Assessment of surface downward longwave radiation in CMIP6 with comparison to observations and CMIP5

Surface downward longwave radiation (SDLR) plays an important role in understanding the greenhouse effect and global warming. The simulated SDLR from 47 coupled models in the Coupled Model Intercomparison Project (CMIP6) general circulation models (GCMs) was evaluated by comparing them with ground measurements and CMIP5 results. The estimated SDLR using all CMIP6 GCMs based on the multimodel ensemble (MME) methods was validated as well. The bias values of the SDLR simulations from individual CMIP6 GCMs averaged over the selected 183 sites around the world varied from-10 to 10 W m(-2), while the root mean squared error (RMSE) values ranged from 20 to 26 W m(-2). Compared to CMIP5 models, the CMIP6 GCMs did not show a significant tendency to underestimate SDLR. However, the SDLR from CMIP6 GCMs exhibited the relatively better precision at low altitude and low latitude sites compared to that at high altitude and high latitude sites. Moreover, the Bayesian model averaging (BMA) method increased the correlation coefficient (R) by approximately 0.02 and reduced the RMSE by approximately 5 W m(-2) on average compared to the individual CMIP6 GCMs. The trend in SDLR was also investigated in this study, which has been related to the changes in air temperature (SAT), and water vapor pressure (WVP).