The Role of the South Pacific Oscillation in ENSO Predictions

Recent advances in tropical Pacific climate variability have focused on understanding the development of and accurately predicting El Niño–Southern Oscillation (ENSO) events. Much work on ENSO predictability rely on precursor patterns from the North Pacific, including the North Pacific Oscillation (NPO) and its role in the seasonal footprinting mechanism / North Pacific Meridional Mode (NPMM). While some have presented evidence of South Pacific influences on ENSO formation, this avenue is less explored. Furthermore, distinguishing the particular flavor (i.e., Central versus Eastern Pacific ENSO events) of the expected ENSO event also remains an open question. This study offers a fresh look at ENSO predictability using South Pacific atmospheric conditions during June – August (JJA; i.e., austral winter) as a precursor signal. The positive phase of the leading mode of South Pacific sea level pressure variability, which we call the South Pacific Oscillation (SPO), exhibits a meridional dipole with a(n) (anti)cyclonic anomaly dominating the subtropical (extratropical) South Pacific. The SPO contains both intrinsic variability and a forced component from tropical Pacific SSTs during JJA. In relations to ENSO development, if the positive SPO mode is present during JJA, cyclonic anomalies in its subtropical node weaken the climatological southeasterly trade winds and promote anomalous equatorward mass transport, effectively charging the eastern tropical Pacific, yielding an Eastern Pacific ENSO event. The absence of a strong positive SPO signature in JJA isolates the warm water anomalies in the central tropical Pacific, increasing the chances of a Central Pacific ENSO event. With this knowledge, we show a high accuracy rate in predicting the type of ENSO event that will result using only the SPO as a predictor for the flavor. The potential role of the SPO in explaining the asymmetry of warm and cold ENSO events and its role in multi-timescale Pacific climate variability (seasonal to decadal) will also be discussed.