Application of GIS-based fuzzy logic and analytical hierarchy process (AHP) to snow avalanche susceptibility mapping, North San Juan, Colorado

Snow avalanches in mountainous terrain are a significant natural disaster that affect roads, structures, and threaten human lives. Mapping of snow avalanche susceptibility has the potential to decrease these risks by modeling, mapping, and visualizing susceptible terrain using Geographic Information Systems (GIS) and remote sensing imagery. The North San Juan area in Colorado, U.S. is an ideal location for studying snow avalanches, with well documented avalanche paths that effect the area around Highway 550, Red Mountain Pass and the town of Silverton. The main goal is this study is produce avalanche susceptibility maps for starting zones or release areas of an avalanche-prone area in North San Juan, Colorado by using both fuzzy logic and analytical hierarchy process (AHP) models. In the first step, avalanche locations are identified by aerial imagery and field surveys, and a total of 70 avalanche locations are mapped from various sources. Then, the avalanche inventory is randomly split into a training dataset ≈70% (50 avalanches) for training the models and the remaining ≈30% (20 avalanches) is used for validation purpose. Six data layers, as the avalanche conditioning factors, are exploited to detect the most susceptible areas for starting zones. These terrain factors are elevation, slope, aspect, plan curvature, profile curvature, and vegetation density. Subsequently avalanche susceptibility maps are produced using fuzzy logic and AHP models. For verification, I developed, receiver operating characteristics curve (ROC). The verification results showed that the fuzzy logic model (89.8%) performed better than AHP (66.9%) model for the study area. These avalanche susceptibility maps would be useful for hazard mitigation purpose and regional planning in remote areas of the world where there is limited field data and with access it GIS and remote sensing imagery.