Machine Learning Improves Debris Flow Warning

Abstract

Abstract Automatic identification of debris flow signals in continuous seismic records remains a challenge. To tackle this problem, we use machine learning, which can be applied to continuous real‐time data. We show that a machine learning model based on the random forest algorithm recognizes different stages of debris flow formation and propagation at the Illgraben torrent, Switzerland, with an accuracy exceeding 90 %. In contrast to typical debris flow detection requiring instrumentation installed in the torrent, our approach provides a significant gain in warning times of tens of minutes to hours. For real‐time data from 2020, our detector raises alarms for all 13 independently confirmed Illgraben events, giving no false alarms. We suggest that our seismic machine‐learning detector is a critical step toward the next generation of debris‐flow warning, which increases warning times using simpler instrumentation compared to existing operational systems.