1. What is the Uncertainty Caused by IC/BCs in the Regional/Urban Ozone Simulations? Linking CMAQ with GEOS-CHEM Global/Regional/Urban Multiscale Study Nan-Kyoung Moon and Daewon Byun Institute for Multidimensional Air Quality Studies (IMAQS) University of Houston Rokjin Park and Daniel Jacob Harvard University

2. - Current usage; mostly rely on “climatological” fixed profiles - Could be different at each side of domain reflecting certain regional differences - Work best when outside the boundary of the domain does not have much direct emissions and no high concentration blobs already existing - Need to study sensitivity of the model simulations to the different IC/BCs - In reality fixed profiles are never accurate! Regional air quality modeling requires prescription of IC/BCs

3. The reality provided by ozonesonde observation

4. - Intermittent data in space and time -Mostly ozone is the only parameter available - There are many other important photochemical precursor species that must be prescribed - The intra-relations among the species must be consistent - Need a systematic IC/BC methods that can address above problems - Let’s try to link global tropospheric model results with regional air quality model Even with ozonesonde or other measurements….

5. This is a global 3-D model of atmospheric chemistry driven by assimilated meteorological observations from the Goddard Earth Observing System (GEOS) of the NASA Data Assimilation Office (DAO). It is being developed by groups at Harvard, Duke, NASA/GSFC, U. Washington, Rutgers U., JPL, EPFL/Lausanne, CNRS/Toulouse, and the University of L'Aquila, Italy, as a versatile tool for application to a wide range of atmospheric chemistry problems. First Look 4.0 Late Look 4.0 Conventional Final Dump - NCEPConventional CDAS Dump - NCEP SSM/I (TPW only)Wentz SSM/I (TPW only) Operational Sea Ice Interactive TOVS retrievals QuickSCAT SBUV Ozone SST - Downstream AverageReynolds SST - Centered Average ----GADS (Passive) NASA GEOS DAO Products Harvard GEOS-CHEM model

6. Air Quality Modeling with US EPA’s Models-3 CMAQ Community Multi-pollutant Multi-scale Air Quality Modeling System

7. First, Horizontal & Vertical Interpolations needed LAT-LON 2 degree X 2.5 degree 20 layers in Sigma P LAMBERT CONFORMAL 108 km X 108 km 23 layers in Sigma Z (Po) Initial & Boundary Condition in IO/API Format in 108km resolution GEOS-CHEM (Goddard Earth Observing System-CHEMisrty) MODEL3 CMAQ (Community Multi-scale/pollutant Air Quality model) To link GEOS_CHEM with EPA’s CMAQ ……

8. 24 species O 3 -NO X -Hydrocarbon chemistry : 24 species CMAQ CB4 O3-NOx-Hydrocarbon chemistry [NO2 ][NOx ] [O3 ][Ox ] - [NOx ] [N2O5] [HNO3] [PNA ][HNO4] [H2O2] [CO ] [PAN ][PAN ] + [PMN ] + [PPN ] [MGLY][MP ] [ISPD][MVK ] + [MACR] [NTR ][R4N2] [FORM][CH2O] [ALD2][ALD2] + [RCHO] [PAR ][ALK4] + [C3H8] + [C2H6] [OLE ][PRPE] [ISOP] GEOS-CHEM 16 species CB4 : 16 species Un-used species : ACET Second, develop procedures matching GEOSCHEM and CMAQ chemistry mechanism CB-4 example

9. GEOS2CMAQGEOS2CMAQ LAT-LON 2 O X 2.5 O w/ GEOSCHEM vertical coordinate Chemistry Mapping CB4 & SAPRC IC/BC Process IC/BC for CMAQ resolution The diagram of linkage between GEOSCHEM and CAMQ LAMBERT CONFORMAL w/ 108 km I/OAPI on CMAQ/MM5 vertical coordinate

10. Possible inconsistencies between the global and regional scale dynamics - inflow conditions at the boundary - differences in the evolution of dynamics with time Remedy (yet to be tested) -Run regional scale model with global scale output as input for initialization and analysis nudging cf: Currently most uses EDAS as basic input for MM5 Issues in linking GEOSCHEM and CMAQ (1)

11. GEOSCHEM ; DAO Comparison of Wind Fields GEOSCHEN : W -NWesterly (inflow) MM5 : Northerly (parallel to grid) Let’s see how big the problem is:

12. GEOSCHEM ; DAO

13. GEOSCHEM : Easterly and northerly MM5 : Clock wise rotation motion

14. GEOSCHEM : inflow MM5 : outflow

15. Chemistry Issues: - coarse resolution concentration distribution in global scale move into fine scale regional grid - differences in the evolution of concentration patterns with time - different chemical mechanism representation - different representation of atmospheric reactivity in the coarse scale Remedy - use as much consistent chemistry mechanisms (TBD) Issues of linking GEOSCHEM and CMAQ (2)

16. Mapping Table SAPRAC O3-NOx-Hydrocarbon chemistry [NO2 ][NOx ] – [NO] [PAN] [CO] [ALK3] [ALK4]+[ALK5[ALK4] [ISOPRENE ][ISOP] [HNO3] [H2O2] [ACET ] [MEK] [CCHO][ALD2] [RCHO] [MRTHACRO][MACR] [MA_PAN][PMN] [MVK] [PAN2][PPN] SAPRAC O3-NOx-Hydrocarbon chemistry [RNO3][R4N2] [OLE1] + [OLE2][PRPE] [ALK2][C3H8] [HCHO][CH2O] [ALK1][C2H6] [N2O5] [HNO4] [COOH ][MP] CMAQ GEOS-CHEM SAPRAC-99 (yet to be simulated)

17. Horizontal distribution of O3 concentration from GEOS-CHEM global output at Layer 1 108km resolution 2 X 2.5 degree resolution

18. Horizontal distribution of CO concentration from GEOS-CHEM global output at Layer 1 108km resolution 2 X 2.5 degree resolution

19. 1. GEOSCHEM vs. CMAQ CONUS 2. CMAQ CONUS vs. CMAQ Regional 36-km 3. Effects of using different IC/BCs - profile vs. GEOS-CHEM for CONUS domain Comparative Study with CMAQ

20. The comparison of vertical cross section between GEOSCHEM and CONUS results GEOSCHEM 108kmCONUS 36km

21. O3.. August 16, 2000, 00UTC (First day of simulation) GEOSCHEMCONUS August 16, 2000 (First day of simulation) September 1, 2000 (Last day of simulation)

22. September 1, 2000. 09 & 21UTC GEOSCHEMGEOSCHEM CONUSCONUS

23. CO.. August 16, 2000, 00UTC GEOSCHEMCONUS

24. September 1, 2000. 09 & 21UTC GEOSCHEMGEOSCHEM CONUSCONUS

25. NO2.. August 16, 2000, 00UTC GEOSCHEMCONUS

26. September 1, 2000. 09 & 21UTC GEOSCHEMGEOSCHEM CONUSCONUS

27. The comparison of vertical cross section for O3, CO, FORM and NO2 between Regional vs CONUS 36-km CMAQ results. ICBC from GEOSCHEM 108km data

28. O3, August 31, 2000. 09 & 21 UTC Comparison of horizontal distribution between CONUS and Regional 36km

29. CO, August 31, 2000. 09 & 21 UTC

30. The comparison of vertical cross section between CONUS and regional domain results CONUS 36kmRegional 36km

31. August 31, 2000. 09 & 21UTC CONUSCONUS R e gi o n al 36 km

32. August 31, 2000. 09 & 21UTC CONUSCONUS R e gi o n al 36 km

33. The comparison for O3, CO and NO2 between profile IC/BC vs. IC/BC from GEOSCHEM 108km data

34. Profile Data Case GEOS-CHEM Data Case The comparison of CMAQ results in different IC and BC (2000.08.25. 09, 21UTC) 03AM CST 03PM CST

35. 08/28/2000 03AM CST09AM CST 12:00 CST03PM CST

36. Profile Data Case GEOS-CHEM Data Case Continued, now at 4-km resolution (August, 26, 2000, 21UTC)

37. Conclusive Remarks Issues related with linking global tropospheric chemistry model with regional air quality model has been studied Problems with current fixed profile method identified Global tropospheric model provides needed dynamic evolution and concentration distribution realism not existing in profile method We observe significant changes in the atmospheric reactivity conditions depending on profile vs. GEOSCHEM IC/BC Global-regional scale linking is the best when outside the regional domain boundary does not have much direct emission sources; e.g., CONUS domain Need to study the issues of harmonization of chemical mechanisms further Need to quantify and minimize the effects of different dynamics between the global and regional meteorological data used