China's carbon emissions have developed swiftly in recent decades, which will not only affect the nation's own sustainable development, but have a potentially negative impact on global climate stability. Given that socioeconomic development is susceptible to regional heterogeneity and geographic scales, a systematic exploration of spatiotemporal variations of carbon emission intensity (CEI) and their drivers across different levels is conducive to enacting more reasonable and efficient measures for emission reduction. However, there is still a lack of comprehensive analysis of these issues. In this paper, we attempted to quantify and compare the spatiotemporal evolution and spatial spillover effects of impact factors on CEI from nighttime light imagery and socioeconomic data at two China's administrative levels by utilizing the variation coefficient, spatial autocorrelation model and spatial econometric methods. The results showed that the spatiotemporal variations of CEI were greater at the prefecture level compared to the provincial level during 2000-2017. There were significant positive spatial autocorrelation of CEI at two administrative levels, and self-reinforcing agglomeration was more substantial at the prefectural level than that provincial level. While the local spatial clustering of CEI of each administrative level altered with scale dependence, the binary spatial structure (High-High and Low-Low) of CEI remained relatively steady in China. Various driver factors not only had direct effects on local CEI, but had spatial spillover effects on neighboring areas. Our findings illustrate that China's CEI is sensitive to the space-time hierarchy of multi-mechanisms, and suggest that "proceed in the light of local conditions" strategies can assist the Chinese government for CEI mitigation.
Keywords: Carbon emission intensity; Multiscale perspective; Spatial autocorrelation; Spatial econometric method; Spatial spillover effect.
Copyright © 2022 Elsevier B.V. All rights reserved.