“Pickart and Hogg (1989) use a model consisting of a simplified gyre, representing the northern recirculation gyre of the deep Gulf Stream, next to a deep boundary current to study the transfer of tracer via diffusion from the boundary current to the recirculation gyre.”
“Spall (1996) uses a primitive equation model with three layers to model the Gulf Stream/DWBC crossover region. He introduces a DWBC in both the second and third layers (the ‘‘upper’’ and ‘‘lower’’ DWBCs) and finds that the upper core of the DWBC (in layer 2) splits into two mean paths as it crosses under the model Gulf Stream. One path flows toward the south along the western boundary while the other is turned offshore under the Gulf Stream. The lower DWBC (in layer 3) continues to flow along the western boundary with very little interaction with the interior. Spall also finds that the strength of the Gulf Stream has a direct impact on the strength of interaction between the upper DWBC and the interior.”
“[Thompson and Schmitz (1989)] demonstrated in a two-layer primitive equation model that the location of the Gulf Stream separation point is shifted southward by the presence of a DWBC. In the absence of the DWBC, the eastward jet separates from the western boundary at exactly mid-basin.”
See Lozier and Riser (1989) for a study of deep recirculation gyres without DWBC?
“In the flat-bottom configuration, the mean DWBC almost completely turns offshore (at approximately y = 2500 km) while in the topography configuration, the mean DWBC is able to remain on the western boundary and continue southward as it crosses under the surface jet. This is because, in the topography configuration, the DWBC is able to conserve potential vorticity by moving up the slope to partially offset the reduction in the Coriolis parameter as it flows southward. In the flat-bottom configuration, the DWBC does not have a slope to move up and thus turns offshore in an attempt to conserve potential vorticity.”
“One striking feature of the tracer contours at the end of a 5000-day run (Fig. 7) is the lack of significant amounts of tracer south of the midbasin f/h contour. This indicates an effective barrier to meridional transport of tracer across this f/h contour. The barrier can be explained in terms of potential vorticity conservation. The surface jet causes the thickness of the upper layer to increase, thus causing a decrease in the thickness of the lower layers. In order to maintain a constant layer thickness, the southward flowing DWBC has to turn offshore as it crosses under the surface jet (Hogg and Stommel 1985; Spall 1996).”
“Because these results are so similar, we feel that for a reasonable range of lateral dissipation coefficients, the results of this study are not highly dependent on [horizontal dissipation].”