Journal Of Physical Oceanography Research Articles

Well-Posedness of Hibler’s Dynamical Sea-Ice Model

This paper establishes the local-in-time well-posedness of solutions to an approximating system constructed by mildly regularizing the dynamical sea-ice model of {\it W.D. Hibler, Journal of Physical Oceanography, 1979}. Our choice of regularization has been carefully designed, prompted by physical considerations, to retain the original coupled hyperbolic-parabolic character of Hibler's model. Various regularized versions of this model have been used widely for the numerical simulation of the circulation and thickness of the Arctic ice cover. However, due to the singularity in the ice rheology, the notion of solutions to the original model is unclear. Instead, an approximating system, which captures current numerical study, is proposed. The well-posedness theory of such a system provides a first-step groundwork in both numerical study and future analytical study.

Comments on “The Interaction of an Eastward-Flowing Current and an Island: Sub- and Supercritical Flow”

AbstractIn regards to the recent paper by Pedlosky and Spall (Journal of Physical Oceanography, November 2015), this comment maintains that the steady-state solutions of Rossby waves in a uniform eastward current past an island have waves on the upstream side that are not caused by the island because of inappropriate boundary conditions and the assumed form of the solution. The solutions are interesting but are not the solutions to the problem as posed in their paper. Similar upstream waves in a time-dependent numerical model are also inconsistent with linear Rossby wave theory, though the reasons for their presence are uncertain.

Suspended sediments due to random waves including effects of second order wave asymmetry and boundary layer streaming

The paper provides a practical stochastic method by which the suspended sediment concentration due to long-crested (2D) and short-crested (3D) nonlinear random waves can be calculated. The approach is based on assuming the waves to be a stationary narrow-band random process, and by using the parameterized formulas valid for regular waves presented in Soulsby, R.L., 1997. Dynamics of Marine Sands. Thomas Telford, London. The Forristall, G.Z., 2000. Journal of Physical Oceanography. 30, 1931–1943. wave crest height distribution representing both 2D and 3D nonlinear random waves is also adopted. The model covers sediment suspension over rippled beds and for sheet flow. Comparisons are made with random wave data from Thorne, P.D., Williams, J.J., Davies, A.G., 2002. Journal of Geophysical Research. 107 (C8), 3178 for flow over rippled beds. An example for sheet flow using data typical to field conditions is also included to illustrate the approach.

Directional recording of swell from distant storms

Almost fifteen years ago in the pages of this Journal, one of us presented power spectra of ocean waves and swell off Pendeen and Perranporth in north Cornwall (Barber & Ursell 1948). The outstanding feature of these spectra is the successive shift of peaks toward higher frequencies. This is the expected behaviour of dispersive wave trains from rather well-defined sources. Storms generate a broad spectrum of frequencies; the low frequencies are associated with the largest group velocity and accordingly are the first to arrive at distant stations. The time rate of increase in the frequency of peaks determines the distance and time of origin. In this way Barber & Ursell were able to identify the dispersive arrivals with a low pressure area in the North Atlantic, a tropical storm off Florida, and a storm off Cape Horn, at distances of 1200, 2800, and 6000 miles, respectively, from the Cornish stations. The measurements were consistent with the simple classical result that each frequency,/, is propagated with its appropriate group velocity, V = g/(47[/*). The present study is in a sense a refinement to the work of Barber & Ursell. The frequency resolution and sensitivity have each been increased by an order of m agnitude, and this makes it possible to detect and resolve meteorological sources that have previously been out of reach. The antipodal swell from the Indian Ocean is a case in point

Circulation and mixing at the New England shelfbreak front: Results of purposeful tracer experiments

We present the results of six dye tracer experiments that measured the mixing and circulation at the shelfbreak front on the New England Shelf. The last three were conducted during the New England Shelfbreak Productivity Experiment (NESPEX) with concurrent isopycnal float deployments. The results are consistent with the Chapman and Lentz [Chapman, D.C., and Lentz, S.J. (1994). Trapping of a coastal density front by the bottom boundary layer. Journal of Physical Oceanography, 24, 1465–1479.] model prediction of the separation and upwelling along the shelfbreak front of bottom boundary layer (BBL) water forced by an Ekman buoyancy flux, but show considerable variability. Cross-shelf velocities at the detachment point are 2–3 × 10 −2 m/s. But seaward, over the slope region, dye tagged water was sheared from the main patch into small filaments that upwelled along the front with cross-shelf speeds up to 0.1 m/s. Cross-shelf diffusion was of order 10 m 2/s in the mixed bottom layer and 1 m 2/s in the interior along the front. Within the stratified front, the mean vertical diffusivity was K z ∼ 4 × 10 −6 m 2/s. The dispersion of shelfwater in the slope region is effected by turbulent flow with advective speeds exceeding the small scale diffusive mixing. The mean flux of the detached BBL water is sufficient to account for the net loss of shelf water during its transit from Cape Cod to Cape Hatteras.

Stability of algebraic non-equilibrium second-order closure models

Stability problems of algebraic non-equilibrium second-moment closure models have given rise to the so-called quasi-equilibrium versions in which turbulence equilibrium is used as an additional constraint. In this paper, we investigate reasons for the failure of the G.L. Mellor, T. Yamada [Reviews of Geophysics 20 (1982) 851] level 2.5 closure model and suggest a remedy for this. We further discuss a new non-equilibrium closure model by V.M. Canuto, A. Howard, Y. Cheng, M.S. Dubovikov (Journal of Physical Oceanography, 2000, accepted for publication) which has proven to allow for stable calculations. All models are then numerically tested with a simple wind entrainment experiment motivated by the H. Kato, O.M. Phillips [Journal of Fluid Mechanics 37 (1969) 643] laboratory experiment, with the aid of which the instability of the Mellor and Yamada (1982) and the stability of the Canuto et al. (2000) model are confirmed. The Canuto et al. (2000) model has three advantages compared to the Mellor and Yamada (1982) which are (i) the symmetry of stability functions, (ii) a higher critical Richardson number, and (iii) that the normalised shear stress increases with normalised shear for turbulence equilibrium. The latter advantage of the new model causes its high physical and numerical stability.

A two-dimensional slice model of the shelf edge region off the west coast of Scotland: model response to realistic seasonal forcing and the role of the M 2 tide

A high resolution, two-dimensional cross-shelf slice model, with a range of turbulence energy sub-models, is applied to the shelf and shelf-edge region off the west coast of Scotland. The model, based upon previous work in the region (Xing and Davies, 1996, Journal of Physical Oceanography, 26, 417–447), emphasises the need for high resolution in both vertical and horizontal directions to study processes over the shelf edge. Using realistic seasonal meterological forcing, a year long integration is performed to investigate the role of mixing upon the formation of seasonal stratification. Results show that, whilst the meteorological forcing is essential to the formation of seasonal stratification and the sub-tidal cross-shelf circulation, the tide (barotropic and associated internal tide) is an important process that enhances mixing and the sub-tidal energetic cross-shelf circulation over the shelf edge.

The dynamic coupling of a third-generation wave model and a 3D hydrodynamic model through boundary layers

Abstract A third-generation wind wave model based on the energy balance equation taking into account the effects of time-varying currents and coupled dynamically with a semi-implicit three-dimensional hydrodynamic model incorporating the influences of time- and space-varying vertical eddy viscosity, bottom topography and wave-current interactions is presented in this paper. The wave model is synchronously coupled with the three-dimensional hydrodynamic model through the surface atmospheric turbulent boundary layer and the bottom boundary layer. The theory of Janssen (1991) (in Journal of Physical Oceanography 21 , 1631–1642) is used to incorporate the effects of waves on the surface boundary layer, while the theory of Grant and Maddsen (1979) [in Journal of Geophysical Research (Oceans) 84 , 1797–1808], which was used by Signell et al. (1990) (in Journal of Geophysical Research 95 , 9671–9678) on the bottom boundary layer for constant waves, is modified for the inclusion of time-varying waves. The mutual influences between waves and currents are investigated through an idealized continental shelf case and hindcastings of storm events in the sea area adjacent to Hong Kong in the northern South China Sea. Calculations are compared with other computed results and observations. Calculations show that the wave-dependent surface stress incorporated in the three-dimensional hydrodynamic model has significant impact on water surface velocities and surface elevations (over 10% higher). The inclusion of wave-dependent bottom stress also shows some effects; however, in the presence of the wave-dependent surface stress, its effect on surge levels becomes negligible. The effect of currents on waves amounts to the reduction of the significant wave height by about 8% and less for wave mean periods. However, the inclusion of the wave-dependent bottom stress in the three-dimensional hydrodynamic model has little effect on wave characteristics whether or not the wave-dependent surface stress is considered although it does reduce the bottom flow velocities, especially in shallow water regions. © 1997 Elsevier Science Ltd

Velocity shear and vertical mixing in the Ekman layer in the presence of a horizontal density gradient

The wind-driven turbulent Ekman layer is studied in the presence of a horizontal density gradient. The stabilizing case, i.e. when the wind brings light water out on top of heavier water, is treated using a 1-1/2 order turbulence model (a k-s model). It is found that the velocity shear within the mixed layer (ML) interacts with the horizontal density gradient and that an equilibrium ML depth is reached in a shorter time than the inertial period. The situation modelled bears some similarities to the competing effects of wind-mixing and surface heating. However, the stabilizing effect arises from the velocity shear, which, in combination with the horizontal density gradient, tends to create a stable stratification within the ML. The non-dimensional equilibrium ML depth and the velocity shear are shown to depend on L E/ L A (= vb∂b/ ∂χ| ∥ −2). [ L E = u ∗ ∥ −1 is the Ekman length and L A = u ∗, f |∂b/∂χ| −1 is a length introduced by J. Rodhe (1991) in the Journal of Physical Oceanography, 21, 1080–1083. u∗ is the friction velocity defined from the wind stress, f is the Coriolis parameter, b = g( ϱ 0 - ϱ)/ ϱ 0 is buoyancy ( g is gravity, ϱ is density and ϱ 0 is a reference density) and x is a coordinate in the direction of the horizontal buoyancy gradient.]

Experiments using a long-time-scale shelf circulation model of relevance to the Labrador Current

Experiments are described using a three-dimensional, shelf circulation model. The model geometry consists of a rectangle in latitude-longitude space with a shelf-slope region bordering the northern and western boundaries and a deep ocean region in the southeast. Relatively light water is flushed in through the northern boundary and allowed to exit through the southern boundary, a situation of relevance to the southward flowing Labrador Current. In an earlier paper, we showed the downstream development of a shelf break current. In that paper, bottom friction was parallel to bottom geostrophic velocity. In this paper, bottom friction is parallel to bottom velocity. This leads to a more diffuse downstream jet. We show that changing the density contrast across the front does not change its width. On the other hand, a sharper front is obtained when a small trough is introduced into the bottom topography. We also describe an experiment in which the density of the inflowing water is varied seasonally. This leads to a seasonal redistribution of the southward transport across the shelf, similar to a suggestion made by Myers et al. [(1989) Seasonal and interannual variability of the Labrador Current and West Greenland Current. Department of Fisheries and Oceans, Canada] for the Newfoundland Shelf. This redistribution results from the seasonal pulsing of fresh water down the shelf, which, in turn, influences transport through the Joint Effect of Baroclinicity And Relief (JEBAR), and is similar to the mechanism proposed by Lazier and Wright [(1993) Journal of Physical Oceanography, 23, 659–678]. Other results concern the splitting of the shelf break jet. We show that in the previous paper, the splitting of the jet was influenced by the numerical formulation of the outflow condition at the southern boundary. We also show that the splitting can be suppressed by specifying the density of water flowing into the model domain through the southern boundary, rather than allowing this to be determined by the previous history of mixing and outflow on the boundary.

Genesis of meteorological disturbances and thermohaline variability of the upper layers in the head of the Bay of Bengal during MONsoon Trough Boundary Layer EXperiment (MONTBLEX-90)

Time-series measurements of surface meteorological elements and near-surface thermohaline profiles made at a station at the head of the Bay of Bengal during the MONsoon Trough Boundary Layer EXperiment (MONTBLEX-90; August–September 1990) were utilized to describe and explain the observed short-term variability in the thermohaline structure during active and inactive regimes of the summer monsoon. The observed near-surface thermal structure showed the occurrence of distinct isolated mini-warm pockets with temperatures in excess of 29°C extending from the surface to about 30 m depth, with a temporal correspondence to the formation of meteorological disturbances over the Bay's head. Accumulation of heat in the upper layers in excess of 0.4 × 108 J m−2 (with respect to the 28.7°C isotherm) appeared to have triggered the formation of meteorological disturbances that extracted this surplus energy to lower the surface temperature to around 28.7°C. The depth-time sections of thermohaline fields suggest the signature of a transient eddy or propagating wave. The small-scale structure in the thermohaline profiles was more prominent during September 1990 compared to August 1990. Two simple one-dimensional numerical models following Kraus and Turner (1967, Tellus, 19, 98–106), Denman (1973, Journal of Physical Oceanography, 3, 173–189), Miller (1976, Journal of Physical Oceanography, 6, 29–35) and Niiler (1975, Journal of Marine Research, 33, 405–422), Niiler and Kraus (1977, Modelling and prediction of the upper layers of the ocean, Pergamon, pp. 143–172) were evaluated to assess the relative importance of the local atmospheric forcing to account the observed variability in the near-surface mixed layer characteristics.

What forces the variability of the southwestern Atlantic boundary currents?

A marked variability in the location of the front originating at the confluence of the Brazil and Malvinas Currents has been observed from both surface and subsurface observations. Modeling experiments using climatological winds predict a seasonal variability on the latitude of separation of the Brazil Current from the coast. During the Confluence Program (November 1988–February 1990) and from data collected with an array of inverted echo sounders, the location of the confluence front and its variability was established. In this paper, the observed oceanic variability is analyzed simultaneously with the wind product from the European Center for Medium Weather Forecast (ECMWF) obtained for the period of the observations. The ECMWF data is validated against in situ indirect wind magnitude observations obtained from a sub-array of the Confluence deployments. The large-scale anomalies are explored through the comparison with the climatological winds field obtained from Hellerman and Rosenstein (1983), Journal of Physical Oceanography, 13, 1093–1104. From the analysis it is concluded that the main source of variability of the Confluence front is the local wind forcing. There is a variability in the location of the front due to the seasonal cycle of the winds in the South Atlantic. In addition to this seasonal variability, the latitude of separation of the Brazil Current from the coast presents a marked interannual variability that is forced from anomalous wind patterns south of the Confluence. There is no apparent correlation between wind-forced pulses in the Antarctic Circumpolar Current and the observed anomalous northward penetration of the Malvinas Current.

Bottom boundary layer spectral dissipation estimates in the presence of wave motions

Turbulence measurements are an essential element of the Sediment TRansport Events on Shelves and Slopes experiment (STRESS). Sediment transport under waves is initiated within the wave boundary layer at the seabed, at most a few tens of centimeters deep. The suspended load is carried by turbulent diffusion above the wave boundary layer. Quantification of the turbulent diffusion active above the wave boundary layer requires estimates of shear stress or energy dissipation in the presence of oscillating flows. Measurements by Benthic Acoustic Stress Sensors of velocity fluctuations were used to derive the dissipation rate from the energy level of the spectral inertial range (the -5/3 spectrum). When the wave orbital velocity is of similar magnitude to the mean flow, kinematic effects on the estimation techniques of stress and dissipation must be included. Throughout the STRESS experiment there was always significant wave energy affecting the turbulent bottom boundary layer. Lumley and Terray [(1983) Journal of Physical Oceanography, 13, 2000–2007] presented a theory describing the effect of orbital motions on kinetic energy spectra. Their model is used here with observations of spectra taken within a turbulent boundary layer which is affected by wave motion. While their method was an explicit solution for circular wave orbits aligned with mean current we extrapolated it to the case of near bed horizontal motions, not aligned with the current. The necessity of accounting for wave orbital motion is demonstrated, but variability within the field setting limited our certainty of the improvement in accuracy the corrections afforded.

The effects of stratification and alongshore currents on the propagation of coastal-trapped waves

A hierarchy of models has been developed to describe coastal-trapped wave (CTW) propagation along continental shelves. In this paper, the results of four models are compared with respect to predictions of phase speeds for low-frequency non-dispersive CTWs propagating through the ACE (Australian Coastal Experiment) study area along the southeast Australian coast. The ACE experiment demonstrated that the stratification there significantly increased theoretical CTW phase speeds. At times substantial mean alongshore currents, apparently associated with meanders of the East Australian Current, were observed to intrude onto the continental shelf. Such currents can alter CTW propagation speeds by advection and by altering background vorticity if the current has shear. Accordingly, the models chosen here for intercomparison incorporate first, no stratification and no current [ Buchwald and Adams, (1968) Proceedings of the Royal Society Series A, 305, 235–250 ]; secondly, barotropic currents and no stratification [ Niiler and Mysak (1971), Geophysical Fluid Dynamics, 2, 273–288] ; thirdly, vertical stratification but no currents [ Huthnance (1978), Journal of Physical Oceanography, 8, 74–92 ]; and fourthly, full stratification and baroclinic currents [ Luther and Bane (1985) Journal of Physical Oceanography, 15, 3–23 ]. The stratification and currents used in these models were determined from CTD casts and ship drift data obtained during six ACE cruises. Between the six cruises the oceanographic conditions on the shelf and offshore were observed to undergo considerable variation. Hence, it has been possible to compare the models over a wide range of parametric variations. The results show that there is considerable variation in the predicted phase speeds between models.

On the role of shear mixing during transient coastal upwelling

A two-dimensional two-layer model for wind-driven transient coastal upwelling is formulated. The momentum equations include the turbulent dynamics and the time-dependent and nonlinear terms in both the cross- and along-shore directions. The continuity and heat equations allow mass and heat turbulent transfer between both layers. The integral form of the momentum, continuity and heat equations are closed using a two-regime parameterization for the entrainment velocity. In the first regime, corresponding to the early stages of upwelling, the interface quickly raises due to flux divergence near the coast. The entrainment velocity is small (0.1-1 m day −1, largely produced through K raus and T urner's [(1967) Tellus, 19, 98–106] slow erosion of the thermocline, and it is estimated using N iiler and K raus' [(1977) In: Modelling and prediction of the upper layers of the ocean, Pergamon Press, Oxford, pp. 143–172] parameterization. When the bulk Richardson number ( Ri) becomes close to its critical value then we switch to the second regime, during which we calculate the entrainment velocity from the continuity equation under the condition that Ri remains near-critical, i.e. the equivalent of P ollard et al. [(1973) Geophysical Fluid Dynamics, 4, 381–404] stability criterion for the upper ocean. The entrainment velocity quickly becomes large (several m h −1), the interface deepens and stratification is eroded. The existence of this regime is supported by observations of persistent near-critical gradient Richardson numbers ( Ri g ) during coastal upwelling [J ohnson (1981) In: Coastal upwelling, American Geophysical Union, Washington, DC., pp. 79–86; J ohnson et al. (1976) Journal of Physical Oceanography, 6, 556–574; K undu and B eardsley, (1991) Journal of Geophysical Research, 96, 4855–4868]. Our model is applied to several initial temperature differences between the surface and bottom layers, with the upper layer depth and forcing parameters realistically chosen. The dynamically important mixing regime corresponds to the second regime, with effective shear-induced mixing being produced through a strong baroclinic coastal jet. A realistic front, formed between the well-mixed water near the coast and lighter offshore surface water, propagates away from the coast. The offshore waters are characterized by the presence of inertial oscillations, overlying the Ekman flow. The inertial oscillations are too weak to produce any significant mixing, but a comparison with de S zoeke and R ichman's [(1984) Journal of Physical Oceanography, 14, 364–377] semigeostrophic model (modified to include the shear-mixing regime) shows that they are important enough to exert some control on the horizontal volume flux divergence near the coast. A relatively fast internal Poincare wave, propagating from the coast, has the effect of slowly dampening the inertial oscillations. The results are in good qualitative agreement with early observations by J ohnson et al. (1976).

Mixing in the Brazil-Malvinas confluence

Conductivity-temperature-depth profile data from the Western Argentine Basin collected from 1984 to 1989 are used to quantify the cross-front heat and salt transfers associated with the vertical finestructure across the Brazil-Malvinas Confluence. The fluxes are estimated following the statistical model of Joyce ( Journal of Physical Oceanography, 7, 626–629, 1977). The data indicate that the upper ocean cross-front structure of the large-scale temperature and salinity fields is constant. The medium-scale finestructure intensity is quantified by the variance of the vertical temperature and salinity gradients in the 10–100 m wavelentgh band. Due to the abundance of intrusions, the upper layer (0–1000 m) variances increase by a factor of four at distances <20 km from the front. Heat and salt flux estimates associated with medium-scale mixing in the upper ocean are of the order 10 −2oC m s −1 and 10 −3 m s −1 respectively. These fluxes are an order of magnitude greater than availabe estimates for other frontal regions. The medium-scale finestructure may therefore play a key role in the dissipation of eddies and intrusive lenses in the region. Heat and salt fluxes between North Atlantic Deep Water and Circumpolar Deep Water are 6.5 × 10 −4 °C m s −1 and 1.8 × 10 −4 m s −1, and agree with existing estimates. Extrapolation of upper layer Brazil-Malvinas Confluence cross-frontal fluxes to the Subtropical Convergence across the South Atlantic suggests that the medium-scale southward heat flux is about 20% of the oceanic northward heat flux at 30°S. Similarly, the freshwater flux balances 20% of the excess evaporation north of 30°S.

Topographic effects on the bottom water circulation of the western tropical North Atlantic Ocean

A simple numerical model of the abyssal circulation of the ocean is used to explore the possibility, first suggested by Warren (MIT Press, pp. 6–41, 1981), that the northward deepening of the sea floor in the western basin of the tropical North Atlantic Ocean is responsible for the observed eastward confinement of the Antartic Bottom Water current, a result contrary to what the Stommel-Arons model predicts. It is shown that (a) a topographic slope of magnitude observed in this region of the Atlantic forces separation of the northward-flowing deep western boundary current at a low latitude, (b) north of this latitude the circulation is dominated by a large-scale cyclonic gyre extending up to where the sea floor becomes flat again, with a northward flow along the eastern boundary as a part of it and a westward flow at about 25°N. Simulated Lagrangian particle trajectories show that new abyssal water from the south is confined to a band next to the eastern boundary in the gyre region as observed. These results are robust under variation of topographic, forcing and frictional parameters, although quantitative aspects and details are sensitive to them. Eastward flow through the Vema Fracture Zone, recently estimated by McCartney et al. ( Journal of Physical Oceanography, 21, 1089–1121, 1991) to be about 2 Sv, diverts some of the northward-flowing eastern boundry current, but otherwise does not affect the circulation. These results compare favorably with recent observations that report southeastward flow along the western boundary at least down to 8°N and suggest a recirculation of abyssal water in this region, as well as with those that suggest a northern limit of the Antarctic Bottom Water flow along the eastern boundary at about 25°N and a possible westward separation there. An analytical solution of the planetary geostrophic equation applied to an abyssal ocean with topography, which captures some features of the numerical model flow, is presented in the Appendix.

On the transport of the gulf stream between cape hatteras and the grand banks

Direct velocity observations from four locations in the Gulf Stream between Cape Hatteras and the Grand Banks are analysed to determine the baroclinic and barotropic transport of the Stream. Three of the sites are moored arrays where a method similar to that proposed by Hall and Bryden (1985, Geophysical Research Letters, 12, 203–206) and Hall (1986, Journal of Physical Oceanography, 16, 1814–1818) is used; the fourth is the Pegasus site of Halkin and Rossby (1985, Journal of Physical Oceanography, 15, 1439–1452). The baroclinic transport is found to change little from site to site while the barotropic changes substantially. Total transport reaches 150 Sv at 60°W and remains at that strength to 55°W, the farthest east that moored information is available. These results, augmented by additional information, are used to deduce a scheme for the total transport steamfunction in the western North Atlantic.

Evidence of internal swash associated with Sulu Sea solitary waves?

Vertical temperature profiles were measured near the shelf edge off the coast of Palawan Island (Philippines) during the passage of several “rip bands” of choppy surface waves. Such rip bands are commonly associated with internal waves. Coincident with the passage of one of the rip bands, the temperature near the bottom decreased by 2.5°C within 1.4 min, becoming colder than any water on the shelf. Furthermore, the apparent depth of our CTD decreased by about 8 m despite the fact that the “line-out” was held fixed. We interpret these limited data, along with a sea-level record, as evidence of internal swash created by breaking internal solitary waves which are generated by tidal flow over a shallow bank in the southeastern Sulu Sea (described by Apel et al., 1985 , Journal of Physical Oceanography, 15, 1625–1651).

Fluid transport through Bass Strait

The flux of water through Bass Strait was measured with two sections of current meter moorings placed in the west of Bass Strait between Cape Otway and King Island, and King Island and Hunter Island, in the autumn and winter of 1984. The flux was found to be highly correlated with the local wind stress and the time gauges on the northern boundary. Mean flux over the period of the experiment was 0.49 Sv from west to east, with a maximum of +3.05 Sv and a minimum of –0.8 Sv. Much of this flux occurred in surges lasting 2–3 days following the passage of atmospheric cold fronts across the region. The variations in flux may be monitored by a single tide gauge on the eastern Victorian coast, at least in winter. From dynamical considerations it is inferred from the available data that the through-flow is forced by a combination of forcing to the west of Bass Strait and wind stress within Bass Strait in the approximate ratio of 3:1. It seems probable that these surges are the principal cause of the coastally trapped waves observed in the Australian Coastal Experiment of 1983–1984 ( H. J. Freeland et al. (1986) The Australian Coastal Experiment: A search for coastal trapped waves. Journal of Physical Oceanography, 13, 1230–1249). It appears that the geometry of Bass Strait causes singularities in linear coastally trapped wave theory in the south-west and north-east corners, and it is probable that this phenomenon is manifested in the stronger currents observed in the south-west than at the other stations further north.