Hydrography and frontogenesis in a glacial fjord off the Strait of Magellan

Current velocity and hydrographic profiles obtained for the first time in a Chilean glacial fjord were combined with under-way surface temperature and salinity measurements to describe the formation of tidal intrusion fronts and plume-like fronts. These fronts formed within several hundred meters from each other in the vicinity of a shallow sill, maximum depth of approximately 3 m, in a glacial fjord off the Strait of Magellan in the Chilean Patagonia. Measurements were obtained in mid-December of 2003 and 2004, during late austral spring, under active glacier melting and calving. The glacial fjord is approximately 18 km long from the face of the glacier to the connection with the Strait of Magellan and typically less than 1 km wide throughout the system. Between the glacier face and the 3-m sill, depths are typically less than 100 m, and seaward of the sill, depths increase to more than 200 m. Velocity and salinity data obtained during flood periods revealed that water with oceanic salinity was aspirated to near-surface levels from depths of approximately 30 m as flood flows accelerated from approximately 10 cm s⁻¹, seaward of the sill, to approximately 60 cm s⁻¹ at the sill crest. The upwelled water was then slightly diluted by mixing at the sill crest before plunging down to the basin between the glacier and the sill. The plunging of salty water over the sill created dramatic tidal intrusion fronts only a few tens of meters from the sill crest and pumping of salt with every flood period. During ebb periods, the low salinity waters derived from the glacier and a small river near the glacier converged at the sill crest. After some mixing, the buoyant waters were released within a thin layer (∼3 m deep) lead by a plume-like front that remained coherent for a few hundred meters seaward of the sill. The main findings of this study were that tidal intrusion and plume fronts were observed within 2 km from each other, and that tidal pumping was the predominant mechanism for salt fluxes into the system.     

Plume rise and spread in a linearly stratified environment

We present analyses of plume rise into a linearly stratified environment, either with or without a uniform horizontal flow. In the case of a still ambient, we collate results on plume spreading height and volume flux, enabling the speed of the spreading intrusion in the buoyancy-inertia regime to be expressed in terms of the fundamental parameters of plume buoyancy flux and ambient buoyancy frequency. A theoretical expression for the final volume flux emanating from the plume-rise region, in terms of maximum rise height, is also derived. Hence it is shown that the ratio of the intrusion radius to the maximum rise height is a simple function of ambient buoyancy frequency and time. In the case of a wind, we analyse the theoretical model for a rising plume to obtain predictions for the downwind volume flux, and subsequent lateral spread, in the limits of strong and weak wind. We identify a regime of very weak wind, which may be modelled as a passive advection of plume flow in the still case as a first approximation. Numerical solutions of a general model are presented which show that it predicts a peak in entrainment, and hence volume-flux growth, in the case of intermediate wind strength. We verify the crosswind model predictions of lateral spread, upwind penetration and entrainment by comparison with large-eddy simulations.     

Interaction of a river plume with coastal upwelling in the northeastern South China Sea

Observational and modeling studies were conducted to investigate the Pearl River plume and its interaction with the southwesterly driven upwelling circulation in the northern South China Sea during the summer. After exiting the Pearl River Estuary, the discharged freshwater generates a nearly stationary bulge of freshwater near the entrance of the estuary. Forced by the wind-driven coastal upwelling current, the freshwater in the outer part of the bulge flows downstream at the speed of the current and forms a widening and deepening buoyant plume over the shelf. The plume axis gradually shifts offshore of the current maximum as a result of currents induced by the contrasting density at the nose of plume and by the intensified Ekman drift in the plume. In this plume–current system, the fraction of the discharged freshwater volume accumulated in the bulge reaches a steady state and the volume of newly discharged freshwater is transported downstream by the upwelling current. Enhancement of stratification by the plume thins the surface frictional layer and enhances the cross-shelf circulation in the upper water column such that the surface Ekman current and compensating flow beneath the plume are amplified while the shoaling of the deeper dense water in the upwelling region changes minimally. The pressure gradient generated between the buoyant plume and ambient seawater accelerates the wind-driven current along the inshore edge of the plume but retards it along the offshore edge. Along the plume, downward momentum advection is strong near the highly nonlinear source region and a weaker upward momentum advection occurs in the far field over the shelf. Typically, the plume is shaped by the current over the shelf while the current itself is adjusting to a new dynamic balance invoked by the plume-induced changes of vertical viscosity and the horizontal pressure gradient. The spatial variation of this new balance leads to a coherent change in the cross-isobath transport in the upper water column during upwelling.     

A model study of the Copper River plume and its effects on the northern Gulf of Alaska

A three-level nested Regional Ocean Modeling System was used to examine the seasonal evolution of the Copper River (CR) plume and how it influences the along- and across-shore transport in the northern Gulf of Alaska (NGoA). A passive tracer was introduced in the model to delineate the growth and decay of the plume and to diagnose the spread of the CR discharge in the shelf, into Prince William Sound (PWS) and offshore. Furthermore, a model experiment with doubled discharge was conducted to investigate potential impacts of accelerated glacier melt in future climate scenarios. The 2010 and 2011 simulation revealed that the upstream (eastward) transport in the NGoA is negligible. About 60 % of the passive tracer released in the CR discharge is transported southwestward on the shelf, while another one third goes into PWS with close to 60 % of which exiting PWS to the shelf from Montague Strait. The rest few percent is transported across the shelf break and exported to the GoA basin. The downstream transport and the transport into PWS are strongly regulated by the downwelling-favorable wind, while the offshore transport is related to the accumulation of plume water in the shelf, frontal instability, and the Alaskan Stream. It takes weeks in spring for the buoyancy to accumulate so that a bulge forms outside of the CR estuary. The absence of strong storms as in the summer of 2010 allows the bulge continue growing to trigger frontal instability. These frontal features can interact with the Alaskan Stream to induce transport pulses across the shelf break. Alternatively as in 2011, a downwelling-favorable wind event in early August (near the peak discharge) accelerates the southwestward coastal current and produces an intense downstream transport event. Both processes result in fast drains of the buoyancy and the plume content, thereby rapid disintegration of the plume in the shelf. The plume in the doubled discharge case can be two to three times in size, which affects not only the magnitude but also the timing of certain transport events. In particular, the offshore transport increases by several folds because the plume appears to be more easily entrained by the seaward flow along the side of Hinchinbrook Canyon.     

Allochthonous subsidies of organic matter across a lake–river–fjord landscape in the Chilean Patagonia: Implications for marine zooplankton in inner fjord areas

Ecosystems can act as both sources and sinks of allochthonous nutrients and organic matter. In this sense, fjord ecosystems are a typical interface and buffer zone between freshwater systems, glaciated continents, and the coastal ocean. In order to evaluate the potential sources and composition of organic matter across fjord ecosystems, we characterized particulate organic matter along a lake–river–fjord corridor in the Chilean Patagonia using stable isotope (δ¹³C) and lipid (fatty acid composition) biomarker analyses. Furthermore, estimates of zooplankton carbon ingestion rates and measurements of δ¹³C and δ¹⁵N in zooplankton (copepods) were used to evaluate the implications of allochthonous subsidies for copepods inhabiting inner fjord areas. Our results showed that riverine freshwater flows contributed an important amount of dissolved silicon but, scarce nitrate and phosphate to the brackish surface layer of the fjord ecosystem. Isotopic signatures of particulate organic matter from lakes and rivers were distinct from their counterparts in oceanic influenced stations. Terrestrial allochthonous sources could support around 68–86% of the particulate organic carbon in the river plume and glacier melting areas, whereas fatty acid concentrations were maximal in the surface waters of the Pascua and Baker river plumes. Estimates of carbon ingestion rates and δ¹³C in copepods from the river plume areas indicated that terrestrial carbon could account for a significant percentage of the copepod body carbon (20–50%) during periods of food limitation. Particulate organic matter from the Pascua River showed a greater allochthonous contribution of terrigenous/vascular plant sources. Rivers may provide fjord ecosystems with allochthonous contributions from different sources because of the distinct vegetation coverage and land use along each river’s watershed. These observations have significant implications for the management of local riverine areas in the context of any human project that may modify terrestrial habitats as well as the productivity, food webs, and community structure of rivers, lakes, fjords, and the coastal ocean in the Chilean Patagonia.     

Vortex pairs and power station cooling water

Cooling water discharged from power stations in the U.K. is frequently released from an outlet in an estuary or the sea. The warm water forms a thermal plume which is slightly buoyant and which spreads horizontally over the water surface while mixing vertically downwards with the cooler ambient water. In this paper, the possibility of vortex pair production at the cooling water outlet is considered as a mechanism contributing to this spreading of the warm water.The motion of a vortex pair contained between two rigid plane boundaries is an idealization of the flow between the water surface and the sea bed. The resulting motion is calculated from potential theory and viscous effects are neglected. The problem of deciding what strength to assign the vortices is discussed and specific consideration of shear and buoyancy at the outlet is detailed. It is observed that bifurcation of the vortex pair is determined by the initial position of the vortices and is unlikely to occur in conditions relevant to U.K. power station discharges.It is calculated that, in the absence of turbulence, the motion of such vortex pairs would result in horizontal spreading of the warm water which is greater than that observed at site surveys. It is concluded that turbulence in the ambient receiving water is sufficient to destroy vortices produced by the discharge during the early stages of the plume development.     

Scales and structure of frontal adjustment and freshwater export in a region of freshwater influence

Sea surface temperature satellite imagery and a regional hydrodynamic model are used to investigate the variability and structure of the Liverpool Bay thermohaline front. A statistically based water mass classification technique is used to locate the front in both data sets. The front moves between 5 and 35 km in response to spring–neap changes in tidal mixing, an adjustment that is much greater than at other shelf-sea fronts. Superimposed on top of this fortnightly cycle are semi-diurnal movements of 5–10 km driven by flood and ebb tidal currents. Seasonal variability in the freshwater discharge and the density difference between buoyant inflow and more saline Irish Sea water give rise to two different dynamical regimes. During winter, when cold inflow reduces the buoyancy of the plume, a bottom-advected front develops. Over the summer, when warm river water provides additional buoyancy, a surface-advected plume detaches from the bottom and propagates much larger distances across the bay. Decoupled from near-bed processes, the position of the surface front is more variable. Fortnightly stratification and re-mixing over large areas of Liverpool Bay is a potentially important mechanism by which freshwater, and its nutrient and pollutant loads, are exported from the coastal plume system. Based on length scales estimated from model and satellite data, the erosion of post-neap stratification is estimated to be responsible for exporting approximately 19% of the fresh estuarine discharge annually entering the system. Although the baroclinic residual circulation makes a more significant contribution to freshwater fluxes, the episodic nature of the spring–neap cycle may have important implications for biogeochemical cycles within the bay.     

The anthropogenic and biogenic origin of low molecular weight halocarbons in a polluted fjord,the Idefjorden

Three halocarbons—chloroform, tetrachloroethylene and bromoform—have been used as tracers to study the water mixing in the Idefjorden, a fjord on the border between Norway and Sweden. Chloroform is discharged into the fjord from a pulp and paper bleachery on the river Tista, and tetrachloroethylene is a constituent of the surface run-off water of the city Halden, located at the middle of the fjord. They are both horizontally distributed in a surface layer separated from the deep water by a pycnocline. Tetrachloroethylene, furthermore, shows an elevation in concentration in the deep water just outside Halden due to the fact that it is dissolved in the sewage water of the city and discharged through a pipe into the deep water. Bromoform has a natural source in algae belts outside the fjord, and its distribution into the fjord and in a transect from the Skagerrak is illustrated.     

Transport of variable-density solute plumes in beach aquifers in response to oceanic forcing

A comprehensive numerical study was undertaken to investigate transport of a variable-density, conservative solute plume in an unconfined coastal aquifer subject to high and low frequency oceanic forcing. The model combined variable-density saturated flow for groundwater and solute transport, and wave hydrodynamics from a 2D Navier–Stokes solver. A sinusoidal tidal signal was specified by implementing time-varying heads at the seaward boundary. The solute plume behavior was investigated under different oceanic forcing conditions: no forcing, waves, tide, and combined waves and tide. For each forcing condition, four different injected solute densities (freshwater, brackish water, seawater, brine) were used to investigate the effects of density on the transport of the injected plume beneath and across the beach face. The plume’s low-order spatial moments were computed, viz., mass, centroid, variance and aspect ratio. The results confirmed that both tide- and wave-forcing produce an upper saline plume beneath the beach face in addition to the classical saltwater wedge. For the no-forcing and tide-only cases (during rising tides), an additional small circulation cell below the beach face was observed. Oceanic forcing affects strongly the solute plume’s flow path, residence time and discharge rate across the beach face, as well as its spreading. For the same oceanic forcing, solute plumes with different densities follow different trajectories from the source to the discharge location (beach face). The residence time and plume spreading increased with plume density. It was concluded that simulations that neglect the effect of waves or tides cannot reproduce accurately solute plume dispersion and also, in the case of coasts with small waves or tides, the solute residence time in the aquifer.     

A model study of tide- and wind-induced mixing in the Columbia River Estuary and plume

A numerical simulation of circulation in the Columbia River estuary and plume during the summer of 2004 is used to explore the mixing involved as river water is transformed into shelf water. The model is forced with realistic river flow, tides, wind stress, surface heat flux, and ocean boundary conditions. Simulated currents and water properties on the shelf near the mouth are compared with records from three moorings (all in 72 m of water) and five CTD sections. The model is found to have reasonable skill; statistically significant correlations between observed and modeled surface currents, temperature, and salinity are all 0.42–0.72 for the mooring records. Equations for the tidally averaged, volume-integrated mechanical energy budget (kinetic and potential) are derived, with attention to the effects of: (i) Reynolds averaging, (ii) a time varying volume due to the free surface, and (iii) dissipation very close to the bottom. It is found that convergence of tidal pressure work is the most important forcing term in the estuary. In the far field plume (which has a volume 15 times greater than that of the estuary), the net forcing is weaker than that in the estuary, and may be due to either tidal currents or wind stress depending on the time period considered. These forcings lead to irreversible mixing of the stratification (buoyancy flux) that turns river water into shelf water. This occurs in both the plume and estuary, but appears to be more efficient (17% vs. 5%), and somewhat greater (4.2 MW vs. 3.3 MW), in plume vs. estuary. This demonstrates the importance of both wind and tidal forcing to watermass transformation, and the need to consider the estuary and plume as part of a single system.     

Numerical study of circulation and temperature-salinity distributions in the Bras d’Or Lakes

The Bras d’Or Lakes (BdOL) are a large, complex and virtually land-locked estuary in central Cape Breton Island of Nova Scotia and one of Canada’s charismatic ecosystems, sustaining ecological and cultural communities unique in many aspects. The BdOL comprise two major basins, many deep and shallow bays, several narrow channels and straits and a large, geologically complex watershed. Predictive knowledge of the water movement within the estuary is a key requirement for effective management and sustainable development of the BdOL ecosystem. A three-dimensional (3D) primitive-equation ocean circulation model is used to examine the estuary’s response to tides, winds and buoyancy forcing associated with freshwater runoff in a series of numerical experiments validated with empirical data. The model results generate intense, jet-like tidal flows of about 1 m s^?1 in the channels between the basins and connecting them to the ocean and relatively weak tidal currents in other regions, which agrees well with previous observations and numerical results. Wind forcing and buoyancy forcing associated with river runoff play important roles in generating the significant sub-tidal circulations in the estuary, including narrow channels, deep basins and shallow bays. The circulation model is also used to reconstruct the 3D circulation and temperature-salinity distributions in the summer months of 1974, when current and hydrographic measurements were made at several locations. The sub-tidal circulation in the estuary produced by the model is characterised by wind and barometric set-up and set-down in different sections of the system, and a classic two-layer estuarine circulation in which brackish, near-surface waters flow seaward from the estuary into the Atlantic Ocean, and deep salty waters flow landward through the major channel. The model results reproduce reasonably well the overall features of observed circulation and temperature-salinity fields made in the BdOL in 1974 but generally underestimate the observed currents and density stratification. The model discrepancies reflect the use of spatially mean wind forcing and spatially and monthly mean surface heat flux and the inability of the coarse model horizontal resolution (～500 m) to resolve narrow channels and straits.     

Feasibility of Deep Nutrients Delivery into a Prince William Sound Beach for the Bioremediation of the Exxon Valdez Oil Spill

Delivery of dissolved chemicals to potentially bioremediate oil from the Exxon Valdez oil spill was investigated at Smith Island, Prince William Sound, Alaska. Two transects for tracer application were installed: one at a clean area and another at an oiled area. Tracer delivery occurred under ambient pressure through manifolds. Lithium in LiBr/sea water solution was the inert tracer, and the solution (82.6 mg/L of lithium) was released at a flow rate of 0.23 L/min for 58.5 h. While maintaining the flow rate, the solution was switched to sea water (i.e., 0.0 mg/L of lithium) for 16 h. The results show that the tracer moved landward with rising tides and seaward with falling tides. The plume got deeper moving landward and shallower moving seaward of the manifold. Thus, in situations where oil biodegradation is limited by the availability of nutrients or oxygen, applying these chemicals by this technique would allow them to reach entrapped oil seaward of the manifold from below. The seaward plume traveling speed (around 2.0 m/d) suggests that this technique is logistically feasible from a hydraulic point of view.     

Physical and dynamical oceanography of Liverpool Bay

The UK National Oceanography Centre has maintained an observatory in Liverpool Bay since August 2002. Over 8 years of observational measurements are used in conjunction with regional ocean modelling data to describe the physical and dynamical oceanography of Liverpool Bay and to validate the regional model, POLCOMS. Tidal dynamics and plume buoyancy govern the fate of the fresh water as it enters the sea, as well as the fate of its sediment, contaminants and nutrient loads. In this context, an overview and summary of Liverpool Bay tidal dynamics are presented. Freshwater forcing statistics are presented showing that on average the bay receives 233 m³ s^− 1. Though the region is salinity controlled, river input temperature is shown to significantly modulate the plume buoyancy with a seasonal cycle. Stratification strongly influences the region’s dynamics. Data from long-term moored instrumentation are used to analyse the stratification statistics that are representative of the region. It is shown that for 65% of tidal cycles, the region alternates between being vertically mixed and stratified. Plume dynamics are diagnosed from the model and are presented for the region. The spring–neap modulation of the plume’s westward extent, between 3.5 °W and 4°W, is highlighted. The rapid eastward erosion of the plume during spring tides is identified as a potentially important freshwater mixing mechanism. Novel climatological maps of temperature, salinity and density from the CTD surveys are presented and used to validate numerical simulations. The model is found to be sensitive to the freshwater forcing rates, temperature and salinities. The existing CTD survey grid is shown to not extend sufficiently near the coast to capture the near coastal and vertically mixed component the plume. Instead the survey grid captures the westward spreading, shallow and transient, portion of the plume. This transient plume feature is shown in both the long-term averaged model and observational data as a band of stratified fluid stretching between the mouth of the Mersey towards the Isle of Man. Finally the residual circulation is discussed. Long-term moored ADCP data are favourably compared with model data, showing the general northward flow of surface water and southward trajectory of bottom water.     

Tidal-induced buoyancy flux and mean transverse circulation

A weakly nonlinear model is used to examine the mean transverse circulation (cross-isobath) driven by tidal-induced buoyancy flux. The mean Eulerian flows driven by both the barotropic and baroclinic tide are presented for a semi-infinite wedge. The mean flow driven by the barotropic tide is significant only near the apex where the thickness of the frictional boundary layer is comparable to the water depth. The mean flow there is characterized by a single-cell circulation with offshore flow near the bottom, and its magnitude can reach a few percentage or a significant fraction of the tidal velocity in oceanic applications. The mean flow driven by the baroclinic tide, on the other hand, is characterized by pairs of half-open (on the seaward side) counter-rotating cells, the number of which equals the vertical mode number. For a baroclinic tide propagating onshore, the mean flow near the top and bottom surfaces is always directed offshore and its magnitude can reach a large fraction of the tidal velocity. Taken together, the model thus predicts a mean offshore flow near the bottom while higher up in the water column the mean flow direction is less definite due to the contribution from different tidal components. The model results are consistent with some current measurements over the Georges Bank.     

Exchange dynamics of a shallow contaminated wetland

Flux pathways are investigated for a wetland system comprised of a river flowing through two shallow forebays and then entering the main basin of a lake. The hydrodynamic exchanges between these three components influence the fate and transport of heavy metals in the system. During non-storm summer flows the river plume, comprised of river water and entrained forebay water, was cooler than the lake surface water and so plunged as it entered the main basin and inserted near the seasonal thermocline. Because the river plume plunged, only a fraction of its metals flux was available to the epilimnion. A return flow into the forebay was always observed immediately above the river inflow, ensuring a predominantly two-layer exchange system. However during days of negative heat flux (surface heating), an additional exchange mechanism existed when the surface waters from the forebay formed a buoyant plume, flowing out into the main basin. This heated outflow produced a significant, but short-lived, arsenic flux from the forebay into the surface waters of the lake.     

Application of aerial photography to the study of small scale upper ocean phenomena

The industrial waste dumped 180 n. miles south of Galveston was monitored in July 1977 by water sampling, hydrographic measurements, acoustic tracking on board two vessels, and by aerial photography. The plume of the waste diffused vertically and horizontally. Photodensitometry of aerial photos of the plume showed lateral dispersion of the plume in agreement with two other methods: acoustic tracking of the waste suspensoid and transmissometer sampling. In addition, the method showed small scale features like the lateral and longitudinal variations in the photodensity, indicating the waste concentration. This waste concentration showed periodic changes in its axial distance, with the spectral peak at about 160 m wave length. It shows a sharp increase at the windward edge of the plume as do the acoustic records. This phenomenon is explained in terms of the shearing current near the surface together with vertical diffusion. The periodic change along the axis is explained in terms of the Langmuir circulation and in terms of internal ship waves.     

A stochastic model for 3-dimensional flow patterns in infiltration experiments

Modelling the 3-dimensional water flux at field scale is important for the design and the analysis of dye tracer experiments. Furthermore, it enables the estimation of the risk to groundwater by pollutants, and the visualisation and classification of flux patterns. A stochastic model is presented that allows for the modelling of a wide range of flow patterns in soils as they appear in dye tracer experiments. The leading idea is that infiltrating water runs along paths, not necessarily preferential ones, and water spreads into the soil uniformly from the paths into the matrix. The model is essentially based on a Poisson point process and three independent random fields. The point process defines the starting points of the paths at the surface. The values of two random fields determine the course of the paths. The third random field governs the depth of the infiltration front. As an extension of the model, we present two simulated examples for stratified soils.     

Quantifying the suspended sediment discharge to the ocean from the Markham River,Papua New Guinea

The Markham River is a small river draining a tropical mountain range with altitudes between 1000 and 3000 m and discharges directly into a submarine canyon, the head of which is at 30 m depth and reaches depths of 500 m only 4 km from the shore. As such, the Markham discharge system serves as a possible analogue for rivers discharging onto margins during low stands of sea-level. Located in a tectonically active area and with high rainfall, sediment supply is high and episodic and is sometimes related to catastrophic mountain landslides. The river has an estimated sediment load of 12 Mt yr⁻¹. Occasionally, high energy flows are generated at the river mouth which is evident from the channel morphology and sediment distribution. Profiles of salinity and suspended sediment concentrations (SSC) show that sediment is dispersed via a plume with components at both the surface, intermediate depth along isopycnal surfaces and near the sea bed. The dispersal pattern of the surface freshwater plume is largely determined by the buoyancy force. The surface plume is very thin with salinity gradients 15 ppt m⁻¹ while a Richardson number greater than unity suggested that the mixing zone is highly stratified. Estimates of the horizontal sediment flux gradient of the surface plume along the estuary axis suggest that about 80% of the sediment discharged is lost from the plume within a distance of 2 km from the river mouth. Particle fall velocities estimated from the vertical flux indicate values less than those of flocculated material. Layers of sediment with SSCs between 500 and 1000 mg l⁻¹ were observed at intermediate depths and near the seabed during periods of both high and intermediate discharge. The mass of sediment in a SSC layer at intermediate depths between 150 and 250 m within the canyon channel was estimated to be equivalent to an average of 2 to 3 days of Markham sediment discharge. SSCs near the seabed of between 250 and 750 mg l⁻¹ suggest that layers of significantly elevated density exist near the seabed, moving under the influence of gravity down steep seabed slopes of the Markham canyon.     

Fate of three major rivers in the Bohai Sea: A model study

Huanghe (Yellow River), Haihe and Liaohe are three major rivers flowing into the Bohai Sea and account for more than 80% of the freshwater and land-drained material inputs annually. The fate of three rivers in the seawaters correlates with the transport and distribution of the riverine sediments and nutrients, and further exerts a profound influence on the local marine ecosystem dynamics. Therefore, the evolution of the river plumes under the influence of the freshwater buoyancy, the tidal forcing and the wind stress are examined using a three-dimensional primitive equation ocean circulation model, independently and jointly. It is found that both tide and wind stirring can deteriorate the stabilization of the water column caused by the freshwater buoyancy; however, the processes are different. The tide stirring originates from the seafloor due to the bottom friction as the tidal wave propagates into the shallow waters, and then the turbulent kinetic energy dissipates upward. On the other hand, the wind stirring proceeds in the up-down direction. The influences of different winds on the evolution of the river plumes are also examined. Since the situation of each river mouth is different, the wind influence is also distinct. At last, the fate of three major rivers driven by the combined tidal forcing and climatology winds is reproduced, and the simulated salinity distribution shows a reasonable agreement with that observed, meaning that the river plume evolution plays a crucial role in shaping the salinity distribution in BS.     

On the three-dimensional oceanic density distribution in a millennium integration with an AO-GCM

 The three-dimensional time-mean density distribution in the ocean is determined not only by the time-mean fluxes of heat and freshwater at the sea surface, but also by time-mean vertical currents and time-mean density fluxes due to oceanic transients excited by fluctuating fluxes at the sea surface. The effects of these various processes on the global density fields are assessed using a balance equation of the variance of spatial density anomalies and a millennium integration with an atmosphere–ocean general circulation model. It is found that spatial density anomalies are generated by the time-mean heat fluxes at the sea surface and destroyed by the time-mean surface freshwater flux, by sinking of dense water and rising of less dense water, and finally by density fluxes associated with transients. The last two processes take place essentially in the oceanic interior. Since density fluxes of transient eddies act to reduce the existing density differences between the Atlantic/Southern Oceans and the other oceans, their presence could affect the global density balance, and from that the thermohaline circulation and the stability of this circulation. Received: 4 October 2001 / Accepted: 10 October 2002 Responsible Editor: Richard J. Greatbatch Acknowledgements I thank Ulrich Cubasch and his colleagues for providing me with the ECHAM3/LSG integration, Peter Müller and Richard Greatbatch for valuable suggestions.