Modeling the air pollution and aerosol-PBL interactions over China using a variable-resolution global model

Modeling air quality has always been a challenge in global models constrained by coarse grids. Here, the variable-resolution Community Atmosphere Model with full chemistry based on the scalable spectral element (SE) dynamical core (MUSICAv0) is applied in simulating air pollution with a finer grid resolution of ∼0.25° over East Asia (SE_VR), in contrast to the same model with a uniform resolution of ∼1.0° (SE_UR). Two nudging experiments and four free-running experiments are conducted to investigate the capabilities of SE_VR in modeling the air pollution and aerosol-planetary boundary layer (PBL) interactions over China. Results show the regional refinement in SE_VR is essential for simulating haze events over complex terrain areas attributed to its better performance in representing the local vertical and horizontal dispersion conditions. SE_VR shows prominent advantages over SE_UR in simulating surface ozone because of better resolving spatial segregation of NOx and volatile organic compounds (VOC) chemical regimes and subsequently representing more detailed chemical processes related to ozone formation, although the model generally overestimates surface ozone over China. Further analysis of SE_VR shows the daytime radiative effect of black carbon (BC) aerosols lowers PBL height by 12.0% (17.9%), and leads to an increase of PM2.5 by 14.5% (10.8%) under the moderate (severe) air pollution conditions over Sichuan Basin. However, SE_UR has deficiencies in simulating BC-PBL interactions due to its inability to reproduce the strong inverse temperature structure caused by BC aerosols in the lower atmosphere layer. Our results highlight the value of variable-resolution global models for simulating air pollution and its interactions with climate.