GIS-Based Flood Vulnerability Assessment in River Basin: A Predictive Modelling Approach

Abstract

Floods remain one of the most devastating natural disasters, particularly in river basins where hydrological and topographic characteristics contribute to varying levels of vulnerability. This study presents a GIS-based flood vulnerability assessment model integrated with predictive modeling techniques to evaluate spatial and temporal flood risks in a selected river basin. Leveraging geospatial datasets—including topography, land use, soil types, rainfall, drainage density, and population data—the model applies multi-criteria decision analysis (MCDA) and machine learning algorithms such as Random Forest (RF) and Support Vector Machine (SVM) to identify vulnerable zones. The vulnerability maps produced were validated against historical flood records to ensure reliability. Results show a strong spatial correlation between flood-prone areas and low-lying zones with high anthropogenic pressures. The findings provide crucial insights for regional planning, early warning systems, and disaster risk mitigation strategies. The study emphasizes the importance of integrating GIS tools with data-driven predictive models to enhance flood vulnerability mapping and decision-making at the river basin scale.