Soil moisture (SM) plays a crucial role in hydrological processes and slope stability. Synthetic aperture radar (SAR) enables all-weather, physically based SM retrievals at spatial and temporal scales relevant to landslide monitoring. This study evaluates multi-frequency, multi-resolution, and multi-depth SM retrievals using SAOCOM (L-band), Sentinel-1 (C-band), and COSMO-SkyMed (X-band) SAR data acquired over the Petacciato landslide, Italy (March–September 2025). The first-order radiative transfer model (RT1) was applied at 50 m and 300 m resolutions and validated against in situ SM measurements from three monitoring stations, with additional observations at 10 cm and 20 cm depths available at one station. Bayesian fusion was implemented to assess the benefits of multi-frequency integration. Results indicate that SAOCOM L-band provided the most reliable retrievals (Formula: see text–Formula: see text at 50 m; Formula: see text–Formula: see text at 300 m), maintaining consistency across depths. Sentinel-1 showed moderate performance (Formula: see text), while X-band retrievals were largely unreliable due to surface roughness and canopy effects. Bayesian L + C fusion improved correlations (Formula: see text), whereas inclusion of X-band reduced accuracy. Overall, L-band SAR demonstrated superior capability for soil-moisture estimation in vegetated, landslide-prone terrain, with coarser resolutions yielding more stable retrievals. These results provide a methodological basis for incorporating multi-frequency SAR-derived soil-moisture dynamics into landslide hazard assessment frameworks.
Rana et al. (Wed,) studied this question.