Labrador Sea sub-surface density as a precursor of multi-decadal variability in the North Atlantic: a multi-model study

• Main Authors: Ortega, Pablo, Robson, Jon I., Menary, Matthew, Sutton, Rowan T., Blaker, Adam, Germe, Agathe, Hirschi, Jöel J.-M., Sinha, Bablu, Hermanson, Leon, Yeager, Stephen
• Format: Text
• Language: English
• Published: 2020
• Subjects: Labrador Sea; North Atlantic
• Online Access: https://doi.org/10.5194/esd-2020-83; https://esd.copernicus.org/preprints/esd-2020-83/

The Subpolar North Atlantic (SPNA) is a region with prominent decadal variability that has experienced remarkable warming and cooling trends in the last few decades. These observed trends have been preceded by slow-paced increases and decreases in the Labrador Sea density (LSD), which are thought to be a precursor of large scale ocean circulation changes. This article analyses the inter-relationships between the LSD and the wider North Atlantic across an ensemble of coupled climate model simulations. In particular, it analyses the link between subsurface density and the deep boundary density, the Atlantic Meridional Overturning Circulation (AMOC), the Subpolar Gyre (SPG) circulation, and the upper ocean temperature in the eastern SPNA. All simulations exhibit considerable multidecadal variability in the LSD and the ocean circulation indices, which are found to be interrelated. LSD is strongly linked with the strength of subpolar AMOC and gyre circulation, and is also linked with the subtropical AMOC, although the strength of this relationship is model dependent. The connectivity of LSD with the subtropics is found to be sensitive to different model features, including: the mean density stratification in the Labrador Sea; the strength and depth of the AMOC; and the depth at which the LSD propagates southward along the western boundary. Several of these quantities can also be computed from observations, and comparison with these observation-based quantities suggests that models representing a weaker link with the subtropical AMOC may be more realistic. This would imply that RAPID AMOC measurements might not be adequate to represent decadal to multidecadal changes in the subpolar overturning circulation.