Application and Evaluation of a NOAA GFS‐Driven Air Quality Model Using CMAQv5.4 and High‐Resolution Emissions: FIREX‐AQ 2019

Abstract The National Air Quality Forecast Capability (NAQFC) provides numerical forecasting guidance of air quality up to 72 hr ahead over the United States. In this study, we evaluate the changes in its prediction skills by updating the Community Multiscale Air Quality (CMAQ) modeling system from version 5.3.1 to 5.4 and utilizing a high‐resolution 1‐km anthropogenic emission data set. The baseline and several sensitivity simulations with emission and science updates in CMAQv5.4 were conducted and evaluated. The results show improved prediction performance among model configurations over the contiguous United States (CONUS) in terms of correlation coefficient, normalized mean bias (NMB), and normalized mean error for fine particulate matter (PM 2.5 ) with NMBs changing from −12.1% to minimum of −6.4% and maximum daily 8 hr average O 3 (MDA8 O 3 ) with an NMB changing from 4.6% to best of 0.6%. Evaluation over CONUS and along the FIREX‐AQ field campaign flight paths suggests that current NAQFC predictions can be improved by using high‐resolution anthropogenic emissions, adjustments to wildfire emissions, and newer versions of CMAQ. This work facilitates future developments of NAQFC by providing detailed information on the factors that contribute to model biases.