Climatic Change http://doi.org/jp8 (2012)

Expert elicitation studies are a frequently used way of assessing scientific knowledge about future climate change. The approach is designed to incorporate experts' understanding of known flaws in computer models to provide a more comprehensive picture of uncertainty than model-driven methods can achieve. In practical terms the goal of expert elicitation is to determine experts' subjective probabilities for the values of key climate variables; making the implicit assumption that experts' knowledge can be captured in this way. However, decision theory has demonstrated this is not necessarily the case when information is ambiguous.

Using a choice experiment, Antony Millner from the Department of Agricultural and Resource Economics, University of California, Berkeley, US, and co-workers empirically determined whether experts' knowledge about climate sensitivity (the surface warming that results from a doubling of atmospheric CO2) can be captured by subjective probabilities. They found that a proportion of the participating climate scientists violated the criteria that must be satisfied for subjective probabilities to adequately describe their beliefs.

These results suggest that existing elicitation studies may qualitatively understate the extent of experts' uncertainty about climate change.