Discerning drivers and future reduction paths of energy-related CO2 emissions in China: combining EKC with three-layer LMDI

Recognizing the process and identifying the drivers of energy-related CO2 emissions can provide suggestions for designing carbon emission reduction paths, and then promote the further reduction of carbon emission. In this paper, the carbon emissions in China and its subordinate provinces during 2005–2017 are firstly divided into four stages, named S1, S2, S3, and S4. The results show that China has just entered the S3, and it is impossible to reach the peak of energy-related CO2 emissions with steady economic growth before 2030. Then, three-layer LMDI is utilized to explore the drivers of CO2 emissions, and the impact of urbanization which is separated from the population is considered innovatively. The economic development increases CO2 emissions, while the other drivers have diverse effects, which may be positive or negative, on carbon emissions in different regions. Therefore, four emission reduction paths with provincial characteristics should be followed in the future: (i) three provinces, namely, Ningxia, Shaanxi, and Xinjiang, should optimize multiple basic objectives in parallel; (ii) four provinces, such as Inner Mongolia and Hainan, should optimize the energy structure; (iii) six provinces, such as Jiangxi and Hunan, should optimize the industry structure; and (iv) the other provinces should develop new clean energy according to regional conditions.