A two-dimensional (depth-averaged) finite volume Godunov-type shallow water model developed for flow over complex topography is presented. The model, SToRM, is based on an unstructured cell-centered finite volume
formulation and on nonlinear strong stability preserving Runge-Kutta time stepping schemes. The numerical discretization is founded on the classical and well established shallow water equations in hyperbolic conservative
form, but the convective fluxes are calculated using auto-switching Riemann and diffusive numerical fluxes. Computational efficiency is achieved through a parallel implementation based on the OpenMP standard and the
Fortran programming language. SToRM’s implementation within a graphical user interface is discussed. Field application of SToRM is illustrated by utilizing it to estimate peak flow discharges in a flooding event of the St. Vrain Creek in Colorado, U.S.A., in 2013, which reached 850 m3/s (~30,000 f3
/s) at the location of this study.