Spatially and temporally varying relationships between ecological footprint and influencing factors in China's provinces Using Geographically Weighted Regression (GWR)

While China enjoys rapid social and economic development, serious ecological and environmental issues have been brought about as well. Furthermore, the cultural, social, economic situation and natural endowments in different provinces in China are varied from one to another, hence, the factors accounting for these ecological environmental issues are also spatially varied. In the perspective of spatial heterogeneity, this paper analysed the spatial variation between ecological footprint (EF) evolution and its influencing factors in the year of 2004 and 2012 in China's 30 provinces. Through regression model comparison, population size, affluence level and technology level were selected as independent variables. Furthermore, regression results comparison was made between GWR and OLS models and it showed that GWR model was superior to OLS in terms of regression goodness of fit, variance comparisons as well as the spatial auto-correlation of residual. Specifically, according to GWR results, population size and affluence level were the main driving forces of EF evolution whereas technological advancement could effectively restrain EF growth. Meanwhile, being spatially and temporally varied, the local coefficients also had their spatial distribution characteristics, namely partial coefficients assuming spatial gradient and spatial agglomeration. Most of the local coefficients of population size and affluence level were increased from 2004 to 2012, and the mean values increased from 0.852 to 0.929 and from 0.457 to 0.571 respectively. Among the three influential factors, technology played an increasingly prominent role. Its scope of influence ascended from 53.33% to 86.67% and the mean value of its local coefficients decreased from -0.113 to -0.178. Finally, the paper provided some suggestions from population, economy, technology and the regional cooperation for China to formulate differentiated ecological and environmental policy based on the features of EF spatial distribution as well as the temporal and spatial differences analysis of its influential factors. (C) 2020 Elsevier Ltd. All rights reserved.