Validation of OMI, GOME-2A and GOME-2B tropospheric NO2, SO2 and HCHO products using MAX-DOAS observations from 2011 to 2014 in Wuxi, China: investigation of the effects of priori profiles and aerosols on the satellite products

摘要

Tropospheric vertical column densities (VCDs) of NO, SO andHCHO derived from the Ozone Monitoring Instrument (OMI) on AURA and theGlobal Ozone Monitoring Experiment 2 aboard METOP-A (GOME-2A) and METOP-B(GOME-2B) are widely used to characterize the global distributions, trendsand dominating sources of these trace gases. They are also useful for thecomparison with chemical transport models (CTMs). We use tropospheric VCDsand vertical profiles of NO, SO and HCHO derived from MAX-DOASmeasurements from 2011 to 2014 in Wuxi, China, to validate the correspondingproducts (daily and bi-monthly-averaged data) derived from OMI and GOME-2A/Bby different scientific teams. Prior to the comparison, the spatial andtemporal coincidence criteria for MAX-DOAS and satellite data are determinedby a sensitivity study using different spatial and temporal averagingconditions. Cloud effects on both MAX-DOAS and satellite observations arealso investigated. Our results indicate that the discrepancies betweensatellite and MAX-DOAS results increase with increasing effective cloudfraction and are dominated by the effects of clouds on the satelliteproducts. In comparison with MAX-DOAS, we found a systematic underestimationof all SO (40 to 57 %) and HCHO products (about 20 %), and anoverestimation of the GOME-2A/B NO products (about 30 %), but goodconsistency with the DOMINO version 2 NO product. To better understandthe reasons for these differences, we evaluated the a priori profile shapesused in the OMI retrievals (derived from CTM) by comparison with thosederived from the MAX-DOAS observations. Significant differences are foundfor the SO and HCHO profile shapes derived from the IMAGES model,whereas on average good agreement is found for the NO profile shapesderived from the TM4 model. We also applied the MAX-DOAS profile shapes tothe satellite retrievals and found that these modified satellite VCDs agreebetter with the MAX-DOAS VCDs than the VCDs from the original data sets byup to 10, 47 and 35 % for NO, SO and HCHO,respectively. Furthermore, we investigated the effect of aerosols on thesatellite retrievals. For OMI observations of NO, a systematicunderestimation is found for large AOD, which is mainly attributed to effectof the aerosols on the cloud retrieval and the subsequent application of acloud correction scheme (implicit aerosol correction). In contrast, theeffect of aerosols on the clear-sky air mass factor (explicit aerosol correction) has asmaller effect. For SO and HCHO observations selected in the same way,no clear aerosol effect is found, probably because for the considered datasets no cloud correction is applied (and also because of the largerscatter). From our findings we conclude that for satellite observations withcloud top pressure (CTP)  900 hPa and effective cloud fraction(eCF)  10 % the application of a clear-sky air mass factor might be a goodoption if accurate aerosol information is not available. Another finding ofour study is that the ratio of morning-to-afternoon NO VCDs can beconsiderably overestimated if results from different sensors and/orretrievals (e.g. OMI and GOME-2) are used, whereas fewer deviations for HCHOand SO VCDs are found.