Modeling Approach and Numerical Simulation of the Headcut Headward Erosion and Breach Collapse Processes in Landslide Dams

Accurate prediction of the landslide dam failure process and outburst floods is crucial for downstream disaster prevention and decision-making. However, current numerical models often overlook the impact of headcut headward erosion on the dam-break process, and the mechanisms of breach side-slope collapse and dam body scouring volume calculation remain insufficiently studied. This study, based on the principles of hydrology, hydraulics, and sediment dynamics, proposes a method for calculating headcut headward erosion velocity and modifies the existing approach for side-slope collapse calculation. A two-stage numerical model for dam-break simulation, incorporating the headcut headward erosion mechanism, was developed. To verify the accuracy of this model, the Baige landslide dam failure, which occurred on November 3, 2018, at the border of Sichuan and Tibet provinces, was analyzed as a case study. The results showed that the relative errors for peak flood flow, total flood volume, and time to peak were 1.96%, 4.04%, and 5%, respectively, meeting hydrological forecasting accuracy requirements. The side-slope collapse phenomenon predominantly occurred during the rapid development stage, with 44 collapse events identified throughout the dam-break process. The scouring volume of the dam body was 4.5767 million m3, with a relative error of 1.7%. Compared with other models, the present model better reproduces the headcut headward erosion process through the moment equilibrium principle, showing higher peak discharge values and earlier arrival times. Additionally, it effectively simulates the side-slope collapse process. This study provides a theoretical basis for rapid flood forecasting and supporting emergency decision-making for landslide dam risks.