A study on spatial clustering of carbon emissions in 265 cities based on regional differences and its influencing mechanisms

Carbon emissions (CE) are a significant cause of global climate change and environmental degradation, with far-reaching impacts on human society and the natural environment. To scientifically formulate targeted carbon reduction strategies, this study takes 265 cities as the research object, and a regional classification framework combining factor analysis, k-means clustering, and CE was used to classify 265 city rows in a secondary way. Based on the Stochastic Impacts by Regression on Population, Affluence, and Technology (STIRPAT) expansion model, we employ the ridge regression analysis method to investigate the mechanism of the influence of factors such as population size, economic scale on CE in various regions. The results indicate that, First, under the regional classification framework, the 265 cities are divided into 16 types of regions, and the cities generally have low levels of energy use. Second, population size, level of economic development, urbanization rate, and carbon productivity are crucial factors that significantly impact CE in each region. Third, energy consumption, level of electricity use, and level of labor force positively affect CE in most regions and to a lesser extent. Finally, the study's findings serve as the foundation for formulating targeted policy recommendations.