Author(s): Jizhixian Liu; Valentin Heller; Yang Wang

Linked Author(s): Jizhixian Liu, Valentin Heller, Yang Wang

Keywords: Landslide-tsunamis; Mass movement types; Numerical simulation; Rock toppling; Smoothed Particle Hydrodynamics

Abstract: Landslide-tsunamis, also called impulse waves, are generated by mass movements, such as landslides, rock falls or iceberg calving, impacting into a water body. Past cases such as the Vajont landslide-tsunami event, which caused approximately 2000 casualties in 1963, demonstrate their significant hazard potential. Landslide-tsunami generation highly depends on the mass movement type involving sliding, falling, flowing, toppling (overturning) and spreading. Nearly all landslide-tsunami studies conducted thus far involve the mass movement type sliding. They resulted in a range of empirical equations to predict the wave characteristics in function of the slide parameters (e.g. volume, density, impact velocity) and water depth for a wide range of conditions such as the slide model (e.g. rigid, granular) or water body geometry (e.g. flume, basin). However, the wave characteristics generated by mass movement types other than sliding are not well understood at present, and the available calculation methods for sliding masses are unlikely to result in accurate predictions of tsunamis generated by e.g. the toppling case. This study investigates the three mass movement types sliding, falling and toppling, for an idealized rock mass, to understand the effect of different mass movement types on landslide-tsunamis. The Discrete Element Method (DEM) coupled with Smoothed Particle Hydrodynamics (SPH), based on the open-source code DualSPHysics v5.0, is applied herein. Firstly, the simulation was calibrated and validated with recently conducted gravity-dominated fall and overturning iceberg-tsunami experiments of Heller et al. (2021) [Heller, V., Tommaso A., Fan, C., Roman, G., and Guido, W. (2021). Large-scale investigation into iceberg-tsunamis generated by various iceberg calving mechanisms. Coastal Engineering 163:103745]. The numerical set-up included a 3D basin with a water depth of 1 m and a falling or overturning mass, respectively. The particle spacing was 0.010 m, the Wendland kernel with a smoothing length radius of 0.035 m was chosen and the symplectic time integrator scheme was applied. The simulated iceberg-tsunamis generally agree well (e.g. 2% deviation for the maximum wave height HM for the toppling case) with the laboratory observations. After this calibration, the density of the block was increased to 2650 kg/m3 to simulate the sliding, falling and toppling mass movement types of rocks and to quantify their tsunamigenic potential. The value for HM generated by the toppling type is the largest, followed by the sliding and finally the falling type. The present simulations indicate that the full range of mass movement types for different mass scenarios may be investigated with the herein applied numerical model to fully understand and quantify the effect of the mass movement type on landslide-tsunamis and to support hazard assessment.

DOI: https://doi.org/10.3850/IAHR-39WC2521711920221680

Year: 2022