Often when we think of the greenhouse effect, we think of islands and coastal cities, of beaches and estuaries--and even one-sixth of Bangladesh--taken over by a rising sea. Such projections rely on a simple logic: as the planet gets warmer, the ice sheets on it should begin to melt and sea level should rise. But that logic, it now seems, is not unassailable. A number of studies over the past few years have revealed evidence for the notion that polar ice sheets may actually grow in the face of greenhouse warming--and sea level may drop, or at least not rise as fast as had been feared.
Some of the most compelling evidence comes from the geologic record of past ice ages. By measuring the proportion of oxygen isotopes in deep-sea sediments, geologists have been able to track changes in the amount of ice on Earth’s surface--and thus changes in sea level--over hundreds of thousands of years. (As water evaporating from the ocean surface gets locked up in continental ice sheets, the water molecules containing the heavier isotope of oxygen tend to remain behind in the ocean.) The oxygen record shows, for instance, that the last glaciation began with a rapid ice buildup some 120,000 years ago.
But this year the circumstances of the ice buildup were put in an interesting new light by Gifford Miller of the University of Colorado in Boulder and Anne de Vernal of the University of Quebec at Montreal. De Vernal, a marine micropaleontologist, has studied the fossils of tiny marine algae that she has culled from 120,000-year-old seafloor sediments in Baffin Bay, the Labrador Sea, and the northwest Atlantic. Different species of algae thrive in water of different temperatures, so the types of algae found in sediments are a measure of the sea-surface temperature at the time the algae rained down onto the seafloor. Similarly, pollen grains in coastal sediments reveal what the climate was like on the neighboring land.
While De Vernal was looking at sediment cores, Miller, working independently, was determining the ages of glacial deposits on the eastern coast of Baffin Island. It was the remarkable parallelism of our two totally independent data sets that got us thinking about the implications, says Miller. What they found was this: Ice was building up 120,000 years ago, all right, and it was building up in the Canadian Arctic--but at a time when the world in general was as warm as it is today. Indeed, the sea surface around the Arctic was warmer.
How can ice sheets grow in a warm climate? The answer is really very simple. The Arctic is so far north that even in a warmer climate than today’s the summers still wouldn’t be warm enough to melt much ice. But the warmer temperature of the sea surface would cause more water to evaporate. Winds would carry this moisture over the land, where in winter it would precipitate out as snow and, as the summers failed to melt it, become transformed into ice. At the beginning of the last ice age, moreover, Arctic summers were getting cooler, thanks to cyclical changes in Earth’s orbit that reduce summer sunlight in the north.
The net effect, say Miller and De Vernal, was that enough water was taken out of the ocean and locked up in ice sheets to cause sea level to drop by more than two feet a century. Once ice sheets formed, they helped cool the planet down by reflecting sunlight back into space. But the ice came before the cold. And it was the warm climate at the beginning of the glaciation that provided the precipitation needed to form the ice in the first place.
Miller and De Vernal also see evidence for this scenario--that warmth leads to ice-sheet growth--in the more recent past. The ice age that began 120,000 years ago reached its final peak around 18,000 years ago. After that the North American ice sheet began to recede. But between 9,000 and 8,000 years ago it expanded again in the Arctic--at a time when the rest of the planet was warmer than it is today.
A few thousand years later, a similar event took place in the Southern Hemisphere. Eugene Domack of Hamilton College in Clinton, New York, and his colleagues have been studying the history of the South Polar ice cap by examining sediment cores hauled up from the Antarctic continental shelf. The sediments dating from periods when the continental shelf was an open sea are rich in plankton; those dating from periods when the shelf was covered with ice consist primarily of rocks and pebbles dropped by the ice. This sediment record shows that between 7,000 and 4,000 years ago, when Earth was well into the present interglacial period and the temperature around Antarctica was about four degrees warmer than it is today, the Antarctic ice sheets were growing again.
This was surprising to us, says Domack. But it is consistent with models that suggest you could warm the area by up to nine degrees before the excess melting would surpass the increase in precipitation and snowfall due to the warmth. And this suggests that under future global warming you would have a net negative contribution to sea level from the Antarctic, rather than a net positive one.
Indeed, that is precisely what is happening right now, according to Charles Bentley of the University of Wisconsin. While De Vernal, Miller, and Domack have been tracking the waxing and waning of ice sheets in the historical record, Bentley and Mario Giovinetto of the University of Calgary have been monitoring the condition of the Antarctic ice sheet today, balancing data on the amount of snow falling over Antarctica against the amount of ice breaking away from the edges of the ice sheet. They calculate that the Antarctic is already sopping up enough water each year to lower the ocean two-hundredths of an inch--apparently, says Bentley, because more snow is falling on the ice cap.
The warmer air is, the more moisture it can hold, he explains. In Antarctica the moisture-carrying air comes in over the continent, and before it leaves again, it drops most of that moisture. So the snowfall over the continent increases as the temperature gets warmer.
The same phenomenon may also have been observed in present-day Greenland. Satellite data compiled by Jay Zwally and his colleagues from NASA’s Goddard Space Flight Center seem to suggest that the southern two- fifths of the ice sheet that covers most of Greenland is thickening at a rate of about nine inches a year. Although the Goddard workers have no data for the northern three-fifths of the ice sheet, they note that it usually receives about half as much new snow as the southern part. If that’s true-- and if the satellite data are accurate, which some researchers doubt--the Greenland ice sheet could be lowering sea level as much as the Antarctic ice sheet is, around two-hundredths of an inch per year.
So what is the bottom line? What is sea level doing now, and what is it likely to do as the greenhouse effect warms the planet? Measuring sea level is tricky because it can only be measured in relation to the land, and the land itself moves--it slowly rebounds upward, for instance, when a glacier recedes and stops pressing it down. But long-term records from tide gauges suggest that sea level has actually risen over the past 50 years by about a tenth of an inch a year. How those measurements jibe with the data from Antarctica and Greenland, however, and what they portend for our future in the greenhouse, are far from clear.
Most researchers agree that the amount of carbon dioxide in the atmosphere has risen by about 25 percent since the world began to industrialize, and may well double within the next 50 years. There is also broad agreement that Earth’s average temperature has risen between half a degree and one and a quarter degrees Fahrenheit over the past century. Researchers are still squabbling, however, about whether that means greenhouse warming--which most now regard as inevitable--has already begun, and about how large a temperature rise we should expect. The estimates range from three to eight degrees.
As Earth warms, sea level will tend to rise for the simple reason that water expands as it heats up. In a greenhouse world, thermal expansion alone could raise sea level by as much as one and a half inches a decade. Warmer temperatures will also melt glaciers on mountains in the temperate and tropical latitudes--indeed, the process seems to have already begun. Geologist Lonnie Thompson of Ohio State University has documented the shrinking of glaciers in the Andes, on the Tibetan plateau, and in Kirghizia, in the former Soviet Union. The evidence is very clear that warming is taking place, Thompson told a Senate committee earlier this year. It is clear that tropical glaciers and ice caps are currently retreating . . . and the rate of retreat seems to be increasing. The Quelccaya ice cap in Peru, for instance, has pulled back 370 feet in just eight years.
But the great unanswered question is the extent to which these two sea-level-raising effects--thermal expansion and the melting of ice at low latitudes--will be balanced by the buildup of ice at high latitudes. The most popular forecast right now is that sea level will rise two to three feet during the next century. If Miller and De Vernal are right, however, that rise could be wiped out entirely by the growth of ice sheets in the north.
It is possible they are right about history--that Arctic ice sheets started expanding during a warm period 120,000 years ago--but wrong about the future. Most of the water that would get locked up in northern ice sheets would have to come from the warming of nearby seas. But as climatologist Stephen Schneider has pointed out, the regional effects of global warming can’t yet be forecast. So we can’t count on the subpolar seas warming in time to prevent a sea-level rise in the next century.
In the long run, probably within a millennium, the fluctuations in Earth’s orbit that control the pace of the ice age cycle will bring the present interglacial period to an end. As ice sheets creep once again over the continents, sea level will surely fall. What Miller and De Vernal are suggesting, in effect, is that global warming might preserve us from drowned coasts by hastening the next ice age along. With alternatives like that, it’s hard to know which outcome to root for.