Spatio-temporal Statistical Study Of Urban Air Pollution In China Based On BSTIM Model

The 21 st century is the era of mankind’s full entry into the urbanization.Since the developed countries have already reached a high level of urbanization,the regions with rapid urbanization level are mainly concentrated in the developing countries at present.As the biggest developing country in the world,China’s urbanization level has been one of the focus of the world in the 21 st century.With the rapid urbanization,there are all kinds of problems.Urban air pollution is undoubtedly one of the hot issues in concern in China and even in the world.Urban air pollution has caused great harm to the health of urban residents and has a great impact on the production activities of the city.Therefore,it is of practical significance to study the urban air pollution problem and analyze the spatial and temporal evolution of air pollution in cities of China.For the study of urban air pollution,domestic scholars generally choose only one city or region to do research,and only adopt classical statistics or geostatistics as the research method.No scholars have used the Bayesian spatiotemporal interaction model to study the air pollution in the city of our country.In the statistical inference of spatiotemporal problems,Bayesian statistics method takes the spatial-temporal correlation in spatio-temporal problems into a model with prior information.It makes Bayesian statistical inference more scientific and reasonable.As one of the Bayesian statistical methods,the Bayesian spatio-temporal interaction model(BSTIM)combines Bayesian hierarchy model with the spatio-temporal interaction model.By introducing transcendental information,Bayesian spatio-temporal interaction model can effectively reflect the trend of temporal and spatial variation of data,so as to effectively solve the problem of spatio-temporal data correlation.In order to conduct research on status and laws about the national urban air pollution,this paper collects air pollutants data from 1497 sites throughout the country from May 13,2014 to December 31,2017.These sites cover all 338 cities except Hong Kong,Macao and Taiwan province.In the empirical analysis,this paper first analyzes the spatial and temporal characteristics of urban air pollution in China through descriptive statistical analysis and spatial autocorrelation study of urban air pollution.Secondly,the air pollution situation in China is analyzed by using the spatio-temporal visualization method.Finally,this paper constructs the spatio-temporal interaction model.Combining with the spatio-temporal evolution map and cold-hot spot distribution map,the results of the total spatial effect parameters are compared and analyzed in this paper.Through the above methods,the spatial and temporal distribution of urban air pollution in China is analyzed and studied.By accomplishing the research above,the following conclusions are obtained:(1)during the study period,the PM2.5 air pollution data in Chinese cities all show the periodic fluctuation law of "four wave peak",and the number of air polluted cities in China every year also presents the spatial distribution law of "expansion-contractionexpansion".(2)Although the urban air quality in China has been greatly improved in 2017,the most serious period of PM2.5 pollution in China is from December to January,while July is the best period of air quality.(3)China’s core air pollution area in the city has the temporal and spatial evolution law of "autumn and winter continuous pollution,spring and summer sporadic pollution,seasonal difference is more obvious".That is,the air pollution in Chinese cities did not show a large spatial difference during the severe pollution period of autumn and winter,while the light pollution period in spring and summer showed a large spatial difference.(4)Comparing the number of hot and cold cities in China,the distribution law of "south less north and west less east" is presented in summer hotspot cities,but the overall trend is decreasing year by year.In winter,most cities belong to the warm spot city,which shows that the air pollution level in winter is relatively small.(5)Comparing the number of cold spot cities and hot spot cities in different provinces,the spatial distribution law is "many hot spot cities in the north and cold spot cities in the south".According to the spatial-temporal distribution characteristics of urban air pollution in China,this paper proposes to improve urban air quality in China from the aspects of developing clean energy,carrying out targeted control measures,establishing a joint prevention and control mechanism,and transforming the economic development mode of heavily polluted cities.