Understanding the Influence of the Tropical Intraseasonal Oscillation in Predicting the Indian Ocean Dipole

Abstract The Indian Ocean dipole (IOD) is an important source of seasonal–interannual climate prediction. However, predicting the IOD itself is a long-standing challenge due to in part the influence of tropical intraseasonal oscillation (ISO). This study revisits the influence of the ISO on the evolution of the IOD. Furthermore, the impact of intraseasonal signals on the IOD prediction is investigated using an atmosphere–ocean coupled model. Compared to the prediction without atmospheric data initialization (nudging toward observed atmosphere), the prediction skill of the IOD at lead times of 1–6 months is improved. This improvement may benefit from a realistic initial condition and prediction in the first month for tropical ISOs such as the boreal summer intraseasonal oscillation (BSISO) and its associated easterly/westerly wind events (E/WWEs). As seen in the observation and predictions, the frequency of the active BSISO in summer is significantly correlated to the SST anomalies over the eastern Indian Ocean (EIO) in autumn, and the WWEs in summer excite downwelling Kelvin waves and accumulate heat in the subsurface of the EIO. However, the effect of atmospheric nudging varies when predicting different IOD events. For instance, it improves the prediction of a positive IOD (pIOD) in 1997, while it degrades the prediction of a pIOD in 2006, which is attributed to the asymmetrical relationship between intraseasonal easterly/westerly wind and SST in the EIO. Our results reinforce the importance of high-frequency signals in the IOD prediction. The limited skill in predicting the ISO and the across-time-scale interactions may restrict the model’s capability in predicting the IOD.