Development and Utilization of Regional Oceanic Modeling System (ROMS). Delicacy, Imprecision, and Uncertainty of Oceanic Simulations: An Investigation with the Regional Oceanic Modeling System (ROMS). Mixing in the Ocean Surface Layer Using the Regional Oceanic Modeling System (ROMS).

Our core objectives are code improvements and oceanographic simulation studies with the Regional Oceanic Modeling System ROMS. The targeted problems are submesoscale wakes, fronts, and eddies nearshore currents internal tides regional and Pacific eddy-resolving circulations and their low-frequency variability mesoscale ocean-atmosphere coupling and planetary boundary layers with surface gravity waves. To address these problems we are making ROMS more of a multi-process, multi-purpose, multi-scale model by including the coupling of the core circulation dynamics to surface gravity waves sediment resuspension and transport biogeochemistry and ecosystems non-hydrostatic large-eddy simulation and mesoscale atmospheric circulation, and by providing a framework for data-assimilation analyses led by others. Our major algorithmic objectives are cross-scale grid-embedding in turbulent flows improved accuracy in the Boussinesq approximation with a realistic Equation of State EOS accurate advection dynamically adaptive, vertical coordinates surface-wave-averaged vortex force and Lagrangian transport and parameterization of wave-breaking and other mixing effects. Finally, we continue to further improve the pre- and post-processing tools and on-line documentation for ROMS. A parallel objective is to establish the characteristics of model uncertainty in ROMS for realistic simulation of complex flows, as an intrinsic model contribution to analysis and to forecast errors. The premise is that defensible alternative model designs -- in parameter values, subgrid-scale parameterizations, resolution, algorithms, topography, and forcing data -- may often provide a range of answers comparable to the model-measurement discrepancies, although as yet this kind of sensitivity is largely undocumented. A corollary is that alternative models may have a sizable degree of mutually irreproducible answers for complex flows.