Flood risk mapping of the flood-prone Rangpur division of Bangladesh using remote sensing and multi-criteria analysis

Abstract

Identification of potential flood risk areas is crucial to reduce flood damage for the frequently flooded and low-lying South Asian developing countries. The present study has prepared district-level (District: second administrative unit of the country) flood risk map for the flood-prone Rangpur Division (Division: first administrative unit) of Bangladesh using the multi-criteria decision analysis along with the application of analytical hierarchy process (AHP) method. Eight physical factors such as elevation, slope, distance from river, drainage density, land cover, rainfall, height above nearest drainage (HAND), and topographic wetness index (TWI), and six social factors such as population density, dependent population, disabled population, female population, agriculture dependent population, and literacy have been assessed to create a final risk map. The flood risk map is divided into five risk zones: very low, low, moderate, high, and very high. Integration of the social factors along with the physical factors reflects the insight of the vulnerability and increases the authenticity of the generated risk map. This study found that 62.46% area of the Rangpur Division resides under the moderate to very high-risk zone of flooding. Using ROC (receiver operating characteristic)-AUC (area under the curve) curve, the risk map is validated with a score of 0.83 from the flood inventory map of 2020 generated from the Sentinel 1 image. This risk map will guide policymakers to easily identify the vulnerable area for flood hazards and suitable areas for development activities necessary to attain sustainable development.