“…The use of ML and DL has replaced other approaches in a variety of fields related to cryoshperic and natural hazard science, e.g., for the study of snow and glacial features (e.g., Haq et al, 2021b;a), snow cover mapping (e.g., Nijhawan et al, 2019), or to detect wet and dry snow (e.g., Tsai et al, 2019). Other approaches used ML to produce avalanche hazard maps that can be used for the prediction of future avalanche events (e.g., Rahmati et al, 2019) or to model snow wetness and snow density using artificial neural networks (e.g., Haq et al, 2019). Several scientific studies have demonstrated the potential of (semi-)automated avalanche detection from SAR (e.g., Vickers et al, 2016;2017;Wesselink et al, 2017;Abermann et al, 2019;Eckerstorfer et al, 2019;Leinss et al, 2020).…”