Future Development of Avalanche Hazard Levels in the Austrian Alps

The Alpine water cycle is profoundly reshaped by anthropogenic climate change. Shifts in spatial distribution, timing and intensity of precipitation, together with rising winter temperatures, are modifying snowfall and snowmelt regimes and alter snowpack structure. Consequently, avalanche risk and regionally assessed avalanche hazard levels are likely to undergo significant changes under ongoing climate warming. 

In this study, we investigate potential future changes in avalanche hazard levels in the Austrian Alps under the Global Warming Levels 2°C, 3°C and 4°C. Our projections are based on the latest generation of EURO-CORDEX models, analysed using a machine-learning algorithm trained on ERA5 meteorological reanalysis data. For training, we use a novel multi-year observational data set compiled from regional avalanche hazard assessments and broadcast teletext archives. Within this data set, individual mountain ranges are grouped into response units that show similar avalanche hazard responses to different weather patterns. This allows us to apply a pattern-recognition algorithm that relates circulation analogues to specific changes in avalanche hazard levels. The algorithm is first validated against long-term observations in Tyrol using the full ERA5 data record and is subsequently applied to the future climate projections.

The results of this study show when, and to what extent, the effects of human-caused climate change exceed the interannual and decadal variability of avalanche hazard over the course of the 21st century. The emergence of a climate change signal has important implications for hazard assessment and risk management in the Alpine region. It furthermore highlights critical warming thresholds beyond which current forecasting practices may require substantial revision.