Inferring Surface Solar Absorption from Broadband Satellite Measurements

Abstract

An atmospheric solar radiation model, in conjunction with a variety of surface albedo models, has been employed to address several issues related to inferring the surface solar radiation budget from satellite measurements. With reference to albedo determinations using narrowband instruments such as GOES and METEOSAT, it is suggested that converting these to broadband quantities, either at the surface or at the top of the atmosphere, could be a formidable task for vegetated surfaces. To do this operationally it would seem necessary to subdivide vegetated surface into a large number of categories, incorporate some appropriate means of scene identification, account for probable seasonal variations in surface albedo and surface anisotropy, and then devise a quantitative method for actually performing the conversions. It would be preferable to evaluate broadband quantities from broadband measurements. A further point of this paper, however, is that a commonly used linear conversion between broadband planetary and surface albedos is also strongly dependent upon vegetation type. It is then alternatively proposed that a linear slope-offset relationship exists between surface and surface-atmosphere solar absorption. For clear skies this relationship is only modestly dependent upon scene type, requires only a rudimentary correction for variations in atmospheric water vapor, and, with the exception of deserts, necessitates only a modest correction for tropospheric aerosols over land areas. No correction for maritime aerosols over ocean area is required. This study also elucidates problems and possible approaches for dealing with overcast areas and regions containing broken clouds.