1Earth, Atmospheric, and Planetary Sciences, Purdue University, West Lafayette, Indiana, USA
2Department of Meteorology (MISU) and Bert Bolin Center for Climate Research, Stockholm University, Stockholm, Sweden
3Purdue Climate Change Research Center, Purdue University, West Lafayette, Indiana, USA
Abstract. In this study, we compare the simulated climatic impact of adding an Antarctic ice sheet (AIS) to the "greenhouse world" of the Eocene and removing the AIS from the modern world. The modern global mean surface temperature anomaly (ΔT) induced by Antarctic Glaciation depends on the background CO2 levels and ranges from −1.22 to −0.18 K. The Eocene ΔT is nearly constant at ~−0.25 K. We calculate an climate sensitivity parameter S[Antarctica] which we define as ΔT divided by the change in effective radiative forcing (ΔQAntarctica) which includes some fast feedbacks imposed by prescribing the glacial properties of Antarctica.
The main difference between the modern and Eocene responses is that a negative cloud feedback warms much of the Earth's surface as a large AIS is introduced in the Eocene, whereas this cloud feedback is weakly positive and acts in combination with positive sea-ice feedbacks to enhance cooling introduced by adding an ice sheet in the modern. Because of the importance of cloud feedbacks in determining the final temperature sensitivity of the AIS, our results are likely to be model dependent. Nevertheless, these model results suggest that the effective radiative forcing and feedbacks induced by the AIS did not significantly decrease global mean surface temperature across the Eocene–Oligocene transition (EOT −34.1 to 33.6 Ma) and that other factors like declining atmospheric CO2 are more important for cooling across the EOT. The results illustrate that the efficacy of AIS forcing in the Eocene is not necessarily close to one and is likely to be model and state dependent. This implies that using EOT paleoclimate proxy data by itself to estimate climate sensitivity for future climate prediction requires climate models and consequently these estimates will have large uncertainty, largely due to uncertainties in modelling low clouds.