Tsunami Squares: Leapfrog scheme implementation and benchmark study on wave–shore interaction of solitary waves

• Published in: Scientific reports 2024-06, Vol.14 (1), p.13053-17
• Main Authors: Zhang, Yu, Yin, Kunlong, Tang, Yang, Xiao, Lili
• Format: Article
• Language: English
• Subjects: 704/242; 704/4111; Benchmark; Eulerian–Lagrangian algorithm; Glaciers; Humanities and Social Sciences; Landslides; Mathematical models; Maximum flow; multidisciplinary; Overland flow; Predictions; Science; Science (multidisciplinary); Sensitivity analysis; Solitary wave; Solitary waves; Tsunamis; Water depth; Waves; Wave–shore reaction
• ISSN: 2045-2322; 2045-2322
• DOI: 10.1038/s41598-024-63010-1

Impulse waves are generated by rapid subaerial mass movements including landslides, avalanches and glacier break-offs, which pose a potential risk to public facilities and residents along the shore of natural lakes or engineered reservoirs. Therefore, the prediction and assessment of impulse waves are of considerable importance to practical engineering. Tsunami Squares, as a meshless numerical method based on a hybrid Eulerian–Lagrangian algorithm, have focused on the simulation of landslide-generated impulse waves. An updated numerical scheme referred to as Tsunami Squares Leapfrog, was developed which contains a new smooth function able to achieve space and time convergence tests as well as the Leapfrog time integration method enabling second-order accuracy. The updated scheme shows improved performance due to a lower wave decay rate per unit propagation distance compared to the original implementation of Tsunami Squares. A systematic benchmark testing of the updated scheme was conducted by simulating the run-up, reflection and overland flow of solitary waves along a slope for various initial wave amplitudes, water depths and slope angles. For run-up, the updated scheme shows good performance when the initial relative wave amplitude is smaller than 0.4. Otherwise, the model tends to underestimate the run-up height for mild slopes, while an overestimation is observed for steeper slopes. With respect to overland flow, the prediction error of the maximum flow height can be limited to ± 50% within a 90% confidence interval. However, the prediction of the front propagation velocity can only be controlled to ± 100% within a 90% confidence interval. Furthermore, a sensitivity analysis of the dynamic friction coefficient of water was performed and a suggested range from 0.01 to 0.1 was given for reference.