Abstract/Summary

Many previous studies have demonstrated a high uncertainty in the relationship between the El Niño-Southern Oscillation (ENSO) and North Atlantic Oscillation (NAO). In the present work, decadal modulation by the Atlantic Multidecadal Oscillation (AMO) is investigated as a possible cause of the nonstationary ENSO-NAO relationship based on observed and reanalysis data. It is found that the negative ENSO-NAO correlation in late winter is significant only when ENSO and the AMO are in-phase (AMO+/El Niño and AMO-/La Niña). However, no significant ENSO-driven atmospheric anomalies can be observed over the North Atlantic when ENSO and the AMO are out-of-phase (AMO-/El Niño and AMO+/La Niña). Further analysis indicates that the sea surface temperature anomaly (SSTA) in the tropical North Atlantic (TNA) plays an essential role in this modulating effect. Due to broadly analogous TNA SSTA responses to both ENSO and the AMO during late winter, a warm SSTA in the TNA is evident when El Niño occurs during a positive AMO phase, resulting in a significantly weakened NAO, and vice versa when La Niña occur during a negative AMO phase. In contrast, neither the TNA SSTA nor the NAO show a prominent change under out-of-phase combinations of ENSO and AMO. The AMO modulation and associated effect of the TNA SSTA are shown to be well reproduced by historical simulations of the HadCM3 coupled model and further verified by forced experiments using an atmospheric circulation model. These offer hope that similar models will be able to make predictions for the NAO when appropriately initialized.