Surface Bi-Directional Reflectance Properties Over the ARM SGP Area from Satellite Multi-Platform Observations

Surface albedo is an important parameter in atmospheric radiation research. Good knowledge of surface bi-directional reflectance distribution function (BRDF) is essential in order to obtain the hemispheric reflectance (albedo) using data from satellite systems, which only view the ground at particular angles. Linear kernel-driven BRDF models have been commonly used for simple operational implementation of inversion procedure for derivation of model parameters. As an alternative to pixel-based approach in deriving BRDF parameters that uses all clear-sky pixel identified at particular point, a landcover-based approach is followed here assuming similarity in BRDF for the same landcover type in similar conditions. This approach is particularly suitable for the agriculture dominant area of the South Great Plain (SGP) where there are a limited number of cover types and within each type surface condition is fairly uniform unlike the majority of natural scenes. The approach can overcome a major limitation of pixel-based approach that is much more demanding on computational resources, and low number of observations for each particular point may impact adversely on the quality of BRDF parameters. Landcover-based approach provides much better sampling in terms of number of observations, requires less computational resources, may be implemented for shorter time intervals; however, may not always capture some spatial differences in the BRDF behavior.