Abstract:We considers the issue of wave breaking closure for Boussinesq type models, and attempt at providing some more understanding of the sensitivity of some closure approaches to the numerical setup. In particular, we are interested in the potential for mesh refinement, and in quantifying the dissipation mechanism active. Two closure strategies are considered. The first is an eddy viscosity approach following some early work by O. Nwogu in the 90's, and some more recent developments by Zhang and co-workers. In this model, a breaking viscosity is computed starting form a turbulent kinetic energy. The latter is obtained from an ad-hoc partial differential equation, which is solved in parallel with the propagation model. The second approach considered consists in suppressing the dispersive terms in breaking regions. The dissipation of total energy obtained in a shallow water shock is used to model the energy dissipation due to breaking. Due to its simplicity and effectiveness, the second approach has gained substantial attention in the coastal engineering community. Here we propose a systematic comparison of the two approaches, to understand more of their sensitivity w.r.t. the type of propagation model used (weakly or fully nonlinear), to the type of breaking wave being simulated, as well as to understand the mesh dependence of the pointwise results obtained, and in particular the potential for achieving mesh converged simulations. Finally, we provide a quantitative analysis of the dissipation introduced by the two closures for a moving breaking bore.