See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/27771849

Earthquakes in India and the Himalaya: Tectonics, geodesy and history

Article in Annals of geophysics = Annali di geofisica · December 2004

DOI: 10.4401/ag-3338 · Source: OAI

CITATIONS READS

425 24,883

1 author:

Roger Bilham
University of Colorado Boulder
311 PUBLICATIONS 19,190 CITATIONS

SEE PROFILE

All content following this page was uploaded by Roger Bilham on 28 May 2014.

The user has requested enhancement of the downloaded file.

 ANNALS OF GEOPHYSICS, VOL. 47, N. 2/3, April/June 2004

Earthquakes in India and the Himalaya:

tectonics, geodesy and history
Roger Bilham
Cooperative Institute for Research in Environmental Sciences (CIRES) and Geological Sciences,
University of Colorado, Boulder, CO, U.S.A.

Abstract
The record of earthquakes in India is patchy prior to 1800 and its improvement is much impeded by its disper-
sal in a dozen local languages, and several colonial archives. Although geological studies will necessarily com-
plement the historical record, only two earthquakes of the dozens of known historical events have resulted in sur-
face ruptures, and it is likely that geological data in the form of liquefaction features will be needed to extend
the historical record beyond the most recent few centuries. Damage from large Himalayan earthquakes record-
ed in Tibet and in Northern India suggests that earthquakes may attain M = 8.2. Seismic gaps along two-thirds
of the Himalaya that have developed in the past five centuries, when combined with geodetic convergence rates
of approximately 1.8 m/cy, suggests that one or more M = 8 earthquakes may be overdue. The mechanisms of
recent earthquakes in Peninsular India are consistent with stresses induced in the Indian plate flexed by its col-
lision with Tibet. A region of abnormally high seismicity in western India appears to be caused by local con-
vergence across the Rann of Kachchh and possibly other rift zones of India. Since the plate itself deforms little,
this deformation may be related to incipient plate fragmentation in Sindh or over a larger region of NW India.

Key words India – earthquakes – history article I shall use the term India to signify both the
Indian tectonic plate and the subcontinent of India.
Perhaps the most disappointing observation is
1. Introduction that despite a written tradition extending beyond
1500 B.C. we know very little about Indian
Throughout the invasions of different ethnic earthquakes earlier than 500 years before the
and religious entitites in the past two millennia the present, and records are close to complete only
Indian subcontinent has been known as Hindoost- for earthquakes in the most recent 200 years. This
an, Hindustan or India in recognition of its unique presents a problem for estimating recurrence in-
isolation imposed by surrounding mountains and tervals between significant earthquakes, the holy
oceans. The northern, eastern and western moun- grail of historic earthquake studies. Certainly no
tains are the boundaries of the Indian plate. The repetition of an earthquake has ever been recog-
shorelines are the echoes of ancient plate bound- nized in the written record of India and the Hi-
aries. Only in recent time have the separate nations malaya, although great earthquakes in the Hi-
of Pakistan, India, and Bangladesh subdivided the malaya should do so at least once and possibly as
continental expression of the Indian Plate. In this much as three times each millennium. The strain
rate within the Indian plate is observed to be less
than 3 ns/yr (Bilham and Gaur, 2000) and the re-
newal time for earthquakes in the sub-continent
Mailing address: Dr. Roger Bilham, Cooperative Insti- may exceed many thousands of years, rendering
tute for Research in Environmental Sciences (CIRES) and
Geological Sciences, University of Colorado, Boulder, CO it unlikely that earthquakes will have repeated
80309-0399, U.S.A.; e-mail:[email protected] during the time of written records.

839
 Roger Bilham

In contrast, trench investigations indicate the arc. Arc-normal convergence across the Hi-
that faults have been repeatedly active both on malaya results in the development of potential
the subcontinent (Sukhija et al., 1999; Rajen- slip available to drive large thrust earthquakes be-
dran, 2000) and within the Himalayan plate neath the Himalaya at roughly 1.8 m/cy, hence
boundary (Wesnousky et al., 1999). The exca- earthquakes associated with, say, 6 m of slip can-
vation of active faults and liquefaction features not occur before the elapse of an interval of at
is likely to play an important role in extending least three centuries (Bilham et al., 1997).
the historical earthquake record of Indian earth- Slip across the 150-300 km wide plate
quakes in the next several decades. boundary between Asia and India in Baluchistan
A feature of Indian earthquakes for which is apparently partitioned between thrust and
numerical deformation data have recently been strike-slip components. For example, the 1931
exhumed is that these data, once analyzed, have Mach Ms = 7.3 earthquake was associated with 1
required substantial revision of earlier in- m of NW directed reverse slip on a fault that may
formed, but speculative, interpretations of the have extended entirely through the crust. It was
causal mechanisms of historic earthquakes. Ge- followed 4 years later by the Ms = 7.7 strike-slip
odetic data have surfaced for the 1819, 1881, Quetta earthquake on a subparallel fault less than
1897 and 1905 earthquakes that have largely 150 km NW of the Mach event. The Mach event
negated the conclusions of many learned arti- slipped in a sense that effectively unclamped the
cles. This obviously raises a cautionary flag: subsequent Quetta earthquake (Ambraseys and
that conclusions concerning felt reports about Bilham, 2003a). Slip on the Chaman fault further
earthquakes in history and prehistory have lim- to the north in Afghanistan in the past century, and
ited value in interpreting subsurface structure. possibly for a longer period, has been much less
I first give a brief overview of Indian tecton- than 42 mm/yr according to seismic moment sum-
ics. I then describe catalogues and data that char- mation of observed seismicity (Ambraseys and
acterize Indian earthquakes, and conclude with a Bilham, 2003b). Although this may be the result of
number of case histories that discuss some of the minor deformation in the northern Afghan moun-
important problems that have surfaced in studies tains, or unreported creep on the Chaman fault, it
of Indian earthquakes, and that may be resolved is quite possible that the northern Chaman fault
by the discovery of further data. I conclude with a system may be overdue for a large earthquake.
discussion of our current understanding of seis- Slip in the IndoBurman ranges is also accom-
mic hazard in India and the Himalaya. panied by strike-slip and thrust seismicity and al-
though no recent large earthquakes have occurred
on land, the north-south Sagaing fault system is
2. Tectonic setting of India clearly strike-slip and the Indo-Burman ranges to
its west the result of distributed east-west conver-
India is currently penetrating into Asia at a gence. Near the Andaman Islands slip is parti-
rate of approximately 45 mm/yr and rotating tioned between thrust earthquakes to the west and
slowly anticlockwise (Sella et al., 2002). This ro- beneath islands, and strike-slip faulting on the
tation and translation results in left-lateral trans- North Andaman fault to their east (Curray et al.
form slip in Baluchistan at approximately 42 1979; 1982; Ortiz and Bilham, 2003)
mm/yr and right-lateral slip relative to Asia in the GPS measurements in India reveal that con-
Indo-Burman ranges at 55 mm/yr (fig. 1). Be- vergence is less than 5 ± 3 mm/yr from Cape
cause of complexities in the structural units at its Comorin (Kanya Comori) to the plains south of
northern, western and eastern boundaries these the Himalaya (Paul et al., 2001). Hence the In-
velocities are not directly observable across any dian Plate should not be expected to host fre-
single fault system. Deformation within Asia re- quent seismicity. However, the collision of India
duces India’s convergence with Tibet to approxi- has resulted in flexure of the Indian Plate (Bil-
mately 18 mm/yr (Wang et al., 2001), and be- ham et al., 2003). The wavelength of this flexure
cause Tibet is extending east-west, convergence is of the order of 650 km and results in an ap-
across the Himalaya is approximately normal to proximately 450-m-high bulge near the central

840
 Earthquakes in India and the Himalaya: tectonics, geodesy and history

Fig. 1. Schematic views of Indian tectonics. Plate boundary velocities are indicated in mm/yr. Shading indi-
cates flexure of India: a 4 km deep trough near the Himalaya and an inferred minor (40 m) trough in south cen-
tral India are separated by a bulge that rises approximately 450 m. Tibet is not a tectonic plate: it extends east-
west and converges north-south at approximately 12 mm/yr. At the crest of the flexural bulge the surface of the
Indian plate is in tension and its base is in compression. Locations and dates of important earthquakes mentioned
in the text are shown, with numbers of fatalities in parenthesis where known. With the exception of the Car Nico-
bar 1881, Assam 1897 and Bhuj 2001 events, none of the rupture zones of major earthquakes are known with
any certainty. The estimated rupture zones of pre-1800 great earthquakes are shown as unfilled outlines, where-
as more recent events are filled white.

841
 Roger Bilham

Indian Plateau, corresponding to the outer rise these thrust events range from tensile just below
of an oceanic collision. Normal faulting earth- the plate interface, to compressional and strike-
quakes occur north of the flexural bulge (e.g., slip at depths of 30-50 km (e.g., M = 6.6 20 Au-
possibly on 15 July 1720 near Delhi) and deep gust 1988, Udaypur).
reverse faulting occurs beneath its crest (e.g., the A belt of microearthquakes and moderate
M = 6.3 21 May 1997 Jabalpur earthquake). earthquakes beneath the Greater Himalaya on
Shallow reverse faulting occurs south of the the southern edge of Tibet indicates a transi-
flexural bulge where the Indian plate is de- tion from stick-slip faulting to probable aseis-
pressed (e.g., the M = 6.3 29 September 1993 mic creep at around 18 km. This belt of mi-
Latur earthquake, fig. 1). croseismicity defines a small circle with ra-
The Indian plate is bent downwards by 4-6 dius 1695 km (Seeber and Gornitz, 1983; Ben-
km beneath the southern edge of the Himalaya dick and Bilham, 2001). Seismicity in Tibet is
attaining depths of 18-20 km beneath the south- largely shallow and is either normal faulting or
ern edge of Tibet (fig. 1). Stresses within the strike-slip faulting.
plate vary from tensile above the flexed neutral- The flexural geometry of the Indian plate is
axis to compressional below it. Where no in- manifest as a standing wave fixed relative to
plane end-loading prevails the position of the southern Tibet. Stresses in the plate vary with
neutral axis lies theoretically half way through time because the Indian plate streams slowly
the thickness of the elastic plate. Since in-plane though this flexural wave, bringing points
stresses of the order of 500 bars exist (necessary within India towards, or away from, compres-
to maintain the height of the Tibetan Plateau) sional or tensile failure. It is for this reason that
this effectively means that the neutral axis rises the earthquakes that occur throughout central
above the plate south of the crest of the central and northern India appear to have no distinc-
Indian bulge. The neutral axis descends into the tive spatial pattern. The flexural stresses are
plate just north of the bulge where it is initially significantly larger than the in-plane stresses
flexed downward. The axis would descend to a needed to sustain the elevation of the Tibetan
path a little above half-way through the plate Plateau, but their change with time is slow
were it perfectly elastic, since the flexural (mbar/yr). Despite this their spatial change is
stresses are much larger than the weak in-plane large (up to 2 bars per km northeastward) and
collisional stresses. However, normal-faulting this results in an important imposed south-north
in the upper surface of the plate near the spatial variation in stress (Bilham et al., 2003).
Ganges Trough weakens the top surface of the Stress changes of less than 1 bar are known to
plate thereby lowering the neutral axis, and trigger earthquakes. Although stresses through-
plastic conditions near the base of the plate both out most of NE India are everywhere close to
raise the neutral axis, both thinning the effec- failure, the triggering of earthquakes occurs
tive elastic thickness and shifting the neutral partly from the movement of India through the
axis to an unknown depth. Eventually, when flexural stress field, and partly from local stress
sufficient focal mechanisms are available from perturbations caused by other tectonic, erosion-
the descending plate, it may be possible to iden- al or dynamic processes.
tify the location of the neutral axis from the ab-
sence of earthquakes near the axis, and from the
difference in mechanisms above and below it. 3. Historical data sources and catalogues
The presence of both flexural stresses and
plate-boundary slip permits all mechanisms of Early earthquakes described in mythical
earthquakes to occur beneath the Lesser Hi- terms include extracts in the Mahabharata
malaya (fig. 1). At depths of 4-18 km great (≈ 1500 B.C.) during the Kurukshetra battle
thrust earthquakes with shallow northerly dip (Iyengar, 1994), and several semi-religious
occur infrequently that permit the northward texts that mention a probable Himalayan earth-
descent of the Indian plate beneath the subcon- quake reputed to have occurred during the time
tinent. Earthquakes in the Indian plate beneath of enlightment of Buddha ca. 538 B.C.

842
 Earthquakes in India and the Himalaya: tectonics, geodesy and history

Fig. 2. Maps in 1690 and 1740 show Debil near the current location of Karachi. Other maps show it on a dis-
tributary of the Indus. An earthquake occurred there in 1668 and another is alleged by Thomas Oldham (1883)
to have occurred in 893 but the event he invokes occurred in an Armenian town with a similar name (Ambraseys,
2004). The city is last mentioned in the 18th century (Yule et al., 1903)

Archeological excavations in Sindh and Gu- earthquake because it is alleged to have been
jarat suggest earthquake damage to now aban- followed by three years of aftershocks, but the
doned Harrappan cities. A probable earthquake absence of reports from other locations renders
around 0 A.D. near the historically important this of little value in estimating its rupture di-
city of Dwarka is recorded as a zone of lique- mensions or magnitude. Similarly the arrival of
faction in archeological excavations of the an- Vasco de Gama’s fleet in 1524 coincided with a
cient city (Rajendran et al., 2003). The town of violent sea-quake and tsunami that caused alarm
Debal (Dewal, Debil, Diul Sind or Sindi) near at Dabul (Bendick and Bilham, 1999). Note that
the current site of Karachi was alleged to have this Portuguese port at latitude 17°34’ on the
been destroyed in 893 A.D. (Oldham, 1883), Malabar Coast is unrelated to Debil above. This
but until recently accounts of its collapse and could have been a local event, but since it was
inundation were considered too vague to be tak- not reported onshore it could have been the
en seriously. Rajendran and Rajendran (2002) tsunami from a remote earthquake that occurred
present a case that the destruction of Debil was along the Makran or Gujarat coastlines. Such
caused by an earthquake linked to the same accounts are thus of fragmentary value in quan-
fault system responsible for the 1819 and 2001 tifying earthquake locations and sizes.
Rann of Kachchh earthquakes, however, Am- The emergence and disappearance of
braseys (2003) notes that Oldham’s sources re- coastal tracts has sometimes been ascribed to
fer to Daibul (Dvin) in Armenia, and that lique- earthquakes. A storm near Cochin in 1341
faction 1100 years ago in western India must be caused an island to emerge, but inspection sug-
attributed to a different earthquake. Figure 2 gests this to be a common accretional feature of
shows the location of Debil west of the Indus storms along the Malabar Coast (Bendick and
delta in a 1690 map drafted by A.D. Winter. Bilham, 1999). An island that sank in 1769
Other maps place it within the distributaries of south of Chittagong (Oldham, 1883) may have
the Indus. Yule and Burnell (1903) describe De- undergone lateral spreading at the time of sig-
bil’s 1000-year-long history, prior to its effec- nificant earthquake near there (Seeber, pers.
tive disappearance from accounts within a cen- comm., 2003).
tury of a second earthquake in its vicinity in In the mid 19th century some of these frag-
1668 (Oldham, 1883; Ambraseys, 2004). mentary data were collected successively in
A single paragraph describes a massive summaries of earthquakes by Mallett, Baird-
earthquake in the Kathmandu Valley in 1255 Smith and Oldham, but there followed more
(Wright, 1877), which may have been a great than a century of archival neglect when little

843
 Roger Bilham

new information surfaced. The seismicity of the en (Chitrakar and Pandey, 1986). Earthquakes
sub-continent has been summarized in compila- in the 18th century are poorly documented. An
tions by Chandra (1977), Srivastava and Ra- earthquake near Delhi in 1720 caused damage
machandram, (1985), Rao and Rao (1984) and and apparent liquefaction but little else is
by Khattri (1987). Recent interest in early known of this event (Kahn, 1874; Oldham,
earthquakes have engaged historians in India 1883). This event, from its location, could have
and elsewhere in a systematic search through been a normal faulting event, but because of the
Urdu, Arabic, Tibetan, Chinese, Nepalese and absence of damage accounts from the Hi-
European languages. malaya it may have been a Himalayan earth-
Two important publications summarize re- quake. In 1713 a severe earthquake damaged
cent findings: Iyengar and Sharma (1998) re- Bhutan and parts of Assam (Ambraseys and
port accounts in Arabic, Sanskrit and Urdu Jackson, 2003).
sources and Ambraseys and Jackson (2003) Thirteen years later, in September 1737, a
provide new data from Tibet and recently col- catastrophic earthquake is alleged to have oc-
lated colonial records. Data presented in these curred in Calcutta. This is the most devastating
publications remain sparse but provide a skele- earthquake to be listed in many catalogues of
tal framework of events on which to build a fu- Indian (and global) earthquakes but is actually a
ture quantitative assessment of historic Indian storm surge that resulted in numerous deaths by
earthquakes as new documents surface. drowning along the northern coast of the Bay of
A list of Indian earthquakes is to be found Bengal. The handwritten ledgers of the East In-
in Bapat et al. (1983) but this contains numer- dia Company in Bengal detail storm and flood
ous entries that have been included uncritically damage to shipping, wharves, warehouses and
from secondary sources, and for these reasons dwellings in Calcutta with an estimate of 3000
can be misleading. Similarly, entries in the un- deaths by drowning (Bilham, 1994). Calcutta’s
critical listing of Dunbar et al. (1992) require population at the time was approximately
careful evaluation before use. A useful and eas- 30 000. A figure of 300 000 fatalities is often as-
ily accessible compilation of information and cribed to this «fake quake» for which earth-
resources for the study of Indian earthquakes is quake shaking was probably invoked in news
a webpage maintained by Stacey Martin, reports as a metaphor for destruction, a possible
http://asc-india.org/menu/gquakes.htm. Relo- description of the buffeting accompanying ex-
cated instrumental earthquakes are listed by treme wind velocities. The spire of St. Annes
Engdhal et al. (1998). church, Calcutta, was blown down by these
An important recent realization is that a se- winds, but the masonry church survived. An ap-
quence of significant earthquakes occurred proximate 10% increase in burials is recorded in
throughout the west Himalaya in the 16th centu- its churchyard for 1737, an increase in deaths
ry. The sequence started in Kashmir in 1501, that year by fewer than two dozen. Although the
followed by two events a month apart in death-toll from drowning along the coast of
Afghanistan and the central Himalaya, conclud- southern Bengal was presumably greater than
ing with a large earthquake in Kashmir in 1555. the official estimates in Calcutta, the fatality
The central Himalayan 1505 earthquake may count of 300 000 is repeated only in accounts
have been Mw ≥ 8.2 based on its probable rup- published in monthly magazines and newspa-
ture area. It destroyed monasteries along a 500 pers in Europe, and is not substantiated by offi-
km segment of southern Tibet, in addition to de- cial documents from any of the several adminis-
molishing structures in Agra and other towns in trative centers then functioning in Bengal.
northern India (Jackson, 2002; Ambraseys and India in the early 19th century was as yet
Jackson, 2003; Bilham and Ambraseys, 2004). incompletely dominated by a British colonial
A Himalayan earthquake that damaged the administration. Remote administrators in dis-
Kathmandu Valley in 1668 is mentioned briefly tant parts of the India subscribed to newspapers
(a single sentence) in Nepalese histories but as and wrote verbose and sometimes extensive de-
with events in 1255 and 1408 no details are giv- scriptions of their experiences which were typ-

844
 Earthquakes in India and the Himalaya: tectonics, geodesy and history

ically printed and circulated to each administra- military service who were in NW India at the
tive outpost. An earthquake in India was some- time. Baird-Smith’s accounts of other earth-
thing of a rarity and generated detailed letters quakes include citations from his sources.
from residents describing its effects. Very often The director of the Geological Survey of In-
the same report would be copied verbatim from dia, Thomas Oldham (1816-1878) published
one newspaper and reported by another. Few of the first real catalog of significant Indian events
the original letters have survived, but the earth- in 1883. His catalog includes earthquakes from
quakes in Kumaon in 1803, Nepal in 1833 and 893 to 1869, and acknowledges the works of
Afghanistan in 1842 were felt sufficiently Mallet and Baird-Smith. His important addi-
widely to lead scientifically inclined officials to tions include verbatim textual extracts with ref-
take a special interest in the physics and geog- erences that permit verification and further
raphy of earthquakes. Mallett’s (1853-1855) work. His notes on some of the earthquakes
global catalogues of earthquakes included sev- form the first detailed studies of individual
eral from India, with a special section devoted earthquakes in India.
to the 1833 earthquake for which he discussed His son, Richard. D. Oldham (1858-1936),
seismic propagation velocities. wrote accounts of four major Indian earthquakes
At about the time of the sequential publica- (1819, 1869, 1881 and 1897). He completed first
tion of Mallet’s global catalogue an army officer his father’s manuscript on the 1869 Silchar,
named Baird-Smith wrote a sequence of articles Cachar, Assam earthquake which was published
(1843a,b,1844) in the Asiatic Society of Bengal under his father’s name (Oldham, 1884). He next
summarizing data from several Indian earth- investigated the Mw = 7.9 December 1881 earth-
quakes and venturing to offer explanations for quake in the Andaman Islands, visiting and map-
their occurrence. He was writing shortly after the ping the geology of some of the islands (Old-
first Afghan war which had coincided with a ma- ham, 1884, 1885). He mistakenly located the
jor 1842 earthquake in the Kunar Valley of NE event deep in the northern Bay of Bengal based
Afghanistan (Ambraseys and Bilham, 2003b), largely on timing data from clocks in Calcutta
which must have impressed him and others in the and Madras. An analysis of the tsunami generat-

Fig. 3. Oldham, father and son, were both geologists in India. Thomas Oldham (left) compiled the first cata-
logue of Indian earthquakes. Richard (right) made definitive studies of individual earthquakes (1819, 1869, 1881
and 1897) in addition to identifying for the first time p-waves and s-waves, and the core of the earth.

845
 Roger Bilham

ed by this earthquake places it on the subduction quake on the northern edge of the Rann of
zone west of Car Nicobar (Ortiz and Bilham, Kachchh close to what is now the India/Pak-
2003). His account of the 1897 Mw = 8.1 Shil- istan border. The earthquake figures prominant-
long Plateau earthquake in Assam (Oldham, ly in Lyell’s Principles of Geology (1830) as
1899) was exemplary, and according to Richter one of the first clear examples of geological up-
provided the best available scientific analyses of lift associated with an earthquake.
available physical data on any earthquake at the Oldham’s 1928 account refers to, but does
time. In contrast to the care with which he in- not reproduce, Baker’s map and profile from a
vestigated the geological, geodetic and geophys- leveling survey crossing the Allah Bund. This
ical aspects of the earthquake, Oldham’s reports profile is key to quantifying the mechanism of
are thin on specific accounts of building damage the earthquake, and it is entirely due to its
which he felt were often exaggerated. Despite serendipitous discovery by Oldham (1898) that
the care with which he interpreted the intensity we have access to it. The map had been acci-
data available to him, his estimated intensities dently omitted in Baker’s original 1846 publi-
for the 1897 earthquake on a modified version of cation by the editor. In a frontispiece to the Ge-
the Rossi-Forel scale are 1.5 to 3 intensity units ographical Society of Bombay in 1846 he
too high in the epicentral region (Ambraseys and apologizes for omitting the map and cross-sec-
Bilham, 2003c). tion and promises to include the figure in sub-
R.D. Oldham’s accounts established a tem- sequent issues, a promise that he failed to ful-
plate for the study of earthquakes that occurred fill. Oldham had discovered the map quite by
in India subsequently. The great earthquakes of accident when supervising a clean-up of the
1905 Kangra (Middlemiss, 1910) and 1934 Bi- Bombay office of the Survey of India. In his
har/Nepal (Dunn et al., 1939) were each as- discussion of the cause of the 1819 Allah Bund
signed to Geological Survey of India special earthquake Oldham speculates that the mor-
volumes, but these never quite matched the in- phology across the natural dam measured by
sightful observations of Oldham’s 1899 volume. Baker in 1846 was caused by subsurface fault-
Investigations of the yet larger Assam earth- ing akin to that reported from Japanese earth-
quake of 1950 were published as a compilation quakes in the early 20th century.
undertaken by separate investigators (e.g., Pod- Assuming the surface morphology to be
dar, 1953; Ray, 1953; Tandon, 1953). In many representative of co-seismic deformation dur-
ways this proved to be the least conclusive of the ing a single earthquake, Baker’s 6 m crest-to-
studies of the 5 largest Indian earthquakes 1819- trough observation is consistent with 11 m of
1950. Information available to Indian authors on slip on a north-dipping reverse fault terminating
the effects of the earthquake were confined 0.5-2 km below the surface (Bilham, 1999).
largely to a narrow corridor of information However, recent geological studies in the re-
along the Brahmaputra valley since access to Ti- gion (Rajendran and Rajendran, 2001 and
bet, Burma, or the tribal regions south of the epi- 2002) have raised the possibility that the ob-
center was unavailable. Regrettably geologists served morphology was a factor of two smaller
did not make a thorough search for surface fault- than that reported by Baker, and that its current
ing in the epicentral region and geodesy near the elevation of < 3 m crest-to-base is caused part-
epicenter was virtually non-existent. ly by the 1819 event and partly by pre-1819
earthquakes. A difficulty in rejecting Baker’s
survey, a canal engineer of repute, is that he
4. Uncertainties associated with the 6 June would have made vertical errors of less than a
1819 Allah Bund earthquake few cm in measuring topography over the 10
km width of the Allah Bund. Thus an error of 2-
Oldham wrote his account of the 1819 3 m can be rejected. The cross-section that was
earthquake in Kachchh in retirement in England intended to accompany Baker’s account was
(Oldham, 1928). His monograph synthesized drafted from a larger scale survey deposited
all the data available for the Allah Bund earth- with the Sind government. The smaller version

846
 Earthquakes in India and the Himalaya: tectonics, geodesy and history

published by Oldham included a typographical of frontal fill and co-seismic slip could be tested
error in the vertical scale, but it is unlikely that with suitable excavations, or seismic profiles, of
gross drafting errors would have been intro- the northern edge of the bed of Lake Sindri.
duced. Moreover, the accompanying map view While excavations of Sindri sedimentation
of the river system is exact in many details might clarify the discrepancy between historic
compared to recent satellite photos suggesting leveling and current morphology, the observa-
that its execution was fastidious. tion by Rajendran and Rajendran that two or
Several explanations can be invoked to rec- more earthquakes caused incremental changes
oncile the leveling data and current morpholo- in the height of the Allah Bund requires down-
gy. The first is that the uplift and subsidence ward revision of the 11 m estimate of coseismic
morphology may have changed since the slip to a more modest 5 m. Any further reduction
earthquake. For example, it is possible that in the coseismic uplift of the Allah Bund can be
Baker’s measurements started at a lower verti- rejected based on Baker’s mapping of the eleva-
cal datum than that available to the Rajendrans tion of the bed of the Narra River since this
in 2000. According to Burnes (1833) the foot- would have been at river base-level before the
wall subsided by 1-3 m, with maximum subsi- earthquake, unaffected by previous earthquakes.
dence near the scarp. Burnes’s two handwrit- The recent 26 January 2001 Bhuj earthquake
ten accounts in the Geological Society of Lon- was associated with 3-6 m of slip (Bendick et
don describe slightly different views of the al., 2001). Since this occurred on a 40 × 40 km
river cut through the Allah Bund in 1827 and rupture, and resulted in isoseismal intensity dis-
1828 that suggest it was evolving in response tributions throughout India similar to the 1819
to the flood of 1826. Currently the sediments earthquake (Hough et al., 2002), it is tempting to
of Lake Sindri slope upwards towards the assume that the two events had similar stress
southern edge of the Allah Bund. In the past drops and local attenuation relationships, and
180 years sediments eroded from the front of somewhat similar geometry and magnitude. This
the scarp, supplemented by sediments from would require the along-strike length of the Al-
the Narra River in flood, would have filled any lah Bund earthquake to be shortened consider-
depression fronting the scarp along the north- ably below the > 100 km length first suggested
ern shore of Lake Sindri resulting in a datum by Oldham and adopted by all later authors. In
possibly 2 m higher than that available to Bak- contrast, Ambraseys and Douglas (2004) favor a
er. The Rajendrans were unable to map verti- Mw = 8.19 magnitude for this event, requiring
cal profiles northward into the Sindh province rupture dimensions consistent with those in-
of Pakistan hence it may not have been possi- ferred by Oldham (1928).
ble to recover Baker’s northern datum.
A second possibility is to assume that the
southern edge of the Allah Bund has now been 5. Himalayan earthquakes 1 September
eroded 1 km or more northward by monsoon 1803 and 26 August 1833
winds and floods driving waves across the 30-
50 km wide fetch of open water to its South. In These earthquakes occurred at the western
1827 the crest of the scarp was fewer than 600 and eastern ends of the inferred 6 June 1505
m from its southern edge. Ablation of the crest earthquake. The first of these events occurred
of the Allah Bund may have also occurred al- during the opening battles of the 2nd war against
though this is considered unlikely because Ra- the Mahrattas. In late August 1803 a British
jendran and Rajendran report the survival of Army had laid seige to the fort and town of Ali-
surface geodetic monuments installed in 1860. garh on the banks of the Calini River (between
The subsidence deformation profile, now the Ganges and Jumna) some 200 km from the
buried beneath Lake Sindri, may in fact be better Himalaya. The commander of the British Army,
preserved than the uplift profile, and this, at Lt. General Lake, writing to Wellesley on 1 Sep-
some future date, may provide additional con- tember indicates that the strength of the defences
straints of slip in the 1819 earthquake. The depth will require a one month seige. Yet, not three

847
 Roger Bilham

days later Lake writes again to Wellesley that not affect western Nepal and its magnitude is
they have successfully stormed the town with too small to have had much effect on releasing
minor loss of life. In contrast to Lake’s silence on strain accumulated since the 1505 earthquake.
the earthquake that occurred between the two let- However, had it occurred on the plate boundary
ters, a member (Thorn, 1818) of the besieging «detachment fault» it could not have released
army describes violent shaking for 2 minutes at much of the slip available to drive the larger
midnight accompanied by collapse several build- magnitude 1934 earthquake a century later.
ings. The earthquake appears in part responsible Since the 1934 earthquake is believed to have
for the successful capture of the fort, either from released up to 8 m of slip, and since potential
damage to its walls or distress to inhabitants, al- plate-boundary slip is renewed at a rate of less
though specific details are lacking. than 2 m per century, the 1833 rupture would
A summary of materials available for the have had to occur on different fault systems or to
1803 event is recorded by Ambraseys and Jack- have slipped on a small patch contiguous to the
son (2003) who assign it an approximate mag- 1934 rupture. One possibility is that one or more
nitude of Mw = 7.5. This was later revised to of the three 1833 earthquakes occurred deep in
Mw = 8.09 by Ambraseys and Douglas (2004) the Indian plate where both strike-slip and thrust
using additional materials, who place it at the faulting can occur, or that all three earthquakes
weatern end of the 1505 rupture. The 1833 were M ≥ 7.5 thrust earthquakes at the Northern
erathquake almost exactly 30 years later oc- edge of the 1934 rupture zone, similar to those
curred at the eastern end of the 1505 rupture. In that have occurred in the past several decades in
contrast to the extensive damage reported from western Nepal.
Tibet in 1505, few accounts of damage have
surfaced from Tibetan sources for these two
earthquakes, suggesting that they were signifi- 6. Cachar 10 January 1869
cantly less severe than the 1505 event. The one
exception to the apparent silence from Tibet for This M > 7 earthquake occurred in the Syl-
the 1833 earthquake are accounts of damage het region (Silchar) of what is now NE
from members of the Nepal quinquennial trib- Bangladesh. Although numerous accounts of
ute delegation returning from Beijing who this earthquake were compiled by the Old-
brought with them accounts of the increasing hams the data are insufficient to estimate a
damage they encountered as they approached causal fault or a precise magnitude. Am-
the northern Nepal border (Bilham, 1995). braseys and Douglas (2004) estimate
The Ms ≥ 7.7 August 1833 earthquake near Mw = 7.39. The most likely fault to be associ-
Kathmandu consisted of three shocks (Bilham, ated with this earthquake is the eastern ex-
1995). The first caused alarm and the second, 5 tremity of the Dauki fault, as hinted by God-
hours later, brought most people from their win-Austin (1869) who was undertaking first-
homes. The mainshock (Mw = 7.69, Ambraseys order triangulation in the region at the time.
and Douglas, 2004) occurred 15 minutes later Few first hand accounts of the event exist out-
causing widespread structural damage in India side the covers of Oldham (1884) but the oc-
and Nepal, but the combined loss of life in India casional letter describing its effects surfaces.
and Nepal was only 500 because most people An example is reproduced below:
were already in the open, alarmed by the two «The earthquake has not been a single shock
foreshocks. Newspaper reports of these events but has lasted, on and off, a month - nay it is
are abundant as are scientific commentaries in said some of the shocks have gone on rocking
journals in India and Europe. The isoseismals for five minutes by the watch till some people
from this earthquake suggest an epicentral re- were literally sea sick. The bazaar at Silchar
gion similar to, or at the western end of, the (the capital of Cachar) is the handsomest street
1934 Ms = 8.1 rupture, which together with the anywhere east of Calcutta and it has been en-
multiple shocks in the event, raises a number of gulfed. i.e. it has gone bodily down not at once
interpretational difficulties. The earthquake did but in a series of descents, some ten feet at a

848
 Earthquakes in India and the Himalaya: tectonics, geodesy and history

time. The river in Silchar in the cold weather ondary earthquake triggered by the mainshock.
runs about 50 feet below the level of its banks Its timing would have to have occurred within a
which are only dried mud, and the country has few minutes of the mainshock for it to have pro-
been so rocked up and down till the river has cut duced the sea wave observed at Port Blair. Local
its banks right down to its own level and the populations were concentrated in only two is-
plain at Silchar is all one debris with no partic- lands and therefore there is no corroboration of
ular river anywhere. The commissioner told me this inferred northern region of submarine fault-
on Monday last that it was officially reported ing which occurred between them. It is probable
that the only thing left standing at Silchar was that offshore corals may be of use in recon-
Clarke’s bridge, and it was the most wonderful structing an extended history of earthquakes in
sight that ever was seen.» (Clarke, 1869) the Andaman-Nicobar Islands. The island of Car
The dangers of speculating on a causal fault Nicobar is believed to have been raised and tilt-
or mechanism for the 1869 earthquake are high- ed during the 1881 event. Deformation models
lighted by radical errors of a century of inter- that do not include this uplift result in an inap-
pretations of the 1897 earthquake that were propriate estimate of the observed tsunami run-
shown to be baseless once the geodetic signal up on the island (Ortiz and Bilham, 2003).
was assessed in 2001. It is possible that enough
of the 1869 geodetic survey network was in
place prior to the earthquake to render its re- 8. 1897 Shillong Plateau Earthquake
measurement even now of value.
The 1897 Great Assam earthquake (Ms = 8.0)
for more than a century was believed to have oc-
7. The 31 December 1881 Mw = 7.9 curred on a thrust fault dipping gently to the
Car Nicobar earthquake north. Some considered it to have been a Hi-
malayan basal thrust. We now recognize that the
This earthquake caused minor damage in earthquake occurred on a reverse fault dipping
the Andaman Island Penal colony and generat- steeply to the south. Slip during the 1897 earth-
ed a tsunami that was observed throughout the quake may have exceeded 16 m, resulting in 10
Bay of Bengal but not along the Burmese m uplift of the northern edge of the plateau.
coast. The tsunami did no damage around the Oldham clearly recognized the value of sur-
Bay of Bengal where tide gauges recorded a face deformation as a quantitative measure of
maximum amplitude of 0.8 m (Oldham, 1884). what happens in an earthquake, but the analyti-
An analysis of five tide gauge records reveals cal tools to interpret these data were not to
that the earthquake was Mw = 7.9 ± 0.2 and oc- emerge for a further half century. In 1897 cor-
curred on an east-dipping thrust fault below respondence with the Surveyor general, Sydney
and to the west of Car Nicobar, an island at Burrard, Oldham requested a geodetic re-survey
9°N midway between the Andaman and Nico- of the Shillong Plateau. The work undertaken
bar islands (Ortiz and Bilham, 2003). GPS by J. Bond covered only the southern half of the
measurements at Port Blair indicate oblique plateau and was considered by Burrard (1898)
convergence of the plate boundary (Paul et al., to be inferior in accuracy to normal survey stan-
2001). The earthquake is believed to have oc- dards because numerous triangles did not close
curred on the interface between the Indian and precisely. (A test of survey accuracy is whether
Andaman Plates and the inferred mechanism of angles in a triangle after correcting for spherical
westward slip of the hanging wall slip is con- excess add up to 180°). We now know that these
sistent with slip partitioning between the dip- misclosures were probably due to postseismic
ping subduction zone, and the strike-slip west adjustments in the epicentral region continuing
Andaman fault east of Car Nicobar. after the earthquake. The 1897 displacement re-
A feature of this earthquake is the inferred sults available to Oldham were ambiguous: ei-
presence of a region of minor slip NE of the ther the plateau had bodily expanded and risen
main rupture zone. This may have been a sec- with no southward motion, or it had risen with-

849
 Roger Bilham

out strain and moved southwards by slip on the raises additional concerns. No historical earth-
Dauki fault bordering its southern edge. Realiz- quakes have been recorded on this fault, and
ing this, Oldham urged resurvey of the northern many previous studies interpret the fault as a
half of the plateau but he was destined never to dextral strike-slip fault. Although the fault may
see the data since it was completed in 1936, the have slipped differently in the past there is little
year he died (Davison, 1936). Analyses of angle doubt that reverse slip is now the prevailing
changes between 1869 and 1936 reveal that mechanism, and has been so for the past one or
Oldham’s instincts were correct. The fault that two million years. Earthquakes beneath the
slipped in 1897 was a 110-km-long blind re- plateau have thrust mechanisms parallel to the
verse fault beneath the northern edge of the strike of the Oldham fault at depths of more
plateau, dipping southward at 45° with 16 ± 5 than 35 km. The 1869 Cachar earthquake de-
m of slip between 9 and 39 km (Bilham and scribed by the Oldhams may have occurred at
England, 2001). We named this unmapped fault the Eastern end of the Dhauki fault (Godwin-
the Oldham Fault in his honour. Austin, 1869; Oldham, 1884).
The earthquake raised the northern edge of A recent review of instrumental records of
the plateau roughly 10 m. The causal fault is be- the 1897 earthquake reveals its teleseismically
lieved to have cut right through the lower crust derived magnitude to be Ms = 8.0 (Ambraseys,
but did not approach closer than 9 km to the 2000) effectively the same as its geodetic seis-
Earth’s surface. Oldham (1899) photographed mic moment of M = 8.1 (Bilham and England,
secondary faulting of up to 10 m at the western 2001). A re-evaluation of Oldham’s 1897 iso-
end of the plateau on the Chedrang fault. The seismal intensity data supplemented by addi-
1.6 km mean-height of the plateau surface ap- tional data from newspapers, diaries, books and
pears to have been driven to its current position government reports unavailable to Oldham, re-
by reverse faults acting on both its northern and veal significantly reduced areas for contours of
southern edges. Three dissected terraces border I > VIII isoseismals, but similar areas for lower
the northern edge of the plateau that may be intensity shaking. The newly evaluated intensi-
separated by active faults, but none have been ties include data from Tibet and Bhutan (Am-
mapped by geologists possibly due to the thick braseys and Bilham, 2003c).
forest cover that makes access difficult.
Enigmatic aspects of this earthquake concern
the uniqueness of the Shillong plateau which 9. Kangra 1905 M = 7.8 earthquake
permits contraction of the Indian plate within 80
km of the Himalaya convergence zone, thereby Occurring just 7 years after the 1897 Assam
reducing the productivity of Himalayan earth- earthquake, the Kangra event found the geolo-
quakes. An uplift rate of 2.5 ± 1 mm/yr can be gists of India eager to map the details of the
calculated from the current elevation of the event. The earthquake had its oddities – in par-
plateau, and from the date of its initial elevation ticular a prominent epicentral region of Rossi-
estimated from changes in sedimentation styles Forel shaking of intensity VIII to X near Kangra
in northern Bangladesh. This convergence re- and Dharmsala and an island of VIII shaking al-
quires a convergence rate of 4 ± 2 mm/yr, or ap- most 250 km to the SE near Dehra Dun. This,
proximately a factor 4 less than the India/Tibet and an artificially-inflated estimate for magni-
convergence rate (Bilham and England, 2001). tude (Richter rounded Gutenberg’s calculated
The only large historical earthquake known in magnitude upward from M = 7.8 to M = 8; see
the Bhutan Himalaya is the 1713 event described Ambraseys and Bilham, 2000), led several in-
in Ambraseys and Jackson (2003) and the pre- vestigators to assume that rupture may have ex-
cise location of this event is far from certain. tended more than 350 km along strike.
The southern edge of the Shillong Plateau is Although geodetic measurements existed
truncated by the Dhauki fault. In order that the along the probable southern edge of the rupture,
surface of the Plateau be horizontal the Dhauki no remeasurements were made after the earth-
fault must also act as a reverse fault, and this quake except near the remote region of high ac-

850
 Earthquakes in India and the Himalaya: tectonics, geodesy and history

celerations near Dehra Dun. No horizontal de- The primary ruptures of the largest mid-plate
formation was detected and a vertical deforma- events of the past two centuries, the 7.8 < M < 8.1
tion signal, though discussed by many subse- 1819 Allah Bund, the M = 8.1 1897 Shillong, the
quent investigators, has recently been dismissed M = 7.3 1931 Mach, and the M = 7.6 2001 Bhuj
as an artifact of the leveling process (Bilham, earthquakes have all been on blind thrust faults,
2001). Hence there is little evidence to believe dipping at approximately 45°, terminating 1-9
that its rupture exceeded 200 km. First-order tri- km below the surface, and extending to the base
angulation prior to the earthquake is limited to of the crust. Thus, although they have caused
the southern edge of the inferred rupture zone widespread destruction in the historical record,
and it appears not to have been re-measured the geological manifestation of their passage is
since its initial measurement in 1845. An inter- limited to secondary cracks and liquefaction phe-
pretation of a GPS occupation of some of these nomena that tell us little about their mechanisms.
points in 2001 is currently underway. Intensities Such knowledge about rupture geometries as we
of this event are re-evaluated by Ambraseys and have obtained for these earthquakes, with the ex-
Douglas (2004). ception of the most recent, has been derived al-
most entirely from sparse geodetic data.
The conclusion to be derived from this ab-
10. Discussion sence of surface ruptures in the subcontinent is
that many historical earthquakes occurred on
The above review of early earthquakes and faults that are currently unmapped, and the
case histories of some of the larger earthquakes, corollary is that there may exist many hun-
omits numerous smaller ones felt by individuals dreds of subsurface faults potentially awaiting
or communities. The larger ones form a patchy re-activation for which we have no geological
history that may be complete for the past 200 intelligence.
years, but which is certainly missing many The mechanisms of the numerous smaller
large earthquakes before then. An important shocks that appear in historical Indian cata-
question is whether there is scant information logues must be inferred from modern focal
on pre-muslim or medieval earthquakes be- mechanisms in those same geographic settings.
cause there were few events, or whether it is be- The inherent problem in doing this is that focal
cause we have no records of them. Although mechanisms in some parts of India, e.g., the Hi-
this question cannot be answered from the his- malayan foothills, vary with depth. Surviving
torical record alone we may consider extreme intensity data are rarely adequate to distinguish
scenarios as a guide to future searches to re- between deep and shallow shocks.
solve the issue.
Aggravating our lack of knowledge of previ-
ous earthquake is the curious observation that 10.1. Intensity and attenuation
none of the numerous earthquakes that have oc-
curred in India and the Himalaya in the past sev- Estimates of intensities for the two largest
eral centuries have produced surface ruptures, earthquakes of the past two centuries (1905 and
with the exception of secondary surface faulting 1897) have revealed that previous estimates of
in the 1897 earthquake (Oldham, 1898), and sur- Rossi-Forel or Modified Mercalli intensity tend
face fractures of the 1993 Latur earthquake (See- to exaggerate high intensity shaking by 1-3 in-
ber et al., 1996). In 1505 and 1892 surface fault- tensity units (Ambraseys and Bilham, 2003c)
ing was observed at the surface along the Pak- whereas lower intensities (V-II) are estimated
istan/Afghanistan border (Ambraseys and Bil- with reasonable accuracy. The reason for this
ham, 2003a) but no surface faulting has ever exaggeration is that the style of building con-
been reported in the Himalayan and Indo-Bur- struction suffers significant damage at intensi-
man plate boundaries, despite geological indica- ties around VII-VIII and that subsequent shak-
tions that surface rupture of the frontal faults has ing produces somewhat imperceptible addition-
occurred in the past (Wesnousky et al., 1999). al damage (Ambraseys and Bilham, 2003b).

851
 Roger Bilham

Even quite recent intensity estimates can be quakes, the 6 June 1505 Kumaon/western Nepal
suspect. For example, the 1989 Udaypur earth- earthquake (Jackson, 2002; Ambraseys and
quake in southern Nepal resulted in both Nepali Jackson, 2003), may have exceeded Mw = 8.2,
(Pandey and Nicolas, 1989; Dikshit and and its recurrence now would result in a similar-
Koirala, 1989) and Indian (Sinha, 1993) inten- sized earthquake (9 m of slip along a 500-600
sity and engineering damage studies. The re- km rupture zone). Damage in northern India was
sulting intensity contours show an abrupt jump considerable during the 1505 event and it is like-
of 1-1.5 intensity units at the Nepal/India bor- ly that its recurrence would damage many of the
der where the two studies abut. large cities along the Ganges and Jumna rivers
The re-evaluation of the felt intensity reports through shaking, and from the effects of exten-
for the 1833, 1897, 1905, 1934 and 1950 earth- sive liquefaction. Smaller seismic gaps are evi-
quakes on a common scale is an important pri- dent in Kashmir, in Sikkim and in Assam for
ority, that has been partly completed by Am- which the historical record is ambiguous or ab-
braseys and Jackson (2003), since it may reveal sent.
the details of seismic hazards in intervening re- Assuming that seven to ten great ruptures
gions where future Himalayan earthquakes are permit the slip of the entire Himalayan Arc, and
anticipated. Currently more than three scales a recurrence interval of 500 years (≥ 9 m slip on
have been used to report these data. Rossi-Forel, 200-300 km long, 70-90 km wide, ruptures) we
Modified Mercalli and MSK intensities, with should anticipate M ≥ 8 earthquakes occurring
caveats imposed by their specific inapplicability every 50-70 years. Insufficient earthquakes
to Indian building methods. In some areas ac- have occurred recently to match this estimate.
celeration damage can only with difficulty be Two great earthquakes only that approach this
distinguished from collapse caused by liquefac- severity have occurred in the past 200 years
tion-induced foundation failure. In 1897 regions (1934 and 1950), and two others are known in
of extensive liquefaction and catastrophic later- the previous 300 years (Kashmir, 1505, 1555).
al spreading follow the banks of the main rivers No great earthquake has occurred for 53 years.
and result in building damage from foundation Almost two thirds of the Himalaya remain un-
collapse, rather than grades of shaking intensity. broken by recent earthquakes, suggesting that
Ambraseys and Bilham (2003c) separated lique- several seismic gaps may currently exist. Final-
faction observations from MSK assignations ly, the summation of seismic moment from all
based on shaking intensity lest they bias the ar- known earthquakes since 1505 along the entire
eas of isoseismal contours. arc yields a slip rate less than 30% of that de-
rived from the current geodetic slip rate (Bil-
ham and Ambraseys, 2004).
10.2. Himalayan recurrence interval From these arguments we may form one of
two conclusions: that one or more great Hi-
The recurrence interval for great Himalayan malayan earthquakes are overdue, or that our
earthquakes remains conjectural since the histor- understanding of the way in which the northern
ical record is probably incomplete even for the plate boundary slips is flawed. The case for the
past 500 years. A summary of those events for imminent failure of a seismic gap is hampered
which we have data is depicted in fig. 4a,b, al- by the absence of any well documented recur-
though both the rupture area and the amount of rence interval, or indeed any evidence for regu-
slip are unknown for each of these events. The lar failure of the Himalayan plate boundary.
figure suggests that the western Himalaya may The absence of constraint permits the extreme
have slipped in a sequence of events between view, for example, that failure occurs in clus-
1501 and 1555, and that since then there have tered sequences, as may have occurred in the
been relatively modest earthquakes, insufficient western Himalaya 1400-1555. If indeed this se-
to release the 1.5-1.8 m per century of accumu- quence released accumulated displacements in
lating convergence revealed from geodetic meas- the western Himalaya five centuries ago, then a
urements. The largest of the pre-1900 earth- case can be made for the immediate recurrence

852
 Earthquakes in India and the Himalaya: tectonics, geodesy and history

Fig. 4a,b. a) The Himalaya describe a small circle that subtends an arc of approximately one radian symmet-
rically about the Thakola Graben in Tibet. Important mid-plate earthquakes named. Microseismicity follows this
small circle (from Engdhal et al., 1998) and is negligible to its south where great earthquakes are located. b) The
arcuate box is expanded (and straightened) and it shows time distance plot of approximate rupture areas of large
earthquakes in the past eight centuries plotted along the arc (approximate transverse-Mercator projection of lin-
ear transverse-km versus angular distance). One or more large earthquakes appear to be overdue in Kashmir, Ku-
maon and Western Nepal. We know of no earthquakes in Sikkim, and the 1897 Assam Shillong earthquake may
have reduced the slip potential in Eastern Bhutan. Pre-1500 earthquakes are known with less certainty. Trench
studies have revealed slip on the frontal thrusts at the beginning of the 15th century at several locations west of
Dehra Dun (Senthil Kumar, per. comm., 2004) and surface rupture on frontal thrusts in eastern Nepal may cor-
respond to the earthquake that destroyed Kathmandu in 1255 (Rockwell, pers. comm., 2004).

853
 Roger Bilham

of one or more 9 m slip events, based on the that have been administered continuously by a
current convergence rate of 18 mm/yr. The re- record-keeping population for the past thousand
gion of the 1505 earthquake has been hitherto years. For this reason, trench investigations of
termed the Central Himalayan seismic gap by faults and liquefaction features will be necessary
Khattri and Tyagi (1983) and Khattri (1987). to fill in the record. In practice, the subsurface
Alternatively, the assumption that great record of strong-ground motion is complete, but
earthquakes are essential features for plate its interpretation may be non-unique, suitable
boundary slip may be incorrect – the Himalaya conditions may not exist everywhere for it to be
may fail in smaller events that incompletely rup- recorded, and it is insensitive to small earth-
ture the plate boundary. These smaller events quakes whose recurrence may be quite damaging.
might be considered to be similar to the ChiChi
earthquake of 1999 that ruptured through a mid-
level segment of the accretionary wedge, rather 10.3. Mid-plate recurrence intervals
than through a basal detachment. Such events
may accommodate convergence without trans- Although numerous micro-earthquakes, and
lating the entire Himalaya southward over India. many damaging shocks have occurred in the
The major 1833, 1885 and 1905 earthquakes past several centuries in India, the geodetic sta-
(7.5 < Mw < 7.8) may have been examples of bility of the plate, and the absence of recent
these «out-of-sequence thrusts». mountain ranges indicates that earthquakes
One of the most troubling observations, that should not recur repeatedly on the same fault
might be accounted for by out-of sequence during the written history of India. Yet archeo-
thrusting or deep events, is that no recent Hi- logical observations in India suggest earth-
malayan earthquake has ever resulted in a quakes may have repeatedly destroyed early
recorded surface rupture. Such ruptures have settlements there, especially in westernmost In-
obviously occurred in recent geological time, dia. Rajendran et al. (1996) present evidence
on the main frontal thrusts for example (Wes- for reactivation of the fault causal to the Latur
nousky et al., 1999), signifying either that re- earthquake. The town of Latur itself, like many
cent earthquakes are anomalously small, or that Indian villages, is a mound city built on the ru-
the search for surface rupture may not have ins of previous cities.
been exhaustive. If some, but not all, great The occurrence of the M = 7.6 Bhuj 2001
earthquakes rupture the Himalayan frontal earthquake less than two centuries after the
thrusts we cannot hope to quantify the recur- M ≥ 7.8 Allah Bund 1819 earthquake has been
rence interval from these events using paleosis- considered by some investigators to represent a
mic fault-trenching methods. short recurrence interval for earthquakes in
Out of sequence thrusts cannot represent a mid-plate India. Such an appraisal would be in-
steady-state condition for Himalayan slip since correct because the two events did not, of
it would not explain the geological observation course, rupture the same fault, nor even the
of occasional slip on the basal thrust systems same fault system. The two earthquakes oc-
and Main Frontal Thrusts. However, it is possi- curred on the ancient Kachchh rift zone, an
ble that excessive recent erosion of the Hi- east-west fault system that can be traced struc-
malayan foothills may have upset the uniform turally from near Karachi to Ahmedabad. In a
taper of the Himalayan accretionary wedge study of the 1819 event it was concluded that
such that adjustments are now underway that contiguous future faulting might be anticipated,
result in a predominance of high-level thrusting with specific concern that rupture to the west
interspersed with infrequent basal thrusts. would create hazards for Karachi (Bilham,
Historical studies have an important role in 1999). As it happened, rupture in 2001 occurred
distinguishing between these various scenarios, 2-4 rupture lengths to the east of the 1819
yet it is unlikely that we shall ever find a history earthquake. Hence there is a possibility that the
that is complete across- and along- the Himalaya, entire Kachchh rift may be converging. In fact
even near the Kathmandu and Kashmir Valleys geodetic data suggest that the rift north of the

854
 Earthquakes in India and the Himalaya: tectonics, geodesy and history

Bhuj region may have converged by more than able records and their loss through fire, war
1 m since 1856 (Sri Devi et al., 2003). Should and decay. Despite their sparseness it is like-
this be the case, additional large earthquakes ly that documents on historical earthquakes
may be anticipated both to the east and west of will surface in Tibetan, Urdu and Arabic
the 1819 and 2001 earthquakes. records that will change current estimates of
The observed geodetic convergence of the the significance of seismic gaps in the Hi-
Rann of Kachchh by 9 ± 3 mm/yr is approximate- malaya, and may change our understanding
ly 2-3 times larger that the entire geodetic con- of earthquakes within the Indian continent.
vergence rate between Northern and Southern In- Our current understanding of Himalayan
dia (Paul et al., 2001). Two explanations for this earthquakes is such that we may calculate po-
have been proposed: one is that a 400 km wide tential slip in several segments of the plate
continental «Sindh flake» is in the process of boundary, but we cannot estimate the timing
fracturing from the NE edge of the Indian plate of future events. Making assumptions about
(Stein et al., 2002), the other is that the intercon- the probable completeness of the historical
nected ancient rift systems of northern India de- seismic record we can estimate estimate min-
fine a small northern plate « the Harappan plate» imum slip potential based on the time since
that allows a large triangle in NE India between the last known earthquake (Bilham et al.,
the central Himalaya and Bhuj to converge with 2001). This has moderate relevance to plan-
the main body of the Indian plate to the south ning for future earthquakes. The eventual es-
(Bilham et al., 2003). Support for either mecha- tablishment of recurrence intervals for Hi-
nism of plate fragmentation is weak, and future malayan ruptures will require a combination
geodetic observations are needed to resolve the of serendipitous historical studies and geolog-
extent of plate deformation in NE India. ical trench investigations of faulting and
earthquake-induced liquefaction features.

11. Conclusions
Acknowledgements
The tectonic setting of India’s collision with
Asia is now reasonably well characterized from R.S. Cox of the American Philosophical
recent seismicity and geodetic studies of rela- Society provided a copy of C.B. Clarke’s let-
tive motion at their plate boundaries. Direct ter concerning the Silchar earthquake. I
measurements across and within the Himalaya thank the organizers of the Erice conference
reveal a locking line beneath the edge of the Ti- on historical earthquakes for inviting my par-
betan Plateau and the absence of creep to its ticipation. I also thank an anonymous re-
south (Bilham et al., 1995, 1998, 2001), imply- viewer for amusing and insightful comments
ing that the advance of the Himalaya over the on the first draft of this article. The study was
Indian plate proceeds largely through the recur- funded by the National Science Foundation
rence of great plate boundary earthquakes. EAR-0003449.
Earthquakes within the Indian plate are at-
tributable to the superposition of the NW com-
pressional stress of collision, on the stresses REFERENCES
arising from plate flexure. The depth and
mechanisms of recent earthquakes reflect the AMBRASEYS, N.N. (2000): Reappraisal of north Indian
sense of these combined stresses. earthquakes at the turn of the 20th century, Curr. Sci.,
79, 101-114.
A several millennia-long written record in AMBRASEYS, N.N. (2004): A note on three little-known ear-
India has revealed few major earthquakes pri- ly earthquakes in India, Curr. Sci., 86 (4), 506-508.
or to the past two centuries. This is partly due AMBRASEYS, N.N. and D. JACKSON (2003): A note on early
to the fact that extant records have yet to be earthquakes in northern India and southern Tibet, Curr.
Sci., 84 (4), 571-582.
searched rigorously for earthquakes, but is in AMBRASEYS, N.N. and R. BILHAM (2000). A note on the
part due to the corruption of potentially valu- Kangra Ms = 7.8 earthquake of 4 April 1905, Curr.

855
 Roger Bilham

Sci., 79 (4), 101-106. (9), 1259-1269.

AMBRASEYS, N.N. and R. BILHAM (2003a): Earthquakes and BILHAM R., P. BODIN and M. JACKSON (1995): Entertaining
crustal deformation in northern Baluchistan 1892- a great earthquake in western Nepal; historic inactivity
2001, Bull. Seismol. Soc. Am. 93 (4), 1573-1600. and geodetic tests for the development of strain, J.
AMBRASEYS, N.N. and R. BILHAM (2003b): Earthquakes in Nepal Geol. Soc., Kathmandu, 11, 73-88.
Afghanistan, Seismol. Res. Lett., 74 (2), 107-123. BILHAM, R., K. LARSON, J. FREYMUELLER and PROJECT IDYL-
AMBRASEYS, N.N. and R. BILHAM (2003c): MSK Isoseismal HIM MEMBERS (1997): GPS measurements of present-
intensities evaluated for the 1897 Great Assam Earth- day convergence across the Nepal Himalaya, Nature,
quake, Bull. Seismol. Soc. Am., 93 (2), 655-673. London, 386, 61-64.
AMBRASEYS, N.N. and J. DOUGLAS (2004): Magnitude cali- BILHAM, R., F. BLUME, R. BENDICK and V. K. GAUR (1998):
bration of North Indian earthquakes, Geophys. J. Int. Geodetic constraints on the Traslation and Deforma-
(in press). tion of India: implications for future great Himalayan
BAIRD-SMITH, R. (1843a): Memoir on Indian earthquakes, earthquakess, Curr. Sci., 74 (3), 213-229.
Part I, J. Asiatic Soc. Bengal, 11, 1046-1054. BILHAM, R., V.K. GAUR and P. MOLNAR (2001): Himalayan
BAIRD-SMITH, R. (1843b): Memoir on Indian Earthquakes, Seismic Hazard, Science, 293, 1442-1444.
Part II, J. Asiatic Soc. Bengal, 12 (2), (New Series, BILHAM, R., R. BENDICK and K. WALLACE (2003). Flexure
136), 1029-1056. of the Indian Plate and Intraplate Earthquakes, Proc.
BAIRD-SMITH, R. (1844): Memoir on Indian Earthquakes, Indian Acad Sci., Earth Planet Sci., 112 (3), 1-14.
Part III, J. Asiatic Soc. Bengal, 156, 964-983. BURNES, L.A. (1833): A memoir of the eastern Branch of
BAKER, W.E. (1846): Remarks on the Allah Bund and on the the Indus giving an account of the alterations produced
drainage of the Eastern part of the Sind Basin, Trans. in it by the earthquake of 1819 and the bursting of the
Bombay Geog. Soc., 7, 186-188. dams in 1826; also a theory of the Runn’s formation
BAPAT, A., R. KULKARNI and S. GUHA (1983): Catalogue of and some surmises in the route of Alexander the Great,
earthquakes in India and neighbourhood from histori- presented by the Hon’ble the Governor in Council to
cal period up to 1979, Indian Soc. Earthquake Eng., the Lit. Soc. of Bombay. Camp at Lucput, 28 March
Roorkee. 1827, 13 August 1828 (manuscript presented by the au-
BENDICK, R. and R. BILHAM (1999): Search for buckling of thor to the Geol. Soc. London 18 December 1833 with
the southwest Indian coast related to Himalayan colli- map inscribed W. Ballantine 1831).
sion, in Himalaya and Tibet: Mountain Roots to Moun- CHANDRA, U. (1977): Earthquakes of Peninsular India - A
tain Tops, edited by A. MACFARLANE, R. SORKHABI and seismotectonic study, Bull. Seismol. Soc Am., 67 (5),
J. QUADE, Geol. Soc. Am. Spec. Pap. 328, 313-323. 1387-1413.
BENDICK, R. and R. BILHAM (2001): How perfect is the Hi- CHITRAKAR, G.R. and M.R. PANDEY (1986): Historical
malayan Arc?, Geology, 29, 791-794. earthquakes of Nepal, Bull. Geol. Soc. Nepal, 4, 7-8.
BENDICK, R., R. BILHAM, E. FIELDING, V.K. GAUR, S. CLARKE, C.B. (1869): Letter to father (American Philo-
HOUGH, G. KIER, M.N. KULKARNI, S. MARTIN, K. sophical Society, Philadelphia), ms.
MUELLER and M. MUKUL (2001): The January 26, 2001 CURRAY, J.R, F.J. EMMEL, D.G. MOORE and R.W. RAITT
«Republic Day» Earthquake, India, Seismol. Res. Lett., (1982): Structure, tectonics and geological history of the
72 (3), 328-335. NE Indian Ocean, in The Indian Ocean: The Ocean
BILHAM, R. (1994): The 1737 Calcutta Earthquake and Cy- Basins and Margins, edited by E.M. NARRIN and F.G.
clone Evaluated, Bull. Seismol. Soc. Am., 84 (5), 1650- SEHLI (New York, Plenum Press), 399-450.
1657. CURRAY, J.R., F.J. EMMEL, D.G. MOORE, R.W. RAITT, M.
BILHAM, R. (1995): Location and magnitude of the 1833 HENRY and R. KIECKHEFER (1979): Tectonics of the An-
Nepal earthquake and its relation to the rupture zones daman sea and Burma, in Geological and Geophysical
of contiguous great Himalayan earthquakes, Curr. Sci., Investigations of Continental Margins, edited by J.S.
69, 101-128. WATKINS, L. MONTADERT and P. DICKERSON, AAPG
BILHAM, R. (1999): Slip parameters for the Rann of Memoir, 29, 189-198.
Kachchh, India, 16 June 1819, earthquake, quantified DAVISON, C. (1936): R.D. Oldham: obituary, Nature, 138,
from contemporary accounts, in Coastal Tectonics, ed- 316-317.
ited by I.S. STEWART and C. VITA-FINZI, Geol. Soc. DIKSHIT, A.M. and A. KOIRALA (1989): Report on the In-
London Spec. Publ. 146, 295-318. tensity Mapping of Udaypur Earthquake of 20 August
BILHAM, R. (2001): Slow tilt reversal of the Lesser Hi- 1988 (HMG Ministry of Industry, Nepal, Dept. of
malaya between 1862 and 1992 at 78°E, and bounds to Mines and Geology, Lainchaur, Kathmandu).
the southeast rupture of the 1905 Kangra earthquake, DUNBAR, P.K., P.A. LOCKRIDGE and L.S. WHITESIDE (1992):
Geophis. J. Int., 144, 1-23. Catalog of significant earthquakes, including quantita-
BILHAM, R. and N.N. AMBRASEYS (2004): Apparent Hi- tive casualties and damage (National Oceanic and At-
malayan slip deficit from the summation of seismic mospheric Adminstration, National Environment Satel-
moments for Himalayan earthquakes, 1500-2000, Curr. lite Data and Information Service, National Geophysi-
Sci. (in press). cal Data Center), Report SE-49, pp. 320.
BILHAM, R. and P. ENGLAND (2001): Plateau «pop-up» in DUNN, J.A., J.B. AUDEN, A.M.N. GOSH and S.C. ROY
the great 1897 Assam earthquake, Nature (London), (1939): The Bihar-Nepal Earthquake of 1934, Mem.
410 (6830), 806-809. Geol. Soc. India, Surv. India, Calcutta, 73, pp. 391
BILHAM, R. and V.K GAUR (2000): Geodetic contributions (reprint, 1981).
to the study of seismotectonics in India, Curr. Sci., 79 ENGDAHL, E.R., R.D. VAN DER HILST and R.P. BULAND

856
 Earthquakes in India and the Himalaya: tectonics, geodesy and history

(1998): Global teleseismic earthquake relocation with SON, M.B. ANANDA, S. JADE, M. MUKAL, T.S. ANUPA-
improved travel times and procedures for depth deter- MA, G. SATYAL and D. KUMAR (2001): The motion and
mination, Bull. Seismol. Soc. Am., 88, 722-743. active deformation of India, Geophys. Res. Lett., 28
GODWIN-AUSTIN, (1869): Notes from Assaloo, North (4), 647-651.
Cachar, on the Great Earthquake of 10 January 1869, PODDAR, M.C. (1953): A short note on the Assam earthquake
Proc. Asiatic Soc. Bengal, 15 March 1869, 91-99. of Aug. 15, 1950, in A Compilation of Papers on the As-
HOUGH, S.E., S. MARTIN, R. BILHAM and G. ATKINSON sam Earthquake of August 15, 1950, edited by M.B.R.
(2002): The January 26 2001, Bhuj India earthquake: RAO, Publ. No. 1 (Central Board of Geophysics, Gov. of
Observed and predicted Ground Motions, Seismol. Soc. India), 38-42.
Am., 92 (6), 2061-2079. RAJENDRAN, C.P. (2000): Using geological data for earth-
IYENGAR, R.N. (1994): Earthquake History of South India, quake studies: a perspective from peninsula India,
The Hindu, 23 January. Curr. Sci., 79 (9), 1251-1258.
IYENGAR, R.N. and D. SHARMA (1998) Earthquake history RAJENDRAN, C.P. and K. RAJENDRAN (2001): Characteristics
of India in medieval times, Rep. Central Build. Res. In- of deformation and past seismicity associated with the
st., Roorkee, pp. 124. 1819 Kutch earthquake, Northwestern India, Bull. Seis-
JACKSON, D. (2002): The great western-Himalayan earth- mol. Soc. Am., 91 (3), 407-426.
quake of 1505: a rupture of the central Himalayan RAJENDRAN, C.P. and K. RAJENDRAN (2002): Historical con-
gap?, in Tibet, Past and Present, edited by H. BLEZER straints on previous seismic activity and morphologic
(Brill’s Tibetan Studies Library I, Leiden), 147-159. changes near the source zone of the 1819 Rann of
KHAN, K. (1874): Muntakhab-ul Lubab, M.H., Bibl. India Kachchh earthquake: further light on the penultimate
Series, Calcutta. event, Seismol. Res. Lett., 73 (4), 470-479.
KHATTRI, K.N. (1987): Great earthquakes, seismicity gaps RAJENDRAN, C.P., K. RAJENDRAN and B. JOHN (1996): The
and potential for earthquake disaster along the Hi- Killari (Latur), Central India, earthquake: an example
malaya plate boundary, Tectonophysics, 138, 79-92. of fault reactivation in the Precambrian crust, Geology,
KHATTRI, K.N and A.K. TYAGI (1983): Seismicity patterns in 24, 651-654.
the Himalayan plate boundary and identification of areas RAJENDRAN C.P, K. RAJENDRAN, K.H. VORA and A.S. GAUR
of high seismic potential, Tectonophysics, 96, 281-297. (2003): The odds of a seismic source near Dwarka, NW
LYELL, C. (1830): Principles of Geology (John Murray, Gujarat: An evaluation based on proxies, Curr. Sci., 84,
London). 695-701.
MALLET, R. (1853): Catalogue of recorded earthquakes RAO, B.R. and P.S. RAO (1984): Historical seismicity of
from 1606 B.C. to 1850, Brit. Ass. Rep., 1853, 1-176; Peninsular India, Bull. Seismol. Soc. Am., 74 (6), 2519-
1854, 118-213; 1855, 2-326. 2533.
MIDDLEMISS, C.S. (1910): The Kangra Earthquake of 4 RAY, S. (1953): Isoseismals for the great Assam Earthquake
April 1905, Mem. Geol. Soc. India, Geol. Surv. India, of August 15 1950, in A Compilation of Papers on the
38, pp. 409. Assam Earthquake of August 15, 1950, edited by M.B.R.
OLDHAM, R.D. (1884): Note on the earthquake of 31 De- RAO, Publ. No. 1 (Central Board of Geophysics, Gov. of
cember 1881, Rec. Geol. Surv. India, 17 (2), 47-53. India), 35-37.
OLDHAM, R.D. (1885): Geology of the Andaman Islands SEEBER , L. and V. GORNITZ (1983): River profiles along the
with references to the Nicobars, Rec. Geol. Surv. India, Himalayan arc as indicators of active tectonics,
18 (3), 135-145. Tectonophysics, 92, 335-367.
OLDHAM, R.D. (1898). A note on the Allah Bund in the SEEBER, L., G. EKSTROM, S.K. JAIN, C.V.R. MURTHY, N.
north west of the Runn of Cuch. Mem. Geol. Soc. In- CHANDAK and J.G. AMBRUSTER (1996): The 1993
dia, Geol. Surv. India, 28, 27-30. Killari earthquake in Central India: a new fault in
OLDHAM, R.D. (1899): Report on the Great Earthquake of Mesozoic basalt flows?, J. Geophys. Res., 101 (B4),
12 June 1897, Mem. Geol. Soc. India, Geol. Surv. In- 8543-8560.
dia, 29, pp. 379. SELLA , G.F., T.H. DIXON and A. MAO (2002): REVEL: a
OLDHAM, R.D. (1928): The Cutch (Kacch) earthquake of model for recent plate velocities from space geodesy, J.
16. June 1819 with a revision of the great earthquake of Geophys. Res., 107, doi: 10.1029/2000JB000033.
12. June 1897, Mem. Geol. Soc. India, Geol. Surv. In- SINHA, K.K. (1993): Isoseismal Studies, Bihar-Nepal Earth-
dia, 46, 71-147. quake August 20 1988, Geol. Surv. India., Spec. Publ.
OLDHAM, T. (1883): A catalogue of Indian earthquakes, 31, pp. 104.
Mem. Geol. Surv. India, Geol. Surv. India, 19, 163-215. SRI DEVI JADE and 12 OTHER AUTHORS (2003): Pre-seismic, co-
ORTIZ, M. and R. BILHAM (2003): Source area and rupture seismic and post seismic displacements associated with
parameters of the 31 December 1881 Mw = 7.9 Car the Bhuj earthquake derived from recent and historic ge-
Nicobar earthquake estimated from Tsunamis recorded odetic data, J. Ind. Inst Sci., 112 (3), 331-346.
in the Bay of Bengal, J. Geophys. Res., 108 (B4), doi: SRIVASTAVA, N.N. and K. RAMACHANDRAM (1985): A new
2002JB001941RR2003. catalogue of earthquakes for Peninsular India during
PANDEY, M.R. and M. NICOLAS (1989): The Aftershocks se- 1839-1900, Mausam, 36 (30), 351-358.
quence of the Indian-Nepalese earthquake of August SUKHIJA, B.S., M.N. RAO, D.V. REDDY, P. NAGABSHANAM,
20, 1988, Report 2 March 1989 (HMG Ministry of In- S. HUSSAIN, R.K. CHADHA and H.K. GUPTA (1999):
dustry, Nepal, Dept. of Mines and Geology, Lainchaur, Timing and return of major paleoseismic events in the
Kathmandu). Shillong Plateau, India, Tectonophysics, 308, 53-65.
PAUL, J., R. BURGMANN, V.K. GAUR, R. BILHAM, K.M. LAR- STEIN, S., G. SELLA and E.A. OKAL (2002): The January 26,

857
 Roger Bilham

2001 Bhuj earthquake and the diffuse boundary of the HAM, K.M. LARSON, XI’AN LAI, X. YOU, Z. NIU, J. WU,
Indian plate, Amer. Geophys. Un. Geodyn. Monogr., Y. LI, J. LIU, Z. YANG and Q. CHEN (2001): Present day
30, 243-254. crustal deformation in China constrained by global po-
TANDON, A.N. (1953): The very great earthquake of August sitioning measurements, Science, 294, 574-577.
15 1950, in A Compilation of Papers on the Assam WESNOUSKY, S.G., S. KUMAR, R. MOHINDRA and V.C.
Earthquake of August 15, 1950, edited by M.B.R. RAO, THAKUR (1999): Holocene slip rate of the Himalaya
Pub. No. 1 (Central Board of Geophysics, Gov. of In- frontal thrust of India - Observations near Dehra Dun,
dia), 80-89. Tectonophysics, 18, 967-976.
THORN, W. (1818): Memoir of the War in India Conducted WRIGHT, D. (1877): History of Nepal (reprint, Ranjan Gup-
by General Lord Lake and Major General Sir Arthur ta, Calcutta, 1966, pp. 271).
Wellesley, Duke of Wellington (T. Egerton Military Li- YULE, H. and A.C. BURNELL (1903): Hobson-Jobson: a
brary, Whitehall, London), 93-94. Glossary of Colloquial Anglo-Indian Words (New Del-
WANG, Qi, PEI-ZHEN ZHANG, J.T. FREYMUELLER, R. BIL- hi, John Murray), pp. 1021 (reprint, 1995, 320-321).

858

View publication stats