Browsing by Author "Mocheka, Matseliso"

Now showing 1 - 1 of 1
  • Thumbnail Image
    ItemOpen Access
    Landslide sucespibility mapping
    (National University of Lesotho, 2025) Mocheka, Matseliso; Mapeshoane, Botle
    Landslide has a serious impact on development and countries’ economy. This study presents a geospatial analysis of landslide susceptibility within the Malibamatšo Sub-Catchment, Lesotho, using remote sensing imagery and GIS-based Frequency Ratio (FR) model. The frequency ratio (FR) was used to assess susceptibility to landslides using ten conditioning factors including topography, climate and distance to streams and roads. The conditioning factors datasets were derived from various sources such as DEM, NASA and cadastral maps. The FR weighted thematic layers were then overlaid through numerical addition using the Raster Calculator to generate the Landslide Susceptibility Index (LSI) map which was then classified into 5 categories. Contributing factors of landslide in Malibamatšo Sub-Catchment based on the total frequency ratio (FR) are distance to roads (FR = 15.09), slope aspect (8.79), LULC (5.99) and rainfall (5.09). Areas located within 200 meters of roads exhibited a high landslide susceptibility, with a Frequency Ratio (FR) of up to 4.82. Northeast-facing slopes, areas with bare or agricultural land cover, and regions experiencing high rainfall distribution are generally more susceptible to landslides. Topographic factors such as slope gradient, curvature and elevation are relatively less important, however, Moderate slopes (11–22ᵒ) are more prone to landslides than steeper slopes. Soil texture is the least contributing factor (FR = 1.15). The resulting Landslide Susceptibility Map (LSM) classifies the landscape into five susceptibility zones ranging from very low to very high risk. Almost 20% of the catchment is susceptible to the occurrence of landslides. Areas most susceptible to landslides are villagesnear the national roads and associated with agricultural activities such as Ha Lejone, Ha Taunyane, Ha Sephapo, Taung, Ha Nkisi, and Liphofung. The model's accuracy was validated using the success rate curve (SRC), yielding an Area Under the Curve (AUC) value of 0.89, indicating a high predictive performance. The study’s findings support informed decision- making for land use planning, infrastructure development, and the implementation of mitigation strategies in landslide-prone areas