Landslide hazard cascades are a major class of geological hazards commonly occurring in high-elevation regions. These cascades typically involve a sequence of processes, including rock slope failure in glacierised terrain, transition into rock-ice avalanches, and downstream propagation into river systems, often resulting in landslide dam and flood hazards. Under the influence of ongoing climate warming, the frequency and magnitude of such cascading events keep rising in recent years.
Such hazard cascades have been widely documented in the Himalaya, such as the Yigong and Chamoli landslides, which caused extensive damage over areas spanning tens of square kilometres. Accurately modelling the full evolution of these cascades is critical for analysing impact areas and improving hazard assessment. The challenges of modelling these cascades lie in capturing key mechanisms that govern mass movement, ensuring momentum conservation across process transition, and maintaining computational efficiency.
In this study, we develop a numerical modelling framework based on a depth-averaged flow formulation, while incorporating multiple mechanical effects, including landslide motion over glacierised surfaces, entrainment, and excess pore pressure effects (Chen et al., 2025). The framework also integrates two-phase interactions and a multi-scale modelling strategy to simulate the complete cascade from slope failure to downstream flooding. The model is validated against the well-documented Yigong landslide cascade in southeast Tibet. Simulated dynamic features, including runout distance, runup height, volume amplification, and inundation area show high agreement with satellite observations and reported data. This proposed modelling framework demonstrates strong potential for application in large-scale landslide cascade simulations under changing climatic conditions.

landslide cascade,numerical modelling,process transition,depth-averaged model