Comparative multi-criteria decision-making approaches for landslide susceptibility mapping in Khagrachhari district of southeastern Bangladesh

Landslides have become common and destructive hydrogeological hazards in the southeastern hilly regions of Bangladesh, particularly in the Chittagong Hill Tracts (CHT). This is mainly due to unstable geology, intense monsoon precipitation, and unregulated human activities. Khagrachhari is recognized as a major landslide hotspot in the CHT; however, insufficient research has been conducted on spatial risk modeling and evidence-based mitigation planning. This study intends to create a detailed landslide susceptibility map of Khagrachhari through a hybrid multi-criteria decision-making framework that combines the Analytical Hierarchy Process (AHP) and the Analytical Network Process (ANP) within a GIS environment. A comparative analysis was performed using both models, integrating high-resolution spatial datasets and assessing thirteen natural and anthropogenic factors affecting slope instability. Factor weights were assigned using expert-driven pairwise comparisons (AHP) and network interdependencies (ANP), with susceptibility maps established by landslide inventory points and validated through a confusion matrix. Findings indicate that both models effectively identified high-susceptible areas, specifically in Matiranga, Mohalchari, and Guimara, with strong predictive performance demonstrated by areas under the curve (AUC). The ANP model, however, provided more continuous spatial patterns and a more accurate risk distribution by considering interrelationships among factors, offering reliable methods for spatial planning, early warning system development, and slope management in regions susceptible to landslides. Local authorities should incorporate these findings to inform zoning regulations and prioritize conservation measures in highly susceptible areas to enhance long-term resilience.