1. A numerical simulation of wind field and air quality above an industrial center Alexander Starchenko Tomsk State University, Tomsk, Russia starch@math.tsu.ru

2. A wide range of problems of atmospheric physics, climate, and environmental protection is solved nowadays using mathematical modeling methods. In major centers of atmospheric research there have been worked out and are successfully applied modeling systems (for example, CMAQ (http://www.cmaqcenter.org), EURAD (EURopean Acid Deposition model, http://www.eurad.uni- koeln.de), European Zooming Model (http://aix.meng.auth.gr)) for carrying out scenario analyses, weather forecasting, and assessing the ambient air pollution level.

3. A computer modeling system was created within the project Integrated System for Intelligent Regional Environmental Monitoring & Management in a city/region (on the example of Tomsk region) of the European Community Framework 5 Program to assist in the analysis of the distribution of meteorological parameters and the concentration of admixtures in the atmospheric boundary layer above a rough inhomogeneous underlying surface. The nonhydrostatic prognostic mesoscale model and the model of pollution transformation and transport make the core of this system.

4. Modeling system of the Tomsk State University and the Institute of Atmospheric Optics (MS TSU-IAO) is intended To simulate meteorology and pollution transport on historical dates for understanding and management of urban air quality To execute scenario analysis for assessment of possible impact of designed plants, factories, airports or motorways

5. Components of the MS TSU-IAO Model initialization block (terrestrial data, ground-based observations, data of vertical distributions of meteorological parameters, data base of point, area and mobile sources of air pollution) Nonhydrostatic meteorological model Pollution transport model Data visualization block

6. Model initialization block Terrestrial data: topography, land use categories (albedo, soil thermal conductivity, heat capacity, density, evaporation, surface roughness, emissivity, deep soil temperature) Ground-based and vertical observations of wind velocity and wind direction, air temperature and humidity, atmospheric pressure

7. Numerical nonhydrostatic model Terrain following (zeta) coordinate system Nonhydrostatic hydrodynamic 3D equations 3D equations of heat and humidity exchange Two-equation “k-l” turbulence model 2D equation for surface temperature Assimilation of observed data Nesting technology

8. Numerical nonhydrostatic model Basic assumptions  Air density depends on pressure, temperature, atmospheric composition, but its temporal variations are negligible.  The processes of water vapor transformation in the atmosphere (formation of rain and cloud water, hail, snow, etc.) are not considered. But condensation and evaporation processes on underlying surface are taken into account.  Thermal radiation fluxes do not modify in the atmosphere.

9. Numerical nonhydrostatic model Governing equations 0

11. Numerical nonhydrostatic model Model of turbulence

12. Numerical nonhydrostatic model Numerical nonhydrostatic model Boundary conditions

13. Numerical method Analytic coordinate transformation Finite-volume method, 2 nd -order approximations for temporal and spatial derivatives, implicit or explicit- implicit schemes Splitting of pressure on base-state, geostropic, hydrostatic and non-hydrostatic parts SIMPLE algorithm derived by Patankar Iterative procedure of solution for each time step Nesting technology Grids: 50x50x30 for meteorology, 100x100x60 for pollution transport High-performance computations on multiprocessor systems with distributed memory

14. Data assimilation in model Geostrofic wind components are constructed on the basis of measured synoptic pressure field Applications of regional-scale values of model parameters (horizontal wind compounds, potential temperature and humidity) in the boundary conditions Adding forcing terms in the prognostic equations

15. Pollution transport model Eulerian 3D equations for basic anthropogenic pollutants of near surface layer (dust, CO, SO 2, NO 2 ) Dry deposition (resistance model) Photochemical reactions of Hurley’s GRS- mechanism of troposphere ozone and PM10 generation (CSIRO) Data base of distributed point, area, mobile (linear) sources

16. Land use classification and surface elevation of South of the Western Siberia Color table of land use categories: blue-water, violet-few vegetation, yellow-farmland, light green-deciduous forest, brown-mixed forest, green-evergreen forest, red-urban area Tomsk region Tomsk (85 o E, 56,5 o N)

17. Nesting technology Domain 200x200km 2 Domain 50x50 km 2

18. Land use categories for the research domain Tomsk city 50x50 km 2 r. Tom Tomsk

19. Surface elevation and roughness of landuse categories Water: 0,0001m Few vegetation: 0,15m Farmland: 0,15m Deciduous forest: 0,5m Mixed forest: 0,7m Evergreen forest: 1,0m Urban area: 1,0m  IAO TOR-station Meteo station  Meteo station

20. Comparison of the predictions and the observed data Meteodata of the IOA TOR-station and the Hydrometeorological Center of RF

21. MEMO Comparison of MEMO and MS TSU-IOA predictions

22. MEMO

23. Lower roughness Horizontal wind field at 10:00 15 August 2000 z=z surf +20m

24. Horizontal wind field at 10:00 29 June 2000 Lower roughness z=z surf +20m

25. 29 June 200015 August 2000 Vertical wind field at 10:00 z=z surf +20m

26. Wind field at 29-30 June 2000

27. Wind field at 15-16 August 2000

28. Prediction of pollutant concentrations in Tomsk 3 inert pollutants: CO, SO 2, NO 2 Point and area sources Emission rate of line sources: Q(h)=Qave*(0.1+1.9*sin(  (h- 6)/18), 6<h<24 hours Computation grid: 100x100x60 10-11 January 2000

29. Index of pollution

30. Fire in toxic waste area 10 January 2003 fire was happened in the Tomsk area of toxic waste, located in the north of Tomsk. Conflagration duration was from 10.00 to 24.00. Inhabitants of Seversk, Svetly and Tomsk felt foxy smell.

31. Toxic waste area Tomsk

32. Toxic waste area Tomsk

33. Toxic waste area Tomsk

34. Toxic waste area Tomsk

35. Toxic waste area Tomsk

36. Toxic waste area Tomsk

37. Toxic waste area Tomsk

38. Toxic waste area Tomsk

39. Toxic waste area Tomsk

40. Conclusions The developed numerical model was applied to investigate a wind field and pollution dispersion nearby Tomsk industrial center The predictions showed a close connection between a local meteorology and air quality in the city and suburban area.Undesirable meteorological situation is calm wind in conjunction with temperature inversion