3-D Mobile Monitoring and CFD Modeling of PM and BC Distributions in Urban Air Pollution Hotspots

Air pollution caused by mobile emission sources in urban areas is typically concentrated in populated areas, and therefore has a direct and wide impact on human health. Based on regional basis, air pollutants emitted from vehicles are often present at a high concentration due to local and on-road contributors and have large variability in time and space. Therefore, there is a limitation in capturing the general spatio-temporal distributions in air pollution hotspots. In this study, we have utilized the three-dimensional (3-D) mobile monitoring technique by installing accurate measurement equipments on a mobile laboratory and an unmanned aerial vehicle (UAV) such as a drone to examine the spatial distributions of air pollution hotspots in densely-populated urban areas. In addition, an atmospheric dispersion modeling technique using a computational fluid dynamics (CFD) model has been introduced to simulate the atmospheric flow and pollutant dispersion in a complex urban area in presence of building and topographical effects. Among various air pollutants, we selected particulate matter (PM) and black carbon (BC) as target species characterizing urban air pollution hotspots. The spatiotemporal variabilities of air pollution hotspots simulated using the CFD model were then compared to the those measured using the 3-D mobile monitoring. In the comparative results, the BC concentration showed larger variability than PM concentration, which is consistent in both 3-D mobile monitoring and CFD modeling. In air pollution hotspots, air quality is severely worsened under the inefficient atmospheric dispersion caused by densely located high-rise buildings and/or stable atmospheric stability especially in the early morning. High-rise buildings and signalized road intersections are dominant contributors in elevating air pollution levels. The comparative study has proven that utilization of two different 3-D techniques at similar scales is feasible to complementarily reduce the knowledge gap between air pollution monitoring and modeling.