Knowledge about our future climate

Whilst there is evidence that our climate is changing as a result of human activities, forecasting future changes remains a significant challenge. Despite the use of complex computer models, there is a high degree of uncertainty in forecasts for the next 10 to 100 years. This uncertainty tends to be greater for small spatial scales but government planners, the private sector and non-governmental organisations need detailed regional or local information. Providing this information is perhaps the biggest challenge to climate science today.

To predict future climate change, computer based models are used to summarise our understanding of the way the climate behaves. These range from simple 'energy balance' models to complex, three-dimensional global circulation models that calculate the changing weather at over 100,000 points around the earth, every 15 to 30 minutes. The results from these models are often compared with observations of climate change over the last 50 years. However, even if a model can simulate the past reasonably well, this does not mean that it can correctly forecast the future.

There are three major sources of uncertainty in climate forecasting. The first source is natural variability, which is studied using identical models with slightly different starting conditions. This accounts for the chaotic 'butterfly effect'. A consequence of this is that a system is sensitive to the precise details of its starting conditions. Second is model uncertainty - do different models give different forecasts? This can be addressed by slightly changing the way the model represents the physics of the climate. Third is uncertainty in the levels of future greenhouse gas concentrations. This depends on natural factors (such as volcanoes) and on anthropogenic factors (such as how society responds to climate change).

In 2001, experts from the Intergovernmental Panel on Climate Change (IPCC) reported on the state of climate science at the time. They predicted a rise in average global temperatures of between 1.4 and 5.8°C by 2100, and a sea level rise of between 0.09 and 0.88 metres. Alongside these, they anticipated changes in extreme weather and climate events including:

• higher maximum and minimum temperatures and more hot days over land areas (very likely)
• more intense rainfall events (very likely, over many areas)
• increase in tropical cyclone peak wind and rainfall intensities (likely, over some areas).

More recently, the climateprediction.net project has developed a method for creating very large numbers of different models. This enables a more comprehensive exploration of possible future climate changes and therefore analysis of the uncertainty in climate forecasts. There are currently over 70,000 participants from over 130 countries. Initial results support the conclusions of other recent studies in suggesting that there is a possibility of a much greater increase in global temperature than considered by the IPCC. Future collaborations with developing country scientists could help develop the project further, with the possibility of providing climate forecasts related to probability for specific countries or sectors e.g. crop yields, flooding risks.

The problem of assessing uncertainty in how the climate will respond to current and future greenhouse gas emissions is not yet solved. However, experiments such as climateprediction.net and developments in climate models will improve our understanding of the sources of uncertainty. The need for mitigation is clear but for the present, planners and policy-makers need to adopt flexible responses to the adaptation problem. Useful, regional climate forecasts related to probability are nevertheless on the horizon and a real possibility during the next decade or so.

Dave Stainforth
University of Oxford
Clarendon Laboratory
Parks Road
Oxford, OX1 3PU
UK
d.stainforth1@physics.ox.ac.uk

See also

www.climateprediction.net

'Climate change 2001: the scientific basis', contribution of Working Group 1 to the Third Assessment Report of the IPCC, Cambridge University Press, IPCC, 2001
www.ipcc.ch/pub/spm22-01.pdf

Views expressed on these pages are not necessarily those of DFID, IDS, id21 or other contributing institutions. Copyright remains with the original authors but (unless stated otherwise) articles may be copied or quoted without restriction, provided id21 and originating author(s) and institution(s) are acknowledged. Copyright © 2005 id21. All rights reserved.