Validation of kernel-driven semiempirical BRDF models for application to MODIS/MISR data

Increasing emphasis on global monitoring makes global modeling of bidirectional reflectance of vegetation an important issue. We study four linear kernel-driven semi-empirical models for use in operational production of the BRDF/albedo product for the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument of NASA's Earth Observing System (EOS). In this paper, we use a variety of multiple angle surface reflectance data sets from field and airborne measurements to evaluate these models. The results show that our semiempirical BRDF models provide a good mathematical description of the bidirectional reflectance of the land covers tested.