Spatial Characteristics Of Air Quality In China’s Provinces

In the process of rapid socio-economic development,urbanization and industrialization gradually intensifies,improving air quality and largely reducing air pollution emissions,and establishing economic-environment friendly society has become a major problem in China’s economic transformation and upgrading.In view of the panel data model considering individual spatial heterogeneity and correlation at the same time,and the previous empirical literature on the air quality is more concentrated in the fixed coefficient spatial panel data model.Dynamic panel data model contains more information than the fixed coefficient model.Therefore,in a scientific point of view,there is of great significance to analysis the economic relation between air quality and economic growth,industrial structure,foreign investment level and urbanization level,and to research spatial characteristics of China’s air quality.And it also plays an important role in the adjustment of regional industrial structure and realization of the coordination between economic development and environmental protection.Besides,it benefits to reduce the pollution of provincial spatial spillover effect and establishment of regional pollution joint-control mechanism.This paper,on the basis of domestic and foreign research experience,attempts to use the traditional panel OLS model,the cross-section model and dynamic panel data model respectively to analyze the causality and spatial dynamic characteristic among exhaust investment,population density,urban green space per capita,per capital GDP.the proportion of the second industry,foreign direct investment,energy consumption per GDP and sulfur dioxide emission(SO2)of 30 provinces of China from 2002 to 2014.The empirical results indicate that there are long-run equilibrium relationships between these variables.There are two main objectives:one is to analysis the provincial differences of air quality;the other is to use both static and dynamic panel estimation model to analysis the spatial correlation and spatial agglomeration of China’s provincial air quality.Firstly,this paper introduces the research background,the domestic and international research status,innovation and deficiency of this paper.Secondly,it briefly introduces the related basic theory of panel model,summarizes several commonly quoted air quality evaluation index,and then introduce the models and indicators selected in this paper.Thirdly,it carries out the empirical analysis of China’s provincial air quality and its influence:firstly,this paper explains the selected variables and data sources,and then focuses on the identification and analysis of the model.Based on unit root test,Granger causality test,co-integration test and Hausman test correspondingly,build the panel model among sulfur dioxide emissions and the effects of governmental investment,energy consumption structure,economic growth,city level,the effects of foreign trade and energy efficiency in use of the linear static panel model and spatial dynamic panel data model respectively.The regression results between SO2 and related influencing factors turn out below,(1)Per capita SO2 emissions of China’s provinces changes with regions.And the per capita SO2 emissions of different regions increase gradually and reduce to the stabilization(2)And apart from the second industries accounted for the proportion of GDP,governance investment,energy efficiency,per capita GDP,foreign direct investment,urban population density,urban garden green area do influence sulfur dioxide emissions because of Granger causality test.In addition,there is a long-term equilibrium relationship among these variables.(3)China’s provincial SO2 emissions Moran’ I index shows SO2 emissions are spatially interacting.There are spatial spillover effects in different provinces,and the lag effect exists in different years.(4)The necessity of the Lagrange multiplier test of the spatial lag effect on per capita SO2 emissions finds that the value of spatial autoregressive test is greater than the spatial error test’s.In hence,choose the spatial autoregressive model to estimate the spatial dynamic effect.It shows that J statistic by applying the generalized method of moments(GMM)model is 71.30,higher than the critical value χ2.The empirical results of the spatial panel data model are statistically insignificant.It shows that the model with spatial lagged factor can better explain the spatial relationship in China’s provincial SO2 emissionsFinally,reasonable suggestions are presented based on the empirical research selected variable indicators.It turns out the air pollutant emissions can be improved by corresponding economic and environmental policies substantially,such as speeding up the restructuring and upgrading of industrial structure,rigorous emission reduction,reasonable introduction of foreign direct investment,to advocate the polices of circular economy and developing new energy and renewable energy industry,and reasonable planning urban green space construction.Thus,to meet the targets of gradually-perfect cooperative prevention and control respond to emissions reduction and achieve sustained and equitable economic growth.Compared with the previous research,this paper adds the explanatory variables,explains the rationality and necessity of introduction of the model variables by Granger causality test,and use Moran I index of air quality in China’s provinces to analysis spatial correlation.Due to the complexity and variability of the air quality in China,and merely use the per capita SO2 emissions to calculate the air quality in China can not fully measure the status of China’s air quality.