Sea bottom topography (bathymetry) exerts significant control over large scale (of scales, say, > 10 km) ocean circulation. A few examples include the steering of boundary currents to flow along continental slopes, e.g., the Gulf Stream, or the California Undercurrent; blocking of currents (even away from the continental slope, and even for surface-intensified currents in some cases) from crossing elevated regions; the formation of rotating eddies ("Taylor Caps") above seamount tops; and a tendency for eddy drift accross large-scale bottom slopes. All the effects mentioned above (though not all topographic effects on currents in general) have to do with the Coriolis effect, as was long ago understood. Yet these and other topographic effects on ocean circulation manifest in myriad ways, and there is much to understand still. Topographic effects were at the center of several of my PhD projects, and they still are. Some of this work is discussed in this page about the DWBC leakiness project, and some in the page on the ongoing work regarding southern ocean overturning circulation. Below, I discuss two additional projects about topographic effects on circulation.
On the role of bottom pressure torques in wind-driven gyres
Topography has also been previously shown to play a large role in basin-scale wind-driven circulation in the ocean, i.e., wind gyres. Specifically, bottom pressure torques (BPT), arising from pressue forces between deep geostrophic flow and topography, have been shown to approximately balance wind stress-curl within zonal strips. In this paper, we looked at the problem from a different angle, since bottom pressure torque vanishes exactly when integrated within an isobath. As shown in the paper using a series of idealized experiments, this implies that non-local circulation and BPT, must result from the compensation of wind stress curl by BPT in the gyre interior. It also means that bottom stress plays a significant role together with BPT, e.g., bottom (frictional) stress becomes important over the continental shelf, and in balancing the non-local BPT response. The relative importance of BPT vs bottom stress in the gyre is quantified, and its depndence on parameters is explained by extending previous theory.
Formation of Anticyclones above Topographic Depressions
While I was working on the DWBC leakiness problem in the Newfoundland Basin, I became aware of the presence of the so called Mann Eddy in the same region. This anticylonic eddy has been first observed in 1967 (by C.R. Mann), has been repeatedly observed since then, and appears to be a quasi-permanent coherent eddy. I became interested in its sustaining mechanism. In a literature survey, we found there are a few additional examples of mesoscale anticyclonic eddies (ACs) repeatedly observed over decades in other locations, such as the Rockall Trough Eddy, the Lofoten Basin Eddy, and Iceland Basin anticyclones. The best studied example is the Lofoten Eddy in the Lofoten Basin of the Nordic Seas. These eddies occur in regions of intense themohaline transport and/or transformations, and understanding a potential role or interaction with these processes is of importance.
The long-liveness of each of the known ACs was previsouly suggested to be due to somewhat different mechanisms (e.g., baroclinic instability, bottom friction, topographic eddy drift, eddy shedding from a turning current) which caused a concentration of negative vorticity material in the eddy location. We noticed that all of these eddies however, occured above wide topographic depressions. That led us to hypothesize that topographic effects specificly related to the presence of a wide, bowl-like, depression may be at play here. This started an investigation that we eventuall published in 2021. Indeed such dynamics were perviously hypothesized and demonstrated to be the case for the Lofoten Eddy. Specifically, ACs first drift down the topographic slope and accumulate above the depressions, and secondly merge to form or sustain the long-lived AC. Drift due to the topographic gradient ("topographic beta drift") contributes to the convergence, although in the case of the Lofoten eddy, "planetary beta drift" (due to the curvature of the earth rather than due to topography) also contributes to the motion of ACs shed from the Norwegian Atlantic Current towards the Lofoten Basin.
To address our hypothesis, we conducted numerical experiments on the spontaneous formation of circulation from random initial conditions above topographic depressions in idealized conditions (in one or several isopycnal layers). The simplicity of the model employed allowed examining a wide parameter regime, including various topographic forms and parameters, scale and amplitude of initial conditions etc. The first finding was that in fact long-lived ACs emerge spontaneously above wide topographic depressions in our experiments, under a wide range of sensible parameter values. The formation generally occured whenever eddy-eddy interaction was weak enough relative to eddy-topograpy effects, as we quantified via a nondimensional parameter. Wide bowls, relative to eddy sizes, were necessary for the convergence of ACs into the bowl to occur, i.e., relatively uniform slopes are necessary on the eddy scale; but we also found the eddy-eddy interaction (in dipole pairs) can cause even faster AC transport towards the bowl. The vertical structure of the trapped ACs was also studied. In observations, bowl-trapped ACs are top-intensified, although with a significant deep component. We find that model (in theory and in the numerical simulation) trapped-ACs have a vertical structure related to that of the original ACs which form it. That is in contrast with Taylor Columns (forming above, e.g., steep seamounts under mean large-scale flow), which are allways bottom intensified.
Evolution of voriticy above a topographic bowl (isobaths shown in black), from a random initial condition. Over time, anticyclones (cyclones) migrate into (outside of) the bowl's central plateau. There the anticyclones merge and form a very long lived (at least a decade in the simulation shown) anticyclone.
Realistic modeling of the Mann Eddy (ongoing work)
In addition to the idealized modeling of ACs above topographic depressions, I am engaged in realistic modeling of the Mann Eddy (described above). Little work has been done regarding the dynamics sustaining the Mann Eddy, although it has been suggested it is spawned from ACs shed from the North Atlantic Current (NAC). The anticyclonic Mann Eddy lies adjacent to the right of the trajectory of the NAC, and hence their transports are difficult to distinguish in observations. Furthermore, the deep water masses of the NAC change in that region, and there is indirect evidence (based on tracer concentrations) in the literature that the Man Eddy has a role in this process, which is part of the Atlantic Meridional Overturning Circulation. The Mann Eddy region is approximately the boundary between the subtropical and subpolar gyres. As such, it hosts large gradients and variability in properties. Previous work has also shown that the North Atlantic Oscillation, a major climate variability mode, has perhaps its strongest signature in the Mann Eddy region as well. All of these processes beg for a better understanding of Mann Eddy dynamics and its connection to other properties of this turbulent region of the ocean.
Therefore we have been examining Mann Eddy properties and dynamics within a high resolution regional simulation, as well as in observations. The first results have confirmed that the Mann Eddy is a coherent and long-lived AC, rather than a diffuse or "statistical" recirculation pattern. It is not completely stationary, but moves within the topographic depression in the center of the Nefoundland Basin. Composites based on the model show that the eddy is deep reaching, with a non-zero expression at the seabed. Its maximal amplitude occurs at a depth of approximately 500 m. Mergers with impinging ACs do occur repeatadly, and seem to sustain the Mann Eddy. All these features appear generally consistent with the ideazlied modeling results described above. The work continues...