This study aimed to determine the depositional effect and improve the identification of debris flow risk zones. To accomplish this goal, we developed a two-dimensional debris flow model (Deb2D) based on a hyperbolic conservation form of the mass and the momentum balance equation with consideration of the erosion–entrainment effect as well as the depositional effect. In this model, we implemented a widely-used rheological equation—the Voellmy equation—and a quadtree adaptive grid-based shallow-water equation. This model was applied to two study sites to assess the depositional effect. The impact area, volume of soil loss, maximum velocity, inundated depth, and erosion depth resulting from the debris-flow modeling were compared with the field data. The simulation results with/without deposition were evaluated using the receiver operating characteristic method. The implementation results of the erosion–entrainment model with deposition showed superior accuracy when estimating the damage range and flow time.

The shallow landslide-generated debris flow on hillside catchments plays a critical role in the change of landscape features caused by natural hazards. When these debris flows occur in dams or reservoirs, they reduce the efficiency of facilities, and when they occur in residential areas, they cause many casualties and property damage. To minimize such damages, some methods can be performed through 1) installation of the warning system and 2) construction of check dam. However, in the case of rainfall-induced debris flow, preparation through a warning system is challenging because debris flows very rapidly. Therefore, to reduce the damage caused by debris flow events, the check dam needs to be installed, and for an efficient installment, a study on numerical modeling needs to figure out. Therefore, in this study, the Deb2D numerical model was used to analyze the mitigation effect through the check dam. This model is a two-dimensional debris flow simulation software based on quadtree-grid. The debris flow was simulated by Voellmy rheology, and the erosion, entrainment, and deposition processes that must be considered for the analysis of debris flow were simulated through the algorithm suggested in our recent study. The Raemian apartment and Galram-ri debris flow events were analyzed which occurred at Mt. Umyeon in 2011 and Gangwon-do in the Republic of Korea. In addition, a check dam was hypothetical by changing the distance from the collapse zone. The efficient location can be suggested through the simulation results.Keywords: Debris flow; Numerical model; Check dam; Mitigation effectAcknowledgmentsThis subject is supported by the Korea Ministry of Environment as “The SS projects; 2019002830001”

Landslides and debris flows occurring in reservoir watersheds can generate tsunami-like waves by inflowing into the reservoir. When such disasters occur during periods of high-water levels, such as the flood season, overtopping due to waves is inevitable. Analyzing such events is essential since they could lead to a dam break, which is particularly significant for Earth-fill dams, where overtopping alone greatly increases the risk of failure. Although several studies have attempted to address these phenomena through numerical modeling, there remains a lack of research that adequately considers the erosion and entrainment processes, which critically influence debris flow dynamics and the amplitude of debris flow-induced waves. In response, this study developed Deb2L, a two-dimensional two-layer numerical model based on shallow-water equations discretized using the finite volume method, capable of considering erosion, entrainment, and deposition processes. The performance of Deb2L was validated using theoretical and laboratory experiment results, demonstrating a quantitative accuracy with an R2 value greater than 0.85 and an RMSE of less than 0.10 m. Its applicability to field-scale events was confirmed by simulating the 2020 Sanyang Reservoir event in Icheon, South Korea. Additionally, to verify the applicability of scenario analysis, the simulation results from the landslide analysis model TiVaSS were used as input data for Deb2L, confirming the potential for coupling these models. According to the results, simulations without erosion and entrainment processes led to an underestimation of the debris flow-induced wave complex disaster.