Comparing observed and modelled components of the Atlantic Meridional Overturning Circulation at 26°N

Abstract

Abstract. The Coupled Model Intercomparison Project (CMIP) allows assessment of the representation of the Atlantic Meridional Overturning Circulation (AMOC) in climate models. While CMIP Phase 6 models display a large spread in AMOC strength by a factor of three, the multi-model mean strength agrees reasonably well with observed estimates from RAPID1, but this does not hold for its various components. In CMIP6 the present-day AMOC is characterised by a lack of lower North Atlantic Deep Water (lNADW), due to the small-scale of Greenland-Iceland-Scotland Ridge overflow and too much mixing. This is compensated by increased recirculation in the subtropical gyre and more Antarctic Bottom Water (AABW). Deep-water circulation is dominated by a distinct deep western boundary current (DWBC) with minor interior recirculation compared to observations. The future decline in the AMOC to 2100 of 7 Sv under a SSP5-8.5 scenario is associated with decreased northward western boundary current transport in combination with reduced southward flow of upper North Atlantic Deep Water (uNADW). In CMIP6, wind stress curl decreases with time by 14 % so that the wind-driven thermocline recirculation in the subtropical gyre is reduced by 4 Sv (17 %) by 2100.  The reduction in western boundary current transport of 11 Sv is more than the decrease in the wind-driven gyre transport suggesting a decrease over time in the component of the Gulf Stream originating in the South Atlantic. 1 RAPID is used here as shorthand for the RAPID-Meridional Overturning Circulation and Heatflux Array-Western Boundary Time Series at 26°N (Moat et al., 2022).