Prediction of movement trajectories and surge disaster-affected areas following instability and failure of a large, high-positioned, loess‒mudstone interface landslide

Instability and failure of landslides near reservoirs often trigger catastrophic landslide-surge disaster chains. However, existing integrated numerical simulation methods struggle to achieve rapid simulation of the entire landslide-surge propagation process. Although the Pudasaini two-phase flow model can simulate the entire process of landslide-debris flow disaster chains, its applicability and predictive accuracy for potential landslide-surge disaster chains require further validation. This study takes the Likan Highway landslide group on the steep slope of the Lijiaxia Reservoir in the upper Yellow River, as a case study. The r.avaflow software was used to simulate the entire process of a potential landslide-surge disaster chain and the two-dimensional motion of the landslide in the reservoir. The results show that ‘Slide block I’ produces a wider affected area but a lower wave heights (11.16 m), whereas ‘Slide block II’ triggers a more concentrated surge with a higher maximum surge height (21.39 m). This indicates that the surge response of high-position reservoir-bank landslides is not determined solely by landslide volume; rather, it is jointly influenced by the movement path, entry location, topographic, and the duration and kinetic energy of landslide entering the reservoir. These factors subsequently control surge propagation and dam-overtopping height. Indirect validation of the surge heights supports simulation credibility and first demonstrates the Pudasaini two-phase flow model’s reliability for rapid assessment of potential landslide-surge disasters. This represents a significant advancement in the integrated numerical simulation method for the rapid evaluation of the entire process of potential landslide-surge disaster chains.