The paper presents an empirical model for inferring soil moisture and surface roughness from radar data using imaging radars. The model is developed using data from the University of Michigan's LCX POLARSCAT and the RASAM radar system, which operate at various frequencies. The backscattering coefficients σ^0_M and σ^0_v are empirically derived and follow specific equations that account for incidence angle, dielectric constant, surface roughness, wave number, and wavelength. The model's performance is assessed by comparing radar-derived soil moisture estimates with in situ measurements, showing an accuracy of better than 4%. The study also discusses the impact of vegetation on the inversion technique and proposes a method to eliminate areas where vegetation impairs the algorithm. The calibration requirements for the algorithm are quantified, and it is shown that current operational multipolarization SAR systems meet or exceed these requirements. The effectiveness of using the ratio of cross-polarized to like-polarized returns to identify suitable areas for inversion is demonstrated, with this ratio positively correlated with the Normalized Difference Vegetation Index (NDVI). The paper concludes with a detailed analysis of inversion results from multiple datasets, including AIRSAR and SIR-C data, and highlights the potential for accurate soil moisture mapping using these sensors.The paper presents an empirical model for inferring soil moisture and surface roughness from radar data using imaging radars. The model is developed using data from the University of Michigan's LCX POLARSCAT and the RASAM radar system, which operate at various frequencies. The backscattering coefficients σ^0_M and σ^0_v are empirically derived and follow specific equations that account for incidence angle, dielectric constant, surface roughness, wave number, and wavelength. The model's performance is assessed by comparing radar-derived soil moisture estimates with in situ measurements, showing an accuracy of better than 4%. The study also discusses the impact of vegetation on the inversion technique and proposes a method to eliminate areas where vegetation impairs the algorithm. The calibration requirements for the algorithm are quantified, and it is shown that current operational multipolarization SAR systems meet or exceed these requirements. The effectiveness of using the ratio of cross-polarized to like-polarized returns to identify suitable areas for inversion is demonstrated, with this ratio positively correlated with the Normalized Difference Vegetation Index (NDVI). The paper concludes with a detailed analysis of inversion results from multiple datasets, including AIRSAR and SIR-C data, and highlights the potential for accurate soil moisture mapping using these sensors.