Geospatial modeling of landslide susceptibility in Debek, South Wollo, Ethiopia: comparative analysis of frequency ratio and analytical hierarchy process models for geohazards management

Introduction Landslides are a major geohazard in the northern Ethiopian highlands, causing significant damage to farmland, infrastructure, and settlements, with profound socio-economic consequences. This study aims to address the pressing need for enhanced natural hazard management by investigating landslide susceptibility in the Debek region of South Wollo, Ethiopia. Methods The study employs advanced geospatial modeling techniques to assess landslide susceptibility. Key causative factors—slope gradient, aspect, elevation, proximity to streams and springs, slope material, distance to lineaments, and land use/land cover (LULC)—were identified and analyzed through field surveys and satellite imagery. A total of 328 landslide events were documented, with data divided into training (75%) and validation (25%) sets. Landslide susceptibility maps were generated using the Frequency Ratio (FR) and Analytical Hierarchy Process (AHP) models. Validation of the models was conducted through landslide density indices (R-index) and receiver operating characteristic (ROC) curves. Results The analysis revealed that slope material and proximity to springs were the most influential factors contributing to landslide susceptibility. The FR model demonstrated a slightly better performance than the AHP model, with an ROC success rate of 0.828 and a prediction rate of 0.835, compared to 0.826 and 0.832, respectively, for the AHP model. The models were validated through the R-index and ROC curves, which showed a high degree of concordance between the predicted and observed landslide events. Discussion This study highlights the effectiveness of GIS-based geomatics approaches in landslide susceptibility mapping in a data-scarce region. The comparative analysis of the FR and AHP models demonstrates the strengths and limitations of each, offering valuable insights for landslide risk mitigation. The findings underscore the importance of integrating geospatial modeling in natural hazard management, supporting more informed land-use planning, targeted mitigation strategies, and comprehensive disaster prevention initiatives. Conclusion This research contributes to advancing the understanding of landslide dynamics in the Ethiopian highlands and provides critical resources for policymakers and stakeholders involved in disaster risk management. The study's findings enhance the capacity for effective landslide-prone area identification and susceptibility reduction, reinforcing the importance of geospatial modeling in improving natural hazard management frameworks.