Improvements in the relationship between tropical precipitation and sea surface temperature from CMIP5 to CMIP6

Precipitation and sea surface temperature (SST) are important variables, the coupling of which is crucial in understanding the variation in the Earth’s climate under the effects of global warming. We evaluated the precipitation–SST (P–SST) and precipitation–SST tendency (P–SST′) correlation by comparing the updated Coupled Model Intercomparison Project Phase 6 (CMIP6) results with observational datasets and results from the CMIP5 models. In general, the multimodel ensemble mean of the CMIP6 models greatly improved the simulation of the annual P–SST′ correlation, with the spatial correlation coefficient increasing by 9.73% and the root-mean-square error decreasing by 7.00% compared with the CMIP5 models. The improvement ratio was greater than the simulation of the P–SST correlation. Air–sea interactions in the tropics vary with both season and location. We found a great improvement for the P–SST correlations in spring and summer, but only a slight improvement in autumn and winter, in contrast with the simulation of the P–SST′ correlations. Specifically, the spread among the CMIP6 models was reduced for the P–SST correlation over the equatorial central-eastern Pacific and for the P–SST′ correlation over the western North Pacific. By contrast, the CMIP6 models displayed poorer results, with a deviated bias over the maritime continent and the western Indian Ocean. We used intermodel empirical orthogonal function analysis to show that the model spread of the P–SST and P–SST′ correlations was mainly determined by the climatological precipitation. These results provide a deeper understanding of the co-variability between tropical precipitation and SST and will improve predictions of the future regional climate.