by Sarah Dry
* * *
As early as the 1820s, Joseph Fourier had recognized the important role played by the atmosphere in trapping the sun’s heat, but he had never imagined that humans might one day materially affect the amount of certain key gases in the atmosphere. John Tyndall had rendered quantitative just how certain molecules absorbed heat, but he too had not considered the possibility that humans might significantly affect the earth’s atmosphere in the future. It was only in 1895 that a Swedish physicist, Svante Arrhenius, had made the first calculation about the effects of an unnatural (or forced) increase in carbon dioxide in the earth’s atmosphere, due to the action of humankind. His prediction, which looks incredibly prescient today, was rejected by other scientists as implausible. In 1938, an English engineer named G. S. Callendar had performed additional calculations on what a doubling of carbon dioxide would mean for the average global temperature.16 Again, his contemporaries considered his findings unlikely. It was only in the 1950s that work by Roger Revelle and Hans Suess brought sustained attention to the effects of carbon in the earth’s oceans and atmosphere. Responding to their work, the geochemist Charles Keeling established a measurement station atop Hawaii’s Mauna Loa volcano to record the percentage of carbon dioxide circulating high in the atmosphere, where it was most evenly distributed. Almost immediately, his device began to reveal a gradual and inexorable rise in the amount of CO2 in the atmosphere as a result of the burning of fossil fuels that released carbon deposited millions of years ago.17
This was the moment, as it were, when the past met the future. While a handful of researchers had foreseen the effects of industrial activity on the climate of the entire planet, their work had not found a receptive audience. It was only in the late 1950s that a particular combination of researchers, instruments, questions, and anxieties came together to prompt sustained and productive research into the nature of the effects of human activity first on the atmosphere and, more recently, on the entire global system. It was also when the term climate shifted from meaning something primarily geographical to a temporal concept connoting change. While men such as Köppen and Hann had worked to “modernize” climatology and render it globally comprehensive, they had remained committed to an understanding of climate as essentially stable on human timescales (i.e., for decades or longer). It was only with the indications of the rapid rise in carbon dioxide, and a growing realization of what that could mean for the climate of the entire planet, that a new kind of climate science came into being, one which emphasized temporal changes on a global scale.18
Gradually, these findings on rising carbon dioxide made their way into public sphere as the findings of individual researchers emerged and certain scientists, notably Roger Revelle, began to speak loudly enough to require responses from those in power. Further investigations were then launched into what the practical impacts might be. In 1965, President Lyndon Johnson convened a panel on environmental pollution. The panelists were tasked with evaluating the direct environmental effects of industrial pollution on air and water quality but decided to include a special report on an “invisible pollutant,” atmospheric carbon dioxide. Chaired by Roger Revelle, the five-man subcommittee on carbon dioxide included Charles Keeling, the man responsible for the CO2 measurements on Mauna Loa, as well as geochemist Harmon Craig, meteorologist Joe Smagorinsky, and a young geologist and geochemist named Wallace Broecker. Referring to earlier work by Svante Arrhenius and T. C. Chamberlin on the effects of rising carbon dioxide on climate, the subcommittee reported on the five years of records then available from Mauna Loa (and another at the South Pole) that revealed a 1.36 percent rise in the carbon dioxide content of the atmosphere. Based on estimates of fossil fuel consumption in the past and projected future use, they guessed that by the year 2000 there would be roughly twenty percent more atmospheric carbon dioxide than preindustrial levels. Considering the extra heat that would be trapped by this extra carbon dioxide, they estimated a rise in the average temperature of the earth’s surface of between 0.6°C and 4°C.
The authors of the report readily admitted that these estimates were based on many simplifying assumptions. Despite the impossibility of making precise predictions, they were still confident that by the year 2000—then thirty-five years in the future—the rise in carbon dioxide would be significant enough to produce “measurable and perhaps marked changes in climate,” including changes in temperature.19 Given the scale of the changes, the effects on human beings could be “deleterious,” and the authors suggested that it was important to explore the possibility of “deliberately bringing about countervailing climatic changes.” The future was close in which it might be advisable to spread reflective particles over large portions of the upper ocean, or to alter cirrus clouds high in the stratosphere.
As word began to spread throughout the scientific community that carbon dioxide was rapidly rising, it became increasingly clear that it would be necessary to understand how Earth’s climate operated at the largest possible scale. Given the complexity of the task at hand, the scientists working in this area started as simply as possible. They created climate “models,” sets of equations that described a few basic features of the climate system.20 They used these toylike models to play with the climate system, testing what changed as they varied a small number of inputs, such as the amount of carbon dioxide in the atmosphere or the amount of radiation being introduced into the system. Since these models were designed to understand the impact of a global metric—the amount of carbon dioxide—they also generated a global output: the global surface temperature.21 This single number usefully reduced the otherwise overwhelming complexity of the planet to something even a child could grasp. It also suggested that something called the “global climate” existed. This so-called global climate was in many ways a useful fiction, born of often necessarily crude averages. The global average temperature did not—and cannot—exist in any one place. It was an imaginary tool for grasping the earth in one gesture, a useful bit of reduction that simultaneously enabled tremendous insights into the working of the global climate system and elided its complexity.
To check if these models were realistic, these early modelers needed to compare them with real data from the earth. This demand for global temperature data prompted scientists who worked with such data—among them climatologists working at the Climatic Research Unit at the University of East Anglia (UEA)—to start producing the first estimates of the average global temperature, based on measurements collected since roughly 1850, when reliable instruments were first introduced.22 Unlike global climate models, which are easy to build as long as equations are kept simple and few data points are entered, there is no easy way to generate a global average temperature. The only way to do so is to take as many measurements around the planet as possible and find a way to combine them that takes into account all the local variations, and gaps in coverage, that would otherwise queer the result. Indices of temperature readings had been compiled, beginning in 1938, by G. S. Callendar, by Mikhail Budyko, and by J. Murray Mitchell Jr. These averages relied on readings from the Northern Hemisphere. It would be decades before data from the oceans was incorporated into these putatively global averages, and even longer before observations from remote polar regions were included.23
The notion of an average global temperature changed what it meant to study climate. It did not, however, spell the end of climatology as it had been long practiced by inheritors of the Humboldtian sensitivity to location, men such as Köppen, Hann, and Hildebrandsson. Indeed, Hubert Lamb, a climatologist with a sophisticated historical and geographical awareness, had founded the Climatic Research Unit and remained an important figure in the field.24 But global temperature indices did help bring about an important shift in what it meant to study climate that ran counter to the geographically oriented strain of climatology Lamb had advocated. Once the planet was considered as a machine for producing average temperature, local or even regional variations became, if not irrelevant, then
subsidiary to the aims of global climate modelers. Here, then, was the birth of a new science of climate—different than climatology—that would concern itself not with the places of the planet but with its past and its future, with the mechanisms that prompted changes that could be measured on a global scale.
Far from abandoning the past, for the scientists studying the effects of carbon dioxide, the newly uncertain future they were facing made the earth’s history an even more precious resource. If carbon released by human activities had the potential to dramatically alter climate, what had happened in the earth’s past potentially became the key to the future. Only by understanding the natural variability in the earth’s past—when it had been warmer, or colder, and why—would it be conceivable to predict what lay in the future. Prediction had been the long-desired mantle of scientific maturity that the meteorologists of the nineteenth century had sought. In the decades following World War II, it seemed as if the new field of climate science might finally be able to deliver on the promise of predicting not only the weather but the earth’s climate as well. At the same time, a small group of scientists grappled with the realization that the climate was already responding rapidly to changes caused by human beings. In this way, the ambition to predict the future climate was always entangled with a newfound awareness of just how variable the climate was, and just how much power humans had to perturb this already unstable system.25
* * *
FIG. 7.5. A storage freezer full of samples from the Camp Century ice core at the Cold Regions Research and Engineering Laboratory in Hanover, New Hampshire, 1965.
FIG. 7.6. Willi Dansgaard in the freezer with an ice core, 1994. Credit: Centre for Ice and Climate, University of Copenhagen.
While these changes were occurring in the discipline of climate science, Willi Dansgaard kept working to gain access to the ice drilled by the U.S. military at Camp Century in Greenland. By 1967, he had prevailed. In that year, ice that had been extracted from the interior of Greenland with great difficulty and expense, and amid much secrecy, was sent in two-meter sections to a New Hampshire government laboratory devoted to studying cold, the U.S. Army Cold Regions Research and Engineering Laboratory (CRREL). Soon after, a Danish colleague arrived in New Hampshire to collect eighty-six samples to be taken back to Denmark for analysis in Dansgaard’s mass spectrometer. He finally had a chance to try out the isotopic time machine he’d been dreaming of.
He now had samples from the entire length of the core. Dansgaard and his team analyzed almost 1,600 different pieces from the nearly mile-long core in their mass spectrometer. When it had been reduced and plotted, the data showed up as a long and squiggly line extending back in time. The core had produced better results at a much finer resolution than anyone had expected.26 The sequence of rises and falls in temperature indicated by the changing amounts of heavy isotopes in the sample were complex, but there seemed to be underlying cycles. The detail revealed in the ice core, especially for the most “recent” 8,300 years, was stunning. It was possible to trace annual temperatures, including seasonal summer/winter variations, as if reading the rings of a tree. The most stunning number was 100,000, the number of years the ice record stretched back. It was by far the longest record of the earth’s climate that had ever been obtained at such a high degree of resolution (the sediment cores taken from lakes, though delivering longer timescales, were much less finely resolved, with a thousand years squeezed into a single centimeter). Ice cores, in contrast, laid out annual bands at a rate of roughly 50 years per meter. Easy to see, and count, they made the past legible in a way it never had been. In a splashy 1969 paper on the “one thousand centuries of climatic record” that appeared in Science, Dansgaard and his team identified a series of cycles, or climatic oscillations, within the results. Roughly every 120, 940, and 13,000 years, they suggested, there were regular changes in the climate.27
The first thing Dansgaard and the team did was compare their findings with other clues about the earth’s past temperature. Around the same time that Dansgaard had been investigating the nature of the isotopic ratio of oxygen in rainwater, other physicists had been using the same ratio to turn muddy cores taken from the ocean floor into thermometers in their own right.28 If they matched, that was a good sign that the ice core was telling them something meaningful about changing temperatures in the past. Otherwise, there was a chance the ice-core results were erroneous or represented temperature changes limited to Greenland. Dansgaard and his collaborators included a figure in their Science paper comparing their results with past temperatures derived from other studies—of ancient pollen from Holland that extended back 80,000 years, Pleistocene deposits that stretched back nearly as far, and deep-sea sediment cores. Two facts immediately stood out when the four traces of climatic variations were placed side-by-side. First, the broad outlines of the four curves, each obtained by independent means, did indeed match up. Second, the detail offered by the ice-core study was phenomenal. The other traces looked like a child’s messy scrawl compared to the minutely saw-toothed edge of the ice-core data.
While Dansgaard and his coauthors stressed that their curve was “primarily valid for the North Greenland area,” they ended their paper by noting the uncanny correspondence between the readings. The correspondence gave credence to their results, but it did something even more important: It gave strong evidence that major changes in past climates had occurred on global, not regional, scales, and that they had occurred much more rapidly than anyone had ever imagined. Their dry language couldn’t conceal the excitement they felt about a technique that provided “far greater, and more direct, climatological detail than any hitherto known method.”29 What this detail revealed was the potential for abrupt changes in the climate of the planet.
Dansgaard’s technique, for all its ingenuity, would have been much less valuable without the remarkable length of ice on which to work. The drill at Camp Century had made it possible along with the unique skills and sheer stubborn-mindedness of the team that operated it. Their success made the drill team hungry for more ice. Soon after hitting bedrock at Camp Century, they embarked on another great drilling expedition, to Byrd Station, a U.S. research station established in Antarctica during the 1957–1958 IGY. Within two years, they’d managed to pull another core from the ice, which went back almost as far in time, and with the same degree of detail, as the Camp Century core. Once again the results were compared, and once again they matched up.
The changes that were preserved in the ice at the top and the bottom of the globe seemed to be the traces of dramatic climatic shifts (up to 8°C) that had affected the entire planet in the space of just a few decades. No one had ever imagined climatic change of such magnitude could happen so quickly.30 Even after the publication of the paper, the meaning of the curves that Dansgaard had produced was not readily apparent. Though Dansgaard and his coauthors knew they had created a powerful new way of looking into the earth’s past, the community of researchers interested in understanding the earth’s past—and the implications it held for the future—did not yet fully understand what the ice-core traces meant. Decoding the jagged curve fully would take an individual with a mind that could range, as had James Croll’s, across disparate data and a tremendous variety of both spatial and temporal scales, to understand fully the implications of the rapid changes revealed by the ice core.
* * *
Wally Broecker was just such a person. Trained as a geochemist with a special interest in ocean isotopes, he had served on the 1965 subcommittee on atmospheric carbon dioxide along with Roger Revelle and Charles Keeling. In 1966, he had noticed “an abrupt transition between two stable modes of operation of the ocean-atmosphere system” in a set of cores taken from the floor of the deep ocean, stretching back 200,000 years.31 Primed to see fast transitions, when Broecker read Dansgaard’s 1969 Science paper, he immediately noticed how abrupt the changes revealed by the cores were. He realized that the cycles that Dansgaard and his coll
eagues had identified in the past could also be used to predict future temperatures with much better resolution than the crude relationship between atmospheric CO2 and temperature that he and his fellow subcommittee members had had to rely on in their 1965 report. The timing of the most recent warming bump in the small-scale cycles reported by Dansgaard suggested that another such warming period was due very shortly. Drawing on the historical record and thinking along the same timescale of decades that the Johnson committee report had established, Broecker extrapolated to the very near future. His paper, published in Science in 1975, was titled “Climate Change: Are We on the Brink of a Pronounced Global Warming?” Today, it is a very familiar headline. At the time, it was not. While Broecker was not the first to use the phrase, his Science article put the term into wide circulation. His paper has since been celebrated as a landmark in the history of global warming. Its thirty-fifth anniversary was marked, and Broecker (to his considerable consternation) was dubbed the “father of global warming.” At the time, however, the adjective global was just as significant as the idea that the climate might be warming. Global change—in any direction—was the signal that Dansgaard’s cores had revealed. Precisely what kind of change remained, as Broecker’s title had suggested, was an open question.
While Broecker’s prediction came true, the inference on which it was based turned out to be groundless. He had taken what he himself called a “gigantic intellectual leap” when he’d assumed that the small-scale cycles in the Camp Century core would also typify the globe. He readily admitted as much. “My prediction was based on the false premise that Dansgaard’s record typified the globe. In reality, it typified only the northern tip of Greenland.” Subsequent ice cores from different locations didn’t reveal additional 120-year cycles. Broecker’s prediction of renewed warming came true, but not because of any underlying global 120-year cycle revealed by the Camp Century core.32 It was a lesson in the challenges of predicting changes in climate that scientists would learn again and again. The global system was a complicated beast, and no single heartbeat could explain the changes it had undergone in the past—nor those it might experience in the future.
