Development of Integrated Air Pollution Modelling Systems for Urban Planning- DIANA project

By Jarkko Niittymaki, Lic.Tech., research group leader, Helsinki University of Technology, Laboratory of Transportation, P.O.Box 2100, FIN-02015, HUT, Finland. Email. [email protected]

Mette Granberg, graduate student, Helsinki University of Technology, Laboratory of Transportation, P.O.Box 2100, FIN-02015, HUT, Finland. Email. [email protected]

Ari Karppinen, Lic. Tech, Senior Research Scientist, Finnish Meteorological Institute Sahaajankatu 20 E, 00810 Helsinki Email. [email protected]

Jarkko Niittymki, Mette Granberg, Ari Karppinen ABSTRACT

This study was done as a part of DIANA (Development of integrated air pollution modelling systems for urban planning) program. The aim was to create a model that would be accurate in modelling urban air quality and its changes depending on the traffic situation. The method was integrating two modelling systems: a microscopic simulation model HUTSIM with an OSPM (Operation Street Pollution Model) model. HUTSIM is a wellcalibrated microscopic simulation model developed at Helsinki University of Technology. HUTSIM has a stochastic nature and its main principles are object-oriented modelling, time-scanning updating, graphical userinterface and rule-based dynamics. OSPM, developed by Hertel and Berkowicz, is a practical street pollution model based on simplified description of flow and dispersion conditions in street canyons. The model achieved was tested with a case study on a busy downtown street in Helsinki. The overall results were promising; the modelled values of NOx, NO2 and CO were equivalent to the measured ones. Traffic is one the main cause of pollution in urban areas. Thus the growing amount of vehicles in cities increases congestion and contributes to deterioration of air quality. Applications such as presented in this paper could be useful tools in aiming towards sustainable development: a challenge that is posed by increasing traffic especially in urban areas. KEYWORDS: emission, urban environment, air quality INTRODUCTION One of the main concerns in the growth of traffic is the state of the environment, which is deteriorating especially in cities. The ever-growing traffic is a highly topical problem in both industrialised and developing countries; declining air quality decreases the quality of life in cities, and exposes people to different kinds of respiratory diseases. The knowledge of the cause and behaviour of emissions helps to evaluate different control strategies in preventing and minimising environmental strain in cities, where the main cause of pollution is traffic. In modelling emissions it is therefore important to understand the impact of vehicle movements and delays in generating accurate urban emission dispersion models. Furthermore, the opportunity of being able to correctly model emissions and the dispersion of air pollution is a means for a greater understanding of the characteristics of urban environment. The aim of DIANA-project (Development of integrated air pollution modelling systems for urban planning) was to create a comprehensive modelling system based on microscopic simulation that could be used in evaluating traffic volumes, emissions, atmospheric dispersion and noise in cities. This paper deals with the

Jarkko Niittymki, Mette Granberg, Ari Karppinen

integration of the traffic microsimulation model HUTSIM and its emission calculation sub-program EMCA, developed and used by Helsinki University of Technology, with an atmospheric dispersion model OSPM (Operational Street Pollution Model), used by Finnish Meteorological Institute. The method and its possible future applications are presented in this paper, along with a case study on a busy downtown street in Helsinki. THE ROLE OF TRAFFIC IN URBAN ENVIRONMENTAL PROBLEMS Traffic is by far one of the most important factors influencing the quality of urban life and environment. The rapid increase of traffic in both industrialised and developing countries has lead to a number of problems, of which decreased traffic safety and growing environmental problems are perhaps the most serious ones. In addition to the growth of traffic, the amount of people living in cities is also growing. In Finland over 60% of the population live in urban areas, and inside EU the percentage is even greater. (1) Land use and transport policies have caused major towns to develop into large extended regions consisting of smaller conurbations and centres. Over the past few decades, the changes in demography and travelling patterns in cities have lead to local and regional economies, increased congestion, deterioration of urban air quality and decrease in the quality of life. Inside EU nearly two-thirds of emissions from pollutants such as carbon monoxide (CO) and nitrogen oxides (NOx) result from road vehicles. (2) Traffic has a few special characteristics as a pollutant. First, the emissions of traffic are caused at street level, which leaves them to easily circulate into respiratory organs. Second, emission concentrations can vary substantially depending on the area and time. Vehicle journeys during rush hours pose particular problems in terms of environmental pollution, and therefore scientific knowledge of their impact on the environment is needed (2). Vehicle driving patterns also influence the amount of pollution caused. Accelerations and decelerations cause more pollution than smooth driving. As an example of this, Figure 1 illustrates the effect of speed and acceleration on the amount of CO caused.

Jarkko Niittymki, Mette Granberg, Ari Karppinen

FIGURE 1

Principal image of CO-emissions as a function of speed and acceleration (3).

Also vehicle type has an effect on emission quantities; heavy vehicles such as buses and lorries release more emissions than personal cars. INTRODUCTION TO THE STUDY

DIANA-project This study was done as a part of DIANA (Development of integrated air pollution modelling systems for urban planning) project. The different sub-models of the program are shown in Figure 2, the bolded arrows show the areas this paper concentrates on.

Graphical User Interface Visualisation of the simulation results

Calculation of emissions from stationary sources

Macro Simulation of Traffic (EMME/2)

Microsimulation of traffic (HUTSIM) Dispersion of pollutants (OSPM)

Numerical weather prediction model (HIRLAM) Dispersion of noise (CITYSIM noise model)

Calculation of emissions from traffic (EMCA)

FIGURE 2

The sub-models of DIANA.

Jarkko Niittymki, Mette Granberg, Ari Karppinen The organisation behind the study was Finnish Meteorological Institute (FMI). FMI accounted also

for the expertise of air quality modelling and management while Helsinki University of Technology (HUT) was in charge of traffic simulation and planning. Simulating traffic Road traffic simulation was introduced in the early days of computer aided calculation of traffic demand and flows on a network system. Computer simulation has become a widely used tool in transportation engineering with a variety of applications from scientific research to planning, training and demonstration. Traffic systems are an excellent application environment for simulation-based research and planning techniques. Traffic systems are also an application area where the use of analytical tools, though very important, is limited to the level of subsystem and sub-problem. The reasons to use simulation in the field of traffic are the same as in all simulation; the problems in analytical solving of the question at hand, the need to test, evaluate and demonstrate a proposed course of action before implementation, to do research (to learn), and to train people. (4) Traditional traffic simulation has been macroscopic. However, the use of macroscopic simulation requires such generalisations in the traffic environment that, it can sometimes be an inflexible and inappropriate way of modelling traffic. In macroscopic simulation, for example, vehicles are treated as traffic flows, and individual vehicles are not separated from it. In microscopic simulation of traffic, on the other hand, each vehicle has its own characteristics. This brings in the real dynamics of the traffic phenomenon, and thus increases the sensitivity of the simulation. The planner and researcher can also follow the traffic phenomenon as it develops as a function of time. Visual presentation and animation of real-time (or faster) action gives the viewer a comprehensive view that helps him or her in interpreting statistical printouts depicting the traffic situation and performance. (5) HUTSIM and EMCA HUTSIM is a traffic simulation program developed at Helsinki University of Technology. HUTSIM is a wellcalibrated microscopic simulation model with stochastic nature. Its main principles are object-oriented modelling, rule-based dynamics, time-scanning updating and graphical user-interface. HUTSIM was designed especially for simulating traffic signal control, and a real signal controller is used as a part of the simulation system. HUTSIM has been developed since 1989 and its enhancement continues (6). EMCA (EMission CAlculation) is a sub-program of HUTSIM, and it was developed to help in environmental analysis of HUTSIM runs. With HUTSIM, it is possible to produce different types of simulation

Jarkko Niittymki, Mette Granberg, Ari Karppinen

output files. The output files contain detailed information on vehicle movements, delays, signal control and other such things. To calculate the emissions of a certain traffic situation the export file of this situation is run through EMCA. (7) (8) Vehicle speeds are updated and stored in output files every 0,1 second. One basic speed step in HUTSIM is 2,5 km/h. Because of this stepwise changing of vehicle speeds the actual speeds, and especially accelerations and decelerations, cannot be seen from the speed-distance curve made directly from HUTSIMs output file. For example, when decelerating from 50 km/h to 0 km/h the deceleration of a vehicle seems to be zero for a while, then almost infinite for a very short while (from 50 km/h to 47,5 km/h in a moment), then zero again and then almost infinite, and so on until after twenty loops the speed is 0 km/h. Vehicle speed is described as a step function of distance, which is discontinuous. In order to form continuous speed and acceleration functions the distance-time curve has to be rounded. In EMCA the rounding is based on cubic spline fit. This ensures that the speed-distance-curve becomes continuous and differentiable. The cubic spline fit is based on polynomial functions in which the curves flow is based on the y-values of the starting (distancei-1) and ending points (distancei) and first derivatives (velocities) of both points. The first and second derivatives of this kind of a function are continuous. (5) The distance-speed table is divided into pieces for the cubic spline fit. The starting point of the piece is placed in the middle of the line segment before the break point, and the endpoint is placed in the middle of the next line segment. New distance values are adapted according to formula 1.

distance new = r00distance i1 + r01distance i + r10 speedi1 + r11speedi r00 = 1 - 3 S2 + 2 S3 r01 = 3 S 2 2 S 3 r10 = (t i t i1 ) (S 3 S2 + 2 S3 ) r11 = (t i t i 1 ) (-S2 + 2 S3 )
where S is an auxiliary variable that gets values evenly from 0 to 1 in the time period of

(1)

S=

time time i1 one piece time i time i1

Vehicle speeds and accelerations are then calculated by differentiation using formulas 2 and 3. (5)

Jarkko Niittymki, Mette Granberg, Ari Karppinen

speed =

distance i +1 distance i1 (2) time i +1 time i1 speedi+1 speedi1 accelerati on = (3) time i+1 time i1 time i+1 time i1 = 2 time division duration = 2 10 0,01 sec = 0,2 sec

The emission calculation of EMCA is based on emission matrices. The formation of emissions correlates with factors describing the operating conditions of a vehicle. These factors are based on the type, speed, acceleration and deceleration of the vehicle. In EMCA emission matrices describe the quantities of emissions as functions of vehicle speed, and acceleration or deceleration. EMCA includes an emission matrix for personal cars and vans (length < 8m), and another one for heavy vehicles such as buses and lorries. In addition to emissions, EMCA also counts the fuel consumption of vehicles. The fuel consumption and emissions are given per individual vehicle, per vehicle kilometre and vehicle hour, and per the whole area modelled. HUTSIM itself creates a report file giving certain figures depicting the traffic situation of the area. Such figures are e.g. average delay, percentage of stops, and average queue and distance driven. (9) Modelling flow and dispersion in a street canyon (OSPM) (10) In modelling flow and dispersion in a street canyon FMI uses an OSPM model, developed by Hertel and Berkowicz (9), (10), (11). The OSPM model is a practical street pollution model based on a simplified description of flow and dispersion conditions in street canyons. Concentrations of exhaust gases are computed using a combination of a plume model for the direct contribution from street traffic, and a box model for the recirculating part of pollutants in the street. The simplified parameterisation of the flow and dispersion conditions in a street canyon has been deduced from extensive analysis of experimental data and model tests. The emission field is treated as an area source at street level. The wind direction at street level is assumed to be mirror-reflected with respect to that of the roof level wind, and attenuated according to a simple logarithmic profile with a roughness length z0 = 0.6 m. The transport wind is calculated at the average height of the vehicles, h0 = 2 m. Due to the presence of the flanking buildings, crosswind diffusion is neglected. The hourly mean concentrations are calculated assuming wind meandering with an angle increasing with decreasing wind speed. The general expression for the direct contribution from street traffic is given by

Cd =

2 Q ln z W w ho

(4)

Jarkko Niittymki, Mette Granberg, Ari Karppinen

here Q is the emission density (per street length and per time), W is the street width and z is the vertical dispersion parameter at the receptor point. The vertical dispersion parameter is modelled assuming that dispersion of the plume is governed solely by mechanical turbulence. Turbulence due to thermal stratification has been neglected, as its influence is usually negligible at street level. The mechanical turbulence is assumed to be generated by the wind and by the traffic in the street.
w = ((u b ) 2 + 2 wo )

(5)

where = 0.1, ub is the street level wind speed and wo is the traffic-induced turbulence. Traffic-induced turbulence plays an important role in the dispersion of pollutants in a street, particularly in low wind-speed conditions. The contribution from the recirculation part is computed using a simple box model. It is assumed that the canyon vortex has the shape of a trapezium, with the maximum length of the upper edge being half the vortex length. The ventilation of the recirculation zone takes place through the edges of the trapeze, but the ventilation can be limited by the presence of a downwind building, if the building intercepts one of the edges (12). The model can also be applied for dispersion in streets with spread-out buildings or buildings on one side only. Considering the chemical transformation in a street canyon, only the fastest chemical reactions can have any significance. For nitrogen oxides, it is therefore sufficient to include only the basic reactions involving NO, NO2, O2 and O3. The model includes the three basic reactions between these compounds:
R NO2 + h K NO + O

O + O2 + M O3 + M
KF NO + O3 NO2 + O2 ,

where M is a molecule and KF and KR are reaction rate constants, which are functions of ambient temperature and solar radiation intensity, as presented by Hertel and Berkowicz (10), (11). However, reaction (2b) is so fast that for all practical purposes (2a) and (2b) can be treated as one reaction. This is also the procedure applied in the OSPM model. (9) For the calculation of NO2 formation in the street, the urban background concentrations of nitrogen oxides and ozone must be given as input. Temperature and total solar radiation are also needed as input values

Jarkko Niittymki, Mette Granberg, Ari Karppinen

for the model in order to compute the chemical reaction coefficients. The urban background concentrations must also be given for the other pollutants (e.g., CO, benzene) calculated by the OSPM model. These are often available from measurements, but an urban background model can also be used to provide these input parameters. (9)
CASE STUDY RUNEBERGS STREET

Measurement campaign in central Helsinki

In 1997 a measuring campaign was conducted in a street canyon (Runeberginkatu; Runebergs St.) in Helsinki. Hourly concentrations of CO, NOX, NO2 and O3 were measured at street and roof level, the latter in order to determine the urban background concentrations. Relevant hourly meteorological parameters were measured at roof level; these included wind speed and direction, temperature and solar radiation. Hourly street level measurements and on-site electronic traffic counts were conducted throughout the whole of 1997; roof level measurements were conducted for approximately two months, from the 3rd of March to the 30th of April 1997. The structure of the project is illustrated in Figure 3.

Measurement campaign in 1997 actual emission concentrations at street and roof level COMPARISON FMI: Calculation of emission quantities based on emission factors dependent on vehicle type and average driving speed HUT: Areal microsimulation of traffic/HUTSIM emission quantities Modelling emission concentrations based on emission quantities acquired by calculations and microscale traffic simulation/OSPM
FIGURE 3 Structure of the project.

The measurement site was in a street segment that had almost uniform building structures over a distance of 175 metres (191.5 yd). The street level measurements were done from a sample intake collected at the height of 4.0 metres (4.4 yd), above the middle of the pavement. Figure 4 shows a cross-section of the streetcanyon.

COMPARISON

Jarkko Niittymki, Mette Granberg, Ari Karppinen

10

Urban background 2.5 m and meteorological measurements 6.0 m

3.4 m

23.0 m

Street level measurements

2.9 m
2.0 m 24.0 m
FIGURE 4 Cross-section of the measured street-canyon (9).

4.0 m

Background has a minor influence for predicted CO and NOx concentrations; it has a substantial influence only for NO2. (see Figures 5-6). This issue has been already quite thoroughly discussed in the reference (10).
1.8

Measured daily CO-concentrations (mg/m3)

1.6

Street level Roof level

1.4

1.2

0.8

0.6

0.4

0.2

03.03.

15.03.

31.03.

15.04.

30.04.

Time (March-April)

FIGURE 5
400

Measured daily CO-concentrations (10).

Measured daily NOX-concentrations

350 300 250

Street level Roof level

(g/m )
3

200 150 100 50 0

03.03.

15.03.

31.03.

15.04.

30.04.

Time (March-April)

FIGURE 6

Measured daily NOx concentrations (10).

Jarkko Niittymki, Mette Granberg, Ari Karppinen

11

For NO2, the background concentrations have a substantial effect and a possibly overestimated background may have influence in model overprediction. The measured NO2-concentrations are shown in Figure 7.
80

Measured daily NO2-concentrations (g/m )

70

Street level Roof level

60

50

40

30

20

10

03.03.

15.03.

31.03.

15.04.

30.04.

Time (March-April)

FIGURE 7

Measured daily NO2-concentrations (10).

The levels of NO2 concentrations in a street with heavy traffic are mainly determined by the efficiency of NO oxidation caused by ozone. Directly-emitted NO2 constitutes only a minor fraction (approximately 5 %) of the nitrogen oxides emitted by traffic. This leads to a strongly nonlinear relationship between the NOx and NO2 concentrations in street canyon air. The meteorological conditions have only minor importance regarding the NO2 formation processes in a street canyon. The qualitative behaviour of the corresponding predicted results is very similar to that of the observed ones, although the OSPM model overestimates the NO2/NOx ratio compared with the measurements. The higher predicted NO2/NOx ratio could be caused by an overestimation of the background concentrations of NO2 or O3. The background concentrations used in the model are given by the measurements from the roof level station. The possibility cannot therefore be dismissed that the vehicular emissions in the street could have somewhat affected the measured roof level background concentrations. The strongest influence of street level emissions on the roof level concentrations is expected when the roof station is downwind of the street; this corresponds to westerly winds. This was actually the dominant wind direction during the intensive measurement campaign. We performed a more detailed analysis of the data in this respect; however, this investigation did not reveal any systematic differences in the NO2/NOx behaviour depending on the wind direction. When analysing the NO2 data, one also has to allow for the potential uncertainties concerning the measured concentrations. Due to the measuring technique, the NO2 concentrations are computed from the difference between the measured NOx and NO concentrations. For example, under circumstances in which NOx concentrations are higher than 400 g/m3, the values of the NO2/NOx ratio are less than about 20 %.

Jarkko Niittymki, Mette Granberg, Ari Karppinen

12

The measuring uncertainties in NO and NOx concentrations can therefore result in substantial uncertainties of the measured NO2 concentrations. This could be the reason for the difference between the observed and modelled NO2/NOx ratios.

The modelled area and time period

The modelled street (Runebergs street; Runeberginkatu) is a busy street in central Helsinki. It has four lanes, and also tram rails. The trams were disregarded in the emission calculations of this study. Figure 8 shows the modelled area within the HUTSIM simulations.
0 100 200 300 400 Traffic signals
tu ianka esper inen H Pohjo katu erian Hesp inen Etel
Ca lon iuk sen

ka tu
Museok atu

tu i nka berg R une

To
Tyki stka tu

at u nk e s iuk pel

Sandelsinkatu

Measurement location

inkatu Dbeln

u kat ln T

FIGURE 8

The test area of Runeberg's street and its representation within the HUTSIM model (9).

Jarkko Niittymki, Mette Granberg, Ari Karppinen

13

The modelled area consisted of various intersections with altogether five traffic signals. Although the measurements were conducted at one point, the modelled area was chosen so that it would describe the premises of the measurement site realistically. Altogether three days and nights were modelled with HUTSIM. The first day modelled was chosen to be Wednesday, the 12th of March 1997. The day was chosen because it was an ordinary working day without any unusual characteristics. The other two days were the Saturday and Sunday, 15th and 16th of March, the weekend of the same week. The days were modelled an hour at a time. Vehicle flows varied from 65 to 2060 vehicles per hour depending on the time of day. The percentage of heavy vehicles varied from 0 to 24%, but was under 10% in most of the cases.
Results

In the first phase the macroscopic (FMI) and microscopic emission models (HUTEMCA) was compared. FMI acquired the hourly emissions of the street segment using emission factors evaluated by the Helsinki Metropolitan Area Council (YTV). The emission factors varied depending on average driving speed and vehicle type. The modelled emissions of both HUT and FMI correlated well with each other. As an example, the modelled NOx emissions with HUTSIM and with the FMI emission model are shown in Figure 9.

N O x e m is s io n s - c o m p a r is o n ; w o r k in g d a y
2500

2000

NOx (g/km/h)

1500
H U T lig h t F M I lig h t

1000

500

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

hour

FIGURE 9

Modelled vehicular emission values of NOx ; HUTSIM compared with the emissions calculated by FMI.

Jarkko Niittymki, Mette Granberg, Ari Karppinen

14

Also the modelled and measured values correlated well with each other. Figures 10 and 11 show the comparisons of the modelled values of NO2 and CO concentrations by HUTSIM-OSPM (HUTSIM output was used as input data for the OSPM-model) and the measurements. The figure shows the results of one week. The modelled Wednesday was used to depict an ordinary working day, and only the meteorological parameters were changed for other working days.
Runeberg Street; 10.-16.3.1997
120 100 80 60 40 20 0
0 6 12 18 0 6 12 18 0 6 12 18 0 6 12 18 0 6 12 18 0 6 12 18 0 6 12 18

HUTSIM-OSPM NO2 measured NO2

NO2 (g/m )

hour

FIGURE 10

Comparisons between the (HUTSIM-OSPM) modelled and measured concentration values of NO2 during the week, 10.-16.3. 1997.
Runeberg Street; 10.-16.3.1997

3.5 3 2.5 CO (mg/m )

HUTSIM-OSPM CO measured CO

2 1.5 1 0.5 0
6 12 18 6 12 18 6 12 18 6 12 18 6 12 18 6 12 18

hour

FIGURE 11

Comparisons between the (HUTSIM-OSPM) modelled and measured concentration

values of CO during the week, 10. -16.3. 1997.

It is remarkable how well the modelled values of emissions follow the measured values. The only day in which the modelled values differ significantly from the measured ones is Friday. This is probably due to the meteorological situation of the Friday in question and different traffic distribution during the rush-hour. When the wind speed is very small or it is calm the OSPM-model tends to overestimate the concentration values.

6 12 18

Jarkko Niittymki, Mette Granberg, Ari Karppinen

15

Figures 12a-c show the correlation of the modelled and measured values of NO2(r2=0.88), NOx(r2=0.66) and CO (r2=0.48).

Comparison betw een measured and modelled hourly NO 2 concentrations for 10.3.-16.3.1997
Modelled NO 2 (g/m3) 120 100 80 60 40 20 0 0 20 40 60 80 100 120 Me a sure d NO 2 (g/m 3)

Series1 x =y

Comparison between measured and modelled hourly NOx concentrations for 10.3.-16.3.1997
M od ell ed N Ox ( g/ m 3) 600 500 400 300 200 100 0 0 100 200 300 400 500 600 Measured NOx (g/m3)

Series1 x=y

Comparison between measured and modelled hourly CO concentrations for 10.3.-16.3.1997

Series1 x =y

FIGURE 12

Correlation between measured and modelled values of NO2, NOx and CO.

Jarkko Niittymki, Mette Granberg, Ari Karppinen

16

The measured and modelled results correlate quite well with each other. The over-estimations of the Friday reduce the correlation between the modelled and measured values, but in spite of this a clear correlation is still visible. Work is still needed in developing the method, but the results of this study have been most encouraging: urban air quality prediction has proven its possibilities.
CONCLUSIONS

The integration of a detailed microscopic traffic simulation model with an accurate meteorological model gives the opportunity of combining two areas of which both are crucial to the quality of urban environment: predicting the behaviour of traffic and circulation of emissions. The modelled values correlated mainly extremely well with the measured ones. The few over-estimations of the Friday still need to be studied carefully, and it is important to find the reasons for them. The aim is to yet improve the correlation between the modelled and measured values of emissions. This intent in mind, another case study will be conducted during the autumn of 2000. Keeping in mind that the results presented here are only the very first ones, it is safe to say that although small modifications and enhancements are still needed, according to the preliminary results accurate modelling of urban air quality is not only possible but an effective instrument in understanding the cause-consequence relationship of traffic control. Air quality is deteriorating especially in cities, where most of the emissions are caused by traffic. Urban air quality prediction methods could be useful tools for example in estimating peoples exposure to pollution and predicting environmental strain caused by increasing traffic. As HUTSIM is especially designed accurate behaviour of traffic as the prime concern, an emission model such as presented here could also be extremely useful in evaluating different traffic control and management strategies. Such strategies could be e.g. adding a separate bus lane, decreasing the amount of heavy traffic, or simply minimising delays and stops in traffic signals. Naturally accurate modelling itself will not solve the problem of deteriorating urban air quality, measures are still needed to constrain the growth of traffic. However, in finding the best environmental solutions, applications such as presented in this paper provide a concrete tool for city and traffic planners all over the world.
ACKNOWLEDGEMENTS

This study was done as a part of DIANA (Development of integrated air pollution modelling systems for urban planning) project. DIANA was conducted in 1997-1999 with the funding of Ministry of Trade and Industry in

Jarkko Niittymki, Mette Granberg, Ari Karppinen

17

Finland. DIANA belonged to MOBILE2-program. The program focused on energy use and environmental impacts of transportation, and the aim was to find means for reducing energy consumption and emissions of both passenger and goods transport in conditions prevalent in Finland.
REFERENCES

1.

Karppinen A., Kukkonen J., Hrknen J., Pohjola M., Kousa A., Aarnio P., Koskentalo T., Niittymki J. and Granberg M., Kaupunkisuunnittelun tarpeisiin sopivien mallijrjestelmien kehittminen ja sovellukset (The development and applications of modelling systems suitable for the needs of urban planning). MOBILE2-report M2K9912-1.1999.

2.

Niittymki, J., Karppinen, A., Kukkonen, J., Ilvessalo, P. and Bjrk, E. CITYSIM - Validated Assessment Tool for Simulating Urban Traffic and Environmental Impacts. In: Sucharov L J (Ed.). Urban Transport and the Environment for the 21st Century V. Fifth International Conference on Urban Transport and the Environment for the 21st Century, Rhodes, Greece, 8-10 September 1999, WIT Press. Southampton, UK, , 1999. pp. 393-404.

3.

Sturm P.J., Hausberger S., Keller M. and de Haan P. Estimating Real World Emissions for Passenger Cars Use and Limitations of Instantenous Emission Data. Paper accepted for publication in International Journal on Vehicle Design. Graz, Austria, 1999. 18 p.

4.

Pursula M. Simulation of Traffic Systems - An Overview. Journal of Geographic Information and Decision Analysis. V ol. 3, No. 1, 1999. pp. 1-8. http://publish.uwo.ca/~jmalczew/gida_5/Pursula/Pursula.html (Accessed June 29, 1999).

5.

Lehmuskoski, V., Niittymki, J. and Silfverberg, B. Microscopic Simulation on high-class Roads: Enhancement of Environmental Analyses and Driving Dynamics, Practical Applications. Preprint CDROM, Paper No. 00550. Transportation Research Board, 79th Annual Meeting, Washington D.C. January 9-13, 2000. 14 pp.

6.

Kosonen I. HUTSIM - Urban Traffic Simulation and Control Model: Principles and Applications. Helsinki University of Technology, Transportation Engineering, Publication 100, Doctoral Thesis. 1999. 249 p.

7.

Sane, K.I., Kosonen I. HUTSIM 4.2 Reference Manual. Helsinki University of Technology, Laboratory of Transportation. 1996. 132p.

8.

Brummer, A. EmCa- EMission CAlculation program for Hutsim. Helsinki University of Technology. A Special Assignment for Transportation Engineering. 1995. 83 p..

Jarkko Niittymki, Mette Granberg, Ari Karppinen

18

9.

Granberg, M., Niittymki, J., Karppinen, A., Kukkonen J., Combined application of traffic microsimulation and street canyon dispersion models, and evaluation of the modelling system against measured data. In Sucharov L.J., Brebbia C.A. (Eds.). Urban Transport and the Environment for the 21st Century, VI International Conference on Urban Transport and the Environment for the 21st Century, Cambridge, UK, 26-28 July 2000. WIT Press, Southampton, UK. 2000. pp. 349-358.

10. Kukkonen, Jaakko, Esko Valkonen, Jari Walden, Tarja Koskentalo, Pivi Aarnio, Ari Karppinen, Ruwim Berkowicz and Raimo Kartastenp. A measurement campaign in a street canyon in Helsinki and comparison of results with predictions of the OSPM model. Atmos. Environ. 35. 2000. pp 231-243. 11. Hertel, O., Berkowicz, R., Modelling NO2 Concentrations in a Street Canyon. DMU Luft A-131. National Environmental Research Institute, Roskilde. 1989. 31p. 12. Berkowicz, R., Modelling street canyon pollution: model requirements and expectations. Int. J. Environment and Pollution, Vol. 8, Nos. 3-6, 1997. pp 609-619. 13. Hertel, O., Berkowicz, R., Operational Street Pollution Model (OSPM). Evaluation of the Model on Data from St. Olavs Street in Oslo. DMU Luft-A135. National Environmental Research Institute, Roskilde. 1989. 34 p.