Evaluating WRF-GC v2.0 predictions of boundary layer and vertical ozone profiles during the 2021 TRACER-AQ campaign in Houston, Texas

Abstract

The Tracking Aerosol Convection Experiment Air Quality (TRACER-AQ) campaign probed Houston air quality with a comprehensive suite of ground-based and airborne remote sensing measurements during the intensive operating period in September 2021. Two post-frontal high-ozone episodes (September 6–11 and 23–26) were recorded during the said period. In this study, we evaluated the simulation of the planetary boundary layer (PBL) height and the vertical ozone profile by a high-resolution (1.33 km) 3-D photochemical model, Weather Research and Forecasting (WRF)-driven GEOS-Chem (WRF-GC). We contrasted the model performance between ozone-episode days and non-episode days. The model captures the diurnal variations of the PBL during ozone episodes (R = 0.72–0.77; normal mean bias (NMB) = 3 %–22 %) and non-episode days (R = 0.88; NMB = -21 %), compared with the ceilometer at La Porte. Land-water differences in PBL heights are captured better during non-episode days than episode days, compared with the airborne High Spectral Resolution Lidar-2 (HSRL-2). During ozone episodes, the simulated land-water differences are 50–60 m (morning), 320–520 m (noon), and 440–560 m (afternoon) in comparison with the observed values of 190 m, 130 m, and 260 m, respectively. During non-episode days, the simulated land-water differences are 140–220 m (morning) and 360–760 m (noon) in comparison with the observed values of 210 m and 420 m, respectively. For vertical ozone distributions, the model was evaluated against vertical profile measurements from the Tropospheric Ozone lidar (TROPOZ), the HSRL-2, and ozonesondes, as well as at the surface from a model 49i ozone analyzer and a site from the Continuous Ambient Monitoring Stations (CAMS) at La Porte. The model underestimates free tropospheric ozone (2–3 km aloft) by 9 %–22 % but overestimates near-ground ozone (< 50 m aloft) by 6 %–39 % during the two ozone episodes. Boundary layer ozone (0.5–1 km aloft) is underestimated by 1 %–11 % during September 8–11 but overestimated by 0 %–7 % during September 23–26. Based on these evaluations, we identified two model limitations: the single-layer PBL representation and free tropospheric ozone underestimation. These limitations have implications for the predictivity of ozone’s vertical mixing and distribution in other models.