Soil Moisture Retrieval Using L-Band Radar Observations

An algorithm for surface soil moisture estimation using L-band radar observations is introduced. The formulation envelops a wide range of land surface conditions based on three limiting cases defined in terms of end-members: smooth bare soil, rough bare soil, and a maximum vegetation covered soil. Parameterizations for these end-members are obtained using forward electromagnetic scattering models. Modulation due to soil surface roughness and overlying vegetation scattering effects between end-members are accounted using the radar vegetation index and the newly introduced radar roughness index. Hence, the retrieval algorithm developed here does not depend on ancillary vegetation or roughness information. The algorithm is tested with ground-based truck-mounted bare soil observations and observations from several airborne field campaigns that represent a wide range of surface conditions.