Debris Flow Susceptibility in the Jinsha River Basin, China: A Bayesian Assessment Framework Based on Geomorphodynamic Parameters

Abstract. Accurately identifying the spatial and temporal locations of debris flow occurrences is a significant challenge in assessing mountain hazard susceptibility and is essential for developing effective disaster mitigation and ecological restoration strategies. The Jinsha River basin, a typical region in China characterized by alpine gorges, frequently experiences debris flow disasters. Due to its vast area and the complex mechanisms underlying debris flow formation, using slope-based indicators alone to assess susceptibility, without considering the "source-sink" process of debris flow formation, results in low accuracy in susceptibility evaluations. To address this issue, we carefully selected a set of geomorphodynamic parameters, designed corresponding quantitative characterization methods, and developed a Bayesian model-based framework to more accurately identify debris flow-prone areas. This framework provides a comprehensive understanding of the spatial distribution and intensity of debris flow events, thereby improving the accuracy and robustness of susceptibility assessments. The model’s evaluation results indicate debris flow susceptibility in the Jinsha River basin for small, medium, and large-scale events, with an average accuracy of 63 %. Furthermore, through an empirical analysis of the catastrophic mountain flood and debris flow event ("8.21") in Jinyang County, Sichuan Province, we found that the model’s predictions closely matched the actual disaster locations, further validating the model’s effectiveness. Our study reveals that the importance of factors contributing to debris flow susceptibility in the Jinsha River basin decreases in the following order: surface material erodibility > connectivity > stream power > frequency and intensity of extreme precipitation. Debris flow-prone valleys are primarily concentrated within a 30 km stretch along the middle and lower reaches of the Jinsha and Yalong Rivers, with approximately 32,000 risk-prone river valleys longer than 200 meters, most of which are small to medium-sized gullies. The distribution of these valleys follows a power function relationship with the distance from the main stream. In areas where debris flow events occur infrequently but with high probability, when such events do occur, they tend to be larger and more destructive. Given that many existing and planned large reservoirs in the Jinsha River basin are in regions densely populated with debris flow-prone valleys, and considering the projected increase in extreme precipitation events, preventing and mitigating debris flow susceptibility remains a significant challenge. The datasets generated in this study, including river power, surface connectivity, and debris flow occurrence probability, provide valuable insights for major construction projects, such as large reservoirs, bridges, and residential developments, helping to improve infrastructure siting and disaster mitigation planning.