 This paper examined two methods used to estimate atmospheric correction for optical remote sensing satellite data. The first method was a multi-spectral approach which assumed a constant relationship between surface reflectance measurements and aerosols. The second method was a temporal approach which assumed that surface reflectance remained consistent over time. Both methods were tested against in situ measurements from the aeronet network. The results showed that the multi-spectral approach provided the best accuracy overall, while the temporal approach performed better when there was no vegetation present or when the landscape changed quickly. When both approaches were combined into a spectrotemporal approach, the results were even more accurate. This shows that combining multiple methods can improve the accuracy of atmospheric correction for optical remote sensing satellite data. This article was authored by Olivier Hegel, Mary Huck, David Villa Pascual and others. We are article.tv, links in the description below.