The upper Jinsha River basin (UJRB) located in the southeastern Qinghai-Tibet Plateau has been undergoing large-scale landslides and subsequent dam floods. However, the absence of a comprehensive inventory has impeded thorough analysis of large landslides in the UJRB. In this study, the first comprehensive inventory of large landslides in the UJRB was compiled, which includes 122 cases, distinguished from previous ones by following the entire UJRB scope, differentiating landslide feature zones, and characterizing path-dependent landslide geometric properties. Landslide geometric properties, spatial distribution, and dam flood hazard were interpreted based on this inventory. The results showed that large landslides having widening longitudinally shaped transition-deposition zones do not have significant numerical advantages compared those with narrowing shaped zones even they are frontally confined. Generally, in determining landslide spatial distribution, topographic-geomorphic factors (e.g. relief and geomorphic unit) are more important than lithologic-tectonic factors (e.g. lithologic unit and distance to fault), and coarser scale determinants (i.e. major rivers and active faults) are more decisive than finer scale determinants (i.e. minor rivers and inactive faults). Mixing up headscarp and transition-deposition zones overestimated the constraint of topographic confinement on landslide geometry, while significantly underestimated the importance of topographic slope and relief in determining landslide spatial distribution, confirming the essentiality of differentiating landslide feature zones. The empirical probability that a large landslide produces a dam lake in valleys with permanent streams in the UJRB is ~25%, and this probability increases to ~70% if landslide scale exceeds 106 m2 (1 km2). The data and results presented in this study provide elementary information for risk analysis of large landslides and subsequent dam floods in the UJRB.
Li et al. (2025) studied this question.