A New ENSO Statistical Prediction Model considering Extratropical Effects and Its Application to the Prediction of the 2015/16 El Niño Event

Abstract Understanding the evolution mechanisms and providing accurate predictions for ENSO continue to be a challenge for the research community. In recent years, some studies pointed out that the extratropical Pacific atmosphere and ocean variations may affect the evolution of ENSO, which is different from the classical theoretical framework of ENSO. In this study, by combining tropical atmosphere and ocean predictors with extratropical precursors, an ENSO statistical prediction model was constructed. It was found that including extratropical predictors can significantly enhance the prediction skills of the model. The Victoria mode index (VMI) can significantly improve the prediction skills in the 10–12-month lead model and can effectively help the model cross the spring predictability barrier (SPB). When predicting 6–8 months in advance, the South Pacific Oscillation index (SPOI) can improve the prediction skills, but the degree of the improvement is relatively small. The model was also applied to conduct prediction experiments on the 2015/16 El Niño event. It was found that the model can successfully predict the event nearly 1 year in advance, mainly due to the inclusion of the VMI. Due to the interdecadal variation in the relationship between predictors and ENSO, when constructing the 6-month lead model, using data since 2000 and replacing the 6-month lead VMI with a 10-month lead, one can effectively improve the model’s prediction skills, especially in predicting the intensity of 2015/16 El Niño events, which has shown significant improvement.