Accurate quantification of landslide dam breaching processes is critical for outburst flood risk assessment. Flow-driven erosion and its cascading effects dominate the geomorphic evolution of landslide dams. However, most existing models either rely on predefined geometric patterns while only considering erosion processes, or use simplistic formulations that decouple geometric updates from erosion-induced dynamic breach evolution. To overcome these limitations, this study developed the Dam Breaching Analysis Model (DBAM), which resolves the bidirectional coupling between erosion processes and breach morphological evolution. We proposed a dynamic partitioning criterion for overtopping and lateral erosion based on in-channel water depth-to-breach depth ratio and breach width-to-depth ratio. Integrated with bank slope instability analysis, the model achieves adaptive updates of both cross-sectional and longitudinal breach profiles without predefined patterns. The model is validated against three benchmark cases: a controlled laboratory experiment, the 2018 Sedongpu dam breach along the Yarlung Tsangpo River, and the 2018 Baige outburst flood along the Jinsha River. Simulations closely match measured hydrographs, confirming the model's ability to capture erosion dynamics and breach evolution. The proposed model provides a robust tool for natural landslide dam breaching risk assessment.

Dam breaching,Landslides dam,Overtopping erosion,bank erosion