| Urban ecological problems have become one of the important factors affecting economic development and population health,and the impacts of air pollution on health have raised increasing concern.Mortality from respiratory diseases caused by air pollution have become the third leading cause of death from chronic diseases,with changes in PM2.5 concentrations affecting the incidence and mortality of respiratory diseases.The causes of respiratory diseases are complex,and it is necessary to understand the characteristics and potential influencing factors of respiratory diseases for the prevention and control of respiratory diseases.In this study,the monitoring data of PM2.5,land use data,road data,etc.were used to construct land use regression model(LUR),and based on this model to inverse the spatial distribution of PM2.5 concentration and analyze its characteristics of seasonal change.Restricted and Controlled Monte Carlo(RCMC)was used to disaggregate the county level respiratory diseases data,the intensity of respiratory diseases was measured by Kernel Ratio Estimators(KRE)using the disaggregated locations and hot spots were identified by Unrestricted and Controlled Monte Carlo(UCMC).Baesd on the above process,correlation analysis and geographical detectors were used to analyze the potential environmental factors of respiratory diseases.The main conclusions are as follows(1)The analysis of PM2.5 influence factors shows that there are differences in the influence direction of different geographical feature variables on PM2.5,and under the same geographical feature there is a spatial scale effect between the variables and PM2.5 concentration.The vegetation area and water area were negatively correlated with PM2.5 concentration,and both of them reached the highest correlation at 1000 m,while the effect of impervious area was the opposite,and the concentration of PM2.5 in areas with large impervious coverage was relatively high.The source of Guangzhou PM2.5 mainly comes from motor vehicle emissions,and there are analysis shows that the main road density in road traffic variables has the greatest influence on PM2.5 concentration,which further proving that the traffic plays a positive role on increasing PM2.5 concentration(2)The LUR predicted the spatial distribution of PM2.5 concentration with good performance(R2 of 0.825-0.992).Compare with the actual monitoring PM2.5 concentration value,the error of simulation value was less than 4 ?g/m3,and the average error rate is 2.96%The spatial distribution of PM2.5 concentration in Guangzhou was high in the central and western regions,while it was low at both north and south area.The high concentration area of PM2.5 was mainly located in the old central urban area of Guangzhou,and the value of PM2.5 concentration is gradually decreasing from the city center to the fringe.The temporal distribution of PM2.5 concentration in Guangzhou has seasonal changes,which show the characteristics of PM2.5 concentration high in autumn and winter,while low in spring and summer.Mean while,the spatial distribution characteristics were consistent with the whole year(3)The intensity of respiratory diseases in Guangzhou was higher in spring and summer,while the number of deaths peaked in June,with the highest total number of cases in Yuexiu district,Liwan district and Haizhu district.From December 1,2014 to November 30,2015,the KRE spatial agglomerate area located at the midwest,norst,northeast and east regions of Guangzhou.The concentration effect of respiratory diseases in the old central city is obvious,and with the increase of distance from the old urban center,the intensity of disease decreases gradually.The first class hot spot of respiratory diseases was located in Liwan district,Yuexiu district and the western part of the Haizhu district.The second class hot spot was located in the southern part of Guangzhou.The disease intensity and risk uncertainty were higher in the hot spot areas compared with other area of Guangzhou.Different from the traditional epidemiological research method which bases on aggregated data,the Kernel-based disease mapping model makes maximum use of spatial information of aggregated data and background population information,which reflects the spatial continuity characteristics of disease(4)The average annual concentration of PM2.5 was positively correlated with respiratory diseases;among the landuse variable,the proportion of vegetation area was negatively correlated with respiratory diseases;and the proportion of water area was positively correlated with respiratory diseases.For the traffic variables,all types of roads and respiratory diseases were positively correlated,but different types of roads have different influential effect,with the highest correlation of the primary road,followed by the total length of roads and density of the secondary roads.The analysis of the relationship between meteorological factors and respiratory diseases showed that temperature and wind speed were slightly positively correlated with the number of respiratory diseases death cases,while precipitation and average air pressure showed the opposite(5)The factor detector of geographical detector showed that vegetation area factor has strong influence on respiratory disease(q=0.1321),followed by the impervious area,and the influence of trunk density(q=0.0654)in traffic variables on disease intensity was statistically significant.The q statistic of PM2.5 concentration was 0.1047,and combined with risk detector,the finding indicateds that there is a correlation between PM2.5 concentration and respiratory disease.When PM2.5 concentration was high,the number of death from respiratory disease was also high.The result of risk detectior of vegetation area was differ from correlation analysis,and the risk detector result showed that in different strata,vegetation area has different influence effects.At the relatively low coverage strata,increasing vegetation area can reduce the risk of respiratory diseases,but when vegetation coverage reaches a certain level,the relieve effect on disease intensity is reduced.Futher more,the interaction detector showed that between the two environmental factors the interaction is enhanced.Among all the environmental fators,the strongest effect happen when vegetation interacted with impervious factor(q=0.2255). |
PDF Full Text Request