Aerosol optical thickness retrieval over land and water using Global Ozone Monitoring Experiment (GOME) data

Abstract

An algorithm for the retrieval of the aerosol optical thickness over land and over water from Global Ozone Monitoring Experiment (GOME) data is presented. The cloud fraction in the GOME pixels is determined using the Fast Retrieval Scheme for Clouds From the Oxygen A Band (FRESCO) algorithm. Surface contributions to the top of atmosphere reflectance are determined from the GOME surface reflectance database. The aerosol retrieval algorithm uses lookup tables that were created using the radiative transfer model 6S. The algorithm allows retrieving the aerosol types characterized by Ångstrom coefficients in the range from −0.1 to 2.8; i.e., the range of values observed by the AERONET ground‐based measurements. Validation of the algorithm done using the AERONET Sun photometer data for 12 sites in Europe and Africa, for the year 1997, shows very good agreement. The correlation coefficient between the satellite retrieval and AERONET data for the wavelength of 440 nm is 97%, and for 670 nm it is 94%. Validation of the algorithm for the year 2000 was done for a few sites with similar results. The algorithm has been successfully tested over an island influenced by Saharan dust (i.e., Cape Verde, 16°N, 22°W) and over a site located near the Saharan desert (i.e., Bondoukoui, 11°N, 3°W). For other sites located near the Saharan desert such as Bidi Bahn (14°N, 2°W) and Banizombou (13°N, 2°E), the agreement was very good at 440 nm. The algorithm has not been tested over other bright surfaces such as ice‐covered regions. Examples of the spatial distribution of the aerosol optical thickness over Europe, north Africa, and the North Atlantic for the year 1997 and 2000 are presented.