An alternative simple approach to estimate atmospheric correction in multitemporal studies

Abstract

Studies that use multitemporal images require the conversion of original digital data into the corresponding physical magnitudes. Atmospheric correction is one of the most important steps in this process, which is usually undertaken using atmospheric radiative transfer models. The main difficulty in these models is the need of atmospheric input data which are not usually available. An alternative approach to atmospheric correction is proposed in this Letter. It is based on the idea that the atmospheric effects over two or more dates can be determined in a relative way, by using the apparent reflectance values of surfaces whose ground reflectance can be considered unchangeable with time. This method has been applied to correct Landsat-5 Thematic Mapper images and satisfactory results have been obtained in comparison with those obtained by a traditional model.