Avalanche Hazard Indication Mapping for Forested Terrain in the Selkirk Mountains, British Columbia, Canada

Abstract:

While forest cover plays a significant role in the potential for avalanche release in mountainous terrain, existing hazard indication mapping methods do not account for the effects of forest characteristics in mediating avalanche release or runout. However, having an in-depth understanding of the relationship between terrain and avalanche hazard below treeline is critical in the mountains of western Canada as substantial amounts of backcountry winter recreation occur below treeline, and avalanches starting within forested terrain have the potential to destroy valuable timber resources.

This research aims to extend the functionality of a hazard indication mapping method developed at WSL Institute for Snow and Avalanche Research SLF to forested terrain by incorporating forest characteristics in probable release area (PRA) models and dynamic avalanche simulations. The focus of our research is the tenure area of a mechanized skiing operation in the Selkirk Mountains of British Columbia where we have combined SPOT 6 stereo satellite imagery and information from the British Columbia Vegetation Resource Inventory (incl. crown cover, stem density) to develop a digital surface model and forest classification layers (5m resolution). Outlines of known startzones drawn by experienced local mountain guides above and below treeline were then used to validate the existing PRA model in the alpine and identify relevant terrain characteristics for forested terrain using a grid search method to test and optimize input parameter values. Runout extents reported by the local guides were used to validate dynamic avalanche simulations and test whether changes to simulation friction parameters would improve accuracy in forested terrain.

The overall goal of this research is to produce a detailed terrain characterization as the foundation for an analysis of decision-making in professional ski guides. However, hazard indication maps specifically targeted at forested terrain can also provide a valuable tool for identifying disturbances in forest ecosystems and mapping hazards to personnel and infrastructure. Since high-quality geospatial data is relatively sparse in the vast mountains of Canada, our objective was to develop a processing pipeline that only relies on satellite imagery and can easily be applied in other areas.