The present art applies a unique conceptual model that uses image statistics of top-of-atmosphere reflectance to assess and then reverse atmospheric effects. The workflow first calibrates the method for an Earth observation satellite pre-operationally. Then, for operational conversion of each image of the satellite, the degree of atmospheric effect is mapped across an image to be converted and then reversed to deliver surface reflectance data for each pixel of the image. The resulting surface reflectance images have haze and thin clouds removed, are clarified for viewing and appropriate for automated analyses.