The influence of the AMOC strength at 26°N on seasonal SST predictability in the North Atlantic

The Atlantic Meridional Overturning Circulation (AMOC) plays a key role in the climate system by transporting heat northwards in the Atlantic. Together with air-sea heat fluxes, these processes are the main contributing factors for the North Atlantic sea surface temperature variability at interannual timescales. Here, we investigate the potential of using the knowledge of the state of the AMOC at 26◦N in the prediction of sea surface temperature anomalies (SSTA) at the seasonal time scale. We evaluate the mechanism proposed by Duchez et al. 2016, which relates anomalies in the AMOC strength, as measured in the RAPID array at 26◦N, with SSTA in the tropical and subtropical North Atlantic as a see-saw pattern. We analyse an ensemble with MPI-ESM-MR, nudged to reanalysis for the atmosphere and ocean component, and to observations for the sea-ice component. Using the AMOC anomalies as a precursor to SSTA, we find that the AMOC leads a SSTA dipole pattern in the tropical and subtropical North Atlantic with maximum correlations at 2-4 months; in agreement with Duchez et al., 2016. In extension to Duchez et al., we find that this pattern varies with the starting season, and is more pronounced when the most recent decade is considered than the entire time series back to 1979. In a set of ensemble hindcast simulations with 30 members, starting each May and November between 1982 and 2014, we also perform a hindcast skill analysis for both SST and AMOC. We find different spatial patterns in the SST hindcast skill at 2 – 4 months lead time for the hindcasts started in November and May, with higher skill over both the subpolar gyre and the eastern North Atlantic for hindcasts started in November over those started in May. We find an improvement of SST hindcast skill at 2 – 4 months lead time, when the strength of the AMOC at the start of the hindcast period is considered, which is in agreement with the see-saw mechanism.