Analysis of the Driving Mechanism of China’s Provincial Carbon Emission Spatial Correlation Network: Based on the Dual Perspectives of Dynamic Evolution and Static Formation

Against the backdrop of China’s commitment to achieving carbon peaking by 2030 and carbon neutrality by 2060, inter-provincial carbon emissions form a complex interconnected spatial network—clarifying its operational mechanisms is crucial for optimizing regional carbon reduction strategies. Based on 2006–2021 data from 30 Chinese provinces, this study constructs the China Provincial Carbon Emission Spatial Correlation Network (CPCESCN) using a modified gravity model. Social Network Analysis (SNA) explores its structural characteristics, while motif and QAP correlation analyses identify endogenous structural and attribute variables. Innovatively integrating Exponential Random Graph Models (ERGM) and Stochastic Actor-Oriented Models (SAOM), it investigates the network’s static formation mechanisms and dynamic evolution drivers. Results show CPCESCN has a stable multi-threaded structure without isolated nodes, with Jiangsu, Guangdong, Shandong, Zhejiang, Henan, and Sichuan as high-centrality core nodes with high centrality. GDP, green technology innovation, urbanization rate, industrialization rate, energy consumption intensity, and environmental regulations significantly influence network dynamics, with reciprocal relationships as key endogenous drivers. While geographic proximity still facilitates network formation, its impact has weakened notably, and functional complementarity has become the dominant evolutionary driver—based on the findings, policy suggestions are proposed, including deepening inter-provincial functional cooperation, implementing differentiated carbon reduction policies, and optimizing multi-dimensional low-carbon transformation systems.