With Speed and Violence: Why Scientists Fear Tipping Points in Climate Change (24 page)

The pendulum moves too fast for any orbital cycles. Some theorists have suggested a role for volcanic eruptions, which shroud the planet with aerosols that can cool it. It is true that at certain times during the little ice age, there were major eruptions. The year after the eruption of Tambora, in Indonesia, in 1815, crops failed from India to Europe and North America. It became known as "the year without a summer." But volcanic dust clouds cool temperatures for only a few years at most. They may from time to time have exacerbated the cooling, but they were not sufficiently frequent or unusual to explain a cold era that lasted on and off for almost half a millennium.

Most climatologists believe that the sun should get the blame. The coldest part of the little ice age, in the mid-to-late seventeenth century, is known as the Maunder Minimum. The popularizing of the telescope by Galileo a few decades before meant that astronomers of the day were able to note the virtual disappearance between 1645 and 171 5 of the by-thenfamiliar spots on the surface of the sun. This is now recognized as a good indicator of a reduced output of solar energy. The best guess is that solar radiation reaching Earth's surface during the Maunder Minimum fell by perhaps half a watt per 10.8 square feet, or around 0.2 percent. But climatologists find it perplexing that such a widespread effect could result from such a modest change.

Enter an idiosyncratic, larger-than-life researcher working at the Lamont-Doherty Earth Observatory, just down the corridor from Wally Broecker. His name was Bond, Gerard Bond. Like Broecker, he hated getting bogged down in detail, and liked seeing the big picture. Like Broecker, he was willing to fly a kite, trusted his intuition, and had the confidence to propose an idea in public just to see if anyone could shoot it down. And, again like his compatriot, he had the intellectual reputation to get his kite-flying published in the often conservative scientific literature.

Bond argued forcefully until his death, in 2005, that the little ice age and the medieval warm period were the most recent signs of a pervasive pulse in the world's climatic system. This pulse, he said, had a cycle that recurred once every 1,500 years or so. It was a pulse, moreover, that seemed largely unaffected by other, apparently bigger influences on global climate, like the Milankovitch orbital cycles that triggered the major glaciations. Ice age or no ice age, he argued, the pulse just kept on going. Bond didn't invent the pulse out of thin air. Other researchers had unwittingly been on its trail for years. But, like his friend down the corridor, Bond was the man who had the confidence to compose a big picture out of the scattered fragments of evidence.

In the early i98os, a graduate student in Germany made the first breakthrough. While at the University of Gottingen, Hartmut Heinrich was examining cores of sediment drilled from the bed of the North Atlantic. He found a number of curious layers of rock fragments that showed up in cores drilled as far apart as the east coast of Canada, the waters west of the British Isles, and around Bermuda. Radiocarbon dating revealed that these rock fragments were laid down in at least six bands over the 6o,ooo years before the end of the last glaciation, at intervals of roughly 8,ooo years.

I looked at some of these rock fragments in the marine sediment store at Bond's old laboratory in New York. They are enormously distinctive. A browse among the trays of sediment revealed fairly subtle differences among the different cores: a change of color here, a slightly different consistency of dust there. Almost everything in these sediments has gone through the mill of being eroded from Earth's surface, discharged down rivers, and dumped in tiny bits on the seabed. But then there are Heinrich's layers. These are a mass of stones the size of gravel or pebbles, but sharp-edged and clearly untouched by the normal processes of erosion and deposition. Researchers soon gave the events that produced them their own name: Heinrich events. There was nothing like them in the sediment record.

Apart from their size and shape, something else was odd about these rock fragments. Though they had been found way out in the middle of the Atlantic Ocean, geologists swiftly established that they came from the Hudson Bay area of northern Canada. How could they have got so far offshore and so far south? What took them there? The only logical explanation, given that all the Heinrich events took place during the last glaciation, was that they had been ripped from the bedrock by great glaciers and carried south on the underside of icebergs. They traveled a long way because the North Atlantic was extremely cold, and were eventually dumped onto the ocean floor as the icebergs melted. That raised other questions. What climatic events would send vast armadas of icebergs sailing south into the tropics? And why the apparent 8,ooo-year cycle?

The next clue came a few years later, in the early 199os, when a distinguished Danish glaciologist, Willi Dansgaard, of the University of Copenhagen, discovered in the Greenland ice-core record a series of large and sudden temperature changes that again punctuated the last glaciation. Several times, temperatures leaped up by 3.6 to i8°F within a decade or so, before recovering after a few hundred years. So far, more than twenty of these warm phases have been identified in the ice-core record. During many of them, temperatures in Europe at least may have been as warm as today.

These warming events, too, seemed to have some kind of periodicity or pulse. Temperatures moved from cold to warm and back again repeatedly, with a cycle ranging between 1,300 and i,8oo years. It was a recognizable pulse, just as a human pulse that races and then slows is recognizable, and averaged a full cycle roughly every 1,500 years. This pulse also swiftly got a name, the rather cumbersome Dansgaard-Oeschger cycle, after Dans- gaard and his Swiss colleague, Hans Oeschger. Some interpret the data as showing a continuous background temperature cycle that on most but not all occasions triggered a more substantial warming episode during its warm phase, and on rather fewer occasions triggered a Heinrich event during its cold phase.

The connection between Heinrich events and the Dansgaard-Oeschger cycle wasn't recognized immediately-understandably enough. They had different time signatures, and one was revealed in the sediments of the mid-Atlantic, while the other emerged from the Greenland ice cores. Both, in any case, seemed at first to be minor local curiosities confined to the last glaciation, and therefore of no relevance to modern climate. But Bond had a hunch that the two were linked in some way, and that they had a global significance. Both, he noted, appeared to coincide with other climate changes, such as the advances and retreat of glaciers in Europe and North America. Like the Younger Dryas event and the climate flip 8,200 years ago, they seemed either to push the world into a different climate mode or to be part of such a process. Down the corridor, Bond's buddy Broecker was on hand to suggest a possible link to the ocean conveyor. The story began to take on a life of its own. But first the pair needed evidence to back up their hunch.

Bond began to re-examine trays of sediment cores from the bed of the North Atlantic that were assembled in his New York archive. Some were old cores, taken years before by the Lamont-Doherty research vessel Vema from beneath the waters off Ireland and the channel between Greenland and Iceland. Others were new, drilled off Newfoundland under Bond's supervision.

As expected, Bond found further evidence of Heinrich's rock fragments roughly every 8,ooo years or so through the last glaciation. But the marine sediment cores also revealed lesser layers of materials normally alien to the seabed of the North Atlantic. Most exciting of all, these lesser layers occurred roughly every 1,500 years, and appeared to coincide with the cold phase of the Dansgaard-Oeschger cycle in the Greenland ice cores. This was pay dirt. Doubly so when it became clear that the iceberg armadas of the Heinrich events occurred during unusually cold phases of the Dansgaard-Oeschger cycle. The pattern seemed to involve a large Heinrich event, followed by five less and less severe i,5oo-year DansgaardOeschger cycles, and then another big Heinrich event. Sometimes this stately progression is influenced by other cycles, such as a solar precession, but otherwise it seems to hold.

Most remarkable of all, perhaps, Bond found that although there have been no Heinrich events during the io,ooo years since the end of the last ice age-the last was 15,000 years ago-the marine imprint of the underlying 1,500-year pulse has not missed a beat. "The oscillations carry on no matter what the state of the climate," he said.

Bond died in 2005, at the age of sixty-five. His longtime colleague Peter deMenocal has continued his work, looking for more signs of the pulse. Examining seabed sediments off Africa's west coast, he has found that every 1,500 years or so there were huge increases in dust particles in the sediments, suggesting big dust storms on land. The sediments also revealed dramatic increases in the remains of temperature-sensitive marine plankton, suggesting a temperature switchback in tropical Africa of as much as 9°F. "The transitions were sharp," deMenocal says. "Climate changes that we thought should take thousands of years to happen occurred within a generation or two."

Bond's final claim, that the pulse can be seen in recurrent climatic events right through to the present, seems to be vindicated, especially by temperatures in Europe and North America. There was an especially strong cooling event in the Northern Hemisphere that ended around 2,000 years ago; it was replaced by the medieval warm period that reached its height perhaps r,roo years ago, and then by another cold era that bottomed out around 350 years ago, during the Maunder Minimum-when temperatures fell by up to 3.6°F in northern Europe, and the Eskimos reached Scotland in their kayaks.

Bond's study was an extraordinary piece of detective work. But it raises more questions than it answers. Two stand out. What, if any, is the relationship between these cycles and other parts of the climate system, such as Broecker's ocean conveyor? And, of course, what causes the mysterious pulse?

Heinrich originally argued that his ice armadas must be the result of some instability in the North American ice sheet that caused periodic collapses into the North Atlantic. There might thus be some link to big freshwater breakouts like that which triggered the Younger Dryas event. Certainly they involved huge amounts of ice. But the timing is fuzzy. Bond argued that while instabilities in the ice sheet could explain Heinrich events, only some of his pulses produced Heinrich events. So instability in ice sheets is unlikely to explain the pulses themselves, which in any case seem to have been unaffected by glaciations. By 2001, Bond believed he had confirmed the answer that many suspected all along.

He went back to the Greenland ice cores to look for evidence of solar cycles. There is no known direct marker for solar cycles in the cores. But other researchers had discovered that isotopic traces of cosmic rays bombarding the atmosphere were left in the ice cores-and that when solar radiation is at its most intense, cosmic rays are literally blown away from the solar system. Thus fewer "cosmogenic" isotopes, like carbon-14 and beryllium-io, are left in the ice cores during periods of strong solar radiation.

Bond came up trumps again. The evidence tallied. Over the past 12,000 years, fluctuations in detritus from the iceberg armadas in the Atlantic coincided with changes in cosmogenic isotopes in the ice cores. Thus there was a solar pulse that translated into a pulse in icebergs, global temperatures, and recurrent climatic events found through both the glacial and the postglacial eras.

Bond was convinced before his death that most climate change over the past io,ooo years had been driven by his solar pulse, amplified through feedbacks such as ice formation and the changing intensity of the ocean conveyor. He worried that people might interpret this as showing that global warming was natural. "But that would be a misuse of the data," he told me in an interview shortly before his death. Rather, he said, the most important lesson from his research is what it shows about the sensitivity of the system itself: "Earth's climate system is highly sensitive to extremely weak perturbations in the sun's energy output." And if it is sensitive to weak changes in solar forcing, it is likely to be sensitive also "to other forcings, such as those caused by human additions of greenhouse gases to the atmosphere."

What, exactly, drives the amplifications is another matter, however. For years, as Bond worked on his ideas, Broecker had declared that the Dansgaard-Oeschger temperature cycle in Greenland was linked to fluctuations in his ocean conveyor. Certainly the geography seemed right. Both appeared to originate in the far North Atlantic. It seemed clear, too, that during the periods when ice armadas were floating south in the Atlantic, temperatures in the North Atlantic were cold, and the amount of deep water being formed around Greenland declined. In extreme casesperhaps during full-scale Heinrich events-the conveyor probably shut down. Perhaps a reduction in solar radiation triggered the entire sequence. But the evidence of what caused what was largely circumstantial. And as we will see later, there is another explanation, producing a large amplification from another quarter entirely.

But whatever the amplifier, the pulse is real and extremely pervasive. In the postglacial era, perhaps only in the past fifty years has something come along with greater power to disrupt climate.

TROPICAL HEAT

26

THE FALL

The end of Africa's golden age

If there was a golden age for humans on Earth-a Garden of Eden that flowed with milk and honey-then it was the high point of the Holocene, the era that followed the end of the last ice age. From around 8,ooo to around 5,500 years ago, the world was as warm as it is today, but there appear to have been few strong hurricanes and few disruptive El Ninos; and it was certainly a world in which the regions occupied today by great deserts in Asia, the Americas, and especially Africa were much wetter than they are now. Optimists suggest that such conditions might await us in a greenhouse world. As we shall see, there are celestial reasons why that might not happen. But the Holocene era, and its abrupt end, may still offer important lessons about our future climate in the twenty-first century.