![]() ![]() Several methods allow for identifying avalanche-prone situations from these profiles. These data typically include a detailed description of the snowpack stratigraphy with vertical profiles of snow properties ( Fierz et al., 2009). Bartelt and Lehning, 2002 Vionnet et al., 2012). Coléou and Morin, 2018) and numerical simulations (e.g. Information on the snowpack evolution can be collected through field observations and measurements (e.g. Prediction of avalanche activity is mainly based on the knowledge of the snowpack evolution and of the mechanical processes leading to avalanches (e.g. Indeed, inferring the relation between avalanche activity and given weather and snow conditions is one of the essential components of operational avalanche hazard forecasting (prediction in the future based on predicted snow and weather conditions). In this work, we focus on the issue of forecasting (estimation of the outcomes of unseen data) of daily avalanche activity from simulated meteorological and snow data. Most of the countries facing such hazards rely on operational services for avalanche hazard forecasting ( LaChapelle, 1977 Morin et al., 2020) and hazard mapping ( Eckert et al., 2018). The mapping ( Keylock et al., 1999 Eckert et al., 2010 b) and forecasting ( Schweizer et al., 2020) of avalanche hazard and related risks are therefore important challenges for local authorities ( Bründl and Margreth, 2021 Eckert and Giacona, 2022). Yet, our study opens perspectives to improve modelling tools supporting operational avalanche forecasting.Īvalanches are a significant issue in mountain areas where they threaten recreationists and infrastructures ( Wilhelm et al., 2001 Stethem et al., 2003). These scores illustrate the difficulty of predicting avalanche occurrence with a high spatio-temporal resolution, even with the current data and modelling tools. ![]() However, due to the scarcity of avalanche events and the possible misclassification of non-avalanche situations in the training dataset, the predicted avalanche situations that are really observed remains low, around 3.3 %. Specifically, using mechanically based stability indices and their time derivatives in addition to simple snow and meteorological variables increases the probability of avalanche situation detection from around 65 % to 76 %. In a region of the French Alps (Haute-Maurienne) and over the period 1960–2018, we show the added value within the machine learning model of considering advanced snow cover modelling and mechanical stability indices instead of using only simple meteorological and bulk information. We develop a rigorous leave-one-out evaluation procedure including an independent evaluation set, confusion matrices and receiver operating characteristic curves. This study combines extensive snow cover and snow stability simulations with observed avalanche occurrences within a random forest approach to predict avalanche situations at a spatial resolution corresponding to elevations and aspects of avalanche paths in a given mountain range. However, it remains unclear how combining snow physics, mechanical analysis of snow profiles and observed avalanche data improves avalanche activity prediction. Several numerical and statistical methods have tried to address this issue. Predicting avalanche activity from meteorological and snow cover simulations is critical in mountainous areas to support operational forecasting. ![]()
0 Comments
Leave a Reply. |