Waters of the World

by Sarah Dry

  Woods Hole in those years was a little utopia of scientific research. Money from the Office for Naval Research (ONR) poured in, the legacy of oceanography’s great wartime contribution and testament to the need to maintain expertise in the ensuing Cold War. During short trips out to sea, Stommel studied the upper layers of the ocean, trying to work out the patterns of hot and cold that enable submarines to hide in the acoustical shadows made when water bends sound. The pacifist in Stommel flinched at the thought of contributing directly to violence, but he saw the practical need for knowledge that could help in defense of his country, and the funds offered by the ONR in aid of a basic physical understanding of the ocean came with few strings attached. To refuse the opportunity to learn more about the ocean would have been perverse.

  With the results of his five-page paper on wind-driven currents in mind, Stommel traveled to Britain in 1948, mimeographed copy of the paper in hand, to see what he could learn by observing rather than theorizing the ocean.8 While his mind’s eye had led him to imagine the effect of the rotation of the earth on the movements of a planet’s worth of water, he was also driven to understand the disorder in the ocean. For this, he needed to start on a considerably smaller scale. He set out to knock on the door (this is almost, but not literally, true) of a man who had studied the behavior of fluids when they break down into turbulence at a range of scales, from the smoke that rose from chimney pots to seeds set loose on the wind and balloons that rose above excited crowds in Hyde Park and Brighton, fluttering tags which members of the public were encouraged to send in once the balloons eventually fell to earth. From these sorts of observations, this man, Lewis Fry Richardson, had come up with a deceptively simple equation that described the speed at which objects in a turbulent fluid separate.

  Richardson’s interest in turbulence arose in tandem with his dream—dating to the early 1920s—of achieving what he called “weather prediction by numerical process”; in other words, of predicting the future by crunching numbers. To do so meant coming up with mathematical equations that described the motion of the atmosphere. This included accounting somehow for the ways in which the movement of air was obstructed, and thereby disordered, through contact with vegetation and mountains at the surface, and through the collision of masses of different kinds of air—hot, cold, wet, and dry. Richardson was aware of the importance of atmospheric turbulence even as he recognized that to make any kind of weather forecast, he would have to dramatically simplify the weather of the planet, to make it fit into imaginary squares which he used to order the globe. Within these squares, 200 kilometers to a side, all messy turbulent phenomena were reduced to a single number.9

  Some thirty years after Richardson had first envisioned numerical weather prediction as a far-off dream, Stommel traveled to visit him. Richardson had devoted the latter part of his life to understanding not natural systems, but human ones. A pacifist like Stommel, he had tried to use mathematics to explain why arms races happen, and why they unfold the way they do. This he considered more important than understanding how the turbulent fluids of the planet behave. But he had consented to join Stommel for a late return back into the nature of physical, rather than human, systems. Previously, Richardson’s observations had been limited to the motion of the atmosphere. The question Stommel and Richardson both wanted to answer now was: How similar is the ocean to the atmosphere?

  They did a watery version of the experiments on diffusion in the air that Richardson had performed two decades earlier using balloons and other objects. Stommel would remember for years to come the damp, heavy soil in the garden from which, on the suggestion of the older man, he dug up the parsnips—parsnips!—and the cold room in which he and Richardson worked together, slicing and weighing the vegetables, before they cycled down to the loch together. There, they walked to the end of the pier and dropped them into the water and tracked the speed at which they separated, using an ad hoc bit of equipment Richardson had worked up out of bits of wood and string to provide some means of accurately gauging the distance traveled by the vegetables.

  Their choice of parsnip as experimental apparatus is telling. If you drop a parsnip into the ocean, its buoyancy is such that it will float low in the water, with just a sliver of itself protruding above the waterline, unaffected by the wind. Floating freely and easy to see, humble parsnips make good oceanographic measuring devices. When given the choice of a simple, robust approach versus a complex, relatively fragile one, Richardson—and Stommel—both chose the former. Later in life, Stommel frequently commented on his own perceived lack of mathematical skill and the effect it had on his working practices. “When I contemplate the superb skills of some of my colleagues as mathematicians, as instrument designers, as masters at squeezing information out of masses of data, as scholars of encyclopedic knowledge, as scientific administrators with considerable power of decision, I realize how limited and amateurish my own ideas are,” he explained. “Therefore when I get an idea, I simply have to pass it on to someone else who has the skills to develop it. That’s not really generosity—it is just being practical.”10 Though he could be defensive about his limited mathematics, he also welcomed the way it forced him to simplify the questions he asked—as well as to seek collaboration with others.

  FIG. 6.3. Henry Stommel, “oracle” mathematician. Though he was self-conscious about what he felt were the limits of his mathematical skills, he credited them with forcing him to simplify problems. Photo by Jan Hahn. © Woods Hole Oceanographic Institution.

  Richardson was less conflicted. He spent his life seeing beyond the limitations that hemmed others in. The power of his imagination outstripped, by far, the power of any contemporary computational abilities when he imagined a great human machine for numerical weather prediction. He estimated that some 64,000 computers—then the only such kind were human beings—would be necessary to make it work, but such detail did not daunt him. He imagined something extraordinary which would eventually come to pass: a means of predicting the future state of the atmosphere based on a knowledge of its current state and a finite set of equations that would describe the motions of its particles.

  He recognized that some features of the weather would need further study. It was not yet possible, he knew, to reduce all the features of the atmosphere to simple equations. He therefore envisioned experiments that would be ongoing—in the basement of the great meteorological theater—to study the motions of eddies, large spirals of whirling water that spun out from and eventually pinched out of major ocean currents, such as the Gulf Stream. Turbulence was too important, and too fascinating, to ignore. At the same time, it was possible—even necessary—to begin the task of crunching the numbers before waiting for turbulence to be fully understood. Richardson set about understanding turbulence by observing it, first with balloons and trails of smoke, then with thought experiments, and finally with Stommel at the end of the pier.

  Together, the two men dropped forty-five pairs of parsnips off the end of the pier, and watched which way they went, hoping to understand what happened as they traveled farther away from each other. Based on what they observed of the bobbing parsnips in the Scottish loch, they concluded that energy diffused in the loch water according to the same principle as it did in the atmosphere. The results they achieved recall a paper Richardson had published nearly thirty years earlier, in 1920, describing the counterintuitive possibility that eddies acted like “thermodynamic engines in a gravitating atmosphere,” which added, rather than dissipated, energy from the system.11 The paper they published together is today remembered as much for the peculiar audacity of its first line—“We have observed the relative motion of two floating pieces of parsnip”—as for its conclusion that the atmosphere and the ocean exhibit similar forms of turbulent diffusion.12 It is important to note that scale mattered to the final result. What happens in a bathtub’s worth of water differs significantly from what happens in the loch, and what happens in the ocean again would require an e
ven larger leap.

  It was Stommel’s Gulf Stream paper, rather than the parsnip paper, that would galvanize the field. It stimulated new work on that most familiar of ocean currents, work that would lead in turn to further understanding of the circulation of all the water in the oceans. The questions raised by Stommel’s parsnip collaboration with Richardson about the role of turbulence in ocean circulation would have to wait considerably longer until they could be addressed. They lay in wait, aspects of the movements of water that could neither be solved nor ignored, like a shadowy creature whose exact dimensions are not understood, and which is only glimpsed in fragments. In the meantime, Stommel’s mind’s eye ranged freely across scales, considering the basin-wide gyre of water of which the Gulf Stream is a mere component and the relatively tiny motions that force those parsnips now this way, now that. It would take time, years and even decades, but eventually those twinned images of the ocean, at the very large and the very small scale, would be brought together once again, not only in Stommel’s mind’s eye, but in the minds of his fellow oceanographers. Then the seemingly quixotic attempt to map the motion of parsnips on a Scottish loch would be revealed as a step in the global understanding of oceans. For now, this all lay in the future.

  * * *

  Pressures in the ocean are almost unimaginably great. They are the reason that the depths of the ocean remain nearly as hostile and unfamiliar a place as the surface of the moon. Just ten meters of water provides the equivalent of an entire atmosphere’s worth of pressure. The pressure at two kilometers below sea level is two hundred times greater than that of the atmosphere at sea level. This pressure is also the reason that it took so long for the discipline of oceanography to catch up with what any individual sailor knows in the gut: that water moves quickly and in ways that are somehow both ordered and chaotic. Experienced sailors know which currents run where, and what kinds of winds are to be expected in a given region, and they also know that the sea is a surprising and fast-changing place. What they know relates to but cannot directly access what lies below. For that, both instruments and ideas are needed which can connect a ship at the roof of the ocean to the mysteries that lie below.

  For a long time, an ocean congruent with the felt experiences of sailors—an ocean that moved in the sometimes chaotic ways they reported—did not emerge from the descriptions of those who tried to study the oceans on the large scale. For most of human history, the ocean was accessed by the ship’s sail and by a sounding line dropped over its side and a variety of instruments deployed at the end of it. Navigation on the surface of the ocean and penetration to its depths were possible but limited by the buffetings of wind and current. Bottles strewn on the surface gave a very rough idea of the speed of the very top layer of the ocean, but no instrument could go deep and stay there. Saltwater, pressure, strong currents, and wildlife conspired to render most instruments useless. Added to these problems was the basic difficulty of tracking and recovering an instrument that might be sent to flow with the currents. Under these circumstances, measuring the flow of deep water directly was impossible. Without the right tools, almost nothing was known about the depths of the ocean for most of human history. The result was that those who studied the oceans assumed that nothing of much import was happening deep below.

  Despite these limitations, there were many aspects of the water that could be sampled. A key episode in the history of the study of ocean currents was the moment in 1751 when Henry Ellis, captain of an English slave ship, noticed that if he sent a bucket down deep enough when the ship was in warm waters near the equator, it was full of cold water when brought back to the surface. The only explanation for how such cold water could be found in such permanently warm places was that it had somehow traveled there from colder places—from the far north or the far south. In 1798, Benjamin Thompson (also known as Count Rumford) published an essay titled “Of the Propagation of Heat by Fluids,” in which he noted that quite unlike freshwater, which begins expanding once it reaches four degrees Centigrade and continues to do so until it freezes, seawater contracts as it cools, right up to the point at which it freezes. Cold saltwater, Rumford saw, would be very dense, dense enough to sink to the depths of the ocean. The idea of a closed circulation—one that returned on itself—seemed to follow directly from the physics of freshwater. Rumford argued that the sinking of cold water in the ocean implied a circulation, or a current, consisting of the equator-ward flow of cold water and a corresponding and opposite flow at the surface.13 The surface winds, which had so long been seen as the main force in moving the oceans, seemed to pale in significance to the great masses of water moving according to their density.

  Much later, in the 1860s, William Carpenter, a physiologist searching for new species of crinoids (feathery echinoderms that were found to live at great depths in the ocean) in the North Atlantic, noted an area between the Shetland and Faroe Islands where warm and cold deep waters were found in close proximity. He developed a theory of what he called “general oceanic circulation,” where the emphasis was squarely on the first word (to distinguish it from local circulation). His “magnificent generalization” was that the waters of the globe moved around the entire planet. Carpenter’s idea was that cold water that sank at the poles was continually replacing the warm water that was transported north by currents like the Gulf Stream. Similar movements could be inferred in the Southern Hemisphere.

  Not everyone agreed. James Croll, the Scottish autodidact who had come up with a grand theory to account for the ice ages, had strong opinions about the relative importance of wind or density to account for the movement of the oceans. His theory of the ice ages depended on the imbalances in the earth’s climate brought about indirectly by very long-term changes in the eccentricity (or shape) of its orbit. For his theory to work, he needed wind to be a significant driver of ocean circulation. He argued that as ice built up at the poles as a result of feedback effects, the trade winds would also strengthen and therefore push the Gulf Stream farther to the north, adding even further to the cooling effect that had been triggered by changes in the earth’s orbit. Due to insufficient evidence, Croll and Carpenter’s disagreement about whether it was winds at the surface or the deep motion of dense water that held the key to ocean circulation reached a temporary stalemate.14

  By the 1870s, mechanical instruments that could withstand the monumental pressures at depth and the corrosive effects of seawater made it possible to accurately measure how warm, how salty, and how deep any particular spot in the ocean was. In a laborious series of surveys, tens of thousands of observations were made, starting with the British expedition aboard the Challenger ship of 1872–1876. At the time the most expensive and comprehensive voyage ever undertaken for the purpose of studying the ocean alone, the Challenger and her crew spent four years traveling 130,000 kilometers (70,000 nautical miles) around the globe. Fifty years later, the German Meteor expedition made an even more systematic survey of a smaller portion of the world’s oceans. Zigzagging between South America and Africa fourteen times, they covered a similar distance to the Challenger.15 Soon, even the treacherous waters of the southern ocean began to be measured, thanks to the efforts of the British Discovery expeditions.

  Challenger, Meteor, and Discovery: the names say it all. These were single-ship operations. No matter how long and ambitious the journeys they made, the outcome of their expeditions was constrained by the simple fact of having been made by a sole ship. Measurements were by necessity taken one after another—serially—and only after the expeditions finished were they were brought together. By plotting the records of temperature and salinity on a map, lines could be drawn between the dispersed data points. These lines traced what were assumed to be the contours of actual bodies of water in the ocean, masses of water defined as having the same properties. But these maps performed a sleight of hand. By combining in a single image observations that had been made years and sometimes decades apart, they created pictures of the ocean
that gave the appearance of being snapshots of the ocean at a particular moment. In fact, they were idealized—and, in a very real sense, therefore imaginary—average bodies of water, based on measurements widely separated in both time and space.

  These atlases suggested the ocean was a very orderly place. Great tongues of water were revealed to be spreading across the ocean. By studying these slablike masses while keeping in mind the basic properties of water—that cold water sinks and spreads and that salty water is heavier than fresher water—it was possible to guess how the water might be flowing. These guesses were the hard-won product of thousands of hours of surveying. They gave specific contours to the theoretical movements of deep water that men such as Rumford and Carpenter had described in the nineteenth century. On the scales of time and space that these expeditions were able to measure, the ocean was a stable place in which certain large-scale features, such as the Gulf Stream and the world’s other western boundary currents, stood out. Water in this averaged ocean acted more like cooling lava or even solid rock than a fast-moving liquid. There was no drama in these deep waters: nothing like the hurricanes and squalls whipped up on the surface, nothing equivalent to the thunderstorms, fronts, and cyclones of the atmosphere. Instead, there was the slow ooze of cold water as it moved across the ocean floor at a distinctly funereal pace—decades, hundreds, and even thousands of years were needed before much would change. Time passed slowly. Any features of the ocean smaller than several hundred kilometers across and shorter in duration than several hundred days could be (and were) easily missed from one data point to the next. The absence of these phenomena from the data led many to assume they must be absent from the oceans themselves.