Observed impact of Atlantic SST anomalies on the North Atlantic oscillation

The large-scale patterns of covariability between monthly sea surface temperature (SST) and 500-mb height anomalies (Z(500)) in the Atlantic sector are investigated as a function of time lag in the NCEP-NCAR reanalysis (1958-97). In agreement with previous studies, the dominant signal is the atmospheric forcing of SST anomalies, but statistically significant covariances are also found when SST leads Z(500) by several months. In winter, a Pan-Atlantic SST pattern precedes the North Atlantic oscillation (NAO) by up to 6 months. Such long lead time covariance is interpreted in the framework of the stochastic climate model, reflecting the forcing of the NAO by persistent Atlantic SST anomalies. A separate analysis of midlatitudes (20degrees-70degreesN) and tropical (20degreesS-20degreesN) SST anomalies reveals that the bulk of the NAO signal comes from the midlatitudes. A dipolar anomaly, with warm SST southeast of Newfoundland and cold SST to the northeast and southeast, precedes a positive phase of the NAO, and it should provide a prediction of up to 15% of its monthly variance several months in advance. Since the "forcing'' SST pattern projects significantly onto the tripole pattern generated by the NAO, these results indicate a positive feedback between the SST tripole and the NAO, with a strength of up to similar or equal to25 m K-1 at 500 mb or 2-3 mb K-1 at sea level. Additionally, a warming of the tropical Atlantic (20degreesS-20degreesN), roughly symmetric about the equator, induces a negative NAO phase in early winter. This tropical forcing of the NAO is nearly uncorrelated with and weaker than that resulting from the midlatitudes, and is associated with shorter lead times and reduced predictive skill.