Item: BACKCOUNTRY TERRAIN USE PATTERNS UNDER VARYING AVALANCHE FORECAST CONDITIONS IN THE CANADIAN ROCKIES: INSIGHTS FROM STRAVA ACTIVITY DATA
-
-
Title: BACKCOUNTRY TERRAIN USE PATTERNS UNDER VARYING AVALANCHE FORECAST CONDITIONS IN THE CANADIAN ROCKIES: INSIGHTS FROM STRAVA ACTIVITY DATA
Proceedings: Proceedings, International Snow Science Workshop, Whistler, BC, Canada, 2026
Authors:
- Marc Pons [ Computational Ecology Group, Canmore, Alberta, Canada ] [ Coldwater Laboratory, Centre for Hydrology, University of Saskatchewan, Canada ]
- Colin Bergeron [ Mount Revelstoke and Glacier National Parks, Parks Canada, Revelstoke, BC, Canada ]
- Julie Heinrichs [ Computational Ecology Group, Canmore, Alberta, Canada ]
Date: 2026-09-28
Abstract: Understanding how backcountry recreationists adjust their terrain choices in response to avalanche hazard remains a key gap in avalanche risk research. Although prior studies have inferred skier behaviour from GPS tracks, online engagement, or mobile phone records, few combine spatially explicit recreation data with avalanche terrain classification and detailed forecast information, including both danger ratings and avalanche problem types. We examine backcountry activity across two case study areas in the Canadian Rockies using an integrated methodology combining Strava activity data, Avalanche Canada forecasts (danger rating and problem type), and Avalanche Terrain Exposure Scale (ATES) maps to quantify how use is distributed across terrain classes under varying hazard conditions. Results show that temporal factors—especially weekends and seasonal progression—are the primary drivers of overall activity, consistent with findings from the Swiss Alps. However, a critical discrepancy emerged between Strava heatmap data, which captures broader user bases across all terrain types and Strava Metro data. While Mount Revelstoke (predominantly road-accessible terrain) shows reliable terrain-danger relationships with ~174,000 recorded trips, Rogers Pass and Glacier National Park (high-exposure backcountry terrain) yielded sparse Metro activity despite heatmap evidence of substantial recreational use. This indicates that terrain selection patterns can be reliably assessed for road-accessible recreation but not for avalanche-exposed backcountry populations using Metro filtering alone. We assess both the potential and critical limitations of Strava as a scalable, privacy-respecting proxy for backcountry use in mountain areas. Despite systematic biases toward accessible terrain and younger, technology-adopting user populations, our analysis demonstrates that evaluation of user behavior relative to temporal avalanche conditions is achievable using crowdsourced activity data. Strava heatmap analysis confirmed that activity is distributed across terrain classes at landscape scales; however, standard Metro data downloads do not provide the temporal granularity required for daily hazard-behavior analysis, particularly in remote backcountry terrain. To overcome this critical limitation, we recommend an API-based workflow for acquiring segment-level activity counts at daily and monthly resolution, combined with heatmap validation and local expert consultation. This integrated approach enables characterization of how diverse user populations respond to avalanche conditions in both road-accessible and high-exposure terrain, supporting robust, scalable monitoring of recreation-hazard dynamics when data limitations are properly acknowledged and addressed. We propose an integrated API-based methodology combining daily/monthly Strava segment data, heatmap validation, and terrain analysis for future studies of backcountry user response to avalanche conditions. Such frameworks can support efforts to incorporate spatially explicit recreation dynamics into avalanche communication frameworks and wildlife–recreation interaction models while maintaining awareness of population-specific data biases.
Object ID: ISSW2026_FS2.6.pdf
DOI: https://doi.org/10.15788/1790098422
Language of Article: English
Presenter(s): Marc Pons
Keywords: avalanche terrain, terrain selection, backcountry users, avalanche problems, mountain parks, crowdsourced recreation data, user-generated content
Page Number(s): 2152 - 2158
-