Spatial-temporal dynamics and influencing factors of city level carbon emission of mainland China

Abstract

Urban areas are major sources of carbon emissions, making it crucial to understand their emission characteristics for effective carbon reduction and sustainable development. Using carbon emission data from mainland China (2001–2021), we analyzed the spatio-temporal dynamics and future trends of city-level emissions and explored influencing factors using machine learning methods. Results indicate significant fluctuations in carbon emissions, with over 40 % of mainland Chinese cities experiencing a doubling in total emissions. Geospatially, cities in South Coast (SC), East Coast (EC), Northeast (NE), and NorthCoast (NC) show stronger intensity and increasing trends in carbon emissions compared to other regions, with over 80 % of cities in these regions experiencing high or higher increases. Additionally, a continued rise in carbon emissions was detected in most Chinese cities, with an average Hurst index of 0.64, indicating persistent trends. Using the XGBoost method, factors such as population density, built-up area, urban green coverage rate, and GDP were found to strongly correlate with urban carbon emissions, exhibiting significant spatial heterogeneity. This research uncovers the characteristics and influencing factors of urban-scale carbon emissions, offering valuable insights for policymakers to tailor carbon reduction strategies to the specific needs and conditions of various urban areas.

Keywords

• Carbon emission
• Spatial-temporal dynamics
• Influencing factors
• XGBoost
• Chinese cities