Wet snow detection by C-band SAR in avalanche forecasting

Abstract

Avalanches usually occur in steep, snow-covered mountain sides under special conditions. The stability of the snowpack depends, among other things, on the melting and freezing of the snowpack. If a snow covered area has not undergone a melt/freeze-process, it could potentially be unstable due to depth hoar or buried layers of frost, and therefore pose an avalanche risk if the area is located at a steep enough mountain side. Additionally, when the snow has been wet early in the winter, it can form a frozen crust later. This crust could constitute the sliding layer when large amounts of new snow accumulates, but the same crust could also contribute to forming unstable layers in the snow by acting as a lid and trapping water vapor from the bottom of the snowpack. These mechanisms are important for avalanche buildup, and improved information from remote sensing could be valuable for avalanche forecasting.

C-band SAR can be used to detect wet snow. Time series of maps with detected wet snow can be made by combining SAR data from periods of melting snow with reference SAR data from a period of dry snow. Since wet snow absorbs radar waves, one can detect wet snow by a scattering coefficient reduction of -3dB or more.

A pattern was found in four of the seven avalanche events studied, which might be used to predict future avalanches.