Remote sensing of aerosols over land from VEGETATION and SeaWiFS

Abstract

The key problem in aerosol retrieval over land is to distinguish between surface and atmospheric contribution to the satellite reflectance. In principle a method similar to the classical Dense Dark Vegetation could be used over brighter surfaces if the surface BRDF could be described with sufficient accuracy. Studying a time series of data, taking into account geometrical conditions, and assuming the ground BRDF to be constant over several days, variations of the satellite signal may be mainly attributed to variations of the atmospheric optical properties. By fitting a subset of satellite observations associated with ground photometer data, the best-fit of BRDF parameters could be determined. Using then this surface characterization as an input of the inversion process, the aerosol optical thickness can in principle be retrieved routinely. Such method has been already explored with a time series of VEGETATION data for several Western European sites. The aerosol optical thickness retrieved from the satellite data and that derived from CIMEL measurements were in good agreement, even for cases of high optical depth. The method has been now improved with the sensor SeaWifS. Aerosol properties retrieved from SeaWifS data have been compared with those measured in-situ by sunphotometers.