Spatio-temporal dynamics and influencing factors of carbon emissions (1997–2019) at county level in mainland China based on DMSP-OLS and NPP-VIIRS Nighttime Light Datasets

Abstract

Global warming caused by extensive carbon emissions is a critical global issue. However, the lack of county-level carbon emissions data in China hampers comprehensive research. To bridge this gap, we employ a deep learning method on nighttime light data sets to estimate county-level carbon emissions in mainland China from 1997 to 2019. Our key contributions include the successful derivation of more reliable data, revealing the evolution of spatial dynamics and emissions epicenters. Moreover, we identify a novel inverted N-shaped relationship between gross domestic product per capita and carbon emissions in the eastern and western regions, as well as an N-shaped relationship in the central region, challenging mainstream wisdom. Additionally, we highlight the significant impacts of population density, industrial structure, and carbon intensity on carbon emissions. Our study also unveils the nuanced effects of government spending, which exhibits both inhibitory and region-specific influences. These findings serve to enhance our understanding of the factors influencing carbon emissions and contribute to informed decision-making in addressing climate change-related challenges.

Keywords

• Nighttime light data
• Deep learning method
• Carbon emissions
• Heterogeneity
• County-level
• Influencing factors