Collaborative Research: Deep Circulation over the Flanks of a Mid-Ocean Ridge
The deep ocean, below the crests of the mid ocean ridges, is a major global reservoir for heat and carbon dioxide, and thus plays an important role in governing the earth's climate. These properties enter the abyssal ocean in plumes of dense water formed at the surface near the poles, and are exchanged throughout the ocean by turbulent diffusion with overlying layers. Deep water flows back toward the poles at lower densities than the plume water. To close the circulation, buoyancy of bottom water must be increased by turbulent diffusion of buoyancy downward, assisted by geothermal heating. Turbulent diffusion has been found to be strongly enhanced over regions of rough topography, but to decrease strongly with height above the bottom. This results, paradoxically, in buoyancy loss in most of the deep ocean, but recent theoretical work suggests that this loss is more than compensated globally by buoyancy gain in a bottom layer and upward flow along the sides of the basins. An example of the processes involved was suggested by an experiment conducted in the Brazil Basin in the 1990's. Turbulence profiles and the evolution of a tracer released above a fracture zone canyon on the flanks of the Mid-Atlantic Ridge showed increased diffusion toward the bottom. Furthermore, though the main tracer patch, which was aloft of most topography, moved poleward and westward, much of the deeper tracer appeared to be drawn eastward and into the canyon, suggesting vigorous mixing in that area. We propose to simulate the evolution of the tracer patch at 2.5-km resolution to understand and quantify the general circulation and lateral eddy stirring, and especially to test measurement-based hypotheses about the distribution of turbulent buoyancy fluxes. Since measurements suggest that buoyancy fluxes in the region are dominated by processes at sills across the canyons, it is necessary to simulate the flow in the region of a well-measured sill at high (250-m) resolution to test hypotheses of how the turbulent buoyancy flux is distributed there in detail. The kind of forcing and bathymetry to be studied in this project typify the flanks of the mid-ocean ridge throughout the Atlantic and much of the Indian and Pacific Oceans. This study will be relevant to a large fraction of the area of the ocean, and will improve our current understanding of the return path of the deep overturning circulation and its role in governing the earth's climate. Generation of turbulence over rough topography and propagation into the interior are complex processes, with many potentially relevant mechanisms of creating velocity shear and large amplitude displacements that provide the ultimate forcing: e.g. internal tides; lee waves; sill flows; and even wind-generated inertial waves. Recent work suggests that this turbulence drives strong upwelling along the sloping ocean topography. The observations collected as part of field programs in the Brazil Basin will provide key information to test the hypothesis. Numerical models will be run to help the investigators interpret the observations and test their theoretical ideas. The main goal is to bring together in a unified picture the numerous Brazil Basin observations, none of which by itself paints a full picture of the abyssal circulation in the Brazil Basin. For example, the sampling of the tracer during Brazil Basin Tracer Release Experiment (BBTRE) and subsequent campaigns was by necessity incomplete. The vertical profiles of microstructure were limited in their spatial and temporal coverage. Mooring velocities were collected at only a few locations. Float data offered a glimpse of the trajectories of a few water parcels. Two sets of numerical simulations will be used in this study. A 2.5-km resolution simulation centered over the region sampled during BBTRE configured to study the dispersion of the tracer, and a 250-m resolution patch embedded in the larger simulation, centered on the BBTRE Canyon to study in detail the diapycnal buoyancy flux and upwelling which appear to be largest along the canyon-and-hills bathymetry.