Abstract:A new theory on the ENSO cycle is advanced in this study:the ENSO is an interannual cycle of SOTA in the tropical Pacific driven by zonal wind anomaly over the equatorial western Pacific which is caused mainly by anomalous East Asian winter monsoon.El Nio(La Nia)or ENSO is really the subsurface ocean temperature anomalies(SOTA)in the tropical Pacific.The interannual cycle of SOTA in the tropical Pacific is driven by zonal wind anomaly over the equatorial western Pacific which is caused mainly by anomalous East Asian winter monsoon/The strongest signal of interannual variations of the tropical atmospheric ISO(intraseasonal oscillation)lies in the western Pacific.As an external forcing,the interannual variations of the ISO play an important role in the ENSO cycle.The occurrence of El Nio(La Nia)is related to the eastward(westward)propagation of ISO.The circulation anomalies are strongly related to the lasting time of El Nio events.Nonlinear optimization method is used to explore the predictability problems for ENSO events.Significant results are obtained.Conditional nonlinear optimal perturbation(CNOP)acts as the initial anomaly pattern that evolves into the ENSO event most probably.And the CNOP-type error superimposed on the ENSO event causes a significant spring predictability barrier(SPB)and has the largest negative effect on the prediction.A possible mechanism for SPB is provided to explain SPB,which suggests the role of nonlinearity in SPB.Besides,by using the CNOP approach,the ENSO amplitude asymmetry is addressed by nonlinearity.The decadal change of ENSO asymmetry is also revealed and an explanation is given to address the mechanism.Finally,the lower bound of the maximum predicable time,the upper bound of maximum prediction error,and the lower bound of maximum allowable initial error for ENSO events are established with three different nonlinear optimization problems,which reveal the SPB for ENSO events from three different perspectives.Furthermore,an ENSO prediction system is developed.In that system,a three-dimensional variational ocean data assimilation system is used.This is capable of assimilating in situ sea water temperature and salinity observations and satellite altimetry data.An empirical parameterization of subsurface entrainment temperature in the coupled model may effectively improve ENSO simulation.An ensemble Kalman filter data assimilation system is implemented to provide the initial ensemble.And balanced multivariate model error is used in the ensemble Kalman filter data assimilation.ENSO prediction is improved well by those ensemble forecasts.

Spatio-temporal data generation based on separated attention for ENSO prediction

ENSONet: a data-driven ENSO forecasting model with concise spatial location learning parameters and temporal embedding

A Residual Network with Geographical and Meteorological Attention for Multi-Year ENSO Forecasts

Spatial-temporal transformer network for multi-year ENSO prediction

Enhancing ENSO predictions with self-attention ConvLSTM and temporal embeddings

ENSO Forecasts with Spatiotemporal Fusion Transformer Network

Echo State Network Based on Phase Space Reconstruction: El Niño 3 Index Forecasting

Arctic Sea Ice Concentration Prediction Using Spatial Attention Deep Learning

Temporal Convolutional Networks for the Advance Prediction of ENSO

El Nino Index Prediction Based on Deep Learning with STL Decomposition

Efficient inference and learning of a generative model for ENSO predictions from large multi-model datasets

A self-attention–based neural network for three-dimensional multivariate modeling and its skillful ENSO predictions

A Hybrid Neural Network Model for ENSO Prediction in Combination with Principal Oscillation Pattern Analyses

Convolutional GRU Network for Seasonal Prediction of the El Niño-Southern Oscillation

A multi-model prediction system for ENSO

Mechanism and Prediction Studies of the ENSO

Prediction of ENSO Using Multivariable Deep Learning

Pacific Decadal Oscillation Forecasting With Spatiotemporal Embedding Network

Deep learning with autoencoders and LSTM for ENSO forecasting

Tendency-and-attention-informed deep learning for ENSO forecasts

The NAO Variability Prediction and Forecasting with Multiple Time Scales Driven by ENSO Using Machine Learning Approaches