GIS-based landslide susceptibility mapping using Frequency ratio method: A case study from Adigrat-Mugulat mountain chains, northern Ethiopia

• Landslide inventory map was prepared from field survey, aerial photos and Google Earth. • GIS-based Frequency ratio method was used to produce landslide susceptibility map. • A total of 283 landslides were inventoried in the study area. • The area under curve (AUC) used to validate the final map. Landslides are common geo-hazards in the mountainous regions of Ethiopian highlands. This study aims to assess the contributions of major causative factors of landslides and produce landslide susceptibility zonation (LSZ) map of the study area using frequency ratio (FR) method. In total, 283 landslides were identified using field campaigns, aerial photos, and Google Earth. Seven causative factors, namely slope steepness, aspect, elevation, lithology, distance to drainage, land use land cover, and distance to lineament, were used for landslide susceptibility mapping. 12.7% of the study area is affected by different types of landslides. The study revealed that slope steepness, lithology, and distance to drainage have significant effect on landslide occurrences. Specifically, slope angles (>35°), distance to drainage (0 to 150 m), land use land cover (bare land), distance to lineament (0 to 400 m), and elevation (>3030 m) are prone to landslides. Moreover, slope aspects with southwest, northwest, and west orientations are more susceptible to landslides. The findings showed that 11.7% of the study area falls under low susceptibility zone, 47.68% under moderate zone, while high and very high susceptibility zones cover 30.83% and 9.79% of the area, respectively. The validation of landslide susceptibility map reveals that 84% of the inventoried landslides fall in high and very high susceptibility zones, while 14% fall in moderate zone and only 2% in low susceptibility zone. Furthermore, verification using the area under the ROC curve resulted in 87% prediction accuracy. Therefore, the generated landslide susceptibility map could serve as a baseline for future infrastructure planning and settlement.