Integrated multi-criteria decision analysis and bivariate statistical modeling for landslide susceptibility zonation in the northwestern Ethiopian highlands

Landslides represent a critical geo-environmental hazard in the Ethiopian highlands, specifically within the Telemt, Janamora, and Beyeda districts, where they frequently result in fatalities and the destruction of vital infrastructure and agricultural land. This study integrated Geographic Information Systems (GIS) and remote sensing data to perform landslide susceptibility mapping (LSM) using two distinct modeling approaches: the Frequency Ratio (FR) and the Analytical Hierarchy Process (AHP). A total of 140 data points (98 for training (70%) and 42 for testing (30%)) were considered for modeling and validation. Eleven geo-environmental conditioning factors: Slope, rainfall, lithology, density of lineaments, curvature, aspect, normalized difference vegetation index, land use/land cover, distance to drainage, drainage density, and soil were analyzed to evaluate their influence on landslide occurrence. According to the AHP model, 13.69% and 26.91% of the study area were classified as very high and high susceptibility zones, respectively. In comparison, the FR model identified 12.09% as very high and 23.78% as high susceptibility zones. Both models underscored that steep slopes, porphyritic basalt unit, high lineament density, proximity to drainage systems, and concave terrain curvature are the primary drivers of slope instability. Conversely, gentle slopes, sandstone unit, and dense vegetation cover were strongly associated with low-susceptibility regions. Model performance was validated using the Receiver Operating Characteristic (ROC) curve, yielding an Area Under the Curve (AUC) value of 0.812 for the AHP model and 0.793 for the FR model, alongside a prediction rate AUC value of 77.4% (0.774) for the AHP model and 70.4% for the FR model, confirming the predictive reliability of both methods. These results provide essential geospatial intelligence for policy-makers and stakeholders to guide disaster risk management, sustainable land-use planning, and targeted rehabilitation efforts in the region.