Interannual ENSO diversity, transitions, and projected changes in observations and climate models

Interannual ENSO diversity, transitions, and projected changes in observations and climate models

Diverse characteristics of El Niño Southern Oscillation (ENSO) events challenge the traditional view of tropical coupled ocean-atmosphere systems. The probability of a transition from one type of event to another is influenced by multiple factors of which many are projected to change. Here we assess the likelihood of ENSO transitions in observations and climate models, including a distinction between events that peak in the Eastern Pacific (EP) and Central Pacific (CP). We find that the initial ENSO state influences the likelihood of certain transitions and that some transitions are not physically possible or stochastically likely. For example, transitions to CP events are more likely than EP events except from a neutral state. We also find that El Niños tend to occur as singular events compared to La Niñas. While consecutive El Niño and La Niña events of EP type are possible, opposing EP events do not occur in succession. We identify several transitions likely driven by internal dynamical processes including neutral conditions to El Niño, CP El Niño to another El Niño, EP El Niño to CP La Niña, CP La Niña to CP El Niño and La Niña, and EP La Niña to neutral and CP El Niño. Projections of future transitions show an increased probability of transitions to CP El Niño events while transitions to EP La Niña events become less frequent under a high-emissions scenario. Accordingly, transitions to these events become more and less likely, respectively. We also find changes in the likelihood of specific transitions in a warming world: consecutive CP El Niño events become more likely while EP El Niño events become less likely to transition into CP La Niña events. These changes are expected to occur as early as 2050 with some changes to be accelerated by the end of the 21st century.