Spatial assessments of landslide hazard vulnerability using decision support system in the Sile-Sago Watershed, Lake Chamo Rift Valley Basin, Ethiopia

Abstract

Landslide hazard (LSH) is a widespread geographical problem, threatening lives and destroying property and infrastructure globally. The focuses of this study were to identify, analyze, and prioritize LSH potential triggering factors and map the vulnerability levels in the Sile-Sago watershed. The data were obtained from satellite images like Sentinel-2, Landsat 8/9, and elevation, key informant interviews, field observation, and ground truth points. Multi-criteria decision-making, weighted, and analytical hierarchy process (AHP) were used to analyze the data in ArcGIS 10.8.2, Terrset 20, and AHP. The individual parameters were identified, analyzed, and ranked in relative importance, ranging from very low, low, moderate, high, to very high. Among the LSH conditioning variables, population density, average rainfall, drainage density, land use/cover, elevation, and vegetation density have a higher magnitude in posing LSH. In contrast, lineament density and rock types have less contribution to the LSH based on AHP, literature, and experts’ opinions. Of the total area (55,468.8 ha), 12.4 % are highly and 6.9 % are very highly vulnerable to LSH. The model performance was checked with field survey data, consistency ratio, and area under the curve (AUC) from receivers operating characteristics (ROC). The CR and ROC-AUC were 3.4 % and 89 %, which are acceptable for the analysis results. The western and higher slopes of the watershed are categorized as highly vulnerable to LSH. Concerted efforts are needed to coordinate conservation of ecosystems, integrated watershed development, sustainable land use, and population settlement planning to mitigate the risk and adverse effects of LSH in the Sile-Sage watershed.