P2C2: Estimating Climate Sensitivity from Temperature Reconstructions of the Last Glacial Maximum

点击翻译按钮获取中文摘要

This is a project to use paleoclimate reconstructions of the Last Glacial Maximum (LGM: 21,000 years ago) and a variety of climate models to evaluate equilibrium climate sensitivity. Schmittner-Boesch published an influential paper in 2011 suggesting that the geological evidence implied an equilibrium climate sensitivity to a doubling of atmospheric CO2 of 1.4 to 2.8 C, lower than the IPCC4 likely range of 2 to 4.5 C and that the geological data ruled out very high sensitivities to doubling of CO2. However there were several issues left unresolved. For instance, several different paleoclimate reconstructions of the LGM produce different average reductions in global temperatures, which have a big effect of the sensitivity estimates. In addition, the range of error on the paleoclimate reconstructions is large and the effects of this error on the climate sensitivity need to be better quantified. Finally the initial study was performed with a "reduce-complexity" climate model. It was not clear the extent to which the result was affected by choice of model. This project will further evaluate these issues by expanding the paleoclimate reconstructions used in the study and incorporating the output from fully coupled climate model simulations of the LGM. Several new reconstructions of the LGM have been published and are now available for analysis. With access to the new CMIP5/PMIP3 models - with specific LGM boundary conditions/reconstructions - it is now possible to perform the analysis of equilibrium climate sensitivity in a more rigorous way by constraining the reduced-complexity model with output from the CMIP5/PMIP3 model results.This project will support a climate modeler from Oregon State University, who will collaborate with a statistician from the Woodrow Wilson School of Public and International Affairs at Princeton University. Mentoring and training of a postdoctoral researcher is included in the project.