Applicability Of C-Band Sar Data For Characterization Of Avalanche Risk
Malnes, Eirik1; Eckersdorfer, Marcus1; Breien, Hedda2; Kristensen, Krister2; Jonsson, Magni Hreinn3; Grimsdottir, Harpa3
1Northern Research Institute (Norut), NORWAY; 2Norwegian Geotechnical Institute, NORWAY; 3Iceland Meteorological Office, ICELAND

Snow pack stability in steep terrain is determined by its seasonal development, external factors such as snow accumulation and wind drift, and internal factors such as temperature driven formation of snow layers. Of special interest is the creation of melt-freeze crusts and associated weak layers, above or beneath them. Due to fluctuating air temperatures, surface snow layers can become wet and consequently refreeze. The formed crusts when buried may act as sliding plans as well as vapor barriers, favoring the formation of faceted crystal layers. Such processes can be monitored with suitable meteorological stations at higher altitudes. However, such a network of meteorological stations is mostly lacking which makes it challenging to monitor melt-freeze cycles and their influence on avalanche activity. In this study we apply historical datasets from Envisat ASAR and Radarsat-2 C-band SAR to characterize wet snow conditions in avalanche regions in Norway and Iceland in order to investigate whether the sensor has capabilities that can be useful in operational scenarios for avalanche monitoring. C-band SAR has good capabilities for detecting wet snow, but is not sensitive to dry snow, nor snow depth. The detected wet snow maps are compared with other available information sources (meteorological data, snow process models and avalanche reports) in order to investigate whether such maps are of value for avalanche monitoring. The temporal resolution of the SAR time series is often a hard restriction, but for Norwegian and Icelandic regions both ASAR and Radarsat-2 ScanSAR has allowed quite dense time series, which allows almost daily observations of the snow cover. The spatial resolution of ScanSAR is on the order of 75-100 m, and allows sufficient detailed information simultaneously with a good spatial coverage. We have investigated several avalanche cycles for two regions in Norway and one region on Iceland in order to assess whether SAR data could have provided additional information which could be valuable in a near-real time scenario for avalanche monitoring. As a preliminary conclusion the method is promising as one of the tools for regional avalanche forecasting. Also, the long and dense time-series of ASAR and Radarsat-2 imagery over Nordic avalanche regions can be used for further studies of historical avalanches. This work has been done in the Snow, Ice and Avalanche Application project (SNAPS) which is funded EU through the Northern periphery programme. Satellite data has been acquired from ESA under AOE Category 1 project 4122 and from the Norwegian Radarsat-2 agreement.