Avalanche Susceptibility mapping of District Shigar, Pakistan using GIS-based MCDA–AHP modeling

Abstract In snow-covered mountainous terrain like the Himalayas, avalanches are responsible for the majority of fatalities and property damage. Statistics and information on its possibility are typically poor and unavailable in complex terrains. Recognizing avalanche-prone locations and making avalanche susceptibility maps available can assist decision-makers and planners in developing successful avalanche risk reduction strategies. The goal of this research is to use a multi-criteria decision analysis–analytical hierarchy process model in a GIS to create an avalanche susceptibility map for the Shigar valley region, North Pakistan. Slope, aspect, curvature, elevation, terrain roughness, and ground cover are the most important avalanche occurrence components in this model. The analyzed factors were created using STRM DEM and Landsat 8 satellite data. The correctness of the findings is determined by looking at the avalanche inventory map of recorded avalanche sites. The area under the ROC curve (ROC-AUC) method was used to calculate the prediction accuracy. Using ROC-AUC, the validation findings have a prediction accuracy of 91%. Decision-makers on a local and regional level can apply the present study approach in the future to improve avalanche preparedness and planning.