British experts have criticized the focus of current climate projections. They say that scientists should shift from models that predict what will happen many decades from now, and concentrate instead on shorter-term forecasts that will aid policy-makers, businesses and the public.

Climate models such as those used in the Intergovernmental Panel on Climate Change (IPCC) reports have been instrumental in convincing the world that climate change threatens ecosystems and human societies, but they do not provide much practical guidance. “We may not be providing what we possibly could,” says Peter Cox, a climate modeller at the University of Exeter, UK, and former chair of climate-system dynamics at the UK Met Office.

Cox and his colleague David Stephenson, also at the University of Exeter, published their argument last week (P. Cox and D. Stephenson Science 317, 207–208; 2007). “The IPCC has nailed many old questions,” says Cox. “It's a done deal, so we had better move on.”

A key question is how to make climate-change models socially relevant. Cox and Stephenson propose having climate forecasters shift their attention to around 2050, rather than trying to predict farther into the future. This would effectively mean that the timescale of climate predictions would match that over which long-term policy and business planning is carried out.

The authors note that climate models are least uncertain for between 30 and 50 years from now. Shorter-term predictions will be less accurate because of uncertainty over initial conditions. Changes that will transpire over the next three decades are essentially “already in the system”, says Cox. For predictions more than 50 years in the future, uncertainty levels in the models increase because no one can accurately forecast the level of carbon dioxide emissions resulting from human activity.

Out of water: practical solutions are needed to the problems of climate change. Credit: J. WOOD/GETTY

Between 30 and 50 years away is thus a sort of sweet spot in which to target policy planning, Cox says. Mitigation policies and plans for associated socioeconomic factors, such as economic growth, energy use and technology needs, could be developed with that time frame in mind.

Some of those involved in the IPCC process do not disagree in principle, but say the inherent uncertainty of climate models will always make forecasting difficult. “Focusing on what 'should be' is a worthy goal going forward, but not a panacea for the uncertainty problem,” says Cynthia Rosenzweig, a climate modeller at NASA's Goddard Institute for Space Studies in New York, and a coordinating lead author of the most recent IPCC review, published this year.

Another report released last week also criticizes the “big gap” between how climate scenarios are currently used and their “potential contributions” to policy making. The report was co-authored by Rosenzweig, and is the second in a series by the US Climate Change Science Program.

The current generation of climate scenarios is still useful for resource managers to guide their preparation for climate change over the next few decades. Planning for such change is becoming “mainstream” in water-management systems, coasts and human health care, says Rosenzweig.