Evolution and prediction of the La Niña conditions in 2024/25

Abstract Skillful El Niño–Southern Oscillation (ENSO) prediction is one of the most important problems in climate science due to its substantial global impacts. There have been many successful examples of predicting ENSO using dynamical climate models since the mid-1980s. It was therefore unexpected that many operational climate models significantly overestimated the likelihood of La Niña conditions in 2024. In this report, we examine the physical processes associated with the arrested development of La Niña conditions in 2024/25, and the possible reasons for overestimated predictions of its strength. Despite favorable subsurface cooling conditions following a strong 2023/24 El Niño, we argue that arrested development of La Niña conditions in 2024/25 resulted from weak episodic easterly wind anomalies and associated weak upwelling Kelvin wave activity, which failed to shoal the thermocline sufficiently to initiate basin-wide air-sea coupling. Furthermore, we find that weaker Kelvin wave activity and ENSO amplitude reduction were linked to the ENSO regime shift with strengthened mean zonal sea surface temperature contrast and enhanced mean trade winds in the tropical Pacific around 2000. Model limitations in capturing atmospheric variability and interdecadal shift contributed to the overestimated strength of the La Niña predictions in 2024/25, underscoring the importance of properly simulating atmospheric variability and the interdecadal regime shift in dynamical models used for predicting ENSO.