AHP multi criteria analysis for landslide susceptibility mapping in the Tellian Atlas chain

Abstract

Algeria's Tellian Atlas, characterized by steep topography and complex geology, is particularly susceptible to landslides, necessitating robust hazard assessment frameworks. This study develops a landslide susceptibility map (LSM) using a GIS-based multi-criteria approach integrating statistical evaluation and expert judgment. A detailed landslide inventory, comprising 501 documented events, was established through high-resolution satellite imagery, aerial photographs, and field verification. Eleven conditioning factors were considered (slope angle, aspect, elevation, curvature, lithology, precipitation, and distances to faults, rivers, and roads, as well as stream power index (SPI) and topographic wetness index (TWI)) derived from remote sensing data, geological maps, and meteorological records. These variables were standardized and analyzed using the analytical hierarchy process (AHP), which generated relative weights through pairwise comparisons. A multicollinearity analysis was conducted to ensure statistical robustness, with factors displaying variance inflation values below critical thresholds. The resulting LSM was validated using a receiver operating characteristic (ROC) curve, yielding an area under the curve (AUC) value of 0.75, indicating good predictive performance. The findings reveal that areas with steep slopes, clay-rich lithologies, and proximity to tectonic features exhibit the highest susceptibility. The LSM offers a valuable tool for spatial planning, early warning systems, and risk mitigation strategies, particularly in the face of increasing climatic extremes that intensify landslide triggers. This integrative approach not only enhances geohazard management in mountainous terrains but also provides a replicable framework for similar environments globally, contributing to the broader goal of sustainable land use and disaster resilience.