08/24/2009
Exploring the Role of Climate Model Quality in Detection and Attribution Studies
Summary
Observed climate change represents a complex mixture of internally generated noise and responses to external forcing. “Fingerprint” studies, which seek to identify the causes of recent climate change, involve rigorous statistical comparisons of modeled and observed climate change patterns. DOE sponsored scientists led by PI Santer in 2007 used a suite of 22 Coupled Climate Models in conjunction with satellite observations to indicate unambiguously that changes in atmospheric water vapor have a human “fingerprint.” Their 2007 study adopted a democratic “one model, one vote” approach in which each of the 22 models received equal weight in the analysis despite large differences in the ability of the models to simulate important features of present-day climate. The group calculated a total of 70 different metrics of model performance, repeating their original fingerprint analysis with various sets of “top ten” and “bottom ten” models. They find that restricting the fingerprint analysis to “better” models does not affect the ability to identify a human-caused fingerprint in satellite records of water vapor changes. This work links and highlights DOE’s expertise in both climate model evaluation and climate change detection and attribution.
References
Santer, B. D., K. E. Taylor, P. J. Gleckler, C. Bonfils, T. P. Barnett, D. W. Pierce, T. M. L. Wigley, C. Mears, F. J. Wentz, W. Brüggemann, N. P. Gillett, S. A. Klein, S. Solomon, P. A. Stott, and M. F. Wehner. 2009. “Incorporating Model Quality Information in Climate Change Detection and Attribution Studies,” Proceedings of the National Academy of Sciences 106(35), 14778–14783. DOI: 10.1073/pnas.0901736106.