Simulating the Delaware Bay Buoyant Outflow: Comparison with Observations

Abstract

Coastal buoyant outflows from rivers and estuaries previously have been studied with field research, laboratory experiments, and numerical models. There is a dire need to evaluate model performance in light of coastal current observations. This research simulates the Delaware Bay outflow and compares results with observations of estuarine and shelf conditions. Observations include an estuarine salinity climatology, a record of freshwater delivery to the shelf, coastal current salinity mappings, and surface drifter data. Simulation efforts focus on spring 1993 and spring 1994, the primary field study period. The simulation is forced with river discharge, winds, and tides; only tidal-averaged results are discussed. Estuarine salinity results are consistent with the observed lateral salinity pattern, vertical structure, and response to river discharge. Salinities within the lower bay agree with observations, but the simulation overestimates the along-estuary salinity gradient. Observed and simulated freshwater delivery exhibit the same amplitude of response to river discharge and winds. The simulation produces a buoyant outflow that is generally consistent with the observed buoyancy signature, width, length, and vertical structure over a variety of river discharge and wind conditions. The simulated coastal current, however, tends to be somewhat shorter and fresher than observed. Simulated surface drifter paths exhibit the observed onshore advection during downwelling winds as well as offshore transport and current reversals during upwelling winds. A statistical evaluation based on shelf salinity mappings indicates that the model reproduces the observed variance and has only a small bias (less than 10% of plume buoyancy signature). The rms error of 1.2 psu is linked to the shorter and fresher nature of the simulated coastal current. Observational comparisons discussed in this paper indicate that the model can simulate many coastal current features and its response to river discharge and wind forcing.

Thermal Recirculation Modeling for Power Plants in an Estuarine Environment

“…The Estuary is 210 km long and comprises tidal reaches of the Delaware River (starting in Trenton, New Jersey) and all of the Delaware Bay [18]. The width of the estuary increases toward the mouth to 18 km, which extends between Cape May and Cape Henlopen.…”

“…Therefore, the application of an average river discharge can be an appropriate representation of the average forcing throughout the power plant life cycle. There are other tributaries flowing into the estuary between the mouth and upstream boundary, but not included, as their effects are considered localized [18] and not significant to the goals of this study.…”

“…Numerous studies have been performed for this water body to address tidal/sub-tidal variations, salinity intrusion, buoyant flow and storm surge [14][15][16][17][18][19]. However, the literature review did not result in any study that addresses thermal recirculation within this water body.…”

Characterizing the effluence near Waikiki, Hawaii with a coupled biophysical model

“…It affects the small-scale discharge of the Connecticut River by moving the plume to the east of the river mouth on an ebb tide and to the west on a flood tide (Whitney and Garvine, 2006). Using a numerical model, Guo and Valle-Levinson (2007) analyzed tidal effects on an outflow plume in Chesapeake Bay and concluded that a bottomadvected plume occurred due to tidal forcing whereas a surface-advected plume developed without tidal forcing.…”

“…For the high-volume river outflow of the Columbia River, plume mixing was strongest near the tidal plume front, which exchanges nutrients and sediments (Orton and Jay, 2005). Although some tidal forcing effects on freshwater plumes have been considered (Wiseman and Garvine, 1995;Orton and Jay, 2005;Whitney and Garvine, 2006;Guo and Valle-Levinson, 2007), the majority of studies have focused on the effects of wind forcing on plumes with large K values.…”