A Further Study on the Simulation of Cloud-Radiative Feedbacks in the ENSO Cycle in the Tropical Pacific with a Focus on the Asymmetry

The El Nino-Southern Oscillation (ENSO) cycle has been used as a surrogate to gauge the accuracy of simulated cloud-radiative feedbacks. However, previous studies have largely ignored the potential differences in the cloud response to sea surface temperature (SST) changes between the two phases of ENSO. Here we underscore this issue by contrasting the differences in the responses of shortwave- and longwave- cloud radiative forcing (SWCRF and LWCRF) to SST changes between the two phases of ENSO in observations and in CMIP5 simulations. In observations, the responses of SWCRF and LWCRF to SST changes in the cold phase had weaker magnitude and more westward positioning than the warm phase. This asymmetric feature in cloud-radiative feedbacks is linked to the corresponding differences in the precipitation feedback. In examining this asymmetry aspect of cloud-radiative feedbacks in the CMIP5 simulations, we divided the models into high-skill-models (HSM) and low-skill-models (LSM). The HSM were found to simulate the cloud-radiative feedbacks equally well in both phases, whereas the LSM had distinctly better simulations in the cold phase than the warm phase, to the extent that it matched the simulations by the HSM. Thus the difference in the simulation of cloud-radiative feedbacks between the HSM and LSM primarily arises from the difference in the warm phase. This is because the cold bias in mean SST in coupled models had a greater impact on the cloud-radiative feedbacks during the warm phase than during the cold phase. Implications for how to further improve cloud-radiative feedback simulations were also discussed.