Bruce Jamieson’s interview with Scott Thumlert revolves around Scott’s career in avalanche research, guiding, as well as planning and engineering. Scott recounts anecdotes during his graduate studies of the stress in the snowpack under skiers and snowmobiles, and implications for tests like the extended column test. He discusses a more recent interest in improving the communication of avalanche likelihood and identifies some challenges in developing a new rating system. One of the challenges is validating the frequency of human-triggered avalanches. Specifically, when human-triggered avalanches are more likely, avalanche practitioners and winter recreationists avoid more avalanche terrain, so the number of avalanches available for validating may not increase as predicted. Scott also outlines the potential role of artificial intelligence in improving avalanche forecasting, but feels manual field work will continue to be needed. With colleagues, Scott is working on hazard mapping projects and related research into the effect of forest fires on extreme avalanche runout and hazard mapping for residential areas. As the chair of the technical committee for the 2026 International Snow Science Workshop, Scott outlines a new strategy – developed by the conference organizers - for merging theory and practice and how they plan to get more practitioners presenting case studies and applied research.  

Key Insights

• Scott’s graduate research focused on measuring the stress in the snowpack exerted by skiers and snowmobiles, highlighting the relationship between stress levels and triggering of slab avalanches.

• Effective communication of avalanche likelihood remains a challenge, with more ratings needed for lower levels of avalanche likelihood. While there is promise for improved validation of likelihood ratings, human triggering – which is very important – will remain more difficult than validating the frequency (and likelihood) of natural avalanches. 

• Artificial intelligence has the potential to enhance avalanche forecasting by improving data analysis, but Scott feels avalanche practitioners will still be required to verify the forecast. 

• The International Snow Science Workshop aims to merge theory and practice, encouraging practitioners to share real-world case studies and applied research.

Links to papers and resources mentioned in the interview:

Assessing and communicating likelihood and probability of snow avalanches. Scott Thumlert, Martin Stefan, Stian Langeland. International Snow Science Workshop 2024. Norway.

The likelihood scale in avalanche forecasting, Scott Thumlert, Grant Statham, Bruce Jamieson. The Avalanche Journal 122 (fall 2020)

Post-wildfire Analysis of Avalanche Hazard. Cam Campbell, Brian Gould and Scott Thumlert. The Avalanche Journal.

Can We Derive an Avalanche Terrain Severity Rating from Observed Terrain Selection of Professional Guides? A Proof-Of-Concept Study. Scott Thumlert and Pascal Haegeli. 2016 International Snow Science Workshop in Breckenridge, Colorado.

Terrain selection tactics in helicopter skiing – Managing avalanche risk during the 2022-23 deep slab season. Scott Thumlert, Craig McGee, Carl Trescher. 2023 International Snow Science Workshop, Bend, Oregon.

How do you stress the snowpack, Scott Thumlert. 2012 International Snow Science Workshop, Anchorage, Alaska.

How do you stress the snowpack (12 minute video)

Measuring the snowpack stressunder a falling skier (11 s video) Legacy Level:

Darren Johnson Avalanche Education Memorial Fund

AVSS

Drone Amplified's Montis

Partner Level:

CIL Avalanche

Safeback

Episode Sponsor:

Propagation Labs

Music: Ketsa

Art: Mike Tea

Production: Bob Keating, Caleb Merrill

PeakVisor Code