With Speed and Violence: Why Scientists Fear Tipping Points in Climate Change (20 page)

Agassiz pondered the significance of this. He realized that he had seen similar marks in the landscape in many parts of Europe that were distant from present-day glaciers. He heard similar reports of glacial scratch marks from across North America. And he read contemporary newspaper stories of perfectly preserved mammoths being dug from the snow in Siberia, their meat so fresh that it was fed to local dogs and scavenged by polar bears. The only explanation, he concluded, was that much of the Northern Hemisphere must once have been covered by ice, and that the event happened very suddenly, in a vast, icy apocalypse. "The land of Europe, previously covered with tropical vegetation and inhabited by herds of great elephants, enormous hippopotami and gigantic carnivores, was suddenly buried under a vast expanse of ice," he wrote. "The movement of a powerful creation was supplanted by the silence of death."

Agassiz's vision was like a creation myth in reverse. Advances in geology soon revealed that not one ice age but a whole series of glaciations had occurred, separated by warm periods like our own. But his picture has oth erwise survived remarkably intact. Indeed, recent evidence has revived his original idea that the onset of the last ice age must have been rather fast, with temperatures crashing in a couple of hundred years at most, and very probably much less.

We now know that two main ice sheets formed. One stretched from the British Isles across the North Sea and Scandinavia, and then west through Russia and western Siberia, and north across the Barents Sea as far as Svalbard. A second, even larger sheet covered the whole of Canada and southern Alaska, with a spur extending over Greenland. A smaller sheet sat over Iceland, and the seas around were full of thick floating ice. Strangely, northern Alaska and eastern Siberia, though deep-frozen, were never iced over. But, combined with the older ice covering Antarctica, these ice sheets contained three times as much ice as is present on Earth today-enough to keep sea levels worldwide some 400 feet lower than they are now-and covered 30 percent of Earth's land surface. The ice sheets were high as well as broad, rising up to 2.4 miles above the land surface. They chilled the air above and acted as a barricade for the prevailing westerly winds, which were forced south, skirting the ice sheets. This perpetuated the ice sheets, since the winds would have been the likeliest source of warmth to melt them.

Temperatures fell by around 9°F as a global average, but were 36 degrees lower than they are today in parts of Greenland, and just 5.4 degrees lower in the western Pacific Ocean. The world beyond the ice sheets became dry and cold. Deserts covered the American Midwest, France, and the wide lands of Europe and Asia between Germany and the modern-day Gobi Desert, in Mongolia. Farther south, the Sahara Desert expanded, the Asian monsoon was largely extinguished, and the tropical rainforests of Africa and South America contracted to a few refuges surrounded by grasslands. At the low point, around 70,000 years ago, even the grasslands were largely extinguished, leaving huge expanses of desert, from which winds whipped up huge dust storms. Humans lived by hunting on the plains and hunkering down in the small areas where lush vegetation persisted despite the cold and arid conditions.

It was clear from the start that something drastic must have triggered all this. Astronomical forces were suggested early on-in particular, the idea that the gravitational pull of other planets in the solar system, such as Jupiter, could influence the steady changing of the seasons, and in that way cause glaciers and ice sheets to grow. Many scientists of the day played with this idea. But the first man to subject it to detailed analysis was the son of a Scottish crofter with virtually no formal learning, but a passion for selfeducation and an extraordinary streak of diligence. James Croll was a shy, large-framed man with big ambitions. He stumbled on the idea of an astronomical cause for the ice ages while reading in libraries; transfixed, he spent most of the i86os and r87os pursuing the idea. He took numerous jobs, from insurance salesman to school caretaker to carpenter, in order to finance his passion.

Astronomical forces, he discovered, have three principal effects on Earth, all of which slightly alter the distribution of the solar radiation that reaches it. The effects are greatest in polar regions, where they can alter the amount of sun by as much as ro percent. First, they change the shape of Earth's annual orbit around the sun. The orbit is not circular but slightly elliptical, and the shape of this ellipse changes according to the gravitational pull on Earth of the other orbiting planets. This "eccentricity" in Earth's orbit has a cycle that repeats itself about every ioo,ooo years.

As well as orbiting the sun once every year, Earth spins, making one revolution every day. But the axis around which it spins is not quite at a right angle to the direction of its orbit around the sun. So looked at from space, Earth appears to be spinning on a slight tilt. The combination of the orbit around the sun and the tilt of Earth's axis is what gives us our seasons, because it means that at certain times of the year the Northern and Southern Hemispheres see more or less of the sun. But this situation is not static. Astronomical forces also gradually alter the tilt of the axis. This change in Earth's "inclination" causes a difference in the intensity of the seasons. It has a 4 r ,ooo-year cycle.

Finally, there is a further wobble in the axis around which Earth rotates, called the precession. This is exactly like the wobble that affects a spinning top. It influences the time of year when the different hemispheres are farthest from or nearest to the sun. It is complicated by its relationship with the other two effects, but it repeats on a cycle of I9,ooo to 23,000 years. Currently the Northern Hemisphere has its summer, and the Southern Hemisphere has its winter, when Earth is farthest from the sun; Io,ooo years ago, it was the other way around.

It turns out that the eccentricity of Earth's orbit around the sun drives the ioo,ooo-year cycles into and out of ice ages. Meanwhile, the other two effects, especially the precession, seem to trigger the short warm episodes that punctuate each ice age.

Croll realized that, averaged over a year, these changes made little difference to the amount of solar radiation reaching Earth. The overall effect was probably less than 0.2 watts per io.8 square feet. But the changes did alter where and when the sun hit. Croll calculated in great detail how these influences waxed and waned over tens of thousands of years. And he established, at any rate to his own satisfaction, that they coincided with what geologists were then discovering about the timing of Earth's progress into and out of ice ages.

Taken together, the changing orbital shape, planetary tilt, and rotational wobble alter the strength of seasonality, making summers and winters more or less intense. And it was this that triggered the growth of ice sheets on land in the Northern Hemisphere, he said. Ice sheets would grow when northern winters were coldest. That would be when Earth was farthest from the sun, and when changing tilt ensured that it received the least sunlight. Once ice sheets started to grow, they would reflect ever more sunlight back into space, intensifying the cooling. Croll realized, too, that there was much less room for ice sheets to spread in the Southern Hemisphere, because they were confined to the continent of Antarctica. So the Northern Hemisphere would dominate events, driving the overall heat budget of the planet. But, he suggested, other feedbacks, such as changes to winds and ocean currents, could help drive the world further into an ice age.

In fact it turned out that Croll was wrong in assuming that it was cold winters that were critical. Later research proved that cold summers gave the world a bigger kick into ice ages, by providing little chance for winter accumulations of snow to melt. Nonetheless, Croll's work was a breathtaking piece of sustained cogent analysis that opened up a new field -much as Arrhenius did later with his examination of the impact of changing carbon dioxide levels on climate.

Croll's theory won him a few medals. But, being of low birth and of a taciturn disposition, he never fitted into the scientific salons of the day. They quickly tired of him and his ideas. Croll spent the last decade of his working life as the resident surveyor and clerk at the Scottish Geological Survey, in Edinburgh. To the last, he had to do his research in his own time. By the end of the nineteenth century, Croll and his ideas were largely forgotten. Even Arrhenius, who might have been expected to understand the importance of his work, dismissed it as an unwelcome rival to his own ideas, though in fact it complemented them.

Today, the idea that astronomical forces influence the formation of ice sheets is back in vogue and probably here to stay. Proof of its worth finally came in the 197os. The British geophysicist Nick Shackleton carried out painstaking isotopic analysis of sediments on the ocean floor and in the process finally dated the glacial cycles sufficiently accurately to make clear their association with astronomical events. But even as the textbooks have been rewritten, Croll has been largely lost from the story. The orbital changes that he analyzed so painstakingly are known universally as the Milankovitch wobbles, after Milutin Milankovitch, a balding, monocled mathematician from Serbia who revived and elaborated Croll's ideas in the early twentieth century.

While Croll and Milankovitch have established to most people's satisfaction that orbital changes are the pacemaker of the ice ages, they did not by any means clear up the processes involved. How did a small change in the distribution of solar heating get amplified into a global freeze on a scale probably not seen since Snowball Earth thawed 6oo million years before? And why, among a series of different wobbles, was it just one, with a return period of roo,ooo years, that had much the greatest impact on global climate? A wobble, moreover, with an apparently weaker effect than the others on solar radiation reaching Earth. It seems, in the words of Dan Schrag, a geochemist at Harvard University, that Earth's system contains "powerful embedded amplifiers that can make it highly sensitive to relatively small forcings." Or, as Richard Alley would put it, we have a drunk on our hands. Identifying those amplifiers is important, not least because it should help answer how Earth's climate system might respond to our interference in its actions today.

Croll believed strongly in the power of growing ice itself to amplify cooling, and there is plenty of evidence to support the strength of this icealbedo feedback. Once snow began to accumulate in the Canadian highlands around Hudson Bay, the ice sheet tended to grow of its own accord by cooling the area around it. Jim Hansen calculates that at the height of the last glaciation, it reduced the amount of heat absorbed by the planet's surface by some 4 watts per io.8 square feet. What has troubled researchers rather more is exactly what limited it. Why, after reaching their greatest extent about 21,000 years ago, did the ice sheets begin to retreat?

Given the power of the ice-albedo feedback, it is far from clear why the ice sheets did not continue to grow until they had covered the entire planet and created a comeback for Snowball Earth. Even a change in the wobble to end the change in seasonality that started the ice growth might not have been enough. And it certainly would not explain the extremely fast collapse of the ice sheets at the end of the last glaciation. They disappeared more than ten times as quickly as they had arrived. Some fast feedback must have taken hold. One suggestion is that the sheer size of the ice sheets shut down further growth and eventually caused their rapid destruction. The main theory is that ice sheets are vulnerable to attack by heat rising from the interior of the planet. Trapped beneath the ice, it would have become of increasing importance as the sheets grew. Eventually, the theory goes, some threshold was passed, and the ice sheets melted from their base, creating a giant, continent-wide version of one of Hansen's "slippery slopes," with great chunks of ice skating into the ocean.

The second feedback that converted a planetary wobble into an ice age was greenhouse gases. Anyone who doubts the role of carbon dioxide in climate change should look at the graphs of atmospheric temperatures and of carbon dioxide levels in ice cores taken from the Greenland and Antarctic ice sheets. They cover the past half-million years, a period that includes several glaciations. Throughout, the two graphs are in lockstep. As carbon dioxide levels fall, so do temperatures, and vice versa. That does not determine which leads, but it clearly shows that they are engaged in a very intimate dance, in which carbon dioxide must amplify temperature changes even where it does not initiate them.

As temperatures fell at the start of each glaciation, around 22o billion tons of carbon left the atmosphere, returning during the brief interglacial periods. Its disappearance was enough to directly reduce Earth's uptake of solar energy by about 2 watts per io.8 square feet. But what triggered this big shift in the planet's carbon cycle, and where did the carbon go? It certainly did not end up in vegetation on land, since that was shrinking as the world cooled. The obvious answer is the oceans. There are today about 44 trillion tons of carbon dissolved in the oceans-fifty times as much as in the atmosphere. So a minor uptake of carbon by the oceans could have had a huge effect on the atmosphere.

How might this have happened? Physics will help. Colder water (as long as it has not frozen) dissolves carbon dioxide better than warmer water. But most researchers believe that there must be some more dynamic feedback involved. To take a cue from Gala, life is the obvious force here. One idea is that the initial cooling made the oceans more biologically productive. Plankton, the meadows of the oceans, do like colder temperatures. That is why the Southern Ocean around Antarctica is today one of the most productive. As the plankton grew, they drew more carbon dioxide out of the atmosphere. This strengthening of the biological pump would probably have been encouraged by enhanced dust storms, created by stronger winds and spreading deserts, which would have distributed mineral dust across the oceans. Even today, iron and other minerals are the limiting factor on the fecundity of much of the ocean food chain.