| Since the reform and opening up,with the rapid urbanization and industrialization,the living standards of urban residents in China have significantly improved.However,a large number of pollutants were concentrated emitted into the urban air environment,exceeding the carrying capacity of the environment,resulting in an extremely serious problem of urban air pollution,and the living environment and health of urban residents were also seriously threatened.After government control,the air pollution in China has improved slightly,but the concentration of pollutants in the air still exceeds the restrictions of the World Health Organization.In addition,the process of urbanization and industrialization in China has accompanied a special suburbanization.Different from the excellent and comfortable suburban environments in developed countries,the environmental resources of the suburban population in China are becoming increasingly unequal.This group of people who suffer from high-concentration air pollution exposure doses during their long-distance commuting every day deserve attention.This study focuses on the construction and result analysis of a model for the dynamic spatial-temporal distribution and population exposure doses of air pollution at the urban scale.The contents are as follows:（1）Past research has quantitatively evaluated the negative impacts of long-term exposure to air pollution.However,the dynamic changes in air pollution levels and population flow during the day are often ignored.We used high-accuracy PM2.5data from micro-air monitoring stations and population heat maps generated based on big data from location services to simulate hourly population flow,and evaluated the hourly active PM2.5exposure levels of typical Chinese large cities.The dynamic"active population exposure"was then compared with the static"census population exposure"based on demographic data from the World Pop data.The results showed that during the two study periods,the weighted population exposure（PWE）of suburban areas was underestimated by 45.83%,while the overestimate of rural PWE and central city PWE was 100%and 34.78%,respectively,with the greatest relative difference reaching-11μg/m3to 7μg/m3.More importantly,during the weekday morning peak hours,the total PWE of the active population was much higher than previously estimated,with approximately 12.41%of people exposed to PM2.5concentrations greater than 60μg/m3,more than twice the level in the Census scenario.Commuters living in suburban areas and working in central cities may be more vulnerable to PM2.5exposure and the unequal allocation of environmental resources.A new method for calculating population exposure was proposed,which can be expanded to quantitatively evaluate other environmental issues and their associated health burdens.（2）A GIS database was constructed,and all information,including land use,urban layout（such as points of interest and building information）,big data from transportation,traffic situation,road network,sources of emissions,and meteorological data,was fully integrated.These data were filtered and reclassified,and a vector dataset and 10m grid dataset containing 186 independent variables were finally created for Changsha.After cleaning and statistical analysis of the PM2.5,PM10,NO2,O3,SO2,and CO monitoring data from 144 micro-air monitoring stations in Changsha from January 1,2019 to December 30,2020,the average concentration of six pollutants during the weekday morning and evening peaks（7:00-9:00,17:00-19:00）was calculated for the year,quarter,and month.The data integrity of SO2and CO was low,and the spatial heterogeneity of the data was also low.Finally,we selected PM2.5,PM10,NO2,and O3as dependent variables for land use regression（LUR）modeling,and a LUR model was established using 10-fold cross-validation for final model validation.Big data from location services（LBS）was used for the first time in LUR modeling,and its advantage is real-time update and high spatial accuracy,which can significantly improve the accuracy of the model.The R2of the final model’s annual average concentration was 0.31-0.75,and the seasonal differences in the modeling performance of PM2.5,PM10,NO2,and O3were significant,with lower modeling performance in summer than in winter.O3had the poorest modeling performance,which was also related to its being a secondary pollutant.In the future,analysis of O3using quantitative chemical models such as WRF-Chem will be needed to improve the explanation level of the model.（3）People’s urban transportation-related air pollution exposure occurs mainly during their daily commute.In previous studies,the simulation of commuting routes usually only considered the fastest route and the lowest air pollution exposure route.Few studies have paid attention to the actual commuting routes of people.This study used a spatial grammar approach,from a human perspective,to simulate a commuting route that was both spatially and behaviorally verified,and compared it with other routes to analyze the actual exposure situation of commuters.The results showed that only 3.4%—13.2%of the fastest route and lowest air pollution exposure route in Changsha were different,and the average exposure difference of PM2.5and NO2was less than 1%.However,more than 52.3%—69.5%of the behavioral verification route was different from other routes,and the average accumulated exposure dose of the behavioral verification route was 7.81%—13.2%higher than that of other routes.Therefore,it is possible that the high dosage exposure of commuters during the initial and last kilometers of their active commute may have been underestimated.In addition,highly connected road networks and non-gated communities are more conducive to people avoiding high-exposure routes.Finally,the high commuting time cost and accumulated exposure dose of residents in the outer areas of the city indicate that the environmental resources of these areas are unequally distributed and deserve attention.The results of this study suggest the possibility of interdisciplinary and multi-disciplinary cooperation between behavior science,urban morphology,and environmental assessment,and confirm the suitability of spatial grammar and other spatial behavior science methods for combining with big data and Geographic Information System（GIS）to predict environmental problems such as air pollution exposure and propose targeted sustainable development interventions. |
PDF Full Text Request