High-resolution CFD modelling of traffic-related air pollutant dispersion in an urban area of Istanbul, Türkiye
Atmospheric Pollution Research, cilt.17, sa.10, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 17 Sayı: 10
- Basım Tarihi: 2026
- Doi Numarası: 10.1016/j.apr.2026.103169
- Dergi Adı: Atmospheric Pollution Research
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, INSPEC
- Anahtar Kelimeler: Air pollution dispersion, CFD modeling, Planetary boundary layer, Traffic emissions, Urban air pollution
- Yıldız Teknik Üniversitesi Adresli: Evet
Özet
Fossil fuel combustion is a major driver of air pollution, threatening environmental quality and human health, with transportation being a dominant urban contributor. This study investigates the dispersion of traffic-related emissions in a densely trafficked urban area of Istanbul, Türkiye. Ten representative test days were selected, during which vehicle counts were conducted at peak traffic hours and categorized into four groups: buses, light-duty vehicles (LDVs), cars, and motorcycles. Based on these counts, Tier-3 mass emission rates of CO, NO, NO2, and PM10 were calculated. The CFD simulations utilized on a three-dimensional geometry and terrain to accurately represent the urban landscape. The CFD model was employed to simulate the traffic-related dispersion of CO, NO, NO2, and PM10, and results were validated against concentrations obtained from the air quality monitoring station. Normalisation of background concentrations was performed according to planetary boundary layer (PBL) height variations. The normalisation procedure incorporating PBL height improved model–measurement agreement, reducing average deviations by approximately 10-15% for NO, NO2, and CO. The analysis showed that normalisation improved the coefficients of determination for NO and NO2 from 0.73 to 0.85 and 0.71 to 0.86, respectively. For CO, although the correlation was already high (R2 = 0.91), normalisation further increased it to R2 = 0.97. The model systematically underestimated NO and NO2 while overestimating CO. These discrepancies were attributed to the inadequate representation of photochemical reactions under high solar radiation and turbulence under low-wind conditions. The weakest model performance was recorded for PM10 due to its multi-source nature.