A multi-sectoral high-resolution emission refinement system for CFD-GIS based air quality modeling

Precise urban air quality management requires a high-resolution understanding of complex emission landscapes. This study proposes a multi-sectoral high-resolution emission refinement system for urban-scale modeling and evaluates its impact through comparative numerical experiments. By leveraging building-specific GIS data and road network information, the Clean Air Policy Support System (CAPSS), South Korea's national inventory, was refined to a 20-m resolution and chemically speciated for the Harvard chemistry mechanism to be used in CFD-Chem modeling. The high-resolution emission experiment (HE) successfully resolved localized concentration hotspots intrinsically linked to urban morphology, which were obscured in the 1-km grid-based conventional emission experiment (CE). Specifically, the HE demonstrated superior performance in reproducing the spatial heterogeneity of NO 2 concentration, showing strong agreement with road-adjacent monitoring stations. The difference ratio analysis between the two experiments revealed that NO 2 concentrations in the HE differed by up to 245.6% from the CE in dense urban corridors, effectively mitigating smoothing errors inherent in coarse-grid emissions. Furthermore, the HE proved pivotal in capturing non-linear photochemical reactions; refined roadside NO emissions in the HE actively triggered local ozone titration, resulting in a reduction in ozone concentrations of up to 50.3% along major road networks compared to the CE. While simulation fidelity for CO and PM 2.5 depended on regional background concentrations, the HE provided a more realistic representation of localized chemical destruction and production cycles. This study establishes a robust scientific foundation for precision air quality management, offering a key instrument for designating intensive management zones and establishing effective pedestrian exposure mitigation measures.