China's energy trilemma: Spatial patterns, evolutionary trends at the city level

This study focuses on China's energy trilemma (ET) and uses ET data from 276 Chinese cities spanning from 2008 to 2021. A comprehensive analysis is conducted on regional disparities, evolutionary trends, spatial patterns, and influencing factors, employing methods such as kernel density estimation, spatial Markov chains, Dagum Gini coefficient decomposition, random forest algorithms, and partial dependence plots (PDP). The results show significant differences in the ET development trends across regions in China. The Yangtze River Delta and the Northeast provinces are relatively stable, while the Guangdong-Hong Kong-Macao area, the Beijing-Tianjin-Hebei region, and the Chengdu-Chongqing region experience more pronounced fluctuations. The spatial Markov chain analysis indicates a "club convergence" phenomenon among cities with high ET indices, but an "exclusion effect" also exists in some areas. The random forest model reveals that the average temperature, population growth, and foreign investment play crucial roles in shaping ET outcomes, and there are threshold effects for fiscal support for research and development and international oil prices. These insights underscore the necessity for tailored regional energy policies and enhanced intercity cooperation to achieve a balanced ET across China.