Credit: NASA

Bull. Am. Meteorol. Soc. doi:10.1175/2009BAMS2607.1 (in the press)

Large uncertainties plague predictions of temperature change at a regional scale. Predicting climate change with greater certainty is of enormous academic and economic interest, as it can inform adaptation and investment decisions.

Ed Hawkins and Rowan Sutton of the UK National Centre for Atmospheric Science in Reading identified the main sources of uncertainty in predicting temperature change across the globe at scales of 2,000 kilometres — the first study to analyze this issue at the regional level. Using data from 15 global climate models and for three emissions scenarios, Hawkins and Sutton found two important sources of uncertainty: first, diversity between models in how they represent the climate, and second, natural variability inherent in the climate system. They also found uncertainty in the extent to which humans will continue to affect the climate, but this was less significant than other sources at small spatial scales and over one to two decades.

The authors say that much work is needed to reduce these sources of uncertainty and that investments made now could improve predictions in the coming decades by 10–20 per cent for the United Kingdom and Europe, and by up to 20 per cent for the rest of the world.