Giant river-blocking landslide dams with multiple failure sources in the Nu River and the impact on transient landscape evolution in southeastern Tibet

Abstract

Giant landslides have frequently occurred in deeply-incised valleys during the topographic development of the Tibetan Plateau, and these landslides commonly blocked rivers and in turn affected fluvial processes. The giant Qianjincun landslide, with a volume of ~110 × 106 m3, created the highest natural dam among all the completely river-blocking landslides identified along the trunk of the Nu River in the southeastern Tibetan Plateau. This giant landslide, which developed on a granite hillslope and dammed the river with a height of ~175 m, was dominated by long-term fluvial incision responding to uplift in one of the deepest valleys in the world. The dammed lake was breached around the end of the Dali glaciation (~10 ka), and the deposits remaining in the valley indicated an earlier landslide from the opposite failure source beside the Qianjincun landslide. The Geduicun landslide on the right bank evolved from gravitational deformation of toppling controlled by schist foliations, while the Qianjincun failure on the left hillslope was governed by the seismically-triggered detachment of granite cut by rock joints. The kinematic process of the Qianjincun landslide was numerically investigated under seismic loading, indicating high-velocity sliding from a thousand-meter-high source. In comparison to the soft rock failure preceded by the long-term gravitational effect, the mass movement on the granite slope was catastrophically induced by dynamics with high initial kinetic energy, and the subsequent overwhelming river blockage transiently caused stronger landscape evolution than any other landslides in the catchment. The inventory of river-damming landslides compiled along the trunk of the Nu River suggested a prominent tendency that all the dammings occurred in the region where the relief is greater than 2000 m. The analysis provided the power-law correlations of the geomorphic parameters of dams that can help predict the disastrous effects of future river blockages. The insight into landscape feedbacks between giant landslides and fluvial processes could contribute to effective risk assessments of disaster chains in high-relief reaches.