Analysis of NO, NO₂, and O₃ between model simulations and ground-based, aircraft, and satellite observations

Regional air quality model simulations with Community Multiscale Air Quality Modeling System (CMAQ) were evaluated using multiple platforms for a 2-month summer period. In this context, inter-comparisons of the model with available satellite-based observations of NO₂, ground- and aircraft-based observations of NO, NO₂, NO _x, NO _y, and O₃ were performed. NO₂ comparisons found low biases in CMAQ results when using both ground- and satellite-based observations. Aircraft-based observations, on the other hand, indicated a higher positive bias and error, but the overall NO₂ vertical profile was captured well by the model. The highest correlation was observed with satellite-based NO₂ observations indicating that the model and satellite found similar spatial gradients. NO concentrations were underestimated in comparison with both ground- and aircraft-based observations, especially near the surface, indicating the limitations of the model to simulate primary pollutant concentrations at point observations when there are sources nearby. NO _y comparisons found positive biases in the model when using both ground- and aircraft-based observations. The main reason for this overestimation was the consistent overestimation of peroxyacyl nitrates (PANs) in CMAQ results. Modeled O₃ concentrations compared well with lowest biases and errors when compared to aircraft- and ground-based observations. The O ₃ vertical profile indicated a small positive bias in the model results near the surface similar to the comparison with ground-based observations. However, a negative bias in the model was observed above 2 km. Comparison with aircraft-based observations revealed significant overestimations in PAN and OH concentrations. Overestimation of the modeled OH concentrations is particularly important considering the effect of OH in atmospheric reactions.