Does The Time Series of the Atlantic Meridional Overturning Circulation (AMOC) at 26°N Enable Improved Sub-Annual Predictions of Atlantic Sea-Surface Temperatures?.
Supervisors: David Smeed, Rob Marsh, Joël Hirschi (NOCS, UK)
The Atlantic Meridional Overturning Circulation (AMOC) demonstrates significant predictive potential for monthly North Atlantic Sea-Surface Temperatures (SSTs). Extending the analysis of Duchez et al. (2016), who identified a strong SST dipole pattern centered at 26.5°N following AMOC anomalies at a 3 to 5-month lag, this study aims to evaluate the predictive skill of AMOC for SST anomalies. Additionally, it explores the potential of AMOC for predicting Sea-Surface Height (SSH), which reflects changes in heat content due to thermosteric sea level variations. We use detrended (trend removed) and deseasonalized (seasonal cycle removed) SST, SSH, and AMOC datasets from April 2004 to February 2022, divided into three-time schemes: the First Decade (April 2004 to March 2014), the Second Decade (April 2014 to February 2022), and All-Years (April 2004 to February 2022). Additionally, we assess the influence of the North Atlantic Oscillation (NAO) and the impact of the 2009-2010 AMOC drop for the AMOC predictive skill. Results for the AMOC and SSH correlations suggest that the AMOC may be a more reliable indicator of SSH variability than of SST variability. Given that satellite measurements of SSH via radar altimetry are directly influenced by ocean heat content variations due to thermal expansion, this underscores AMOC’s effectiveness as a predictor for SSH. The predictive regression of AMOC towards SST and SSH also support this finding, with the SSH has a higher prediction percentage. The analysis also confirms that the NAO has a complex impact on the relationship between AMOC and SST/SSH, influencing both short-term and longer-term dynamics. High NAO indices impact AMOC significantly within the first 5 months, while low and middle NAO indices affect longer lags. Consequently, SSH predictions are particularly robust during high NAO periods, with distinct patterns emerging based on NAO indices and time lags. In summary, with longer time series of RAPID array at 26.5oN from 2004 to 2022, we reveal that AMOC is a more effective predictor for SSH compared to SST, emphasizing the utility of SSH in reflecting AMOC’s influence on ocean heat content.

Keyword(s): AMOC’s predictive skill, RAPID array, North Atlantic SST and SSH, lagged correlation analysis