the importance of computer models in improving our understanding of climate change and global warming has remained a controversial issue. By feeding data on temperature, winds, solar radiation, ocean currents and a host of other climatic factors into a model, scientists can simulate what could happen over time if emissions of heat-trapping greenhouse gases increase.
These simulations are the basis of predictions that the average surface temperature of the earth will rise substantially and disrupt the world's climate in coming decades unless emissions of greenhouse gases are reduced.
The forecast of various models has been unanimous and has been widely accepted as gospel by some environmentalists and scientists. But critics -- especially the opponents of emissions reductions -- have dismissed the forecast as useless. Richard S Lindzen of the Massachusetts Institute of Technology, has always considered the models worthless as prognosticators. "I'm not saying the model output bears no resemblance at all to nature; in the gross figures, it looks plausibly similar," he says. But that, he adds, "does not give you forecast ability."
But most experts take a middle stance, saying that while the models are getting more realistic all the time, they remain relatively crude versions of the real state of the Earth's environment. There is a general agreement that any finding should be interpreted with greater care.
If there is disagreement on the effectiveness of the models, how seriously should scientists take these predictions? One hint emerges from a computer model being operated by Makiko Sato, physicist and climate researcher in New York, and James E Hansen, director of the us National Aeronautics and Space Administration's ( nasa 's) Goddard Institute for Space Studies. Sato's computer monitor has world maps in vivid colours ranging from yellow to orange to dark red (for degrees of warmth) through shades of blue (for degrees of cold). One map shows the pattern of global temperature changes the computer climate model "predicted" for the period 1979-95. Maps on the left show the changes that actually took place.
The picture presented by the comparison makes it clear that the computer got it right in broadest terms. The world as a whole warmed, and the Northern Hemisphere warmed significantly. But the model also got a lot of regional details wrong. The computers predicted that eastern Canada should have warmed substantially, but, instead, it cooled a lot.
Not a lot of stock can be placed in a single "run" of a single model, or even in many runs. But many runs of many models present the same picture suggested by the maps on Sato's screen: that the models show a general global fidelity to reality but also pervasive regional error. Significantly, the forecast of the most recent and advanced computer models matches that of the earlier, much cruder models that came into use nearly two decades ago: that a doubling of carbon dioxide concentrations in the atmosphere would raise the Earth's average surface temperature by one to four degrees c . The wide range is a measure of the differences among model projections, and thus of the uncertainty inherent in climate modelling. A doubling of carbon dioxide concentrations is expected late in the next century if the current pace of industrial development continues.