Future Changes in Extreme El Nino Events Modulated by North Tropical Atlantic Variability

The extraordinarily strong El Nino events, such as those of 1982/1983, 1997/1998, and 2015/2016, are known to cause critical socio-economic impacts worldwide by disrupting global weather patterns, cyclones, drought/floods, and ecosystems. Therefore, it is a critical question how often such extreme El Nino will take place in the future. Although many climate models tend to simulate more frequency extreme El Nino under the greenhouse warming, there is a large intermodel diversity, with a range as large as the multimodel averaged change. The cause for this intermodel uncertainty is not known. Here we show that the north tropical Atlantic (NTA) mean precipitation plays a significant role in controlling changes in the extreme El Nino frequency under global warming; that is, much frequent extreme El Nino events under the greenhouse warming are simulated in the models whose climatological precipitation over the NTA is largely decreased. Relatively drier climatology over the NTA can induce a wetter climatology over the equatorial eastern Pacific through an atmospheric teleconnection, and it provides background condition that the El Nino-related convective responses are amplified to push El Nino to the extremes. Plain Language Summary There is still large intermodel diversity in the simulated increase of extreme El Nino frequency, and it is still unclear what controls occurrences of extreme El Nino. In this study, we demonstrate that north tropical Atlantic condition is a key factor in explaining the intermodel diversity of the frequency change of the extreme El Nino, suggesting that the occurrence of extreme El Nino is considerably affected by the state of the north tropical Atlantic. Our study implies that a better understanding of the climate change over the Atlantic may significantly improve the accuracy of the best estimate and reduce the uncertainties in projection the devastating extreme climate events.