[1] This paper investigates the interplay between river discharge and winds in forcing coastal buoyant outflows. During light winds a plume influenced by the Earth's rotation will flow down shelf (in the direction of Kelvin wave propagation) as a slender buoyancy-driven coastal current. Downwelling favorable winds augment this down-shelf flow, narrow the plume, and mix the water column. Upwelling favorable winds drive currents that counter the buoyancy-driven flow, spread plume waters offshore, and rapidly mix buoyant waters. Two criteria are developed to assess the wind influence on a buoyant outflow. The wind strength index (W s ) determines whether a plume's along-shelf flow is in a wind-driven or buoyancy-driven state. W s is the ratio of the wind-driven and buoyancy-driven along-shelf velocities. Wind influence on across-shelf plume structure is rated with a timescale (t tilt ) for the isopycnal tilting caused by wind-driven Ekman circulation. These criteria are used to characterize wind influence on the Delaware Coastal Current and can be applied to other coastal buoyant outflows. The Delaware buoyant outflow is simulated for springtime high-river discharge conditions. Simulation results and W s values reveal that the coastal current is buoyancy-driven most of the time (jW s j < 1 on average). Wind events, however, overwhelm the buoyancy-driven flow (jW s j > 1) several times during the high-discharge period. Strong upwelling events reverse the buoyant outflow; they constitute an important mechanism for transporting fresh water up shelf. Across-shelf plume structure is more sensitive to wind influence than the along-shelf flow. Values of t tilt indicate that moderate or strong winds persisting throughout a day can modify plume width significantly. Plume widening during upwelling events is accompanied by mixing that can erase the buoyant outflow.

No abstract

No abstract

No abstract

No abstract

Observations of estuarine low subtidal sea level and current fluctuations have often shown domination by the remote effects of the wind, acting on the adjacent coastal ocean, over the local surface stress, acting on the estuary itself. The remote effects are transmitted to the estuary by impressment on its mouth of sea level change induced by the onshore component of coastal Ekman transport and become increasingly dominant as the frequency decreases. A simple, barotropic model is developed to investigate the joint action of these two wind forcing mechanisms. The relative shortness of most estuaries relative to low subtidal estuarine wavelengths explains the dominance of the remote effect for both sea level and barotropic current fluctuations in the estuary. For the same reason, however, surface slope is dominated by local wind setup. For an estuary with axis nearly parallel to the coast the two effects will operate either in concert or in opposition, depending on hemisphere and orientatipn of the estuary axis relative to the coast. For the geometries of both Chesapeake Bay and the Delaware Estuary the model predicts opposition with the remote effect dominant at lower frequencies, consistent with recent observations.

Garvine and Monk [1974] have contributed to the body of brackish water through the tank. The arrow at the top inknowledge of river plumes. Their results fill in many details of dicates the convergence line of the front. Ahead of this point some earlier work done at our laboratory in which the outflow the brackish (darker) water has increased its thickness. (At the of the Nid river in Trondheimsfjord was treated as a problem , same time, a very thin layer of freshwater that escaped from in hydraulics.The side spreading of the lighter river water acts like a density front for which the (two-dimensional) front velocity computed by means of the Bernoulli equation yields a densimetric Froude number Fr = 2 [Moshagen, 1972]. The degree to which shear stresses enter into the propagation velocity is a matter of speculation. Benjamin [1968] implies that it is important, at least for small-scale laboratory experiments. This is not necessarily true on the length scale associated with large river plumes.Considering the difficulties of measuring the thickness and velocity, which tend to be highly unstationary near a front, the values of 1.7 and 3.3 appear to support the purely dynamical computation. These values have nothing to do with the problem discussed by $tommel and Farmer [1952], as is suggested by Garvine and Monk.The Nid river plume was computed by considering the time history of a cross-sectional slice of the river in a quasi-onedimensional model. By preserving the river momentum and allowing the river to spread sideways with Fr = 2 the width of

The outflow of buoyant waters from major estuaries affects the dynamics of inner continental shelves profoundly as lateral density gradients force an alongshore current. Often the Coriolis force causes the outflow to remain trapped near the coast. We observed one such current, the Delaware Coastal Current, on the inner shelf near the Delaware Estuary on the eastern seaboard of the United States. The spatial variability along the shelf, however, suggests at least two dynamically distinct regions that we term source and plume regions. In the source region we find fronts, a current whose width scales well with the internal deformation radius, and a ratio of relative to planetary vorticity that reaches unity, that is, the Rossby number is O(1). As nonlinear inertial forces in the across-shelf momentum balance are weak, we suggest that such forces contribute to the along-shelf momentum balance only. Farther downstream in the plume region, we find much reduced lateral density gradients, a current much wider than the deformation radius, and relative vorticities that are much smaller than the planetary vortiCity. From our observations we compute nondimensional dynamical parameters, with which we discuss our observations. The Burger, Rossby, and Ekman numbers for the Delaware Coastal Current suggest that most models of buoyancy-driven coastal currents do not apply to this coastal flow. I. INTRODUCTIONThe lateral flux of buoyancy from estuaries into the coastal ocean constitutes a forcing agent influencing both circulation and mixing. The density differences between brackish estuarine and salty oceanic waters force a flow on the shelf. Density gradients induce pressure gradients which are often balanced by Coriolis forces. In the northern hemisphere the Coriolis force turns the estuarine outflow to the right looking seaward and traps the buoyant water on the inner shelves. Across-shelf mixing with ambient shelf water is thus reduced.Such currents distribute riverborne nutrients, larvae, sediments, sewage, toxic chemicals, and oil from accidental spills dominantly along the shore. This study will describe one such current, namely the Delaware Coastal Current. Woods and BeardsIcy[1988] studied barotropic estuarine outflow problems with a set of analytical, numerical, and laboratory experiments. The outflowing fluid enters a uniformly sloping shelf, where vortex tube stretching and friction determine its path. Their studies relate indirectly to the discharge of water from major rivers, since they isolate barotropic from baroclinic effects. For small Rossby numbers, they found that Csanady's [1978] arrested topographic wave dynamics resulted, whereas for moderate Rossby numbers they reproduced the one-layer results of BeardsIcy and Hart [1978]. Beardsley and Hart prescribed an outflow from a point source, while Woods and Beardsley [1988] modeled the shelf flow that was forced by an outflow from a finite gap in the coastline. In both studies estuarine waters turn to the right in the northern hemisphere. Most outflows, however, ...