Salt transport in partially mixed estuaries

Date of Completion

January 2005


Physical Oceanography




Simple models are used to investigate the subtidal exchange of salt between estuaries and adjacent waters. Inverse methods are used to fit dispersion coefficients to observations in Long Island Sound, Delaware Bay and Chesapeake Bay, estimate uncertainties and measure model performance. A salt budget box-model is developed which allows for the inclusion of classical gravitational circulation, varying as the 1/3-power of the runoff, in addition to a constant axial dispersion due primarily to interaction of currents with bathymetry. The two-layered exchange flow in Long Island Sound accounts for 12%±10% of the exchange of salt with shelf waters. Exchange through the East River with New York Harbor is constrained by transport estimates made by Blumberg and Pritchard (1997) and is determined to be less than 10% of the exchange through The Race. A one-dimensional model is developed for tapered channel geometry, allowing for a net flux of salt through the estuary. Age, the time a water-parcel has been in the estuary since entering at a boundary, is used as a passive tracer for comparison with flushing time estimates evaluated using a freshwater fraction approach and residence times evaluated as the amount of salt in the estuarine segment divided by the rate of influx due the dispersion coefficient acting on the salinity gradient. Results suggest that dispersion increases with cross-sectional area only for higher values of the dispersion coefficient. The negative curvature of the axial salinity profile required for the salinity to asymptotically approach shelf-water salinities is consistent with an increase of dispersion near the mouth, and requires a channel which decreases in cross-sectional area in the up-estuary direction in the vicinity of the mouth. Application to the three estuaries suggests that channel geometry as well as the magnitude and distribution of dispersion and runoff can help explain the differences in their salinity profiles. A simpler tapered model is used to demonstrate the influence of channel convergence on the runoff-dependence of the salinity intrusion in the Connecticut River and the Delaware and Chesapeake Bays. ^