Some parts of the world could actually benefit from climate change, while others could suffer tremendously. But for the foreseeable future the effects will he uncertain. No nation can plan on benefiting, and so, says Schneider, we must all "hedge our global bets," by reducing emissions of greenhouse gases. "The longer we wait to take action," he says, "and the weaker the action, the larger the effect and the more likely that it will be negative." Says meteorologist Howard Ferguson, assistant deputy minister of the Canadian Atmospheric Environment Service, "All the greenhouse scenarios are consistent. These numbers are real. We have to start behaving as if this is going to happen. Those who advocate a program consisting only of additional research are missing the boat."
While the greenhouse effect threatens to make life on Earth miserable, it is also part of the reason life is livable in the first place. For at least the last 100,000 years atmospheric carbon dioxide, naturally generated and consumed by animals and plants, was in rough equilibrium, at a couple of hundred parts per million. Without this minute but critical trace to hold in heat, the globe's mean temperature would be in the forties instead of a comfortable 59 degrees. The amount of carbon dioxide has risen and fallen a bit, coinciding with the spread and retreat of glaciers as ice ages have come and gone. But until the Industrial Revolution, atmospheric carbon dioxide levels never rose above a manageable 280 parts per million.
Then, beginning early in the nineteenth century, the burning of fossil fuels, especially coal, took off. By 1900, carbon dioxide levels in the atmosphere had begun to rise steadily, reaching 340 parts per million last year.
Levels of the other greenhouse gases have also risen. Methane, for example, is generated primarily by bacterial decomposition of organic matter—particularly in such places as landfills, flooded rice paddies, and the guts of cattle and termites— and by the burning of wood. Methane concentration in the atmosphere has grown steadily as Earth's human population has grown, rising one percent a year over the last decade. Levels of chlorofluorocarbons, which are used as refrigerants, as cleaning solvents, and as raw materials for making plastic foam, have climbed 5 percent annually.
The amount of nitrous oxide in the atmosphere has quickly increased as well, with about a third of the total added by human activity— much of that emitted by nitrogen-based fertilizers, and half of that from just three nations: China, the Soviet Union, and the United States. This gas is also released by the burning of coal and other fossil fuels, including gasoline. And ozone, which forms a beneficial shield against ultraviolet radiation when high in the stratosphere, is an efficient greenhouse gas when it appears at airliner altitudes— as it increasingly does, since it too is a by-product of fossil fuel burning.
All these gases are far more efficient at absorbing infrared energy (the invisible radiation that ordinarily carries Earth's excess heat into space) than is carbon dioxide. Indeed, atmospheric chemists have estimated that the combined warming effect of these trace gases will soon equal or exceed the effect from carbon dioxide. And even as growth has slowed in the industrialized nations, the Third World is rushing full tilt into development. All told, billions of tons of greenhouse gases enter the atmosphere each year.
The big question is, given the inexorable buildup of these gases—a growth that even the most spirited optimists concede can only be slowed, not stopped— what will the specific effects be? It's hard to say, because the relationship between worldwide climate and local weather is such a complex phenomenon to begin with. The chaotic patterns of jet streams and vortices and ocean currents swirling it around the globe and governing the weather still confound meteorologists; in fact, weather more than two weeks in the future is thought by some to he inherently unpredictable.
So far, the best answers have come from computer models that simulate the workings of the atmosphere. Most divide the atmosphere into hundreds of boxes, each of which is represented by mathematical equations for wind, temperature, moisture, incoming radiation, outgoing radiation, and the like. Each mathematical box is linked to its neighbors, so it can respond to changing conditions with appropriate changes of its own. Thus, the model behaves the way the world does—albeit at a very rough scale. A typical model divides the atmosphere vertically into nine layers and horizontally into boxes that are several hundred miles on a side.
Climate modelers can play with "what if" scenarios to see how the world would respond to an arbitrary set of conditions. Several years ago, for example, computer models were used to holster the theory of nuclear winter, which concluded that smoke and dust lofted into the atmosphere in a nuclear war would block sunlight and dangerously chili the planet. To study the greenhouse effect, climatologists first used models to simulate current conditions, then instantly doubled the amount of carbon dioxide in the atmosphere. The computer was allowed to run until conditions stabilized at a new equilibrium, and a map could be drawn showing changes in temperature, precipitation, and other factors.
But Hansen's latest simulations—the ones he used in his startling congressional testimony—are more sophisticated. In them he added carbon dioxide to the atmosphere stepwise, just as is happening in the real world. The simulations, begun in 1983, took so much computer time that they were not completed and published until this summer.
Even the best climate model, however, has to oversimplify the enormous complexity of the real atmosphere. One problem is the size of the boxes. The model used at the National Center for Atmospheric Research, for example, typically uses boxes 4.5 degrees of latitude by 7 degrees of longitude— about the size of the center's home state of Colorado— and treats them as uniform masses of air. While that's inherently inaccurate—the real Colorado contains such fundamentally different features as the Rocky Mountains and the Great Plains— using smaller boxes would take too much computing power.
Another problem is that modelers must estimate the influence of vegetation, ice and snow, soil moisture, terrain, and especially clouds, which reflect lots of sunlight back into space and also hold in surface heat. "Clouds are an important factor about which little is known," says Schneider. "When I first started looking at this in 1972, we didn't know much about the feedback from clouds. We don't know any more now than we did then."
