The spatio-temporal evolution of provincial carbon emissions in China based on machine learning

With the increasingly severe global climate change issue, controlling carbon emissions has become a global focus. Taking the provincial carbon emission dataset of China as the research object, and combining the STIRPAT model with machine learning methods, this paper analyzes the spatio-temporal evolution characteristics and driving factors of carbon emissions in China. The research results show that there are significant regional differences in carbon emissions in different regions of China, with the eastern and northern regions having significantly higher carbon emissions than the southern and western regions. At the same time, there is an inverted U-shaped relationship between carbon emissions and economic growth, verifying the EKC hypothesis. The carbon emission prediction results using machine learning models show that if the current policy intensity is maintained, China’s carbon emissions will tend to stabilize and gradually decline in the coming years. In addition, improving energy conversion efficiency and promoting low-carbon technologies play an important role in carbon reduction. This paper provides a reference basis for the formulation of regional carbon reduction policies and low-carbon transformation in China.