For the Yangtze River Delta (YRD) region of China, exploring the spatio-temporal characteristics of carbon emissions from energy consumption (CEECs) and their influencing factors is crucial to achieving carbon peaking and carbon neutrality as soon as possible. In this study, an improved LMDI decomposition model based on the Tapio model and Kaya's equation was proposed. Combined with the improved LMDI and k-means cluster analysis methods, the energy structure, energy intensity, unit industrial output value and population size were selected as the driving factors, and the contribution of each driving factor to the CEECs of prefecture-level cities was quantitatively analyzed. Our study found that: (1) By 2020, the total amount of CEECs in the 26 prefecture-level cities in the YRD will stabilize, while their intensity has shown a downward trend in recent years. (2) The decoupling relationship between CEECs and economic development generally showed a trend from negative decoupling to decoupling. The dominant factor in decoupling was generally the shift of DEL values towards urbanization rate and energy intensity and the open utilization of energy technologies. (3) From 2000 to 2010, the dominant factors affecting CEECs in 26 cities were energy intensity and energy structure, followed by industrial output value and urbanization rate. In general, the promotion effect of economic development on carbon emissions in the YRD region was greater than the inhibitory effect. After 2010, the restrictive effect of various factors on CEECs increased significantly, among which the role of gross industrial output was crucial. The research results can provide a scientific policy basis for the subsequent spatial management and control of carbon emission reduction and carbon neutrality in the YRD region at a finer scale.
Keywords: carbon emissions from energy consumption; decoupling elasticity; improved LMDI model; k-means clustering; map visualization; spatio-temporal characteristics.