Advancing urban air quality modeling with solar radiation-included computational fluid dynamics simulations

Abstract

Air quality modeling is a challenging task due to the complex interactions among meteorological, chemical, and physical processes. Accurate predictions require sophisticated models that can simulate these interactions under varying environmental conditions. This study advances micro-scale urban air quality modeling knowledge by considering solar radiation-included Computational Fluid Dynamics (CFD) simulations. Using data on air pollution concentration monitored in Antwerp, Belgium, this research evaluates the benefits of incorporating non-isothermal conditions and solar radiation into CFD air pollution dispersion models compared to the traditional neutral boundary layer assumption. The study analyzes twelve hourly meteorological case studies under non-isothermal conditions with solar radiation included to assess the accuracy and performance of both approaches. Findings reveal that the non-neutral CFD model outperforms the neutral model in 75% of the scenarios, with differences in the concentration maps ranging from 8% to 32%. These disparities significantly affect the identification of concentration hotspots and pollutant level predictions. Accuracy assessments against in-field monitored data show notable reductions in relative errors for the non-neutral model, averaging 21%, compared to 40% for the neutral model. While both models perform similarly under low wind speed and low solar radiation conditions, the non-neutral model generally demonstrates superior performance under higher solar radiation and more variable wind speeds. Overall, the non-neutral model offers an improvement factor of around 2 compared to the neutral model. The integration of solar radiation in CFD simulations represents a significant advancement in urban air quality modeling, with promising implications for enhancing air quality management and urban planning decisions.