GIS MULTI-CRITERIA ANALYSIS FOR LANDSLIDE RISK ASSESSMENT APPLIED TO KENYA

Assessing risk due to active landslides is a multi-criteria process that can be accomplished by using reliable historical and spatial data. Kenya is experiencing frequent landslide events that resulted in 233 casualties between 1986 and 2018 and destructed both the precious environment and the costly infrastructure. Just in 2013, the government announced that about $18 million was needed to fix the landslide damage caused in the road network. In this study, we considered several landslide causative factors, including precipitation, lithology, slope, elevation, soil, land-cover, distance to fault traces, distance to earthquake occurrences, distance to main streams, and distance to roads, for landslide vulnerability modeling. The population density, urban density, land cover, and infrastructure were considered for assessing landslide risks via weighted overlay modeling. All thematic layers were standardized, and the Analytical Hierarchical Process (AHP) was applied in order to derive the respective weights for input layers. The risk index was created using a weighted GIS – Geographic Information System – overly. The high and very high-risk classes make up 16% of the total study area, and the validation step incorporating 130 documented landslides demonstrated that about 86% of the landslide events occurred in the high and very high-risk classes. To the best of our knowledge, this is the first landslide risk index produced for Kenya, which we anticipate to help mitigate its landslide hazards and support its future developmental planning.