Theories of ice ages past and future

What causes a warm world of flowing water and verdant growth to become a cold world of dry winds raking arid landscape? This question has been the object of a long and continuing rumination, of more than a century of field investigation, theorizing, and debate. Along the way, a new history of Earth's climate has been written and a science of paleoclimatology invented. And yet, for all of the new techniques and powers of reasoning at their command, researchers cannot satisfactorily answer the basic question: Why do such large changes in climate periodically overtake the planet?

The difference between one world and the other is about as different as worlds can be. The difference is mile-thick ice burying 11 million square miles of continent—Asia, North America, and Europe sagging under its weight. The difference in the level of the oceans is 400 feet. The difference is a mere 9°F in mean global temperatures. What these parameters of modern climate science fail to describe, of course, is the difference that is most obvious and most germane to humankind—that one is a world with much more life in it than the other. Like the great swayings of the ice ages themselves, theories about them have come and gone. Early in the century, the British meteorologist C. E. P. Brooks borrowed a line from Rudyard Kipling to describe the state of the art. "There are at least nine and sixty ways of constructing a theory of climatic change," he wrote, "and there is probably some truth in quite a number of them." U.S. Weather Bureau physicist William J. Humphreys lamented that "no one can form the slightest idea" about past and future ice ages—when they might begin, how cold they may be, or how long they may last. Some fine minds have visited the problem, although in the first half of the century, Brooks and Humphreys were among only a handful of researchers who were knowledgeable in the ways of the atmosphere. For most, climate change was not a serious issue, and ice ages were ancient, irrelevant history. Most people who called themselves climatologists were busy tabulating the statistics of weather for the benefit of local farmers and engineers, and most meteorologists were practicing a trade of weather prognostication founded on maps of pressure differentials and intuition.

In Brooks's day, climate research was a poor, young, backwater branch of geology in competition for students with the lucrative oil exploration business, a venture that takes a different interest in strata bearing signs of ancient vegetation. Without the data to test an idea or the technical wherewithal to pursue it, researchers proposed theories of climate that, for many years, were what scientists politely call qualitative—little more than hand-waving arguments and just about anybody's game. Carl A. Zapffe, a Baltimore metallurgist, mixed his physics with esoteric theology and the legend of the Lost Continent of Atlantis to bolster his "Submarine Vulcanism Theory." Periodically, volcanoes would erupt along the floor of the Atlantic and spew vast blankets of cooling dust into the atmosphere at the same time they warmed the sea, promoting evaporation, a boost in water vapor, and increased snowfall.

To finally solve such a large and enduring puzzle would be a mark of the greatest scientific distinction, worthy of a Nobel Prize. The hunt would be joined by astronomers, mathematicians, physicists, geologists, oceanographers, biologists, botanists, chemists, climatologists, meteorologists, and computer modelers. They would look everywhere for explanations—deep in the earth and far out in space. They would consider the multifaceted character of the climate system itself—how the oceans, the atmosphere, and the ice cover affect one another—and any number of external astronomical possibilities.

For all of the early uncertainties, most of the main theories of climate change were proposed remarkably soon after Louis Agassiz's famous "Discourse" in 1837, when in Neuchatel, Switzerland, he unveiled his theory of the ice ages. As different types of data accumulated over the years, theories would shift in and out of favor among the researchers of the day. Much of the history of the thinking about ice ages has followed this pattern of the debunking of one venerable theory and the resurrection of another.

Among the first explanations were the geological ones. The nineteenth-century British geologist Charles Lyell and later the American James D. Dana pointed to mountain building and other changes in the heights of Earth's crust as possible explanations. One theory held that the poles have changed location over time, due to changes in the planet's tilt, exposing different flanks of Earth to warm and cold latitudes. In the 1920s, Alfred Wegener and his father-in-law, climatologist Vladimer Koppen, offered a variation on that theme. The Germans explained ice age evidence in terms of Wegener's theory of continental drift. Like the "wandering pole" theory, climates of the continents were altered as they drifted through different climate-controlling latitudes. At the U.S. Weather Bureau, Humphreys persuaded himself—although not very many of his colleagues—that the cooling effect of dust from volcanic eruptions was to blame.

Several theories were linked to processes and features internal to Earth's climate system. The role of the gas carbon dioxide in the atmosphere was first proposed as a cause of ice ages by Swedish researchers Svante Arrhenius and Nils Eckholm at the turn of the century. Pointing to the "greenhouse effect" of CO2, which is transparent to the Sun's short-wave radiation but traps long-wave (i.e., heat) energy radiating back from Earth's surface, they argued that a loss of carbon dioxide could send the planet into an ice age and that the buildup of CO2 could stave off such a catastrophe. A century later, of course, the close relationship between temperatures and greenhouse gases would become critical in the context of another climate puzzle.

Early in the century, meteorologist Brooks suggested a meteorological explanation—the idea that the formation of ice provoked a feedback mechanism that progressively cooled the planet by reflecting more and more sunlight from its surface. In 1956, Maurice Ewing and William Donn gave this idea another twist when they proposed that ice ages were triggered by rising Arctic temperatures. When ocean temperatures warmed to the point at which the Arctic became free of ice, a cascade of atmospheric effects—heightened evaporation and increased water vapor—led to a surge of snowfall over nearby lands, to the ice reflection feedback loop and the spawning of a new ice age.

Widely publicized at the time, Ewing and Donn's theory was the first expression of a concept that would become a common scenario in a globally warming world. Columbus Iselin, director of the Woods Hole Oceanographic Institution, where Ewing and Donn were doing their work, set the stage in a 1957 interview with Christian Science Monitor journalist Robert C. Cowen. Noting that the modern era was "literally flooding the atmosphere with carbon dioxide," Iselin asked, "Are we making a tropical epoch or are we, perhaps, starting another ice age? We don't know enough about the oceans and the weather yet to be sure which way the effect will go." Later scientists and journalists would sketch a similar scenario. Warming temperatures in the far north would not necessarily usher in an ice age, but could spread colder-than-normal temperatures over much of northern Europe.

In 1964, the New Zealand scientist Alex T. Wilson suggested that the Antarctic ice sheet occasionally reaches critical mass, collapses, and sends large sections surging into the surrounding ocean. Sea levels rise and sunlight bounces off the spreading sheets of floating ice, cooling the atmosphere. While researchers found no evidence of such an event in the sediments of the Southern Ocean, Wilson was among the first to recognize the ability of ice sheets to fail catastrophically, an idea that would become important years later in the search for mechanisms of abrupt climate change.

Among the most powerful and popular of the early theories were those that looked beyond Earth to astronomical circumstances. One proposal was from the realm of celestial mechanics. Certain geometrical features of Earth's orbit of the Sun could account for the big swayings, the long timescale cycles in the ice age data. More directly, another proposal suggested that changes were due to variations in the Sun's luminosity, suggesting a link between the appearance of sunspots and differences in the amount of heat reaching Earth.

At the beginning of the nineteenth century, even before Agassiz began to establish Earth's ice age past, the English astronomer William Herschel was speculating about solar variations, although the very idea encountered deep-seated resistance. Like their climate, what humans want of their life-giving star is stability. By the middle of the nineteenth century, astronomers at least offered regularity, establishing the cyclical nature of sunspots and identifying an 11-year period, a pattern that suggested a link to weather if not climate.

A century later a long, tedious, star-crossed campaign to find correlations between sunspot cycles and changes in weather was under way. Led by American astronomer Charles Greeley Abbott, 1920s and 1930s researchers identified weather-sunspot correlations just about everywhere they looked. For many years, Abbott insisted that he could detect variations in the Sun's luminosity and that, before long, these variations would lead to great improvements in weather prediction. Yet every attempt at prediction was a failure. As Theodore S. Feldman wrote in a recent essay on the subject, "These attempts got a well-deserved bad reputation."

For climate researchers, the search for links to variations in the Sun's output has been more respectably cautious, although only a little more rewarding. While the Greenland ice sheet record would prove to be too short to help resolve many puzzles about long-term ice age patterns, its clarity of detail promised valuable insights into changes on the order of 10 years, that is, almost the duration of the 11-year sunspot cycles. The Camp Century ice core team weighed in on the subject in 1970 with an article by Sigfus J. Johnsen, Willi Dansgaard, and Henrik B. Clausen of Denmark and Chester Langway of the U.S. Army's Cold Regions Research Laboratory. The stable isotope variations going back to the year AD 1200 revealed oscillations with periods of 2,400, 400, 181, and 78 years, they wrote, and all of the periods "seem to originate from changing conditions on the Sun." In a graph depicting their isotope results, the Greenland team boldly extended the cycle out into the future, drawing a curve that showed a cooling of global temperatures to the year 2000.

At Lamont-Doherty Geological Observatory in New York, Wallace S. Broecker looked at these Greenland results and even more ambitiously predicted a different result. "It is possible that we are on the brink of a several-decades-long period of rapid warming," Broecker wrote in 1975. The cooling of global temperatures that had been under way since 1940 was "one of a long series of similar natural climate fluctuations" seen in the Camp Century core that had been counteracting the warming effects of the increasing concentration of carbon dioxide. But the natural trend was about to "bottom out," he wrote. "Once this happens, the CO2 effect will tend to become a significant factor and by the first decade of the next century we may experience global temperatures warmer than any in the last 1,000 years." Looking back over 25 years of warming, Broecker would acknowledge years later that he had "made a large leap of faith" in assuming that the Camp Century cycles of 80 and 180 years were global. Science historian Theodore S. Feldman described the 1975 prediction as "one of several cases where Broecker's scientific instincts were sounder than his evidence."

The short-term Camp Century oscillations, whatever their cause, did not materialize as global phenomena in other data, and the accumulation of other evidence did not back up their supposed link to variations in the Sun's brightness. Decades later, scientists were willing to assume that the minor variations in solar radiance that had been measured by orbiting satellites are among a variety of forces contributing to changes in a climate that already is unstable. As a mechanism that explains ice ages, however, the variability of the Sun had given way to a more powerful astronomical theory.

Since the mid-1970s, most researchers have satisfied themselves that the pattern of climate change at the geological timescale is best explained by one of the oldest theories. Its main elements are features of celestial geometry—the shape of Earth's orbit and the tilt in the planet's axis, that is, its angle of rotation relative to the Sun.

As Johann Kepler established in the seventeenth century, Earth's orbit of the Sun is slightly elliptical rather than circular. The Sun occupies a focus of the ellipse that is offset from the center, which means that during one time of the year the Earth is slightly closer to the Sun than it is at another. As Earth orbits the Sun annually, its elliptical orbit itself also rotates about the Sun, but on a much longer cycle. On another timescale, the orbit changes shape slightly, becoming more or less circular. Every 24 hours, of course, the Earth completes its rotation around its axis, which is currently tilted at a 23.5 degree angle in relation to its orbital plane. That angle, the feature that accounts for the different seasons, varies slightly, between 22 and 25 degrees, at one timescale. On yet another cycle of different duration, Earth's tilting axis wobbles, like a top spinning down. All of these geometric features and their different rates of change—complex in combination but mathematically predictable—alter the intensity of sunlight striking Earth.

The interplay of these properties of celestial geometry produces three periods of oscillation in the climate record—at about 22,000 years, 41,000 years, and 100,000 years—and forms the basis of the generally accepted explanation of the ice ages. This is what scientists refer to as the theory of orbital or astronomical forcing, or the Milankovitch theory. Agreement about it has been a long time coming, and it is still incomplete. As a theory of climate change, it is not very satisfying. One of the most important scientific papers on the subject describes these elements collectively not as the cause but as the "pacemaker" of the ice ages.

In 1842, just five years after Agassiz's "Discourse," Joseph Adhemar, a mathematics teacher in Paris, was first to suggest that the ice ages were related to Earth's orbit of the Sun. The changes of Earth's wobbly tilt through its elliptical orbit meant that the hemispheres alternately experience somewhat shorter summers and longer winters through a 22,000-year cycle. Focusing on this cycle, Adhemar surmised, mistakenly, that one hemisphere would get colder as the other warmed. He sketched an improbable scenario in which ice ages came and went with the waxing and waning of the Antarctic ice sheet. About all that survives of this first theory is the basic premise that astronomical circumstances are involved.

In the 1860s the Scotsman James Croll extended and strengthened the astronomical theory to include very slow changes in the elongated shape, or eccentricity, of Earth's elliptical orbit. Croll argued that ice ages are triggered during epochs when Earth's orbit is most elliptical, in its winter hemisphere, when it is farthest from the Sun. During these epochs, he theorized, ice ages occurred in alternating hemispheres every 11,000 years. Calculating these periods of high eccentricity of the orbit, Croll estimated that the last glacial epoch ended about 80,000 years ago. Although meteorologists questioned the power of relatively minor changes in the shape of Earth's orbit to provoke such dramatic climate changes, Croll identified two amplifying effects—the reflection of sunlight off growing ice sheets and the increasing strength of the equatorial trade winds. Geologists debated the theory for years, but support for it waned as evidence slowly accumulated that the last ice age had ended about 10,000 years ago, rather than the 80,000 years estimated by Croll. By the end of the nineteenth century, scientists had discarded the astronomical theory as an explanation of ice ages.

The theory was resurrected in the 1920s by the Serbian mathematician Milutin Milankovitch, who carefully developed a set of curves that measured the orbital effects on radiation at different latitudes over the past 650,000 years. His work gained widespread recognition when the German climatologist Vladimer Koppen and his son-in-law, Alfred Wegener, incorporated Milankovitch's calculations into the publication of their own research into the geological history of climate. To Koppen and Wegener, the Milankovitch diagrams seemed to fit the evidence, which German geographers had gathered from gravels left by alpine glaciers, that identified four major ice ages during the past 650,000 years. Also, rudimentary estimates that the last European ice sheets disappeared about 20,000 years ago seemed to agree with the Milankovitch curves. For 50 years the pattern accepted by geologists was one of long, warm periods interrupted by four relatively brief ice ages.

Powerful new technologies developed after World War II changed the character of the ice age debate, allowing geologists to more fully test the astronomical theory on their own terms. As Milankovitch himself had observed in 1941: "These causes—the changes in insolation [that is, solar radiation reaching Earth] brought about by the mutual perturbations of the planets—lie far beyond the vision of the descriptive natural sciences." To geologists, who were unfamiliar with the realms of celestial physics and mathematics, the postwar technical advances brought the theory "down to Earth," in a sense, by giving them new geological tools. Nuclear technology brought radiocarbon dating and stable isotope analysis. The Swedish geologist Borje Kullenberg designed a new piston-core device that could retrieve much longer ocean sediment cores. And the discovery of magnetic reversals in the geological record made it possible to correlate data around the globe.

Radiocarbon dating gave researchers their first relatively accurate age estimates going back 40,000 years. In 1957, Wallace S. Broecker had made the application of radiocarbon dating to climate questions the subject of his Ph.D. thesis. He would spend 10 years as director of Columbia University's radiocarbon laboratory at Lamont Geological Observatory and would play a key role in developing theories about both the slow roll of ice ages and abrupt climate change mechanisms.

As it happened, the first radiocarbon results provoked another sudden collapse in scientific support for what had become widely known as the Milankovich theory. In the mid-1950s, geologists discovered a peat layer in Farmdale, Illinois, that was dated at 25,000 years old. Here was a relic of a warm climate that came precisely at a time when Milankovich's calculations called for a solar radiation minimum. Similar discrepancies followed elsewhere in North America and Europe, and by the early 1960s the astronomical theory of the ice ages again was in eclipse.

Broecker did not give up on the concept, however. In 1965, in a conference presentation entitled "In Defense of the Astronomical Theory of Glaciation," Broecker argued for a new model of climate that made important changes to Milankovitch. The climate system has "two stable states, glacial and interglacial," Broecker proposed, and "rapid transitions between these states" are triggered when Earth is receiving maximum radiation from the Sun.

Developing ideas he had proposed in his Ph.D. thesis, Broecker was challenging two tenets of conventional thinking on the subject. The German model of four relatively brief ice ages widely separated by much longer warm periods had given rise to the idea that Earth's climate was in some way naturally inclined to be warm unless perturbed by some exceptional circumstances. And almost invariably, whatever their magnitude, changes could be depicted along a timeline of gently sloping curves. In the years ahead, these two ideas—of two stable climate states and rapid transitions between them—would become central to understanding abrupt change.

In 1968, Broecker helped revive interest in the Milankovitch theory with radioactive isotope tests he conducted on samples of ancient coral taken in Barbados by Robley K. Matthews. Terraces of coral reefs that stand far above the present sea level cover much of the Caribbean island, and Matthews had established that the different terraces were relics of eras of higher sea level. Applying his nuclear laboratory methods to a slow-decaying isotope of thorium, Broecker established that the different terraces of coral reef were formed 125,000, 105,000 and 82,000 years ago. According to the Milankovitch curve for the latitude of 45° North, these were times of maximum solar radiation intensity, when melted ice sheets had caused exceptionally high sea levels. Reporting these results, Broecker called the Milankovitch theory "the number-one contender in the climatic sweepstakes."

At the time, even more compelling information was coming from the sea itself. The technical improvements in the retrieval of ocean floor sediments had brought a whole new archive of ancient climate into the laboratories. These sediments, accumulating at a rate of a few centimeters per thousand years, were proving ideal for studying a problem on the timescale of ice ages. Although their laminations did not reveal the detail found in polar ice, especially Greenland ice, the seafloor sediments were much older, and the new technology allowed researchers to extract long cores that revealed large-scale climate changes going back millions of years.

Researchers used two techniques to tease climate changes out of the sediments. The cores are composed primarily of calcium carbonate skeletons of minute ocean creatures known as foraminifera, or forams, which settle to the seabed over the eons, and these tiny skeletons serve as proxies for changes taking place in the climate over time. Like the polar ice investigators, geologist Cesare Emiliani and geophysicist Nicholas Shackleton employed the new nuclear technique of mass spectrometry that would prove so valuable in Greenland. In the calcium carbonate, they measured the changing proportions of the stable isotopes of oxygen, believing that variations between the common 16O atom and the rare and heavier 18O atom revealed ocean temperatures at the time when the forams lived. Like the German Wolfgang Schott, who pioneered the field in the 1930s, geologist David B. Ericson at Lamont Geological Observatory sampled populations of the forams at different depths and built a temperature profile according to the assemblages of warm-water and cold-water species.

In 1966, Emiliani reported sediment results that he described as a record of ocean temperatures through the last 400,000 years. This research, begun in the mid-1950s, fundamentally changed the picture of ice ages held by two generations of geologists, although it was mired in controversy for years. Eventually, most scientists agreed with Shackleton and others that Emiliani's results reflected changes in global ice sheet volume as well as ocean temperatures. The critics pointed out that because the lighter 16O isotope evaporates more easily from the ocean's surface, the buildup of great ice sheets caused the subsiding seas, whatever their temperatures, to be enriched with 18O. As Willi Dansgaard pointed out, however, a profile of "paleoglaciation" was just as valuable as a profile of "paleotemperatures" For identifying ice ages, what could be better than a reliable record of the comings and goings of glaciers and ice sheets?

For a time the controversy over the meaning of the data tended to obscure Emiliani's more fundamental discovery. His sediment profiles identified many more ice age periods than the conventional four ages described by the German geographers at the end of the nineteenth century. It challenged the very basis of Koppen's and Wegener's support for Milutin Milankovitch in the 1920s. Although the Emiliani curves agreed with the Milankovitch diagrams, they turned the prevailing conception of past climate on its head. Rather than four ice ages, there were many. Rather than an equable climate occasionally interrupted by ice ages, the new pattern was the opposite. More frequent and longer-lasting ice ages were punctuated by much shorter warm periods.

In 1968, the Czechoslovakian geologist George Kukla demolished the old order. Kukla had studied layers exposed at a brick quarry and found ice age layers of windblown silt more frequent and numerous than would be expected by the German model. In fact, he found a frequency of 100,000 years in the ice age cycle. Using the new techniques, Kukla returned to the alpine terraces the Germans had used to construct their ice age profile and found that the gravels contained material of very different ages than they had supposed.

In 1970, Wally Broecker and his Lamont colleague Jan van Donk reported results of isotopic dating of a core taken from the Caribbean that agreed with Emiliani's multiple ice age data as well as the Milankovitch curves. They identified a saw-toothed pattern in the profile, representing long, slow descents into ice age cold that ended in "rapid terminations." The dominant cycle lasted about 100,000 years. Another important breakthrough came in 1973, when Nicholas Shackleton presented a climate profile he had meticulously developed from one of the best sediment cores ever extracted from the seabed. Taken in 1971 by the Lamont research vessel Vema from shallow water in the western equatorial Pacific, the core reached back more than a million years, beyond the time of the last magnetic reversal 700,000 years ago. A new isotope analysis method using radioactive potassium detected evidence of the event in the core. This critical feature allowed Shackleton to calibrate climate events along the time profile of the core, establishing the first accurate chronology of the last 1.5 million years of climate. This "Rosetta Stone" of ice age research also showed a dominant 100,000-year cycle.

In 1976, Shackleton, John Imbrie, and James D. Hays put it all together in the classic paper on the subject in the journal Science. Their study of two cores from the South

Atlantic showed Milankovitch curves at 23,000, 41,000, and 100,000 years that coincide with changes in the geometry of Earth's orbit. The scientists concluded that "changes in earth's orbital geometry are the fundamental cause" of the succession of ice ages, although they frankly acknowledged that they could not explain just how these changes affected climate.

The study confirmed a set of particularly puzzling facts. The cycles that imposed the strongest differences in the strength of sunlight reaching the planet, the 23,000- and 41,000-year cycles, together accounted for about 35 percent of the climate variation over the past 500,000 years, they estimated, and probably were strong enough to influence climate directly. On the other hand, the dominant 100,000-year cycle which they associated with fully 50 percent of past climate variation was too weak to cause these changes directly. In the vernacular, the response was nonlinear, that is, the rules of traditional physics don't apply. One unit of push does not necessarily produce one unit of shove; the system is chaotic and huge changes are possible from minor variations. Something about the climate system caused it to respond all out of proportion to the minor difference in solar radiation that results from slight changes in the elliptical shape of Earth's orbit. Or something else was causing the 100,000-year cycle.

Scientists like to use the word elegant to describe a theory that has the power to reduce complex circumstances to simple explanations—in a single stroke, to bring order from chaos. Plate tectonics is thought of as elegant, for example, as is Darwin's theory of evolution by natural selection. As it happens, however, on the basic question of the causes of ice ages, even when beautifully predictable properties of celestial mechanics are invoked, Earth's climate evidently is not congenial to elegant explanations.

Still, there are lingering doubts, especially about the cause of the 100,000-year cycle, and major mysteries remain unsolved. If anything, this best explanation for ice ages raises the mystery to a new level of complexity. There would be no Eureka! moment for the ice age problem, no occasion when someone would jump up with a solution that would transform the science. Researchers instead were left with the unsatisfying prospect of plowing through the data to identify any number of mechanisms that could amplify this insignificantly weak radiation signal so powerfully from afar. Nonlinear meant that the future, for the time being at least, was just about anybody's guess.

The Milankovitch orbital timetable led Shackleton, Imbrie, and Hays to predict the onset of an ice age; in their words, "the long-term trend over the next several thousand years is toward extensive Northern Hemisphere glaciation." They warned, however, that forecasts based on the astronomical theory came with two critical qualifications. "First, they apply only to the natural component of future climatic trends—and not to such anthropogenic [or human-caused] effects as those due to the burning of fossil fuels. Second, they describe only the long-term trends, because they are linked to orbital variations with periods of 20,000 years or longer."

By the early 1970s, the future of the climate was becoming an issue of increasingly widespread concern. A powerful El Niño in 1972 so wrecked the grain harvest in the

Soviet Union and elsewhere that modern leaders for the first time began to wonder about the security of the world's food supply. A distinct cooling trend had been under way since 1940, and magazine and newspaper articles were raising the specter of a new ice age in the minds of the general public. Moreover, recognition of a climate history dominated by successive ice ages was giving rise to a different picture of the future than most everyone had assumed. The concept of abrupt change had appeared on the horizon, like an uninvited guest. Dansgaard had reported early data from the Camp Century ice core. And two contending ice age theories—Ewing and Donn's Arctic ice scenario and Wilson's Antarctic ice surge—suggested that changes could be both rapid and devastating.

In January 1972, leading climate scientists gathered at Brown University in Providence, Rhode Island, to pool their thinking about the future of the current 10,000-year-old warm "interglacial" climate episode that had cradled the rise of civilization and agriculture. The conference—The Present Interglacial: How and When Will It End?—attracted most of the major climate scientists. Among the contributions was "Speculations About the Next Glaciation," a brief paper by Willi Dansgaard and both American and Danish members of the Greenland ice core team.

Like scientists before and since, the researchers who threw open the Camp Century window and first saw the fine detail in the landscape of climates past had to set their sights over the tops of the canyons and ridges of abrupt change in search of the more distant ranges and divides left by the advances and retreats of glaciers over geological time. Even then, they could not avoid the most striking feature that the variations in the oxygen isotope ratio revealed in the Camp Century ice—a pace of change that had never been seen before.

The Dansgaard team noted that an event they estimated to have taken place 89,500 years in the past had plunged the climate "from warmer than today into full glacial severity" within just a century or even less. In fact, the drop in the oxygen isotope ratio, and thus in temperatures, might have occurred "almost instantaneously." The curve in the climate profile, they wrote, "suggests that it took 1000 [years] to recover from this catastrophic event." Like no other record before it, the Camp Century ice profile was recording climate events on a timescale of centuries and decades, the units of human history. "Catastrophic" and "instantaneous" were not words that geologists were accustomed to using when describing changes in climate.

What could have caused such an event? The Greenland team ruminated on the possibilities. Could ice surges from the Antarctic have triggered such swift changes? It seemed unlikely that events in an ocean at the opposite pole could transmit such changes so abruptly to northern Greenland. Could intense volcanic activity have caused it? No telltale volcanic layer showed up in the Camp Century ice at that depth.

Dansgaard noticed that the event was one of three similarly rapid climate collapses that were recorded deep in the Camp Century ice. All three had occurred during periods when orbital features diminished the radiation from the Sun. Yet only the last event, 73,000

years ago, ushered in an ice age, he observed. The mechanism may be a mystery, but the link to Earth's orbit was irresistible. Among the three abrupt changes, only the last was followed by 60,000 years of slightly lower summer midlatitude solar radiation.

As if setting an agenda, Dansgaard then posed a series of questions that would occupy many climate scientists for the rest of the twentieth century. "Were the sudden decreases in [oxygen isotope ratios] triggered by low insolation in the Northern Hemisphere, where extensive areas of land can be covered by ice? If so, the conditions for a catastrophic event are present today Or, are we faced with more or less accidental events, such as ice surges or intense volcanic activity, that trigger a full glaciation, if the insolation conditions favor such development? Is man's present activity equivalent to such [an] accidental event?"

Most of the conference participants followed the Milankovitch curves and the contemporary cooling trend and suggested that the end of the current warm period and the onset of a new ice age was a strong possibility. Whether it would come in a few centuries or a few thousand years was anybody's guess. Uncertainties about the impact of what Dansgaard called "man's present activity" hung over the proceedings. Cesare Emiliani observed that scientists could not predict the amplitude of change it might provoke or even the direction in might take. But his own research into the ice age problem had established a new and important fact, "that temperatures as high as those of today occurred for only about 10% of the time during the past half million years." After 10,000 years of warmth, the system was in a state of "precarious environmental balance" that makes man's interference "extremely critical."

Everyone agreed that they needed to know more. The profile of change in the Camp Century ice and the increasingly sharp focus emerging from the laminations in ocean sediment cores were revealing a truly baffling history of climate. Although new instruments and ingenuity were shedding new light on the subject, the fundamental problem of the past, the ice age problem, had not been resolved very satisfactorily. Like a microscope turned on a drop of pond water, the new records revealed details of events that were totally unexpected. Rather than answering old questions, the increasing refinements of vision raised entirely new mysteries.

Along the way, a new science of paleoclimatology had been born. Some of it was fashioned from the old disciplines of geology and astronomy, to be sure, but other elements were spanking new, invented in the richly financed postwar laboratories of nuclear technology. Engineering breakthroughs made possible the exploration of climate records in ice sheets and ocean sediments that opened up new archives of climate history. And human ingenuity was devising new ways to tease information from proxies in fossil relics of epochs past.

Many theories had been proposed. Leads had been followed down a lot of "blind alleys." Breakthroughs had been celebrated prematurely. Earth does not reveal its ways willy-nilly, especially the ways of its climate. Besides, that's the way it goes in science. As historian of science Spencer R. Weart has observed, "Every great scientific paper is written at the outside edge of what can be known, and deserves to be remembered if there is a nugget of value amid the inevitable confusion."

+1 0

Post a comment