Analysis On Spatiotemporal Characteristics And Driving Factors Of Air Pollution In China Based On Interprovincial Embodied Transfer Measurement

With the rapid development of economy and continuous acceleration of urbanization,the air pollution in China has entered a regional composite stage of multi-source impact,multi-regional correlation,and multi-process evolution,which seriously affects the health of residents as well as poses great challenges to environmental governance.On the one hand,there exist imbalances in the status quo of economic development and pollution control intensity in different regions,where the pollution sources and emission causes vary in diverse respects.On the other hand,regional trade may lead to a spatial dislocation between production and consumption as the consequent embodied transfer mechanism decouples environmental costs from economic benefits.Therefore,clarifying the current status of atmospheric pollution,quantifying the dynamic paths of embodied transfer,examining the characteristics of spatiotemporal evolution,and analyzing the spatial effects of driving factors all serve as the essential basis for scientific formulation in regional collaborative governance measures.According to the summarization of relevant literature,this study refines the connotation on embodied pollution transfer and driving factor indicators so as to explore inter-provincial linkage paths and spatial interaction effects of air pollution.First,the consumption-based transfer of embodied air pollution across provinces in China in 2012 is quantified with the multi-regional input-output model.The revised direct exhaust emissions are compared under the regional and industrial caliber to reveal static volume distribution and dynamic transfer paths.Then,the inverse-distance based spatial weight matrix and the geo-social nested weight matrix,which is combined with the embodied pollution transfer in inter-provincial trade,are both constructed.Based on the panel data of 30 provinces,municipalities and autonomous regions in China from 2006 to 2017,the spatiotemporal evolution of atmospheric pollution equivalents is investigated through descriptive statistics and spatial autocorrelation analysis.After a series of tests and comparison,the indicators and spatial lags are included to estimate the static Spatial Durbin Model with individual fixed effect,which is applied to examine the direction and degree of influence on air pollution emissions in the certain region and neighboring areas.Derived from the conclusions,specific suggestions are initiated on regional coordination management and pollution symptomatic governance.The main conclusions are summarized as follows.(1)The overall performance is characterized by a net inflow of atmospheric pollution pressure.The spatial distribution of embodied net flow and direct exhaust emissions is generally consistent,where the quantitative amount presents a trend of polarization.Specifically,the net outflow areas are comparatively uniform and dispersed,while the net inflow areas possess local clustering characteristics.(2)The main transfer paths are from east to west and south to north,of which the areas in North Coast and Middle Yellow River(Hebei,Shanxi,Inner Mongolia and Henan etc.)are the primary inflow concentration and East Coast(Zhejiang,Jiangsu,and Shanghai)outputs pollution pressure as the main force.(3)The major industries have both contribution commonality and regional differences.The proportion of industry contribution basically conforms to the Pareto Principle,and different resource endowment may render comparative advantages and thus distinct distribution.(4)About the temporal trend from 2006 to 2017,nationwide atmospheric pollution equivalents start with a steady decrease followed by the sudden rebounding in 2011 and then a rapid drop.As for the overall positive spatial autocorrelation,the emissions in the east and north are generally heavier than those in the west and south.Moreover,the inflection points in pollution equivalents and Moran’s I index basically correspond to the period nodes of China’s Five-Year Plan.(5)The geo-social nested weight matrix tends to illustrate a better interpretation of spatial interaction effects.In view of estimation results,economic development,industrial structure,R&D intensity,and unit exhaust investment are conducive to reduce equivalent emissions inside and outside the certain region,whereas urbanization and population impose adverse influence.Besides,energy consumption has positive direct effects and negative indirect effects.Public transportation may exacerbate local pollution.Furthermore,the impacts of foreign direct investment and exhaust treatment investment are not significant.This study attempts to optimize the research process from the following three aspects of innovation.Firstly,the selection of indicators focuses on both representativeness and comprehensiveness,such as weighted conversion of major air pollutants into atmospheric pollution equivalents to carry out spatial analysis.Secondly,for model application,the results of embodied pollution measure are employed to construct a nested weight matrix,which is explored based on dual perspectives of regional and industrial comparison as well as dual dimensions of static and dynamic analysis.Thirdly,in terms of disciplinary frameworks,the theoretical basis contributes to a reasonable discussion of research conclusions,including Pareto Principle,Comparative Advantage Theory,Heckscher-Ohiln Theory,Ecological Compensation Mechanism,and classification criteria on eight comprehensive economic regions.In addition,there are certain limitations in this study.For one thing,subject to the update progress of inter-regional input-output tables,the corresponding data for 2012 as the latest release may lead to a slight lack of timeliness in the embodied transfer measurement.For another,the selection of spatial weight matrix fails to take into account the endogenous character and the ability to adjust according to social development trends,where the dynamic transmission mechanism of spatial effects may be ignored.