An Approach for Mapping Frozen Soil of Agricultural Land under Snow Cover using RADARSAT-1 and RADARSAT-2

A frozen soil map is a key tool to assess environmental impacts of agricultural practices on water quality, because the frost penetration in the ground has a direct impact on runoff and nutrient losses at spring melt in Eastern Canada. SAR images data have a great potential to provide this information due to there sensibility to the soil dielectric properties. The goal of this study is to develop a classification model by analyzing interactions between the different parameters resulting from in situ field data and SAR images under snow cover and to produce frozen soil maps at watershed scale. Two issues will be tackled, the mapping of frozen soil using RADARSAT-1 images and the polarimetric scattering mechanisms generated by frozen/unfrozen soil status. In this paper we present some initial results of polarimetric radar measurements using the C-band Convair-580 SAR. The analysis is addressing polarimetric signatures and entropy-alpha space distributions.