A Parameterization Scheme for Westerly Wind Bursts in the Western Tropical Pacific Based on the U-Net Model and its Relationship with ENSO

El Niño–Southern Oscillation (ENSO) refers to the strongest interannual climate mode in the atmosphere–ocean coupling system over the tropical Pacific. Westerly wind bursts (WWBs) serve as important precursors of ENSO events. In the past decades, several statistical and dynamics-based models have been used to simulate and predict ENSO events. However, these models suffer from some defects related to the representation of WWBs, which lead to their limited simulation and prediction performance of ENSO events. This study constructs U-Net models that are driven by several atmospheric and oceanic data for zonal wind anomalies (u_a) in the tropical Pacific. Furthermore, the spatio–temporal characteristics of WWBs are identified, and their relationship with ENSO is analyzed. The results indicate that the U-Net model can effectively simulate the WWB characteristics in spatiotemporal distribution during the testing period (2003–2022). The deviation between the occurrence frequency and accumulated days of WWBs reconstructed by the model and the observed values is less than 4%, and the correlation coefficient between the time series of WWB occurrence probability (P) by the model and that observed reaches 0.87. Meanwhile, the U-Net model can effectively capture the nonlinear relationship between WWBs and ENSO events, and a significant lead–lag correlation is detected between them in the U-Net models and reanalysis data. Meanwhile, the duration, zonal width, and average maximum amplitude of WWBs reach their peak during El Niño events, and the peak probability of WWBs during El Niño events reconstructed by the U-Net model is considerably closer to that of observation. In conclusion, compared with the traditional WWB parameterization scheme, which relies on the establishment of an approximate linear relationship between WWBs and sea surface temperature fields, the U-Net model offers important advantages in representing WWBs.