In this article, we are going to investigate the effects of snow, rain, temperature and wind on the number of backcountry and off-piste avalanche accidents. The data base of our survey is restricted on the western part of Austria (federal states Tyrol and Vorarlberg) within the winter periods 1987/88--2008/09. We are able to stratify the daily data for municipalities in Tyrol and Vorarlberg.
Employing spatial kriging and hurdle models, we found a positive significant effect of the snow water equivalent measurement on avalanche accident counts (if we consider the running average over the past 3 days). The variables rain and temperature 1800 meter above sea level showed negative effects on the number of accident counts. In the case of the variable wind - ERA5 global reanalysis data turned out not to be reliable -- we had a focus on the 3 avalanche accident hot spots of Austria St. Anton am Arlberg, Lech and Sölden observing wind data of the weather stations Galzig, Warth and Obergurgl. At least in the case of St. Anton and Lech, we found significant positive effects (daily velocity totals and west wind component) on the number of avalanche counts. Calculating the daily mean wind load showed a positive effect only in the case of St. Anton am Arlberg.
Finally, we tried to find conclusions in connection with `avalanche problems' such as used by several avalanche information services only finding (beside `new snow') some evidence for a `spring scenario'.