by Sarah Dry
In the ocean, resolution is a matter of both time and space. The biggest challenge is to observe multiple locations in the ocean simultaneously. This kind of view—called synoptic—was pioneered by meteorologist Robert FitzRoy in the 1850s when he linked coastal observers via a telegraphic network. Some 120 years later, oceanographers were finally on the cusp of doing the same for the ocean. The task was much harder not only because the ocean is a wild and unforgiving place to live and work, but because water is so much denser than air. As a result, it packs much more turbulence into an equal area, and that turbulence—the eddies—lasts much longer than storms in the atmosphere. Ocean eddies are roughly a tenth the size of atmospheric storms, and they last weeks and even months instead of days.
The task that now faced Stommel was to understand how eddies fit into the large-scale structure of the ocean. “We are not interested in describing these eddies in isolation,” he wrote in a piece intended to serve as a wake-up call to oceanographers, “we are concerned with discovering whether they play a significant role in driving the large-scale circulation. Is there interaction of eddies and large-scale circulation in the ocean as there is in the atmosphere?”29 What, for example, is the relationship between these newly discovered underwater storms and the large-scale circulation of the ocean? Do eddies work to dissipate energy, to counterintuitively add energy back into the system, or do they do a bit of both? Stommel had demonstrated that the ocean was susceptible to simple physical explanations. He had given the burgeoning field the confidence to start drawing—sketching—a representation of the movements of the ocean. And he insisted that to know the ocean it must be observed repeatedly, with patience and determination.
For the past twenty years, the way scientists had imagined the atmosphere and the ocean had differed dramatically. The atmosphere had become, thanks to the network of radiosondes, and to the work of Joanne Simpson and others, a turbulent, fast-changing environment. The ocean had remained, in theoretical terms, remarkably still: a “steady smooth flow,” in contrast to the “highly non-linear fluid-dynamical flow, with large eddies [storms] playing an essential and dominant role.”30 It was time, argued Stommel, to determine once and for all whether the ocean was as nonlinear in its motions as the atmosphere. The questions Stommel and Richardson had investigated on the loch were about to be explored on a much bigger scale. “We anticipate that the eddies will be found to play a dominant role in the dynamics of the ocean circulation,” predicted Stommel, “and that our whole theoretical concept of ocean currents, developed over the past twenty years, will be changed.”31 It was a risky undertaking that threatened to consign to the dustbin the hard-won theories on which many of the participating scientists had worked, but “if anybody was going to correct or demolish our old theories, we wanted to be the ones who did it.”32
The best way to determine the answer to this question was not, as Stommel made clear on numerous occasions, to set up a vague and unfocused survey that would passively gather data. “After all,” explained Stommel, “anyone can sprinkle dots on a map of the world and call it a plan for future measurement.”33 To really understand the ocean would require a series of experiments with clear hypotheses to test and protocols for assessing their outcome. The planners called the project MODE, for Mid-Ocean Dynamics Experiment. Every word of that tidy acronym was a signal of the reach of its ambitions: to bring motion and the methods of physical experiments to a part of the ocean that had previously evaded observation. The name was a sword to slay the legacy of the descriptive atlases of the kind produced by the Meteor that even in the 1970s still haunted oceanography. For this was, quite self-consciously, an experiment—not a survey, not a series of hydrographic stations, but an experiment. The well-defined question this experiment sought to ask was “Do eddies exist on this scale in the deep ocean?” and it set out to do so in a similarly well-defined length of time.
Stommel had identified the “problem” of experiment in oceanography as early as 1963, when he’d written an important paper for Science, titled “Varieties of Oceanographic Experience,” in a sly reference to William James’s classic study on religion. In it, he had made the case for considering every oceanographic expedition as a scientific experiment (“if we regard an expedition as a scientific experiment, then we must propose to answer certain specific questions . . .”),34 and for taking careful account of the range of scales on which the ocean varies—its varieties of experience. The ocean, it was becoming more and more clear, varied across an astonishing range of time and space. Designing good experiments—experiments that had the possibility of yielding definitive results—meant taking careful account of just what kind of variety existed in the ocean. As the initial results from Aries had made clear, it wasn’t possible to simply use statistics to average out energy in the ocean across convenient scales. Answering particular oceanographic questions—say the variation in sea level in a particular ocean basin—would require asking questions at the right scale. To this end, Stommel included a diagram which served as a visual index of the range of scales on which energy varied in the ocean, from the hundreds of meter-long gravity waves that lasted just minutes to the tidal variations that happened on a daily and monthly basis, to meteorological effects that transpired on similar scales with much less regularity, to the mammoth variations, occurring over thousands of years and thousands of kilometers, that constituted the ice ages. This diagram was classic Stommel—a deceptively simple tool for ordering complexity. It was both a map of the energy of the ocean and, as Stommel tried to argue, should also be a road map for oceanographers if they ever hoped to catch the most meaningful oceanographic phenomena with the instruments at their disposal. Though the task seemed overwhelming, if matters of scale received their proper due, Stommel thought there was reason to hope that in the future “theory and observation will at last advance together in a more intimately related way.”35
Working alongside colleagues including Carl Wunsch, Francis Bretherton, and Allan Robinson, Stommel developed a plan to catch an eddy in the ocean.36 If the holes in the experimental net were too big, an eddy could pass through it undetected. Too small, and they would only see a fraction of it. The entire ocean was, from Stommel’s point of view, neither more nor less than a problem in what he called hydrodynamics—the movement of water—on a scale “larger than a laboratory, smaller than a star.” Similarly, an eddy was an object of definite size (though what size precisely they could only guess) to which appropriately scaled detectors would need to be directed.37 Stommel and colleagues determined that a parcel of ocean approximately 300 kilometers square, stretching from the surface to a depth of roughly four kilometers, would be the right size. Like a hunter waiting for its quarry, the scientists would have to set up a trap of detectors and then simply wait, and hope that an eddy would pass in the time allocated to the experiment.
They reckoned it would take six ships, two airplanes, dozens of moorings, neutrally buoyant floats, free-fall velocity profilers, air-current dropped probes, and 121 pressure gauges arrayed on the ocean floor to capture the eddy.38 To create such a complex system of instrumentation and monitor it during the study would require the coordination of fifty oceanographers from fifteen institutions, including, for the first time, modelers who would be integral to the design and execution of the project.39 The experiment itself would last four and a half months at a location between Bermuda and Florida. The experiment made use of a new kind of floating instrument—called SOFAR, for SOund Fixing And Ranging floats—that could be sent to specific locations in the deep ocean and moored there, taking measurements of the temperature, speed, and salinity of the water that flowed by them. At the same time, instruments were deployed to follow the currents in an area, floating freely.
If MODE was successful, it could be used to refine the most fundamental theories of the general circulation of the ocean, and would have a corresponding impact on the climate model that combined, or “coupled,” the ocean and atmosphere that h
ad been published in 1969 by Syukuro Manabe and Dick Wetherald, at the GFDL at NOAA. Another hoped-for dividend was a better understanding of the ocean—including the possible discovery of the ocean’s “weather”—that would improve forecasting for the kinds of weather that affected people on land and at sea. The plan was bold, and the timing seemed auspicious. There was consensus among those whose voices mattered that this was a worthy use of time and money. Nevertheless, there was real uncertainty about whether it would be successful. It was entirely possible that the “eddy trap” they laid would fail to catch anything.
A film made at the time captures the uncertainty (and copious facial hair) of the moment. Scientists sit scattered across a lawn, seemingly performing oceanography in the open air, relaxed, engaged, and egalitarian. “We may all be retired before something’s happened,” cautioned one scientist, while another worried they might find nothing at all. Deciding where to site the experiment required the scientists to make assumptions about how the oceans operated that the experiment itself was designed to test. There was thus real uncertainty not only about the kind of results that would emerge but about whether anything at all would be measured. A crucial decision concerned whether the experiment should be centered over a relatively flat piece of ocean or a rocky piece of ocean. The answer depended on how much you thought the landscape of the ocean floor—its topography—affected the mixing of the water above it. To what extent could any piece of ocean give the results that oceanographers were hoping for? Carl Wunsch, for one, suggested that “There is no such thing as a typical piece of the ocean, every piece of the ocean is different.” When someone countered that the experiment could be centered at the boundary between a rocky and a smooth ocean floor, Wunsch replied that “maybe nobody’s going to be satisfied with that compromise.” The men laughed, but Wunsch wasn’t really joking.40
FIG. 6.5. Early sketch of instrumentation for the Mid-Ocean Dynamics Experiment. The experimental area measured 300 miles square and four kilometers deep. On the seafloor, 121 pressure gauges were arrayed with thirty-mile spacing, to be augmented by three to four moored hydrophones and a set of pinging constant-level floats floating at four different depths. These physical systems were to be augmented by a “computer-numerical-model” for prediction of float positions to enable the experiment to be tracked and adjusted in real time. Source: Memo from Henry Stommel, August 11, 1969, Mid-Ocean Dynamics Experiment records, AC 42 Box 2, Massachusetts Institute of Technology Institute Archives and Special Collections, Cambridge, MA. Courtesy of Institute Archives and Special Collections, MIT Libraries.
FIG. 6.6. The MODE research zone, straddling an area of smooth and rough ocean floor topography. Source: MODE-1: The Program and the Plan, March 1973, Mid-Ocean Dynamics Experiment, AC 42, Box 1, Massachusetts Institute of Technology Institute Archives and Special Collections, Cambridge, MA. Courtesy of Institute Archives and Special Collections, MIT Libraries.
Eventually, the MODE team compromised on a location in the Atlantic (the Pacific was never an option, for reasons that were primarily logistical), choosing a spot that was both rocky and smooth. The experiment itself ran remarkably smoothly, making use of special “hot-line” phones that enabled the on-ship scientists to communicate with a headquarters. The only major setback was the disappearance of the central mooring, a hazard of leaving large pieces of equipment unsupervised in the open sea but a puzzling disappearance nevertheless.
The audacious plan paid off. An eddy was found and tracked. MODE showed that eddies were common and widespread and, most importantly, contained a staggering ninety-nine percent of all the kinetic energy in the ocean. Eddies, in this sense, were the dark matter of the oceans, until MODE revealed that oceanography had been missing the biggest story of all. The mysteries of turbulence and flow were beginning to be fully revealed, if not yet solved—or resolved.
That was a clear success. Less clear-cut was what the legacy of MODE would be for oceanography. MODE had “established a reputation as one of the most tightly run field programs ever, much to the dislike of some of the participants,” a writer for Science noted. The need for consensus on planning among such a large group meant that, as one scientist put it, “on sticky points, we would take people into the back room and intellectually beat each other into agreement.”41 This was a far cry from the intellectual independence that Stommel had so appreciated. The success of MODE broadcast to the wider world that oceanography could now operate on big scales, with big budgets and dozens, if not hundreds, of researchers working together. It was, as the Science article put it, “big science, new technology.” But with institutional heft came what many lamented as the loss of individual freedom, and the success of the project was a source of deep ambivalence for many of those who had participated in it. Francis Bretherton, a theoretician who clearly enjoyed the chance to participate directly in a field experiment, nevertheless remarked that “it would be disastrous if the success of oceanographic big science persuades people that it is the only way to do things.”42
MODE created new ways to think about energy and motion in the ocean. It also, and just as fundamentally, created a whole new way of doing oceanography. Dozens of scientists from several countries had collaborated for an intense and short-lived moment in which an enormous amount of data was generated. For Richardson, for whom the image of a massive calculating organization, consisting of some 64,000 human computers, represented a utopian fantasy, such data would have signaled the arrival of a long-hoped-for future for meteorology. Whether he, an intensely independent-spirited man, would have liked to actually live in such a brave new world is unknowable. Stommel was there to see the day arrive and found himself recoiling from both the administrative pitfalls of big projects and the use of brute force rather than physical theories to crack oceanographic problems. For him, scientific breakthroughs were an intensely personal achievement. “Breaking new ground in science is such a difficult process that it can only be done by an individual mind,” he wrote in a cri de coeur titled “Why We Are Oceanographers.” “For some of us, this is the main attraction of doing scientific work. In this respect, it is like the art of painting or musical composition or poetry . . . it all begins with an individual’s choice of medium, choice of theme and style and subject.”43
Stommel sought to maintain and celebrate the contribution that a single individual (or a group of like-minded individuals choosing freely to work together, as upon an idealized ship expedition) could make to the biggest questions about the ocean. He believed that the biggest questions about the ocean—how it circulated on the basin scale or even the global scale—were accessible only to the best individual minds. But he needed data, not too much or too little, but just enough. To get the data required organization, grant-writing, logistical planning, all of which threatened to consume the time needed for thinking and doing science. The paradox was that the bigger Stommel’s ideas about circulation got, the more he needed data to test them, and the bigger the projects got. He could not repeat the 1948 achievement, when he essentially deduced the Gulf Stream from a few equations.
The success of MODE was a source of abiding ambivalence for Stommel. MODE made it possible, even necessary, to think about the old sluggish ocean in an exciting new way. It also raised the possibility that the kind of science that he loved—driven by mental pictures and intense collaboration with a few individuals—would become increasingly difficult to justify. This was a transitional moment, when the discipline stood poised between the old ways (already the object of nostalgia), when it had been possible for individuals to make their own destiny, scientifically speaking, and the future, in which oceanography would be defined by its large projects rather than the ideas of its individual practitioners. This had not yet come to pass, but with MODE, Stommel saw the metamorphosis beginning.
* * *
Stommel concluded that the best way—the only way—forward was to “take up various oceanographic phenomena separately as though they were
mutually independent (which of course they are strictly not).”44 To study the machinery meant taking it apart, conceptually, all the while remembering that the machine only functions when it is whole. The questions raised by MODE, in other words, could only be answered with other experiments similar to it which could probe different elements of the ocean system. And so, despite his deep, almost constitutional, misgivings about the new “bigness” of oceanography, Stommel continued to involve himself in projects similar in scale to MODE, including a follow-on project, a contentious collaboration with the Russians, called POLYMODE. The pace with which these new experiments occurred was intense. Between 1973 and 1978, a total of nine field experiments were planned, an entire alphabet soup of acronyms from MODE to GARP, NORPAX, JASIN, CUEA, SDO, INDEX, ISOS, and GEOSECS.
Thanks to these experiments, it became clear through the course of the 1970s that the answer to the question of where eddies existed was relatively straightforward. Almost everywhere people looked, they found eddies.45 Eddies in the North Pacific. Eddies in the Arctic. Eddies in the Indian Ocean. Even eddies in the Antarctic. The question become not where eddies were, but where they weren’t. By 1976, John Swallow was wondering whether there was any place in the ocean from which they were absent. It appeared there was not, and this fact, added to the energy that they contained, made it conceivable that eddies were not incidental to ocean circulation but essential to it.46