Landslide susceptibility zonation mapping using GIS-based frequency ratio model with multi-class spatial data-sets in the Adwa-Adigrat mountain chains, northern Ethiopia

Abstract

Different organizations reported that natural hazards related disasters killed hundreds of people and damaged infrastructures estimated millions of dollars every year worldwide. The purpose of this research was to generate landslide susceptibility zonation (LSZ) map for the first time and understand the main landside conditioning factors in the Adwa-Adigrat mountain chain (about 2540 km2). At initial stage, landslide inventory map was prepared using classical fieldwork, stereo aerial photographic analysis and interpretations (API) and Google Earth™ (GE) image interpretations. A total of 546 landslides were identified, located and delineated. Rock fall was the dominant type of landslide (170), while rock slide was the smallest (only 28 rock slides). Nearly 7.6% of the total study area is found to be affected by past and active landslides. Five spatial data-sets or landslide conditioning factors: lithology, land use and land cover, distance from streams, distance from lineament/faults, slope aspect and slope angle maps were prepared and analyzed. Frequency ratio (FR) model was used to generate LSZ map. The generated LSZ map was classified into very low, low, moderate, high and very high susceptibility zones. The validation and verification of the LSZ map was carried out using the accuracy statistics approach and the area under the curve (AUC) methods. Both methods confirmed the reliability and good fit of the FR method for the present study area. It was observed that 73.8% of the past landslides fell in the high and very high susceptibility zones. In mountain chains, field studies for landslide inventory and LSZ mapping is important. The produced LSZ map can help to predict and mitigate landslide related damages and risks, and future land use planning.