Surface Soil Moisture Retrieval based on Model-Based Decomposition of Dual-Polarimetric Sentinel-1 Data

In this study, a physical model-based decomposition of dualpolarimetric C-band Synthetic Aperture Radar (SAR) data of Sentinel-1 was utilized to estimate surface soil moisture. The model-based decomposition using toke’s vector was used in this approach to determine the generalized surface parameter. The surface permittivity was estimated as a function of this generalized surface parameter, local incidence angle, and the permittivity (real). The permittivity estimates were used to empirically model the soil moisture. For validation, we collected in-situ measurements using the Stevens HydraGO soil moisture sensor corresponding to the satellite pass. The estimated observations and in-situ measurements exhibit a correlation of 0.82 with an RMSE of 0.02.