Mapping of snow parameters with polarimetric and multifrequency SAR data in a subarctic environment

Snow plays an important role in subarctic environments and its spatial dispersion is notably linked to vegetation cover. The objective of this paper is to present an approach to estimate snow depth from SAR data using classification of vegetation types from the same SAR dataset. The polarimetric RADARSAT-2 data are used to produce a classification of the vegetation cover types and the effect of the snow cover on SAR backscattering is then studied within each vegetation class. The results show that classes with higher vegetation tends to have a deeper snow cover, but the correlation between SAR backscattering and snow depth is greater for classes with low or scattered vegetation.