Comprehensive Evaluation of Simulation Performance of Nonmethane Hydrocarbons (NMHCs) and Oxygenated VOCs in China Using a Three-Dimensional Numerical Model

Abstract

Volatile organic compounds (VOCs), particularly oxygenated VOCs (OVOCs), critically influence ozone (O₃₎ formation through free radical production. However, comprehensive evaluations of three-dimensional models in simulating both nonmethane hydrocarbons (NMHCs) and OVOCs remain scarce, hindering O₃ control strategies. This study systematically evaluates the Community Multiscale Air Quality model (CMAQv5.3.3) using hourly observations from 12 monitoring sites across China from 2017 to 2021. Results reveal that the model underestimates NMHCs by 22.0% and OVOCs by 43.7%, on average. Within the 14 abundant OVOC components, formaldehyde (HCHO) and ketones (PRD2) show overpredictions, while other OVOCs are underpredicted (4.7–94.2%), with simulated OVOC contributions to total VOCs (8.8–36.7%) being substantially lower than observations (16.6–60.8%). The model exhibits persistent underestimation in daytime concentrations across seasons and shows stronger declines in daytime concentrations from nighttime peaks compared to observations, which suggests overestimated atmospheric physical diffusion or vertical mixing processes. Residual error analysis after subtracting the simulation bias of NMHCs (primarily attributed to biases in emissions and atmospheric physical processes) highlights inadequate secondary chemical formation as a major bias source for OVOCs. Sensitivity experiments demonstrate that conventional adjustments to boundary layer and surface parametrization schemes, minimum turbulent diffusivity (Kzmin), or primary emissions failed to resolve daytime OVOCs underestimation. On the contrary, adjusting eddy diffusivity values during the day, the local vertical gradient in an unstable atmospheric state, and the yield coefficient for OVOCs chemical formation effectively enhance the simulated daytime concentrations of OVOCs and reduce the model bias. This study highlights that better descriptions of the atmospheric diffusion or vertical mixing processes during the daytime and the secondary chemical formations of OVOCs should be prioritized to improve the performance of NMHCs and OVOCs modeling.