The estimation of the carbon dioxide emission and driving factors in China based on machine learning methods

Global warming can be reduced and the ecological environment can be enhanced by reducing carbon dioxide emissions. Therefore, it is imperative to determine how to calculate urban carbon dioxide emissions. Besides, as the country with the largest carbon dioxide emissions, exploring the influencing factors of carbon dioxide emissions is conducive to providing support for emission reduction actions. In the first place, the paper makes use of China's provincial carbon dioxide emissions data and nighttime light data to build an inversion model from 2000 to 2019 that calculates carbon emission data for prefecture-level cities. Furthermore, this paper employs machine learning methods such as decision tree and random forest to determine the factors affecting carbon dioxide emissions. The main conclusion is that carbon dioxide emissions are highest in the eastern regions with higher economic development. Additionally, cities which are dependent on resources to develop have higher carbon dioxide emissions and a rising trend. Factors such as gross domestic production, financial general budget revenue and foreign investment can influence carbon dioxide emissions. According to the random forest results, the feature importance of GDP, financial general budget revenue and foreign investment is 0.45, 0.12 and 0.08, respectively. Accordingly, different regions cannot ignore carbon dioxide emissions when developing their economies. As the growth rate of emissions has slowed in recent years in part due to government policies, China's ongoing implementation of low-carbon transformation must continue to be implemented, including low-carbon city pilot programs, carbon trading markets, etc. As well, areas with serious carbon dioxide emissions, such as Shanghai, Tianjin, and Chongqing, should be prioritized as areas for low-carbon economic development.