Numerical Simulation Of Tsunami In South China Sea

Tsunami is one kind of sudden and most devastating ocean disasters. The 2004 Indian Ocean tsunami and 2011 Japan Tohoku tsunami in the 21 th century have highlighted inherent vulnerabilities of coastal communities to extreme hazards, and caused worldwide attention. Tsunami is triggered by landslides, submarine earthquakes, volcanic eruptions and so on, which lead to the severe deformation of seafloor and huge wave on sea surface. When tsunami waves propagate into nearshore region, the wave amplitude increases significantly due to the decrease of water depth and bring damage to coastal area. The objective of this study is to simulate tsunami generation, propagation and effect on coastal regions by the nonlinear shallow water equations. To provide scientific basis, this study focuses on the analysis of impact of potential tsunami in South China Sea, building tsunami early warning system in South China Sea and fast forecast method for tsunami from Manila Trench.Firstly, the Okada model is used to analyze the sensitivity of earthquake source parameters, including strike angle, dip angle, slip angle, focal depth, and slip dislocation, and to evaluate the relationship with uplift and subsidence of seafloor. To describe the seafloor deformation of a real earthquake, the fault can be the linear superposition of the unit sources, or give the time of movement of every fault.The numerical model used in this study is based on the nonlinear shallow water equations(GeoClaw). The finite volume scheme, for one dimensional shallow water equations is studied, combined hyperbolic conservation law and linearization method. The adaptive refined mesh technique is advantage of this model, which can discretize part of computational domain with a finer mesh, if wave heights in the region exceed a threshold value. For transoceanic tsunami propagation of such large-scale problem, this technique can improve computational efficiency.The numerical model is firstly used to simulate dam break wave, propagation of cylindrical wave, water sloshing in a parabolic container, and Gaussian wave propagation on slope. Comparisons with analysis results and experiment measurments available show a good computational accuracy for modeling long wave propagation. The adaptive refined mesh can ensure accuracy and improve computational efficiency.The 2004 Indian Ocean tsunami and the 2011 Japan Tohoku tsunami are chosen to validate the numerical model. Comparisons with the buoy measurements show this model can reproduce accurate results. The propagation characteristics are analyzed based on numerical results. Two kinds of earthquake source parameters have been adopted for Japan tsunami, which has shown the source model with higher resolution will produce better results. It is concluded that 2011 Tohoku tsunami did not generate a significant influence on China coasts, and the maximum wave height on China Coast is smaller than 0.5 m. Finer grids are used to simulate tsunami flooding on Sendai Bay, and compare the computed results with satellite and survey data.The source parameters for normal earthquake and excessive earthquake have been summarized, according to fault distribution in Manila Trench. The propagation characteristics and influence domain of tsunami triggered by six fault plates respectively are presented, including the impacts on Hainan Island, Taiwan Island, and Lingding Bay. The worst case scenario of M_w=9.3 has been simulated and the impacts on coasts around South China Sea are analyzed, included the wave height distribution and flow field. The dispersion effects of tsunami in South China Sea have been studied with the Boussinesq model(FUNWAVE). The numerical results with and without dispersion term show that the dispersion effect of potential tsunami in South China Sea is weak.For the potential earthquake source in Manila Trench, the numerical model is used to simulate the elementary wave fields triggered by unit sources of potential tsunami source with unit slip. Using the elementary wave fields as the tsunami database, an inversion method for determining the plate slip based on multi-buoy measurements and the least squares method is adopted to compute the time series of surface elevation at the given locations around the coasts in South China Sea. Comparisons with direct computation results are used to validate the inversion method. It is found that the multi-buoy inversion method can provide better result than single buoy for extreme earthquake. Because of the expensive cost of deploying a buoy in ocean, the priority buoy is given. Based on characteristics of fault distribution in Manila Trench, a fast early warning method for potential tsunami in Manila Trench is presented, which can provide numerical results in five minutes.