We compile citations and summaries of about 400 new articles every week.
Email Signup | RSS Feed

HELP: Tutorials | FAQ
CONTACT US: Contact info

Search Results

Journal Article


Bellaire S, van Herwijnen A, Mitterer C, Schweizer J. Cold Reg. Sci. Technol. 2017; 144: 28-38.


(Copyright © 2017, Elsevier Publishing)






Wet-snow avalanches are relatively poorly understood and difficult to forecast. By definition, water is required in the snow cover, thus assessing the liquid water content of the snow cover is of paramount importance for wet-snow avalanche forecasting. While evaluating wet-snow instability through field measurements is difficult, physically based snow cover models can be used to estimate the amount of liquid water within the snow cover using meteorological input. Recently, an index based on the liquid water content of the snow cover was suggested showing high potential to predict the onset of wet-snow avalanche activity. However, as the snow cover model was forced with data from automated weather stations (AWS) only a now-cast was possible. As snow cover conditions quickly change during snow melt, a forecast would be useful. For this study, we therefore force the 1-D physically based snow cover model SNOWPACK with data from the high-resolution numerical weather prediction model COSMO and investigate whether forecasting regional patterns of the onset of wet-snow avalanche activity is feasible. To validate the index, we compared simulations performed at the location of numerous AWS in the Swiss Alps with wet-snow avalanche observations from the corresponding region. Only by forcing SNOWPACK with data from automated weather stations up to the actual day and then adding the forecasted input data to produce a forecast led to results comparable to the simulations with station data only. While using this setup, the index was able to predict the onset of wet-snow avalanching with a probability of detection of >80% for three winters between 2013 and 2016 and for two different climate regions in Switzerland. However, the false alarm ratio was high (up to 80%), suggesting that further refinements of the classification method are needed.

Language: en


Avalanche forecasting; Numerical weather prediction; Snow cover modelling; Wet-snow avalanche


All SafetyLit records are available for automatic download to Zotero & Mendeley