A greedy non-intrusive reduced order model for shallow water equations

In this work, we develop Non-Intrusive Reduced Order Models (NIROMs) that combine Proper Orthogonal Decomposition (POD) with a Radial Basis Function (RBF) interpolation method to construct efficient reduced order models for time-dependent problems arising in large scale environmental flow applications. The performance of the POD-RBF NIROM is compared traditional nonlinear POD (NPOD) model by evaluating accuracy and robustness test representative riverine flows. Different greedy algorithms are studied determine near-optimal distribution points RBF approximation. A new power-scaled residual (psr-greedy) algorithm proposed address some primary drawbacks existing approaches. relative performances these numerical experiments using realistic two-dimensional (2D) shallow water applications involving coastal dynamics.