Landslide Hazard Zonation Using Geospatial and Multi Criteria Decision Analysis Techniques in Gechi District, Western Ethiopia

Abstract

Landslides are considered as the main cause for mass deterioration and distraction of topographic features. It is an ongoing problem in every rainy season where the mass of soil is detached from its original location and moves down the hills thereby causing loss of life and public properties. Therefore, this study aimed to produce landslide hazard susceptibility map based on certain environmental parameters like Elevation, slope, aspect, curvature, lithology, soil texture, land use Land cover, distance to lineament, distance to drainage, distance to road and rainfall by using Geospatial and Multicriteria Decision Analysis at Gechi district, Oromia regional state, western Ethiopia. The weights of selected susceptibility factors were performed using pair wise comparison matrix of analytical hierarchical process (AHP) method. The sum total of all parameters was weighted using Weighted Linear Combination (WLC) to prepare the Landslide Hazard Zonation (LHZ) map. In this study probability and non-probability sampling techniques were employed to collect ground truth data which identifies about the existing landslide and overlaid with the prepared landslide hazard zonation map for validation. The study clearly reveals that about 18.4 km2 (1.3%) remains very high hazard zones; whereas about 299.6km2 (20.5%) was falls in very low hazard zone. Consequently, from the total number of 29 landslide inventory points 82.7% of past landslide events were falls in maximum landslide hazard zone and only 17.3% of landslide events were falls in low hazard zone. So, this shows satisfactory agreement the rationality of considered parameters, the adopted MCDA technique, tools and procedures with the prepared landslide hazard zonation map. Therefore, it is recommended that local governments must develop appropriate plans to reduce landslide effects through land use and land cover policies and regulations.