Item: Regional stability evaluation with modelled snow cover data
Title: Regional stability evaluation with modelled snow cover data
Proceedings: International Snow Science Workshop, Davos 2009, Proceedings
Authors: Michael Schirmer, Jürg Schweizer, Michael Lehning WSL Institute for Snow and Avalanche Research SLF, Davos, Switzerland
Abstract: Stability (or the probability of avalanche occurrence or release) is probably the key factor defining the avalanche danger level. Most snow stability evaluations are based on field measurements which are time-consuming and sometimes dangerous. Numerical modelling of snow cover stratigraphy and stability offers a solution to the problem of having only sparse information about the regional stability. We compared numerical model output with observed stability. Over 700 snow profiles combined with rutschblock score and release type in the surroundings of five weather stations were rated into three stability classes. Snow stratigraphy data were produced for the locations of these five weather stations using the snow cover model SNOWPACK. We determined whether (i) existing physically based stability interpretations implemented in SNOWPACK are applicable for regional stability evaluation; (ii) modelled variables equivalent to those, which were identified as significant for real snow covers, contain similar classification power; (iii) additional modelled variables, which cannot be measured in the field perform better. Finally, with objective feature selection a set of variables was chosen to obtain an optimal link between the modelled snow stratigraphy data and the stability rating using classification trees. Cross-validation was used to assess the quality of the classification trees. Another aspect of goodness, the consistency with experts’ judgement, was considered. The interpretation derived can be further developed into a support tool for avalanche warning services to predict the regional avalanche danger.
Keywords: avalanche forecasting, snow stability, avalanche danger, snow cover, numerical modelling
Digital Abstract Not Available