Recorded presentation4.5
Development of a coupled CFD-photochemistry model and its use in studying reactive pollutant dispersion in street canyons
Kyung-Hwan Kwak, Seoul National University, Seoul, South Korea; and J. J. Baik
A computational fluid dynamics (CFD) model coupled with a photochemistry model is developed to study the dispersion of reactive pollutants in street canyons. The CFD model is a Reynolds-averaged Navier-Stokes equations (RANS) model with the renormalization group (RNG) k-epsilon turbulence closure model and the photochemistry model includes the carbon bond IV mechanism (36 gaseous species). A two-dimensional street canyon with a canyon aspect ratio of one is considered and NOx and volatile organic compounds (VOCs) are emitted near the street bottom. Four distinct types of pollutant concentration distribution in the street canyon are classified in the presence of polluted ambient air: an emission type, a core type, a shear-layer type, and an entrainment type. These types are associated with different chemistries of pollutants in and above the street canyon, and within a shear layer formed between in-canyon and above-canyon regions. The relative importance between NO titration and NO2 photolysis determines a net O3 production or loss. NO titration (NO2 photolysis) dominates over NO2 photolysis (NO titration) in (above) the street canyon, resulting in a net O3 loss (production). The shear layer is a chemically active region. In the shear layer, the activities of short-lived species (such as NO3, HO2, and RO2) are much higher than those of long-lived species. O3 concentration in the street canyon decreases as the NOx and VOCs emission ratio increases. The sensitivity of O3 concentration to NOx and VOCs emissions ratio is affected mostly by NO titration and slightly by VOC chemistry, even though the polluted area we consider is regarded as a NOx-saturated regime.
Session 4, Atmospheric Transport and Dispersion in Urban Areas
Tuesday, 3 August 2010, 3:30 PM-5:15 PM, Crestone Peak I & II
Previous paper Next paper