C53D-06
The Impact of Multidecadal NAO Variations on Atlantic Ocean Heat Transport and Rapid Changes in Arctic Sea Ice

Friday, 18 December 2015: 14:55
3007 (Moscone West)
Thomas L Delworth, Geophysical Fluid Dynamics Laboratory, Princeton, NJ, United States
Abstract:
The Arctic and North Atlantic have experienced pronounced changes over the 20th and early 21st centuries, including a rapid loss of Arctic sea ice over the last several decades and prominent multidecadal variability in both ocean temperatures and sea ice. Here we use suites of climate model simulations to probe some of the factors responsible for the multidecadal variability in the Atlantic/Arctic system. We show that multidecadal fluctuations of the North Atlantic Oscillation (NAO) induce multidecadal fluctuations of the Atlantic Meridional Overturning Circulation (AMOC). A positive phase of the NAO is associated with strengthened westerly winds over the North Atlantic. These winds extract more heat than normal from the subpolar ocean, thereby increasing upper ocean density, deepwater formation, and the strength of the AMOC and associated poleward ocean heat transport. In model simulations the observed negative phase of the NAO in the 1960s and 1970s led to a weaker than normal AMOC, reduced poleward ocean heat transport, a cold North Atlantic, and an increase in Arctic sea ice extent in both winter and summer. The NAO strengthened from the 1970s to the mid 1990s, leading to an increase of the AMOC and a warming of the North Atlantic. The increased heat transport extended throughout the North Atlantic, into the Barents Sea, and finally into the Arctic, contributing to a rapid reduction of sea ice in the 1990s through the 2000s. Feedbacks involving shortwave radiation are an important component of the overall changes. In these model simulations as much as 1/3 of the recent reduction of Arctic sea ice is associated with the NAO-induced AMOC and heat transport increase. Since the mid 1990s the NAO has changed from a strong positive phase to a more neutral phase. In our model simulations this weakens the AMOC and poleward ocean heat transport, and diminishes the contribution of ocean heat transport to the reduction of Arctic sea ice extent. Considered in isolation, the reduction in ocean heat transport implies a possible moderation in the rate of Arctic sea ice loss in the coming decade.