Towards a model of Atlantic Ocean circulation: the plumbing of the climate's radiator.
Those of us who make ocean circulation observations often get involved in a different sort of model. While in some sheer numerical sense there are a lot of ocean circulation measurements, the circulation is remarkably complex, so the measurements are actually rather sparse relative to what is needed to describe the phenomena unambiguously. Thus we also become modelers, for even a basic description of ocean circulation is a model: an attempt to provide in words and pictures a synthesis, or simplified interpretation, of the measurements.
One such model, the popular "conveyor belt" interpretation (overleaf), has emerged from the increasing focus (over roughly the last decade) on the ocean's role in global climate. Its underpinnings are, however, traced to some of the oldest observations and their early interpretations as expressed by Benjamin Thompson, Count of Rumford, in 1797 and elaborated by William Carpenter and Joseph Prestwich in the 1870s. This very simple conveyor-belt model is based on a few fundamental circulation observations:
* Cold water with identifiable North Atlantic characteristics can be traced in diluted form through most of the world ocean.
* To conserve mass, an equal volume of warm water must replenish the cold water draining out of the North Atlantic.
* There is a large liberation of heat from the ocean to the atmosphere at middle and high North Atlantic latitudes.
The model, as far as it goes, is consistent with these observations. In particular, the estimated replenishment rate, the estimated intensity of heat loss, and the observed temperature change of the warm-to-cold conversions are in harmony.
An observationist builds on a simple model like the conveyor belt by checking its consistency with observations other than the ones on which it was based. The broader the base of such consistency, the more "real" the model is perceived to be, and the more valuable it becomes as a benchmark for testing a theoretician's model or as a guide for theoretical models--for example, what is the circulation that the model should duplicate? But when consistency with observations is lacking, the observationist must augment the simple model by building in missing pieces--or perhaps discard it as a flawed starting point.
The figure opposite is an example of an augmented version of part of the conveyor-belt model. WHOI Senior Scientist Bill Schmitz and I developed it over the last few years to describe the field of horizontal warm-water circulation in the North Atlantic. Such circulation schematics attempts to integrate all relevant measurements and interpretations, in this case both our own and those of our many colleagues around the world, present and past. Such circulation schematics have an even longer history than the conveyor-belt interpretation, for we follow the footsteps of centuries of ocean explorers who have tried to make maps of Atlantic Ocean surface circulation. One of the oldest of these, by the Jesuit Athanasius Kircher, dates from 1678.
It is important to recognize two things about circulation schematics. First, they are almost instantly rendered obsolete by new measurements, for our science is still in a basic exploration phase. Second, they represent a subjective synthesis: Other oceanographers might emphasize different aspects, that is, draw the picture and estimate the transport amplitudes somewhat differently--the product is very much "in the eye of the beholder!"
The schematic opposite points out a problem inherent in the process of sorting out ocean circulation's role in climate. We show 13 million cubic meters per second of warm water entering the North Atlantic across the equator; this is our estimate of the magnitude of cold-water production in the North Atlantic and thus of the intensity of the conveyor belt. However, the Gulf Stream south of Nova Scotia transports about six times this amount! The "recirculating gyre" components of measured North Atlantic warm-water transport thus greatly exceed the amount that is estimated to convert from warm to cold. (The recirculating gyre can be seen in the figure opposite as the Gulf Stream water that moves across and down the middle of the Atlantic to rejoin the northward flowing currents.) Thus, the part of the circulation field most important for climate issues is a small fraction of the field actually being measured. The physical phenomenon responsible for the northward transport of warm water through the North Atlantic is a linked set of gyre flows rather than the simple image of the upper limb of the conveyor belt. The physics of the western intensification of these gyres was first deduced by Henry Stommel in 1948, and their relationship to the wind and thermal and hydrological forcing remains a very active research topic to this day. The synthesis of observations, their interpretation, and their physical modeling proceed in parallel.
Similar degrees of complexity are emerging from studies of the cold-water part of the system, studies that illustrate the model-building process. The dominant components of the meridional flow of cold water are western intensified both within the Atlantic and within its subbasins, which are defined by abyssal ridges. In Kircher's old circulation-system map on page 1, water was imagined to descend into subterranean passages and to reemerge in faraway places. The real plumbing of the system is internal to the ocean itself, but it achieves a similar result: Waters sink in a few places to mid-depth or the seafloor (but not with the violence of maelstroms!) and are carried by current systems to the far reaches of the world ocean by deep currents. The width of these currents is restricted by the internal physics of the ocean and steered by the abyssal topography (not subterranean passages!)
The physics of this deep western intensification was first elucidated by Henry Stommel in the mid 1950s. Such intensified flows are "easier" to measure because they are geographically small. Measurements of the intensity of these southward flows show a magnitude often twice or more than that expected from the simple conveyor-belt interpretations. Flow measurements in the basins' interior regions have revealed that a gyre component is responsible for the larger-than-expected boundary current flows. This system of gyres is more broken up by abyssal bathymetry than are the large warm-water gyres. This model preserves the intensity of the conveyor-belt conception's net meridional flows of cold water, but it is also consistent with measurements of the intensity of deep western boundary currents and interior flows. Combining this cold-water schematic with the warm-water circulation of the figure below gives a representation of the full three-dimensional ocean circulation that the quasi-two-dimensional conveyor belt simplifies. Perhaps it is more aptly described as a baggage carousel than a conveyor belt, with water parcels mostly going round and round rather than directly towards their destinations.
The latest work on the Atlantic circulation model involves the cold-water system near the equator. For this, MIT/WHOI Joint Program graduate student Marjy Friedrichs, Associate Scientist Mindy Hall, and I examined the southward cold-water transport in the tropics of the North and South Atlantic. We found the North Atlantic's transport dominated by a colder water mass than the South Atlantic's. The shift does not occur as a smooth north-to-south transition, for the contrast in the western basin persists even quite close to the equator. Measurements indicate that a zonal upwelling cell in the cold water along the equator achieves the transition.
One of the difficulties we face in interpreting our data is visualization of the three-dimensional structure of the circulation. In the top figure overleaf, the vertical structure of the deep circulation was averaged to focus on the overall horizontal flow. In the bottom figure, part of the vertical structure of that deep flow is shown in a perspective view, with the ocean above 2,000 meters removed, and with the view looking northwest towards the continental slope of Brazil from a point in the middle South Atlantic south of the equator. The result looks a bit like an expressway cloverleaf, but with no connections between the two levels. We believe that the indicated seven million cubic meters per second eastward flow of the colder level upwells in the eastern tropical Atlantic to supply the indicated seven million cubic meters per second of westward flowing water on the warmer level of the cloverleaf.
The temperature change involved in this equatorial upwelling cell, about 0.5 [degrees] C, is small, but the recognition of the circuitous route that cold water from the North Atlantic takes in order to move southward across the equator has implications for the mode of the system's response to climate change. A perturbation in the warm- to cold-water conversion process that provides the "northern source" water to the cold-water limb of the conveyor belt does not simply move through the Atlantic relatively undiluted in a deep western boundary current "filament." The recirculation gyres dilute the source water's characteristics with those of the basin's interior water mass. The diluted deep water that reaches the equator is mostly diverted into the interior and further altered before returning to the western boundary to continue southward through the South Atlantic--where we are finding additional recirculation gyres. These new observations and interpretations present a new modeling challenge to the theoreticians, and, for we observationists, it points to new measurements that will improve our interpretive circulation model.
Mike McCartney came to WHOI 20 years ago with degrees in theoretical fluid dynamics. Within two weeks of his arrival Fritz Fuglister, one of the real pioneers of Atlantic physical oceanography, took him to sea on the institution's research vessel Chain. It changed his outlook! Recently Mike added up his research cruise time and found that he has averaged a month per year, including eight crossings of the Atlantic. He says that the main change over the years is that the cruises come in clumps, for example none in 1993 and a scheduled 90 days in 1994 for crossings of the South Pacific and of the Antarctic Circumpolar Current. Between the cruises and the fun of untangling the data's message about the circulation and the physics, he further relaxes by taking long solo cruises on his old gaff cutter and untangling all her rigging. He doesn't understand why people think it odd that he spends so much time on the water.
|Printer friendly Cite/link Email Feedback|
|Author:||McCartney, Michael S.|
|Date:||Mar 22, 1994|
|Previous Article:||A primer on ocean currents: measurements and lingo of physical oceanographers.|
|Next Article:||Dynamics and modeling of marginal sea outflows.|