1. Two Decades of Tropospheric Ozone Observations from Satellite Measurements, the Use of these Data in Models, and Some Insight into Future Capabilities Jack Fishman, 1 John K. Creilson, 1,2 Amy E. Balok 1,2 and Fred M. Vukovich 1,2 1 Atmospheric Sciences NASA Langley Research Center Hampton, Virginia USA 23681 2 also at SAIC International, Inc. Hampton, Virginia USA 23666 Presented at: 10th CACGP Scientific Conference 7th IGAC Scientific Conference Heraklion, Crete, Greece September 20, 2002

2. Separate Stratosphere from Troposphere to Compute Tropospheric Ozone Residual (TOR)

3. Climatological Comparison of Ozonesonde Data with SBUV Measurements at Wallops Island

4. Information Contained in SBUV Measurements

5. Schematic Diagram of Empirical Correction Input SBUV Measurement: Output * for TOR Calculation (A + B + C)C * = Z 1 (A + B + C)/(X + Y + Z 1 ) B * = Y (A + B + C)/(X + Y + Z 1 ) A * = X (A + B + C)/(X + Y + Z 1 )

6. Comparison Using Empirical Correction with Ozonesondes

7. Other Data Sets Are Required To Separate Tropospheric Ozone from Total Ozone Measurements SAGE: Good Vertical Resolution; Poor Spatial Coverage HALOE: Good Vertical Resolution; Poor Spatial Coverage MLS: Vertical Resolution Only >68 mb; Relatively Good Spatial Coverage Only One Archived Layer below 100 mb SBUV: Poor Vertical Resolution; Good Spatial Coverage Archived Layers: 1000–253 mb; 253-126 mb; 126-63 mb Stratospheric Fields Generated from 5 Days of Data SAGE/TOMS TOR: ~ 30,000 Coincident Observations 1979-1991 [Fishman & Brackett, 1997]  ~ 10 data points per 5° x 10° grid box for seasonal climatology SAGE/SBUV TOR: Use Every TOMS Observation (up to 28,800 per day)  ~ 1500 data points per 1° x 1.25° grid box for seasonal climatology Tropopause Heights: Archived Gridded Data Sets 2.5° x 2.5°

8. Comparison of Pixel Size for Computing TOR SAGE/TOMS TOR (5° x 10°) (1° x 1.25°)

9. DU July 1988 Monthly TOR Captures High Ozone During Major Pollution Episode July 1988 TOR July 3-15 Average Daily Max. O 3 (from Schichtel and Husar, 1998)

10. Seasonal Depictions of Climatological Tropospheric Ozone Residual (TOR) 1979-2000 SBUV Tropospheric Ozone Residual (TOR) SON 1979-2000 SBUV Tropospheric Ozone Residual (TOR) DJF 1979-2000 SBUV Tropospheric Ozone Residual (TOR) JJA 1979-2000 SBUV Tropospheric Ozone Residual (TOR) MAM 1979-2000

11. Comparison of TOMS/SAGE TOR with TOMS/SBUV TOR

12. Global TOR Averages Change with TOMS Archive Fishman et al. [1990]: 32.7 DU (pseudo-Version 6/SAGE) Version 6 corrected for instrument drift Fishman & Brackett [1997]:27.5 DU (Version 7/SAGE) Version 7 incorporates ISCCP cloud climatology for correction This Study:31.5 DU (pseudo-Version 8/SBUV) Version 8 includes aerosol and scan-angle dependence corrections

13. Total Ozone along Latitude Line Spectral Contribution to Total Ozone at 14.5°S Wavenumber Unfiltered V.7Total Ozone from TOMS ArchiveTOMS Total Ozone After Fourier Filter Fourier Filter Applied to TOMS Archive to Remove Scan-Angle Dependence

14. Population and Ozone Pollution Strongly Correlated in India and China TOR in Dobson Units Summer Climatological Distribution

15. GOME NO 2 Measurements Also See Enhancements over India and China

16. DU July 1988 Monthly TOR Captures High Ozone During Major Pollution Episode July 1988 TOR July 3-15 Average Daily Max. O 3 (from Schichtel and Husar, 1998) Lower TOR within box due to terrain artifact Use terrain information for global validation

17. Lower TOR over North African Desert Regions Coincident with Higher Elevations December-February TOR 3400 m 2000 m 3100 m 4600 m 3800 m 2900 m > 2000 m Implications : TOMS Capable of Isolating Small (Regional) Scale Features ~3 DU for   2km dz  20 ppb in pbl Information can be used to validate O 3 backscatter sensitivity in boundary layer over cloudless unpolluted area TOR in DU

18. Inferred Ozone Profile over North Africa Desert Region:   2 km [O 3 ] dz = ~3 DU    4 km [O 3 ] dz = ~6 DU   Trop. (~17 km) [O 3 ] dz = ~25 DU 3400 m 2000 m 3100 m 4600 m 3800 m 2900 m > 2000 m Higher Elevation Differences (3-4 km) Coincident with Greater O 3 Deficits (5-7 DU) 18 20 22 24 26 TOR in DU Altitude (km) O 3 (ppbv)

19. Ozone Enhancement over India TOR in Dobson Units Summer Climatological Distribution June 1982June 1999 How does the Amount of Ozone over India Compare with the Amount Observed over the Eastern United States?

20. TOR and Surface O 3 Depiction During July 3-15 Pollution Episode July 1988 June 1982 Comparison of Indian and U.S. Air Pollution Episodes

21. 1988 Worst Year

22. Examine Interannual Variability of Ozone Enhancement over India TOR in Dobson Units Summer Climatological Distribution June 1982June 1999 1982: El Niño Year and El Chichon Eruption on 6 April Did aerosol loading affect TOMS total ozone retrieval?

23. Monthly TOR Values Over Northern India 1979-1999 Jan29.8Feb29.9Mar34.6Apr44.May47.3Jun48.2Jul46.4Aug42.0Sep36.8Oct32.7Nov30.5Dec27.9 199131.5199233.3198940.5198247.2198252.9198252.1198248.3199243.7199040.1199935.0198133.2199730.0 198431.2198733.0198238.1198447.1198150.0198951.3199247.8199043.5198837.9199834.4198832.1198529.5 199830.9198432.6199037.9199145.9199049.8199251.2198747.6198743.4199237.6198533.2199731.8199928.8 199030.7197931.8198736.7197945.7198949.5199049.9199047.6199143.1199137.3198633.1199231.7198328.7 198630.7198331.0198435.5198144.8199249.2199149.3199147.5198242.9198937.1198033.0199131.5198928.4 198730.7198831.0198134.9198944.7198348.1198748.5198946.9198942.4198637.0199032.9199931.3198828.4 197930.3198630.8199834.8199244.6198647.5198448.5198446.6198842.4199836.8198332.9197931.2198128.3 198830.1199330.7198834.7198044.5199147.4198047.5198846.6198342.3198536.7198932.9198730.7199028.3 199929.8199030.5199334.7199344.4197947.2198847.5198146.6198442.2198736.7199132.7198230.1199227.8 198129.8198530.2197934.0198644.4198446.6198147.4198346.0197942.0198336.6197932.7198330.0197927.7 198329.7198129.4198634.0198744.0199946.2197946.8198646.0198141.4199736.4198832.6198529.8198027.7 199329.7199829.1198033.9199843.6198046.1198346.8198045.7198541.1198036.3199732.5198929.6198227.3 198529.4199929.0198532.8199043.5198845.6198546.6197945.1198040.8198136.0198132.4199029.2198727.1 198229.1198228.6198332.6198341.6198544.4198646.4198544.9198640.6198436.0198232.1198028.8199126.7 198929.1198925.8199232.3198841.1198744.2199846.0199844.8199840.5199935.8199232.0198628.6198626.1 199227.6198025.5199926.7199940.8199842.4199945.4199944.0199940.4198235.5198431.8198428.6198425.8 198025.7199125.1198540.5197935.2198730.6 Monthly Averages for Each Year are Rank-Ordered: 1982 Highlighted in Red 1999 Highlighted in Blue

24. Tahiti Darwin Definitions of ENSO Indicators Classic SOI:  P between Darwin & Tahiti Other definitions include Sea Surface Temperature Anomalies (SSTA) in various regions of the Pacific: Niño 1+2: Off coast of Ecuador; Niño 3: Eastern Pacific; Niño 4: Western Pacific; Niño 3.4: Central Pacific

25. ENSO Year  Summer India TOR and SSTA-Niño 4 from 1979-1999 TOR (DU) SSTA (°C)

26. Correlation Coefficients Between Northern India Monthly TOR Values and Monthly/Seasonal ENSO Indicators (1979-1999) Most Significant Relationship between Summer TOR and Seasonal ENSO Indicators Shaded Values Statistically Significant (>.9 confidence level) Note: Monthly Average for each year comprised of >7500 TOR measurements (252 points x ~30 days)

27. Springtime TOR Variability Over Atlantic Mid-Latitudes Linked to Differences in Prevailing Transport Patterns Spring 1992 Spring 1980

28. North Atlantic Oscillation Determines Intensity of Transport Across Atlantic

29. Strong Correlation between TOR and NAO Index

30. Summary of 20 Years of SBUV/TOMS TOR Measurements Data located at: http://asd-www.larc.nasa.gov/TOR/data.html Strong Correlations Apparent: - Pollution and Population Distributions - N.E. India and ENSO - Atlantic and NAO High Resolution Data Delineate Elevated Terrain - Possible Use for Validation Can ENSO or Other Indicators be Used as Predictors?

31. Part II: The Use Satellite Data to Improve Atmospheric Chemistry Models

32. :Lon Outcomes: Assess effects of emissions control options. Evaluate development options and emissions strategies to set policies and construct attainable State (air quality) Implementation Plans. Outcomes: Quantify contributions of physical & chemical processes to pollutant concentrations. Extend ozone forecasting to regional transport for urban to rural areas. Outcomes: Determine source and destination of long range dust and pollutants. Route airplanes. More accurate forecasts of haze & pollution episodes. Warn hospitals & farmers. Outcomes: Reassess ozone and precursor transport across state boundaries. Implement air quality strategies & related development policy based on detailed data and models. Outcomes: Accurate (regional, multi-day) pollution forecasts. NAAQS planning and mitigation based on validated models. Current trajectory: Steady improvement in documenting the chemical content of the lower atmosphere, Steady improvement in the physical accuracy of modeled processes for pollution episode warnings. Road Map to Use Satellite Measurements In Conjunction with Global and Regional Models to Develop Air Quality Forecast Capability CMAQ / Forecasts: State/regional planning. Same-day air quality predictions. 2000200220042006200820102012 Simultaneous, high time & space resolved multi- pollutant (O3, CO, NOx, SO2, HCHO, aerosol) data enables sound decision making Outcomes: Evaluate exceptional events for effect on NAAQS violations; provide exceptions for attainment. Large scale transport of aerosols (TOMS aerosol index) Earth System Modeling Framework Forecasts by 2012: Robust emissions control planning and management. Routine warnings of elevated pollution episodes. Accurate 3-day air quality forecasts. Improve boundary conditions (ozone residual) Validate measurements Ozone, SO 2 & NO 2 profiles & regional transport (Build on TOMS & GOME) Continental inflow/outflow Monitor long range transport of mineral and pollution aerosol (CALIPSO) Couple chemistry & aerosol models Assimilate satellite data for trace constituents TOMSAQUA SAGE III AURA TERRA CALIPSO Cloud Sat NPOESS Impacts: Reduce impaired lung function and use of medications. Reduce hospital admissions and lost work/school days. Impacts: Reduce asthma & lung related diseases. Improve visibility. Improve crop health & yields.

33. Part III: New Satellites Will Provide Better Measurements

34. Methodology to Derive TOR with Instruments from Aura Satellite Using HIRDLS (High-resolution Infrared Dynamic Limb Sounder) and OMI (Ozone Monitoring Instrument) Define Distribution of Ozone in the Stratosphere and Upper Troposphere from Profile Information Use HIRDLS Profiles with 4°-latitude x 5°-longitude Resolution Normalize HIRDLS Stratospheric Column Ozone (SCO) to OMI SCO Derived from Convective Cloud Differential (CCD) Method Subtract Stratospheric Ozone Fields from Higher Resolution OMI Total Ozone Fields to Derive TOR Distribution

35. HIRDLS Daily Profile Coverage Will Provide Sufficient Information to Derive 3-Dimensional Stratospheric Ozone Distribution Down to 1 km Below Tropopause

36. Objective Provide Quasi-Global Daily Maps of Tropospheric Ozone with 48 x 52-km Resolution

37. OMI Goal is to Produce Daily TOR Distribution with Better Resolution Map of Houston and surrounding area From Fishman and Balok (1999) OMI Total O 3 Resolution OMI TOR Resolution

38. Geostationary Observations Will Provide Hourly Observations with 5-km Resolution TOMS (Daily) OMI (Daily) GeoTRACE (Hourly) Map of Houston and surrounding area

39. Summary 2-Decade Record of TOR Now Available Strong Correlation between Population and Pollution Interannual Variability over Northern India Linked to ENSO Transport of Pollution across Atlantic Linked to NAO Challenge to Use Satellite Measurements with Models to Understand/Forecast Global and Regional Pollution New Satellites Promise Much Better Tropospheric Measurement Capability within Next Few Years