EGU23-12774, updated on 10 Jan 2024
EGU General Assembly 2023
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Next steps to a modular machine learning-based data pipeline for automated snow avalanche detection in the Austrian Alps

Kathrin Lisa Kapper1, Thomas Goelles1,2, Stefan Muckenhuber1,2, Andreas Trügler1,3,4, Jakob Abermann1, Birgit Schlager1,2, Christoph Gaisberger1, Jakob Grahn5, Eirik Malnes5, Alexander Prokop6, and Wolfgang Schöner1
Kathrin Lisa Kapper et al.
  • 1Institute of Geography and Regional Science, University of Graz, Graz, Austria (
  • 2E/E & Software, Virtual Vehicle Research GmbH, Graz, Austria
  • 3Know-Center GmbH, Graz, Austria
  • 4Institute of Interactive Systems and Data Science, Graz University of Technology, Graz, Austria
  • 5NORCE Research Institute, Bergen, Norway
  • 6SnowScan GmbH, Vienna, Austria

Snow avalanches pose a significant danger to the population and infrastructure in the Austrian Alps. Although rigorous prevention and mitigation mechanisms are in place in Austria, accidents cannot be prevented, and victims are mourned every year. A comprehensive mapping of avalanches would be desirable to support the work of local avalanche commissions to improve future avalanche predictions. In recent years, mapping of avalanches from satellite images has been proven to be a promising and fast approach to monitor the avalanche activity. The Copernicus Sentinel-1 mission provides weather independent synthetic aperture radar data, free of charge since 2014, that has been shown to be suitable for avalanche mapping in a test region in Norway. Several recent approaches of avalanche detection make use of deep learning-based algorithms to improve the detection rate compared to conventional segmentation algorithms.

          Building upon the success of these deep learning-based approaches, we are setting up a modular data pipeline to map previous avalanche cycles in Sentinel-1 imagery in the Austrian Alps. As segmentation algorithm we make use of a common U-Net approach as a baseline and compare it to mapping results from an additional algorithm that has originally been applied to an autonomous driving problem. As a first test case, the extensive labelled training dataset of around 25 000 avalanche outlines from Switzerland will be used to train the U-Net; further test cases will include the training dataset of around 3 000 avalanches in Norway and around 800 avalanches in Greenland. To obtain training data of avalanches in Austria we tested an approach by manually mapping avalanches from Sentinel-2 satellite imagery and aerial photos.

          In a new approach, we will introduce high-resolution weather data, e.g., weather station data, to the learning-based algorithm to improve the detection performance. The avalanches detected with the algorithm will be quantitatively evaluated against held-out test sets and ground-truth data where available. Detection results in Austria will additionally be validated with in situ measurements from the MOLISENS lidar system and the RIEGL VZ-6000 laser scanner. Moreover, we will assess the possibilities of learning-based approaches in the context of avalanche forecasting.

How to cite: Kapper, K. L., Goelles, T., Muckenhuber, S., Trügler, A., Abermann, J., Schlager, B., Gaisberger, C., Grahn, J., Malnes, E., Prokop, A., and Schöner, W.: Next steps to a modular machine learning-based data pipeline for automated snow avalanche detection in the Austrian Alps, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12774,, 2023.

Supplementary materials

Supplementary material file