Comparative analysis of analytical hierarchy process (AHP) and frequency ratio (FR) models for landslide susceptibility mapping in Reshun, NW Pakistan

Abstract

The Reshun area of the Chitral District, NW Pakistan, is a part of the Hindu Kush region and is frequently subjected to cataclysmic landslides. A frequency ratio (FR) coupled with the analytical hierarchy process (AHP) has been used in conjunction with geographical information system (GIS) to make a susceptibility map of the potential landslides in Reshun. Seven landslide causing factors (i.e., slope, lithology, land cover, aspect and distance to stream, road, and fault) were selected for the susceptibility assessment. Using the AHP and FR methods, the historical landslides were linked to the causing factors to compute their importance on landslide susceptibility. A susceptibility map was extracted from each method, classified into five classes, and validated by utilizing the area under curve (AUC). Results reveal that FR with AUC ​= ​0.84 performed better than AHP with AUC ​= ​0.78 in zoning the study areas with respect to historical landslides. The FR method classified 18% of the area as having very high susceptibility to landslides, 27% as high susceptibility, 28% as medium susceptibility, 18% as low susceptibility, and 9% as very low susceptibility. These maps would support further information-based land-use planning in the region.