Revisiting the Inconsistent Influence of El Niño–Southern Oscillation on the Equatorial Atlantic

Abstract The impact of El Niño–Southern Oscillation (ENSO) on the equatorial Atlantic zonal mode (AZM) is investigated using two atmospheric reanalysis products and 44 models from phase 6 of the Coupled Model Intercomparison Project (CMIP6). While the impact of ENSO on equatorial Atlantic SST is inconsistent, as previously reported, there is a very robust influence of decaying ENSO events in boreal spring on the contemporaneous surface zonal winds over the equatorial Atlantic, with El Niño events associated with anomalous easterlies that cool the equatorial Atlantic. This dynamic impact is counteracted by a thermodynamic impact, in which El Niño events cause changes in surface heat fluxes that lead to warming over the tropical Atlantic. The thermodynamic influence is active throughout the life cycle of the ENSO event, but the dynamic influence is highly seasonal with a pronounced peak in boreal spring. Thus, a quickly decaying El Niño event will lead to warming, while a slowly decaying event will lead to cooling in the equatorial Atlantic. The ENSO–AZM relation is further complicated by partially independent Atlantic variability, in particular that associated with the South Atlantic subtropical dipole (SASD). Prediction experiments with a simple linear regression model support the role of the SASD as an AZM precursor. Many CMIP6 models show skillful predictions of the June–August AZM based on SST indices in the preceding December–February, with correlations above 0.5. When predictions are restricted to ENSO years, even more models exceed this threshold, but skill in the reanalyses remains relatively low, possibly due to insufficient training data. Significance Statement El Niño is the strongest climate signal on interannual time scales. Its influence on the equatorial Atlantic, however, is surprisingly inconsistent. Here, we show, using an ensemble of climate models and observation-based data, that El Niño has a very robust influence on equatorial Atlantic surface winds, but only in boreal spring. This leads to a large difference in how El Niño events influence the equatorial Atlantic, based on how quickly they decay in boreal spring. We also find that variability in the South Atlantic, which is partly independent of El Niño, has a strong influence on the equatorial Atlantic, supporting the result of previous studies. These findings suggest that the predictability of equatorial Atlantic temperature anomalies may be higher than previously thought.