Numerical investigations of reactive pollutant dispersion and personal exposure in 3D urban-like models

With satisfactory validation by experimental data, we perform computational fluid dynamic(CFD) simulations with the standard k-ε model to investigate how NO–NO2–O3 photochemistry and turbulent mixing influence reactive pollutant dispersion and vehicular NOx exposure in 21-row(neighborhood-scale~1 km) three-dimensional(3D) medium-dense urban models with an approaching wind parallel(perpendicular) to the main(secondary) streets. Personal intake fraction P_iF and its spatially-averaged values for the entire building (i.e. building intake fraction B) are adopted for reactive/passive exposure analysis with/without NOx-O3-photochemistry. Some meaningful findings are proposed: 1) There are flow adjustment processes coupling turbulent mixing and chemical reactions through urban areas (i.e. secondary Street 1 to 20). NO–NO2–O3 photochemistry induces O3 depletion and NO conversion into NO2 producing significant increase in NO2 exposure and slight decrease in NO exposure compared with passive dispersion. 2) With span-wise NOx sources, Street 10 in the fully-developed region experiences weaker wind and subsequently greater B(0.207 ppm) than Street 3(0.135 ppm) in the upstream flow-adjustment region. B descends exponentially from the target building toward downstream, and Street 10 experiences quicker decay rates. 3) With stream-wise NOx sources along the main street, B first ascends, then reaches equilibrium values (e.g.0.046–0.049 ppm). 4) If background O3 concentration [O3] rises from 20 ppbv to 40 and 100 ppbv, more NO is oxidized by O3 to generate NO2. As [O3] = 20 ppbv, if NO–NO2 emission ratio decreases from 10 to 5, NO2 exposure is partly offset but NO exposure change little. Present methodologies are confirmed effective to investigate impacts of more complicated meteorological conditions and chemical mechanisms on exposure in urban districts