The effect of sea-ice extent in the North Atlantic on the stability of the thermohaline circulation in global warming experiments

PDF: http://data.dhkim.info/monograph/CD/M1XGN5AALN3P9LTF.pdf

Article  

  Climate Dynamics
Publisher: Springer-Verlag Heidelberg
ISSN: 0930-7575 (Paper) 1432-0894 (Online)
DOI: 10.1007/s00382-004-0414-0
Issue: Volume 22, Numbers 6-7

Date:  June 2004
Page: 689  
The effect of sea-ice extent in the North Atlantic on the stability of the thermohaline circulation in global warming experiments
O. A. Saenko1, 2, 2 , M. Eby1 and A. J. Weaver1

(1)  School of Earth and Ocean Sciences, University of Victoria, Victoria, Canada
(2)  Present address: Canadian Centre for Climate modelling and analysis, Meteorological Service of Canada, University of Victoria, Victoria, Canada

Received: 8 August 2003  Accepted: 27 January 2004  Published online: 30 April 2004

Abstract   Different climate models simulate different behavior of the Atlantic meridional overturning circulation (MOC) under the same global warming scenario. We propose a plausible explanation for this and argue that a proper simulation of the present-day climate in the subpolar North Atlantic is important. This is illustrated using results from idealized global warming experiments, in which both the radiative forcing scenario and the model employed are the same, with the only major difference being the initial subpolar North Atlantic climate. The initial conditions are made progressively colder, with more extensive sea-ice cover in the northern North Atlantic.The key result is that starting from conditions which are too cold in the North Atlantic and with sea-ice that is too extensive leads to an MOC that is more stable to the radiative forcing. Furthermore, under considerably underestimated sea surface temperatures in subpolar regions, the MOC can even intensify. A reduction of freshwater flux associated with the reduction of sea-ice melt is shown to be important for such unusual behavior of the MOC. Other mechanisms are also considered, but not deemed as important in explaining published inter-model differences.