Assessment of quasi-global surface ozone simulations with WRF-chem using multi-dataset evaluation

This study performs a quasi-global surface ozone (O₃) using the Weather Research and Forecasting model coupled with Chemistry (WRF-Chem). The primary goal is to evaluate the O₃ simulation performance of WRF-Chem. The model reproduced surface meteorology reasonably well, with the average correlation coefficients (R) and index of agreement (IOA) exceeding 0.7 when compared to ground-based observations from March 2019 to February 2020. The Tropospheric Ozone Assessment Report (TOAR) daily O₃ observational data and two reanalysis datasets, CAMS and GEOS-CF hourly data, are used in this study to evaluate O₃. WRF-Chem captured surface O₃ variations in Europe when compared to TOAR data, with R greater than 0.6. Asia and Africa had an average R of 0.5 and a mean bias (MB) of -4.6 ppb, while the Americas did worse, with South America exhibiting an average overestimation of 11.2 ppb. The differences in nitric oxide (NO), nitrogen dioxide (NO₂), and O₃ between RADM2 and RACM are further examined. The largest NO₂ differences occur over eastern China, northern India, and Europe, whereas the O₃ differences are concentrated over tropical rainforest regions and the equatorial oceans, with mean absolute differences exceeding 10 ppb. Overall, the findings indicate that more advancements in model structure, chemical mechanisms, and meteorological parameters are still required to improve WRF-Chem's capacity to simulate quasi-global O₃.