Retrieving aerosol characteristics and sea-surface chlorophyll from satellite ocean color multi-spectral sensors using a neural-variational method

► An algorithm was proposed to retrieve aerosol/oceanic properties from satellite data. ► The algorithm, named SOM-NV, was applied off the west African coast. ► SOM-NV provides new product and accurate aerosol optical thickness (AOT). ► It was found that the Saharan dust was at maximum in summer and minimum in November. ► A 13 year climatology of aerosol type, AOT and chl-a concentration is available.