Datensatz

Zusammenfassung

The multi-model ensemble mean sea surface temperature (SST) of CMIP5 models shows an El Niño-like Pacific warming (PW) trend, contradictory with the observational result which manifests a La Niña-like PW pattern. Here, we demonstrate that these two SST PWs coexist in the CMIP5 models and they are largely determined by the model's tropical Indo-Pacific SST gradient. When the Pacific warms faster than the Indian Ocean (IO), the model tends to project an El Niño-like PW pattern. In contrast, a La Niña-like warming trend prevails if a more rapid IO warming is simulated. We suggest that the PW pattern in an individual model is nonstationary and may transform from a La Niña-like to an El Niño-like when its interbasin SST gradient changes with more robust warming in the Pacific, and vice versa. Moreover, based on large ensembles from both individual models and intermodels of CMIP5, we confirm that the interbasin SST gradient between IO and Pacific is tuned by the model internal oceanic and atmospheric processes. Our conclusions shed great light on the future SST PW pattern change projected by the coupled models.