Evaluation of spatiotemporal effects of soil depth on shallow landslides and debris flows via coupled numerical analysis

Abstract

Soil depth critically influences the timing and magnitude of shallow landslides and subsequent debris flows, yet its spatiotemporal effects remain insufficiently understood. This study investigates these effects through landslides–debris flows coupled numerical analysis, employing three soil depth configurations: the slope-based S model, the elevation-based Z model, and the U model, which assumes uniform soil depth across the study area. The 2011 Mt. Umyeon landslides in South Korea, a well-documented event with extensive field survey data, were simulated to validate the simulation results. Model performance was evaluated using the receiver operating characteristic method. Results show that increasing soil depth enlarges the scale of landslides and debris flows while delaying their onset. However, when soil depth greatly exceeds rainfall intensity and cumulative rainfall, slope failure susceptibility decreases, emphasizing the spatiotemporal control of soil thickness. In debris flow simulations, soil depth strongly influenced erosion and entrainment processes, substantially affecting downstream residential damage estimates. Furthermore, our results indicate that when soil depth data are unavailable, the slope-based S model provides more spatiotemporally stable predictions of landslide and debris flow behavior. This study highlights the importance of soil depth in geohazard modeling and advances understanding of rainfall-induced landslide–debris flow hazards in ungauged mountainous regions.