El Niño-Southern Oscillation (ENSO) is the dominant mode of inter-annual climate variability. This climate mode is excited by ocean-atmosphere interactions in the tropical Pacific but its global teleconnections extend far beyond the tropics. Yet, it is extreme El Niño events that have the largest impacts on global climate and weather patterns, ecosystems, and human health. Three such extreme El Niño have occurred in recent decades (1982, 1997, and 2015), leading to billions of dollars in damages worldwide. Several other El Niño events in the observational record were only slightly weaker. With rising temperatures, should we expect more frequent and/or even stronger extreme El Niño events? Studies disagree however on whether the probability and magnitude of extreme El Niño events would increase with global warming, which can be largely explained by major differences in the physical mechanisms that control extreme events across climate models and their divergent responses to rising CO₂ concentrations in the atmosphere.

Three key ingredients contribute to the occurrence of extreme El Niño events. The first one is westerly wind bursts (WWBs). Lasting several weeks, they occur in the western and central equatorial Pacific at the onset and during the development of El Niño events. Several strong initial bursts in winter, followed by a sequence of WWBs coupled to the underlying sea surface temperature (SST) anomalies during El Niño development, can result in a strong positive Bjerknes feedback, leading to a strong El Niño. This positive feedback is responsible for the rapid growth of SST anomalies during El Niño development. Recharge and the subsequent discharge of ocean heat content (OHC) in the western equatorial Pacific is another factor – a larger initial OHC may lead to a stronger El Niño event. Finally, the nonlinearity of wind stress response to SST anomalies related to shifts in atmospheric convection can also amplify El Niño by further strengthening the Bjerknes feedback. When these factors operate together, conditions can become right for the development of an extreme El Niño. Nevertheless, even a strong initial OHC and strong WWBs early in the year do not always lead to an extreme event (as happened for example during the weak 2014 El Niño). Some strong events appear to develop even in the absence of a strong initial OHC (e.g., as in 1982). Furthermore, how climate models replicate these factors can be different between the models and the observations, and across different climate models.

Accordingly, the overarching goal of this project is to quantify how the interplay between WWBs, ocean heat content, and convective nonlinearities controls the occurrence of extreme El Niño events in Energy Exascale Earth System Model (E3SM) and other Earth system models in the present and future climates. To achieve this goal, firstly the project will systematically analyze the temporal and spatial WWB characteristics and their links to extreme El Niño events, using preindustrial, historical, and future climate simulations, and available observations. Then, a broad hierarchy of numerical experiments will be conducted with the aim to investigate the mechanisms and impacts of WWBs. Specifically, mechanism denial experiments will be used to identify the main factors responsible for WWBs, how they are related to tropical cyclones, and how they affect ENSO events. Further, WWB-ENSO interactions will be examined in modified coupled simulations (using the so-called “wind shaving” approach) limiting the effect on the ocean of wind stress induced by WWBs, allowing us to rigorously quantify the impacts of WWBs on ENSO statistics as well as on individual events. We will be able to assess the relative roles of WWBs, OHC, and convection-related nonlinearities in the development of extreme events. We will also investigate the effects of climate warming on WWBs and on their interaction with ENSO. The contribution of WWBs to the skewness and kurtosis of ENSO and their changes with global warming will be examined (these important nonlinear parameters are controlled by strong El Niño events, and many models fail to realistically reproduce them).

Ultimately, this project will lead to a better understanding of El Niño dynamics, especially of the occurrence and predictability of extreme El Niño events, with broad socioeconomic benefits stemming from the vast impacts of such events. It will also investigate potential future changes in extreme events with global warming and may help improve ENSO simulation by E3SM and other climate models. Finally, the project will support a postdoctoral associate at Yale University and a graduate student at Harvard University.