• The Journal of Engineering Geology
• /
• v.32 no.3
• /
• pp.327-338
• /
• 2022

This study aims to empirically and statistically predict soil depths across areas affected by landslides. Using soil depth measurements from a landslide area in Korea, two sets of soil depths are calculated using a Z-model based on terrain elevation and a probabilistic statistical model. Both sets of calculation results are applied to derive landslide risk using the saturated infiltration depth ratio of the soil layer. This facilitates analysis of the infiltration of rainfall into soil layers for a rainfall event. In comparison with the probabilistic statistical model, the Z-model yields soil depths that are closer to measured values in the study area. Landslide risk assessment in the study area based on soil depth predictions from the two models shows that the percentage of first-grade landslide risk assessed using soil depths from the probabilistic statistical model is 2.5 times that calculated using soil depths from the Z-model. This shows that soil depths directly affect landslide risk assessment; therefore, the acquisition and application of local soil depth data are crucial to landslide risk analysis.