Developing a 3D NO2 Land-use Regression (LUR) Model for Air Pollution Exposure Study in Mega-city

Project: Research

View graph of relations


As air quality continues to worsen in the city in response to increasing traffic emissionsin the urban area, the health effect of air pollution will be more noticeable and severe. Therecent development of the Land-use regression (LUR) model aims to derive a cost effectivesolution for evaluating air pollution exposure for epidemiological study, which accounts forthe variability of air pollutant concentration within the city. The current methodologyallows one to apply a number of local air quality measurements with land-use relatedpredictors (traffic, land-use and topography) to derive a high-resolution spatial distribution(2D) of air pollutant concentrations within the city.In this proposal, a new 3D Land-use regression model will be developed using thetraditional 2D LUR model, urban/building morphology, and vertical profiles of airpollutant. The selected pollutant for this study is NO2, one of the predominant pollutantsfrom traffic emission, while the target area of the study is Mongkok district, the downtownarea of Hong Kong. First, the study integrates local geographical data (e.g., road network,topography, land-use type, urban/building morphology and vehicle fleet characteristics) tobuild a 2D LUR model. Different experimental data including local traffic count, mobilemeasurement of roadside NO2, and vertical profiles of NO2 are then added to formulate a 3DLUR model. With the assist of the Computational Fluid Dynamic (CFD) model, measuredvertical profiles of NO2 are converted into standard/generic vertical profiles for differentbuilding environments through building parameterization.This new LUR approach allows one to consider several additional parameters beyondthe existing 2D LUR model, which includes building height, frontal area index, and buildingmorphological parameter, specifically designed for the mega-city. The vertical profiles ofNO2 for different built environments are incorporated to achieve a 3D LUR model, whichprovides a way to account for the actual floor height of where an individual is located.Since the vertical profiles of NO2 in urban high-rise buildings exhibit exponential/thirdorder-like decay with height, the inclusion of vertical profiles improves the overallperformance of actual air pollution exposure of an individual in the epidemiological study.It also provides valuable information of general indoor/outdoor NO2 air quality in HongKong. Success of the proposed research will lead to a better method of predicting airpollution exposure for epidemiological studies in the mega-city.


Project number9042555
Grant typeGRF
Effective start/end date1/10/1710/09/19