Landslide susceptibility zonation using geospatial techniques and Analytical Hierarchy Process: A case study in Muong Lay town and its vicinity

Abstract

This study demonstrates an integrated approach of remote sensing, geographic information system (GIS), and Analytical Hierarchy Process (AHP) method to create a landslide susceptibility map for Muong Lay town and its vicinity in Northern midland and mountainous of Vietnam. Nine landslide-related factors, including petrological composition, active fault density, slope, drainage density, the difference in height per unit area, land cover, soil texture, maximum daily rainfall, and earthquake density were created using ground or remotely sensed data in a GIS environment. Weight for each factor was assigned using AHP depending on its relative importance in landslide occurrence in the study area through literature review. The landslide susceptibility map was generated using a weighted linear combination method in GIS and categorized into five susceptible classes namely, very low, low, moderate, high, and very high using quantile classification. The results revealed that 29% of the study area is at very low susceptibility, 24% at low susceptibility, 21% of moderate susceptibility, 15% of high susceptibility, and 11% of very high susceptibility area coverage. The effectiveness of these results was checked by computing the area under Receiver Operating Characteristic curve (AUC) which showed a satisfactory result of 63.3%. Most of the recorded landslide events were located in high and very high susceptibility areas. These findings could be useful to planners and decision-makers in land use planning and slope management to prevent or reduce future landslides.