Spatial Distribution Analysis Of Urban Traffic-induced Particulate Matter Using LUR Model

Traffic-related air pollutants,especially particulate matter,has a significant negative impact on the human living environment,as well as long-term health.Knowing the distribution and intensity of different traffic-related air pollutants at a community level would provide the reference for both the reduction of human exposure to these pollutants and improvement in urban planning and regulations.In recent years,although among many studies about air pollution,land use regression(LUR)method has been used effectively in dealing with mapping city’s or region’s air pollutant levels,the applicability of using the LUR method in a smaller area yet need further research.Hence in this paper,we conducted a mobile sampling using portable dust monitors(GRIMM)and weather&environmental meter(Kestrel),as well as GPS loggers in an extent of few blocks in Xi’an,and developed LUR models for PM10,PM2.5,and PM1.The dispersion model AERMOD was also used for the same study area,so as to make a performance comparison of dispersion model and LUR model.The dependent variable in the establishment of the models are the concentrations(or the logarithms of them)of the 3 types of PM.Independent variables in the regressions involved traffic intensity,meteorological data,and land use types,which were extracted from satellite image analysis of the studied area in ArcMap and classified into dwellings,green space,road space,and open space.The final PM10,PM2.5,and PM1 model accounted for 50.1%,73.4%,and 39.3%respectively of their variance;relative humidity,wind speed and green space within150m radius,were indicated as some of the influential variables;due to the limited scope of the study area,traffic intensity did not become an influencing factor.With the map-processing and analysis software ArcMap,the distribution of PM concentration in the study area has been visualized.The visualization results are close to the measured values,and therefore are able to provide information to the residents and city decision makers about the pollution hot spots,so they can understand and assess the pollutants exposure levels and health risks better.Compared with AERMOD,LUR method has advantages in source setting,data preparation,results and its wide range of application.This research indicated that in a dense city area,LUR model can perform well with an acceptable R~2.Along with mobile monitoring,mapping spatial distribution of air pollutants using LUR method can be more effective and flexible than other methods like dispersion models.