Abstract
Snow avalanches are potentially dangerous phenomena in many mountainous areas worldwide. Most avalanche related research concentrates on a detailed analysis of single events at local scale. This research focuses at regional scale (i.e., about 7,400 km2) and aims to (i) delineate the static likelihood of snow avalanche occurrence for South Tyrol, (ii) to detect snow avalanches on the basis of optical remote-sensing data and to (iii) integrate approach (i) and (ii) to achieve a better performing avalanche detection model. In the context of the first activity (i), the methodological framework consists of a statistically based mapping of zones that are prone to avalanche release. The derived avalanche release likelihood model is supposed to be static in time and subsequently coupled with a random walk approach to estimate snow avalanche paths at regional scale. For the second activity (ii) optical remote sensing imagery of the Sentinel-2 Sensor are used to detect avalanches with a change detection approach. Finally, the third activity (iii) involves the combination of the (static) avalanche model (activity i) with the avalanche detection approach (activity ii). First results indicate that the proposed integration of the static statistical avalanche model with a remote sensing based change detection approach has the potential to enhance the data-driven multi-temporal mapping of avalanche occurrence at regional scale.