The structure and transport of the Brazil Current between 27⁰ and 26⁰ South

Abstract

A set of four hydrographic sections through the Brazil Current are analyzed to identify downstream changes in the Brazil Current. The data, from the Thomas Washington Marathon Cruise, Leg 9, are at 27, 31, 34 and 36°S. The region they span details the change of the current from a relatively small near surface feature to a large, deep current. While the Brazil Current does not appear to develop transports as large as those found in the Gulf Stream, the calculated transports greatly exceed previous estimates. At 27°S the current extends down to approximately 700 m and transports 12 Sv southward; this value is consistent with previous estimates farther north. Downstream, surface layer transport increases, the current deepens, and the transport reaches a maximum of 80 Sv at 36°S. Part of the growth comes from the tight recirculation found just offshore of the Brazil Current. The recirculation strengthens and deepens to the south, with a minimum transport of 4 Sv north at 27°S and a maximum of 33 Sv at 36°S. The change in the current is also reflected in its shear profiles. At 27°S Brazil Current shear is found only in the upper portion of the water column, over the continental slope. Downstream, the current moves off the slope into deeper water and develops top-to-bottom, monotonic shear. To obtain velocity from the shear profiles, zero velocity surfaces are chosen based on conservative use of tracer information. A simple basin-wide model is used at 31°S to tie limits on the size of the Brazil Current and recirculation to various limits on layer-to-layer exchanges south of the section. The multi-layer model - based on changes in depth of several isotherms is used to extend the interpretation of the current beyond that of an isolated ocean feature. The model is required to conserve mass in each layer, either by applying barotropic transports or by allowing layer-to-layer exchanges south of the section. Solutions are deemed acceptable if the sense, or direction, of the various layer-to-layer conversions are consistent with accepted ideas of water mass formation. Initially, a two layer model is employed. Governed by the conservation of mass in each layer, the two layer model has only one constraint on the resulting solutions: a conversion of cold-to-warm water in the south (or the surface layer flowing north and the deep layer flowing south). Such a meridional flow pattern is consistent with the equatorward heat flux in the South Atlantic. The single constraint, however, is not strong enough to limit the solution region in any significant way. The final model presented has four layers, and acceptable solutions have the net transports of the surface layer and the bottom water northward and form intermediate water from North Atlantic Deep Water in the south. The resulting solution set has a fairly small range of transports for the Brazil Current, with surface layer transports between 20 and 35 Sv; this range is consistent with the value calculated from hydrographic data at 31°S. Given the complex interleavings of the South Atlantic water masses, the four layer model performs remarkably well. Finally, total potential vorticity is calculated from the hydrographic sections. Contrary to what one might expect, the reference level choice is not a significant problem: where currents are large, most of the signal in relative potential vorticity comes from the measured shear, and where currents are small, the relative potential vorticity is not significant compared to the planetary vorticity. Unfortunately, the process of taking two horizontal derivatives of the density field results in a jittery relative potential vorticity signal. As a result, a clear potential vorticity profile could not be constructed for the current. This variablitiy may be real -the ocean is frequently much noisier than one imagines. It may also be possible, though, to smooth the data sufficiently so that a cleaner picture emerges. Despite the problems involved in obtaining a quantitative profile of the potential vorticity, qualitative changes are useful in detecting different flow regimes. By comparing the downstream changes in total and planetary potential vorticity, one can deduce frictional and inertial regimes in the different layers. The presence of a frictional regime at the inshore edge suggests that care should be taken in assuming that potential vorticity is conserved in western boundary currents. In addition the potential vorticity sections trace a pattern of the recirculation feeding into the Brazil Current in the upper layers; other tracers did not provide a clear image. The final picture which emerges is not of a small, surface-trapped Brazil Current; rather, it is that of a classic western boundary current, increasing in strength and depth before turning east into the interior ocean.