1. 1 Relationship between ground-level NO 2 concentrations and OMI tropospheric NO 2 columns 11.09.2007 Martin Steinbacher, Empa; Edward Celarier, SGT Inc. Eric Bucsela, NASA GSFC; Edward Dunlea, CIRES Joseph Pinto, U.S. EPA Lok Lamsal and Randall Martin Tropospheric NO 2 workshop, 10-12 Sept 2007, KNMI, De Bilt

2. 2 Motivation  Surface NO 2  Toxic pollutant  Associated with mortality  Surface NO 2 data to provide information on level of toxicity in air  Surface measurements of NO 2  Sparse measurement sites  Large gradient in NO 2 field  May not reflect personal exposure to NO 2  Interference from reactive nitrogen Correlation of NO 2 and PM 2.5 with PAH and VOC [Brook et al., 2007] PAH VOC 0.0 0.2 0.4 0.6 0.8 Correlation coefficient

3. 3 Tropospheric column as a proxy for surface NO 2  Tropospheric columns related with surface NO x emissions  NO 2 in the urban mixed layer can make 70-90% contribution to tropospheric column  Tropospheric column a proxy to infer surface NO 2 San Francisco Los Angeles Phoenix Houston Dallas Chicago Toronto

4. 4 GEOS-Chem and our approach to derive surface NO 2 from OMI GEOS-CHEM In situ 100% difference in GEOS-Chem surface NO 2  < 10% error in derived surface NO 2 in polluted areas Tropospheric NO 2 column (standard product) S NO 2 → Surface NO 2 T NO 2 → Tropospheric NO 2 column

5. 5 Ground-based NO 2 measuring instruments Chemiluminescent NO x analyzer Molybdenum converter Photolytic converter  Commonly used instrument  Specific to NO  Indirect measurement of NO 2  Significant interference from other reactive nitrogen DOAS, LIF, TILDAS (research grade instruments)  Specific to NO 2  Some interference from HONO

6. 6 Interference in molybdenum converter analyzer: Comparison with DOAS measurements ► Interference~50%, HNO 3 ~60%, Alkyl nitrates~10-30% ► Mexico City, Apr 2003 [ Dunlea et al. 2007] OMI overpass time (12.00 to 14:00 local time) Molybdenum Conv. DOAS

7. 7 Interference in molybdenum converter analyzer: Comparison with photolytic converter measurements Taenikon/Switzerland Jan-Dec 2000 Steinbacher et al. 2007 Interference:20-60% PAN:30-50% HNO 3, particulate nitrates

8. 8 Interference in molybdenum converter analyzer: Experiments CompoundsConversion efficiencyExperiments NO 2, ethyl nitrate~ 100%Winer et al., 1974 PAN92%Winer et al., 1974 HNO 3, PAN, n-propyl nitrate, n-butyl nitrate ≥98% Grosjean and Harrison, 1985 Ammonia, gas phase olefins, particulate nitrate No significant interference Dunlea et al., 2007 Difficult issue: Loss of HNO 3 on stainless steel of inlet Difficult to quantify the conversion efficiency

9. 9 Correction for interference Bias = 40% Bias = 8% Taenikon, Switzerland

10. 10 Correction for interference for US and Canadian sites

11. 11 Comparison between corrected in situ and OMI-derived surface NO 2  Selected 214 stations  201 US and 13 Canada  Collocation criteria  Radius 10 km, time = 12:00 to 14:00 local time ► Mean correlation 0.51, maximum 0.86 ► Stronger correlation in polluted areas Correlation map

12. 12 Comparison between corrected in situ and OMI-derived surface NO 2 DJF MAM JJA SON uncorrected in situ corrected in situ OMI-derived

13. 13 Comparison between corrected in situ and OMI-derived surface NO 2 J F M A M J J A S O N D

14. 14 Comparison between OMI and in situ measurements according to land type  Urban, suburban, rural  Mean bias -21 to -48%  Seasonal bias in OMI retrievals DJF MAM JJA SON 33% -20% -58% -15%

15. 15 Conclusions  Ground-level NO 2 from OMI using  Tropospheric NO 2 columns  NO 2 profile shapes from GEOS-Chem model ► Derived NO 2 not sensitive to model profile  Interference in chemiluminescent NO x analyzer  Interfering species: PAN, alkyl nitrates, and HNO 3  Developement of a correction algorithm following lab/field measurements and using GEOS-Chem  Comparison between corrected in situ and OMI-derived surface NO 2  OMI-derived surface NO 2 less than corrected in situ by 21 to 48%  Seasonal bias in OMI-derived surface NO 2