How warm was the last interglacial? New model–data comparisons

• Published in: Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences
• Main Authors: Otto-Bliesner, Bette L., Rosenbloom, Nan, Stone, Emma J., McKay, Nicholas P., Lunt, Daniel J., Brady, Esther C., Overpeck, Jonathan T.
• Format: Article in Journal/Newspaper
• Language: English
• Published: The Royal Society 2013
• Subjects: General Physics and Astronomy; General Engineering; General Mathematics; Antarc*; Antarctic; Antarctica; Arctic; Ice Sheet; North Atlantic; Sea ice; Southern Ocean
• Online Access: http://dx.doi.org/10.1098/rsta.2013.0097; https://royalsocietypublishing.org/doi/pdf/10.1098/rsta.2013.0097; https://royalsocietypublishing.org/doi/full-xml/10.1098/rsta.2013.0097

A Community Climate System Model, Version 3 (CCSM3) simulation for 125 ka during the Last Interglacial (LIG) is compared to two recent proxy reconstructions to evaluate surface temperature changes from modern times. The dominant forcing change from modern, the orbital forcing, modified the incoming solar insolation at the top of the atmosphere, resulting in large positive anomalies in boreal summer. Greenhouse gas concentrations are similar to those of the pre-industrial (PI) Holocene. CCSM3 simulates an enhanced seasonal cycle over the Northern Hemisphere continents with warming most developed during boreal summer. In addition, year-round warming over the North Atlantic is associated with a seasonal memory of sea ice retreat in CCSM3, which extends the effects of positive summer insolation anomalies on the high-latitude oceans to winter months. The simulated Arctic terrestrial annual warming, though, is much less than the observational evidence, suggesting either missing feedbacks in the simulation and/or interpretation of the proxies. Over Antarctica, CCSM3 cannot reproduce the large LIG warming recorded by the Antarctic ice cores, even with simulations designed to consider observed evidence of early LIG warmth in Southern Ocean and Antarctica records and the possible disintegration of the West Antarctic Ice Sheet. Comparisons with a HadCM3 simulation indicate that sea ice is important for understanding model polar responses. Overall, the models simulate little global annual surface temperature change, while the proxy reconstructions suggest a global annual warming at LIG (as compared to the PI Holocene) of approximately 1 ° C, though with possible spatial sampling biases. The CCSM3 SRES B1 (low scenario) future projections suggest high-latitude warmth similar to that reconstructed for the LIG may be exceeded before the end of this century.