Diagnostic evaluation of ozone production and horizontal transport in a regional photochemical air quality modeling system

Abstract

Abstract A diagnostic model evaluation effort has been performed to focus on photochemical ozone formation and the horizontal transport process since they strongly impact the temporal evolution and spatial distribution of ozone (O3) within the lower troposphere. Results from the Community Multiscale Air Quality (CMAQ) modeling system are evaluated against surface and upper air measurements from field studies during summer 2002 when several high O3 episodes occurred in the eastern United States. Modeled O3 and winds are compared to research aircraft measurements and wind profiler data, respectively, to investigate whether model underestimates of daily maximum 8-h ozone concentrations during high O3 episodes might be attributable to discrepancies in either or both of these modeled processes. Comparisons of 10 AM surface O3 concentrations, which are representative of O3 levels in the residual layer aloft, revealed that model underestimation was greater at higher observed ozone levels. Mid-morning vertical ozone profiles corroborated this surface-level finding, as modeled concentrations tended to be lower than observed O3 aloft. Net ozone production efficiency (OPE) results suggested photochemical ozone formation was comparable between the model and observations with composite OPE values of 6.7 and 7.6, respectively, within the afternoon planetary boundary layer. Evaluation of wind profiles revealed modeled wind speeds with the base four-dimensional data assimilation (FDDA) approach underestimated observed speeds by more than 2 m s−1 and direction was biased by about 20° in the nocturnal residual layer aloft as coarse resolution analysis fields involved in FDDA were found to inhibit modeled winds. These differences could produce large spatial displacements in modeled and observed ozone patterns within the region. Although sensitivity simulation results with the WRF meteorological model with FDDA using all available upper air profile observations displayed improvements in capturing wind fields aloft, CMAQ maximum 8-h O3 results using the improved wind fields also underestimated observations.