The University of Auckland
10_6_Buehler_et_al.pdf (5.17 MB)

Automated large scale hazard indication mapping for snow avalanches. GeoComputation 2019

Download (5.17 MB)
Version 2 2019-12-01, 22:40
Version 1 2019-09-13, 02:41
conference contribution
posted on 2019-12-01, 22:40 authored by Yves Buehler, Daniel von Rickenbach, Marc Christen, Stefan Margreth, Andreas Stoffel, Lucas Stoffel, Roderick Kuehne
Snow avalanches are a major threat for people and infrastructure in alpine regions. In Switzerland and other countries in the Alps, long term experience lead to an integral avalanche protection approach covering organizational measures, hazard maps and structural measures such as avalanche barriers or dams. However, in numerous regions around the world reliable hazard maps, combining avalanche cadaster information with terrain analysis, information on snow climatology, numerical modelling and expert knowledge are not yet elaborated because of uncomplete information and high costs. Even in Switzerland, detailed hazard maps with different danger levels only exist for selected areas where people and infrastructure are at high risk. In the canton Grisons, which is entirely covered by mountains, this is only the case for approximately 10% of the total area.

To provide reliable information on potential avalanche hazard in regions without detailed hazard maps, we develop an automated approach combining the delineation of potential avalanche release areas (PRA) with snow depth information and avalanche dynamic simulations. This approach, producing hazard indication maps over large regions for different scenarios (e.g. frequent and extreme), can be applied to any regions where high spatial resolution (2 – 10 m) digital elevation models (DEM) with an adequate quality and snow depth information are available. To assess the significance of the automated approach, we compare the result to detailed hazard maps available in the settlement area of Davos, Grisons, Switzerland where good avalanche cadastres exist. From the preliminary evaluation in collaboration with the canton Grisons, we conclude that the proposed approach has a big potential to produce large scale hazard indication maps, providing valuable first information on potential hazard. This is of particular interest in regions where no or only limited information on avalanche hazard exists, for example if new infrastructure is planned outside the settlement area.



University of Auckland