Spatial distribution analysis of the OMI aerosol layer height: a pixel-by-pixel comparison to CALIOP observations.

A global picture of atmospheric aerosol vertical distribution with a high temporal resolution is of key importance not only for climate, cloud formation and air quality research studies, but also for correcting aerosol radiation effect in absorbing trace gas retrievals from passive satellite sensors. Aerosol layer height (ALH) was retrieved from the OMI 477 nm O₂-O₂ band and its spatial pattern evaluated over selected cloud-free scenes. Such retrievals benefit from a synergy with MODIS data to provide complementary information on aerosols and cloudy pixels. We used a neural network approach previously trained and developed. Comparison with CALIOP aerosol level 2 products over urban and industrial pollution in east China shows consistent spatial patterns with an uncertainty in the range of 462-648 m. In addition, we show the possibility to determine the height of thick aerosol layers released by intensive biomass burning events in South-America and Russia, and of a Saharan dust outbreak over sea from OMI visible measurements. Complementary detailed analyses show that the assumed aerosol properties in the modeling are the key factors affecting the accuracy of the results, together with potential cloud residuals in the observation pixels. Furthermore, we demonstrate that the physical meaning of the retrieved ALH scalar corresponds to the weighted average of the vertical aerosol extinction profile. These encouraging findings strongly suggest the potential of the OMI ALH product, and in more general the use of the 477 nm O₂-O₂ band from present and future similar satellite sensors, for climate studies as well as for future aerosol correction in air quality trace gas retrievals. [ABSTRACT FROM AUTHOR]