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

net/publication/252112220

A modelling system for predicting urban air pollution

Article  in  Atmospheric Environment · December 2000

DOI: 10.1016/S1352-2310(00)00073-X

CITATIONS READS

64 106

5 authors, including:

Ari Karppinen Jaakko Kukkonen

Finnish Meteorological Institute Finnish Meteorological Institute
177 PUBLICATIONS   3,404 CITATIONS    269 PUBLICATIONS   5,947 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

NordicWelfAir View project

Dissertation View project

All content following this page was uploaded by Jaakko Kukkonen on 18 October 2017.

The user has requested enhancement of the downloaded file.

 AEA=2815=Durai=Venkatachala=BG

Atmospheric Environment 34 (2000) 3723}3733

A modelling system for predicting urban air pollution:

model description and applications in the Helsinki
metropolitan area
A. Karppinen *, J. Kukkonen , T. ElolaK hde
, M. Konttinen , T. Koskentalo
,
E. Rantakrans
Finnish Meteorological Institute, Air Quality Research, Sahaajankatu 20 E, 00810 Helsinki, Finland

Helsinki Metropolitan Area Council, Opastinsilta 6 A, 00520 Helsinki, Finland
Received 24 November 1999; accepted 11 December 1999

Abstract

We have developed a modelling system for evaluating the tra$c volumes, emissions from stationary and vehicular
sources, and atmospheric dispersion of pollution in an urban area. The dispersion modelling is based on combined
application of the urban dispersion modelling system (UDM-FMI) and the road network dispersion model (CAR-FMI).
The system includes also a meteorological pre-processing model and a statistical and graphical analysis of the computed
time series of concentrations. The modelling system contains a method, which allows for the chemical interaction of
pollutants, originating from a large number of urban sources. This paper presents an overview of the modelling system
and its application for estimating the NO and NO concentrations in the Helsinki metropolitan area in 1993.
V 
A companion paper addresses comparison of model predictions with the results of an urban measurement network. This
modelling system is an important regulatory assessment tool for the national environmental authorities.  2000
Elsevier Science Ltd. All rights reserved.

Keywords: Model; Urban; Air pollution; Nitrogen oxides; Chemical transformation

1. Introduction On the other hand, various local scale Gaussian models

using the Pasquill (or equivalent) stability classes are still
Urban scale dispersion modelling systems have been widely used in practical applications in many European
developed in many European countries; these have re- countries.
cently been reviewed by Moussiopoulos et al. (1996). The The urban scale modelling systems should be able to
latest generation of local scale models is used in combina- allow for the various local scale e!ects, for instance, the
tion with meteorological pre-processing models, which in#uence of buildings and obstacles, downwash phe-
are based on scaling theories of the atmospheric bound- nomena and plume rise, together with chemical trans-
ary layer (ABL). Examples of such models are the Danish formation and deposition (e.g., Kukkonen et al., 1997).
OML model (Olesen, 1995a), the UK-ADMS system of The modelling systems also need information concerning
the United Kingdom (Carruthers et al., 1995) and the the hourly tra$c volumes, travel speeds and emissions
models applied in this study, UDM-FMI (Karppinen et from various urban mobile pollution sources.
al., 1998c) and CAR-FMI (HaK rkoK nen et al., 1995, 1996). This paper describes an integrated urban pollution
modelling system and discusses predicted concentra-
tion distributions of nitrogen oxides in the Helsinki
metropolitan area in 1993. Some selected results have
* Corresponding author. Fax: #358-9-1929-5403. been published previously by Karppinen et al. (1997a,
E-mail address: ari.karppinen@fmi." (A. Karppinen). Karppinen et al. (1997a, 1998a, b). A companion paper

1352-2310/00/$ - see front matter  2000 Elsevier Science Ltd. All rights reserved.
PII: S 1 3 5 2 - 2 3 1 0 ( 0 0 ) 0 0 0 7 4 - 1
 Aea=2815=Durai=VVC=BG

3724 A. Karppinen et al. / Atmospheric Environment 34 (2000) 3723}3733

addresses the testing of the model against the results of as though they resulted from a time sequence of di!erent
an air quality monitoring network. steady states (commonly taken as 1 h). However, this
This atmospheric dispersion modelling is based on assumption can be invalid particularly during peak con-
a combined application of the urban dispersion model- centration episodes, caused by accumulation of air pollu-
ling system UDM-FMI and the road network dispersion tion in an urban area. In contrast, in Eulerian grid or
model CAR-FMI (contaminants in the air from a Road), Lagrangian models, meteorology, dispersion and chem-
both of these models have been developed at the Finnish istry can be described (at least in principle) in real time,
Meteorological Institute (FMI). Both dispersion models selecting a suitable numerical time step.
include a treatment of the chemical transformation of Clearly, modelling of chemical transformation and de-
nitrogen oxides. The dispersion modelling system was position processes is not so straightforward in Gaussian
re"ned to take into account the chemical interaction of modelling systems, compared with Eulerian grid models
pollutants from a large number of individual sources. or Lagrangian models. However, some solutions have
This novel modelling method allows for the interdepen- been presented (for instance, Benson, 1984, 1992; HaK r-
dence of urban background NO, NO and O concentra- koK nen et al., 1996).
 
tions, and NO and NO emissions from various sources.

In a previous study by Valkonen et al. (1995, 1996), the 2.2. The integrated modelling system
spatial distribution of concentrations and the statistical
parameters of CO, NO and SO concentrations were
  Fig. 1 shows the overall structure of the modelling
numerically evaluated in the city of Espoo in southern system applied in this study. The system includes the
Finland in 1990, with the UDM-FMI model. The com- following models: (i) the estimation of tra$c volumes and
bined dispersion modelling system UDM-FMI and travel speeds with the EMME/2 transportation planning
CAR-FMI has been applied in air quality assessments, system (INRO, 1994), (ii) the computation of emissions
which have been conducted nationally in numerous cities from vehicular sources, using the EMME/2 and LIISA
(for instance, Pietarila et al., 1997). systems (MaK kelaK et al., 1996), (iii) the model for evaluat-
ing the emissions from stationary sources, (iv) the me-
teorological pre-processing model MPP-FMI, developed
2. The mathematical models at the FMI (Karppinen et al., 1997b, 1998c), (v and vi) the
dispersion models for stationary and mobile sources,
2.1. Comparison of various modelling approaches UDM-FMI and CAR-FMI, and (vii) post-processing
models, including a statistical and graphical analysis of
An advantage of Gaussian modelling systems is that the computed time series of concentrations.
these can treat a large ensemble of emission sources, The urban dispersion modelling system UDM-FMI
dispersion situations, and a receptor grid network, which takes into account all stationary sources and the road
is su$ciently dense spatially (of the order of tens of network dispersion model CAR-FMI all tra$c sources,
meters). For instance, an hourly time series of one year is respectively. The programs have been executed on the
required, in order to evaluate statistical concentration Cray C94 supercomputer.
parameters, which have been de"ned in air quality guide-
lines (for instance, various percentile values).
Computations of the present study contain approxim-
ately 5000 line sources, a couple of hundred stationary
sources, an hourly time series of one year (8760 meteoro-
logical and emission situations) and about 10 000 recep-
tor points. The dispersion equations therefore had to
be solved separately of the order of 5;10;10;10"
5;10 times. The transformation chemistry computa-
tions increased the computational time even further. The
computations therefore required 24 h CPU-time on
a super-computer Cray C94. The corresponding compu-
tations using an Eulerian grid model (for instance,
Yamartino et al., 1992; Nikmo et al., 1999) would not be
numerically possible using the presently available com-
puter resources.
Limitations of Gaussian urban modelling systems in-
clude that these are based on so-called quasi-steady-state
assumptions (for instance, Seinfeld and Pandis, 1998). It
is assumed that pollutant concentrations can be treated Fig. 1. An overview of the modelling system.
 Aea=2815=Durai=VVC=BG

A. Karppinen et al. / Atmospheric Environment 34 (2000) 3723}3733 3725

2.3. Trazc volumes and emissions The emission factors of cars in city tra$c are based on
tra$c cycle measurements in Helsinki. The emission fac-
We conducted an extensive emission inventory in the tors of the LIISA system (MaK kelaK et al., 1996) were used
Helsinki metropolitan area in 1993, which included the for road tra$c and for heavy duty vehicles.
emissions from various mobile sources (road tra$c, har- The stationary sources considered include energy pro-
bours and marine tra$c, and aviation) and stationary duction, industry and residential heating systems. These
sources (power plants, other point sources and residential are considered as point or area sources. The computa-
heating). Table 1 presents a summary of the total emis- tions included 5000 line sources, 169 point sources, area
sions of NO . Approximately, one half of the total NO sources and the regional background concentrations.
V V
emissions originates from stationary sources.
In the spatial concentration distributions presented in 2.4. Atmospheric boundary-layer scaling
this paper, we have neglected the emissions from marine
tra$c and from aviation, as their in#uence is local and in The relevant meteorological parameters for the models
general small (about 6% of all the emissions of NO
V are evaluated using data produced by a meteorological
within the study area), compared to emissions from roads pre-processing model (Karppinen et al., 1997b, 1998c).
and streets. The in#uence of marine tra$c and aviation The model is based mainly on the energy budget method
on the NO and NO concentrations in the Helsinki
V  of van Ulden and Holtslag (1985). The model utilises
metropolitan area has been discussed in more detail by meteorological synoptic and sounding observations, and
Pesonen et al. (1996). its output consists of estimates of the hourly time series of
The emissions from vehicular tra$c were estimated as the relevant atmospheric turbulence parameters (the
follows. First, the so-called tra$c demand matrices Monin-Obukhov length scale, the friction velocity and
(which describe the vehicle trips from every spatial zone the convective velocity scale) and the boundary layer
to all the other zones within the area considered) were height.
formed, based on a comprehensive inquiry in 1988. The We have made use of the meteorological database of
matrices were "rst evaluated for three speci"c times: our institute, which contains routine weather and sound-
morning and afternoon tra$c peak hours and an average ing observations. Fig. 2 shows the location of the study
non-peak hour. We then utilised the regression coe$- area, the meteorological stations and the background
cients, which were based on tra$c countings performed air quality measurement stations. We used a combina-
by the city of Helsinki and the Finnish National Road tion of the data from the stations at Helsinki-Vantaa
Administration. We formed 24 new matrices: 10 for airport (about 15 km north of Helsinki downtown)
weekdays, seven for Saturdays and seven for Sundays. and Helsinki}Isosaari (an island about 20 km south of
For a more detailed description, the reader is referred to Helsinki). The mixing height of the atmospheric bound-
ElolaK hde and Koskentalo (1996). ary layer was evaluated using the meteorological pre-
The tra$c volumes and average travel speeds of each processor, based on the sounding observations at
tra$c link were computed using EMME/2 transporta- Jokioinen (90 km northwest) and the routine meteoro-
tion planning system (INRO, 1994), which assigns the logical observations.
trips (from zone to zone) to links of a network model. The
model allows for the diurnal and daily variations both in
tra$c volumes and speeds, and tra$c emissions.

Table 1
A summary of the NO emissions in the Helsinki metropolitan
V
area in 1993

Source Emissions (as NO )


t a\ %

Public power generation 13 305 47.7

Other point sources and residential 855 3.0
heating systems
Road transport 12 090 43.3
Aviation 437 1.6
Marine tra$c and harbours 1223 4.4 Fig. 2. The location of the cities of Helsinki, Vantaa, Espoo
Total 27 910 100 and Kauniainen and the meteorological stations of Helsinki}
Vantaa, Helsinki}Isosaari and Jokioinen.
 Aea=2815=Durai=VVC=BG

3726 A. Karppinen et al. / Atmospheric Environment 34 (2000) 3723}3733

2.5. Road network dispersion model (CAR-FMI) substantial in episodic conditions, during prevailing
stable atmospheric strati"cation and low wind speed.
The dispersion from a road network is evaluated with The system of Equations (2a)}(c) can be solved analyti-
the Gaussian "nite-line source model CAR-FMI (Con- cally (Benson, 1984, 1992). Appendix A presents
taminants in the Air from a Road; HaK rkoK nen et al., 1995, a method for including the chemical transformation
1996). The model includes an emission model, a disper- module corresponding to these equations into a Gaus-
sion model and statistical analysis of the computed time sian line source dispersion model.
series of concentrations. The advantage of using such fairly simple chemical
The dispersion equation is based on an analytic solu- modules is that dispersion and chemistry computations
tion of the Gaussian di!usion equation for a "nite line are su$ciently e$cient numerically. For the number of
source (Luhar and Patil, 1989): sources and receptor grid points applied in this study, the
numerical requirements of more complex numerical



Q (z!H) chemistry schemes would be unreasonable, even for nu-
C" J exp !
2 (2p p u sin h 2p
X
merical program execution on a super computer.
X The predictions of the CAR-FMI model have been



(z#H) sin h (p!y)!x cos h previously compared with the results of two measure-
#exp ! erf ment campaigns, conducted near major roads (i) in a sub-
2p
X (2 p
W urban area in the city of Espoo in 1994 (Walden et al.,


sin h (p#y)#x cos h 1995; HaK rkoK nen et al., 1997) and in a rural area in
#erf , (1) southern Finland in 1995. For both of these measure-
(2 p
W ment campaigns, the predicted NO and NO concen-
V 
where C is the concentration, Q is the source strength trations agreed well with the experimental data in most
J cases considered; however, there were some speci"c me-
per unit length, u is the average wind speed, h is the angle
between the wind direction and the road, x, y and z are teorological and tra$c conditions, in which there was
the coordinates, H is the e!ective source height, p is the substantial disagreement.
half-length of the line source, erf is the error function and
p and p are the vertical and lateral dispersion para- 2.6. Urban dispersion modelling system (UDM-FMI)
X W
meters, respectively. The solution (1) allows for any wind
direction with respect to the road. The urban dispersion modelling system (Karppinen et
The dispersion parameters are modelled as function of al., 1998c, 1997b) includes a multiple source Gaussian
the Monin}Obukhov length, the friction velocity and the plume model and the meteorological pre-processor. The
mixing height (Gryning et al., 1987). These quantities are dispersion model is an integrated urban-scale model,
computed by the meteorological pre-processing model taking into account all source categories (point, line, area
(Karppinen et al., 1997b, 1998c). Tra$c-originated tur- and volume sources).
bulence is modelled with a semi-empirical treatment For the most general case involving volume sources, it
(Petersen, 1980). is assumed that these are evenly distributed within the
The model includes the basic reactions of nitrogen volume *x *y *z, i.e. +x, y, z " x 3 [x , x ]  y 3
 
oxides, oxygen and ozone: [y , y ]  z 3 [z , z ],. Assuming a total re#ection of
   
pollutants from the ground, the resulting ground-level
)0 NO#O, concentration can be written as (Karppinen et al., 1998c):
NO #hl P (2a)

Q
C(x, y, 0)"
O#O #MPO #M,
 
(2b) p *x *y *z u

 
)$ NO #O , W V X exp (!y/ 2p(x))
NO#O P (2c) ; W
   p (x)
W V X W
where M is a molecule and K and K are reaction rate exp(!H/2p(x))
$ 0 ; X dz dx dy, (3)
constants, which are functions of ambient temperature p(x)
and solar radiation intensity, as presented by Hertel and X
Berkowicz (1989). where Q is the total emission strength of the individual
The in#uence of hydrocarbons on the transformation sources. If the sources are of equal height, the volume
of nitrogen oxides is important in the regional and long- source reduces to an area source, and Eq. (3) can be
range transport scales. In the urban scale, their in#uence written in terms of error functions.
is less signi"cant, due to short transport times. Their As previously, the dispersion parameters are modelled
in#uence in Northern European urban areas may be as function of the Monin}Obukhov length, the friction
 Aea=2815=Durai=VVC=BG

A. Karppinen et al. / Atmospheric Environment 34 (2000) 3723}3733 3727

velocity and the mixing height (Hanna, 1985; Gryning et

al., 1987). These quantities are computed by the meteoro-
logical pre-processing model.
The transformation of NO}NO is described by the

simple relation (Janssen et al., 1988)

C(NO )/C(NO )"B (1!exp (!b x)), (4)

 V  
where C(i) is the concentration of the species. B and

b are parametrized in terms of local atmospheric condi-

tions, ozone concentration, wind speed and season of the
year (Janssen et al., 1988). Eq. (4) has been derived from
measurements over a period of ten years in the Nether-
lands.
The model also includes a treatment of dry and wet
deposition for nitrogen oxides and SO , plume rise, Fig. 3. A schematic presentation of the model for evaluating the

downwash phenomena, the dispersion of inert particles chemical interaction of pollutant plumes originating from vari-
and the in#uence of a "nite mixing height. The system ous urban sources.
computes an hourly time series of concentrations and
statistical parameters, which can be directly compared to
air quality guidelines. centrations for each month. This procedure produc-
The model predictions have been compared with the ed a matrix of 12 times 24 regional background
tracer experiments of Kincaid, Copenhagen and Lilles- concentration values, for each pollutant and each
troK m, presented by Olesen (1995b). The predictions were measurement station.
well in agreement with the Kincaid data; somewhat (ii) Secondly, we summed the spatial pollutant distrib-
larger di!erences were found for the Copenhagen and ution from all urban stationary sources to the
LillestroK m data. regional background. This simpli"cation is reas-
onable, as most of the emissions of stationary so-
2.7. Chemical interaction of pollutants originating from urces are released from higher altitudes. Their
various sources chemical transformation therefore mostly takes
place before substantial interaction with pollutant
Both dispersion models applied allow for the chemical plumes from other urban sources. This procedure
transformation of nitrogen oxides, by considering produces a so-called "rst-order spatial background
a single plume in a background air with a uniform distribution for each pollutant (NO, NO and O ).
 
constitution. However, the plumes originating from vari- (iii) Thirdly, the system evaluates the contribution on
ous sources interact chemically also with each other. For the background caused by the mobile sources. The
instance, it is not uncommon that urban background urban mobile sources and the receptor points are
O concentration vanishes, caused by the depletion of sorted out in terms of their location with respect to

O in the oxidation of NO into NO . However, many the wind direction. The plume from the most up-
 
regulatory modelling systems assume for simplicity that wind mobile source (a link of road or street) is
the urban background O concentration is equal to the allowed to interact chemically with the "rst-order

regional O background concentration. spatial background distribution. This produces

We have therefore developed a modelling system, a second-order spatial background distribution.
which allows for the chemical interdependence of the Correspondingly, the second most upwind mobile
NO concentrations originating from various sources source is allowed to interact with the second-order
V
and the O concentrations. Fig. 3 illustrates the main spatial background distribution. All the mobile

principles of the modelling system. sources are then treated consecutively.

(i) First, the modelling system evaluates the time vari- The system properly takes into account, for instance,
ation of the regional background concentrations. the depletion of O in the urban area. The in#uence of

The regional background concentrations are based the chemical interactions of the sources is largest during
on the data from the monitoring station of Luukki, the rush hours and air quality episodes, in which atmo-
situated in the North-Eastern part of the Helsinki spheric di!usion conditions are unfavourable. We have
metropolitan area. In order to "lter out temporally tested this assumption with numerical computations, us-
high episodic concentrations, originating from local ing the modelling system (i) including the above-men-
sources, we computed diurnal hourly average con- tioned modelling method, and (ii) excluding it from the
 Aea=2815=Durai=VVC=BG

3728 A. Karppinen et al. / Atmospheric Environment 34 (2000) 3723}3733

computations. For the Helsinki Metropolitan Area in

1993, the in#uence of the chemical plume interactions on
the hourly NO concentrations during rush hours ex-

ceeds 2 and 5% at 10 and 2% of the grid points, respec-
tively. For O concentrations, the relative in#uence

of the chemical plume interactions can be substantially
larger.

3. The numerical results

Helsinki metropolitan area is situated by the Baltic Sea

at the latitude of 603N. The climate is relatively milder
compared with many other areas in the same latitudes,
largely because the Gulf Stream and the prevailing global
atmospheric circulation have a warming e!ect. The Hel-
sinki metropolitan area comprises four cities: Helsinki,
Espoo, Vantaa and Kauniainen. The population of the
Helsinki metropolitan area is 850 000 and the area covers
743 km.
The dispersion models applied are Gaussian and these
do not generally take into account the in#uence of indi-
vidual buildings on the atmospheric dispersion (although
the UDM-FMI model allows for the in#uence of the
source building itself on the dispersion). However, the
terrain in the area is relatively #at and the average height
of the buildings is fairly low (most buildings are lower
than 15}20 m). There is only a moderate number of street
canyons in the area. We have used the roughness length
of 0.7 m in the numerical computations.
We have computed the concentrations of nitrogen Fig. 4. NO emissions from mobile (upper "gure) (lg m\ s\)
oxides (NO ) and nitrogen dioxide (NO ) in the Helsinki V
V  and stationary (lower "gure) (t a\) sources in the Helsinki
metropolitan area for one year, 1993. The concentration metropolitan area in 1993.
time series were computed on a receptor grid, which
contains approximately 10 000 receptor points. The re-
ceptor point network covers the whole area, and the
largest grid intervals are 500 m. A more densely spaced of the "gures shows the absolute values of the pollutant
grid was applied in the Helsinki downtown area, the grid emissions. In Fig. 4b, the height of each rectangle is
interval being 100 m. In the vicinity of the major roads in proportional to the stack height, and its width is logarith-
the area, the smallest grid interval was 50 m. The variable mically proportional to the emission rate. The dotted
receptor grid is required in order to evaluate isoconcen- lines indicate the city borders.
tration curves with adequate accuracy from the com- The emissions from mobile sources are mainly depen-
puted data. dent on tra$c densities and driving speeds. The emis-
The NO concentrations in Helsinki are generally sions of NO increase continuously with increase in
 V
comparable with those in the major Central European vehicle travel speed in the speed range from 40 to
cities (Jol and Kielland, 1997; Kukkonen et al., 1999). 120 km h\ (MaK kelaK et al., 1996), caused by the enhanced
In other Finnish cities the NO concentration levels oxidation of air-originated nitrogen with the increasing

are usually somewhat lower than those in the capital combustion temperature. The largest emission densities
(Kukkonen et al., 1999). occur in the Helsinki city centre area, and along Ring
Road 1, situated at a distance of 8}10 km from the city
3.1. The emissions centre. There are also substantial emissions along the
major roads leading to the Helsinki city centre, and at
Figs. 4a}b show the evaluated NO emissions of mo- major crossroads.
V
bile and major stationary sources in the Helsinki metro- All the largest stationary sources are coal-"red power
politan area in 1993. The size of the depicted area is plants, also using heavy fuel oil as an additive fuel. The
35 km;25 km and the legend in the top left-hand corner stack heights of the three largest power plants are about
 Aea=2815=Durai=VVC=BG

A. Karppinen et al. / Atmospheric Environment 34 (2000) 3723}3733 3729

150 m; the emissions of the largest individual unit

(`Hanasaari Ba) are approximately 6200 tonnes NO (as
V
NO ) annually. The contribution of minor residential

heating plants and industrial plants (not shown in Fig.
4b) on the total NO emissions is only a few per cent.
V

3.2. The annual average spatial concentration distributions

Figs. 5a}b show the computed annual means of NO

V
and NO concentrations at the ground level in the Hel-

sinki metropolitan area in 1993. The legend in the top
left-hand corner shows the absolute values of the pollu-
tant concentration.
Clearly, the tra$c emissions have a larger relative
in#uence on the ground-level concentrations, compared
to the stationary emissions, which are mostly released

Fig. 6. Predicted spatial distribution of the yearly mean concen-

tration of NO (lg m\) in the Helsinki downtown area in 1993.


from higher altitudes. Although the contribution of traf-

"c on the total emissions is slightly less than a half,
approximately 80}95% of the ground level NO concen-
V
trations originate from tra$c sources. The modest con-
tribution of stationary sources is caused by the strong
concentration of the NO emissions to large power
V
plants, which have substantial stack heights, within the
study area. The concentrations of NO and NO are the
V 
highest in the vicinity of the main roads and streets, and
in the downtown area of Helsinki. The "gures show the
distinct in#uence of the ring roads (situated at the distan-
ces of about 8 and 15 km from the city centre), the major
roads leading to the Helsinki city centre, and the junc-
tions of major roads and streets. The concentrations of
NO are strongly distributed along major roads, with

higher vehicle travel speeds, due to the increase of the
Fig. 5. Predicted spatial distribution of the yearly means of
NO emissions with the travel speed, and in the down-
NO (upper "gure) and NO (lower "gure) concentrations
V 
V
(lg m\) in the Helsinki metropolitan area in 1993. The location town area, caused by the largest tra$c volumes within
of the monitoring stations of YTV (ToK oK loK , Vallila, LeppaK vaara the area.
and Tikkurila) has also been indicated. The size of the depicted Fig. 6 illustrates the NO concentrations at the ground

area is 35 km;25 km. The lower "gure shows an outline of the level in the Helsinki downtown area in 1993. Again, the
Helsinki downtown area, presented in Fig. 6. in#uence of the main tra$c routes is clearly visible.
 Aea=2815=Durai=VVC=BG

3730 A. Karppinen et al. / Atmospheric Environment 34 (2000) 3723}3733

3.3. The chemical transformation caused by the computational procedure (with grid sizes
of 50}100 m).
The ratio of NO }NO is approximately 10% in traf-
 V
"c emissions, and somewhat smaller in power plant emis- 3.4. The spatial distribution of the statistical concentration
sions. Most of NO released from tra$c is oxidised to parameters
NO within a time scale of a few minutes (e.g., HaK rkoK nen

et al., 1996), while for stationary releases this time scale is Figs. 7a}b show the predicted distribution of statistical
of the order of 10}20 min (Janssen et al., 1988). Far from parameters for the highest hourly 99-percentile and the
the sources, an approximate photochemical equilibrium second highest daily mean NO concentrations in 1993.

is reached, resulting in an NO /NO ratio of about 90% These parameters can be compared to the corresponding
 V
in typical ambient conditions. hourly (150 lg m\) and daily (70 lg m\) national
The predicted concentration ratio NO /NO varies health-based air quality guidelines. The guideline values
 V
from approximately 50% in the vicinity of busy roads to for the daily NO concentrations were exceeded fairly

approximately 90% at the outer edges of the computa- extensively, in the vicinity of the main roads and in the
tional regime. As expected, the NO /NO concentration Helsinki downtown area. The corresponding guideline
 V
ratios are substantially higher than the corresponding values for the hourly NO concentrations were exceeded

ratios in the emissions, even close to the tra$c emission only at a few limited areas.
sources. This is caused by the in#uence of the more
`ageda air masses from urban background and station-
ary sources, and partly also by the numerical averaging 4. Conclusions

The modelling approach adopted has certain inherent

limitations, both concerning the evaluation of emissions
and atmospheric dispersion. Gaussian dispersion model-
ling does not allow for the detailed structure of buildings
and obstacles. However, the terrain in the area is #at and
the average height of the buildings is fairly low (most
buildings are lower than 15}20 m). The computed con-
centrations should be interpreted as spatially averaged
values (on the scale of the grid spacing, varying from 50
to 500 m), while for instance, inside a street canyon the
actual concentrations can vary substantially on the scale
of tens of meters.
On the other hand, the use of fairly simple dispersion
models facilitates the evaluation of the hourly time series
of meteorological and emission conditions for one year,
which is required for the computation of statistical con-
centration parameters, de"ned in national health-based
air quality guidelines. It is not at present possible to
conduct such an analysis for an agglomeration of cities
using, for instance, street canyon dispersion models or
computational #uid dynamics models.
It was also possible to include emissions from a large
number of sources (this study included 5000 line sources,
169 point sources and area sources), a substantial num-
ber of receptor points (10 000), and to use a su$ciently
dense computational grid. We also allowed for the essen-
tial chemical transformation processes in the computa-
tions. The central processing (CPU) time required for the
computations on the Cray C94 supercomputer was ap-
proximately 20 h.
Estimation of emissions also contains inherent limita-
Fig. 7. Predicted spatial distribution of the second highest daily tions. Near the junctions of roads and streets there is
mean NO concentration (lg m\) (upper "gure) and of the acceleration and deceleration of the tra$c #ow, as well as

highest 1-h 99-percentile NO concentration (lg m\) (lower stops and occasional congestion, which causes increased

"gure) in the Helsinki metropolitan area in 1993. emissions. The emission modelling takes properly into
 Aea=2815=Durai=VVC=BG

A. Karppinen et al. / Atmospheric Environment 34 (2000) 3723}3733 3731

account the in#uence of vehicle acceleration and deceler- environmental impact assessment of the di!erent trans-
ation on the emissions. The emissions, however, are as- portation system scenarios in the Helsinki Metropolitan
sumed to be distributed evenly along each line source in Area (HaK mekoski and Sihto, 1996).
the numerical computations, although these can be
strongly concentrated in the immediate vicinity of the
junctions. This e!ect can cause an underprediction of
Acknowledgements
tra$c emissions near major junctions.
The modelling system contains a novel method, which
This study is by the cooperation between the Finnish
allows for the chemical interaction of pollutants, origin-
Meteorological Institute (FMI), the Helsinki Metropoli-
ating from a large number of urban sources. The system
tan Area Council (YTV) and the Technical Research
properly takes into account, for instance, the depletion of
Centre of Finland (VTT). We would like to thank our
O in the urban area. This phenomenon can have a sub-
 coworkers in this study, Ms. PaK ivi Aarnio (Helsinki met-
stantial in#uence on the computed results particularly in
ropolitan Area Council) and Mr. Juhani Laurikko (VTT
episodic conditions, in which the atmospheric di!usion
Energy). We also wish to acknowledge Mr. Harri
conditions are unfavourable. However, the state-of-the-
Pietarila, Mr. Risto Pesonen, Mr. Jari HaK rkoK nen and Mr.
art modelling systems do not take these dependencies
Esko Valkonen (FMI) for their cooperation. The study
into account.
has been part of the national research programme `MO-
It can be shown numerically that although the contri-
BILE } Energy and the environment in transportationa.
bution of tra$c on the total emissions is slightly less than
The funding from the Technology Development Centre
a half, approximately 80}95% of the ground level NO
V (TEKES) is gratefully acknowledged.
concentrations originate from tra$c sources. The modest
contribution of stationary sources on ground level con-
centrations is caused by the strong concentration of the
NO emissions to large power plants within the study Appendix A. Inclusion of the chemical transformation
V module into a Gaussian line source dispersion model.
area.
The concentrations of NO and NO are strongly
V  The chemical transofrmation equations are included to
distributed in the vicinity of the main roads and streets,
and in their junctions. The NO emissions increase with the dispersion model based on a modi"ed version of the
V discrete parcel method (DPM); the original version has
the vehicle travel speed, which are substantially higher at
major roads, compared with minor roads and streets. been presented by Benson (1984, 1992). This method
The concentrations of NO and NO were also higher in considers air parcels, in which the emissions and back-
V  ground air are assumed to be instantaneously uniformly
the downtown area of Helsinki, caused by the large tra$c
volumes. The national air quality guidelines of the daily mixed. The chemical reactions in each parcel are then
NO concentrations were exceeded fairly extensively, in assumed to proceed independently of the dispersion pro-
 cess.
the vicinity of the main roads and in the Helsinki down-
town area. In this model, the initial mixing zone concentration
The annual average of the NO /NO concentration from a line source can be written as
 V
ratio varies from approximately 50% in the vicinity of
busy roads to approximately 90% at the outer edges of sQ
C "C # G . (A.1)
the computational regime. The latter limit corresponds G G u2H

approximately to photochemical equilibrium conditions.
The NO /NO ratio is substantially higher in the pre- where i refers to the species, C is the ambient back-
 V G
dicted concentrations, compared with the corresponding ground concentration, s is the mass fraction of the
G
ratio in the emissions, even near the sources. This is species i in the emissions, Q is the emission strength per
caused by the in#uence of the more &aged' air masses from unit length of the source (kg/(ms)), u is the wind velocity
urban background and stationary sources, and partly at the height H and 2H is the source height. The mass
 
also by the numerical averaging caused by the computa- fraction s is needed, for instance, to take into account
G
tional procedure. the fraction of NO in the total NO emissions.
 V
The modelling system developed has been an impor- The size of the reaction volume is therefore determined
tant assessment tool for the local environmental authori- by the height of the source and the wind velocity. The
ties. The system has been applied in order to evaluate the chemical reactions are then allowed to proceed during
compliance of air quality with the guidelines and limit the travel time, i.e., the time of transport from the source
values (together with the measured concentrations), the to the receptor.
in#uence of various emission categories on air quality, We suggest a revised version of the discrete parcel
and the representativity of the urban monitoring sta- method, in which the reaction volume is de"ned separ-
tions. The modelling system has also been used in the ately for each receptor location, instead of the source
 Aea=2815=Durai=VVC=BG

3732 A. Karppinen et al. / Atmospheric Environment 34 (2000) 3723}3733

location. This method is named receptor-oriented DPM ment. International Journal of Environment and Pollution 5,
(R-DPM); we call the original method source-oriented 382}400.
DPM (S-DPM). ElolaK hde, T., Koskentalo, T., 1996. Calculating emissions of
The relevant length scales of the reaction volume in the a transport network. Magazine of the Finnish Air Pollution
Prevention Society 5}6, 26}28.
R-DPM are determined by the length of the "nite line
Gryning, S.E., Holtslag, A.A.M., Irwin, J.S., Sievertsen, B., 1987.
source, and the lateral and vertical dispersion parameters
Applied dispersion modelling based on meteorological scal-
p and p . We de"ne the vertical length-scale simply as
W X ing parameters. Atmospheric Environment 21, 79}89.
H (x)H #kp , where k is a constant (selected here as
  X Hanna, S.R., 1985. Air quality modelling over short distances.
k"3) and x is the distance fro the source to the receptor. In: Houghton, D.D. (Ed.), Handbook of Applied Meteoro-
The horizontal length scale is de"ned as H (x)"2 logy. University of Wisconsin, Wisconsin (Chapter 25).

(p sin h#k p ), where p is the half-lengtht of the line HaK mekoski, K., Sihto, S., 1996. Transportation System 2020 and
W
source and h is the angle between the wind direction and Air quality in the Helsinki metropolitan area, Finland. In:
the road. Baldasano Recio, J.M., Sucharov, L.J. (Eds.), Urban Trans-
The initial concentration in the reaction volume in the port and the Environment II. Computational Mechanics
R-DPM is equal to the source mass #ux (2 ps Q) divided Publications, Southampton.
G HaK rkoK nen, J., Valkonen, E., Kukkonen, J., Rantakrans, E., Jal-
by the volume #ux perpendicular to the line source at the
kanen, L., Lahtinen, K., 1995. An operational dispersion
receptor location, H (x) H (c) (u sin h), i.e.,
  model for predicting pollution from a road. International
Journal of Environment and Pollution 5, 602}610.
sQ HaK rkoK nen, J., Valkonen, E., Kukkonen, J., Rantakrans, E.,
C (t)"C # G N . (A.2)
G G (H#kp )(p sin h#kp )(u sin h) Lahtinen, K., Karppinen, A., Jalkanen, L., 1996. A model for
 W the dispersion of pollution from a road network. Finnish
Meteorological Institute, Publications on Air Quality 23.
The e%uents are then assumed to be instantaneously Helsinki, 34 pp.
mixed into the reaction box, and the chemical reactions HaK rkoK nen, J., Walden, J., Kukkonen, J., 1997. Comparison of
are assumed to proceed during the travel time. model predictions and measurements near a major road in
In bothe versions of the model, the time available for an urban area. International Journal of Environment and
reactions is a function of the transport distance. How- Pollution 8 (3}6), 761}768.
Hertel, O., Berkowicz, R., 1989. Modelling NO concentrations
ever, for longer transport distances, more background air 
in a street canyon. DMU Luft-A131. National Environ-
is available for the reactions. This is allowed for in the
mental Research Institute, Roskilde, 31pp.
R-DPM, as the reaction volume is a function of the INRO, 1994. EMME/2 User's Manual. INRO Consultants Inc.,
transport distance, while in the S-DPM, the reaction box MontreH al, Canada.
volume is determined only by source properties. This is Janssen, L., van Wakeren, J., van Duuren, H., Elshout, A., 1988.
the main physical di!erence between the two models. A classi"cation of NO oxidation rates in power plant plumes
Bothe DPMs do not explicitly allow for the interaction based on atmospheric conditions. Atmospheric Environ-
of the chemical reactions and the physical mixing pro- ment 22, 43}53.
cesses, or the in#uence of the pollutant concentration Jol, A., Kielland, G. (Eds.), 1997. Air Pollution in Europe 1997.
pro"les within the plume on the chemical transforma- EEA Environmental Monograph No. 4. European Environ-
tion. In order to allow for these e!ects, a substantially ment Agency, Copenhagen, 107 pp.
more complex numerical transformation model would be Karppinen, A., Kukkonen, J., Konttinen, M., Rantakrans, E.,
Valkonen, E., HaK rkoK nen, J., Koskentalo, T., ElolaK hde, T.,
required (for instance, Kerminen and Wexler, 1996).
1997a. Comparison of dispersion model predictions and the
For a more detailed discussion of these models, the results from an urban air quality measurement network. In:
reader is referred to HaK rkoK nen et al. (1996). Power, H., Tirabassi, T., Brebbia, C.A. (Eds.), Air Pollution.
V. CMP, Southampton, pp. 405}411.
Karppinen, A., Jo!re, S., Vaajama, P., 1997b. Boundary
layer parametrization for Finnish regulatory dispersion
References models. International Journal of Environment and Pollution
8(3}6), 557}564.
Benson, P.E., 1984. CALINE4 } a dispersion model for predic- Karppinen, A., Kukkonen, J., Konttinen, M., HaK rkoK nen, J.,
ting air pollutant concentrations near roadways. Valkonen, E., Rantakrans, E., Koskentalo, T., ElolaK hde, T.,
FHWA/CA/TL-84/15, California Department of Transpor- 1998a. The emissions, dispersion and chemical transforma-
tation, Sacramento, CA. tion of tra$c-originated nitrogen oxides at the Helsinki
Benson, P.E., 1992. A review of the development and application metropolitan area. International Journal of Vehicle Design
of the Caline3 and 4 Models. Atmospheric Environment 20 (1}4), 131}136.
26B, 379}390. Karppinen, A., Kukkonen, J., Konttinen, M., HaK rkoK nen,
Carruthers, D.J., Edmunds, H.A., Ellis, K.L., McHugh, C.A., J., Valkonen, E., Koskentalo, T., ElolaK hde, T., 1998b. De-
1995. The atmospheric dispersion modelling system velopment and veri"cation of a modelling system for
(ADMS): comparisons with data from the Kincaid experi- predicting urban NO concentrations. In: Gryning, S.-E.,

 Aea=2815=Durai=VVC=BG

A. Karppinen et al. / Atmospheric Environment 34 (2000) 3723}3733 3733

Chaumerliac, N. (Eds.), Air Pollution Modelling and its Pietarila H., Pesonen R., Rasila, T., Rantakrans, E., 1997. Air
Application XXII, Vol 22. NATO, Challenges of Modern quality in the surroundings of the city of Turku (in Finnish).
Society, New York, pp. 567}574. Finnish Meteorological Institute, Air Quality, Report. Hel-
Karppinen, A., Kukkonen, J., Nordlund, G., Rantakrans, E., sinki, 81 pp.#A pp.
Valkama, I., 1998c. A dispersion modelling system for urban Pesonen, R., Rantakrans, E., Pietarila, H., Karppinen, A., 1996.
air pollution. Finnish Meteorological Institute, Publications Computations of the dispersion of nitrogen oxides in the
on Air Quality 28. Helsinki, 58 p. Helsinki metropolitan area, (in Finnish). Finnish Meteoro-
Kerminen, V.-M., Wexler, A.S., 1996. The occurrence of sulfuric logical Institute, Air Quality, Report. Helsinki, 61 pp.#A pp.
acid}water nucleation in plumes: urban environment. Tellus Petersen, W., 1980. User's guide for HIWAY2, a highway
B 48, 65}82. air pollution model. EPA-600/8-80-018, U.S. Environ-
Kukkonen, J., HaK rkoK nen, J., Valkonen, E., Karppinen, A., mental Protection Agency, Research Triangle Park, NC,
Rantakrans, E., 1997. Regulatory dispersion modelling in 69 pp.
Finland. International Journal of Environment and Pollu- Seinfeld, J.H., Pandis, S.N., 1998. Atmospheric Chemistry and
tion 8 (3}6), 782}788. Physics. ISBN 0-471-17815-2. Wiley, New York, USA,
Kukkonen, J., Salmi, T., Saari, H., Konttinen, M., KartastenpaK aK , pp. 945}947.
R., 1999. Review of urban air quality in Finland. Boreal Valkonen, E., HaK rkoK nen, J., Kukkonen, J., Rantakrans, E.,
Environment Research 4 (1), 55}65. Jalkanen, L., Haarala, S., 1995. Application of dispersion
Luhar, A.K., Patil, R.S., 1989. A General "nite line source model models for evaluating the in#uence of urban air pollution on
for vehicular pollution prediction. Atmospheric Environ- human health. International Journal of Environment and
ment 23, 555}562. Pollution 5, 557}566.
Moussiopoulos, N., Berge, E., Bohler, T., de Leeuw, F., Gron- Valkonen, E., HaK rkoK nen, J., Kukkonen, J., Rantakrans, E.,
skei, K.-E., Mylona, S., Tombrou, M., 1996. Ambient air Jalkanen, L., Haarala, S., 1996. Modelling urban air pollu-
quality, pollutant dispersion and transport models. Euro- tion in Espoo, Finland. The Science of the Total Environ-
pean Environment Agency, Topic Report 19, Air Quality. ment 189/190, 205}211.
Copenhagen, 94 pp. van Ulden, A., Holtslag, A., 1985. Estimation of atmospheric
MaK kelaK , K., Kanner, H., Laurikko, J., 1996. Road tra$c exhaust boundary layer parameters for di!usion applications.
gas emissions in Finland } LIISA 95 calculation software. (in Journal of Climate and Applied Meteorology 24,
Finnish). VTT Communities and Infrastructure, Transport 1196}1207.
Research, Research Notes 1772, Technical Research Center Walden, J., HaK rkoK nen, J., Pohjola, V., Kukkonen, J., Kartasten-
of Finland, Espoo, 45 pp.# App. 51 pp. paK aK , R., 1995. Vertical concentration pro"les in urban condi-
Nikmo, J., Tuovinen, J.-P., Kukkonen, J., Valkama, I., 1999. tions } comparison of measurements and model predictions.
A hybrid plume model for local-scale atmospheric disper- In: Anttila, P., et al. (Ed.), Proceedings of the 10th World
sion. Atmospheric Environment 33, 4389}4399. Clean Air Congress, Espoo, Finland, May 28}June 2, 1995,
Olesen, H.R., 1995a. Regulatory dispersion modelling in De- Vol. 2. The Finnish Air Pollution Prevention Society,
nmark. International Journal of Environment and Pollution Helsinki, pp. 268 (4 p.).
5, 412}417. Yamartino, R.J., Scire, J.S., Carmichael, G.R., Chang, Y.S., 1992.
Olesen, H.R., 1995b. Datasets and protocol for model valida- The CALGRID mesoscale photochemical grid model } I.
tion. International Journal of Environment and Pollution 5, Model formulation. Atmospheric Environment A (26), (8),
693}701. 1493}1512.

View publication stats