An adaptive Finite Point Method (FPM) for solving shallow water problems is presented. The numerical methodology we propose, which is based on weighted-least squares approximations on clouds of points, adopts an upwind-biased discretization for dealing with the convective terms in the governing equations. The viscous and source terms are discretized in a pointwise manner and the semi-discrete equations are integrated explicitly in time by means of a multi-stage scheme. Moreover, with the aim of exploiting meshless capabilities, an adaptive h-refinement technique is coupled to the described flow solver. The success of this approach in solving typical shallow water flows is illustrated by means of several numerical examples and special emphasis is placed on the adaptive technique performance. This has been assessed by carrying out a numerical simulation of the 26th December 2004 Indian Ocean tsunami with highly encouraging results. Overall, the adaptive FPM is presented as an accurate enough, cost-effective tool for solving practical shallow water problems.

Peer Reviewed