Laboratory studies of the effects of interrupted, sloping topography on intermediate depth boundary currents in linearly stratified fluids

https://doi.org/10.1016/S0377-0265(00)00064-6Get rights and content

Abstract

Laboratory experiments are described which provide insight into the interaction of intermediate depth boundary currents (IDBCs) with interrupted sloping topography. Specifically, they contribute to the debate over meddy formation on the Iberian continental slope. The experiments were performed in a rectilinear rotating tank filled initially with a linearly-stratified fluid. A false bottom sloped away from the side-wall along which the current flowed, and was interrupted by a gap of variable length. The effects of varying gap length and rotation rate on the boundary current were observed.

In the first of two sets of experiments, the current flowed above the slope, along the vertical sidewall. In the second, the current flowed along the sloping bottom. In the former, current nose speed was consistent with geostrophic predictions, but decreased in the presence of a gap in the topography. Kelvin wave radiation is postulated as a reason for this. The IDBCs exhibited vortical lateral intrusions at values of the Burger number Bu=(N0/Ω)2 at which counterpart flat-bottom studies had been stable, implying that the sloping topography had a de-stabilising effect. Energy measurements and qualitative observations suggest the intrusions were due to mixed barotropic/baroclinic instabilities, the latter dominating at higher rotation rates.

In the second configuration, four distinct flows were observed, distinguished by the deformation radius:gap width ratio RD/G. For a range of values of RD/G, attached eddies formed at the upstream end of the gap. They remained at this position, unlike those in similar studies of surface boundary currents (Klinger, 1993). Their persistence and ability to move downstream  salient factors for meddy  formation were greater for a finite gap size than a permanent change from sloping to flat bottom.

Introduction

Boundary currents are common in both the oceans and atmosphere. They occur when three conditions are met, namely: a boundary exists that is not aligned normal to the local gravitational potential gradient; this boundary contains a fluid in which density is non-uniform; and background rotation is dynamically important. In these situations, the fluid reaches dynamic equilibrium when a flow of anomalous density is set up parallel to the boundary.

The present work considers intermediate depth boundary currents (IDBCs herein). These have a density which lies between the extremes of density found in the ambient fluid, and flow at their level of neutral buoyancy. The simplest case is that of a two-layer ambient in which the IDBC density is the mean of the two-layer densities (Fig. 1a). Typically, however, in the environment, the ambient is continuously-stratified (Fig. 1b).

IDBCs are often formed when a fluid is disgorged into a stratified regime at a level where its density is not that of the ambient. The resulting intrusive flow mixes with the surrounding fluid, altering its density and depth until the former corresponds to the ambient density at the latter. If a boundary is present, the anomalous fluid forms a current along it, acting under the same forces which produce surface boundary currents (see for example, Griffiths, 1986). A prime example of this is found in the gulf of Cadiz, where Levantine intermediate water (LIW) and western Mediterranean deep water (WMDW) flow westward into the Atlantic ocean beneath a surface counter-flow (e.g. Ambar et al., 1976; Baringer and Price, 1997a, Baringer and Price, 1997b). After surmounting the shallow sills in the strait of Gibraltar, these anomalously-dense waters plunge into the basin of the gulf. Ultimately, they reach their level of neutral buoyancy at ∼1200 m and form an IDBC along the southern Iberian coast.

The boundaries along which oceanic IDBCs flow are usually continental slopes. These are often interrupted by submarine canyons that are likely to affect a current’s passage. Examples of this topographic constraint are the Portimao Canyon on the southern Iberian coast (Prater and Sanford, 1994) and the Sakarya Canyon in the Black Sea (Ozsoy et al., 1993). The need to understand these complex situations is taken as motivation for the work presented. Further motivation comes from the need to improve general understanding of IDBCs (see for example, Davies et al., 1991), particularly their de-stabilisation.

Section snippets

Previous work

Buoyancy-driven boundary currents have been researched intensively (Griffiths, 1986, gives a review of fundamental understanding of their dynamics). Most of this work concerns surface or bottom currents — the most common configurations in nature — although important work on IDBCs has been reported. Relevant studies of current instability and topographic interaction — the topics of interest here — are now reviewed to provide a context for the results presented.

Surface boundary currents (SBCs)

Apparatus

The experimental apparatus is shown schematically in Fig. 2. The working chamber — a perspex channel — measured 2.10m×0.46m in plan-view cross-section and 0.31 m in depth. A rigid lid 0.24 m above the bottom covered the stratified fluid. At one end of the tank a grid was suspended at an intermediate depth. The shaft of the grid was attached, off-centre, to a motor-driven rotating wheel causing the grid to oscillate vertically with an amplitude of ∼15 mm. The grid consisted of square elements made

General observations

Examples of the flows observed are shown in Fig. 5. The illustrated runs used the solder visualisation method, and illustrate the sharp definition achieved. Fig. 5a shows a typical configuration for Bu=59.3. The boundary current is characterised by two laterally-growing, vortical features (A and B). These begin to develop as soon as the boundary current is formed, apparently growing from perturbations caused when the mixed patch impinges on the side wall.

For Bu=14.8 runs (Fig. 5b), the lateral

General observations

Examples of Case II flows are shown in Fig. 12. In all cases, the boundary current (B in Fig. 12a) is observed separating from the slope at the upstream edge of the gap (U in Fig. 12a), and moving to towards the wall (since the Coriolis force normal to the wall becomes unrestrained at that point). Initially, this causes the current to form an eddy that remains at the upstream end of the gap. Throughout the Case II runs, strong vertical oscillations were observed in the current when it separated

Case I runs

IDBCs were studied flowing along a vertical sidewall over sloping topography interrupted by gaps revealing a flat bottom. The following deductions were made.

  • The boundary current flowed with a nose speed consistent with that predicted by geostrophic balance in all cases. The nose speed/geostrophic speed ratio was independent of background rotation rate, but decreased in the presence of a finite gap in the bottom topography. Radiation of energy by a Kelvin wave is postulated as a reason for this

Acknowledgements

The authors acknowledge the support of the UK Natural Environment Research Council (NERC) through the provision of a postgraduate studentship to AMF. Stimulating discussions with Dominique Renouard and Gabriel Chabert d’Hières at the Coriolis Laboratory, Grenoble and Joe Fernando at Arizona State University were extremely beneficial for the progress of the study and the help and generosity of these individuals is acknowledged with gratitude. The detailed comments of two anonymous referees are

References (35)

  • M.O. Baringer et al.

    Momentum and energy balance of the Mediterranean outflow

    J. Phys. Oceanogr.

    (1997)
  • Besley, P., 1989. Eddies and turbulence in stratified fluids an experimental investigation. Ph.D. Thesis, University of...
  • G. Chabert d’Hières et al.

    A laboratory study of boundary currents: application to the Algerian current

    J. Geophys. Res. (Oceans)

    (1991)
  • S.A. Condie et al.

    Convectively driven coastal currents in a rotating basin

    J. Marine Res.

    (1988)
  • S.A. Condie

    Descent of dense water masses along continental slopes

    J. Marine Res.

    (1995)
  • E. D’Asaro

    Generation of submesoscale vortices: a new mechanism

    J. Geophys. Res. (Oceans)

    (1988)
  • P.A. Davies et al.

    Generation and spreading of a turbulent mixed layer in a rotating stratified fluid

    J. Geophys. Res. (Oceans)

    (1991)
  • Cited by (1)

    View full text