Improving snow avalanche forecasting by data-driven automated predictions
The overall goal of the proposed research is to improve avalanche forecasting by developing a decision support tool that provides data-driven automated predictions of avalanche hazard. We hypothesise that by applying modern data science and machine learning methods on the diverse (in time and space) snow and avalanche data, snow avalanche hazard can automatically be forecast – with at least the accuracy of present experience-based forecasts.
Operational avalanche forecasting – issuing warnings to the general public – is still by and large an experienced-based process. The lack of appropriate numerical or statistical methods has hence prevented (1) knowledge extraction required for a sustainable operation, and (2) numerical forecasting i.e. data-driven decision support crucially important for consistent and objective forecasts.
Develop statistical learning techniques for support decision making in snow and avalanche disciplines
Develop data-driven probabilistic methods from multi-year snow and avalanche datasets to support operational avalanche forecasting
Assist avalanche forecasters and operational decision support tools with:
Predictive methods of avalanche danger level (a score from 1 to 5)
Predictive models of avalanche danger type (the type of likely avalanches)
A better understanding of the spatio-temporal relationships between measurement stations, and potentially improve the monitoring network
M. Hendrick, C. Pérez-Guillén, A. van Herwijnen and J. Schweizer (2020). Machine learning as a tool for avalanche forecasting. 22nd EGU General Assembly, held online 4-8 May, 2020, id.13419, Machine learning as a tool for avalanche forecasting , 2020
C. Pérez-Guillén, M. Hendrick, F. Techel, A. van Herwijnen, M. Volpi, O. Tasko, F. Pérez-Cruz, G. Obozinski, and J. Schweizer (2021). Data-driven automatic predictions of avalanche danger in Switzerland, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-6154, CO Meeting Organizer EGU21 , 2021