In consequence of the huge spatial scale, the computation load of Tsunami simulation tends to be too large to afford. In order to overcome this difficulty, a rapid numerical simulator was constructed by applying the GPGPU technology to solve the 2-dimensional shallow water equations. This simulator is based on an explicit finite volume scheme combined with unstructured triangular meshes. With the power of GPGPU, this simulator can be used to simulate large scale Tsunami movement. Furthermore, by using unstructured meshes, complex boundary could also be represented well. And, the Recursive Spectral Bisection (RSB) method was used to enable GPU computations. The simulation results for the run-up Tsunami in Tone River caused by Tohoku earthquake (2011) are verified and validated, and show that it is sufficiently accurate for real-world simulations.
Features of numerical errors of Tsunami simulation using unstructured triangular meshes with finite volume methods are investigated quantitatively. A series of test simulations of the Tsunami wave propagating in the deep area shows that the numerical diffusion errors strongly depend on the spatial size of mesh, kind of the adopted numerical scheme and the travel distance to be solved. Furthermore, introduction of an Adaptive Mesh Refinement (AMR) technique to the triangular unstructured mesh system is investigated. By conducting numerical simulation of the Tsunami propagation for 2011 Tohoku Earthquake, it is found that the AMR technique can suppress the numerical errors and reduce the required computational loads.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.