Mapping seasonal snow with SIR-C/X-SAR in mountainous areas

The objective of this study is to examine the capability of SIR-C/X-SAR to map seasonal snow covers in alpine regions. We develop two classification trees. The first classifier can map snow and discriminate dry from wet snow, but it requires a digital elevation model (DEM) for radiometric terrain correction and reduction of the effect of local incidence angle. The second classifier can map wet snow without using a DEM. We demonstrate some processing techniques that are necessary to discriminate wet snow front smooth, bare surfaces and to discriminate dry snow from short vegetation. For validation, we compare our results with snow maps of both binary and snow-fraction classifications obtained from the Landsat Thematic Mapper (TM) at about the same time. The SAR is about 77% as accurate as the TM binary classification, but it underestimates snow cover when compared with snow-fraction classification, with the major differences occurring for mixed pixels of snow and trees. (C)Elsevier Science Inc., 1997.