Spatio-temporal dynamics of per capita energy consumption carbon emissions in the central Yunnan urban agglomeration based on the ESTDA model

Abstract

China is currently the world’s largest emitter of carbon dioxide and also one of the countries making the greatest efforts to reduce emissions. The Central Yunnan Urban Agglomeration, located in southwest China, sits at the geometric center connecting China with South and Southeast Asia. Positioned at the convergence of the Belt and Road Initiative and the Yangtze River Economic Belt, it represents a typical plateau-based, ecologically livable urban cluster. Anthropogenic emissions at the county administrative level are crucial for achieving carbon neutrality goals, as reduction targets can be effectively decomposed to subnational units. However, existing research has primarily focused on the provincial or national level, with limited studies examining the spatiotemporal interaction characteristics of carbon emissions at the county level. This paper examines the Central Yunnan Urban Agglomeration, employing Exploratory Spatio-Temporal Data Analysis (ESTDA) and Tapio spatial econometric methods. Based on a remote sensing image inversion dataset, it quantifies the spatio-temporal dynamics of county-level carbon emissions within the agglomeration from 2006 to 2021, along with the decoupling of emissions from economic growth during this 15-year period. Spatio-temporal interaction patterns of per capita carbon emissions across counties were analyzed using LISA metrics (path length, curvature, mean activity direction), spatiotemporal transition matrices (transition probabilities, transition types, transition indices), and spatiotemporal network graphs. Results indicate that per capita energy consumption carbon emissions in counties within the Central Yunnan Urban Agglomeration exhibit strong spatial clustering stability and path dependency characteristics. From 2006 to 2021, Type IV transitions (self-sustaining transitions where neither the region itself nor its adjacent units undergo spatial association type changes) dominated, accounting for 65.31%. This phenomenon may be linked to the rigidity of local energy consumption structures and the slow pace of industrial restructuring. However, the proportion of such transitions has shown a declining trend in recent years. By constructing a synergy index based on the LISA time-path covariance correlation coefficient of per capita carbon emissions in adjacent counties and visualizing it through the LISA spatiotemporal network, it was found that the region predominantly exhibited positive correlations (synergistic growth) from 2006 to 2021, with a pronounced trend of synergistic evolution. This formed a weak synergistic development network centered on Chenggong County, reflecting the core county’s significant radiating and driving role in the regional low-carbon synergy process. Furthermore, this study identifies four decoupling states between per capita carbon emissions and per capita GDP: weak decoupling, strong decoupling, negative growth decoupling, and strong negative decoupling. Among these, weak decoupling is the most prevalent state, indicating that economic growth in most counties still relies to some extent on increased carbon emissions, and the low-carbon transition process requires further deepening.