Landslide Susceptibility Assessment Using Frequency Ratio: A Case Study of Kiliba (Sud-Kivu/DR Congo)

Abstract

The conversion of natural ecosystems into agricultural or urban areas can alter geomorphological processes, particularly in landslide-prone regions. Landslides in such areas can be triggered by natural events like heavy rainfall or earthquakes, as well as human activities such as deforestation and unplanned urbanization. Their impacts can be severe, resulting in significant socio-economic damage. Uvira Territory, in the western part of the East African Rift Valley, frequently experiences these events. It is located between the Ruzizi Plain to the east and the Mitumba Mountains to the west, with diverse geology comprising precambrian formations and quaternary sediments. The topography has a stepped relief with altitudes ranging from 770 to 3250 meters. The climate is tropical and humid, with a rainy season from September to May and a dry season from June to August. The area features coastal plains and mountain slopes, with many waterways flowing into Lake Tanganyika or the Ruzizi River. Detailed studies on landslide susceptibility mapping in this area are limited. This study aimed to map landslide susceptibility in the Kiliba River catchment to assist policymakers in land management. It used Google Earth images, GPS surveys, and field observations, applying a Frequency Ratio (FR) model that considered seven geo-environmental factors: slope, aspect, elevation, distance to watercourses, topographic wetness index, vegetation cover, and land use/landcover. The inventory identified 106 landslides in the study area, with densities of up to 11.25 landslides per km². Key factors in predicting landslide susceptibility were slope, elevation, and vegetation cover. The prediction model had an accuracy rate of 72.2%. The study shows that regions at medium elevation with steep slopes and low vegetation cover are mostly at risk for landslides. These findings are key for land management and disaster prevention. Future studies should consider more factors and a broader geographic range to enhance risk management