Simulations of the North Tropical Atlantic Mode-ENSO Connection in CMIP5 and CMIP6 Models

The North Tropical Atlantic (NTA) mode is a dominant component of sea surface temperature (SST) variability over the northern tropical Atlantic, which is crucial to the occurrence of El Nino-Southern Oscillation (ENSO). This study utilizes the historical simulation of the Coupled Model Intercomparison Project Phase 5 and 6 (CMIP5 and CMIP6) to evaluate the simulations of the NTA mode and NTA-ENSO connection. We found that the CMIP5 and CMIP6 could reproduce the spatial pattern of the NTA mode similarly. Most of them could reasonably reproduce the spatial pattern of the NTA mode similar to the observation. Still, there are some differences in the central values and position of the NTA mode. In addition, the analysis further suggests that the CMIP5/CMIP6 models have a large diversity in the NTA-ENSO connection, which may be due to their differences in simulating the spring climatological mean Atlantic Intertropical Convergence Zone (ITCZ) precipitation and spring-to-summer NTA SST persistence. Models with stronger spring climatological mean Atlantic ITCZ precipitation in favor of the spring NTA-related SST anomalies stimulate local precipitation anomalies, resulting in a more powerful atmospheric circulation reaction over the tropical Pacific. Additionally, the stronger spring-to-summer NTA SST persistence for sustaining the teleconnection, which constantly transmits impacts from the tropical Atlantic into the tropical Pacific. The CMIP6 models show improvements to the CMIP5 models regarding simulations of the spring climatological mean Atlantic ITCZ precipitation and spring-to-summer NTA SST persistence. Therefore, the CMIP6 models simulate the NTA-ENSO connection better than the CMIP5 models.

Automated summary

This study evaluates the simulations of the NTA mode and NTA-ENSO connection using CMIP5 and CMIP6 historical simulations. The results show that most models can reproduce the spatial pattern of the NTA mode reasonably, although there are some differences in the central values and position. Furthermore, the analysis indicates that the CMIP5/CMIP6 models have a large diversity in the NTA-ENSO connection, which may be attributed to their differences in simulating the spring climatological mean Atlantic ITCZ precipitation and spring-to-summer NTA SST persistence. The CMIP6 models perform better in simulating the NTA-ENSO connection compared to the CMIP5 models.