3.2.3. Hydrological Factors

Hydrological factors played a determinant role in affecting slope stability. For this study, the TWI and the SPI were utilized as predictors of landslides. Both the TWI and the SPI are secondary attributes of the DEM and can be calculated using Equations (1) and (2), respectively. The TWI measures the topographic control on groundwater flow and accumulation. Terrains with a higher TWI values are more likely to become saturated during rainfall events. The SPI quantifies the erosion power of flowing water. Gullies are more likely to form at locations with high SPI values. As indicated in Figure 3f,g, the TWI of the study area ranged from 6.80 to 34.72, and the SPI ranged from −2.41 to 25.13. The factors of TWI and SPI were finally divided into five categories through the "natural breaks" function in GIS.

$$\text{TWI} = \ln\left(\frac{A}{\tan\beta}\right) \tag{1}$$

$$\text{SPI} = \ln(A \cdot \tan \beta) \tag{2}$$

where *A* is the unit upstream accumulation area, β is the slope gradient, and ln is the natural logarithm.

## 3.2.4. Environmental Factors

Given the poor data availability in Kenya, three environmental LCFs were considered in this study to produce LS maps. These two factors were soil texture and land use. Soil texture indicates the proportional composition of sand, silt, and clay content in the soil. Because high clay soils usually contain high organic matter content, which is favorable for soil resistance against detachment, soils characterized by high sand or loam content are more susceptible to land sliding than clayey soils [32]. A nationwide soil property GIS database developed by the International Soil Reference and Information Centre (ISRIC) [33] was utilized in this study. Types of soil texture were classified as follows: (1) very clayed, (2) clayed, (3) loamy, and (4) sandy (Figure 3h). Land use and landform type contribute significantly to slope stability. Specifically, vegetation roots may enhance slope stability by altering the cohesive forces and hydrologic properties. The degradation of forests and vegetation increases the degree of susceptibility of the area to landslides [34]. As for the factor of landform, steep and hill/mountain terrains are prone to landslides compared with flat terrains such as plains, valley floor and foot slope. In addition, from the perspective of geomorphology, landslide itself also plays as a driving role in landform evolution [35,36]. The Kenya National Land Use Dataset (KNLD) [37] was utilized in this study to obtain the LCFs of land use and landform type. As indicated in Figure 3i, the KNLD contains ten land use types (Figure 3i) and eleven landform types (Figure 3j).
