Sensitivity of Sentinel-1 observations to snow properties, comparing radar backscatter, polarimetric decomposition parameters, and interferometric phase changes

Remote sensing can offer important information on snow properties at the global scale. However, the sensitivity of satellite radar measurements at C-band (5.4 GHz) from the ESA and Copernicus Sentinel-1 (S1) mission to snow properties still requires further investigation. This study provides the first results of a detailed sensitivity analysis, carried out over the European Alps at 1-km spatial resolution. It includes three processing types of radar measurements: radar backscatter observations in vertical-vertical (VV) and vertical-horizontal (VH) polarizations, polarimetric decomposition parameters (e.g., H-Alpha dual polarization decomposition), and interferometric phase change and coherence between successive S1 acquisitions from the same relative orbit. The sensitivity of the different radar measurements is investigated with respect to snow properties (snow depth, snow water equivalent, wet-dry snow state), soil properties (surface soil moisture, soil temperature), and vegetation properties (LAI), and furthermore stratified by snow climatology, land cover, and elevation. Preliminary results suggest that, in regions with significant snowfall and limited vegetation, the VH backscatter correlates strongest with snow depth and SWE, whereas the VV backscatter is more strongly correlated with soil properties. The Alpha polarimetric decomposition parameter increases with snow accumulation, indicating increased contributions of volume scattering and multiple scattering. The often low interferometric coherence is confounding the interpretation of interferometric phase changes in mountainous regions. In conclusion, the first results of this sensitivity study indicate the usefulness of S1 radar backscatter and polarimetric decomposition parameters for snow retrieval algorithm development.