SEPARATION OF VEGETATION AND ROCK SIGNATURES IN THEMATIC MAPPER AND POLARIMETRIC SAR IMAGES

A technique for separating classes of surface roughness and vegetation in semiarid regions using polarimetric Synthetic Aperture Radar (SAR) and Thematic Mapper (TM) data has been developed. We conclude that polarimetric SAR data may be used to separate areas with < 10% vegetation cover where one suite of model inversions may be applicable for derivation of surface roughness. In more vegetated regions, visible/near-infrared (NIR) data will provide a mechanism for determining the relative contribution of vegetation to multiple scattering in the SAR data. Thus, the synergistic use of the two data sets provides a more complete characterization of surface units and cover. This result, while important for geologic interpretations, is even more critical when trying to monitor dielectric constant variations related to changes in moisture content of soil and vegetation cover in arid to semiarid regions. A strategy for automating a procedure for measurement of surface roughness is proposed.