Polarimetric Soil Moisture Retrieval at Short Wavelength

Soil moisture was assessed with quad-polarimetric TerraSAR-X and dual-polarimetric (HH/VV) TanDEM-X data from the agricultural region of Wallerfing, Germany. The investigations included a bare soil case in April 2009 (TerraSAR-X) and a time series of 04-07/2011 (TanDEM-X) to estimate soil moisture on bare and vegetated soils including different agricultural crop types at various phenological stages. Therefore polarimetric decomposition methods, incorporating a synthetic cross-polarization, were developed for X-band to separate ground scattering from vegetation scattering. Afterwards the IEM model was used to invert the ground or bare soil scattering components using the dominant scattering alpha angle. For both campaigns, comprising several crop and soil types and different phenological stages along the growth cycle, a root mean square error of <6vol.% (bare soil/TerraSAR-X) and of 5-14vol.% (vegetated soil/TanDEM-X) can be achieved for the soil moisture estimation.