With Speed and Violence: Why Scientists Fear Tipping Points in Climate Change (33 page)

This apparently obscure debate could matter a great deal in the twentyfirst century. Right now, the world has become worried that melting ice in the Arctic could freshen the far North Atlantic and shut down the Gulf Stream. This is a real fear. But maybe, while we are researching that pos sibility, we are ignoring the risk that large stores of freshwater in the Antarctic might break out and disrupt deepwater formation there. Arguably, the risks are far greater in the South, where, besides the potential breakout of ice from Pine Island Bay, recent radar mapping studies have revealed a large number of lakes of liquid water beneath the ice sheets of Antarctica. They might set off a cascade of freshwater into the Southern Ocean, similar in scale to the emptying of Lake Agassiz. Yet nobody, so far as I am aware, has studied what the effects of such a breakout might be for deepwater formation and the Southern arm of the ocean conveyor.

Or, rather than shutting down deepwater formation in Antarctica, might we be about to trigger a switch in the bipolar seesaw, so that deepwater formation in the South takes over from that in the far North? Could that switch be flipped in the South, rather than in the North? And if so, how? And what might happen? It would certainly lead to the Southern Hemisphere's hanging on to very large amounts of heat that currently head north on the Gulf Stream. The Southern Ocean might warm dramatically while the North Atlantic froze. And if the Southern Ocean were to warm substantially, says Will Steffen, the former head of the International Geosphere-Biosphere Programme, "it could result in the surging, melting, and collapse of the West Antarctic ice sheet." Ouch.

If anybody doubts that plenty of new surprises are waiting to be discovered, then the work by Drew Shindell, of the Goddard Institute for Space Studies (GISS), should offer food for thought. His story starts with an apparent success for climate modelers. Since the days of Arrhenius, most climate models have predicted that global warming will be greatest at high latitudes, where known feedbacks like ice-albedo are most pronounced. So rises in temperatures of up to 5°F over parts of the Arctic and the Antarctic Peninsula in recent decades have often been taken as the first proof of man-made climate change.

But there has been a persistent and troubling counterargument. The warming in the polar regions appears to be linked to two natural climatic fluctuations, one in the North and one in the South. In the North, the fluctuation is known as the Arctic Oscillation, an extension of the betterknown North Atlantic Oscillation. It is the second largest climate cycle on Earth, after El Nino. The oscillation itself, as measured by meteorologists, is a change in relative air pressure, but its main impact is to strengthen or weaken the prevailing westerly winds that circle the Arctic. Like El Nino, the Arctic Oscillation flips between two modes. In its positive mode, air pressure differences between the polar and extrapolar regions are strong, and winds strengthen. Especially in winter, the winds take heat from the warm oceans and heat the land. So, during a positive phase of the Arctic Oscillation, northern Europe, Svalbard, Siberia, the Atlantic coast of North America, and Alaska all warm strongly. Likewise, when the oscillation is in its negative phase, the winds drop and the land cools.

The strength of this effect depends on the warmth of the oceans, and in particular on the Gulf Stream and the health of the ocean conveyor. But for most of the past thirty-five years, the Arctic Oscillation has been in a strongly positive mode, helping sustain a long period of warming. Modeling studies suggest that at least half of the warming in parts of the Northern Hemisphere is directly due to its influence, leaving global warming itself apparently a bit player. Except that there is growing evidence that global warming is driving the Arctic Oscillation, too. And it does so from a surprising direction.

Enter Shindell. He likes to occupy the unpopular boundaries between scientific disciplines. His particular interest is the little-studied relationship between the stratosphere, the home of the ozone layer, and the troposphere, where our weather happens. He studies this with the aid of the GISS climate model, one of the few that can fully include the stratosphere in its calculations. Most models show little relationship between global warming and the Arctic Oscillation. The GISS model is the same when the stratosphere is not included. But Shindell discovered that when the stratosphere is hooked up, the result is a huge intensification of the Arctic Oscillation and the westerly winds around the Arctic. In fact, with current levels of greenhouse gases, he has reproduced a pattern very similar to the current unusually strong positive state of the oscillation.

What is going on? One of the problems with climate models is that it is not always easy to pinpoint exactly which of the elements in the model is causing the effects that you see in the printout. But here the role of the stratosphere is clear. And Shindell reckons he has the links in the chain ex plained, at least. As greenhouse gases cool the stratosphere, this cooling alters energy distribution within so as to strengthen stratospheric winds. In particular, a wind called the stratospheric jet, which swirls around the Arctic each winter, picks up speed. This wind, in turn, drives the westerly winds beneath, in the troposphere. So they go faster, too. In this way, a stratospheric feedback is amplifying global warming in the Arctic region by pushing the Arctic Oscillation into overdrive and strengthening the winds that warm the land. It is a brilliant, startling, and, until recently, entirely unexpected feedback.

Might the same apply to events in Antarctica? The GISS model suggests so. There, the dominant climatic oscillation is the Southern Hemisphere annular mode, or SAM. Like the Arctic Oscillation, the SAM is a measure of the air pressure difference between polar and nonpolar air that drives westerly winds sweeping around Antarctica. The geography is somewhat different from the Arctic's. The winds whistle around the Southern Ocean and hit land only on the Antarctic Peninsula, which juts out from the Antarctic mainland toward South America.

The climatologist John King has studied the SAM for the British Antarctic Survey. He says that, like the Arctic Oscillation, it has been in overdrive since the mid-i96os, driving stronger westerly winds. And, again like the Arctic Oscillation, it is amplifying warming along its path. The Antarctic Peninsula has seen air temperatures rise by 5'F since the 196os-the only spot in the Southern Hemisphere to show warming on this scale. The effects include the melting of the peninsula's glaciers and the dramatic collapse of its floating ice shelves, such as the Larsen B. Additionally, by bringing more warm air farther south, the SAM winds are warming the waters that wash around the edges of Antarctica and beneath its ice-helping destabilize the West Antarctic ice sheet.

Here again, Shindell's model suggests that the strengthening of the SAM is the product of a cooling stratosphere and a strengthening of stratospheric jets. There is an important additional element here in the thinning ozone layer, which makes an additional contribution to stratospheric cooling.

All this is alarming evidence of a new positive feedback that intensifies warming in two particularly sensitive regions of the planet, where that ex tra warming could unleash further dangerous change. Glaciologists say that the Greenland ice sheet could collapse if warming there reaches 5°F. The huge stores of methane beneath the Siberian permafrost and the Barents Sea could be liberated by similar warming. And "the SAM warming now includes parts of the West Antarctic ice sheet, as well as the Antarctic Peninsula," says Shindell's boss, Jim Hansen. "This is a really urgent issue."

The discovery of the stratospheric feedback also helps answer another question that has long bothered climate scientists: Why do variations in solar output that are probably no more than halfa watt per i o.8 square feet cause the big climate fluctuations in the North Atlantic identified by Gerard Bond in his analysis of the 1,5oo-year solar pulse? Conventional climate models without a stratospheric dimension suggest that such a solar fluctuation shouldn't produce temperature changes of more than o.35°F. But, although the global temperature change may well have been close to that, in parts of Europe and North America the pulses produce changes ten times as great.

Researchers have struggled to find amplifying mechanisms that might have caused that. Sea ice, the ocean conveyor, and tropical flips like El Nino have all been suggested, but none seems up to the task. Shindell says the answer is his stratospheric feedback. The heart of the mechanism this time is ultraviolet radiation. While the total solar radiation reaching Earth's surface during Bond's pulses varies by only a tenth of a percentage point, the amount of ultraviolet radiation reaching Earth changes by as much as io percent. Most of the ultraviolet radiation is absorbed by the ozone layer in the stratosphere, so its impact at ground level is small. But the process of absorption causes important changes in energy flows in the stratosphere. These eventually change the stratospheric jets, and with them the Arctic Oscillation in the Northern Hemisphere and the SAM in the South.

Shindell modeled the likely effects of the last reduction of solar radiation at the Maunder Minimum in the depths of Europe's little ice age, 350 years ago. The GISS model without the stratosphere was unmoved by the tiny change in solar radiation. But with the stratosphere included, it delivered a drop in temperatures of i.8 to 2.6°F in Europe, but only a tenth as much globally-results remarkably close to likely events in the real world. The declining flows of ultraviolet radiation into the stratosphere triggered a slowdown in the westerly winds at ground level, says Shindell. That, in turn, caused winter cooling, particularly over land, in the higher latitudes of the Northern Hemisphere.

The stratosphere and its influence on polar and midlatitude winds thus seem to be a hidden amplifier that can turn small changes in solar radiation into larger changes in temperature in the polar regions of the planet. This is not the only amplifier in those regions. Ice and snow are important, along with the ocean conveyor and, maybe, methane. But it appears to be the critical ingredient that turns minor solar cycles into big climatic events. It makes sense of Bond's solar pulse and, perhaps, of tiny short-term variability in solar radiation.

Climate skeptics have sometimes argued that sunspot cycles correlate so well with warming in the twentieth century that greenhouse gases could be irrelevant. Mainstream climate scientists dismissed this idea because they could not see the mechanisms that might make this happen. The changes in solar radiation seemed much too small. Shindell's finding of a powerful stratospheric feedback to the solar signal have forced a rethink. But Shindell has not joined the climate skeptics. Far from it.

His conclusion is that for the first half of the century, the correlation between estimated solar output and Earth's temperature is not bad. And the stratospheric feedback might show how the sun could have driven some warming early in the century, followed by a midcentury cooling that made some fear an oncoming ice age. But since then, there has been no change in the solar signal that could be amplified to explain the recent warming. During the final three decades of the twentieth century, average solar output, if anything, declined, while global temperatures-not just at high latitudes but almost everywhere-surged ahead at what was probably a record rate. So, Shindell says, "although solar variability does impact surface climate indirectly, it was almost certainly not responsible for most of the rapid global warming seen over the past three decades."

For that most recent period, he says, it is clear that rising concentrations of greenhouse gases are the primary driver. But besides producing a general global warming, they have generated changes in the stratosphere that have produced a specific positive feedback to warming in the polar regions and the midlatitudes. The positive feedback has manifested itself through the apparently natural Arctic Oscillation and the SAM-cycles that appear to have gone into overdrive.

Only a fool would conclude from this that we don't need to worry so much about man-made climate change. On the contrary, Shindell's dramatic discovery of the stratospheric feedback suggests that the natural processes of temperature amplification are much stronger than those in most existing climate models. His newly discovered feedback seems set to continue, driving up temperatures in Arctic regions beyond the levels previously forecast. That additional warming is likely to unleash other feedbacks that will melt ice, raise sea levels, release greenhouse gases trapped in permafrost and beneath the ocean bed, and perhaps cause trouble for the ocean conveyor.

Relieved? I don't think so.

CONCLUSION: ANOTHER PLANET

Over the past ioo,ooo years, there have been only two generally stable periods of climate, according to Richard Alley. The first was "when the ice sheets were biggest and the world was coldest," he says. "The second is the period we are living in now." For most of the rest of the time, there has been "a crazily jumping climate." And now, after many generations of experiencing global climatic stability, human society seems in imminent danger of returning to a world of crazy jumps. We really have no idea what it will be like, or how we will cope. There is still a chance that the jumps won't materialize, and that instead the world will warm gradually, even benignly. But the odds are against it. There are numerous feedbackswaking monsters, in Chris Rapley's words-waiting to provide the crazy jumps. Climatically, we are entering terra incognita.

The current generation of inhabitants of this planet is in all probability the last generation that can rely on anything close to a stable global climate in which to conduct its affairs. Jim Hansen gives us just a decade to change our ways. Beyond that, he says, the last thing we can anticipate is what economists call "business as usual." It will be anything but. "Business as usual will produce basically another planet," says Hansen. "How else can you describe climate change in which the Arctic becomes an open lake in the summer, and most land areas experience average climatic conditions not experienced before in even the most extreme years?"