Large sensitivity to freshwater forcing location in 8.2ka simulations

• Published in: Paleoceanography
• Main Authors: Morrill, Carrie, Ward, Ellen M., Wagner, Amy J., Otto-bliesner, Bette L., Rosenbloom, Nan
• Format: Article in Journal/Newspaper
• Language: English
• Published: Amer Geophysical Union 2014
• Subjects: Atlantic Meridional Overturning Circulation; abrupt climate change; deglaciation; 8; 2 ka event; freshwater forcing; coupled climate model simulation; Labrador Sea; north atlantic current; North Atlantic
• Online Access: https://archimer.ifremer.fr/doc/00496/60735/65213.pdf; https://doi.org/10.1002/2014PA002669; https://archimer.ifremer.fr/doc/00496/60735/

The 8.2ka event is a key test case for simulating the coupled climate response to changes in the Atlantic Meridional Overturning Circulation (AMOC). Recent advances in quantifying freshwater fluxes at 8.2ka from the proxy record have improved the realism of the forcing magnitude in model simulations, yet this forcing is still generally applied in an unrealistic geographic manner, across most of the Labrador Sea rather than just along the Labrador coast. Previous simulations with eddy- resolving ocean models have come to conflicting conclusions regarding the ability of such a coastally confined flow to impact the AMOC. These simulations have also not incorporated full atmosphere models nor have they used the new meltwater forcing values for 8.2ka. We use the Community Climate System Model, version 3, with an ocean model resolution only slightly coarser than that used in previous eddy-resolving simulations, to test the sensitivity to freshwater forcing location. When revised freshwater forcing is applied across the Labrador Sea, the AMOC is reduced by similar to 40% and climate anomalies compare well with proxy records of the 8.2ka event in terms of magnitude and duration. When the forcing is added just along the Labrador coast, however, most meltwater joins the subtropical gyre and travels to the subtropics with minor impact to the AMOC (similar to 10% decrease). It is likely that model biases in the placement of the North Atlantic Current remain an important limitation for correctly simulating the 8.2ka event.