A NEW METHOD OF DERIVING TIME-AVERAGED TROPOSPHERIC COLUMN OZONE OVERTHE TROPICS USING TOTAL OZONE MAPPING SPECTROMETER (TOMS) RADIANCES -INTERCOMPARISON AND ANALYSIS USING TRACE-A DATA

Error analysis of archived total O-3 from total ozone mapping spectrom eter (TOMS) (version 6) presented in earlier studies [Hudson and Kim, 1994; Hudson et al., 1995] is extended to include scan angle effects. Daily total O-3 maps for the tropics, from the period October 6-21, 19 92, are derived from TOMS radiances following correction for these err ors, These daily maps, averaged together, show a wavelike feature, whi ch is observed in all latitude bands (10 degrees N to 14 degrees S), u nderlying sharp peaks which occur at different longitudes depending on the latitude. The wave pattern is used to derive both time-averaged s tratospheric and tropospheric O-3 fields, The nature of the wave patte rn (stratospheric or tropospheric) cannot be determined with certainty due to missing data (no Pacific sondes, no lower stratospheric Strato spheric Aerosol and Gas Experiment (SAGE) ozone for 18 months after th e Mt. Pinatubo eruption) and significant uncertainties in the corrobor ative satellite record in the lower stratosphere (solar backscattered ultraviolet (SBUV), microwave limb sounder (MLS)). However, the time-a veraged tropospheric ozone field, based on the assumption that the wav e feature is stratospheric, agrees within 10% with ultraviolet differe ntial absorption laser Transport and Atmospheric Chemistry near the Eq uator-Atlantic) (TRACE A) O-3 measurements from the DC-8 [Browell et a l., this issue] and with ozonesonde measurements over Brazzaville, Con go (4 degrees S, 15 degrees E), Ascension Island (8 degrees S, 15 degr ees W), and Natal, Brazil (5.5 degrees S, 35 degrees W), for the perio d October 6-21, 1992, The derived background (nonpolluted) Indian Ocea n tropospheric ozone amount, 26 Dobson units (DU), agrees with the cle anest African ozonesonde profiles for September-October 1992. The assu mption of a totally tropospheric wave (flat stratosphere) gives 38 DU above the western Indian Ocean and 15-40% disagreements with the sonde s. Tropospheric column O-3 is high from South America to Africa, owing to interaction of dynamics with biomass burning emissions [Thompson e t al., this issue (a, b)]. Comparison with fire distributions from adv anced very high resolution radiometer (AVHHR) during October 1992 sugg ests that tropospheric O-3 produced from biomass burning in South Amer ica and Africa dominates the O-3 budget in the tropical southern hemis phere during the study period.