Quantifying the relative contributions of vegetation and soil moisture conditions to polarimetric C-Band SAR response in a temperate peatland

Effective modeling of many hydrological and climatological processes requires accurate spatial characterization of soil moisture, often over large regions and across different spatial scales. Synthetic Aperture Radar (SAR) has been shown to be sensitive to surface soil moisture, and is therefore a promising alternative to field data campaigns. However, the presence of spatially-variable vegetation and surface roughness also affect SAR backscatter. In this research, empirical models were developed to both predict soil moisture from SAR and assess the relationship between LiDAR-derived vegetation and surface conditions, and polarimetric SAR parameters in a vegetated peatland environment. Importantly, the low predictive strength of soil moisture models was only evident through a process of model cross-validation (bivariate regression R2 ranged from 0.14 to 0.66 for fitted models and 0.05 to 0.41 for independently cross-validated models). The LiDAR-derived vegetation density was found to explain a large amount of variance in the SAR data, and models to predict soil moisture from SAR from only the least vegetated sites within the peatland demonstrated much higher predictive strength (R2 = 0.11 to 0.71). Soil moisture within the vegetated and least-vegetated sites was not significantly different. Therefore, non-vegetated areas may be useful as representative imaging locations for remotely monitoring surface moisture conditions in large peatland complexes with heterogeneous vegetation.