A comparative modeling of landslides susceptibility at a meso-scale using frequency ratio and analytic hierarchy process models in geographic information system: the case of African Alpine Mountains (Rif, Morocco)

Landslides represent a major natural hazard for all countries in the world. The Rif mountains in Morocco suffer from different types of landslides. Some of them are very active and present a significant risk to urban areas and transport systems. Consequently, in terms of sustainable development, landslide susceptibility mapping is essential to assess the levels of danger posed by these phenomena. This study aims at evaluating landslide susceptibility using two different approaches based on a statistical method (Frequency Ratio, FR) and on a heuristic method (Analytic Hierarchy Process, AHP). The second purpose is to compare them to select the most relevant and reproducible one with a view to applying it to areas having a similar geomorphological context. This study includes a precise inventory map representing the spatial distribution of three landslide categories within 892 sites. Rock falls, flows and landslides were studied using field survey and satellite imagery. Nine thematic layers of predisposing factors controlling landslides occurrence were prepared. The final result is presented in the form of six susceptibility maps of rockfalls, flows and landslides for FR and AHP models. The result of the success rates (AUC) indicates that the FR method is better with an AUC of 88% for rock fall, 89% for flows and 87% for landslides, while the AUC is 83%, 84% and 76%, respectively for the AHP method. Moreover, the results indicate which method to use for similar regions to produce indicative mapping and help users select priority areas prone to landslides.