Landslide Susceptibility Mapping Using Multi-criteria Decision-Making Methods in Sylhet Division

Landslides, among the most destructive natural disasters, pose a severe threat, especially in the Sylhet Division, Bangladesh with its varied topography. Moulvibazar, Habiganj, and Sylhet districts are characterized by elevated terrains called ‘Tilla,’ which are highly susceptible to landslides triggered by environmental factors like rainfall, flash floods, and seismic activity. In order to assess the landslide-prone areas in the Sylhet Division, this chapter employed One statistical method and one Multi-Criteria Decision-Making (MCDM) method, namely the Frequency Ratio (FR) and Analytical Hierarchy Process (AHP). Various factors, including slope, aspect, elevation, curvature, land use/land cover (LULC), Normalized Difference Vegetation Index (NDVI), precipitation, stream density, distance from roads, population density, and soil types, were used as triggers for landslides. Among all these parameters, slope angle and precipitation were identified as the most significant factors contributing to landslide occurrences. Out of 103 landslide points gathered from various sources, 70% were utilized for model training and 30% for validation. ArcGIS V10.8 was utilized to create a map depicting the susceptibility of landslides. According to the results obtained from AHP and FR, it is evident that about 40 and 50% of the Sylhet Division’s total area is exposed to the threat of landslides. The findings suggest that AHP is the more suitable method for this study area, boasting an accuracy rate of 82%. However, the FR also exhibited robust performance, achieving an accuracy of approximately 74% in this study. These results can provide valuable insights for policymakers when making more precise decisions regarding land use and urban planning.