Suppressed multi-centennial climate variability in EC-Earth3 high CO2 simulations

A distinct multi-centennial climate variability signal is apparent in the EC-Earth3 model 2000-year pre-industrial control simulation. This variability arises primarily in the North Atlantic basin and appears to be closely associated with Atlantic Meridional Overturning Circulation (AMOC). It is mainly modulated by the ocean heat transport and freshwater exchange between the Arctic Ocean and the North Atlantic. When a stronger AMOC occurs, it is coherent with anomalous anticyclonic surface currents in the Arctic and cyclonic surface currents in Greenland, Iceland, Norwegian Seas and Labrador Seas. The increased heat in the subpolar gyre region strengthens the oceanic surface evaporation, resulting in a saltier deep convection region and hence strengthens the deep-water formation. Meanwhile, stronger AMOC transports more ocean heat into the Arctic and melts the sea-ice, causing more freshwater to enter the Arctic. The AMOC strength and freshwater accumulation in the Arctic both reach their peaks in about 50 years. Then, in the following 50 years, the freshwater in Arctic slowly pours into the Greenland-Iceland-Labrador Seas, weakens the subpolar gyre, inhibits deep-water formation and eventually weakens the AMOC. Finally, the oscillation shifts to the opposite phase. These physical processes sustain a 160-200 year variability of AMOC, which is considered as the main driver of the multi-centennial climate variability signal in our simulation.

 

In high CO2 forced climates, here simulated with climate sensitivity experiments with alternate CO2 levels of 400 and 560 ppm respectively, the multi-centennial variability of AMOC is present but has a suppressed amplitude. AMOC variability under 400 ppm CO2 forcing shows a similar frequency band as that in the pre-industrial simulation, with enhanced Arctic-Atlantic salinity anomaly exchange. Under 560 ppm CO2 forcing, the AMOC variability shows a lower frequency band. Here, alongside the Arctic-Atlantic salinity exchange, there are also salinity anomalies propagating from the south Atlantic to the north Atlantic. This leads to a longer maintaining of meridional inter-basin exchanges in the entire Atlantic and Arctic. The decrease of Arctic sea-ice under stronger radiative forcing will cause more freshwater to enter the North Atlantic, slow down the deep-water flow, and thus suppress the AMOC strength. Meanwhile the mechanism that sustains AMOC variability, which was inferred from the pre-industrial simulation, will also change as less sea ice in the North Atlantic and Arctic lead to a more well mixed Arctic-Atlantic salinity anomaly exchange.

These experiments indicate that the dynamics of the meridional inter-basin exchange in the north Atlantic and its influence on the salinity are essential components to the centennial climate variability and should be considered when assessing future North Atlantic climate.