COASTAL OCEAN MODELING ACROSS LENGTH SCALES: FROM ESTUARINE CIRCULATION TO BOUNDARY CURRENTS

From the shoreline to the continental slope, numerical models of coastal physics provide forecasts of storm impacts, inform environmental policy, and drive key processes in climate predictions. The coastal zone is a highly dynamic region characterized by tides, buoyancy-driven flow, bathymetric steering, upwelling, and boundary currents. This complex mix of competing forces has necessitated the development of increasingly advanced realistic numerical models. Realistic models can be used to analyze ocean physics in ways that would not be possible with observational data alone. The first two chapters investigate water mass exchange between Narragansett Bay and Rhode Island Sound using current meters and regional models. The first chapter focuses on a single hurricane event which led to an intrusion of cold shelf water into the bay. Not only is this the first study to examine storm-driven residual transports in Narragansett Bay, but it also advances the broader literature on estuarine storm response by addressing the questions of local versus non-local wind forcing as well as the importance of baroclinic effects. Results show that local winds drive most of the exchange flow but non-local winds drive most of the storm surge. Additionally, baroclinic effects were necessary to establish realistic vertical shear, despite strong winds vertically mixing the water column. Chapter 2 looks at estuary-shelf exchange more broadly under a range of wind and tide conditions. Our improved calculations of Narragansett Bay exchange flow are at least two times larger than previous estimates and suggest that offshore inputs are a larger source of nitrogen in the bay than previously reported. The addition of a simulated dye to track shelf water nutrients reveals substantial variations in the dye concentration of the inflow. We find that southward winds drive the strongest exchange flow but that eastward winds drive shelf upwelling which increases the concentration of dye/nutrients entering the bay. Chapter 3 looks to the future of climate modeling with a new variable-resolution global ocean model which can resolve small-scale coastal processes that would be too computationally expensive for traditional models. We find that enhancing resolution in a band around the coast of North America improves representation of eastern boundary upwelling but creates unrealistic Gulf Stream behavior. We show that the Gulf Stream gets trapped in the enhanced resolution region and impinges on the Labrador Current which prevents deepwater formation, leading to a dramatic weakening in the thermohaline circulation. We provide recommendations on how to change the mesh to avoid this impingement, and how to improve the mesh design process in the future