Balmy forecast. Projected temperature differences between the 2020s and the 1990s.

Can't Beat the Heat

The world will continue to warm for decades no matter what people do to curb greenhouse gas emissions. This gloomy conclusion, published in the 18 April issue of Nature, was based on computer models developed by two research teams. Combined, the models make up the first thorough forecast of climate change in the next few decades.

During the last century, average global temperature increased by approximately 0.6 K. Although that may not sound like a heat wave, it's enough to melt glaciers and throw rainfall patterns out of whack. Most climate models indicate that the world is only going to get warmer. But many variables in these predictions, such as the cooling effect of tiny particles called aerosols or wildcards such as volcanic eruptions, still rely on educated guesses by experts.

The hot question is how to estimate the probable impact of these uncertainties. Reto Knutti and his colleagues at the University of Bern in Switzerland chose a relatively simple climate model and ran a total of 25,000 simulations of past climate changes, each based on a random combination of various factors known to contribute to climate change. Another team, led by Peter Stott of the Hadley Centre in Bracknell, U.K., and Jamie Kettleborough of the Rutherford Appleton Laboratory in Didcot, U.K., ran fewer simulations on a more comprehensive climate model. Both teams then chose the versions of their models that best predicted past climate changes and used them to predict future change for various levels of greenhouse gas emissions.

Despite their different approaches, the teams came up with similar results. Stott and Kettleborough predict that the global mean temperature during the 2020s will be 0.3 to 1.3 K higher than during the 1990s. Knutti's team expects a rise between 0.5 and 1.1 K. Both studies suggest that global warming is inevitable in the short term. "Until 2040 it doesn't make too much difference which scenario [for emission rates] you choose," Stott says. But decisions made now will affect climate in the second half of the century, the researchers say.

The two studies should be "very useful for policy-makers," says Francis Zwiers of the Meteorological Service of Canada. Most climate models focus on 100-year forecasts because the effects of different emission scenarios are clearer on that time scale. This makes sense scientifically, but it reduces the forecasts' political value, because policy plans seldom go further than a few decades. The new models bridge the gap. But Ulrich Cubasch of the Max Planck Institute of Meteorology in Hamburg says their "political strength is two-edged." Those who oppose curbs on greenhouse gas emissions, he points out, can argue that temperatures will rise anyway, whereas environmentalists can stress that the findings prove that acting now will pay off in the long run.

Related Sites:
Climate Modeling at University of Bern
Climate Models at Hadley Centre Intergovernmental Panel on Climate Change