283, P. 6]. One liter of air at a temperature
20 C and absolute humidity contains 1.8  1017 deuterium atoms. At a full
expenditure of these atoms in the reaction of thermonuclear synthesis, energy
Wnc ¼ 2.9  1013  1.8  1017 J ¼ 52 kJ can be released, which corresponds
to the energy density rnc ¼ 5.2  107 J/m3. This value is more than two orders
smaller than the energy density of “Goodlet” BL [13, 14]. Therefore, in the
specified case, thermonuclear energy cannot be the basic supplier of BL energy.
292 V.L. Bychkov et al.

In 1994, V.L. Bychkov published an article [251] with the description of a model
of a polymeric BL. His “theory is a development of theories [31, 146] in application
to polymeric structures. Fractality is a natural property of a polymeric BL. The
model is electric, as [31], and chemical as [146], and includes all achievements of
[146] in case of polymeric materials.” Bychkov found that in the summer “plants
throw out hundred million tons of organic molecules and macromolecules; and
there is a process of destruction of vegetation of type of leaves and a grass. The
characteristic sizes of such hydrocarbon structures are 105–103 cm. In presence
of electric fields of a thunderstorm or the electric equipment these macromolecules
gather in the easy connected porous structures of the macroscopic sizes.” Associa-
tion of globules occurs as follows: “electric field induces the electric moments on
globules, and interaction of these dipole moments at some spatial configurations
leads to an attraction of globules.” In storm conditions “at growth of fractal
polymeric formations there is a process of their charging by ions and electrons.
Polymeric units can accumulate a considerable quantity of charges because of
existence of traps – emptiness inside or between macromolecular chains or sub-
chains.” The size “of an ionic current on the unit is the sum of diffusion and
hydrodynamic currents in a direction to a surface. Charging can be carried out by
simultaneously positive and negative ions, and because of various speeds of drift of
ions generally the polymeric structure should not be electrically neutral. Thanks to
dielectric properties of polymer structure the recombination of captured positive
and negative charges occurs extremely slowly, therefore the object can accumulate
considerable quantities of electric energy. The charged structure can cause occur-
rence of discharge near its surface. Presence of electron current to a surface of BL in
the field of the discharge leads to heating of a polymeric material. This heating
causes acceleration of processes charges recombination in polymer and to
corresponding additional heating of structure therefore the structure collapses.”
Having accepted that the intensity of breakdown electric field for polymers is
equal to Ebrp ¼ 109 V/m, Bychkov found that for an object with a radius of 10
cm, its fractal polymeric structure at an uncompensated charge of Q ¼ 103 C can
store electric energy of up to 106 J.
Let us make a simple estimation of a polymeric BL-energy content. Bychkov
considered BL substance consisting of islets of charges of a different polarities, that
is, from small electric capacitors. We will accept that all the volume of a sphere is
filled by such capacitors, and that the electric-field strength between their facings is
equal to Ebrp ¼ 109 V/m. The energy density of the electric field is rE ¼ ee0Ebrp2/2.
Having put a value of dielectric permeability of polymer e ¼ 2.5, we found that
rE ¼ 1.11  107 J/m3. In a sphere of radius R ¼ 10 cm, energy Wei ¼ 4.65  104 J
can be stored. Now, we will image that this sphere also possesses the unipolar
electric charge created on its surface electric-field strength of Ebrp ¼ 109 V/m.
A charge of this sphere is Q ¼ 4pe0R2Ebrp ¼ 1.11  103 C, and the energy of the
external electric field is Wee ¼ 2pe0Ebrp2R3 ¼ 5.56  104 J. The sum of the internal
and external field energy is 105 J, and the density of the stored electric energy is 2.4
 107 J/m3. Bychkov, at the charge of 103 C, found that the energy of the sphere
was close to 106 J, which is by ten times greater than what we found. However, the
6 Ball Lightning Investigations 293

estimation made showed that even in the considered extreme case, when the
electric-field strength on a sphere surface exceeds the value of the breakdown
field strength of the air by 33 times (Ebra ¼ 3  106 V/m), BL-energy store of
10–100 MJ could not be reduced to the energy of its electric field. We noticed that
though Bychkov considered himself as a follower of Stakhanov and Smirnov, he
brought elements in the BL model, which were rejected by his predecessors: the
large electric charge (as well as the one connected with it) and the possibility of
increase in its weight. In fact, in the presence of a charge in atmospheric electric
fields, objects can levitate not only with a mass of 3 g but also with a weight of
1–10 kg.
In [284], by supplementing calculations and earlier estimations [251], Bychkov
found that because of the recombination charges in the polymeric structure, its
surface can heat up to temperature of 550 K, that is, to a temperature of ignition of a
majority of polymeric materials (500–600 K). In [285], Bychkov informed that
breakdown electric fields in thin polymeric threads can reach 1010 V/m, and the
value of the limiting density of energy of a polymeric fireball can be raised to 4.5 
108 J/m3.
In 2004, Bychkov completely changed the approach to a structure and an energy
source of a polymeric BL [286]. Thus, according to his ideas, the BL does not arise
because of aggregation of the small parts weighed in air, but “as the result of
unipolar stroke of a linear lightning in any material at which there is its heating and
fusion. Thus occurs formation of the loose unipolarly charged structure – BL.” It
can be “the ball of threads, fused sphere or a bubble.” Thus, the electric field on the
BL surface will be of the order E 3  106–108 V/m and will be defined by the
density of the charge that has passed from the charge of a linear lightning to an area
of melt. “On BL surface the discharge is developed (during 1–102 s), it heats up the
surface to ignition temperature, and the object collapses. The size of a charge of
such BL is defined by conditions of its levitation in the field of a storm cloud.” It
also “leads to development of electrostatic instabilities on a surface of BL . . . with
formation of threads and drops on it. . . . Limits of ignition of drops and threads in
air are much lower, than for a massive material.” Ignition and burning of BL
material leads to energy storing whose density can reach 109 J/m3. In [287, 288],
Bychkov considered conditions of explosion of BL at heating of its surface,
occurrence of burns when a hot BL comes into contact with people, and features
of its passing through holes.
In 1998, Nikitin published an article [139, 140] with the new concept of BL.
A year later, he described the model at the Sixth International Symposium on Ball
Lightning [289]. Similar to other authors of modern models, he had a predecessor –
de Tessan, who proposed, in 1859, a BL model in the form of the spherical electric
capacitor [211]. As we have shown, even at the replacement of air by an insulator
with an electric-field strength Ebrp ¼ 109 V/m, the energy stored in the capacity is
insufficient for the explanation of high-specific energy density of BL. However,
strong dependence of energy density on the electric field (rE E2) makes the use of
this electric capacitor as the keeper of energy an attractive option. Besides, in the
nature, there is an ideal insulator – vacuum. The complexity consists in that in real
294 V.L. Bychkov et al.

electric capacitors the field pulls out the electrons at super-strong electric fields
from metal electrodes, and there is an electric breakdown even at the use of a
vacuum as insulator. Therefore, for capacitor “designing,” it is necessary to refuse
application of solid-state electrodes (as the nature hardly has them during BL
creation). Therefore, it is necessary to search for a configuration whose existence
would be supported by the separated electric charges of different polarity. It is clear
that such charges cannot remain motionless. During this reasoning, the author
developed the “design” termed as “the dynamic electric capacitor.” In the capacitor,
electrons are located in the center, moving in a ring orbit, and positive ions rotate
around them (the lightest of them – protons) under the action of the electric field
created by electrons charge. The current created by the movement of protons
generates a magnetic field inside their orbits, whose vector, in the location of an
electronic ring, is perpendicular to a vector of electric-field strength. In these
“crossed” fields, electrons move over a cycloid (as electrons in a magnetron do).
It appeared that at equality of total charges of protons and electrons (i.e., at an
electric neutrality of the system), a value of the magnetic field strength, created by
the movement of protons, is not sufficient to make electrons move in the closed
orbit. Thus, the system turned out to be positively charged, which added difficulties
in the maintenance of system stability: a force of Coulomb repulsion of the like
charges was added to the obviously unstable combination of two coaxial rings of
rotating charges of different polarities. Besides, for the maintenance of vacuum in
the capacity, the developed “design” should have a shell, isolating it from the
atmosphere. Fortunately, it appeared that this shell can exactly provide stability
to the system. If it is made of the dielectric (e.g., water) in nonuniform electric field
of the charged core (system of moving protons and electrons), then there will be a
polarization of a shell substance, and it will begin to be shrunk along the radius,
interfering with the core expansion. The shell has to be positively charged so that
the core is reflected inside the sphere at the contact of an internal surface of a shell
with external “proton facing” the condenser. It can obtain a charge owing to the fact
that the core gradually loses a charge despite the presence of a vacuum. This charge,
“having filtered” through a cover, flows down into the atmosphere. Estimations
have shown that the main energy in such capacitor is stored not in the form of
electric and magnetic fields, but in the form of kinetic energy of protons. The
density of this energy can reach the limiting values, which have been found out in
high-energy BL (rW ¼ 1011 J/m3). The electric charge of BL plays a double role.
On the one hand, its presence causes expansion of the system because of Coulomb
repulsion of charges, and on the other hand, it creates the force that draws up the
shell. Therefore, at certain values of BL energy and a thickness of its shell, its
existence is possible both at small and large values of the charge [136, 290, 291].
The nature, apparently, chooses the first variant. Estimations show that the energy
density stored in the dynamic electric capacitor increases with the reduction of its
size (connected with increase in orbital speed of protons). Therefore, it is “energet-
ically favorable” to BL to include elements of small size (tiny BL with own core
and shell), placed in the common shell of large size. In the considered system,
electrons are relativistic; therefore, they are a source of synchrotron radiation.
6 Ball Lightning Investigations 295

Fast decay of electrons emission is hindered because they are many in number and
are homogeneously distributed in the orbit. Such BL can be formed at a course of a
strong current pulse in the conductor bend (e.g., in a bend of the linear lightning
channel) [136, 291–293]. Thus, in air, for short time, the vacuum cavity is formed in
which separation and acceleration of charges take place under the action of a
magnetic field of a lightning current and eddy electric field.
In this collection, the ideas of Bychkov, Nikitin, and Dijkhuis – authors of three
high-energy BL models (with respect to their mathematical complexity) – about BL
construction are detailed.
Hence, one can consider that comprehension of the importance of BL’s basic
properties, such as the ability to accumulate large amount of energy, was a push to
working out the full BL theory. From the modern point of view, the BL is
considered as a strongly nonequilibrium high-energy open system capable of self-
organizing. The considered three models, each in their own way, reflect the general
principles of existence of this system. Undoubtedly, a definitive conclusion about
the advantages and disadvantages of these models can be made only after its
experimental verification.

6.1.5 Conclusion

The Tenth jubilee Symposium served as an occasion to make the review of BL

researches’ history, and in particular, the analysis of work carried out in the last
20 years – an activity time of the International Committee on Ball Lightning
(ICBL). It is possible to state that the Committee managed to fulfill its primary
goal: to give the chance to the scientists working in the field of BL researches, to
meet and discuss the results of their researches periodically (once in 2 years). It can
be seen from the above-stated facts that the last 20 years witnessed active
researches on BL problem. Data banks of BL observations were collected and
replenished, experiments on long-lived shining formations were carried out, and
studies on the creation and verification of BL models were conducted. Unfortu-
nately, these activities have not yet brought notable results. The reason for failure,
apparently, is owing to the fact that we could not choose “the main link” in the
properties of BL, and as a result, have incorrectly chosen the ways of its experi-
mental modeling. Also, the hope that a dialog of researchers on these symposia and
conferences will help to somehow smooth these disadvantages, cannot be justified.
For some reason, practically there is no reaction to criticism, and corrections are not
made to the developed models. Constantly, new models of BL are appearing, not as
a result of criticism or the development of existing models, but as though “from
nowhere.” As a rule, authors of new models should explain only separate charac-
teristics of BL and should not pretend to explain all its properties. At the heart of
many experiments, no accurate plan and theoretical base exist. It leads to the fact
that researchers give primary attention to reproduction of only certain properties of
BL: the spherical form of an object, presence of luminescence, and a lifetime of
296 V.L. Bychkov et al.

about 1 s. Thus, the basic properties of BL, such as high density of energy, the
presence of uncompensated electric charge, the complex nature of its structure, the
nonthermal mechanism of radiation, and the features of its movement, have been
neglected.
In the field of BL-observation data collection, there was the same situation: new
data practically added nothing to a “portrait” of an average BL, which for some
reason, has not yet helped to create an adequate model of this phenomenon. Simi-
larly, the time has come to shift to the publication of full descriptions of BL
observations. Quite probably, they may contain data that can become the “key,”
opening its secrets. The time has come to seriously consider the data file of UFO
observations, more than half of those objects possess properties of BL. This,
however, will require an explanation on the rare characteristics of BL and UFO, as
objects of some tens of meters in size, existing for tens of minutes, moving with high
speeds and accelerations, and also their ability to emit out narrow beams of light.
The history of scientific development shows that as a rule, it is difficult to
interpret the consequences of a particular discovery. In the nineteenth century, no
one would have guessed that the opening of an electromagnetic induction by
Faraday could become the basis of the electric power industry in the future. In
the beginning of the twentieth century, only science-fiction writers expected tech-
nical revolution, which took place in the second half of the century, with the
creation of lasers, development of an atomic energy, and a flight in the cosmic
space. In the same way, now, it is impossible to predict what advantage mankind
will gain after the secret of BL will be solved. For now, it has not occurred, and BL
researches will proceed as before  with the efforts of the small group of enthu-
siasts aspiring “to satisfy their curiosity.” It is necessary to realize that BL science is
the interdisciplinary science requiring the participation of experts working in
various areas of physics, chemistry, power, synergetics, biology, and psychology.
Let us hope that the International Committee on Ball Lightning can help to organize
a teamwork of experts from these areas of science.

6.2 Ball Lightning as an Object of Condensed Substance

A BL model has been considered in this subchapter. According to it, BL appears as

the result of linear lightning impacting different materials. At this impact, an heated
unipolarly charged condensed substance area is created, from which some mass is
thrown out. This mass is transformed into a sphere. The sphere’s surface cools
down leading to the creation of a quasi-solid-vitrified cover. The field in the charged
substance of BL is distributed, such that its greater value is situated in the thin layer
under the surfaces. As BL substance is created in natural conditions, there are pores
with captured air and other defects in it. Hence, the favorable conditions for the
creation of discharges exist in this layer. Discharges begin to heat captured air and
surface. At some temperature of heating, the surface reaches the temperature of
melting (some structures such as jets and waves appear on the object surface), and
6 Ball Lightning Investigations 297

then, the temperature of ignition. The BL disappears either owing to explosion or

combustion. It releases a chemical energy of combusted material. Its ability to
levitate is defined by the high value of electrical charge on its surface.

6.2.1 Introduction

One of the theoretical directions of BL description comprises models of BL with a

solid framework (e.g., see [294–300]). They are constructed based on the assump-
tion that these objects at storm activity are formed of different sorts of inorganic
[294–296, 298–300] or organic [297] particles, specks of dust, products of various
materials’ destruction and soil, arising at a lightning stroke, or at the interaction
of charged water clusters [295]. They can have aerosol, drop, or a bubble-like
structure.
BL, whose models are described in [294–296, 298–300], have low values of
energy density  30–100 MJ/m3. These models are not capable of explaining the
cases of BL observation with energy density in the range of 103–106 MJ/m3,
described by observers (e.g., see [301]). In [297], to explain the nature of BL with
its organic origin, an attempt was made with the polymeric threads in the form of a
ball formation that accumulates electric charges on the surface of the threads. Thus,
the density of the polymeric threads in the ball influenced the ability to store energy.
In the case of unipolar charging, it could store energy up to 300 MJ/m3, which is
obviously smaller than those given out by most of the interesting observations. To
develop a theory [302, 303], experiments with the application of organic and
inorganic materials to create long-lived shining objects (LSO), were conducted.
It was possible to create LSO with a lifetime of 5–7 s and size of up to 1 cm at an
insertion of cotton wool or charred wood particles in the channel of erosive plasma
generator with walls of organic glass [302], and LSO with a lifetime of up to 2 s and
size of up to 2 cm in a plasma generator with walls of a mixture of wax, rosin, and
small particles of wood [303]. During the experiments, LSO was observed in which
fused metal particles and pieces of metallic electrodes were found in its base. The
existence time of such objects was defined by their combustion. In case of metal
particles’ oxidation, bang sounds have been detected. These experiments showed that
the composite particles consisting of metal and organic particles can be BL analogs.
In [298], successful experiments on LSO reproduction were carried out at
discharge blow in soil. The authors started from an idea that BL is born in fulgurites
area – in the place of a linear lightning blow where there is a soil fusion, and the
tangle of silicon nanodimensional chains takes off – and the life span of such an
object is basically defined by the time of reduction and oxidation reactions in the air.
Experiments carried out in [304, 305] were the ideological continuation of an earlier
work [298]. Under the influence of an electric arc or MW discharges, subjects of
SiO2 were heated, they turned into the fused shining objects, and perished as a result
of high-temperature oxidation–reduction processes; the upper time of arc discharge
excited doped silicon wafers combustion processes [305] was about 8 s and the sizes
of the objects reached several centimeters.
298 V.L. Bychkov et al.

The results of these works show the perspectives of models where BL exists in
the form of the composite condensed substance created at the impact of a linear
lightning on various materials.
The collection and analysis of BL observations and successful experiments on
the creation of BL analogs have allowed to single out the following BL features that
should be explained using BL theory: a high BL-energy density of up to 105–106
MJ/m3, BL’s unipolar charge and own electric field, long lifetime – about tens of
seconds, an ability to levitate, an ability to ignite subjects, an ability to blow up, an
ability to be transformed from sphere to tape, an ability to penetrate through holes,
an ability to decay not only in balls but in pieces, its manifestation in the form of a
subject with liquid surface with jets ejecting from it, and a condensed object that, at
jumping, can sound like a solid sphere. This theory has to take into account BL’s
appearance in the summer period when thunderstorms occur, and the interaction of
linear lightning with different materials leads to BL’s appearance in accordance
with the observations. This object can have a hot surface, and hence, the developed
theory has to include the possibility of different materials for BL framework.
The model presented here includes some features of [306–308] BL-melted
unipolarly charged object and has developed it further.

6.2.2 General Model

Let us consider the following scenario of BL’s origination and life. The possibility
of BL’s existence follows from the fact that linear lightning leader and the return
stroke are unipolarly charged objects, and they can transfer energy directly and
inductively to the substances. In the simplest case of analysis, which we take as the
starting point, the linear lightning impacts the earth (with soil consisting of inor-
ganic and organic components). At this impact, heated unipolarly charged area is
created. Some mass of melted–charged substance is thrown out from it. This mass is
transformed into a sphere mainly under Coulomb force of charges in this substance
and the counteraction of surface tension force. The sphere’s surface cools down
leading to the creation of quasi-solid-vitrified cover. Convective air flows create a
convective boundary layer over the surface, and hence, observers do not experience
heat from this object. The electric field in the charged substance of BL is
distributed, so that its greater value is situated in the thin layer under the surfaces.
As BL substance is created in natural conditions, there may be pores with captured
air and other defects in it. Hence, favorable conditions for the creation of discharges
exist in this layer. Discharges begin to heat captured air and surface. At some
temperature of heating, the surface reaches the temperature of melting (some
structures such as jets and waves appear on the object surface), and it can be
separated into two balls. Then, at the rising of temperature, the BL surface reaches
ignition temperature. As a result, the BL disappears either owing to explosion or
combustion (with extinction).
6 Ball Lightning Investigations 299

6.2.3 Ball Lightning Energy

The main feature of BL presenting scientific and applied interest is its high-energy
density. As we consider the combustion of BL substance, we have to account for
two main parameters playing their roles at the release of BL energy. The first one is
the mass msph of BL material, and the second is the combustion energy DUcomb
release of BL substance. They allow estimation possible at the upper levels of BL-
energy density. For this, let us consider the natural material at that was impacted by
linear lightning. Its density r can be widely ranged due to its different conditions of
impact from porous to condensed one so r (0.110)
103kg/m3. The combustion-
energy release of the substances changes in a wide range with respect to the
properties of definite substance, DUcomb (2–20)  107 J/kg [309]. Hence, it is
possible to estimate energy W and energy density Wspec of BL released at the
combustion of these materials:

W ¼ msph
DUcomb ; (6.6)

Wspec ¼ msph
DUcomb =ð4pR3sph =3Þ ¼ r
DUcomb ; (6.7)

r ¼ msph =ð4pR3sph =3Þ; (6.8)

where Rsph is the radius of the sphere. By inserting values of density and combus-
tion-energy release of materials, one can obtain the following range of energy
density possible to be released at combustion of BL material, Wspec ¼ 2  102 to
2  106 MJ/m3. These values lie in the range of observed BL-energy densities.
Hence, appearing under linear lightning impact, combusting material can produce
the energy necessary to explain high-energy BL events, and the combustion of
material excited by linear lightning can be the reason for its high energy.
As BL has to levitate, its mass has to be dependent on a value of the electric force
allowing it to levitate. Hence, the force has to be dependent on BL charge value qsurf
and the external electric field Eext. At a definite charge on BL’s surface and high
thunderstorm electric field (including electric fields of the Earth and the cloud, or
their combination), BL would levitate. This follows from the fact that high thun-
derstorm electric fields of 3  103 V/m were detected [310]. The second fact
confirming this is that: usually soil consists of 50% of SiO2 [298], and samples of
SiO2 can accumulate high values of electrical charge [311] (with the creation
of electric fields of 107 V/m and higher over them). Hence, at multiple strokes
of linear lightning at some region on earth, there is a possibility of realization of
extremely high electric fields over it, owing to the accumulation of charges from
these strokes. If we consider BL to be a charged sphere, then the electric charge
value on the surface of it is proportional to the electric field on the surface.
However, one has to remember that the electric field Esurf on the surface of BL
can be very high up to inter-atomic fields that are about 5  1011 V/m.
300 V.L. Bychkov et al.

Let us write down the known equation [312, 313] describing BL levitation over
the conducting surface corresponding to this model (in SI units system):

q2surf
qsurf Eext ¼ msph
g þ ; (6.9)
16px2 ee0
where g is the gravitation constant, e ¼ 1 is the dielectric permeability of the air,
e0 is the vacuum constant, x is the BL height over the earth (here, we consider that
BL and earth surface have an electric charge of same polarity). Furthermore, qsurf
and electric field Esurf are connected by Coulomb law in the form of

qsurf ¼ 4pee0 Esurf
R2sph : (6.10)

Using these equations, it is easy to obtain the following equation connecting the
density of BL material with BL charge, qsurf, external electric field, Eext, and the
electric field, Esurf, on the surface of BL:
!
2
3ee0 Esurf Eext Esurf Rsph Wspec
r¼ 1 ¼ : (6.11a)
gRsph Eext 4x2 DUcomb

This equation allows estimation of the parameters of BL and its height over the
earth with respect to atmospheric thunderstorm conditions.
For example, BL with Rsph ¼ 0.1 m, density r ¼ 3.8  103 kg/m3, and the
electric field on its surface Esurf ¼ 1.44  109 V/m, can levitate at a height x ¼ 2 m
in a thunderstorm atmosphere with the field Eext ¼ 1.0  106 V/m, and its energy
density can be in the range Wspec ¼ (1.6–4)  105 MJ/m3. It can be observed
that the obtained values are about those observed for high-energy BL [301] and are
only by an order of factor 5–7 smaller than those of the extreme cases (1.1–2.5) 
106 J/m3 [301].
Similarly, BL with Rsph ¼ 1 m, density r ¼ 3.8  103 kg/m3, and electric field
on its surface Esurf ¼ 1.44  108 V/m, can levitate at height x ¼ 2 m in a
thunderstorm atmosphere with the field Eext ¼ 1.0  106 V/m, and its energy
density can be in the range Wspec ¼ 1.6  103–4  103 MJ/m3.
If 1>>Esurf =Eext
R2sph =ð4x2 Þ, then one has the most interesting case when

3ee0 Esurf Eext DUcomb

Wspec ¼ ; (6.11b)
gRsph

and

3ee0 Esurf Eext Wspec

r¼ ¼ : (6.11c)
gRsph DUcomb

Let us use the formulas (6.11b and 6.11c) for the analysis of different situations.
Let us consider an organic charged sphere with a typical density of organic material
r  2.0  103 kg/m3, DUcomb 5  107 J/kg, and Wspec  1011 J/m3. This means
6 Ball Lightning Investigations 301

that the electric field strength on the surface of this BL has to be Esurf 2.5  1011
R V/m (R is in m) at e 3 and Eext 103 V/m.
Let us consider some natural mineral-type material with r  2.5  103 kg/m3,
DUcomb 108 J/kg, and Wspec  2.5  1011 J/m3, and it corresponds to Esurf 2.0 
1011·R V/m at e 5 and Eext 103 V/m.
We can observe that the obtained values are about those observed for high-
energy BL [301] and are only by an order of magnitude smaller than those of the
extreme cases of (1.1–2.5)  1012 J/m3 (see Section 6.1.2). Electric fields on BL’s
surface in these cases are of inter-atomic values. In such anomalous conditions,
where the BL consists of excited atomic components, DUcomb can be much higher.
According to [314], BL components in the form of a powder or little particles,
will be more effective from the point of view of combustion. In this case, (6.11c)
can be changed to
!
3ee0 Esurf Eext Esurf R2sph   Wspec1
r1 ¼ 1 1 þ n
ri2 =R2sph ¼ : (6.11d)
gRsph Eext 4x 2 DUcomb

Here, n is the number of particles of radius ri inside the sphere Rsph. For
simplicity, we consider that small particles are identical and that the electric-field
strength Esurf on the surface of the large sphere and on the internal spheres is the
same. These formulas also show that large energy effects can be realized in
anomalous conditions at large Esurf and DUcomb. At n  1 formula (6.2.6d) limits
to the result of (6.2.6c), but it shows that almost all the chemical energy in this case
is located in small particles. Hence, the process of heating and destruction of the BL
surface can lead to an inflammation of small particles resulting in their ignition or
explosion in the air.
Estimates also show that the electric energy of the considered objects is much
smaller than their chemical energy. BL, of course, can exist at the violation of
(6.11b); in this case, they will not exhibit such a unique property as levitation.
From the point of view of the presented theory, the objects obtained in experi-
mental works with ignition and combustion of materials in a plasma represented
variants of BL, such as burning objects of condensed melted material, and they
were analogous to the natural BL.

6.2.4 Electric-Field Parameters

The question of BL charge and electric-field distribution is important for BL, since
it determines the behavior of BL. There are no questions when the electric-field
strength on the BL surface E0 is below the air breakdown value Ebr 3  106 V/m,
and the electric-field strength distribution over the radial direction is
EðrÞ ¼ E0 ðRsph =rÞ2 ðr Rsph Þ: Usually, in models [295, 296], the authors accepted
fields on BL’s surface below the breakdown value.
302 V.L. Bychkov et al.

According to the previous section, electric fields on BL’s surface can be much
higher. In this case, the appearance of plasma has to be realized over the BL’s
surface as a result of ionization, attachment, recombination, and other processes
[315]. This plasma is characterized by electron temperature Te in the range of
3–10 eV and concentrations of electrons Ne 1013–1016 cm3. According to
[316], electrons and ions of plasma will create a layer in the region between E0
and Ebr directed oppositely to the electric field of the sphere, and essentially,
decrease the initial field of the sphere. After that, the plasma decays and the initial
electric field is restored, and the whole process is repeated. The distance at which
1=2
the field decreases is the Debye radius [315, 316] rD ¼ ðee0 Te =ð2e2 Ne ÞÞ ; it
4 5
ranges within the limits rD 4.0  10 –2.0  10 cm. Furthermore, the time
of switching off of the initial field is approximately equal to the time of electron
drift over this distance: t rD =wdr , where electron drift velocity is wdr 107–108
cm/s, or the corresponding frequency n 3  1010–5  1012 s1 ionization times
and creation of electrons may be even smaller. Hence, the frequency of field pulses
will be of n.
The presence of pulsed electric field and appearance of the plasma-change
processes are connected with the attraction to the image of charged objects [317].
Hence, the appearance of plasma can substantially decrease the influence of charge
image and rise the electric-field strength on the BL surface.
The presence of such pulsed field can cause luminescence of lamps with cut
wires (see below). At the introduction of easily ionized particles in this area, change
in the electric field near their surfaces and breakdown can occur. This can explain
the appearance of large luminescent clouds that surround BL.
A distribution of electric field inside the sphere depends on the distribution of
charges in it and the thickness of the cover. For example, in the case of a uniformly
charged sphere with charge Q, its internal field distribution is well known [312, 313]

Eint ðrÞ ¼ Q
r=ð4pee0 R3sph Þ;

In particular, this formula shows that the maximum values of electric-field

strength in the case of the charged sphere are realized near its surface. In the
presence of pores or defects in such a sphere, the breakdowns of air inside them
will occur, as it usually takes place in real-charged dielectric substances at an
applied voltage to their surface [318], and this can explain the luminescence of
BL and heating of its surface.

6.2.5 Ball Lightning Surface

Evidently, the surface of BL will be cooled owing to radiation and thermal

conductivity, whose durations are of minutes for their typical sizes of 0.1–1 m.
Also, there is a process of convection that may be connected with BL cooling.
6 Ball Lightning Investigations 303

A convective boundary layer creation takes place over its surface. At the typical
temperature of melted material of BL, which is 400–600 K for organic and
1,200–1,500 K for inorganic substances, a thin boundary layer owing to convection
is created near a spherical body. The thickness of this layer can be estimated as [319]

pffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
L 2Rsph = Re ¼ 2Rsph = u
2Rsph =nd ;

here Re is the Reynolds number, u is the air velocity, and nd is the dynamic viscosity
(nd ¼ 0.15 cm2/s for air at room temperature). Considering BL with a radius in the
range Rsph ¼ 0:525 cm, and air velocity (typical for air flows in quiet room
conditions) in the range u ¼ 1–10 m/s, one can obtain that the width of BL
boundary is L 0.1–0.3 cm. This means that the drop of the gas temperature
from the BL’s wall temperature to the ambient gas temperature takes place in this
region, and observers will not feel BL’s surface temperature.
A very important mechanism of modification of BL surface was considered in
[320]. It was shown that in the case of charges inside a sphere with water cover, a
polarization of molecules in the cover appeared, which realizes a strong force that
subtends the molecules to the center. Thus, for the charge value of Q 102 C
(corresponding electric-field strength on the surface E 1010 V/m) of a sphere of
6-cm radius, the force was about F 6  108 N. Evidently, this force can be
realized in our case that created a solid cover over BL.
Quick processes of cooling and compression of the surface can not only lead to
the creation of a solid surface, but also a vitrified surface [321, 322] whose
temperature of vitrification is 200–300 K smaller than those of melting.
In the case of BL with a heated surface, when it represents the charged liquid, the
instability phenomena of charged liquid surfaces can be realized [308]. The insta-
bility of the charged melted sphere with the formation of structures on its surface
(corrugation), according to [300], was developed with the following condition for
the electric field on the surface:

2
1=4
Esurf > pffiffiffiffiffiffi at
Patm =Rsph ;
ee0

where at is the surface tension coefficient of the melted material and Patm is the
atmospheric pressure. This formula gives fields of Esurf ¼ 9  106 V/m in the case
of paraffin at Rsph ¼ 0.1 m and Esurf ¼ 1.2  107 V/m in the case of glass at the
same Rsph.
Another instability leading to the separation of particles [299] from the charged
sphere (the electrostatic instability) realizes at

2
1=2
Esurf > pffiffiffiffiffiffi at =Rsph :
ee0
304 V.L. Bychkov et al.

For example, it realizes at Esurf ¼ 3  105 V/m for the sphere of paraffin with
Rsph ¼ 0.1 m and at Esurf ¼ 1.3  107 V/m of glass at the same Rsph.
Development of these instabilities leads to the destruction of the surface and its
separation into pieces.
Let us discuss the question about the type of BL form. First, let us consider BL
with continuous material. As we consider BL consisting of liquid, melted material,
Laplace surface tension, pressure of charges, and atmospheric pressure can impact
BL. In this case, an equation to balance the forces acting on charged BL sphere in
the air has the form [300]

kQ2 2at
¼ Patm þ ; (6.12)
8pR4sph Rsph

where k ¼ 1=ð4p
ee0 Þ. Using the fact that Patm >> 2at =Rsph for materials like
silicon, alumina and paraffin, by which we can model BL substance at typical
sizes, one can obtain the following formula connecting the main parameters

 1=4
kQ2
Rsph ¼ : (6.13)
8
p
Patm

From this formula, it can be observed that a sphere of radius Rsph ¼ 0.1 m can be
formed at Q ¼ 1.8  104 C and of radius Rsph ¼ 1 m at Q ¼ 1.7  102 C. One
can see from this equation that the radius decreases with the loss of electric charge
by BL, for example, in the result of streamers or charge runoff, and hence, the
collapse of the object takes place.
Second, let us consider BL as a bubble with a charged liquid surface. In
equilibrium, the following equation is valid [300]:

kQ2 4at
Pvap þ Pgas þ ¼ Patm þ ; (6.14)
8
p
Rsph
4 Rsph

where Pvap, Pgas describe vapor and gas pressure inside a cavity of the bubble,
respectively, which depend on the conditions of bubble formation. Their sum can be
either higher than the atmospheric pressure or smaller than that with respect to the
conditions of formation.
If the BL’s surface is heated, then Pvap, and Pgas values can rise and lead to a
variation of the BL’s radius. In the case of molecular gas in the cavity, one can show

Pvap þ Pgas ¼ Pint 0
ðT=T0 Þ
ðR0 =Rsph Þ3 ; (6.15)

where Pint 0, T0, R0 are the initial pressure inside a sphere, initial temperature, and
initial radius of the bubble, respectively.
Let us consider different situations using the fact that Patm >> 4at =Rsph for
materials like silicon, alumina, paraffin. If the ball was created so that
Pvap þ Pgas << Patm ; then its radius is determined using (6.13).
6 Ball Lightning Investigations 305

Table 6.1 Ball lightning (BL) Esurf Si Glass Paraffin

radius (m) for different
106 0.8 0.12 0.04
materials (6.16b) at electric
3  106 0.09 0.013 0.004
field Esurf (V/m) on BL
107 0.008 0.0012 0.00045
surface

If the ball was created so that the initial pressure in it was close (by its value) to
the external pressure value (e.g., blowing out of BL) Pvap þ Pgas  Patm , then its
initial radius is determined by the equation

 1=3
kQ2 32
pat
k
Rsph  ¼ : (6.16a)
32
p
at E2surf

It must be noted that usually at ¼ a0
ðT0 =TÞb ; where a0 and T0 are the initial
values of surface tension coefficient and a temperature of a liquid, respectively, and
b > 0, then

!1=3
kQ2
ðT=T0 Þb
Rsph  : (6.16b)
32
p
a0

In Table 6.1, one can observe the values of BL radii obtained by (6.16b) at
different electric-field values on BL surface. These values reasonably agree with the
observed radii of BL.
Now, let us consider BL explosion possibility, for example, when the surface of
BL is heated as the result of gas-discharge development over it, so that it can be
ignited. Thus, in (6.15), we
consider that the temperature inside BL and on its
surface rises as T T0
exp t=tign ; where tign is an ignition time. Thus, from
(6.15) and (6.16b), one can obtain the following equation:

 


3 a R0
Pint 0
exp t=tign
R0 =Rsph  4 :
R0 Rsph

Or


Rsph exp t= 2
tign ; (6.17)

BL’s radius rises sharply in time, and this leads to BL’s destruction.
The appearance of small melted particles due to the electrostatic instabilities of
the BL’s surface facilitates ignition and accelerates the combustion of disperse
material on the BL’s surface and also leads to its explosion.
306 V.L. Bychkov et al.

6.2.6 Ball Lightning Lifetime

Let us estimate the maximal life span of BL. As this model is represented by a
unipolarly charged sphere, charges from air in a form of oppositely to sphere
charged ions will move to its surface. Hence, the time of the sphere’s discharging
is connected with the recombination of atmospheric ions with charges of the sphere.
Let us write down an equation describing the process of decrease in the charge Q on
the sphere, to which ions come through an area 4pR2:

dQ
¼  4pR2 wNi e ¼ 4pR2 bi E
Ni e
dt (6.18)
¼  4pR2 bi
Ni eQ=ð4pR2 ee0 Þ ¼ bi
Ni eQ=ee0 ;

where w is the drift velocity of ions, e is the elementary charge, bi is the mobility of
ions, and Ni is their concentration. Solution of (6.18) gives the typical time of the
charge decrease t:

t ¼ e e0 =ðbi e Ni Þ: (6.19)

Assuming that in atmospheric air, charged particles are mainly produced by

background ionization with a rate S and disappear through ion-ion recombination
with a rate constant aii one obtains that a average number of ions in air is
pffiffiffiffiffiffiffiffiffiffi
Ni  S=aii : (6.20)

By inserting (6.20) into (6.19), one obtains the following formulae for a typical
time of charge decrease in the sphere (maximal value of its lifetime)
pffiffiffiffiffiffiffiffiffiffi
t ¼ ee0 =bi
aii =S: (6.21)

The average value of ion mobility is bi ¼ 2.8  104 m2/(V s) [296, 315],
ion-ion recombination coefficient at room temperature in air aii ¼ 2
106 cm3 =s
[296, 315], and S ¼ 4 cm3 s1 (S changes with respect to location on earth)
[312]. By inserting these values in (6.13), one obtains the maximal BL lifetime
t  1.4  102 s.
As noted in the previous paragraph, at the surface of BL in pores and defects, an
electric field with a high strength is realized and there is an ionization of air and
plasma formation in a thin layer at the surface. The time of life of BL is defined by
the time of a warming up of a thin layer of the BL’s surface. It is either the heating
temperature of a dielectric material to a temperature of destruction of solid BL
cover at which electrostatic instabilities or temperature of ignition of a dielectric
material occur (Tign – T0 300 K, Tign – temperature of air corresponding to
temperature of ignition of BL material). It is supposed that the time of destruction or
burning of this material is much smaller than the time of BL surface warming.
6 Ball Lightning Investigations 307

The simplest equation describing the heating of a surface layer has the form:

@T
Cp ¼ divðw grad TÞ; (6.22)
@t

where w is the thermal conductivity coefficient, Cp is the heat capacity of air, T is the
temperature in the layer of air, and t is the time. In a spherical case, (6.22) has the
form [323]:
 
@Tðr; tÞ w @ 2 Tðr; tÞ 2 @Tðr; tÞ
¼ þ ; (6.23)
@t Cp @r 2 r @r

with the following initial conditions

t > 0; 0 < r < Rsph ;

@Tðr; tÞ qef @Tð0; tÞ

 ¼ 0; ¼ 0; Tð0; tÞ 6¼ 1:
@r w @r

Heating of the surface of the sphere is determined by a flux of heat qef to the
surface. If it has a gas discharge origination, then it is connected with the para-
meters of a discharge:

qef ¼ Ne
w
ee ; (6.24)

where Ne is the electron concentration in the area of the plasma, w is their mobility,
and ee is their mean energy. For estimations, we supposed uniformity of parameters
for all the surface. At qef  100 W/m2, typical for air plasmas, the parameters were
Ne  10 9–1010 cm3, w  10 7 cm/s, and ee  3–4 eV. In principle, qef can vary in
wide limits with respect to external conditions, and for comparison, we also
considered qef  1,000 W/m2.
There is an analytical solution for (6.23) [323]:

qef Rsph
Tðr; tÞ  T0 ¼
Fðr; tÞ; (6.25)
w

where F(r, t) is some function depending on radius and time. Calculations in [324]
were carried out for a condition of a flux to the surface r ¼ Rsph : The results of the
calculations are represented in Fig. 6.8.
Lines a and b correspond to heating of the surface layer for Tign  T0 ¼ 300 K,
a at qef  100 W/m2; and b at qef  1,000 W/m2.
The obtained values allowed comparing them with the observed data on BL’s
life span; solid line corresponds to decaying BL, dashed lines indicate to exploding
BL, and the dashed-point line corresponds to going out (with extinction). As one
308 V.L. Bychkov et al.

Fig. 6.8 Ball lightning (BL) lifetime – radius dependence with respect to different ways of
elimination. (a, b) – BL surface heating by discharges with a power of 100 and 1,000 W/m2,
respectively; solid line – decay of real BL; dashed line – explosion of real BL; dashed-point
line – extinction of real BL; two-point – dashed line – discharging of BL

can see that the addition of qef  100 W/m2 to the surface can, in general, explain
the destruction of BL, the corresponding curve is not far from the set of observed
BL curves. It also explains the tendency of BL’s life span, which increases with
size. However, the values of heating of the BL’s surface at a delivered additional
heat flux of qef  1,000 W/m2, are very small. But, they can explain quick BL
destructions at getting into stoves.
From (6.25) at small BL radii (<5 cm), one obtains

ðTmax  T0 Þ
t Rsph : (6.26)
T0
qef
6 Ball Lightning Investigations 309

In other words, BL’s life span is directly proportional to a temperature interval of

heating, the size of BL, and is inversely proportional to the power released near its
surface. This result can explain the lifetime-diameter dependence for BL observed
during thunderstorms [325].
This section shows that the considered mechanisms that determine BL’s life
span limit its lifetime by the value of 102 s and that the life span increases with
its size.

6.2.7 Examples

Let us consider examples of the presented model application.

The first one is connected with BL observation in Novosibirsk in 1996 [326]. BL
had appeared on a pine-tree at a height of about 12 m. Then, it descended down,
almost directly. At that, it created a furrow, whose width was about 5 mm at the top
and 50 mm near the ground. Then, it moved and covered about 62 m and exploded
in the end. Subsequently, its total observation time was about 8–9 s (we represent
only the necessary data for our analysis).
Before our analysis, we have to mention that a theory of vapor explosion created
by linear lightning and BL was delivered in [327]. According to it, the electric field
of the lightning creates electric current in water, Joule heating takes place, it heats
water to vaporization, and vapor pressure breaks the tree surface.
We also consider this reason for BL’s bark destruction. We accept [328] that
water in trees moves upward exactly under the bark, and hence, there should be
destructions. Also, one has to note that water in trees is purified due to its
transportation through the molecular membranes of a root system [328]. Hence,
its electrical conductivity is close to those of distilled water.
Let us determine the amplitude value of pulsed electric field strength on the
surface of BL, at which evaporation of water takes place under the bark and causes
vapor explosion.

Wv  lv
r
V; (6.27)

where lv is the specific water vaporization heat, lv  2.26
106 J/kg, r is the water
density, and V is the volume occupied by water.
The electrical energy spent for water vaporization is:

We  s
E2
V
t; (6.28)

where E is the pulsed electric field strength on BL’s surface, s is the water electrical
conductivity, s  5  104 S/m (S is siemens) is the average value of distilled
water conductivity, and t is the time of BL interaction with water in the layer under
310 V.L. Bychkov et al.

a bark. For estimates, we took t  1.5 s, proportionally to the covered distance.

Combining (6.27) and (6.28), one has:
rffiffiffiffiffiffiffiffiffiffiffi
lv
r
E : (6.29)
s
t

By inserting necessary data into (6.29) one obtains E 1.7  106 V/m.
This result shows that electrical-field strength was about air breakdown field
strength; it impacted water under the bark and led to the explosion of vapors there.
Let us consider a second example. There are well-known cases of BL passing by
the lines of lighting, where there was damage of wires connecting lamps, and the
moment when BL was near them, the lamps began to shine.
We can explain this through the well-known effect of a capacity in a circuit of
alternating current [329]. In this case, the amplitude value of voltage U0 on the
capacity C, current i0, circular frequency of the current o; and capacity are
connected by the equation

U0 ¼ i0 =ðo
CÞ: (6.30)

By inserting the circular frequency of pulsed electric field o ¼ 2  1011–2 

10 s , a small capacity of the electric lamp of an order C 1012 F, and a current
12 1

i0 10 A necessary to ignite the lamp, from (6.30), one can conclude that a voltage
applied to the lamp has to be U0 5–50 V, which can be realized by BL.
Considering a distance at which this voltage has to be applied, L 10 cm, one
can conclude that BL has created a pulsed electric field with the amplitude value of
E 50–500 V/m, which can be realized by BL or a radius of 0.1 m at a distance of
2 m if the electric field on the BL’s surface remains in the range E0 2  104–2 
105 V/m.
Let us consider the third example – BL blowing out of a socket. For this purpose,
we consider the simplest case – BL as a bubble with a charged liquid surface. Its
radius via charge dependence is described by (6.16b).
In Fig. 6.9, we represent the consequence accounting for the processes taking
place with a bubble created inside the socket (see Fig. 6.9a), which we obtained on
the basis of the equations (6.14)–(6.16). The appearance of a bubble takes place in
the socket at an overloading voltage in the electrical circuit: melting and charging
of some material piece (wire, solder material and the composite material of a
socket) takes place inside the socket, a charged liquid drop occurs, then at motion,
as a result of cavitation, it transforms into the charged bubble [307]. It is covered
with a plasma layer, and hence, there is no direct contact with the walls of the
socket.
If the internal bubble volume is greater than those of the cavity inside the socket,
then it begins to propagate outside through the holes in the socket (Fig. 6.9b).
Charges in the melted, conductive material are concentrated at the edges of the
structure [329], and hence, in (6.15), (6.16), conditions for the outside local
bubbles’ appearance on these edges appear (Fig. 6.9c). The electric-field conditions
6 Ball Lightning Investigations 311

Fig. 6.9 Ball lightning (BL) blowing out process

creating plasma in the cavities of the surface are improved on the surface of the
outside bubbles, their temperature rises owing to joule heating of plasma layer (see
above), and the conductivity rises. Charges begin to flow to these outside bubbles
from the internal bubble, and hence, the size of the internal bubble decreases.
The processes of plasma joule heating in the surface of the outside bubbles
continue to rise their temperature and sizes (6.16b) (Fig. 6.9d). Being melted, the
large, liquid bubbles outside coalesce at touching, producing a single large outside
sphere. As the temperature of the sphere is the highest, all the charges penetrate to
it, and the internal bubble disappears and its connecting part closes. This description
coincides with the pictures represented in [330].

6.2.8 Conclusions

The development of a unipolarly charged melted structure BL theory has been

made. In the present model, in most interesting cases, BL is a heavy condensed
object with a high storage of chemical energy, and the electric fields in this object
are extremely high. In the frames of the presented model, one can explain BL’s
lifetime, explosion and its impact on some materials (wood).
Being quasi-solid, it can be separated into parts, jump, and make the sound of a
solid body, and when it consists of small particles, it can resemble a container of
caviar. As it is made of vitrified materials, it can be transparent and semi-transparent.
When its surface is heated up to melting temperature, BL has an ability to
penetrate through holes, and manifest its liquid surface in the form of jets. Such
BL appearance is most probable in the summer period when thunderstorms take
place and an interaction of linear lightning with different materials leads to BL’s
appearance.
We believe that this work is only the beginning of investigations of such unusual
material object developed by the presented hypothesis.
Estimates of BL’s energy have been made using equilibrium data on energy
released at combustion. These values can be different in the case of excited
material.
312 V.L. Bychkov et al.

For a sequential analysis of the electrical characteristics of charged BL, it is

necessary to solve a problem on plasma characteristics in a strong nonstationary
electric field of BL.
In the analysis of condensed melted substance, we could not find data on its
hardening and vitrification. Hence, there are questions about the plasma layer near
the BL surface, because we do not know exactly what type of surface the BL has.
There are also questions on the combustion and explosion of BL substance;
however, the experimental results of [298, 302, 304, 305] showed that combustion
and explosion take place with condensed melted materials leading to their disap-
pearance during several seconds.
At this stage of investigations, we did not consider penetration of such BL
through windows in this work. The explanation of rare BL properties for penetra-
tion through window panes without visible destruction requires an additional
analysis of observation reports, properties of melted glasses, and the formulation
of new experiments. The last part of the work was started only recently. The same
can be said about the disappearance of golden adornments from the bodies of
people.

6.3 Electrodynamical Model of Ball Lightning

A BL model, as a dynamic electric capacitor, has been presented. According to this

model, the core of BL is formed of a great number of elements, defined as “dynamic
electric capacitors”. In the center of this capacitor, an electrons’ ring is located, and
protons rotate around this ring. Because of protons’ movement, a magnetic field is
generated; this field provides a circular motion of electrons. To make this field
sufficiently strong, it is necessary that the number of protons exceeds the number of
electrons, so that this system has an excess of positive charge. The charges’
movement takes place in vacuum, which can exist owing to the presence of a
shell separating the vacuum cavity from atmosphere. In a nonuniform electric field
of the unipolarly charged core, a force that tends to shrink the shell appears. The
core, in its turn, tends to expand because of Coulomb repulsion of charges and the
centrifugal force of rotating protons. The balance of the oppositely directed forces
maintains the stable existence of this system.

6.3.1 Introduction

A key to solving the BL problem is in an explanation of the cases showing its

extremely high energy content. There are some cases when BL had fallen into
vessels with water. On the basis of these events, through water heating, its limit
energy density was estimated to be in the region from 1010 to 1012 J/m3 [331–335].
For the autonomous BL, which does not receive energy from the outside, such high
6 Ball Lightning Investigations 313

values of energy density cannot be explained either by the chemical energy of

gaseous reagents or by the electric energy of ionized gas. The most probable form
of BL energy is the kinetic energy of its particles. However, in this case, for the
holding of moving particles in the restricted volume, there has to exist some
container, preventing the expansion of the power core of BL.
Among the numerous BL models, only these that possess a potential ability to
explain its high energy content deserve attention. It is the author’s opinion that there
are three such models [336–338]. The first one is the BL composed of the closely
stowed nanotubes. On the axis of this nanotube, an electron thread, compressed by
quantum exchange forces, is located, and the surface of the tube comprises ions,
rotating around the electron thread [339–341]. The second one is the model of BL in
the form of a charged bubble with hard walls (of polymer, metal, or silicon), in
which the energy is produced owing to the combustion of solid shell [342, 343].
And, lastly, the third one is the author’s model, named as the electrodynamical
model of BL [344–347]. According to this model, the core of BL is formed of a
great number of elements, defined as “dynamic electric capacitors”. In the center of
this capacitor, an electrons’ ring is located, and protons rotate around this ring.
Because of protons’ movement, a magnetic field is generated, and this field
provides a circular motion of electrons. To make this field sufficiently strong, it is
necessary that the number of protons exceeds the number of electrons; hence, this
system must have an excess of positive charge. The charges’ movement must take
place in vacuum, which can exist owing to the presence of a shell, separating
vacuum cavity from the atmosphere. In a nonuniform electric field of unipolarly
charged core, a force appears, which tends to shrink the shell. The core, in its turn,
tends to expand because of Coulomb repulsion of charges and the centrifugal force
of rotating protons. The balance of the oppositely directed forces maintains the
stable existence of this system.

6.3.2 Dynamic Electric Capacitor

The considered system may be schematically represented by two concentric flat

rings, the inner one (of radius r) comprises the electrons, and the outer (of radius R)
comprises the protons (see Fig.6.10).
Electrons, located in the radial electric field E, with a magnetic field H orthogo-
nal to it, drift with a velocity vd [348, 349]:

vd ¼ E=m0 H; (6.31)

where m0 is the magnetic constant, m0 ¼ 1.2566  106 V s/A m. The condition must
be fulfilled in the order in which the electrons move in a round orbit [350, P. 75]:

H=E r ðe0 =m0 Þ1=2 ; (6.32)

314 V.L. Bychkov et al.

Fig. 6.10 Diagram of charges movement in dynamic electric capacitor. The inner rings includes
electrons, the outer ring comprises protons

where e0 is an electric constant, e0 ¼ 8.8542  1012 F/m. By substituting

H ¼ (e0/m0)1/2E in (6.31), we obtain that electrons must move with a speed near
the light velocity:

vd  ðe0 m0 Þ1=2 ¼ c: (6.33)

While moving in a closed orbit, the electrons execute oscillations along the
direction of the electric field with amplitude [351]:

2 2 
b ¼ ð2mEÞ em0 H ; (6.34)

where m and e are mass and charge of electron, respectively. The value b must be
less than the gap R  r between the orbits of electrons and protons. By substituting
the expression H2 ¼ e0E2/m0 from (6.32) in (6.34), we obtain:

b  2m=ðem0 e0 EÞ: (6.35)

To validate the possibility of existence of the discussed system, let us make

some estimates. For the convenience of calculations, let us replace our system,
composed of negatively charged electron ring and positively charged spherical
shell of rotating protons, with a cylindrical electric capacitor with external positive
electrode of radius R, negative inner electrode of radius r, and cylinder height h.
The capacitance of this condenser is C ¼ 2pe0h/ln(R/r) [352]. Let the total charge
of the electrons be Qe. In this case, a potential difference on its electrodes is
6 Ball Lightning Investigations 315

U ¼ Qe/C ¼ Qeln(R/r)/2pe0h, and an average electric field in the space between

electrodes is


U Qe ln R=r
E¼ ¼ : (6.36)
ðR  r Þ 2pe0 hðR  r Þ

Let us assume that protons are retained in the orbit because of the action
of electric field E. By equating centripetal force Fe ¼ eE to centrifugal force
Fc ¼ (mpvp2)/R (mp is a proton’s mass, and vp is a proton’s velocity), we find:

 
eER 1=2
vp ¼ : (6.37)
mp

The frequency of the proton rotation in an orbit is

 1=2
vp eE
fp ¼ ¼ : (6.38)
2pR 4p2 mp R

Let the total charge of protons be Qp, then a current, produced by their motion, is

!1=2
Q2p eE
Ip ¼ Qp
fp ¼ : (6.39)
4p2 mp R

The protons’ movement generates a magnetic field in the center of orbit

!1=2
Ip Q2p eE
Hp ¼ ¼ : (6.40)
2R 16p2 mp R3

Taking into account (6.32), we obtain:

Qe em0 hðR  rÞ hðR  rÞ


¼ 4:7889
3 : (6.41)
Q2p 8pmp R3 lnðR=rÞ R lnðR=rÞ

For the performance of numerical estimates, let us set specific relationships

between the dimensions of the capacitor. Let h ¼ R and r ¼ 0.7R. In this case,
(6.41) may be rewritten as

Q2p 0:2483RQe : (6.42)

The velocity of proton is (see (6.36) and (6.37)):

316 V.L. Bychkov et al.

1=2
eR Q lnðR=rÞ
vp ¼
: (6.43)
mp 2pe0 hðR  rÞ

For the accepted dimensions of the condenser

1=2
vp ¼ 2:0469  1018 Qe =R : (6.44)

The kinetic energy of proton is

mp v2p eRQe lnðR=rÞ

wp ¼ ¼
: (6.45)
2 2 2pe0 hðR  rÞ

The total number of protons is Np ¼ Qp /e, and hence, the kinetic energy of Np
protons is
RQp Qe lnðR=rÞ
Wp ¼ Np
wp ¼
: (6.46)
2 2pe0 hðR  rÞ

At h ¼ R and r ¼ 0.7R

Wp ¼ 1:0686  1010 Qp Qe =R: (6.47)

Expressing Qp from Qe and R (see (6.42)), we can obtain:

Wp ¼ 5:3248  109 Qe 3=2 =R1=2 : (6.48)

By assuming the volume of dynamic capacitor as Vc ¼ 4pR3/3, one can find an

average density of protons’ kinetic energy:

Wp 3Qp Qe lnðR=rÞ
rW ¼ ¼
: (6.49)
Vc 16p2 R2 e0 hðR  rÞ

For the accepted dimensions of the capacitor

rW ¼ 2:551  109 Qp Qe =R4 : (6.50)
Using (6.42), we can obtain:
rW ¼ 1:271  109 Qe 3=2 =R7=2 : (6.51)
In Table 6.2, values of the total protons’ charge Qp, uncompensated positive
charge of the system Qp  Qe, proton’s velocity vp, total kinetic energy of protons
Wp, and density of protons’ kinetic energy rW for values of the total electrons’
charge from Qe ¼ 1012 to Qe ¼ 103 C and for values of condenser radius from
R ¼ 1010 to R ¼ 101 m, are presented. Only these cases were considered when
6 Ball Lightning Investigations 317

Table 6.2 Dynamic electric capacitor parameters with respect to its dimensions R ¼ h, r ¼ 0.7R at
different values of electrons’ charge and protons’ charge
R (m) Qp (C) Qp  Qe (C) vp (m/s) Wp (J) rW (J/m3)
12
Qe ¼ 10 (C)
101 1.58  107 1.58  107 4.52  103 1.68  108 4.02  106
102 4.98  108 4.98  108 1.43  104 5.32  108 1.27  102
103 1.58  108 1.58  108 4.52  104 1.68  107 4.02  101
104 4.98  109 4.98  109 1.43  105 5.32  107 1.27  105
105 1.58  109 1.58  109 4.52  105 1.68  106 4.02  108
106 4.98  1010 4.97  1010 1.43  106 5.32  106 1.27  1012
107 1.58  1010 1.57  1010 4.52  106 1.68  105 4.02  1015
108 4.98  1011 4.88  1011 1.43  107 5.32  105 1.27  1019
109 1.58  1011 1.48  1011 4.52  107 1.68  104 4.02  1022
1010 4.98  1012 3.98  1012 1.43  108 5.32  104 1.27  1026
Qe ¼ 1011 (C)
101 4.98  107 4.98  107 1.43  104 5.32  107 1.27  104
102 1.58  107 1.58· 107 4.52  104 1.68  106 4.02  101
103 4.98  108 4.98· 108 1.43  105 5.32  106 1.27  103
104 1.58  108 1.58· 108 4.52  105 1.68  105 4.02  106
105 4.98  109 4.97· 109 1.43  106 5.32  105 1.27  1010
106 1.58  109 1.57· 109 4.52  106 1.68  104 4.02  1013
107 4.98  1010 4.88· 1010 1.43  107 5.32  104 1.27  1017
108 1.58  1010 1.48· 1010 4.52  107 1.68  103 4.02  1020
109 4.98  1011 3.98· 1011 1.43  108 5.32  103 1.27  1024
Qe ¼ 1010 (C)
101 1.58  106 1.58· 106 4.52  104 1.68  105 4.02  103
102 4.98  107 4.98· 107 1.43  105 5.32  105 1.27  101
103 1.58  107 1.58· 107 4.52  105 1.68  104 4.02  104
104 4.98  108 4.97· 108 1.43  106 5.32  104 1.27  108
105 1.58  108 1.57· 108 4.52  106 1.68  103 4.02  1011
106 4.98  109 4.88· 109 1.43  107 5.32  103 1.27  1015
107 1.58  109 1.48· 109 4.52  107 1.68  102 4.02  1018
108 4.98  1010 3.98· 1010 1.43  108 5.32  102 1.27  1022
Qe ¼ 109 (C)
101 4.98  106 4.98  106 1.43  105 5.32  104 1.27  101
102 1.58  106 1.58  106 4.52  105 1.68  103 4.02  102
103 4.98  107 4.97  107 1.43  106 5.32  103 1.27  106
104 1.58  107 1.57  107 4.52  106 1.68  102 4.02  109
105 4.98  108 4.88  108 1.43  107 5.32  102 1.27  1013
106 1.58  108 1.48  108 4.52  107 1.68  101 4.02  1016
107 4.98  109 3.98  109 1.43  108 5.32  101 1.27  1020
Qe ¼ 108 (C)
101 1.58  105 1.58  105 4.52  105 1.68  102 4.02
102 4.98  106 4.97  106 1.43  106 5.32  102 1.27  104
103 1.58  106 1.57  106 4.52  106 1.68  101 4.02  107
104 4.98  107 4.88  107 1.43  107 5.32  101 1.27  1011
105 1.58  107 1.48  107 4.52  107 1.68 4.02  1014
106 4.98  108 3.98  108 1.43  108 5.33 1.27  1018
Qe ¼ 107 (C)
101 4.98  105 4.97  105 1.43  106 5.32  101 1.27  102
102 1.58  105 1.57  105 4.52  106 1.68 4.02  105
(continued )
318 V.L. Bychkov et al.

Table 6.2 (continued)

R (m) Qp (C) Qp  Qe (C) vp (m/s) Wp (J) rW (J/m3)
103 4.98  106 4.88  106 1.43  107 5.32 1.27  109
104 1.58  106 1.48  106 4.52  107 16.8 4.02  1012
105 4.98  107 3.98  107 1.43  108 53.2 1.27  1016
Qe ¼ 106 (C)
101 1.58  104 1.57  104 4.52  106 16.8 4.02  103
102 4.98  105 4.88  105 1.43  107 53.2 1.27  107
103 1.58  105 1.48  105 4.52  107 168.0 4.02  1010
104 4.98  106 3.98  106 1.43  108 532.0 1.27  1014
Qe ¼ 105 (C)
101 4.98  104 4.88  104 1.43  107 5.32  102 1.27  105
102 1.58  104 1.48  104 4.52  107 1.68  103 4.02  108
103 4.98  105 3.98  105 1.43  108 5.32  103 1.27  1012
Qe ¼ 104 (C)
101 1.58  103 1.48  103 4.52  107 1.68  104 4.02  106
102 4.98  104 3.98  104 1.43  108 5.32  104 1.27  1010
Qe ¼ 103 (C)
101 4.98  103 3.98  103 1.43  108 5.32  105 1.27  108

the protons’ charge exceeded the electrons’ charge, and proton’s velocity was less
than the light velocity. The second restriction was a limit of the number of electrons
in the orbit because of quantum mechanics laws.
According to the uncertainty principle, distance Dx between the electrons cannot
be less than h=2p (where h is the Planck’s constant, h ¼ 1.05457  1034 J·s, and
p is the electron’s pulse). For relativistic electrons, p ¼ we/c (where we is the
electron’s energy and c is the light velocity), hence

Dx r hc=2we : (6.52)

Let us assume that electrons may occupy a space inside a ring tube with a cross-
section b2 (see (6.35)) and a length 2pr. The volume of this tube is

8pm2 r r
Vtb ¼ 2prb2 ¼
¼ 6:564  1012 2 : (6.53)
e2 m20 e20 E2 E
By expressing E in the charge of electron ring (see (6.36)), we have
32p3 m2 rh2 ðR  rÞ2
Vtb ¼
: (6.54)
e2 m20 e20 Q2e ½lnðR=rÞ2

By assuming h ¼ R, r ¼ 0.7R, we find

Vtb ¼ 1:006  108 R5 =Q2e : (6.55)
6 Ball Lightning Investigations 319

Let the volume that an electron may occupy in space be

ve ¼ ðDxÞ3 ¼ ð
hc=2we Þ3 : (6.56)

Then, the limit number of electrons inside the ring tube is Nmax ¼ Vtb·ve and the
limit charge of these electrons is
8:048  108 eR5 w3e
ðQe Þm ¼ eNmax ¼ : (6.57)

h3 c3 Q2e

From (6.57), a ratio of maximum electron charge to real electron charge can be
obtained as
ðQe Þm =Qe ¼ 4:08  1050 R5 w3e =Q3e : (6.58)

Let we ¼ 1 MeV ¼ 1.6  1013 J. Then, (Qe)m/Qe ¼ 1.677  1012 R5/Qe3. In

the Table 6.2, only those combinations of R and Qe are allowed when (Qe)m/
Qe ¼ 1. The analysis showed that the most severe criterion is the requirement
that the proton’s velocity should be less than light velocity, rather than an electron’s
density in the ring should be less than its limit value.
One can see from Table 6.2 that the energy, stored in the dynamic capacitor
increases with decreasing radius R. This is explained by the corresponding increase
in the proton’s velocity, and hence, its kinetic energy. The second conclusion based
on the table is that the greatest possible energy may be stored in the system
consisting of a great number of small-sized condensers. Let us estimate the amount
of energy contained in the system of dynamic condensers with a total proton’s
charge QS ¼ 101 C. If the proton’s charge of the unit condenser is Qp, then the
number of such condensers is Nc ¼ QS /Qp and these total energy is (see (6.47)):
WS ¼ Wp
Nc ¼ Wp QS =Qp ¼ 1:0686  1010 Qe QS =R: (6.59)

For the lower lines of Table 6.2 at different Qe, relation Qe/R ¼ 102 C/m. At
QS ¼ 101 C, WS ¼ 10.689  106 J. Thus, it can be observed that this system is
able to store about 10 MJ of energy in the form of the kinetic energy of protons.

6.3.3 Features of Ball Lightning, Following from

Electrodynamical Model

6.3.3.1 Energy of Ball Lightning

The main goal of the performed estimates was to demonstrate the capability of the
discussed system to store energy, comparable with the energy that may be
contained in the natural BL. However, we did not take into account forces
320 V.L. Bychkov et al.

expanding the proton’s orbits owing to their Coulomb repulsion, as well as the
possibility of the fact that their orbital (or “quasi-orbital”) motion, besides attrac-
tion to central negative charge, may be maintained by their reflection from the walls
of some container. Now, let us turn our attention to the discussion of properties of
this container – a shell of BL.
Let us consider that inside the spherical shell of some dielectric (e.g., water), a
power core (one or some dynamical capacitors) with uncompensated charge Q is
located. This charge generates electric field in the place, where the shell is located,
Esh ¼ Q/4pe0Rsh2 (Rsh is the inner radius of the shell). In such field, polarization of
the shell will take place: on its inner surface, negative charges will appear, and on
the outer surface, positive ones will occur. Under the total polarization of the shell’s
material in a strong electric field, the surface density of the charges s may be
assumed to be equal to Ns , which is the number of water molecules on the unit
of the shell surface. The distance g between the water molecules in ice crystal is
g ¼ 3  1010 m, and hence, Ns ¼ (g)2 ¼ 1019 m2 and the surface density of
charges is s ¼ Ns
e ¼ 1.6 C/m2. The dipole moment of a surface unit of thickness a
is D ¼ sa ¼ 1.6a C/m. Therefore, the force Pcp, attracting a unit of surface towards
the charge Q, is equal to

Pcp ¼ D
grad E ¼ 2DQ=4pe0 R3 ¼ 2DE=R: (6.60)

By multiplying Pcp by an area of the inner surface of the shell S ¼ 4pR2, we can
find a force shrinking the shell towards the center:

Fcp ¼ Pcp
4pR2 ¼ 2saQ=e0 R: (6.61)

As in the above-discussed numerical example, let the uncompensated charge

(whose value is near the total charge of the protons) be Q ¼ 101 C, and the inner
radius of the shell be Rsh ¼ 101 m. In this case, Fcp ¼ 3.614  1011a (N).
A centrifugal force of protons with a total mass Mp, possessing the energy Wp, is

2Mp v2p 2Wp

Fcf ¼ ¼ : (6.62)
2R R

Fcf ¼ 2  108 N at Wp ¼ 107 J. By equating Fcp to Fcf, we can find that for the
compensation of centrifugal force, the shell thickness must be equal to a ¼ 0.55
mm. For the shell to reflect positively charged core at impact, it must also be
positively charged. It can acquire this charge from a continuous loss of charge of the
energic core. This charge, passing through the shell’s wall, leaks to the atmosphere
from its outer surface.
Besides the centrifugal force Fcf, the shell is stretched by the force Fel, produced
by Coulomb repulsion of charges and by magnetic field, generated by protons’
current. This force [353, P. 46] is:
6 Ball Lightning Investigations 321

kQ2
Fel ¼ : (6.63)
8pe0 R2

where k is a coefficient, depending on the proton’s velocity and parameters of cross-

section of the current’s layer, k  2. Thus, two forces shrink the shell: a “gradient”
electric force (see (6.61))

4saQ
Fcp ¼ ; (6.64)
e0 ð2R þ aÞ

(where R is replaced by R þ a/2), and a force of atmospheric pressure Pa:

Fa ¼ Pa
4pR2 : (6.65)

The condition of shell equilibrium is a balance of all forces acting on it:

2Wp kQ2 4saQ

F ¼ Fcf þ Fel þ Fcp þ Fa ¼ þ   4pPa R2 ¼ 0: (6.66)
R 8pe0 R 2 e0 ð2R þ aÞ

Taking the derivative dF/dR and accounting the constancy of shell’s volume as
well as the fact that the increase in the shell’s radius takes place because of the
work of forces expanding the energic core, one can show that at any values of k,
Q, a, and R derivative dF/dR < 0. This means that the equilibrium of the shell is
stable [354].
In Fig. 6.11, the dependencies of the shell thickness a on the energic core charge
Q, constructed by (6.66) for inner radius R ¼ 3 cm, k ¼ 2, and kinetic energy of
protons Wp from 1 to 107 J are shown. One can see that when the energy exceeds
16.9646 J, the curves of dependence a ¼ f(Q) have a minimum value. From the
condition da/dQ ¼ 0 from (6.66), one can find values of minimum wall thickness
am and the corresponding values of charge Qm:

1=2
16pe0 R

Qm ¼
Wp  2pPa R3 ; (6.67)
k

2R’
am ¼ ; (6.68)
1’

where


1=2
ke0 Wp  2pPa R3
’¼ :
16ps2 R3

By rewriting (6.66) in the form:

322 V.L. Bychkov et al.

3
11
10

8
1
7

6
Iga, mm

–1

5
–3 1
2
4 3

–5
–8 –6 –4 –2
IgQ, C

Fig. 6.11 The dependence of shell thickness a for the ball lightning (BL) with inner shell’s radius
R ¼ 3 cm on the charge Q of power core for values of kinetic energy of protons Wp equal to 1 J
(curve 1), 10 J (curve 2), 16.9 J (curve 3), 16.9646 J (curve 4), 17 J (curve 5), 102 J (curve 6), 103 J
(curve 7), 104 J (curve 8), 105 J (curve 9), 106 J (curve 10), and 107 J (curve 11)

2saRQ kQ2
Wp ¼ 2pPa R3 þ  ; (6.69)
e0 ð2R þ aÞ 16pe0 R

we can find a relation between energy Wp and radius R. By substituting a value

Qm in (6.69) (see (6.67)), we can obtain:



2 e0 ð2R þ aÞ 2 16pe0 R

Wp  2pPa R3 ¼ Wp  2pPa R3 : (6.70)
saR k

The apparent solution of this equation is (Wp)1 ¼ 2pPaR3. At this value of Wp,
according to (6.66) and (6.67), Qm ¼ 0 and am ¼ 0, and thus, the equilibrium is
sustained by the atmospheric pressure without the participation of electric forces.
The second solution of (6.70) is
6 Ball Lightning Investigations 323

 16ps2 a2 R3
Wp 2 ¼ 2 pPa R3 þ : (6.71)
ke0 ð2R þ aÞ2

The first term in the right part reflects the contribution to energy of the atmo-
spheric-pressure force, and the second one reflects the contribution to energy of
electric forces. The ratio of these terms is

ke0 Pa ð2R þ aÞ
¼ : (6.72)
8s2 a2

If R ! 0, then  ! 0, and if R ! 1, then  ! 1. This means that at small

radii of the shell, the core’s energy is sustained predominantly by electric forces,
while at large values of radius, the atmospheric force becomes a determining factor.
When a ¼ 102 m, the action of both the forces becomes equal ( ¼ 1) at R ¼ 17
m; on the other hand, when a ¼ 103 m,  ¼ 1 at R ¼ 1.7 m.
There are some cases described in the literature, when the energy of BL that fell
in a vessel with water was measured. The range of the measured energy values was
from 10 to 4,000 MJ. The considered model allows explaining the observed cases of
extremely high energy content of BL.
In Fig. 6.12, the plots of dependencies of outer radius R þ a, the shell wall
volume Ven, and power core charge Q on the protons’ energy Wp are shown. The
graphs are constructed based on (6.66) [347, 354]. Adjacent to the straight line of
dependence R þ a (the observable parameter), points of measured values of BL
energy are plotted. One can see that the experimental data fit the calculated line
reasonably well. It is self-evident that the model, which is able to explain a
mechanism of energy storage for highly-energetic BL, can also explain the reasons
for the energy content for usual BL with energy from 10 to 1,000 J.

6.3.3.2 Radiation of Ball Lightning

The model, trying to approach fact as closely as possible, must be able to explain
not only the nature of its energy content, but also its other characteristics: long life
span, peculiarities of its radiation and movement, and demonstration of its electric
features. According to the considered model, every element of the core of BL –
dynamic capacitor – has electrons, moving in a circular orbit of radius r with a
velocity near to light velocity (see (6.33)). Such electrons are the source of
synchrotron radiation. It is directed at a tangent to the electron’s orbit and is
gathered in a narrow conical beam with an angle of divergence (in radians) [357,
P. 97]:

1=2
dC ¼ 1  b2 ¼ mc2 =we ; (6.73)
324 V.L. Bychkov et al.

2
Ig(R + a), cm; Ig Ven , I; IgQ, C

R+a
0

–2
Ven
Q

–4

–6
3 4 5 6 7 8 9 10
Ig Wp , J

Fig. 6.12 Dependence of outer radius of ball lightning (BL) R þ a, the shell’s wall volume Ven,
and power core charge Q for minimum value of Ven on the kinetic energy of protons Wp. Points are
values of the external radius of the observed BL with energy (from left to right): 10 MJ (Goodlet,
B.L. [14]), 260 MJ (Imianitov, I., Tikhii, D. [32]), 1,000 MJ (Dmitriev, M.T., Bakhtin, B.I.,
Martynov, V.I. [355]), 2,300 MJ (Batygin,A., Mosin, I. [43]), and 4,000 MJ (Balyberdin, V.V.
[20, 356])

where b is the ratio of electron’s velocity v to the light velocity c, b ¼ v/c, m is the
rest mass of the electron, and we is the electron energy. For the synchrotron
radiation, the number nmax of the harmonic of basic frequency of orbital movement
of electron n ¼ c/2pr, corresponding to a maximum of intensity, is defined by the
expression [357, P. 115]:

3  we 3 3
3=2
nmax ¼ ¼ 1  b2 : (6.74)
2 mc2 2

For example, let us consider an electron rotating in the orbit of radius r ¼ 105 m
(see Table 6.2). The frequency of its rotation is n ¼ 4.775  1012 s1. Frequency,
corresponding to a middle of visible radiation spectrum (l ¼ 500 nm), is f ¼ c/l
6 Ball Lightning Investigations 325

¼ 6  1014 Hz. For this frequency, number nmax ¼ f/n ¼ 126. From (6.74), we can
find we ¼ 1.1941  1013 J ¼ 0.7453  106 eV. For the condenser with a radius of
electron orbit r ¼ 104 m for wavelength l ¼ 500 nm, nmax ¼ 1256, for which
we ¼ 2.5726  1013 J ¼ 1.6057  106 eV. For r ¼ 103 m, nmax ¼ 12,565 and
we ¼ 5.5424  1013 J ¼ 3.4593  106 eV. The ratios b of the electron’s velocity to
the light velocity for three considered values of radius r are equal: for r ¼ 105 m,
b ¼ 0.72794; for r ¼ 104 m, b ¼ 0.94801, and for r ¼ 103 m, b ¼ 0.98903.
Power of synchrotron radiation (W) [357, P. 95] is

ce2 b4
Pe ðWÞ ¼ 5:99  109
2
: (6.75)
r 2 ð1  b2 Þ

By substituting values of b for different radii r in this formula, we can find that at
r ¼ 105 m, Pe ¼ 5.86  1010 W, at r ¼ 104 m, Pe ¼ 3.63  1010 W, and at
r ¼ 103 m, Pe ¼ 0.93  1010 W. If the total charge of the electrons is Qe ¼ 108
C (see Table 6.2), their number Ne ¼ Qe/e ¼ 6.24  1010, and the power emitted by
them will be in the range from 5.8 W (for r ¼ 103 m) to 36.6 W (for r ¼ 105 m).
With regard to Table 6.2, we can see that for radius R ¼ 103 m, a store of protons’
kinetic energy Wp is 0.168 J, for R ¼ 104 m, Wp ¼ 0.532 J, and for R ¼ 105 m,
Wp ¼ 1.684 J. Supposing that some effective mechanism of transferring of protons’
kinetic energy to energy of electrons exists, this store will be exhausted: for r ¼ 105
m at 4.6  102 s, for r ¼ 104 m at 2.3  102 s, and for r ¼ 103 m at 2.8  102 s.
The obtained result is at variance with the observation data, according to which,
glowing of BL may last from 1 to 100 s, which is 102–104 times longer. The reason
for the inconsistency lies in the fact that (6.75) describes the emission of a single
electron or a bunch of electrons. If the electrons are uniformly distributed in the orbit,
the power of their collective emission intensity may become considerably lower [358].
With the absolutely uniform distribution of electrons, the power of electrons’ ring
emission may become equal to zero (it is known that the coil with a constant current
does not radiate any electromagnetic energy).
The spectrum of synchrotron radiation consists of a set of electron’s orbital
rotation harmonics nn. It extends from value nnmax of the critical frequency,
corresponding to the maximum of radiation intensity, towards long waves (up to
l ¼ 2pr/c), which is also observed with regard to the short waves. Decrease in the
harmonics intensity with the lowering of the frequency proceeds rather slowly – as
(nn)1/3, for frequencies above the critical value, the emission power falls exponen-
tially with the harmonic number. At some n, it becomes equal to zero [357, P. 115].
The minimum wavelength lmin relates to the maximum intensity wavelength lmax
as lmin ¼ 0.42lmax [357, P. 212]. For lmax ¼ 350 nm, lmin ¼ 147 nm.
If the main reason of BL radiation is, indeed, a synchrotron emission of
electrons, one can predict some basic characteristics of its glowing. Radiation
of BL is not monochromatic, but is a set of frequencies, whose intensity slowly
decreases in the region of low frequencies and sharply falls in the region of high
frequencies. The frequency of emission, corresponding to the maximum
326 V.L. Bychkov et al.

intensity, depends on the electron’s energy. As it decreases (e.g., before decay of

BL), the BL color must become red. Furthermore, the existence of BL emitting
only in the infrared region is possible (“black BL”). The ultraviolet part of the
emission spectrum, penetrating through the shell of BL, may cause ionization and
excitation of air molecules. This may show up in darkness as a halo around the
BL. In the case of transparency of the shell’s wall and of ordering of the
electrons’ orbits, a BL emission may have the form of a conical beam with a
partly polarized light. Synchrotron emission in a range of infrared and millimeter
waves may resonantly interact with the metallic objects of small size (rings
and chains), heating them. As it was discussed earlier, the radiation intensity of
BL must sharply depend on the degree of ordering of the emitting electrons in
the orbits. With the destruction of uniformity of the electrons’ distribution in the
orbit (owing to external action or before its death), the radiation intensity (in the
whole range of waves) may increase by some orders.

6.3.3.3 Motion of Ball Lightning

The character of movement of BL is determined by the availability of noncompen-

sated electrical charge. At fine weather near the earth’s surface, an electric field Efw
presents with vector directed upward (the earth is charged negatively) and the field
intensity is Efw ¼ 130 V/m [359, 360]. In this field, BL with positive charge Q
will be caused by the force Ffw ¼ QEfw to fall from the sky to earth. During a
thunderstorm, the intensity of the electric field near the earth under a cloud may
increase to Ecl ¼ 104–105 V/m, and the direction of vector Ecl is opposite to that of
vector Efw. In this field, BL will experience a lifting force Fcl ¼ QEcl. At charge Q
>103 C, Fcl ¼ 10–100 N. This force is able to lift an object with a mass of 1 – 10
kg. Just as the direction of the electric field vector, its intensity near the earth’s
surface may vary over a wide range. These variations are caused by relief of the
country, the presence of electro-transmission lines and buildings, as well as the
currents flowing in earth. External electric fields may “sag” inside buildings
through windows. Because of this BL moving along electric field lines, can enter
inside the shielded facilities. Electric fields in the near-earth layer are inconstant in
time. Because of this, often the movement of BL is irregular. On the basis of the
described mechanism of charge motion, it has become possible to explain the
behavior of the falling BL in the photo [361, 362].
With the approach of charged BL to earth (or to other conductor), on its surface
charges of opposite polarity are induced. Hence, the force, attracting BL, must
increase at Fat ¼ Q2/4pe0(2l)2 (Here l is the distance between BL and the earth’s
surface). However, the observations show that often BL does not fall on conductor,
but hovers over it at a distance of some of its diameters. One of the hypothesis
explaining such behavior of BL, is an assumption that BL is repelled from the
conductor because of the interaction of eddy currents induced in the conductor, with
currents flowing in it [344, 347]. However, the experimental testing of this hypoth-
esis had shown that forces, generated by these currents, rapidly diminish when
6 Ball Lightning Investigations 327

moving away from the conductor surface (as l7) [363]. The hypothesis, according
to which the reason for the decrease in the attraction of charged BL to a conductor is
the creation of space charge between BL and conductor, is closer to reality. The
polarity of this charge coincides with that of the charge of BL. This charge may
appear due to corona discharge on the surface of BL [364].
When approaching the dielectric, for instance, glass, BL can be attracted to it
because of gradient forces arising due to the dielectric’s polarization. In the
electric field created by the charge of BL, on the glass surface a discharge appears.
This discharge is fed by the energy flow from BL. As a rule, this discharge takes
place on the side of glass, opposite to the side, where BL is located [365]. Local
heating and subsequent cooling results in the creation of a ring hole in glass
[366–368]. BL is able to pass through chinks or small holes even when their
dimensions are less than BL diameter. These evidences are in favor of the
assumption that BL’s core consists of small-size elements (104–103 m, see
Table 6.2), which are able to change mutual position under the action of external
force. Because of the oppositely directed forces of Coulomb repulsion and mag-
netic attraction present in this system, the possibility of the indicated rearrange-
ment of the core seems quite possible. As the shell’s form is determined mainly by
the pressure of the power core, its rearrangement may be accompanied by the
rearrangement of the shell.
It is a considerable challenge for explanation (for all models of autonomous BL)
of the cases of BL passing through glass without its destruction [333, P. 173; 369,
P. 77–85]. In the context of the proposed model, it is assumed that this process
consists of the following steps: (1) BL adhered to the glass fires discharge at the
opposite side of the glass. (2) In the nonuniform electric field of this discharge,
charges are pushed away from the center to the periphery, carrying with them
neutral particles; near the glass surface, a vacuum cavity is created. (3) The dying
“mother” BL transfers its energy to the “daughter” BL in the form of a pulse of
high-frequency radiation. The low-frequency envelope curve of this pulse favors
formation of its power core (this process is described below). (4) From particles
(water, dust) evaporated from the inner surface of glass, the vacuum shell is formed.
A circumstantial evidence of the described scenario’s possibility is the fact that
BL that had passed through the intact glass, as a rule, had the small energy
content [369].

6.3.3.4 Electric Effects Caused by Ball Lightning

By shooting out of the vacuum gap between the shell and power core, the discharge
current may melt conductors, heat water in capillaries of trees, or injure people
or animals. In the last case, the consequences of the impact of BL differ from
the analogous consequences of impact of usual linear lightning. Let the electrical
capacitance of BL be equal to the capacitance of a sphere with a radius R ¼ 101
m: Cbl ¼ 4pe0R ¼ 1011 F. At the resistance of human body, Rb ¼ 103 O, the
328 V.L. Bychkov et al.

duration of the current pulse passing through the body at a discharge of BL is

tbl ¼ RbCbl ¼ 108 s. The depth d of skin layer for the current’s pulse of duration t
is [344]:

d ¼ 0:5ðrtÞ1=2 ðmÞ; (6.76)

where r is the resistivity of the conductor. By substituting tbl ¼ 108 s and resistiv-
ity of body rb ¼ 10 O mm2/m in (6.76), we obtain d ¼ 0.16 mm. This means that at
the impact of BL, the great part of current will flow at the body’s surface, not getting
to the heart area. In contrast, the pulse duration of linear lightning is much larger – up
to 105 s. In this case, according (6.76), d ¼ 0.5 cm. This makes the discharge of
linear lightning more dangerous than that of BL.
The microwave part of BL radiation, whose intensity may be noticeably
increased due to external disturbance of BL, may cause skin burn at close contact
with it. As fast particles (protons and electrons) enter into the composition of the
power core, they may have been thrown out through the torn shell in the form of
streams during the death of BL that did not completely exhaust its energy [332, P.
106; 40, P. 65]. These streams may cause radiation injury to the breathers, who
occasionally found themselves near the exploding BL [333, 370].
According to the observations, the magnetic field of a majority of BL is very
weak. It was reported that BL did not produce any influence on iron nails passing
over them [371, P. 120], and did not cause any movement of steel objects lying in
the pocket of a passenger of aircraft, when it had passed near him at a distance of 50
cm [333]. However, a case is known when BL had lifted a permanent magnet with a
mass of about 0.5 kg [334, P. 72]. In the context of the discussed model, one can
explain the magnetic features of BL with the fact that its core consists of many
elements, and each of them is a magnet [372]. In a normal state, all elementary
magnets are closed to each other, and the external magnetic field of BL is absent.
However, under the influence of the external magnetic field, some parts of the
magnets may change their orientation, and as a result, a magnetic field will appear
outside of BL.
As discussed earlier, BL is a complex ordered system. Therefore, it is reasonable
to suggest that it may be created in nature only as a sequence of certain events. As a
result of these events, there must be: (1) the creation of vacuum cavity in air; (2)
producing of ions and electrons; (3) separation of positive and negative charges in
space; (4) acceleration of protons and electrons; (5) formation of the highly ordered
energy core with excess of positive charges; and (6) creation of a shell, separation
of vacuum cavity from the atmosphere and preventing expansion of the energy
core. It would appear reasonable that the source of BL energy and entropy is a
discharge of linear lightning. The energy of the average lightning is about 5  109 J
[333, 360, 373, 374]. This is a sufficient value even for the generation of BL with an
extremely high energy content. Besides, the creation of a low-entropy, highly
ordered object such as BL is possible only in the system with low entropy. The
linear lightning, undoubtedly, is such a system.
6 Ball Lightning Investigations 329

6.3.4 Creating of Ball Lightning in Nature

A germ of BL – a vacuum cavity in the air may appear due to the vortical motion
of the air near the channel bend of linear lightning [347, 375, 376]. The estimates
show that the lifetime of the spindle-like vortex may be about 40 ms. Let us
consider that a new lightning discharge passes along the coil bend of the channel
surrounding the vortex, after a time of vortex formation. The ultraviolet light of
this discharge will cause the creation of electrons and ions in the wall of the
cavity. Besides, the current of this discharge will produce a pulse of magnetic field
inside the vacuum cavity, directed perpendicular to the plane of the current coil.
Let the radius of this coil be Rcl ¼ 0.2 m. Consider that the lightning current
I increases linearly up to Imax ¼ 4  105 A in a period Dt1 ¼ 10 ms, and after
that, it falls to zero in a time Dt2 ¼ 50 ms [373]. The magnetic field strength at
the coil center is H ¼ I/2Rcl, at t ¼ 10 ms, and it reaches a maximum value
Hmax ¼ Imax/2Rcl ¼ 106 A/m. In the period Dt1 of increase in the magnetic field
inside the coil, an eddy electric field is induced. The vector Eed of this field lies in
a plane of the current coil and is directed along the tangent to a circle of radius r.
The intensity of this field is

pr 2 m0 dH rm Hmax
Eed ¼ 
¼ 0
: (6.77)
2pr dt 2 Dt1

During the period Dt1 ¼ 10 ms at r ¼ 0.2 m, Eed ¼ 1.26  104 V/m. Let us
consider the behavior of charges placed in a small volume of gas in the vortex’s
wall, adjusted to the vacuum cavity. Let us direct the magnetic field vector H along
the negative values of the vortex’s axis Z, and vector Eed in the direction of positive
values of X-axis. In the system of coordinates rotating with an angular velocity
equal to vortex’s rotation velocity, the motion of electrons may be described with
an equation [377]:

mðdv=dtÞ ¼ ðEed þ EÞ  em0 ½vH  mvnm ; (6.78)

where v, m, and e are velocity, mass, and charge of electron, respectively, nm is the
rate of electron collisions with molecules of gas, Eed and E are the x- and y-
components of the vector of the electric field, respectively, and H is the z-compo-
nent of the magnetic field vector. Equation (6.78), being written in coordinates X
and Y, can be split into two equations:

mðdvx =dtÞ ¼ eEed þ em0 vy H  mvx nm ; (6.79)


m dvy =dt ¼ eE  em0 vx H  mvy nm : (6.80)
330 V.L. Bychkov et al.

(Here, a condition H < 0 is taken into account). Equating the derivative dv/dt to
zero, we can find the velocity components of the settled motion of electrons from
(6.79) and (6.80):

e2 m0 EH þ emEed nm
vx ¼ ; (6.81)
ðmnm Þ2 þ ðem0 HÞ2

e2 m0 Eed H  emEnm
vy ¼ : (6.82)
ðmnm Þ2 þ ðem0 HÞ2

At the beginning of the process, E ¼ 0 and the velocity of electrons’ movement

along the Y-axis to the center of vortex is

e2 m0 Eed H
ðvy Þ0 ¼ : (6.83)
ðmnm Þ2 þ ðem0 HÞ2

Using the relation nm (s1) ¼ 2  109P (Torr) [377], for Eed ¼ 1.26  104 V/m,
H ¼ 105 A/m (H ¼ 0.1 Hmax), and P ¼ 10 Torr (1.33  103 Pa), we can find that
(vy)0 ¼ 5.51  104 m/s. Let the thickness of the vortex wall from which the
electrons’ emission takes place be 2 cm [347, 375]. To pass this way, electron
needs time tpn ¼ 2  102 m/5.51  104 m/s ¼ 0.36  106 s. Finding themselves
in the vacuum region, electrons will drift to the vortex’s axis with a velocity
vdr ¼ Eed/m0H ¼ 105 m/s. Moving with this velocity, they can reach the system’s
axis at a time tdr ¼ R/vdr ¼ 2  106 s.
The motion of positive ions (we shall consider only the lightest of them –
protons) is described by (6.51) and (6.52), where one must change the polarity of
the terms involving the charge of the particle to the opposite, change the mass of the
electron m to mass of the proton mp, and replace nm by nm,p – the rate of protons’
collisions with air particles. By assuming the equality of electrons’ and protons’
temperatures and the equality of cross-sections of their collisions with the gas
particles, we can find that nm,p is in (mp/m)1/2 ¼ 40 times less than nm. Because
of this, one can assume nm,p (s1) ¼ 5  107 P (Torr). At E ¼ 0, Eed ¼ 1.26  104
V/m, H ¼ 105 A/m, and P ¼ 10 Torr, we can find from (6.53) that (vp)y ¼ 58 m/s.
This value is 0.95  103 times lower than the drift velocity of electrons. At time
Dt1 ¼ 105 s, protons can shift only by a distance (vp)y Dt1 ¼ 0.6 mm; thus, they
have no chance to come out to the vacuum cavity from the dense layers of rotating
gas. Hence, because of great difference of drift velocities of electrons and protons at
the first study of process, the separation of charges will take place and the electric
field E, directed along the axis Y, will appear. As the strength of E is increasing, the
drift velocity vy of the electrons is lowered, until according to (6.82), it becomes
equal to zero at

Emax ¼ ðem0 Eed H Þ=ðmnm Þ: (6.84)

6 Ball Lightning Investigations 331

By assuming P ¼ 5  103 Torr (0.67 Pa), Eed ¼ 1.26  104 V/m, and Hmax
¼ 106 A/m, we can obtain Emax ¼ 2.8  108 V/m.
The increase in the radial electric field E must produce protons’ drift along the Y-
axis. According to (6.82), the velocity of this drift at Eed ¼ 1.26  104 V/m, H ¼ 106
A/m, Emax ¼ 2.8  108 V/m, and nm,p ¼ 5  108 s1 is (vp)y ¼ eE/mpnm,p ¼ 5.36 
107 m/s. As a result of this, a part of protons will pass through dense layers of gas
and will come to the vacuum cavity. After reaching the vacuum cavity, the protons
will stop their motion in the direction of E. Under the action of the axial magnetic field
H, vector vp will be turned and the protons will drive in the direction perpendicular to
the vectors E and H. Moving in the field E, protons gain energy, at the same time, field
Eed, directed opposite to proton’s velocity, brakes their motion.
At the stage Dt2 ¼ 50 ms of decreasing magnetic field H, the eddy electric field
vector Eed changes its direction to the opposite. This will favor the acceleration of
protons. At this period, the emission of charges into vacuum cavity will be absent.
As a result of the discussed processes inside the vacuum cavity, separation of
charges and protons’ acceleration will occur. The electrons will be gathered in
the center, and the protons will rotate around the negative center.
An axial magnetic field has a marked effect on the charges’ motion in the plane
XY, but it essentially does not prevent their motion in the direction of axis Z. In
principle, owing to their high mobility, electrons at the period of lightning stroke
are able to move away from the coil axis to larger distances than protons. During
their travel, the electrons may come across positive charges (e.g., in cover of
streamer of lightning [374]) and recombine with them. As a result, the system
will lose a part of negative charge and will acquire excess positive charge.
Because of the described processes inside the vacuum cavity, a double flat ring
of moving charges will be formed: the outer ring of protons (at the excess of one)
and the inner ring of electrons (see Fig. 6.10). Owing to the deformation type of
twisting to eight, this double ring may fall to many small rings. From these rings,
dynamical electric capacitors will be formed. The shell of BL may be gathered from
water containing in the vortex’s wall.

6.4 Ball Lightning as the Thermodynamic Limit of the Periodic

System

In the lightning discharge plasma, we find a regime for negative dielectric response
from microwave transmission through tenuous wire grids. From photonic response
in vortex plasma, we treat BL as a charged eigenstate fully defined by its central
potential, and evolving without change in the radius and brightness. Helical plas-
mon waves in spherical vortex crystals enhance self-confinement by interfering like
worm gears packed with octahedral symmetry. The thermodynamic limit packs
taped tori in concentric gear shells at energies beyond those of electron shells in the
natural elements.
332 V.L. Bychkov et al.

By Madelung’s method for Coulomb interaction in ionic crystals, we derived the

potential energy in periodic arrays of line sources with logarithmic interaction
potentials. For straight rows, our gauge choice ensured convergence to finite and
scale-free lattice sums with lowest energy expressed as Wallis product. For vortex
streets, we obtained a smooth energy surface via the complex plane, extendable to
3D lattices via quaternion space. We presented a matrix formulation of quaternion
inversion that maps vortex flow in a wedge on potential flow driven by an equatorial
vortex row on the surface of a sphere.
We obtain 3D potential fields and flows from quaternion functions satisfying six
local and six global conditions for a smooth closure. For quaternion matrices and
Jacobians, we found real eigenvalues, orthogonal eigenvectors, nonzero determi-
nants, and inverses by transposition. Group structure includes 8D octavian space
resolving into three quaternion sub-spaces, in turn, resolving into three complex
sub-planes.
Our matrix method plots equipotential surfaces in parametric form, orthogonal
to streamlines as an intersection of two complementary sets of stream surfaces. By
quaternion inversion, we calculated the atmospheric circulation patterns as belts,
lattices, and Von Karman vortex streets mapped on the unit sphere. The quaternion
sine function bends the atmospheric electricity into elliptic coordinates for trig-
gered lightning near the ground level or for spark gaps in high-voltage test facilities.

6.4.1 Introduction

Maxwell’s equations endow materials combining negative permittivity and perme-

ability with surprising properties, including reversal of refraction angles, Doppler
shifts, and Cherenkov radiation [378]. Such photonic response is not found
in natural substances, but various designs predicted this behavior for periodic
microstructures made from dielectric or metallic materials, as shown in Fig. 6.13
[379, 380].

a b

Fig. 6.13 Two photonic

microstructures. (a) Square
array of cylinders in microwave
beam. (b) Capacitive ‘Swiss
roll’ design enhances negative
magnetic response
6 Ball Lightning Investigations 333

Meanwhile, several classes of artificial structures called “photonic crystals”

have conclusively demonstrated their negative electromagnetic constitution, rang-
ing from microwave to optical frequencies [381]. Ionospheric plasma provides a
natural example of negative permittivity causing reflection of radiowaves with long
enough wavelength. Here, we will argue for a crucial role of photonics in the
electromagnetic constitution of BL.
Despite two centuries of scientific investigation covered in Singer’s inspirational
monograph, the true nature of BL remains undecided [382]. Smirnov’s comprehen-
sive review combines BL physics with electrochemistry on fractal skeletons [383].
The leading database now stores several thousand eyewitness reports collected in
the Stakhanov–Keul–Bychkov data bank [384, 385]. Their statistical analysis has
fostered consistency and consensus regarding the mean properties of the global
BL population. At the same time, laboratory efforts are still far removed from
reproducing the full range of its observed characteristics [386, 388–390]. Two BL
events reported by professional scientists highlight the features challenging the
conventional plasma theory.
In 1964, an all-metal airliner carrying British radio-astronomer, R.C. Jennison,
encountered an electrical storm on a late night flight from New York to Washington
[391]. Some seconds after a bright and loud discharge, a glowing sphere emerged
from the pilot’s cabin, passing the observer at an arm’s length in steady motion
along the aisle. The blue-white ball measured 22 cm with perfect 3D symmetry, free
of any polar or toroidal structure. Its surface had a solid appearance with some limb
darkening, radiating light at 5–10 W, but not heat. Three years later, the Russian
plasma chemist, Dmitriev, standing on the Onega River bank, witnessed a BL with
a lifetime of over 1 min, following an intense flash of ordinary lightning [382, 392].
The yellow-white core with 6–8 cm diameter of this slightly elongated ball was
covered in violet and blue envelopes measuring 1–2 and 2 cm, respectively.
A transistor receiver recorded its radio emission as a continuous rumble with
maximum loudness even at 65 m from the ball. As it passed overhead, the observer
heard a noise emission with some crackling, and smelled the acrid odor of its bluish
mist trail from high-energy radiation in air. On his gamma scintillation radiometer,
Dmitriev read 1.2 Mrad/h as the dose rate at 2 m from the source, but a photo film in
a holder remained practically unexposed.
Nowadays, carefully screened scintillation counters have established energetic
gamma radiation as a common feature of natural or triggered lightning in Florida,
USA [393]. Previously, BF3 counters had already detected neutron generation by
lightning bolts in the Himalayan foothills much above the background levels from
cosmic rays [394]. For particle acceleration by turbulent vortex dynamics, our helix
string model details a planar version for the lightning channel and a spherical
version for BL [395, 396]. For both the geometries, we included generic formulas
from microwave photonics rescaled to vortex lattices in atmospheric discharge
plasma. Through reversed Coulomb forces, we connected the central concentration
of charge with evolutionary tracks of BL resembling stellar evolution driven by
gravity. Negative dielectrics enable standard balls to consume their energy content
without a visible change at the surface as in Jennison’s case, or explode into sparks
334 V.L. Bychkov et al.

when disturbed prematurely. And in exceptional cases such as Dmitriev’s event,

circulation velocities make fusion reactions of light nuclei as inevitable as in main
sequence starsi such as our Sun [397].
Tenuous metallic wire grids, as used in fast-particle detectors, demonstrate
plasma behavior in microwave propagation analysis and tests reported by Pendry
et al. [398]. Their micron-size wires downscale the plasma threshold for wave
reflection or transmission from optical to gigahertz frequencies. Transmission data
and Maxwell theory agree on photonic band gaps within 5% and negative dielectric
constant as a robust feature of 3D wire arrays, as shown in Fig. 6.14.
Transverse modes incident on wires excite the surface waves turning into
massive plasmons through the Higgs mechanism for the mass of elementary
particles. These would range down to the DC level with wires made loss-free by
superconductivity.
Pre-Higgs models link high-energy storage in BL with plasmons mediating
dielectric attraction between like charges, similar to pions in atomic nuclei [382,
399, 400]. Bergström’s electromagnetic model freezes bare nuclei into a uniform
atmospheric cloud by an immediate dielectric reaction effect [399]. In [400], long
lifetime and high energy storage arise from the bosonic nature of microscopic
vortex filaments in turbulent BL plasma. Early support for high-energy processes
in the air came from lightning-triggered neutron counts recorded over a 3-year
period in the Himalayan foothills [394]. New evidence for such intriguing processes

Fig. 6.14 Cubic thin-wire grid with photonic response to microwave transmission
6 Ball Lightning Investigations 335

is provided by energetic radiation from rocket-triggered lightning and high-energy

gamma rays emitted by thunderclouds [393, 401]. Recent BL data reviewed by
Bychkov revealed significant thermal and electric impact causing fires (body)
burns, melts, and water heating in a quarter of the reviewed events [402]. Nikitin
et al. concluded that the Sochi 1960 photo tracks a charged BL descending towards
a lamppost while hovering during repeated field reversals by a nearby lightning
discharge [403]. Advanced statistical analysis by Keul revealed a time-constant and
space-invariant set of BL phenomena recorded over several centuries in Central
Europe [385].
Comparison of Pendry’s wire grid with vortex plasmas modeling BL and its
laboratory analogs showed common ranges for their critical parameters. An early
ruby laser facility at Philips Research Laboratories in Eindhoven, the Netherlands,
regularly produced such analogs from an open induction coil upon discharge of a
capacitor bank. Data on wire gauges, mesh size, source voltage, and pulse duration
supplied by the operator comply with Plasmon-regime parameters [404]. More
recently, lightning impact formed a bright ball at a tall lamp post supporting traffic
lights and power lines at a T-crossing for trams in Rijswijk, the Netherlands.
A Saturday shopping crowd watched this event for about 1 min, and two of them
recorded the static ball on their digital cameras. Our site investigation revealed
erosive damage on several nearby support wires [405]. Again, estimates for wire
gauges and mesh sizes in the complex wiring structure locate this event inside the
regime for plasmon waves, as shown in Fig. 6.15.
A Japanese test facility for the study of BL generates gigahertz microwaves at
much higher power levels than that generated in Pendry’s transmission tests [386,
398]. Instead of thin wires, ceramic plates cause local field amplification approach-
ing breakdown levels in the air by photonic interference [406]. Current Russian
experiments achieve reproducible fireball formation by erosive discharges, water
surfaces, thin films, and exploding wire rings, with afterglow durations approaching
1 s [388–390, 407, 408].
Function theory of complex variables provides electrodynamics and aerodynam-
ics with a powerful tool for the calculation of planar potential fields [409–411]. Two
Cauchy–Riemann relations keep fields and flows irrotational, and the complex
function, smooth. The level curves of their real and imaginary parts draw equipo-
tential lines and field- and stream-lines at right angles. By swapping these results, a
complementary field pattern with line charges turned into currents, and vice versa
could be obtained. The conformal property preserves the angles in maps by regular
complex functions. A large class of such functions is expressible as polynomials
and power series with real coefficients. In Hamilton’s 4D quaternion space [412],
an earlier study [396] presented smoothness conditions for 3D potential field and
flow solutions, through such functions. By quaternion inversion and exponentials,
another study [413] obtained spherical fields and flows, and cast in real matrix
format for the three division algebras [414]. Underneath, we applied this format for
modeling global atmospheric circulation and pre-stroke potential field calculation
for rocket-triggered lightning.
336 V.L. Bychkov et al.

1.5

1.0

ε(ω)
ε0 ε(ω) ε2 / ε2p – 1
=
ε0 ε2 / ε20 – 1
0.5

0.0
ωp /ω0
→

–0.5
ω
→
ω0

–1.0
0 1 2 3 4 5

Fig. 6.15 Generic formula for dielectric response turns negative in the shaded frequency range.
Replacing e by m gives the equivalent formula for magnetic response

6.4.2 Photonic Response

Our bosonic model treats BL as cool, dense plasma threaded by microscopic vortex
filaments in coherent rotation at the quantum limit for electron circulation. Its
conduction electrons primarily interact by exchange forces as in alkali metals.
Ambient vorticity merges them into a superfluid ground state around hollow vortex
cores also seen in cavitation of normal fluids. Coulomb forces dress electronic cores
in a co-rotating ion mantle restoring overall charge neutrality around vortex loops.
This core-mantle structure confines electromagnetic fields between conducting
cylinders as in coaxial waveguides. Their elastic stretching, bending, and torsion
extend reversible energy storage to the relativistic regime [397]. Here, we show that
local contraction of their cross-section nicely fits tapered vortex rings into a
spherical potential solution for BL.
Let a threshold electric field Eth as in (6.85a) dissociate air molecules into equal
numbers of electrons, ions, and neutral particles in the cavitation regime of vortex
plasma [397]. With uniform current I in the lightning channel, the field lines of
magnetic induction B encircle discharge plasma by Ampère’s law, as in (6.85b).
Screening by electrons in loss-free layers restricts magnetic fields to the London
penetration depth lL, as in (6.85c). Let us consider the cavitation thread quantum
6 Ball Lightning Investigations 337

units of flux as hollow tori through the plasma. Inertial mass of vortex rings comes
mainly from entrainment of ambient fluid, resulting in migration times as in
(6.85d). Together, these relations constitute our vortex plasma through:

ne2
Eth ¼  23 MV/m; (6.85a)
80pe20 Vi

m0 I
B¼  0:4 T; (6.85b)
2p R
rffiffiffiffiffiffiffiffiffiffiffiffi
m
lL ¼  1:1 mm; (6.85c)
m0 ne2

4peEth R2
t¼  230 ms; (6.85d)
rk3

with fundamental constants e  1.6  1019 C, electron mass m  9.11  1031 kg,
e0  8.85 pF/m, m0 ¼ 400p nH/m, and k ¼ 7.3  104 m2/s as quantum unit for
electron circulation. As material parameters, we used mass density r  1.3 kg/m3 of
air and ionization potential Vi  14 V of nitrogen. At 50% ionization, our electron
concentration reached n  2.5  1025 m3, with peak current I ¼ 105 A and
channel radius R ¼ 0.05 m as discharge parameters. As the length and timescales
for field gradients, l and t in (6.85c) and (6.85d) are set as conditions for photonic
response to electromagnetic waves in our vortex plasma.
Microwave transmission through Pendry’s cubic wire grid demonstrates nega-
tive dielectric response at gigahertz frequencies [380]. His micron-sized wires
reduce mean electron concentration by a factor of 106 relative to solid metal, and
raise the canonical electron mass to baryon values. Thus, his grid cuts off micro-
wave propagation below the plasma frequency at 8.2 GHz, where dielectric
response turns negative. By generic formulas for dielectric response, loss-free
waves still propagate along wire surfaces in between resonance at o0 and plasma
frequency op, as shown in Fig. 6.15. More refined microstructures in Fig. 6.13 show
negative magnetic response for microwaves, confirming Veselago’s prediction of
refraction at reversed angles from Snell’s law [378, 381].

6.4.3 Charged Eigenstates

Superconducting wires in Pendry’s grid will extend negative dielectric response down
to the static limit, where plasmon wavelength merges with London penetration depth.
Their longitudinal and transverse modes share dispersion relations with elementary
particle models creating Higgs bosons in a vacuum [380]. By an attractive dielectric
reaction between like charges, Coulomb forces cause self-confinement of nuclei and
338 V.L. Bychkov et al.

BL in Bergstrom’s electromagnetic theory of strong interaction [399]. For spherical

charge distributions, we derived smooth potential profiles via conformal mapping in
quaternion space [396]. Fixed radius and net charge serve as boundary conditions
recasting these profiles into charged eigenstates of BL with the central potential as
an evolutionary parameter. As the conserved and observable surface field strength,
the threshold electric field in (6.85a) sets reference values for other electric-state
variables.
Most lightning strokes carry negative charge down to earth, but positive strokes
are more powerful on the average. The 4D quaternion space inverts a uniform pre-
stroke field into a spherical potential function F w/(w2 þ r2), parametrized by
w for its evolution [396, 400]. Equation (6.86a) shifts its zero from infinity to r ¼ R
at the surface, with central potential FC as the reference value. From standard
vector relations, the profiles for electric field E and charge density r with their
reference values ES and rC are as follows:
 2
F w 2 R2  r 2 E r w2 þ R2 r w4 3w2  r 2
¼ 2 2 ; ¼ ; ¼ ;
FC R w þ r2 ES R w2 þ r 2 rC 3 ðw2 þ r 2 Þ3
(6.86a)
   2
1 R2 ne2 MV 3e0 ES R2 C
FC ¼ ES R 1 þ 2 V, ES ¼ 2
 2:3 ; r C ¼ 1 þ 2
;
2 w 80pe0 Vi m R w m3
(6.86b)

with minus signs denoting a negative well potential and inward field directions
from negative charge density balanced by positive surface charge. In (6.86b), the
central potential and charge density evolve via w without change in the surface
potential and field strength. Figure 6.16a and b plot successive curves meeting in
F(R) ¼ 0 for potentials, and with a common beginning and end for the electric

a b c
Potential Electric field Charge density
0.0 0 0
w=1

–10 0.7
–0.5 w=1 w=1
0.5
–1
0.7 –20
0.7 0.4
–1.0 0.5
0.5 –30
0.4 –2
0.4 –40 w=0.3
–1.5
w=0.3 p
E –50
Φ –3 pc
–2.0 Es
Φc r w=0.3 r –60 r
R R R
–2.5 –4 –70
0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2

Fig. 6.16 Electric profiles in ball lightning (BL) plasma evolve without change in radius and net
charge. (a) Potential curves decay smoothly towards ground level. (b) Photonic vortex plasma
requires field strength below the dashed line. (c) Flattening charge density curves release energy
for radiative output
6 Ball Lightning Investigations 339

field. Each charge density curve plotted in Fig. 6.16c integrates to the same net
charge by volume, but they lack a common point.
Steep initial profiles shown in Fig. 6.16 gradually flatten as BL evolution
increases w from an initial value below 1. Small w values concentrate space charge
near the center of a deep potential well, much like what the nucleus does in the
atoms, but with opposite polarity. A proportional rise in electric-field strength will
increase the initial ionization levels above single ionization as used in (6.85a). The
top curve in Fig. 6.16b defines our final stage of BL, where photonic response just
vanishes. The bottom curve defines an initial stage with a compact core region
dressed in a wide photonic envelope. Successive curves show that its thickness and
peak field diminish towards the surface. Chemical reactions or atomic Rydberg
states can increase the life span by releasing additional energy [387, 401–403], and
likewise, for light nuclei when central voltages in (6.86b) enter the regime for their
fusion reactions [395, 397].

6.4.4 Coherent Rotation

For dynamic effects in vortex plasma, we need much stronger photonic response
than tenuous wire grids shown in microwave transmission experiments. Hence, we
seek compact vortex structures with constructive interference of their photonic
plasmon waves. These structures will serve as a unit cell for periodic vortex lattices
extending to infinity in the thermodynamic limit. By the inversion method used for
electric profiles in (6.86), we can map various lattice structures onto a spherical
vortex state conserving their photonic response. These inversion maps straighten
tube lattices into tapered vortex loops packed in concentric shells with geometries
resembling electron orbitals in the periodic system of elements. Nature allows
stable atoms with at most 92 electron orbitals, but the loop count provides a free
parameter for energy storage in such spherical vortex crystals.
We used coherent rotation of various gear wheel combinations as visualization
for constructive interference of photonic plasmon waves on vortex tubes. With gear
wheels packed on spheres, as in Fig. 6.17, coherent rotation visualizes constructive
interference of plasmon waves propagating along their outer edges.
On each sphere, units of four or six connect wheels in even numbers are required
for coherent rotation. Spherical combinations connecting gear wheels in odd num-
bers must remain static. Clearly, the same distinction between static and dynamic
grip applies to planar gear-wheel combinations.
The Euclidean plane packs identical gear wheels most densely in triangular unit
cells, but coherent rotation requires square-based or hexagonal lattices. Spherical
surfaces admit just one fully regular co-rotation pattern, packing eight wheels in
units of four in Cartesian octants, as shown in Fig.6.17a. Vertices of the other four
Platonic solids have threefold or fivefold symmetries that block co-rotation. Out of
13 Archimedean solids, 3 trace out semi-regular co-rotation patterns. The truncated
octahedron, shown in Fig.6.17a, copies the partition of its Platonic parent. Figure
340 V.L. Bychkov et al.

Truncated octahedron Zonohedron Small rhombicuboctahedron

Red : blue=4:4 Red : blue=6:8 Red : blue=6:12
a b c

Fig. 6.17 Some (semi-) regular solids and some zonohedra partition of a spherical surface for
coherent rotation. (a) Octahedral symmetries connect eight gear wheels in six loops of four with
balanced 4:4 ratio of spin sense; wheel centers define the edges of a cube. (b) Unbalanced 6:8 spin
ratio results from octahedral symmetries connecting 14 gear wheels in 12 loops of 4; the wheel
centers define a rhombic dodecahedron. (c) Unbalanced 6:12 spin ratio results from octahedral
symmetries connecting 18 gear wheels in eight loops of 6; a 45 twist of the top section restores
spin balance by changing the Archimedean partition into the Miller solid

6.5c packs 18 wheels in 8 units of 6. As a third Archimedean solid, the snub cube
partitions spherical surfaces for co-rotation of 32 wheels in 6 units of 12. The wheel
count sequence 8, 18, 32 obeys the 2n2 formula counting electron orbitals in
successive shells of noble gas atoms. The l ¼ 3, m ¼ 2 spherical harmonic solution
of Schrödinger’s wave equation traces out the octant partition in Fig. 6.17a by its
zero lines, but the other two have no such connection. While atomic physics gives
most electron states a non-zero orbital angular momentum, vortex tubes have none
in the static co-rotation patterns of Fig. 6.17.
Remarkably, the 18 gear wheels on the Archimedean solid in Fig. 6.17c show a
2:1 spin imbalance absent from the Miller solid produced by a 45 twist of its top
section. At a slightly lower zonohedral level of symmetry, co-rotation with a 4:3
spin imbalance appears with 14 wheels packed in 12 rhombic units of 4, as shown in
Fig. 6.17b. In the periodic system, 14 is the number of M-shell electrons in iron
atoms, distinguished by their magnetism, nuclear binding energy, and complicated
optical spectrum.

6.4.5 Vortex Crystals

On vortex tubes in 3D Euclidean space, we visualized photonic plasmon waves by

rotating worm wheels, as shown in Fig. 6.18. We constructed regular gear units with
cell faces at right angles to worm-wheel axes. Axis directions of a twin worm pair
have two bisectors that define the complementary values for pitch angles with
opposite sign, as shown in Fig. 6.18.
6 Ball Lightning Investigations 341

a b c
tan a = –1 1 ½ 2 ½ 2
a = –45° 45° –35° 55° –27° 63°

Fig. 6.18 Helical plasmon waves on vortex tubes look like the surface of rotating worm wheels.
Dihedral angles |2a| in (a) cubic, (b) octahedral, and (c) dodecahedral cells determine the screw
pitch tana

4
∞, 2a = 90° Dynamic
tan 2a =∞ tan 2a= 8, 2a » 71° Dynamic tan 2a = 3, 2a » 53° Static
a b c

Fig. 6.19 Three Platonic solids stack gripping worm wheels normal to cell faces at dihedral angles
2a. Symmetry admits co-rotation in (a) cubic and (b) octahedral lattices, but not in (c) dodecahe-
dral cells

For strong photonic response by constructive interference between plasmon

waves, we produced a coherent rotation of worm gear units, extendable to an
infinite lattice in the thermodynamic limit.
From their face and vertex symmetries, we labeled the cubic cells by {4, 3},
octahedral cells by {3, 4}, and dodecahedral cells by {5, 3}, as in Schl€afli’s {p,q}
symbol [415]. Figure 6.19a shows the left-handed worm wheels packed in right-
handed triplet units centered on a body diagonal to the cubic unit. This left–right
combination allows co-rotation, because its triplet units have continuous normal
velocities in all three tangent points, and likewise, for right–left triplets. Conversion
to either right–right or left–left triplets blocks co-rotation in the unit cell and
throughout the vortex lattice. Compactness doubles when helicities are packed
alternately in a filled-up static version.
The octahedral {3,4} cell shown in Fig. 6.19b packs worm wheels in triplet units
by face symmetry and in quartet units by vertex symmetry. Twelve large-pitch and
left-handed worm wheels are packed in right-handed triplet units and left-handed
342 V.L. Bychkov et al.

quartets, resulting in overall co-rotation. This dynamic grip will survive inversion
to right-handed worm wheels packed in left-handed triplets and right-handed
quartets. Other conversions give static grip at best. In the dodecahedral {5,3} cell
shown in Fig. 6.19c, vertex symmetry suffices to pack worm wheels at equal angles
in large and small triplet units. Here, large pitch allows co-rotation in large triplets,
but blocks co-rotation in small triplets, reducing the unit to static grip. The same
holds for the same cell packing thinner wheels in quintets by face symmetry in
addition to triplets by vertex symmetry.

6.4.6 Tapered Torus Shells

The 4D inversion map used for our charged eigenstates in (6.86) also bends
cylinders into toroidal tubes with variable cross-section. At infinite cylinder length,
the torus end points converge to the center of inversion, where cross-sections shrink
to zero. In general, any torus mapped from an unbounded vortex lattice converges
on this center as a tapered vortex ring. Conversely, cylinder sections nearest to the
center of inversion end up farthest and largest on image tori. Tubes at equal distance
from the center of inversion define a circumsphere by their outermost image points.
A smaller concentric circumsphere for end points of tube images encloses a vortex-
free core region. Together, these circumspheres enclose a photonic plasma enve-
lope threaded by toroidal vortex tubes, as shown in Fig. 6.20.
The octahedral unit of Fig. 6.19b packs 12 worm wheels with midpoints at equal
distances from its center. Their inversion to a spherical gear packs 12 tapered worm

Dynamic grip Static grip

a b

Fig. 6.20 Tapered toroidal tubes retain symmetry, helicity, and grip of worm gear units. (a)
Octahedral {3, 4} grip enhances photonic response of plasmon waves in gear sphere. (b) Dodeca-
hedral {5, 3} grip weakens photonic plasma response of 30 toroidal tubes
6 Ball Lightning Investigations 343

tori, as shown in Fig. 6.20a. Clearly, this inversion preserves {3, 4} face and vertex
symmetries. As an improper map, it also reverses pitch angles and packing heli-
cities, preserving dynamic grip of gear wheels and photonic response in vortex
lattices. Likewise, for dodecahedral units, packing six worm wheels as shown in
Fig. 6.19c, will preserve their static grip upon inversion onto tapered tori. Static grip
also survives for 30 worm wheels packed in a dodecahedral cell, with {5, 3} face
and vertex symmetries clearly visible in the image sphere of Fig. 6.20b.
Periodic vortex lattices pack the central unit in copies weaving successive tube
layers around it. In cubic lattices shown in Fig. 6.19a, parallel tubes spin with
alternating sense in three square-based grids. More compact octahedral lattices,
shown in Fig. 6.19b, align vortex tubes in four triangular grids. Successive tube
layers invert to concentric torus shells inside the circum-sphere of Fig. 6.20a. As
cross-sections at end points approach zero near the center, their circulation speeds
up for the conservation of angular momentum. By this vortex acceleration, ions
gain more kinetic energy than electrons in proportion to their masses. Circulation in
the lowest Bohr orbit gives light nuclei enough energy for fusion reactions [397].

6.4.7 Thermodynamic Limit

The thermodynamic limit of Pendry’s wire grid fills space with identical cubic cells
extending to infinity in all directions. Its wire directions are either parallel or cross
at right angles by alignment with cube edges. Such unbounded straight wires
interact by logarithmic potentials for electric or magnetic forces. Hence, their
interaction energy diverges to infinity when moving two parallel wires apart
without bound. Grid response with negative epsilon follows Maxwell theory
given in [398] gauged for noninteracting wire currents. Here, we included (Cou-
lomb) interactions between (charged) vortex tubes, as Madelung sums do for
spheres packed in ionic crystals.
The logarithmic potentials also hold for hydrodynamic interaction between
straight and parallel vortex lines in potential flows. These arise as twin vortex–
anti-vortex pairs shed by thin wires moving through fluid at Reynolds numbers near
transition from laminar to turbulent flow [395]. Von Karman vortex streets arrange
such wake flows of source–sink pairs into rows with alternating polarity, as shown
in Fig. 6.21.
Planar vortex lattices also form in liquid helium by rotation of its container, and
in one-component electron plasmas inside magnetic traps [396]. Coherent 3D forms
of vortex lattices arise in tenuous Bose–Einstein condensates of ultra-cold atoms.
For straight-line sources [397], 3D vortex crystals with symmetries and topologies
for minimal dissipation of their circulation patterns, were constructed.
For a stable equilibrium, all systems must settle into states with minimum
energy. In ionic crystals, Coulomb’s law arranges ions as hard spheres in rows
with alternating polarity. Their inverse-square interaction profile ensures conver-
gence to finite row energies by the Leibniz criterion for alternating series. Such
344 V.L. Bychkov et al.

Fig. 6.21 Von Karman vortex

streets align alternating source
polarities in straight, staggered,
or symmetric rows

series converge to finite sums when absolute values of their terms decrease monot-
onously to zero. Underneath, we derived corresponding energy sums for logarith-
mic potential profiles, showing their convergence to finite-row energies. Our
dimensionless form scales potential sums relative to nearest-neighbor interaction
as a Madelung lattice constant for wire grids and vortex lattices.

6.4.8 Wallis Product

Let a reference distance d separate adjacent vortices with opposite circulation in an

unbounded straight row, as shown in Fig. 6.13. Each vortex finds itself halfway
between identical vortex pairs at successive distances d, 2d, 3d, 4d, . . . exerting
6 Ball Lightning Investigations 345

potentials 2 ln (d), 2 ln (2d), 2 ln (3d), 2 ln (4d), . . . . Combining successive loga-

rithms readily gives a scale-free potential series 2 lnð1=2Þ; 2 lnð3=4Þ; 2 lnð5=6Þ; . . . with
terms that converge monotonously to zero without alternating signs. However, the
closely related alternating series lnð1=2Þ; lnð3=2Þ; lnð3=4Þ; lnð5=4Þ; . . . does meet both
Leibniz’ conditions for convergence. Hence, we secured finite potential sums for
alternating line rows by defining their Madelung constant a through:

X
1
a2 ð1Þnþ1 lnðndÞ  lnðdÞ  F1  lnðdÞ; (6.87)
n¼1

where ln(d) shifts the row potential F1 by a gauge transformation without the
consequence for classical electromagnetic fields [410]. This achieves a convergent
and scale-free potential sum for the straight vortex row, expressed in the classical
Wallis product through:

F1  lnðdÞ ¼ lnð1dÞ  2 lnð2dÞ þ 2 lnð3dÞ  2 lnð4dÞ þ


(6.88a)

   
13 35 57 79 2
a ¼ ln






¼ ln (6.88b)
22 44 66 88 p

showing sign alternation in (6.88a), and monotonous decrease to zero for absolute
values jlnð1=2Þj; jlnð3=2Þj; jlnð3=4Þj; . . . in (6.88b). Thus, we can find a finite Made-
lung constant for the straight vortex row, much like corresponding ion rows
summing to 2ln(2) via the alternating harmonic series 1, 1/2, þ 1/3, 1/4, . . .,
also named as Brouncker series.
Inside the unit cells, we may shift the positive (or negative) vortices horizontally
without affecting the convergence of their potential sums. For a row with period 2,
two complementary distances a and 2  a give the potentials lnð1  a=2Þ;
lnð1 þ a=2Þ; lnð1  a=4Þ; lnð1 þ a=4Þ; . . . ; which for 0 < a < 1, continue to meet the
Leibniz conditions. Likewise, for rows with period 2d, complementary distances ad
en (2  a)d between (next-)nearest neighbors rearrange and sum to a Madelung
formula:
 
ð2  aÞð2 þ aÞ ð4  aÞð4 þ aÞ ð6  aÞð6 þ aÞ
a ¼ ln






22 44 66
 
sinðap=2Þ
¼ ln ; 0 b a b1 (6.89)
a p=2

which is easily seen to reproduce the Wallis product in (6.88b) by the substitution of
a ¼ 1. Moreover, when vortices and anti-vortices coincide and cancel, the limit
a ! 0 reduces a in (6.89) to zero by the standard limit sin(x)/x ! 1 for x ! 0. Thus,
(6.89) ensures a smooth transition from zero energy in fluid at rest to negative
energies when vortex–anti-vortex separation increases from 0 to 1 for the straight
346 V.L. Bychkov et al.

row in Fig. 6.21. This contrasts with ionic crystals, where ions cannot neutralize
their charges by merging, zero energy requires infinite separation between all ions,
and bound states have negative energy relative to infinity.

6.4.9 Vortex Streets

What remains to be done is translations in other than horizontal directions, as shown

in Fig. 6.21. This can be done by expressing potential flow patterns around line
vortices in complex coordinates and functions, scaled to agree with the previous
section. Accordingly, stream function and potential around a line vortex at the
origin follow as real and imaginary parts of a logarithmic complex potential z(z)
¼ ln((x + Iy)p/2). Likewise, a compound potential function ln(sin(z)) creates a
periodic row of identical vortices along the x-axis with period 2. By sign reversal,
the potential ln(sin(zp)) creates a complementary row of anti-vortices, shifted
over p ¼ (a + Ib)p/2 in arbitrary directions and distances. In this way, setting a ¼ 1
and b ¼ 0 recovers the straight row, with lowest Madelung sum as given in (6.88a).
For general Von Karman vortex streets, we combine the vortex and anti-vortex
rows in a single complex potential z as:
   
sinðzÞ sinðpÞ
B ¼ ln ; 
a ¼ ln  (6.90a, b)
sinðz  pÞ p 

Periodic units of their streamline patterns are plotted in Fig. 6.10a–c as level
curves of the real part from (6.90a). The straight row pattern in Fig. 6.22a stands out
by the near absence of open streamlines.
For Madelung sums at complex vortex–anti-vortex separation, we rely on sum
limits returned by our Mathematica program. For the symmetric Von Karman street
shown in Fig. 6.22a, an imaginary separation vector Ib replaces a in (6.89). Writing
sin(Ib) ¼ I sinh(b) gives ln(2 sinh(p/2)/p)  0.382 as positive Madelung sum for
the symmetric row with b ¼ 1, shown in Fig. 6.22b. As stated earlier, merging
vortices reproduce the zero level of Madelung energy, now by the limit sinh
(x)/x ! 1 for x ! 0. For the staggered vortex street in Fig. 6.22 with complex p,
the lattice sum re-arranges into a Madelung formula given in (6.90b) with a ¼ ln
(2/p))  0.452 as the lowest value for vortex–anti-vortex separation p ¼ p/2.
When compared with (6.90a), the general potential sum of (6.90b) differs only by
its modulus bars that make sure that a is real for any complex vector p.
The Madelung surface, shown in Fig. 6.23, illustrates a saddle point on the left,
uniting two separatrix lines that trace out the zero energy level of fluid at rest.
Our lattice sum for a converges for separation vectors in a vertical strip between
a ¼ 0 (solid line) and a ¼ 1 (dashed line) in Fig. 6.23. In this domain, the straight
row, shown in Fig. 6.22, serves as an attractor for other configurations. Instead of
neutralization by vortex–anti-vortex recombination, gradients of our Madelung
6 Ball Lightning Investigations 347

Fig. 6.22 Streamlines for straight, staggered, and symmetric vortex streets from Eq. (6.90a)

surface, shown in Fig. 6.23, favors evolution of turbulent systems towards a bound
state resembling a vortex crystal.
The staggered row, shown in Fig. 6.22b, enables construction of an unbounded
square-based vortex lattice by periodic vertical translations, and similar oblique
translations do the same for the two other rows. Now, the resulting flow pattern is
periodic in two directions, and elliptic functions will replace sine functions in the
complex potential of (6.88). Thus, the same replacement in (6.89) suggests an
expression for the corresponding Madelung formula.

6.4.10 3D-Lattice Sums

For corresponding 3D-lattice sums, we can preserve the great convenience of com-
plex coordinates via Hamilton’s 4D quaternions [412]. By his definition, quaternions
q combine a scalar real part w with a vector imaginary part Ir ¼ ix þ jy þ kz, with
four real variables r, x, y, z, and four imaginary units I, i, j, k connected by:
348 V.L. Bychkov et al.

Fig. 6.23 Level curves for Madelung energy surface of vortex streets

q ¼ w þ ix þ jy þ kz; i2 ¼j2 ¼k2 ¼  1; ijk ¼ 1 (6.91a)

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
q ¼ w þ Ir; I = (ix þ jy þ kzÞ=r; r ¼ x2 þ y2 þ z2 (6.91b)

with r as norm of the imaginary part ix + jy + kz for short. The i, j, k units readily
follows i ¼ jk, etc., as in outer products of Cartesian unit vectors. For this quater-
nion algebra, we give an equivalent matrix formulation, which is more convenient
for the calculation of potential flows with Mathematica.
Construction of polynomial and other standard functions in quaternion space
only requires addition and multiplication formulas valid for complex functions.
Thus, Von Karman vortex street potentials as given in (6.91a) separate into real and
imaginary parts for their sine and logarithmic functions as:
6 Ball Lightning Investigations 349

a b c

Fig. 6.24 Three unit cells stack identical cylinders with axes at equal angles and distances;
(a) left-handed vortex trio at normal angles; (b) right-handed quartet at tetrahedral angles;
(c) quartet with mixed helicities

sinðqÞ ¼ sin(w þ IrÞ ¼ sinðwÞ coshðrÞ þ I cosðwÞ sinhðrÞ (6.92a)

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
lnðqÞ ¼ ln(w þ IrÞ ¼ lnðmÞ þ Iy; y ¼ arctanðr=wÞ; m ¼ w2 þ r 2 (6.92b)

with m as the norm of the full quaternion w þ Ir and angle y for its argument.
Hereby, the potential flows and Madelung formulas for our vortex streets acquire
4D freedom in quaternion space. For 3D lattices, an earlier study [398] developed a
cubic unit cell with three orthogonal line sources, as shown in Fig. 6.24a. The unit
cells shown in Fig. 6.24b, c stack four line sources at equal angles, like diagonals
inside a cube or valences of carbon atoms. Reference [411] classified symmetries,
topology, and compactness of their respective lattices.

6.4.11 Regular Quaternions

Parametric plots of potential flows need explicit formulas for z  x + Iy in terms of

complex potentials z  F þ IC, with potential F and stream function C as
parameters. A single vortex–anti-vortex pair with complex potential z ¼ ln
((z + 1)/(z  1)) turns into a circular row of four pairs upon substitution of z4 for
z. In this way, we obtain the complex potential for vortex circulation confined by a
wedge, and its inverse function as:
4   
z þ1 1 þ expðF þ ICÞ 1=4
z  F þ IC  ln 4 ; z  x þ Iy ¼ (6.93a, b)
z 1 1  expðF þ ICÞ

The planar part of Fig. 6.25 shows one of the eight sectors in the circular vortex
row with flow pattern as parametric plot of the inverse complex potential in (6.93b).
350 V.L. Bychkov et al.

Fig. 6.25 Vortex flow pattern

inside p/4 wedge mapped on
sphere by quaternion inversion

In 4D quaternion space, an earlier study [416] connected position vectors q ¼ w

+ ix + jy + kz with square matrices Q having q as leading column, as written in
(6.94a, b). Matrix products of Q with four vectors copy the results of quaternion
algebra. Q’s determinant (w2 + x2 + y2 + z2)2 ensures inverse matrices for all
nonzero matrices Q, similar to what q’s norm (w2 + x2 + y2 + z2) does for the
vectors. Quaternion space splits functions f(q) ¼ W + iX + jY + kZ into four real
functions W, X, Y, Z of four real coordinates w, x, y, z. Their 16 partial derivatives
Ww, Wx, . . ., Zz enter the Jacobian matrix J with quaternion derivative df/dq as
leading column through:

0 1 0 1
w w x y z
BxC Bx z y C
B C B w C
q  B C; QB C;
@yA @y z w x A
z z y x w
0 1 0 1 0 1
Wðw; x; y; zÞ Ww Wx Wy Wz Ww
B Xðw; x; y; zÞ C BX Xz C BX C
B C B w Xx Xy C df B wC
f ðqÞ  B C; JB C; B C:
@ Yðw; x; y; zÞ A @ Yw Yx Yy Yz A dq @ Yw A
Zðw; x; y; zÞ Zw Zx Zy Zz Zw
(6.94)
6 Ball Lightning Investigations 351

By selective removal of rows and columns from matrix Q, we can recover the
corresponding matrix pairs for three different complex planes with axes (w, x),
(w, y), and (w, z). Each plane has a 2  2 J-matrix imposing one off-diagonal
Cauchy-Riemann-condition, collected in the gradient column of (6.95a). Plane-
wise, the second Cauchy–Riemann condition emerges from the eigenvalues column
in (6.95). Determinant and transpose of Q together yield its inverse matrix. Q’s
eigenvalues are quaternion vector q and its conjugate, both twice. The moment
equation is indispensable for inductive proofs of all 12 conditions in (6.95c) covering
quaternion polynomials and power series with real coefficients:

0 1 0
Gradient 1 0 1 0 1
Rotation
Xw Wx Yz Zy
@ Yw A ¼ @ Wy A; @ Zx A ¼ @ Xz A;
Zw Wz Xy Yx
(6.95a, b, c, d)
0 1 0 1
Moment 0 Eigenvalues
1 0 1 0 1
yZ zY Xx Xy Xz x x
@ zX A ¼ @ xZ A; @ Yx :
Yy Yz A @ y A ¼ Ww @ y A
xY yX Zx Zy Zz z z

We find explicit coordinate dependence in six global conditions, and six local
conditions without it. For smooth closure in quaternion space, they leave
1612 ¼ 4 partial derivatives needed for uniqueness of quaternion derivative
given by (6.94). Gradient entries anti-symmetrize the leading row and column of
matrix J, and rotation entries symmetrize the remaining sub-matrix S, linking their
determinants as follows:

Ww2 þ Xw2 þ Yw2 þ Zw2 fw; x; y; zg

Det(JÞ ¼ Det(SÞ; fW; X; Y; Zg ¼ 2 :
Ww w þ x 2 þ y 2 þ z2
(6.96a, b)

The vector expression (6.96b) for quaternion inversion f(q) ¼ 1/q satisfies all
smoothness conditions (6.95). It enables conformal maps of planar potential solu-
tions on spherical potential solutions by stereographic projection as shown in Fig.
6.25 for wedge flow.

6.4.12 Matrices for Quaternion Maps

For 3D potential solution by quaternion calculus in real matrix form, we defined

three elementary matrices that reproduce the algebra of Hamilton’s imaginary triple
i, j, k in (6.91). We obtained a unit matrix E0 and three elementary matrices E1, E2,
E3 as partial derivatives of quaternion matrix Q with respect to w, x, y, and z as:
352 V.L. Bychkov et al.

0 1 0 1
1 0 0 0 0 1 0 0
B
@Q B 0 1 0C B
@Q B 1 0 C
0 C 0 0 C
B C  E0 ; ¼B C  E1 ;
@w @ 0 0 1 0A @x @ 0 0 0 1 A
0 0 0 1 00 1 0
0 1 0 1
0 0 1 0 0 0 0 1
B 1C @Q B C
@Q B 0 0 0 C B 0 0 1 0 C
¼B C  E2 ; ¼B C  E3 :
@y @ 1 0 0 0A @z @ 0 1 0 0 A
0 1 0 0 1 0 0 0
(6.97a--d)

The elementary E’s in (6.97a–d) and identity matrix E0 copy all product –rules
E12 ¼ E0, E1E2 ¼ E3> E2E1, etc. of Hamilton’s imaginary triple i, j, k in (6.91).
Using these real matrices simplifies linear algebra for 3D potential solutions and
programming for their plots.

6.4.13 Atmospheric Circulation

Geometrical inversion maps circles on circles and spheres on spheres. Lines and
planes turn into circles and spheres passing through the center of inversion.
Stereographic projection, as shown in Fig. 6.25, maps planar patterns conformally
on a spherical surface, or vice versa in cartographic applications. We reproduce this
map as parametric plot of the inverse quaternion function f(q) ¼ 1/q defined in
(6.96b). Let a potential F and a stream function c serve as independent variables.
Substitute the wedge flow functions f (F, c)  x2 and g(F, c)  y2 from Eq. (6.93a)
in (6.96a, b) for x and y, and remove the z-component. Let parameter w take its
place by cyclic permutation of remaining 3D coordinate axes (w, x, y). Our inverse
quaternion map on a unit sphere now reads:

w ¼ 1; x ¼ f ðf; cÞ; y ¼ gðf; cÞ; z ¼ 0; (6.98a)

ðx; y; wÞ
ðX; Y; Z Þ  ; 0 b c b 2; 0 b f b 4; (6.98b)
w2 þ x2 þ y2 þ z2

with flow parameters F and c in domains as for vortex circulation inside the wedge
shown in Fig. 6.25. An eightfold rotation about the z-axis completes the spherical
flow pattern shown in Fig. 6. 26b as an equatorial vortex–anti-vortex belt.
Inversion maps by (6.98a); (6.98b) apply to arbitrary planar flows. Multiple
source or vortex combinations add up as in linear vector spaces. Parametric plots of
a complex potential (F, c) use its inverse function as coordinates (x, y) in (6.98b).
Thus, confined wedge flow of source–sink doublets maps into spherical versions of
6 Ball Lightning Investigations 353

Fig. 6. 26 (a) Streamlines close locally in octahedral vortex lattice. (b) An equatorial vortex row
closes its stream-lines locally. (c) Symmetric vortex street sends equatorial circulation around the
globe. (d) Staggered vortex street gives ‘wavy’ equatorial circulation

some classical planar potential flows. Equation 6.98 maps the regular octahedral
vortex lattice in Fig. 6.26a from a vortex–anti-vortex doublet inside a p/2 wedge.
Polarities alternate between neighbors, and each streamline encloses just one
singularity, as in the vortex belt in Fig. 6.26b. Opposite polarities in the upper
and lower hemispheres of Fig. 6.26c, d send equatorial flows around the globe.
They have planar equivalents in the symmetrical and the staggered Von Karman
vortex street, respectively.

6.4.14 Triggered Lightning

Reference [393] reported about lightning discharges triggered by grounded copper

wires carried into overhead thunderclouds by rockets launched from a platform in
Florida, USA. Nearby scintillation counters consistently recorded energetic emis-
sions that challenged conventional models for breakdown in air by lightning. With
high voltage along the vertical wires above the ground level, their electric field
matched the standard spark gap configurations as used in high-voltage technology.
Currently, scintillation counters also record similar energetic radiations from atmo-
spheric spark discharges in controlled laboratory conditions.
Parametric plots of complex sine functions draw field patterns for planar spark
gaps in elliptic coordinates. For 3D parametric plots, we can rewrite the quaternion
sine function in (6.92a) with potential and field parameters F, r, s, t rescaled for
convenient notation and computation as:
0 1 0 1 0 1 0 1
W sinðwÞ coshðrÞ w fp=15
B X C B x cosðwÞ sinhðrÞ=r C B x C B r=2 C
B C¼B C B C¼B C
@ Y A @ y cosðwÞ sinhðrÞ=r A; @ y A @ s=2 A: (6.99a, b)
Z z cosðwÞ sinhðrÞ=r z t=2

From 4D quaternion space in (6.99a); (6.99b) we remove the stream function t

and its rescaled substitute z. A 3D coordinate frame (X, Y, W) serves for parametric
354 V.L. Bychkov et al.

Fig. 6.27 (a) Quaternion sine plots hyperbolic equipotential surfaces around spark gap. (b) Field
surfaces are orthogonal to equipotential surfaces in (a). (c) Complementary field surfaces branch at
the gap electrode, as in (6.99b). (d) Combination of potential and field surfaces around a vertical
spark gap

plots of the potential function F and two remaining stream functions r and s.
Substitution of F ¼ 0 gives W ¼ 0 for the surface of zero potential shown in
Fig. 6.27a as XY-plane for a vertical spark gap scaled to unit length at unit vector
e2. The real function cosh(r) has no zeroes and increases monotonously from unity
in the first quadrant, and likewise, for sinh(r)/r [409, 417]. The range between zero
and p/2 for parameter w allows seven unit steps for the rescaled potential F.
Starting from zero, the unit steps separate equipotential surfaces drawn by field
parameters (r, s) on a square domain given in (6.100a). They dress the spark gap in
seven equipotential surfaces, shown in Fig. 6.27a, as confocal hyperboloids of
revolution.
Obvious rotation and reflection symmetries map these potential surfaces from
the first to any other octant.

Potential: f ¼ f0; 1; :::; 8g; 0 b r b 6; 0 b s b 6; t ¼ 0; (6.100a)

Field: 0 b f b 8; r ¼ f0; 1; 2; 3g; 0 b s b 6; t ¼ 0; (6.100b)

Surfaces: 0 b f b 8; 0 b r b 6; s ¼ f0; 1; 2; 3g; t ¼ 0: (6.100c)

Equations (6.100b, c) define the parameter ranges for complementary sets of

field surfaces on square domains shown in Fig. 6.27b and c. One side of their open
octant shows streamlines along confocal ellipses. The same holds for interior field
lines traced by field surfaces that are not orthogonal. A common branch line above
the electrode gap unites all field surfaces, securing their uniqueness, which is also
needed for inverse sine potentials in the complex plane.
Figure 6.27d combines the potential and field surfaces of the quaternion sine
function in a 3D mesh for spark gap configurations. The conjugate solution will
dress a charged bar in confocal ellipsoids of revolution as equipotential surfaces. In
[413], our matrix format derives a spherical potential distribution from exponential
quaternion functions.
6 Ball Lightning Investigations 355

6.4.15 Conclusions

We have connected plasma conditions in lightning channels and BL using the

following scenario:
– Electric field threshold initiates cavitation by exchange forces in cool, dense
discharge plasma.
– Quantum limit for plasma circulation sets length and time-scales for its photonic
response.
– Condensation of photonic plasmon waves suddenly accumulates free charge in
central regions.
– A charged eigenstate ensures compactness and coherent circulation by octahe-
dral symmetry.
– Steady radiative loss at fixed radius slowly flattens the radial electric profiles
until extinction.
For wire grids or charged vortex lattices interacting by logarithmic potentials,
we derived the following:
– A gauge condition securing finite lattice sums by the Leibniz condition for
convergence
– A smooth Madelung curve with a finite minimum expressed by the Wallis
product
– Extension of complex functions to 4D quaternion space for 3D lattices sums
– Smoothness conditions on quaternion functions enabling 3D potential field
calculations
– Quaternion inversion maps on tapered tori with nuclear potential in the thermo-
dynamic limit
Our matrix format for quaternion functions and their Jacobians:
– Ensures nonzero determinants, real eigenvalues, and orthogonal eigenvectors.
– Maps planar flow patterns on the unit sphere by stereographic projection.
– Maps lightning potential distributions near ground level in elliptic coordinates.
– Extends corresponding smoothness conditions to 8D octavian space.
As touchstones for the validation of our BL model by laboratory reproduction,
we put forward:
– Preservation of net charge content, radius, and luminosity throughout its life span
– Increased life span resulting from chemical reactions and atomic processes in
outer layers
– A steady state resulting from ignition of fusion reactions between light nuclei in
the core region

Acknowledgement The author wishes to thank J. de Graaf for the helpful and stimulating
discussions on field theory.
356 V.L. Bychkov et al.

References

1. Ohtsuki, Y.-H. (ed.): Science of Ball Lightning (Fire Ball). World Science, Singapore (1989)
2. Singer, S.: The first decade of International Symposia on Ball Lightning. In: Proceedings of
the 5th International Symposium on Ball Lightning (ISBL97), Tsugawa-Town, Niigata,
Japan, pp. 1–5 (1997)
3. Singer, S.: Direction in recent research and miscellaneous activities in ball lightning. In:
Proceedings of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp,
Belgium, pp. 1–2 (1999)
4. Singer, S.: Consideration of selected ball lightning models. In: Proceedings of the 8th
International Symposium on Ball Lightning (ISBL04), Chung-li, Taiwan, pp. 1–5 (2004)
5. Singer, S.: Progress toward a solution of ball lightning. In: Proceedings of the 9th Interna-
tional Symposium on Ball Lightning (ISBL06), Eindhoven, The Netherlands, pp. 1–2 (2006)
6. Bychkov, V.L., Golubkov, G.V., Nikitin, A.I.: Contemporary state of investigations of ball
lightning. Khim. Fiz. 25(3), 3–6 (2006)
7. Nikitin, A.I., Bychkov, V.L.: 2nd International Symposium on Unconventional Plasmas
(ISUP06) and 9th International Symposium on Ball Lightning (ISBL06). Khim. Fiz. 27(2),
87–96 (2008)
8. Leonov, R.A.: Ball Lightning Enigma. Nauka, Moscow (1965)
9. Lomonosov, M.V.: Selected Works on Chemistry and Physics, pp. 220–256. AS USSR
Publishers, Moscow (1961)
10. Arago, F.: Sur le Tonnerre. Annuare au Roi par le Bureau des Longitudes, Notices Scient.
p. 221 (1838)
11. Brand, W.: Der Kugelblitz. Verlag von H. Grand, Hamburg (1923)
12. Baturin, V.V., Khanzhonkov, V.I.: Air circulation in premises with respect to location of
intake and air vent holes. Heating and Ventilation (4–5), 29–33 (1939)
13. Morris, W.: A thunderstorm mystery. Daily Mail (London). November 5 (1936)
14. Goodlet, B.L.: Lightning. J. IEE, London 81, 1 (1937)
15. Chirvinsky, P.N.: Ball lightning. Priroda (6), 14–20 (1949)
16. McNally, J.R., Jr.: Preliminary Report on the Ball Lightning. Oak-Ridge National Labora-
tory, Report N. ORNL3938 (1966)
17. Rayle, W.D.: Ball Lightning Characteristics. NASA Technical NOTE-D- 3188 (1966)
18. Singer, S.: The Nature of Ball Lightning. Plenum, New York (1971)
19. Dmitriev, M.T.: Ball lightning nature. Priroda (6), 98 (1967)
20. Balyberdin, V.V.: Determination of ball lightning energy. Samoletostroenie i Tehnika
Vozdushnogo Flota. KhGU Publishers, Kharkov N.3, 102–104 (1965)
21. Charman, W.N.: Ball lightning. Phys. Rep. 54(4), 261–306 (1979)
22. Altschuler, M.D., Houste, L.L., Hildner, E.: Is ball Lightning a nuclear phenomenon? Nature
228, 545–547 (1970)
23. Covington, A.E.: Ball lightning. Nature 226, 252 (1970)
24. Zimmerman, P.D.: Energy content of Covington’s lightning ball. Nature 228, 853 (1970)
25. Tompkins, D.R., Rodney, P.F., Gooding, R.: Photographic observations of ball lightning.
Bull. Am. Phys. Soc. 20, 659 (1975)
26. Eriksson, A.J.: Videotape recording of a possible ball lightning event. Nature 268, 35 (1977)
27. Mills, A.A.: Ball lightning and thermoluminescence. Nature 233, 131 (1971)
28. Fleming, S.J., Aitken, M.J.: Radiation dosage associated with ball lightning. Nature 252, 220
(1975)
29. Stakhanov, I.P.: Physical Nature of Ball Lightning. Atomizdat, Moscow (1979)
30. Stakhanov, I.P.: On Physical Nature of Ball Lightning. Energoatomizdat, Moscow (1985)
31. Stakhanov, I.P.: On Physical Nature of Ball Lightning. Nauchny Mir, Moscow (1996)
32. Imyanitov, I., Tikhiy, D.: Beyond Boundary of Science Laws. Atomizdat, Moscow (1980)
33. Bychkov, V.L.: On observation features of ball lightning. Khim. Fiz. 25(3), 7–17 (2006)
6 Ball Lightning Investigations 357

34. Dmitriev, M.T., Bakhtin, B.I., Martynov, B.I.: Thermal factor of ball lightning. Zhur. Tech.
Fiz. 51(12), 2567–2572 (1981)
35. Barry, J.D.: Ball Lightning and Bead Lightning. Plenum, New York (1980)
36. Brovetto, P., Maxia, V., Bussetti, G.: On the nature of ball lightning. J. Atmos. Terr. Phys.
38, 921–934 (1976)
37. Moigno, F.: Le Cosmos. 14, 672 (1859)
38. Endean, V.G.: Ball lightning as electromagnetic radiation. Nature 263, 753 (1976)
39. Menzel, D.H.: Flying Saucers. Harvard University Press, Harvard (1965)
40. Cowgill, W.: Curious phenomenon in Venezuela. Sci. Am. 55, 389 (1886)
41. Grigorjev, A.I.: Fiery Killers. Mysteries and Secret of Ball Lightning. Debiut, Yaroslavl
(1990)
42. Grigorjev, A.I.: Ball Lightning. YarGU Publishers, Yaroslavl (2006)
43. Batygin, A., Mosin, I.: Visit of “fiery lady”. Pravda (Moscow). 220 (25938), August,
8 (1989)
44. Egely, G.: Analysis of Hungarian ball lightning observations. In: Keul, A.G. (ed.) Progress in
Ball Lightning Research. Proceedings of the VIZOTUM, Salzburg, Austria (1993)
45. Avramenko, R.F., Bychkov, V.L., Klimov, A.I., Sinkevich, O.A. (eds.): Ball Lightning in a
Laboratory. Khimiya, Moscow (1994)
46. Novosti UFOLATS http://www.ufo.lv/index. php? id= 13323112072006125332
47. Grigorjev, I.S., Meilikhov, E.Z. (eds.): Physical Values. Reference Book. Enrgoatomizdat,
Moscow (1991)
48. Stenhoff, M.: Ball Lightning. An Unsolved Problem in Atmospheric Physics. Kluwer, New
York (1999)
49. Chernobrov, V.A.: Chronicle of UFO Visits. Drofa, Moscow (2003)
50. Vostrukhin, D.: At 4.10 sharp. Trud (Moscow), January, 30 (1985)
51. Ofuruton, H., Ohtsuki, Y.-H., Kondo, N., Kamogawa, M., Kato, M., Takahashi, T.: Nature of
ball lightning in Japan. In: Proceedings of the 5th International Symposium on Ball Light-
ning (ISBL97), Tsugawa-Town, Niigata, Japan, pp. 17–19 (1997)
52. Keul, A.G.: The Southern Bavaria ball lightning car collision. In: Proceedings of the 5th
International Symposium on Ball Lightning (ISBL97), Tsugawa-Town, Niigata, Japan,
pp. 27–34 (1997)
53. Stakhanova, I.G.: Ball lightning pictures from the I.P. Stakhanov’s archive and their inter-
pretation. In: Proceedings of the 5th International Symposium on Ball Lightning (ISBL97),
Tsugawa-Town, Niigata, Japan, pp. 35–41 (1997)
54. Amirov, A.Kh., Bychkov, V.L.: Observational and computer assisted analysis of data from
SKB data base. In: Proceedings of the 5th International Symposium on Ball Lightning
(ISBL97), Tsugawa-Town, Niigata, Japan, pp. 42–46 (1997)
55. Stakhanova, I.G.: Observed properties of the ball lightning depending on weather conditions.
In: Proceedings of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp,
Belgium, pp. 3–9 (1999)
56. Ofuruton, H., Ohtsuki, Y.-H., Kondo, N., Kamogawa, M., Kato, M., Takahashi, T.,
Sato, S., Shintani, K.: Nature of ball lightning in Japan (Part 2). In: Proceedings of the
6th International Symposium on Ball Lightning (ISBL99), Antwerp, Belgium, pp. 10–12
(1999)
57. Keul, A.G.: Attempted ball lightning tracking with a Lightning Detection System. In:
Proceedings of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp,
Belgium, pp. 13–18 (1999)
58. Amirov, A.Kh., Bychkov, A.V., Bychkov, V.L.: Ball lightning nature in respect to depen-
dence lifetime diameter. In: Proceedings of the 6th International Symposium on Ball
Lightning (ISBL99), Antwerp, Belgium, pp. 19–26 (1999)
59. Amirov, A.Kh., Bychkov, V.L.: On the ball lightning interaction with airplanes. In: Proceed-
ings of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp, Belgium,
pp. 34–37 (1999)
358 V.L. Bychkov et al.

60. Toselli, P., Fedele, R.: Project for a database of ball lightning observations in Italy. In:
Proceedings of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp,
Belgium, pp. 27–33 (1999)
61. Timoshuk, A.S.: Observations of two consequently formed and decayed ball lightning. In:
Smirnov, B.M. (ed.) Sharovaya Molniya, vol. 1, p. 9. Institute for High Temperature AS
USSR, Moscow (1990)
62. Bychkov, V.L.: Observations of ball lightning. In: Materials of 13-th Russian Conference on
Cold Nuclei Transmutation of Chemical Elements and Ball lightning, 2005. Dagomys, city
of Sochi, pp. 237–245. NITs FTP “Ersion”, Moscow (2006)
63. Bychkov, V.L., Bychkov, D.V., Sedov, Y.B.: Some new ball lightning observation data. In:
Proceedings of the 8th International Symposium on Ball Lightning (ISBL04), Chung-li,
Taiwan, pp. 6–10 (2004)
64. Arora, R.: Ball lightning without lightning strike. In: Proceedings of the 8th International
Symposium on Ball Lightning (ISBL04), Chung-li, Taiwan, pp. 11–15 (2004)
65. Keul, A.G., Schrattenecker, R., Baumegger, S.: The Zwoenitz, Germany, ball lightning
webcam record. In: Proceedings of the 8th International Symposium on Ball Lightning
(ISBL04), Chung-li, Taiwan, pp. 44–50 (2004)
66. Kawano, S.: Is a spatial stem a ball lightning? In: Proceedings of the 8th International
Symposium on Ball Lightning (ISBL04), Chung-li, Taiwan, pp. 79–82 (2004)
67. Bychkov, V.L.: On observation properties of ball lightning. In: Proceedings of the 9th
International Symposium on Ball Lightning (ISBL06), Eindhoven, The Netherlands, pp.
18–25 (2006)
68. Keul, A.G., Hentschel, K.H., Stummer, O.: German ball lightning data banks results. In:
Proceedings of the 9th International Symposium on Ball Lightning (ISBL06), Eindhoven, the
Netherlands, pp. 96–105 (2006)
69. Nikitin, A.I., Velichko, A.M., Vnukov, A.V., Nikitina, T.F.: Estimation of ball lightning
characteristics based on the analysis of its photo. In: Proceedings of the 9th International
Symposium on Ball Lightning (ISBL06), Eindhoven, The Netherlands, pp. 148–156 (2006)
70. Papaelias, Ph.M.: Photographs of ball lightning support evidence of antimatter. In: Proceed-
ings of the 9th International Symposium on Ball Lightning (ISBL06), Eindhoven, the
Netherlands, pp. 167–171 (2006)
71. Tar, D.: Observation of lightning ball – a new phenomenological description of the phenom-
enon. In: Proceedings of the 9th International Symposium on Ball Lightning (ISBL06),
Eindhoven, The Netherlands, pp. 222–225 (2006)
72. Bychkov, V.L., Nikitin, A.I.: Recent ball lightning investigations: short review. In:
Proceedings of the 10th International Symposium on Ball Lightning (ISBL08) and 3rd
International Symposium on Unconventional Plasmas (ISUP08), Kaliningrad, Russia, pp.
36–38 (2008)
73. Keul, A.G.: European ball lightning statistics. In: Proceedings of the 10th International
Symposium on Ball Lightning (ISBL08) and 3rd International Symposium on Unconven-
tional Plasmas (ISUP08), Kaliningrad, Russia, pp. 66–71 (2008)
74. VanDevender, J.P., VanDevender, A.P., Wilson, P., VanDoorn, P., McGinley, N.: Extreme
ball lightning event of August 6, 1868 in County Donegal, Ireland. In: Proceedings of the
10th International Symposium on Ball Lightning (ISBL08) and 3rd International Symposium
on Unconventional Plasmas (ISUP08), Kaliningrad, Russia, pp. 142–148 (2008)
75. Fitzgerald, M. Notes on the occurrence of globular lightning and of waterspouts in County
Donegal, Ireland. Quarterly Journal of the Meteorological Society. First Quarter of 1878.
Quarterly Proceeding at the March 20, 1878, Proceedings at the Meeting of the Society,
pp. 160–161 (1878)
76. Shelkunov, G.P., Nikitin, A.I., Bychkov, V.L., Nikitina, T.F., Velichko, A.M., Vasiliev, A.
L.: Examination of a window pane exposed to ball lightning. In: Proceedings of the 10th
International Symposium on Ball Lightning (ISBL08) and 3rd International Symposium on
Unconventional Plasmas (ISUP08), Kaliningrad, Russia, pp. 127–134 (2008)
6 Ball Lightning Investigations 359

77. Mashkovich, M.D.: Electric Properties of Inorganic Dielectrics in MW Range. Soviet Radio,
Moscow (1969)
78. Nikolayev, G.V.: Mysteries of Electromagnetism. “Znamya mira”, Tomsk (1999)
79. Ford, R.A.: Homemade Lightning. McGraw-Hill, New York (2002)
80. Planté, G.: Electrical Phenomena in Atmosphere. Paris (1891)
81. Silberg, P.A.: A review of ball lightning. In: Coroniti S.C. (ed.) Problems of Atmospheric
and Space Electricity, pp. 436–466. Proceedings of the 3rd International Conference on
Atmosphere and Space Electricity, 1963, Montreux, Switzerland. Elsevier, Amsterdam
(1965)
82. Dijkhuis, G.C.: Threshold current for fireball generation. J. Appl. Phys. 53(5), 3516–3519
(1982)
83. Kapitsa, P.L.: On nature of ball lightning. Doklady AN SSSR 101, 245–248 (1955)
84. Kapitsa, P.L.: Free plasma filament in high frequency field at high pressure. Zhur. Experim. i
Teoretich. Fiziki. 57, N. 6(12), 1801–1866 (1969)
85. Ohtsuki, Y.-H., Ofuruton, H.: Plasma fireballs formed by microwave interference in air.
Nature 350, 139–141 (1991)
86. Zhiltsov, V.A., Leitner, J.F., Manykin, E.A., Petrenko, E.A., Skovoroda, A.A., Handel, P.:
Spatially-localized discharge in the atmosphere. Zhur. Experim. i Teoretich. Fiziki. 108. N. 6
(12), 1966–1985 (1995)
87. Ofuruton, H., Kondo, N., Kamogawa, M., Aoki, M., Ohtsuki, Y.-H.: Experimental condition
of artificial ball lightning by using microwave and discharge. In: Proceedings of the 5th
International Symposium on Ball Lightning (ISBL97), Tsugawa-Town, Japan, pp. 215–217
(1997)
88. Ofuruton, H., Kondo, N. Kamogawa, M., Aoki, M., Ohtsuki, Y.-H.: Experimental condition
for ball lightning creation by using air gap discharge embedded in microwave field. J. Geoph.
Res. 106. N. D12. 12367–12369 (2001)
89. Anderson, R.W.: Absence of diffusion in creation random lattices. Phys. Rev. 109(5).
1492–1505 (1958)
90. Tanaka, K., Tanaka, M.: Is ball lightning “Anderson localization”? In: Proceedings of the 5th
International Symposium on Ball Lightning (ISBL97), Tsugawa-Town, Japan, pp. 137–141
(1997)
91. Kamogawa, M., Tanaka, H., Ofuruton, H., Ohtsuki, Y.-H.: Possibility of microwave locali-
zation to produce an experimental plasma fireball. Proc. Jap. Acad. 75(10) Ser. B, pp.
275–280 (1999)
92. Ofuruton, H., Kamogawa, M., Tanaka, H., Ohtsuki, Y.-H.: Experiment of the localization of
the electromagnetic wave in the waveguide. In: Proceedings of the 6th International Sympo-
sium on Ball Lightning (ISBL99), Antwerp, Belgium, pp. 207–211 (1999)
93. Kamogawa, M., Ofuruton, H., Liu, J.Y., Tanaka, H., Ohtsuki, Y.-H.: Study of microwave
localization mode for appearance of plasma fireballs. In: Proceedings of the 8th International
Symposium on Ball Lightning (ISBL04), Chung-li, Taiwan, pp. 118–119 (2004)
94. Ofuruton, H., Kamogawa, M., Tanaka, H., Ohtsuki, Y.-H.: Experiments on microwave
localization for ball lightning research. In: Proceedings of the 8th International Symposium
on Ball Lightning (ISBL04), Chung-li, Taiwan, pp. 120–122 (2004)
95. Ofuruton, H., Kamogawa, M., Ohtsuki, Y.-H.: Experiments on ball lightning by using
microwave localization. In: Proceedings of the 9th International Symposium on Ball Light-
ning (ISBL06), Eindhoven, The Netherlands, pp. 157–158 (2006)
96. Ofuruton, H., Kamogawa, M., Ohtsuki, Y.-H.: Experiments for artificial ball lightning by
using microwave localization. In: Proceedings of the 10th International Symposium on Ball
Lightning (ISBL08) and 3rd International Symposium on Unconventional Plasmas
(ISUP08), Kaliningrad, Russia, pp. 99–100 (2008)
97. Chukanov, K.B.: Ball Lightning – source of free energy. In: Proceedings of the 10th
International Symposium on Ball Lightning (ISBL08) and 3rd International Symposium on
Unconventional Plasmas (ISUP08), Kaliningrad, Russia, pp. 22–26 (2008)
360 V.L. Bychkov et al.

98. Golka, R.K., Jr.: Laboratory-produced ball lightning. J. Geophys. Res. 99, N. D 5,
10679–10681 (1994)
99. Dikhtyar, V., Jerby, E.: Fireball ejection from a molten hot spot to air by localized micro-
waves. Phys. Rev. Lett. 96, 045002-1–045002-4 (2006)
100. Corum, K.L., Corum, J.F.: Tesla’s production of electric fireballs. Tesla Coil Builder’s
Association News 8(3), 13–18 (1989)
101. Koloc, P.M.: Formed PLASMAK artificial ball lightning results. In: Proceedings of the 6th
International Symposium on Ball Lightning (ISBL99), Antwerp, Belgium, pp. 219–226
(1999)
102. Koloc, P.M.: Comparison between PLASMAK BL model and formed PMKs. In: Proceed-
ings of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp, Belgium, pp.
227–232 (1999)
103. Kunin, V.N.: Research experiments on method of ball lightning laboratory model obtaining.
Khim. Fiz. 25(3), 94–95 (2006)
104. Kunin, V.N., Pleshivtsev, V.S., Furov, L.V.: A method of obtaining of independent long-
lived plasma formations in free atmosphere. Khim. Fiz. 25(3), 96–100 (2006)
105. Kunin, V.N., Furov, L.V.: Investigations of plasma toroidal vortices appearing at electric
explosion of diaphragms in air. Izvestiya VUZov. Fizika 33(6), 119 (1990)
106. Kunin, V.N., Pleshivtsev, V.S., Furov, L.V.: Experimental laboratory research of the ball
lightning nature. In: Proceedings of the 5th International Symposium on Ball Lightning
(ISBL97), Tsugawa-Town, Japan, pp. 225–229 (1997)
107. Kunin, V.N., Pleshivtsev, V.S.: On the possible mechanism of transformation of toroidal
plasma vortex into the laboratory analog of ball lightning. In: Proceedings of the 6th Interna-
tional Symposium on Ball Lightning (ISBL99), Antwerp, Belgium, pp. 233–235 (1999)
108. Kunin, V.N., Kondakov, V.P., Pleshivtsev, V.S., Furov, L.V.: Experimental results of
anomalous decay of independent long-lived plasma formation in free atmosphere. Khim.
Fiz. 25(3), 101–103 (2006)
109. Kunin, V.N., Furov, L.V.: On refractional mechanism of energy accumulation in a volume of
long-lived plasma formation. In: Materials of 13-th Russian Conference on Cold Nuclei
Transmutation of Chemical Elements and Ball Lightning, 2005. Dagomys, city of Sochi, pp.
284–289. NITs FTP “Ersion”, Moscow (2006)
110. Alexeff, I., Parameswaran, S.M., Thiayagarajan, M., Grace, M.: An experimental study of
ball lightning. IEEE Trans. Plasma Sci. 32(3), 1378–1382 (2004)
111. Alexeff, I., Parameswaran, S.M., Thiayagarajan, M., Grace, M.: An observation of synthetic
ball lightning. IEEE Trans. Plasma Sci. 33(2), 498–499 (2005)
112. Avramenko, R.F., Bakhtin, B.I., Nikolayeva, V.I., Poskachheva, L.P., Shirokov, N.N.:
Investigation of plasma formations initiated by erosive discharge. Zhur. Tekhnich. Fiziki.
60(12), 57–64 (1990)
113. Ershov, A.P., Rozanov, V.A., Sysoev, N.N., Timofeev, I.B., Chuvashev, S.N., Shibkov, V.
M.: Subsonic plasma jets flowing into atmosphere created by capillary. Physical faculty of
M.V. Lomonosov MSU. Physical Hydrodynamics. N. 4. Preprint N. 8 (1994)
114. Bychkov, V.L., Gridin, A.Y., Klimov, A.I.: On nature of artificial ball lightning. Teplofiz.
Vysok. Temper. 32(2), 190–194 (1994)
115. Emelin, S.E., Semenov, V.S., Bychkov, V.L., Belisheva, N.K., Kovshik, A.P.: Some objects
formed in the interaction of electrical discharges with metals and polymers. Tech. Phys. 42
(3), 269–277 (1997)
116. Bychkov, A.V., Bychkov, V.L., Timofeev, I.B.: Experimental modeling of long-lived
luminescent formations in air on a basis of polymeric organic materials. In: Materials of
10-th Russian Conference on Cold Nuclei Transmutation of Chemical Elements and Ball
Lightning, 2002. Dagomys, city of Sochi, pp. 125–148. NITs FTP “Ersion”, Moscow (2003)
117. Bychkov, A.V., Bychkov, V.L., Timofeev, I.B.: Experimental modeling of long-lived
shining formations in air on a basis of polymeric organic materials. Zhur. Tekhnich. Fiziki.
74(1), 128–133 (2004)
6 Ball Lightning Investigations 361

118. Klimov, A.I.: Plasmoid generation in airflow and its physical properties. In: Materials of 10-
th Russian Conference on Cold Nuclei Transmutation of Chemical Elements and Ball
Lightning, 2002. Dagomys, City of Sochi, pp. 229–235. NITs FTP “Ersion”, Moscow (2003)
119. Klimov, A., Moralev, I., Zhirnov, K., Kutalaliev, V.: Investigation of physical properties of
plasmoids in air and airflow. In: Materials of 13-th Russian Conference on Cold Nuclei
Transmutation of Chemical Elements and Ball Lightning, 2005. Dagomys, City of Sochi, pp.
298–306. NITs FTP “Ersion”, Moscow (2006)
120. Klimov, A.I.: Investigation of physical properties of plasmoids in air and air flow. Khim. Fiz.
25(3), 104–109 (2006)
121. Klimov, A., Moralev, I., Zhirnov, K., Kutalaliev, V.: Investigation of physical properties of
plasmoids in air and airflow. In: Proceedings of the 9th International Symposium on Ball
Lightning (ISBL06), Eindhoven, The Netherlands, pp. 112–121 (2006)
122. Abrahamson, J., Dinnis, J.: Ball lightning caused by oxidation of nanoparticle networks from
normal lightning strikes on soil. Nature 403, 519–521 (2000)
123. Abrahamson, J.: Ball lightning from atmospheric discharges via metal nanosphere oxidation:
from soils, wood or metals. Phil. Trans. Roy. Soc. Lond. 360(1790), 61–88 (2002)
124. Piva, G.S., Pavão, A.C., de Vasconcelos, E.A., Mendes Jr., O., Da Silva Jr., E.F.: Production
of ball-lightning-like luminous balls by electrical discharge in silicon. Phys. Rev. Lett. 98,
048501-1–048501-4 (2007)
125. Shabanov, G.D.: Optical properties of long-lived luminescent formations. Pis’ma v Zhur.
Tekhnich. Fiziki. 28(4), 81–86 (2002)
126. Shabanov, G.D., Zherebtsov, O.M.: Experimental modeling of ball lightning analog. In:
Materials of 10-th Russian Conference on Cold Nuclei Transmutation of Chemical Elements
and Ball Lightning, 2002. Dagomys, City of Sochi. pp. 285–301. NITs FTP “Ersion”,
Moscow: (2003)
127. Emelin, S.E., Pirozerskyi, A.L., Egorov, A.I., Stepanov, S.I., Shabanov, G.D., Bychkov, V.
L.: Modeling of ball lightning with a help of electric discharge through a surface of weak
water solution. In: Materials of 9-th Russian Conference on Cold Nuclei Transmutation of
Chemical Elements and Ball Lightning, 2001. Dagomys, City of Sochi, pp. 240–248. NITs
FTP “Ersion”, Moscow (2002)
128. Shabanov, G.D., Zherebtsov, O.M.: Electric properties of independent luminous formations.
In: Materials of 11-th Russian Conference on Cold Nuclei Transmutation of Chemical
Elements and Ball Lightning, 2003. Dagomys, City of Sochi, pp. 279–290. NITs FTP
“Ersion”, Moscow (2004)
129. Shabanov, G.D., Zherebtsov, O.M.: Electric discharge to air half-space. Optich. Zhurl. 71(1),
6–8 (2004)
130. Shabanov, G.D., Sokolovskyi, B.Y.: Macroscopic separation of charges in the pulsed electric
field. Fizika Plasmy 31(6), 560–566 (2005)
131. Shabanov, G.D., Zherebtsov, O.M., Sokolovskyi, B.Yu.: Experimental modeling with a help
of pulsed electric discharge to air half-space of linear lightning leader and ball lightning. In:
Materials of 13-th Russian Conference on Cold Nuclei Transmutation of Chemical Elements
and Ball Lightning, 2005. Dagomys, City of Sochi. pp. 246–265. NITs FTP “Ersion”,
Moscow (2006)
132. Shabanov, G.D., Krivshich, A.G., Sokolovsky, B.Yu., Zherebtsov, O.M.: Ball lightning
nature. In: Materials of 14-th Russian Conference on Cold Nuclei Transmutation of Chemi-
cal Elements and Ball Lightning, 2006. Dagomys, City of Sochi, pp. 243–254. NITs FTP
“Ersion”, Moscow (2008)
133. Shabanov, G.D., Krivshich, A.G, Sokolovsky, B.Yu.: In: Materials of 14-th Russian Confer-
ence on Cold Nuclei Transmutation of Chemical Elements and Ball Lightning, 2006.
Dagomys, City of Sochi, pp. 255–266. NITs FTP “Ersion”, Moscow (2008)
134. Shabanov, G.D., Krivshich, A.G., Sokolovsky, B.Yu., Zherebtsov, O.M.: Performance of the
laboratory ball lightning. In: Proceedings of the 9th International Symposium on Ball
Lightning (ISBL06), Eindhoven, The Netherlands, pp. 202–209 (2006)
362 V.L. Bychkov et al.

135. Shabanov, G.D., Zherebtsov, O.M., Sokolovskyi, B.Y.: Independent long-lived luminescent
formations in open air. Experimental check of ball lightning formation by the linear lightning
leader hypothesis. Khim. Fiz. 25(4), 74–88 (2006)
136. Nikitin, A.I.: Electrodynamic model of ball lightning. Khim. Fiz. 25(3), 38–62 (2006)
137. Chalmers, J.A.: Atmospheric Electricity. Clarendon, Oxford (1949)
138. Emelin, S.E., Pirozersky, A.L., Vassiliev, N.N.: The dust-gas fireball as a special form of the
electric erosive discharge afterglow. In: Proceedings of the 9th International Symposium on
Ball Lightning (ISBL06), Eindhoven, The Netherlands, pp. 51–61 (2006)
139. Nikitin, A.I.: Electric capacitor as element of energy core of ball lightning. Elektrichestvo.
(11), 14–23 (1998)
140. Nikitin, A.I.: Electrical capacitor as the element of the power core of ball lightning. Electr.
Technol. Russ. (4), 70–85 (1998)
141. Emelin, S.E., Astafiev, A.M., Pirozerski, A.L.: Investigation of space-time structure of the
discharge with an electrolyte anode and face-type, air half-space directed cathode (Gat-
china’s discharge). In: Proceedings of the 10th International Symposium on Ball Lightning
(ISBL08) and 3rd International Symposium on Unconventional Plasmas (ISUP08), Kalinin-
grad, Russia, pp. 42–45 (2008)
142. Emelin, S.E., Pirozerski, A.L.: Some questions of power-consuming plasma-chemical ball
lightning. Khim. Fiz. 25(3), 83–89 (2006)
143. Pirozerski, A.L., Emelin, S.E.: Long-living plasmoids generation by high-voltage discharge
through thin conducting layers. In: Proceedings of the 9th International Symposium on Ball
Lightning (ISBL06). Eindhoven, The Netherlands, pp. 180–190 (2006)
144. Juettner, B., Noak, S., Versteegh, A., Fussmann, G.: Long-living plasmoids from a water
discharge at atmospheric pressure. In: Proceedings of the 28th International Conference on
Phenomena in Ionized Gases, Prague, Czech Republic, pp. 2229–2234 (2007)
145. Versteegh, A., Behringer, K., Fantz, U., Fussmann, G., J€ uttner, B., Noak, S.: Long-living
plasmoids from an atmospheric water discharge. Plasma Sour. Sci. Technol. 17(2),
02414–02421 (2008)
146. Smirnov, B.M.: Problem of Ball Lightning. Nauka, Moscow (1988)
147. Kuz’min, R.N., Miskinova, N.A., Shvilkin, B.N.: Laboratory model of ball lightning. In:
Materials of 10-th Russian Conference on Cold Nuclei Transmutation of Chemical Elements
and Ball Lightning, 2002. Dagomys, City of Sochi, pp. 236–242. NITs FTP “Ersion”,
Moscow (2003)
148. Kuz’min, R.N., Miskinova, N.A., Shvilkin, B.N.: Laboratory model of ball lightning. Khim.
Fiz. 25(3), 90–93 (2006)
149. Vlasov, A.N.: A ball lightning is a natural nuclear reactor? In: Proceedings of the 5th
International Symposium on Ball Lightning (ISBL97), Tsugawa-Town, Japan, pp. 75–79
(1997)
150. Vlasov, A.N.: Bose-Einsten condensed electron catalyzed fusion as a possible mechanism of
the ball lightning energy supply. In: Proceedings of the 6th International Symposium on Ball
Lightning (ISBL99), Antwerp, Belgium, pp. 133–138 (1999)
151. Vlasov, A.N.: Ball lightning as the current layer induced in a vortex. In: Proceedings of the
9th International Symposium on Ball Lightning (ISBL06), Eindhoven, The Netherlands, pp.
243–249 (2006)
152. Vlasov, A.N.: Magnetohydrodynamical model of plasma object capable to be generated at
impact of ordinary lightning. In: Proceedings of the 10th International Symposium on Ball
Lightning (ISBL08) and 3rd International Symposium on Unconventional Plasmas
(ISUP08), Kaliningrad, Russia, pp. 149–150 (2008)
153. Vlasov, A.N. Experimental modeling of ball lightning on a basis of electric explosion of a
wirespiral rolled in torus. In: Materials of 13-th Russian Conference on Cold Nuclei
Transmutation of Chemical Elements and Ball Lightning, 2005. Dagomys, City of Sochi,
pp. 290–297. NITs FTP “Ersion”, Moscow (2006)
154. Vlasov, A.N., Kolesnikov, S.A.: Calculations of an installation parameters for obtaining of
plasmoids at electric explosions of wire spirals rolled in torus. In: Materials of 14-th Russian
6 Ball Lightning Investigations 363

Conference on Cold Nuclei Transmutation of Chemical Elements and Ball Lightning, 2006.
Dagomys, City of Sochi, pp. 162–179. NITs FTP “Ersion”, Moscow (2008)
155. Vlassov, A., Vlassov, A.: Getting a fireball by means of electric explosion of wire spiral
twisted into a ring. In: Proceedings of the 9th International Symposium on Ball Lightning
(ISBL06). Eindhoven, The Netherlands, pp. 234–242 (2006)
156. Golubnichiy, P.I., Gromenko, V.M., Krutov, V.M.: Formation of long-lived luminescent
objects at decay of dense low temperature water plasma. Zhur. Tekh. Fiziki. 60(1), 283–186
(1990)
157. Golubnichiy, P.I., Gromenko, V.M., Krutov, V.M., Nikitin, E.V.: Plasma of electric dis-
charge in water as a raw material for laboratory analogue of a ball lightning. In: Proceedings
of the 9th International Symposium on Ball Lightning (ISBL06), Eindhoven, The Netherlands,
pp. 62–66 (2006)
158. Golubnichiy, P.I., Krutov, V.M., Nikitin, E.V.: Long-living luminous objects, formed
from the products of disintegration of water plasma. In: Proceedings of the 9th Inter-
national Symposium on Ball Lightning (ISBL06), Eindhoven, The Netherlands, pp.
67–73 (2006)
159. Matsumoto, T.: Ball lightning during underwater spark discharges and the Matsumae earth-
quakes. In: Proceedings of the 5th International Symposium on Ball Lightning (ISBL97),
Tsugawa-Town, Japan, pp. 193–201 (1997)
160. Matsumoto, T.: Micro ball lightning during underwater spark discharges. In: Proceedings of
the 6th International Symposium on Ball Lightning (ISBL99), Antwerp, Belgium, pp.
249–254 (1999)
161. Matsumoto, T.: Transport of micro ball lightning. In: Proceedings of the 6th International
Symposium on Ball Lightning (ISBL99), Antwerp, Belgium, pp. 255–262 (1999)
162. Urutskoev, L.I., Liksonov, V.I., Tsinoev, V.G.: Experimental discovery of “strange” radia-
tion and transformation of chemical elements. Prikl. Fiz. 4, 83–100 (2000)
163. Klimov, A., Baranov, D., Kutlaliev, V., Moralev, I., Zhirnov, K., Tolkunov, B.: Study of
anomalous “traces” of moving bright particles created by erosive pulsed plasma generator.
In: Materials of 14-th Russian Conference on Cold Nuclei Transmutation of Chemical
Elements and Ball Lightning, 2006. Dagomys, city of Sochi, pp. 199–208. NITs FTP
“Ersion”, Moscow (2008)
164. Turner, D.J.: The investigation of ball lightning with glass window panes. J. Meteorol. 22
(216), 52–64 (1997)
165. Shelkunov, G.: Ball lightning: observations and analysis of traces. Nauka i Zhizn. (10),
52–53 (2001)
166. Kolosovsky, O.A.: Investigations of ball lightning trace on a window pane. Zhur. Tekh. Fiz.
51(4), 856–858 (1981)
167. Nikitin, A.I., Bychkov, V.L., Nikitina, T.F., Velichko, A.M.: Modeling of ball lightning
interaction with window panes. In: Materials of 11-th Russian Conference on Cold Nuclei
Transmutation of Chemical Elements and Ball Lightning, 2003. Dagomys, City of Sochi, pp.
254–268. NITs FTP “Ersion”, Moscow (2004)
168. Nikitin, A.I., Bychkov, V.L., Nikitina, T.F., Velichko, A.M.: Modeling of ball lightning
interaction with window panes. In: Proceedings of the 8th International Symposium on Ball
Lightning (ISBL04), Chung-li,Taiwan, pp. 23–31 (2004)
169. Nikitin, A.I., Bychkov, V.L., Nikitina, T.F., Velichko, A.M.: Modeling of ball lightning
interaction with window panes. Khim. Fiz. 25(4), 98–105 (2006)
170. Nikitin, A.I., Leipunsky, I.O., Nikitina, T.F.: A search for the reasons explaining hovering of
ball lightning over conductor surface. In: Materials of 13-th Russian Conference on Cold
Nuclei Transmutation of Chemical Elements and Ball Lightning, 2005. Dagomys, City of
Sochi, pp. 307–319. NITs FTP “Ersion”, Moscow (2006)
171. Nikitin, A.I., Leipunsky, I.O., Nikitina, T.F.: A search for the reasons explaining hovering of
ball lightning over conductor surface. In: Proceedings of the 9th International Symposium on
Ball Lightning (ISBL06), Eindhoven, The Netherlands, pp. 139–146 (2006)
364 V.L. Bychkov et al.

172. Aleksandrov, V.Y., Podmoshenskyi, I.V., Sall, S.A.: Experiments on electric field impact on
gas discharge model of ball lightning. Zhur. Tekh. Fiz. 60(1), 73–76 (1990)
173. Nikitin, A.I., Nikitina, T.F., Velichko, A.M.: Corona discharge as a means to levitation of
ball lightning. In: Proceedings of the 10th International Symposium on Ball Lightning
(ISBL08) and 3rd International Symposium on Unconventional Plasmas (ISUP08), Kalinin-
grad, Russia, pp. 89–98 (2008)
174. Bychkov, V.L., Bychkov, A.V., Bychkov, D.V.: Some new observations of ball lightning. In:
Materials of 10-th Russian Conference on Cold Nuclei Transmutation of Chemical Elements
and Ball Lightning, 2002. Dagomys, City of Sochi, pp. 109–124. NITs FTP “Ersion”,
Moscow (2003)
175. Bychkov, V.L., Ershov, A.P., Chernikov, V.A.: Corona discharge modeling of some ball
lightning features. In: Proceedings of the 10th International Symposium on Ball Lightning
(ISBL08) and 3rd International Symposium on Unconventional Plasmas (ISUP08), Kalinin-
grad, Russia, pp. 7–11 (2008)
176. Kuz’min, R.N.: On physical and chemical nature of ball lightning. In: Materials of 9-th
Russian Conference on Cold Nuclei Transmutation of Chemical Elements and Ball Light-
ning, 2001. Dagomys, City of Sochi, pp. 235–239. NITs FTP “Ersion”, Moscow (2002)
177. Vvedenskyi, B.L. (ed.): The Big Soviet Encyclopedia, vol. 43, pp. 280–281. BSE Publishers,
Moscow (1956)
178. Bychkov, V.L., Nikitin, A.I.: Observations of ball lightning. In: Materials of 12-th Russian
Conference on Cold Nuclei Transmutation of Chemical Elements and Ball Lightning, 2004.
Dagomys, City of Sochi, pp. 165–169. NITs FTP “Ersion”, Moscow (2005)
179. Uman, M.A.: Decaying lightning channels, bead lightning and ball lightning. In: Coroniti, S.
C., Hughes, J. (eds.) Planetary Electrodynamics, vol. 2, p. 199. Gordon and Breach, New
York (1968)
180. Bostick, W.H.: Experimental study of ionized matter projected across magnetic field. Phys.
Rev. 104(2), 242–299 (1956)
181. Bostick, W.H.: Experimental study of plasmoids. Phys. Rev. 106(3), 404–412 (1957)
182. Shafranov, V.D.: On equilibrium magneto-hydrodynamic configurations. Zhur. Exper. i
Teor. Fiz. 33(3(9)), 710–722 (1957)
183. Shafranov, V.D.: On magnetohydrodynamical equilibrium configurations. Sov. Phys. JETP.
6, 545 (1957)
184. Finkelstein, D., Rubinstein, J.: Ball lightning. Phys. Rev. 135(2A), A390–A396 (1964)
185. Powell, J.R., Finkelstein, D.: Ball lightning. Am. Scientist. 58, 262 (1970)
186. Endean, V.G.: Electromagnetic field energy models – some recent developments. In: Stenh-
off, M. (ed.) Proceedings of the 4th TORRO Conference: Ball lightning, Tornado and Storm
Research Organization (TORRO), pp. 75. Oxford Brooks University, Oxford (1992)
187. Wooding, E.R.: Ball lightning. Nature 199, 272 (1963)
188. Bergström, A.: Electromagnetic theory of strong interaction. Phys. Rev. D3, 4394 (1973)
189. Neugebauer, Th.: Zu dem Problem des Kugelblitzes. Zeit. f€ ur Phys, 106(7), 8.474–8.484
(1937)
190. Neugebauer, Th: Zu der Quanten-mechanischen Theorie des Kugelblitz. Acta Physica 42(1),
29 (1977)
191. Dijkhuis, G.C.: A model for ball lightning. Nature 248, 150–151 (1980)
192. Dijkhuis, G.C.: Scaling law for fusion power from ball lightning. In: Janiszewski, J., Moron,
W., Sega, W. (eds.) Proceedings of the of International Wroclaw Symposium on Electro-
magnetic Compatibility, Wroclaw, Poland, pp. 21–25 (1988)
193. Hill, E.: Globular lightning. Nature 56, 293 (1897)
194. Coulvier-Gravier. Recherches sur les meteors et sur les lois qui les regissent, p. 185. Paris
(1859)
€
195. Meissner, A.: Uber Kugelblitze. Meteorol. Zeit. 47, 17–20 (1930)
196. Dawson, G.A., Jones, R.C.: Ball lightning as a radiation bubble. Pure Appl. Geophys. 75,
247–262 (1969)
6 Ball Lightning Investigations 365

197. Besnou, M.: De l’état de l’iode dans l’atmosphere, et de la possibilité de la formation de

l’iodide d’azote dans les orages. Mem. Soc. Sci. Nat. Cherbourg. 1, 103 (1852)
198. Thornton, W.M.: On thunderbolts. Fortsch. Phys. 67(3), 342 (1911)
199. De la Rive, A.: Traite de l’electricité theoretique et appliqué. Paris. 3, 197 (1858)
200. Hildebrandsson, H.H.: Kugelblitz. Fortsch. Phys. 39, H. 3518 (1883)
201. Schonland, B.F.J.: The Flight of Thunderbolts. Oxford University Press, Oxford (1950)
202. Benedicks, C.: Theory of lightning balls and its application to the atmospheric phenomenon
called ‘flying saucers’. Ark. Geof. 2(1), 1 (1951)
203. Chirvinsky, P.N.: Materials on ball lightning observations, collected by V.K. Cherkas.
Klimat i Pogoda. N. 5, 49 (1936)
204. Barry, J.D.: Ball lightning. J. Atmos. Terr. Phys. 29, 1095 (1967)
205. Barry, J.D.: Fireball, ball lightning and St. Elmo’s fire. Weather 23, 180 (1968)
206. De Tastes, M.: L’orage du 1-er fevrier 1884, a Torus. Météorologie 32, 105 (1884)
207. Frenkel, Y.I.: On nature of a ball lightning. Zhur. Exper. i Teor. Fiz. 10, 1424–1426 (1940)
208. Aleksandrov, V.Y., Golubev, E.M., Podmoshensky, I.V.: Aerosol nature of ball lightning.
Zhur. Exper. i Teor. Fiz. 52(10), 1987–1992 (1982)
209. Mukharev, L.A.: The nature of ball lightning. Sov. J. Commun. Technol. Electron. 30(9), 77
(1986)
210. Dauviller, A.: Foudre globulaire et reactions thernonucleaires. Comp. Rendu Hebd. Séances
Acad. Sci. 245, 2155 (1957)
211. De Tessan, M.: Sur la foudre en boule. Comp, Rendu Hebd. Séances Acad. Sci. 49, 189–191
(1859)
212. P€uringer, A.: Comments. In: Coroniti, S.C. (ed.) Problems of Atmospheric and Space
Electricity, pp. 460. Elsevier, Amsterdam (1965)
213. Hill, E.L.: Ball lightning. Am. Scientist. 58, 479 (1970)
214. Crew, E.W.: Ball lightning. New Scientist 56, 764 (1972)
215. Stakhanov, I.P.: On the nature of ball lightning. JETP Lett. 18, 114 (1974)
216. Watson, W.K.R.: A theory of ball lightning formation. Nature 185, 449–450 (1960)
217. Tonks, L.: Electromagnetic standing waves and ball lightning. Nature 187, 1013–1014
(1960)
218. Smirnov, B.M.: Physics of ball lightning. Uspekhi Fiz. Nauk. 160(4), 1–45 (1990)
219. Tamm, I.E.: Theory of magnetic thermonuclear reactor. In: Physics of plasma and problem
of controlled thermonuclear reactions, vol. 1, pp. 3–19; 31–41. Izdatelstvo AN SSSR,
Moscow (1958)
220. Sakharov, A.D.: Theory of magnetic thermonuclear reactor. In; Physics of plasma and
problem of controlled thermonuclear reactions, vol. 1, pp. 20–30. Izdatelstvo AN SSSR,
Moscow (1958)
221. Rosenbluth, M.N., Bussac, N.N.: MHD stability of Spheromak. Nucl. Fusion 19(4), 489
(1979)
222. Koloc, P.M.: “PLASMAK” star power for energy intensive space applications. Fusion
Technol. 15(3), 1136–1141 (1989)
223. Nikitin, A.I.: The requirements for elaboration of theory of ball lightning. In: Proceedings of
the 8th International Symposium on Ball Lightning (ISBL04), Chung-li, Taiwan, pp. 16–22
(2004)
224. Nikitin, A.I.: Analysis of the models of highly energetic ball lightning. In: Proceedings of the
8th International Symposium on Ball Lightning (ISBL04), Chung-li, Taiwan, pp. 57–63
(2004)
225. Nikitin, A.I.: The principles of developing the ball lightning theory. J. Rus. Laser Res. 25(2),
169–191 (2004)
226. Nikitin, A.I.: Will ball lightning problem be resolved in 21-st century? Khim. Fiz. 25(3),
18–37 (2006)
227. Lowke, J.J.: A theory of ball lightning as an electric discharge. J. Phys. D. (Appl. Phys.) 29
(5), 1237–1244 (1996)
366 V.L. Bychkov et al.

228. Handel, P.H.: Maser theory of ball lightning. Bull. Am. Phys. Soc. Ser. II. 20(1), 26BF8
(1975)
229. Handel, P.H., Leitner, J.F.: Theory of the stationary nonlinear ball lightning system of
fireball and atmospheric maser. In: Proceedings of the 5th International Symposium on
Ball Lightning (ISBL97), Tsugawa-Town, Japan, pp. 114–119 (1997)
230. Handel, P.H., Leitner, J.F.: Development of the maser-caviton ball lightning theory.
J. Geophys. Res. 99, 10689–10691 (1994)
231. Singer, J.R.: Masers. Wiley, New York (1959)
232. Vuylsteke, A.A.: Elements of maser theory. D. Van Nostrand, New York (1960)
233. Oraevsky, A.N.: Radiation echo. Uspekhi Fiz. Nauk. 91(2), 181–191 (1967)
234. Akulin, V.M., Karlov, N.V.: Intense Resonant Interactions in Quantum Electronics. Nauka,
Moscow (1987)
235. Gordiets, B.F., Osipov, A.I., Shelepin, L.A.: Kinetic processes in gases and molecular lasers.
Nauka, Moscow (1980)
236. Meek, J.M., Craggs, J.D.: Electrical Breakdown of Gases. Clarendon, Oxford (1953)
237. Wieder I.: On the absence of water maser at atmospheric pressure. In: Proceedings of the 9th
Intern. Symp. on Ball Lightning (ISBL06), Eindhoven, The Netherlands. pp. 250–255 (2006)
238. Cheung, A.C., Rank, D.M., Townes, C.H., Thornton, D.D., Welch, W.J.: Detection of water
in interstellar regions by its microwave radiation. Nature 221, 626–628 (1969)
239. Letokhov, V.S.: Astrophysical masers. Kvantovaya elektronika 32(12), 1065–1079 (2002)
240. Handel, P.H. Carlson, G.A., Grace, M., Leitner, J.F.: Electric ponderomotive forces cause
explosive ball lightning damages. In: Proceedings of the 8th International Symposium on
Ball Lightning (ISBL04), Chung-li, Taiwan, pp. 95–100 (2004)
241. Handel, P.H., Carlson, G.A., Grace, M., Leitner, J.F.: Motion of a BL discharge fed by an
atmospheric maser. In: Proceedings of the 8th International Symposium on Ball Lightning
(ISBL04), Chung-li, Taiwan, pp. 89–94 (2004)
242. Handel, P.H., Carlson, G.A.: Rise time of maser-caviton ball lightning energy spikes. In:
Proceedings of the 9th International Symposium on Ball Lightning (ISBL06), Eindhoven,
The Netherlands, pp. 74–80 (2006)
243. Smirnov, B.M.: On analysis of ball lightning nature. Uspekhi Fiz. Nauk. 116(4), 731–739
(1975)
244. Smirnov, B.M.: Origination of ball lightning. Doklady AN SSSR 226(4), 806–808 (1976)
245. Kadomtsev, B.B.: Ball lightning as a phenomenon of self-organization. J. Moscow Phys.
Soc. 1(4), 335–340 (1991)
246. Stepanov, S.I., Sall, S.A., Arutjunan, A.V.: Electric machine in ball lightning. In: Proceed-
ings of the 5th International Symposium on Ball Lightning (ISBL97), Tsugawa-Town,
Japan, pp. 183–187 (1997)
247. Stepanov, S.I.: Further development of electrochemical model of ball lightning. In: Proceed-
ings of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp, Belgium, pp.
96–101 (1999)
248. Stakhanov, I.P.: To a question of ball lightning nature. Zhur. Tekh. Fiz. 57(8), 1575–1582
(1987)
249. Turner, D.J.: Ball lightning and other meteorological phenomena. Phys. Rep. 293(1), 1–60
(1998)
250. Igolkin, S.I., Savelyev, S.K.: Condensation model of the ball lightning. In: Proceedings of the
5th International Symposium on Ball Lightning (ISBL97), Tsugawa-Town, Japan, pp. 80–86
(1997)
251. Bychkov, V.L.: Polymer ball lightning model. Physica Scripta. 50, 591–599 (1994)
252. Manykin, E.A., Ozhovan, M.I., Poluektov, P.P.: On collective electronic state in a system of
strongly-excited atoms. Doklady AN SSSR 260, 1096–1098 (1981)
253. Manykin, E.A., Ozhovan, M.I., Poluektov, P.P.: Theory of the condensed state in the system
of excited atoms. ZhETF 84(2), 442–453 (1983)
254. Manykin, E.A., Ozhovan, M.I., Poluektov, P.P.: To the question of ball lightning nature.
ZhTF 52(7), 1474–1476 (1982)
6 Ball Lightning Investigations 367

255. Manykin, E.A., Ojovan, M.I., Poluektov, P.P.: Rydberg matter and ball lightning. In:
Proceedings of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp,
Belgium, pp. 114–119 (1999)
256. Biberman, L.M., Norman, G.E.: On existence possibility of overcooled dense plasma.
Teplofiz. Vysok. Temper. 7(5), 822–831 (1969)
257. Norman, G.E.: Ball lightning as overcooled non-ideal plasma. Khim. Fiz. 18(7), 78–86
(1999)
258. Norman, G.E.: Ball lightning as a supercooled nonideal plasma. Chem. Phys. Rep. 18(7),
1335–1352 (2000)
259. Manykin, E.A., Norman, G.E.: Ball lightning as supercooled plasma condensed phase. In:
Proceedings of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp,
Belgium, pp. 120–125 (1999)
260. Mesenyashin, A.I.: Electrostatic and bubble nature of ball lightning. Appl. Phys. Lett. 58
(23), 2713–2715 (1991)
261. Mesenyashin, A.I.: Spherical formations in the atmosphere as a physical phenomenon.
J. Electrostatics 38, 139–150 (1995)
262. Zaitsev, I.V., Zaitsev, S.V.: Electrostatic model of ball lightning. Pisma v ZhTF. 17(7),
34–37 (1991)
263. Fedele, R.: A possible quantum-like approach to non-conventional plasmas. In: Proceedings
of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp, Belgium, pp.
126–132 (1999)
264. Pavlovsky, A.I., Dolotenko, M.I.: Pisma v ZhTF 38(9), 473 (1983)
265. Marsh, G.: Force free magnetic fields: solutions, topology and applications. World Scientific,
Singapore (1996)
266. Rañada, A.F., Trueba, J.L.: Ball lightning as an electromagnetic knot? Nature 383, 32 (1996)
267. Rañada, A.F., Soler, M., Trueba, J.L.: Ball lightnings and force-free magnetic knots. In:
Proceedings of the 6th International Symposium on Ball Lightning (ISBL99), Antwerp,
Belgium, pp. 102–107 (1999)
268. Jennison, R.C.: Ball lightning. A general critique. In: Proceedings of the 5th International
Symposium on Ball Lightning (ISBL97), Tsugawa-Town, Japan, pp. 6–11 (1997)
269. Callebaut, D.K.: Energy storage by force-free magnetic field in the initial phase of ball
lightning. In: Proceedings of the 8th International Symposium on Ball Lightning (ISBL04),
Chung-li, Taiwan, pp. 32–37 (2004)
270. Callebaut, D.K., Karugila, G.K., Khater, A.H.: Ball lightning with force-free magnetic fields
and runaway current. In: Proceedings of the 9th International Symposium on Ball Lightning
(ISBL06), Eindhoven, The Netherlands, pp. 33–38 (2006)
271. Shakirzjanov, F.N.: Electromagnetic model of ball lightning. Elektrichestvo. N. 10, 74–77
(1999)
272. Arnhoff, G.H.: On the spheric radiation. In: Proceedings of the 5th International Symposium
on Ball Lightning (ISBL97), Tsugawa-Town, Japan, pp. 101–107 (1997)
273. Arnhoff, G.H.: On the spheric radiation. In: Proceedings of the 6th International Symposium
on Ball Lightning (ISBL99), Antwerp, Belgium, pp. 160–165 (1999)
274. Dijkhuis, G.C.: State equation and phase diagram for fractal growth in ball lightning. In:
Kikuchi, H. (ed.) Environmental and Space Electromagnetics, pp. 535–546. Springer, Tokyo
(1991)
275. Feynman, R.F., Leighton, R.B., Sands, M.: The Feynmann Lectures on Physics. 9, p. 246.
Mir, Moscow (1967)
276. Dijkhuis, G.C.: Plasmoid confinement by the charged particles micro-fields. Nature 290, 166
(1981)
277. Dijkhuis, G.C.: Verhulst dynamics and fractal stretching of transition layer vorticity. In:
Kikuchi, H. (ed.) Dusty and Dirty Plasmas, Noise, and Chaos in Space and in the Laboratory,
pp. 163–176. Plenum, New York (1994)
368 V.L. Bychkov et al.

278. Dijkhuis, G.C.: Random walk and fractal deformation of transition layer vorticity. In:
Proceedings of the 1st International Symposium on Heat and Mass Transfer under Plasma
Conditions, Çesme, Turkey, pp. 611–621 (1994)
279. Dijkhuis, G.C.: Constructions for scale-invariant and kink-free vortex stretching. Physica B.
Condensed Matter 228, 144–152 (1996)
280. Dijkhuis, G.C.: Helix string field in ball lightning. In: Proceedings of the 5th International
Symposium on Ball Lightning (ISBL97), Tsugawa-Town, Japan, pp. 144–151 (1997)
281. Dijkhuis, G.C.: Exponential particle acceleration by recursive stretching of vortex filaments
in turbulent discharge plasma. In: Proceedings of the 8th International Symposium on Ball
Lightning (ISBL04), Chung-li, Taiwan, pp. 70–75 (2004)
282. Dijkhuis, G.C.: On Madelung sums for bound states in a plasma vortex system. In: Proceed-
ings of the 10th International Symposium on Ball Lightning (ISBL08) and 3rd International
Symposium on Unconventional Plasmas (ISUP08), Kaliningrad, Russia, pp. 27–33 (2008)
283. Artsimovich, L.A.: Controlled Thermonuclear Reactions. GIFML, Moscow (1961)
284. Bychkov, V.L., Bychkov, A.V., Stadnik, S.A.: Polymer fire balls in discharge plasma. Phys.
Scripta 53, 749–759 (1996)
285. Amirov, A.K., Bobkov, S.E., Bychkov, V.L., Emelin, S.E., Klimov, A.I., Semenov, V.S.:
Modern theoretical and experimental approaches to the problem of ball lightning. In:
Proceedings of the 5th International Symposium on Ball Lightning (ISBL97), Tsugawa-
Town, Japan, pp. 52–60 (1997)
286. Bychkov, V.L.: Unipolar ball lightning model. In: Proceedings of the 8th International
Symposium on Ball Lightning (ISBL04), Chung-li, Taiwan, pp. 64–69 (2004)
287. Bychkov, V.L., Bychkov, D.V.: Ball lightning as unipolarly charged object with hot surface.
In: Proceedings of the 9th International Symposium on Ball Lightning (ISBL06), Eindhoven,
The Netherlands, pp. 26–32 (2006)
288. Bychkov, V.L.: Unipolarly charged ball lightning. Khim. Fiz. 25(3), 63–71 (2006)
289. Nikitin, A.I.: The dynamic capacitor model of ball lightning. In: Proceedings of the 6th
International Symposium on Ball Lightning (ISBL99), Antwerp, Belgium, pp. 81–84 (1999)
290. Nikitin, A.I.: Stability and limiting energy content of autonomous ball lightning. Elektri-
chestvo. 3, 29–36 (2004)
291. Nikitin, A.I.: Substance of ball lightning as a certain form of unconventional plasma. Int. J.
Unconven. Electromag. Plasmas. 1(1–2), 101–108 (2008)
292. Nikitin, A.I.: How ball lightning can be created in nature. In: Proceedings of the 6th
International Symposium on Ball Lightning (ISBL99), Antwerp, Belgium, pp. 85–90 (1999)
293. Nikitin, A.I.: Origination of a ball lightning at development of a linear lightning. Elektri-
chestvo. (3), 16–23 (2000)
294. Aleksandrov, V.Y., Golubev, E.M., Podmoshensky, I.V.: Aerosol nature of ball lightning.
Zhur. Tech. Fizi. 52(10), 1987–1992 (1982)
295. Stakhanov, I.P.: On Physical Nature of Ball Lightning. Energoatomizdat, Moscow (1985)
296. Smirnov, B.M.: Problem of Ball Lightning. Nauka, Moscow (1988)
297. Bychkov, V.L.: Polymer ball lightning model. Physica Scripta. 50, 591–599 (1994)
298. Abrahamson, J., Dinniss, J.: Ball lightning caused by oxidation of nanoparticle networks
from normal lightning strikes on soil. Nature 403, 519–521 (2000)
299. Grigoriev, A.I., Shiryaeva, S.O., Grigorieva, I.D., et al.: On possibility of one ball lightning
division in two. Z. Tekh. Fiz. 61, 25–31 (1991)
300. Saranin, V.A.: Equilibrium of Liquids and its Stability. Institute of Computer Investigations,
Moscow (2002)
301. Bychkov, A.V., Bychkov, V.L., Abrahamson, J.: On the energy characteristics of ball
lightning. Phil. Trans. Roy. Soc. Lon. 360(1790), 97–106 (2002)
302. Emelin, S.E., Semenov, V.S., Bychkov, V.L., Belisheva, N.K., Kovshyk, A.P.: Some objects
formed in the interaction of electrical discharges with metals and polymers. Tech. Phys. 42
(3), 269–277 (1997)
6 Ball Lightning Investigations 369

303. Bychkov, A.V., Bychkov, V.L., Timofeev, I.B.: Experimental modeling of long-lived
shining formations in air on a basis of polymeric organic materials. Z. Tekh. Fiz. 74(1),
128–133 (2004)
304. Piva, G.S., Pavão, A.C., de Vasconcelos, E.A., Mendes Jr., O., Da Silva Jr., E.F.: Production
of ball-lightning-like luminous balls by electrical discharge in silicon. Phys. Rev. Letts. 98,
048501-1–048501-4 (2007)
305. Dikhtyar, V., Jerby, E.: Fireball ejection from a molten hot spot to air by localized micro-
waves. Phys. Rev. Lett. 96, 2–4 (2006)
306. Bychkov, V.L.: Unipolar ball lightning model. In: Proceedings of the 8th International
Symposium on Ball Lightning (ISBL04), Chung-li, Taiwan, pp. 64–69 (2004)
307. Bychkov,V.L., Bychkov, D. V.: Ball lightning as unipolarly charged object with hot surface.
In: Proceedings of the 9th International Symposium on Ball Lightning (ISBL06), Eindhoven,
the Netherlands, pp. 26–32 (2006)
308. Bychkov, V.L.: Unipolarly charged ball lightning. Khim. Fiz. 25(3), 63–71 (2006)
309. Drysdale, D.: An Introduction to Fire Dynamics. Wiley, Chichasler/New York/Brisbane/
Toronto/Singapore (1985)
310. Anderson, F.J., Freier, G.D.: Relation of electric fields to thunderstorm days. J. Geophys.
Res. 78(27), 6359–6363 (1973)
311. Sessler, G.M. (ed.): Electrets. Springer-Verlag, New York (1980)
312. Terletskyi, Y.P., Rybakov, Y.P.: Electrodynamics. Vysschaya Schkola, Moscow (1980)
313. Panofsky, W.K.H., Phillips, M.: Classical Electricity and Magnetism. Addison-Wesley,
Cambridge, MA (1962)
314. Merzbacher, C.I.: Materials that emit light by chemical reaction. Phil. Trans. Roy. Soc. Lon.
360(1790), 89–96 (2002)
315. Raizer, Y.P.: Physics of gaseous discharge. Nauka, Moscow (1992)
316. Aleksandrov, A.F., Bogdankevich, L.S., Rukhadze, A.A.: Basics of Plasma Electrodynam-
ics. Vyschaya Schkola, Moscow (1988)
317. Nikitin, A.I., Nikitina, T.F., Velichko, A.M.: Corona discharge as a means to levitation of
ball lightning. In: Proceedings of the 10th International Symposium on Ball Lightning
(ISBL08) and 3rd International Symposium on Unconventional Plasmas (ISUP08), Kalinin-
grad, Russia, pp. 89–98 (2008)
318. Blythe, T., Bloor, D.: Electrical Properties of Polymers. Cambridge University Press,
Cambridge (2005)
319. Abramovich, G.N.: Applied gas dynamics. Nauka, Moscow (1986)
320. Nikitin, A.I.: Electrodynamic model of ball lightning. Khim. Fiz. 25(3), 38–62 (2006)
321. Rawson, H.: Inorganic Glass-Forming Systems. Academic, London/New York (1967)
322. Bartenev, G.M.: Structure and Mechanical Properties of Inorganic Glasses. Publishers of
Building Literature, Moscow (1966)
323. Lykov, A.V.: Theory of Thermal Conductivity. Vysschaya Schkola, Moscow (1967)
324. Amirov, A.Kh., Bychkov, V.L., Bobkov, S.E.: On the dependence lifetime- diameter for ball
lightnings. Physica Scripta. 54, 13–14 (1998)
325. Amirov, A.K., Bychkov, A.V., Bychkov, V.L.: Ball lightning in respect to dependence
lifetime diameter. In: Proceedings of the 6th International Symposium on Ball Lightning,
23–25 August, Antwerp, Belgium, pp. 19–26 (1999)
326. Pudovkin, A.K.: Ball lightning in Novosibirsk Akademgorodok. Uspekhi Fiz. Nauk. 166
(11), 1253–1254 (1996)
327. Saranin, V.A.: The theory of electrothermal explosion produced by lightning. Teplofiz.
Vysok. Temper. 37(1), 31–36 (1999)
328. Mineev, A.: About high trees. In: Tikhomirova,V.A., Chernoutsan, A.I. (eds.) Physics and
Biology. Bureau Quantum, Moscow (2001)
329. Kalashnikov, S.G.: Electricity. Nauka, Moscow (1964)
330. Likhosherstnykh, G.U.: 138 approaches to a mystery of nature. Tekhnika Molodezhi. (3),
38–43 (1983)
370 V.L. Bychkov et al.

331. Goodlet, B.L.: Ball Lightning. J. Inst. Elect. Eng. 81, 1 (1937)
332. Barry, J.D.: Ball Lightning and Bead Lightning. Extreme Forms of Atmospheric Electricity.
Plenum, New York (1980)
333. Stenhoff, M.: Ball Lightning. An Unsolved Problem in Atmospheric Physics. Kluwer/
Plenum, New York (1999)
334. Imianitov, I., Tikhii, D.: Beyond the Laws of Science. Atomizdat, Moscow (1980)
335. Batygin, A., Mosin, I.: Visit of “Fairy Lady”. Pravda (Moscow), N. 220 (25938), pp. 6 (1989)
336. Nikitin, A.I.: The principles for developing of ball lightning theory. J. Russian Laser Res. 25
(2), 169–191 (2004)
337. Nikitin, A.I.: The requirements for elaboration of theory of ball lightning. In: Proceedings of
the 8th International Symposium on Ball Lightning, Chung-li, Taiwan, pp. 16–22 (2004)
338. Nikitin, A.I.: Analysis of the models of highly energetic ball lightning. In: Proceedings of the
8th International Symposium on Ball Lightning, Chung-li, Taiwan, pp. 57–63 (2004)
339. Dijkhuis, G.C.: A model for ball lightning. Nature 248, 150–151 (1980)
340. Dijkhuis, G.C.: Plasmoid confinement by the charged particles microfields. Nature 290, 166
(1981)
341. Dijkhuis, G.C.: Helix string model for turbulent vorticity and cavitation in shearing arc
plasma. Ann. NY Acad. Sci. 891, 259–272 (1999)
342. Bychkov, V.L.: Polymer ball lightning model. Physica Scripta 50, 591–599 (1994)
343. Bychkov, V.L.: Ball lightning as unipolarly charged object with hot surface. In: Proceedings
of the 9th International Symposium on Ball Lightning, Eindhoven, The Netherlands. pp.
26–38 (2006)
344. Nikitin, A.I.: An electrical capacitor as the element of the power core of ball lightning.
Electri. Technol. Russia (4), 70–85 (1998)
345. Nikitin, A.I.: The dynamic capacitor model of ball lightning. In: Proceedings of the 6th
International Symposium on Ball Lightning, Antwerp, Belgium, pp. 91–95 (1999)
346. Nikitin, A.I.: Substance of ball lightning as a certain form of unconventional plasma. Int. J.
Unconven. Electromag. Plasmas. 1(1–2), 101–108 (2008)
347. Nikitin, A.I.: Electrodynamical model of ball lightning. Khim. Fiz. 25(3), 38–62 (2006)
348. Artsimovich, L.A.: The Elementary Physics of Plasma. Atomizdat, Moscow (1969)
349. Frank-Kamenetskii, D.A.: Lectures on Plasma Physics. Atomizdat, Moscow (1968)
350. Landau, L.D., Lifshits, E.M.: Field Theory. GIFML, Moscow (1961)
351. Artsimovich, L.A.: Controlled Thermonuclear Reactions. GIFML, Moscow (1961)
352. Kalashnikov, S.G.: Electricity. Nauka, Moscow (1985)
353. Sarantsev, V.P., Perel’shtein, E.A.: Collective Acceleration of Ions by Electron Rings.
Atomizdat, Moscow (1979)
354. Nikitin, A.I.: Stability and limit energy content of an autonomous ball lightning. Elektri-
chestvo. (3), 29–36 (2004)
355. Dmitriev, M.T., Bakhtin, B.I., Martynov, V.I.: Investigation of a thermal factor of ball
lightning. J. Tech. Phys. 51(12), 2567–2572 (1981)
356. Balyberdin, V.V.: Estimate of the inner energy of ball lightning. In: Samolyotostroenie i
Tekhnika Vozdushnogo Flota, vol. 3, pp. 102–104. Publishing of Kharkov State University,
Kharkov (1965)
357. Ternov, I.M., Mikhailin, V.V., Khalilov, V.R.: Synchrotron Radiation and Its Applications.
Publishing of Moscow State University, Moscow (1980)
358. Artsimovich, L.A., Lukianov, S.Y.: Motion of Charged Particles in the Electric and Mag-
netic Fields. Nauka, Moscow (1978)
359. Frenkel’, Y.I.: Theory of the Phenomenon of Atmospheric Electricity. GITTL, Leningrad-
Moscow (1949)
360. Uman, M.A.: Lightning. McGraw-Hill, New York (1969)
361. Nikitin, A.I.,Velichko, A.M., Vnukov, A.V, Nikitina, T.F.: Estimation of ball lightning
characteristics based on the analysis of its photo. In: Proceedings of the 9th International
Symposium on Ball Lightning, Eindhoven, the Netherlands, pp. 148–156 (2006)
6 Ball Lightning Investigations 371

362. Nikitin, A.I., Velichko, A.M., Vnukov, A.V., Nikitina, T.F.: Estimate of the ball lightning
parameters on the base of analysis of its photo. Khim. Fiz. 26(8), 80–89 (2007)
363. Nikitin, A.I., Leipunsky, I.O., Nikitina, T.F.: A search for the reasons explaining hovering of
ball lightning over conductor surface. In: Proceedings of the 9th International Symposium on
Ball Lightning, Eindhoven, The Netherlands, pp. 139–147 (2006)
364. Nikitin, A.I., Nikitina, T.F., Velichko, A.M.: Corona discharge as a means to levitation of ball
lightning. In: Proceedings of the 10th International Symposium on Ball Lightning and 3rd
International Symposium on Unconventional Plasmas, Kaliningrad, Russia, pp. 89–98 (2008)
365. Schelkunov, G.P., Nikitin, A.I., Bychkov, V.L., Nikitina, T.F., Velichko, A.M., Vasiliev, A.
L.: Examination of a Window Pane Exposed to Ball Lightning. In: Proceedings of the 10th
International Symposium on Ball Lightning and 3rd International Symposium on Uncon-
ventional Plasmas, Kaliningrad, Russia, pp. 127–134 (2008)
366. Turner, D.J.: The interaction of ball lightning with glasses window panes. J. Meteorol. 22
(216), 52–64 (1997)
367. Nikitin, A.I., Bychkov, V.L., Nikitina, T.F., Velichko, A.M.: Modeling of ball lightning
interaction with window panes. In: Proceedings of the 8th International Symposium on Ball
Lightning. Chung-li, Taiwan, pp. 23–31 (2004)
368. Nikitin, A.I., Bychkov, V.L., Nikitina, T.F., Velichko, A.M.: Modeling of interaction of ball
lightning with window panes. Khim. Fiz. 25(4), 98–105 (2006)
369. Grigor’ev, A.I.: Ball Lightning. Publishing of Yaroslavl’ State University, Yaroslavl’ (2006)
370. Singer, S.: The Nature of Ball Lightning. Plenum, New York (1971)
371. Stakhanov, I.P.: On a Physical Nature of Ball Lightning. Nauchny Mir, Moscow (1996)
372. Nikitin, A.I., Velichko, A.M., Nikitina, T.F.: Principles for search of conditions of ordered
plasma structures creation. Izvestia Akademii Nauk, Energetics. N.2. pp. 115–132 (2008)
373. Meek, J.M., Craggs, J.D.: Electrical Breakdown in Gases. Clarendon, Oxford (1953)
374. Bazelian, E.M., Raizer, YuP: Physics of Lightning and of Protection Against Lightning.
Nauka, Moscow (2001)
375. Nikitin, A.I.: Creation of ball lightning at developing of linear lightning. Elektrichestvo. (3),
16–23 (2000)
376. Nikitin, A.I.: How ball lightning can be created in nature. In: Proceedings of the 6th
International Symposium on Ball Lightning, Antwerp, Belgium, pp. 85–90 (1999)
377. Raizer, YuP: Physics of Gas Discharge. Nauka, Moscow (1987)
378. Veselago, V.G.: The electrodynamics of substances with simultaneously negative values of e
and m. Sov. Phys. Uspekhi. 10, 509–514 (1968)
379. Yablonovitch, E., et al.: Photonic band structure: the face-centred cubic case employing non-
spherical atoms. Phys. Rev. Lett. 67, 2295–2298 (1991)
380. Pendry, J.B., et al.: Magnetism from conductors and enhanced nonlinear phenomena. IEEE
Trans. Microwave Theor. Tech. 47, 2075–2084 (1999)
381. Shelby, R.A., et al.: Experimental verification of a negative index of refraction. Science 292,
77–79 (2001)
382. Singer, S.: The Nature of Ball Lightning. Plenum, New York (1978)
383. Smirnov, B.M.: Physics of ball lightning. Phys. Rep. 224, 151–236 (1993)
384. Amirov, A.K.h., Bychkov, V.L.: Ball lightning diameter-lifetime statistical analysis of the
SKB data bank. Physica Scripta. 50, 413–416 (1994)
385. Keul, A.G., et al.: German ball lightning data bank results. In: Proceedings of the 9th
International Symposium on Ball Lightning, Eindhoven, The Netherlands. pp. 96–105 (2006)
386. Ofuruton, H. et al.: Experimental conditions for ball lightning creation by using air gap
discharge embedded in a microwave field. J. Geophys. Res. 106, 12,367–12,369 (2001)
387. Golubkov, G.V., et al.: Collision of Rydberg atom A** with ground-state atom B. JETP Lett.
75, 314–316 (2002)
388. Klimov, A. et al.: Investigation of physical properties of plasmoids in air and airflow. In:
Proceedings of the 9th International Symposium on Ball Lightning. Eindhoven, The Netherlands.
pp. 112–121 (2006)
372 V.L. Bychkov et al.

389. Shabanov, G.D.: Performance of the laboratory of ball lightning. In: Proceedings of the 9th
International Symposium on Ball Lightning. Eindhoven, The Netherlands. pp. 202–209
(2006)
390. Vlassov, A.: Getting a fireball by means of electric explosion of a wire spiral. In: Proceedings
of the 9th International Symposium on Ball Lightning, Eindhoven, The Netherlands. pp.
233–241 (2006)
391. Jennison, R.C.: Ball lightning. Nature 224, 895 (1969)
392. Dmitriev, M.T.: Stabilility mechanism for ball lightning. Sov. Phys. Tech. Phys. 14, 284–289
(1969)
393. Dwyer, J.R., et al.: Energetic radiation produced during rocket-triggered lightning. Science
299, 4 (2003)
394. Shah, G.N., et al.: Neutron generation in lightning bolts. Nature 313, 773–775 (1985)
395. Dijkhuis, G.C.: Helix string model for turbulent vorticity and cavitation in shearing arc
plasma. Ann. NY Acad. Sci. 891, 259–272 (1999)
396. Dijkhuis, G.C.: Exponential particle acceleration by recursive stretching of vortex filaments
in turbulent discharge plasma. In: Proceedings of the 8th International Symposium on Ball
Lightning, Taiwan, Chung-li, pp. 70–76 (2004)
397. Dijkhuis, G.C.: Scaling law for fusion power from ball lightning. In: Janiszewski, J., Moron,
W., Sega, W. (eds.). Proceedings of the International Wroclaw Symposium on Electromag-
netic Compatibility, pp. 21–25 (1988)
398. Pendry, J.B., et al.: Low frequency plasmons in thin-wire structures. J. Phys. Condens.
Matter. 10, 4785–4809 (1998)
399. Bergström, A.: Electromagnetic theory of strong interaction. Phys. Rev.D. 8, 4394–4402
(1978)
400. Dijkhuis, G.C.: A model of ball lightning. Nature 284, 150–151 (1980)
401. Tsuchiya, H. et al.: Detection of high-energy gamma rays from winter thunderclouds. Phys.
Rev. Lett., 99, 4165002, 4 (2007)
402. Bychkov, V.L.: On observation properties of ball lightning. In: Proceedings of the 9th
International Symposium on Ball Lightning. Eindhoven, The Netherlands. pp 39–42 (2006)
403. Nikitin, A.I. et al.: Estimation of ball lightning characteristics based on the analysis of its
photo. In: Proceedings of the 9th International Symposium on Ball Lightning, Eindhoven,
The Netherlands, pp. 148–156 (2006)
404. Noten, L.C.: Private communication supported by his original laboratory journal. (1995)
405. Dijkhuis, G.C.: Site report and evaluation of Dutch ball lightning event. In: Proceedings of
the 9th International Symposium on Ball Lightning, Eindhoven, The Netherlands, pp. 39–42
(2006)
406. Ofuruton, H., Kamogawa, M., Ohtsuki, Y.-H.: Experiments on ball lightning by using
microwave localization. In: Proceedings of the 9th International Symposium on Ball Light-
ning (ISBL06), Eindhoven, The Netherlands, pp. 157–158 (2006)
407. Emelin, S.E., et al.: The dust-gas fireball as a special form of the electric erosive discharge
afterglow. In: Proceedings of the 9th International Symposium on Ball Lightning, Eindho-
ven, The Netherlands, pp. 51–61 (2006)
408. Pirozerski, A.L., et al.: Long-living plasmoids generation by high-voltage discharge.
In: Proceedings of the 9th International Symposium on Ball Lightning, Eindhoven, The
Netherlands, pp. 180–190 (2006).
409. Spiegel, M.R.: Theory and Problems of Complex Variables. McGraw-Hill, New York (1964)
410. Panofsky, W.K.H., Phillips, M.: Classical Electricity and Magnetism. Addison-Wesley,
Cambridge, MA (1962)
411. Dijkhuis, G.C.: Crystal lattices for 3D vortex systems with uniform curvature and torsion at
minimal dissi-pation. In: Doerffer, P. (ed.) Internals Flows. IFFM Publishers. Gdansk,
Poland, pp. 661–668 (2001)
412. Hamilton, W.R.: On a New Species of Imaginary Quantities Connected with the Theory of
Quaternions. Proc. Roy. Irish Acad. 2, 424–434 (1844)
6 Ball Lightning Investigations 373

413. Dijkhuis, G.C.: On potential flow solutions from the division algebras. In: Proceedings of the
6th European Symposium on Aerothermod. for Space Vehicles, ESA SP-659 (2009)
414. Conway, J.H., Smith, D.A.: On Quaternions and Octonions. A.K. Peters, Natick, MA (2003)
415. Coxeter, H.S.M.: Regular Polytopes, 3rd ed. Dover, New York (1973)
416. Kuipers, J.B.: Quaternions and Rotation Sequences. Princeton University Press, New York
(1999)
417. Abramowitz, M., Stegun, I.A.: Handbook of Mathematical Functions. Dover, New York
(1968)
Index

A G
Aerosols, ix, xi, 69–92, 99–101, 117–122, 130, Global theoretical models, 181
131, 135, 170, 248, 252, 261, 273, 283, Growth, 70, 75, 80–83, 87–92, 100, 101, 111,
284, 297 112, 116, 121–123, 125, 131, 134, 135,
Atmosphere, ix–xii, 3, 20, 21, 41, 49, 56, 237, 239, 285, 286, 292
69–92, 97–135, 150, 161, 163, 164, 170,
175–197, 211, 237, 249, 250, 255, 257, I
262, 264, 266, 273, 274, 277, 278, 280, Ionosphere, ix–xii, 97–135, 148, 150–152,
287, 294, 300, 312, 313, 320, 328 156–161, 164–171, 175–197
Atom-molecular processes, 2 Ionosphere perturbations, 106, 131, 181–183,
185, 192, 193, 195
B
Ball lightning
experiments, xii, 201, 202, 208, 216, 219, N
224, 246–271, 295, 297, 298, 301, 312 Nucleation, 70, 72–80, 82, 87–91, 162
observations, xii, 201–246, 271, 275, 295,
296, 298, 326, 328 P
theoretical models, 201, 202, 222, 227, Plasma disturbances, 98, 105
270–295 Precursors, 83, 87, 89, 100, 117, 127,
132, 148, 150, 151, 154, 161–171,
C 191–195, 236
Condensation, 69, 72, 75, 77–83, 85, 87–89,
203, 256, 257, 264, 273, 286, 290,
291, 355 S
Satellite experiments, 148–156
E
Earthquake prediction, 164 U
Electromagnetic field, 98, 107, 154, 156, 231, Upper atmosphere, ix, x, 3, 175–197, 237
244, 273, 277, 287–290, 336, 345 Upper ionosphere, 100, 101, 105, 109–114,
Elementary chemical reactions, 189 125–127, 129, 132, 134, 135, 178

375