Work

Discrepancy in the Connection between the Indian Ocean Dipole and Following Year’s ENSO in Coupled Climate Models

Year: 2025

Type: article

Abstract: Observations indicate that the Indian Ocean dipole (IOD) in autumn is an important precursor to El Niño–Southern Oscillation (ENSO) 14 months later. Our study explores the performance of CMIP5 and CMIP6 models in simulating this IOD–ENSO relationship. The models exhibit significant spread in the IOD–ENSO relationship, which can be attributable to the diversity in the simulated mean precipitation over the tropical South Pacific Ocean (TSPO) and tropical South Indian Ocean (TSIO). Models with higher mean precipitation over the tropical South Pacific Ocean produce stronger surface easterly (westerly) wind anomalies over the western–central equatorial Pacific in response to the negative (positive) IOD event via modulating the tropical Walker circulation. Stronger surface easterly wind anomalies over the western–central equatorial Pacific cause more warm water to be accumulated in the western tropical Pacific through strengthened anticyclonic wind stress curl and westward downwelling Rossby waves, providing favorable conditions for the El Niño occurrence. Furthermore, higher mean precipitation enhances the air–sea interaction over the tropical Pacific, which is also favorable for ENSO development. Mean precipitation over the tropical Indian Ocean also contributes to the coupling feedback within the Indian Ocean, but this effect is relatively small. Further analysis shows that the prediction skill of the Niño-3.4 index according to the IOD index 14 months earlier reaches 0.6 with a lead time of 14 months for the high correlation models, whereas for the low correlation models, it starts to decline from May of the following year and drops to 0.4 at the peak of ENSO. This suggests that the physical connection between IOD and ENSO should be carefully considered when using predictors outside the tropical Pacific to forecast ENSO. Significance Statement Studies have suggested that IOD events, which peak in the boreal autumn, have a significant influence on the development of ENSO in the following year. Particularly, inclusion of the IOD signal in coupled climate models is expected to improve ENSO prediction skills and mitigate the ENSO’s spring predictability barrier. With the aim of utilizing IOD as a useful predictor for ENSO prediction, climate models should be able to simulate the leading relationship between IOD and ENSO as well as the associated physical processes realistically. Therefore, models’ ability in simulating the IOD–ENSO relationship is examined in this study. The results indicate that a great diversity exists among the models in reproducing this IOD–ENSO relationship. Some models well capture the observed IOD–ENSO linkage and the corresponding physical mechanism, whereas others show weak skills. We further investigate the factors that lead to the intermodel spread. The findings of this study have inspiration for enhancing our comprehension of interbasin interactions and improving the seasonal prediction capabilities of ENSO.

Source: Journal of Climate

DOI: 10.1175/jcli-d-24-0195.1