International Snow Science Workshop Proceedings - Montana State University Library

Item: TOWARDS AN AUTOMATED AVALANCHE THALWEG ANALYSIS TO IDENTIFY AND COMPARE AVALANCHE TERRAIN

• • Title: TOWARDS AN AUTOMATED AVALANCHE THALWEG ANALYSIS TO IDENTIFY AND COMPARE AVALANCHE TERRAIN

    Proceedings: Proceedings, International Snow Science Workshop, Whistler, BC, Canada, 2026

    Authors:

    • Paula Spannring  [ Austrian Research Centre for Forests (BFW), Department for Natural Hazards, Innsbruck, Austria ]
    • Christoph Hesselbach  [ Austrian Research Centre for Forests (BFW), Department for Natural Hazards, Innsbruck, Austria ]
    • Andreas Huber  [ Austrian Research Centre for Forests (BFW), Department for Natural Hazards, Innsbruck, Austria ]
    • Anna Wirbel  [ Austrian Research Centre for Forests (BFW), Department for Natural Hazards, Innsbruck, Austria ]
    • Felix Oesterle  [ Austrian Research Centre for Forests (BFW), Department for Natural Hazards, Innsbruck, Austria ]
    • John Sykes  [ Simon Fraser University Avalanche Research Program, Burnaby, BC, Canada ]  [ Chugach National Forest Avalanche Center, Girdwood, AK, USA ]
    • Alberto Mariani  [ Alpsolut S.r.l., Livigno, Italy ]  [ University of Insubria, Como, Italy ]
    • Declan Knies  [ Simon Fraser University Avalanche Research Program, Burnaby, BC, Canada ]
    • Jan-Thomas Fischer  [ Austrian Research Centre for Forests (BFW), Department for Natural Hazards, Innsbruck, Austria ]

    Date: 2026-09-28

    Abstract: To reduce complexity of avalanche simulation outputs in three-dimensional terrain, individual avalanches can be represented by thalwegs. A thalweg describes avalanche properties and quantitative metrics along the main flow direction. Deriving such metrics at the regional scale requires an automated workflow. We present an automated, objective model-based approach for identifying two-dimensional thalweg representations of avalanches at the regional scale, extending an open-source model chain that delineates potential release areas and simulates their potential intensity and runout using AvaFrame::com4FlowPy. For each release area, the thalweg is computed as the flux-weighted (similar to mass-weighted) main flow direction, from which runout length and maximum impact pressure are derived. Applying this model chain with identical input parameters across four study areas, Davos (Switzerland), Livigno (Italy), Rogers Pass (Canada) and Sellrain (Austria), we compare the resulting distributions to characterize regional differences in avalanche release area, simulated runout length and impact pressure, which are translated to avalanche size. Results show that the terrain in Davos produces the smallest avalanches. 50% of release areas are assigned to dimension size 2, with a median release area of 2 600 m². For runout length and impact pressure, Rogers Pass has the largest proportion of avalanches assigned to size 5. Release area values are highest in Livigno across all percentiles, with a median of 4 200 m². Sellrain, while having the highest simulated avalanche density per unit area, shows the highest median impact pressure (56.3 kPa) among all study areas. These terrain-driven differences demonstrate the ability of the presented approach to objectively quantify and compare avalanche terrain across regions. Additionally, our results support the plausibility of the technical scheme for avalanche size classification used here, underlining its applicability for regional-scale avalanche terrain analysis.

    Object ID: ISSW2026_FS2.1.pdf

    DOI: https://doi.org/10.15788/1790098407

    Language of Article: English

    Presenter(s): Paula Spannring

    Keywords: Avalanche terrain identification, Avalanche size, Avalanche mobility, Automated model chain

    Page Number(s): 755 - 762

  • Get Full Article

Back to Home page