Carbon dioxide in the atmosphere warms Earth. But just how much warming you get depends on where you put your continents.
If adding carbon dioxide to the atmosphere creates a greenhouse effect that warms Earth, it must have happened in the past. That’s why paleoclimatology, once a small and esoteric field, is such a growth industry these days, with legions of geologists trying to glean past temperatures and CO2 levels from rocks, and legions of climate modelers trying to tell us what it all means--not only for the past but also for the future of Earth’s climate. On the whole, the results have been what you’d expect. When carbon dioxide levels were low, the climate was cold, and when they were high, the climate was warm, says climatologist Thomas Crowley of Texas A&M; University.
But lately two glaring exceptions to that simple rule have turned up. During the Ordovician Period, 440 million years ago, there seems to have been 16 times as much carbon dioxide in the atmosphere as there is today--and yet, judging from the gravelly deposits it left behind, there was also an ice sheet near the South Pole that was four-fifths the size of present-day Antarctica. The second exception is even more troubling. The Cretaceous Period, when dinosaurs ruled the Earth and CO2 levels were about eight times what they are today, has been one of the most popular case studies for global warming forecasters. And everyone knows what the climate was like during the dinosaurs’ heyday: steamy. Or was it? The latest evidence, reported just this past summer by British researchers, suggests that temperatures in the tropics 95 million years ago were no higher than they are now; and while it was a lot warmer at the poles than it is today, it was still freezing cold.
What happened to Earth’s greenhouse during these two periods? Climate modelers are beginning to believe the solution to both puzzles may be the same: geography. Carbon dioxide does tend to warm the planet--no one is questioning that--but the climate you actually end up with depends to a great degree on how you arrange your continents.
In the case of the Ordovician, Crowley thinks, the solution is fairly straightforward. He and his co-worker Steven Baum have spent the past couple of years recreating the Ordovician Earth in their computer and trying to understand how it could have supported glaciers. They got help from the sun: according to astrophysicists it was 4 percent dimmer 440 million years ago than it is today, which means that although the Ordovician greenhouse had 16 times as much heat-trapping CO2, it also had less heat to trap. But Crowley and Baum calculated that the net greenhouse effect was still equivalent to what you’d get by quadrupling CO2 levels today. In other words, an ice sheet should have stood little chance of surviving.
The survival of a permanent ice sheet depends less on whether it gets cold enough to snow in winter than on whether it gets warm enough in summer to melt all the previous winter’s snow. And the key to the Ordovician ice sheet, says Crowley, is the fact that most of the continents were joined together into one roughly circular landmass called Gondwanaland, whose southern edge was just over the South Pole 440 million years ago. In the interior of the supercontinent the climate was extreme- Midwestern: cold winters and hot summers and no permanent ice. But along the coast, the ocean--which warms far more slowly than the land because of its tremendous heat-storage capacity--put a damper on this seasonal cycle. The moderating effect of the water mutes the amount of summer warming you get, says Crowley.
People who live in places like Maine today know what muted summers are like; along the southern coast of Gondwanaland 440 million years ago, the summers would have been worse than muted. The wind blowing in from offshore was blowing over water that was right at the South Pole, and it was cold indeed. The global average temperature in Crowley and Baum’s simulated Ordovician was 64 degrees--14 degrees hotter than today. But along the southern margin of Gondwanaland, three feet of snow survived the summer each year, becoming part of thickening ice sheets.
Then around 430 million years ago the ice sheets disappeared. Crowley and Baum’s simulations point to an explanation that sounds paradoxical only at first: Gondwanaland was drifting southward at the time, and the farther south it went, the smaller the ice sheets became. As the center of the supercontinent came closer to the South Pole, the winters in the interior became colder--but the land still warmed up enough in summer to melt the snow. Meanwhile, what had been the southern, glaciated edge of Gondwanaland moved north into warmer waters. The glaciers soon vanished.
Thus the Ordovician mystery no longer seems so mysterious. The Cretaceous Period, though, is another story. Over the years, various analyses of the carbon locked in Cretaceous limestones--including the great chalk beds, formed from the corpses of countless plankton, that gave the period its name--have convinced researchers that the CO2 level in the Cretaceous atmosphere was eight times what it is today. Analyses of oxygen isotopes, meanwhile, have suggested that the Cretaceous climate was appropriately toasty--perhaps 20 degrees warmer than today. A hothouse climate would explain why paleontologists have found fossils indicating that tropical plants and reef corals survived at higher latitudes in the Cretaceous than they do today.
Yet the evidence for the hothouse, says geologist Bruce Sellwood of the University of Reading, has never been conclusive. As a snapshot of global climate history, the rocks that went into the isotopic analyses were far from perfect. For one thing, they differed in age by as much as 30 million years. For another, most of them came from the northern mid- latitudes, and almost none from the tropics. That just reflects the availability of chalk outcrops in Europe and North America, explains Sellwood.
He and his colleagues Greg Price and Paul Valdes set out to get a sharper picture of the Cretaceous climate. They limited their analysis to sedimentary rocks dating from a relatively small window of time around 95 million years ago. We still have 7 million years to play with, but it’s a hell of a lot better than the generalizations you got before, says Sellwood. The Reading workers also got some of the first solid information on tropical temperatures in the Cretaceous by analyzing 95-million-year-old sediments drilled recently from the ocean floor.
The results suggest that the Cretaceous was much cooler than previously thought. Temperatures at the poles were supposed to have averaged around 50 degrees, but Sellwood’s group claims they actually hovered around freezing. And the tropics, previously thought to have been 10 degrees warmer than today, were no warmer at all.
Why was the planet so cool, given all that CO2 in the atmosphere, and why was there so little difference between the poles and the tropics? Like Crowley and Baum, Valdes is looking for answers in computer models. And like them, he is finding that geography matters.
Earth 95 million years ago was a world of shallow seas. The landmasses no longer formed a supercontinent, but they remained close together, and because sea levels were high during the Cretaceous, the oceans flooded into the interiors of many continents. One of these shallow seas, for example, cut North America in half from Canada to Mexico. And what all of them did, according to Valdes, was carry humid air into the heart of continents, where it created heavy cloud cover that blocked sunlight and cooled the Earth. There are a lot of arguments and uncertainties about clouds, Valdes acknowledges--clouds can also trap heat rising from Earth’s surface--but in our model the clouds increase and create cooling.
Geography may also help explain why polar temperatures in the Cretaceous were so close to tropical ones. In today’s world, warm water is carried north in the Atlantic to just south of Greenland, where it cools, sinks to the bottom of the ocean, and flows back south. Valdes thinks that in the Cretaceous this conveyor belt circulation may have been shallower-- which would have made it faster at transporting heat out of the tropics. In the Cretaceous, the gap between America and Europe was very small, he says. There wasn’t much of an Atlantic. So maybe the water didn’t go down deep, it just recirculated and warmed up more and more, and so you didn’t have such a big temperature gradient. I’m waving my hands around, but it’s a strong hypothesis you can test--which is what Valdes hopes to do in the near future with a new model that incorporates realistic ocean currents.
The computer models that have been used to forecast global warming don’t incorporate realistic ocean currents, and their geography is pretty crude, too. Sellwood thinks that’s a reason to be more skeptical of their forecasts--although no less cautious about polluting the atmosphere. This research tells us that the link between carbon dioxide and global warming isn’t as secure as we all imagine, he says. But what I would hate to see happening is governments leaping on this and saying, ‘Ah, there is no link between carbon dioxide and climate, and therefore we can go on with what we’re doing now or even worse.’ This work shows the whole climate system is much more complex than we imagined. That’s actually a word of warning. To pump out huge amounts of carbon dioxide when we don’t know how the system works is a pretty dangerous thing.
