Influence of surface and vegetation characteristics on C-band radar measurements for soil moisture content

Soil moisture estimation using microwave remote sensing faces challenges of the segregation of influences mainly from roughness and vegetation. Under static surface conditions, it was found that Radarsat C-band SAR shows reasonably good correlation and sensitivity with changing soil moisture. Dynamic surface and vegetation conditions are supposed to result in a substantial reduction in radar sensitivity to soil moisture. A C-band scatterometer system (5.2 GHz) with a multi-polarization and multi-angular configuration was used 12 times to sense the soil moisture over a tall vegetated grass field. A score of vegetation and soil parameters were recorded on every occasion of the experiment. Three radar backscattering models Viz., Integral Equation Model (IEM), an empirical model and a volume scattering model, have been used to predict the backscattering phenomena. The volume scattering model, using the Distorted Born Approximation, is found to predict the backscattering phenomena reasonably well. But the surface scattering models are expectedly found to be inadequate for the purpose. The temporal variation of soil moisture does show good empirical relationship with the observed radar backscattering. But as the vegetation biomass increases, the radar shows higher sensitivity to the vegetation parameters compared to surface characteristics. A sensitivity analysis of the volume scattering model for all the parameters also reveals that the radar is more sensitive to plant parameters under high biomass conditions, particularly vegetation water content, but the sensitivity to surface characteristics, particularly to soil moisture, is also appreciable.