Since 1991, the UN Intergovernmental Panel on Climate Change (IPCC) report has predicted a 1.5-4.5°C increase in average temperatures due to a doubling of CO2 levels. Now, new results from distributive computing projects suggest that the temperature increase could go as high as 11°C. Although the IPCC has been reluctant to embrace these estimates, some scientists say that new developments make the wider temperature range nearly impossible for the panel to ignore.
“Our experiment shows that increased levels of greenhouse gases could have a much greater impact on climate than previously thought,” says David Stainforth, a climate scientist at Oxford University. He and his colleagues at Climateprediction.net (www.climateprediction.net), a collaborative effort of British universities and government agencies, will be discussing their results at the European Geosciences Union meeting in April. The researchers enlisted 95,000 people from 150 countries to download a general circulation model (GCM) and run it using the idle processing capacity on their personal computers. The distributed computing project enabled the researchers to run thousands of versions of the model, in which 21 parameters were set to alternative values considered plausible by experts (Nature 2005, 433, 403-406).
Stainforth and his associates analyzed more than 2000 of the simulations that perturbed 6 parameters, such as the threshold of relative humidity, the cloud-to-rain conversion rate, and the ice fall speed. When CO2 concentration doubles from pre-industrial levels-as is expected to happen by about 2050-the simulations predict that global mean temperature could rise over a period of many years anywhere from 1.9 to 11.5°C. “An important message is that we didn’t get any measures of climate sensitivity less than 1.9°C,” Stainforth says. This means that policy makers can be increasingly confident that global temperature will rise by no less than 2°C in response to doubling CO2 levels. But it is also possible that temperatures could rise by as much as 11°C, Stainforth cautions.
The Climateprediction.net experiment is the first time that a GCM has produced climate responses substantially greater than 5°C, says Michael Schlesinger, a climate scientist at the University of Illinois at Urbana-Champaign. “When the IPCC set the climate sensitivity range at 1.5-4.5°C in 1990, it had a strong constraint on what science could do,” says Schlesinger, who helped write the first IPCC report that set the range. “If climate modelers, myself included, got a value higher than 4.5, they changed the model to make it otherwise,” he says. That action, Schlesinger says, was driven in part because climatologists did not want to provide fuel for climate change skeptics to argue that large ranges of uncertainty in model outcomes undermine any rationale for taking action on greenhouse gas emissions.
Nevertheless, during the past five years, a handful of researchers have used simplified climate models to generate statistical estimates of global temperature response to doubled CO2 that range from 1 to 9.3°C, he says. IPCC scientists have been reluctant to acknowledge these estimates because they did not come from the more complex GCMs. “Now, because of Climateprediction.net, the IPCC would get torn to shreds if they tried to say that global mean temperatures couldn’t rise above 4.5°C,” Schlesinger asserts.
Stainforth agrees: “Our results demonstrate the wide range of behavior possible within a GCM and show that high sensitivities cannot yet be neglected as they were in the headline uncertainty ranges of the IPCC Third Assessment Report.” The range of possibilities for future climate evolution needs to be taken into account when planning climate change mitigation and adaptation strategies and the Climateprediction.net results will be an important input to the current IPCC assessment process, he adds.
“We don’t design models to have a narrow range of climate sensitivity, and my colleagues have never tuned their models in response to critics or politics,” says Bill Collins, an atmospheric scientist at the National Center for Atmospheric Research in Boulder, Colo., and a member of the group preparing the fourth scientific assessment for the IPCC in 2007. The accepted range of climate sensitivity has remained at 1.5-4.5°C because it reflects the diversity of results obtained by the modeling community, he says. The next IPCC report will quote the Climateprediction.net results to characterize the range of uncertainty produced by GCMs, he says. However, because the Climateprediction.net model is not as complex as other GCMs, it will not join the 15 models used to evaluate parameters such as precipitation over the North Atlantic, Collins says.
Stainforth says his model versions are “as realistic as other state-of-the-art climate models.” Although his model is not scaled to predict regional impacts, he emphasizes that “models with such extreme sensitivities are critical for the study of the full range of possible responses of the climate system to rising greenhouse gas levels, and for assessing the risks associated with specific targets for stabilizing these levels.”