P62 Air Quality Modeling over the U.S.: Multi-Model Evaluation and Intercomparison

Jena, Chinmay, Yang Zhang, and Kai Wang, Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University

Online-coupled meteorology and chemistry models can simulate feedbacks of chemistry into climate, thus more realistically representing the real atmosphere compared to offline-coupled models. Accurate predictions of the concentrations and spatiotemporal variations of multi-pollutants using online-coupled models are crucial for their health effect assessment in a changing climate. In this work, four advanced online-coupled models (WRF/Chem-CAM5 v3.4, WRFv3.4-CMAQ v5.0.2, WRF/Chem v3.7.1, WRF/Chem-ROMS v3.6.1) are applied and inter-compared for five full years of 2008 – 2012 over North America (NA) centered in the U.S. Comprehensive evaluation of meteorological and chemical predictions is performed using surface observations and satellite retrievals. The objectives are to (1) evaluate current regional models' capability in reproducing the observations of major pollutants such as ozone (O3) and fine particulate matter (PM2.5) for their health impact assessment, (2) identify the areas of potential improvements to enhance the models' skills, and (3) generate the best possible model predictions for 2008 – 2012 to serve as the baseline for future-year simulations under a variety of energy transition and climate change scenarios. A comprehensive evaluation shows overall good performance for meteorological variables except for wind speed at 10-m and precipitation against National Climatic Data Center. Concentrations are overall well predicted for O3, PM2.5, elementary carbon, total carbon, and column NO2 over NA. However, moderate-to-large biases exist for other species in some regions. These biases indicate uncertainties in the model representations of boundary layer processes (e.g., surface roughness), cloud processes (e.g., microphysics and cumulus parameterizations), emissions (e.g., biogenic, wildfire, and dust emissions), chemistry and aerosol treatment (e.g., winter photochemistry, aerosol thermodynamics), as well as inaccurate boundary conditions (e.g., carbon monoxide, O3). The preliminary results show that WRF-CMAQ performs the best for surface O3, and WRF/Chem performs the best for surface PM2.5 among all four models evaluated in this work. Diagnostic evaluation and sensitivity simulations will be performed to pinpoint the likely causes for potential improvement of the models' skills. Improved model results will be used to identify pollution hot spots for air quality and human health impact studies under a variety of energy transition and climate change scenarios.