Integrating the Analytical Hierarchy Process (AHP) and Geographic Information System (GIS) for flood susceptibility mapping in Sunamganj district

Abstract

Flood susceptibility mapping is vital for disaster risk reduction in Bangladesh’s northeastern region, where recurrent floods severely affect livelihoods and ecosystems. The study focuses on the Sunamganj district in Bangladesh, one of the most flood-prone areas in the country because of its low-lying terrain, proximity to the Meghalaya hills and its extensive network of rivers. Despite numerous flood assessments, few studies have systematically integrated multi-criteria decision analysis with high-resolution geospatial data to produce reliable flood risk maps. Therefore, this research aims to develop a robust flood susceptibility map using an integrated Analytical Hierarchy Process (AHP) and Geographic Information System (GIS) framework. Nine flood-conditioning factors- Slope, Elevation, Topographic Wetness Index (TWI), drainage density, soil texture, rainfall, land use/land cover (LULC), and distances from rivers and roads- were assessed using Sentinel-2 imagery, SRTM DEM, BARC soil data, CHRS rainfall, and OpenStreetMap databases. Factor weights were assigned using an AHP-based pairwise comparison matrix, with a consistency ratio of 0.08, confirming the reliability of the expert judgments. Results revealed that slope (26.12%), elevation (19.63%), and TWI (15.58%) were the most influential parameters. Spatial analysis showed that 17% of the study area falls within very high and 20% within high susceptibility zones. The obtained AUC value of 0.844 indicates very good predictive performance of the model. This means the model has an 84.4% probability of correctly distinguishing between flood-prone and non-flood-prone areas. The findings provide spatially explicit insights to guide land-use planning, infrastructure design, and climate-resilient adaptation strategies. The integrated AHP-GIS approach used in this study demonstrates significant potential for regional-scale flood susceptibility assessment in monsoon-dominated haor environments.