In Colorado, avalanches are a serious and persistent danger during the snowy months throughout the mountains. Millions of dollars are spent on avalanche mitigation and safety, yet Colorado suffers more avalanche fatalities than any other state. Of the money/resources spent, very little is used for backcountry forecasting.
For this project, I wanted to see if I could create an avalanche probability map for a specific location in the backcountry using publicly available data. I downloaded lidar data from the USGS for the Lone Eagle area in Grand County and converted the data into .las files in ArcGIS Pro that would allow for further analysis.
Once converted, I made a .5 meter DEM that was the main data source for the project. I researched the most important GIS data used for avalanche analysis and found slope, elevation, aspect, profile curvature, and plan curvature were the main areas to focus on. Using the lidar-based DEM, I was able to calculate the slope, aspect, elevation, and both curvatures for the area.
Using a fuzzy analysis approach for this project seemed most appropriate, given many of the data types had ranges (vs a specific weight) that could affect the final analysis. The fuzzy membership tool, specifically the linear, large, and gaussian, were used for each of the five data types. The final step was using a fuzzy overlay to aggregate all the fuzzy layers into a final dataset and determine which areas were most prone to avalanches.
Despite extensive searching, no other analysis was found for this area (or any similar area) making any comparisons of the overall data difficult. However, the final results were consistent with anecdotal information, online image searches, and personnel experience from the area. In the future, more analysis of the backcountry will be important as its usage increases.