Regression-based soil moisture estimation from SMAP polarimetric radar data with Aquarius derived coefficients

Soil moisture radar retrieval algorithms generally utilize substantial ancillary data to parametrize complex electromagnetic models. In this work, we utilize an existing empirical approach as an alternative; it describes radar backscatter of a vegetated scene as a linear function of soil moisture, hence reducing dependence on ancillary data. We use 2.5 years of L-band Aquarius radar and radiometer derived soil moisture data to calculate two coefficients of a linear model function on a global scale; the coefficients are derived on a per pixel and on an IGBP land cover classification basis. These coefficients are used to estimate soil moisture from 2.5 months of L-band SMAP radar data and are evaluated with the SMAP Level 2 radiometer-only soil moisture product.