Spatio-temporal dynamics of human-induced carbon emissions in Southeast Asia (1992-2022) based on nighttime light.

Understanding regional carbon emissions from human activities, particularly their spatio-temporal patterns, is essential for implementing decarbonization strategies and cultivating a low-carbon economy. This study develops a spatial visualization model to estimate carbon emissions in Southeast Asia using calibrated nighttime light data, with DMSP-OLS (Defense Meteorological Satellite Program Operational Linescan System) and NPP-VIIRS (National Polar-orbiting Partnership Visible Infrared Imaging Radiometer Suite) standardized through polynomial regression and machine learning to ensure consistency. Emissions in Southeast Asia increased by 2.51 times from 1992 to 2022, shifting from gradual to rapid growth. Validation against Open-source Data Inventory for Anthropogenic CO (ODIAC) and Emissions Database for Global Atmospheric Research (EDGAR) shows strong agreement in high-emission urban areas but discrepancies in low-emission rural regions due to data sparsity and satellite sensor limits. Spatial analysis reveals that major Southeast Asian cities and their peripheries exhibit robust, sustained growth, while rural, less-developed areas show slower trends, highlighting persistent urban-rural disparities. These urban regions demonstrate a "circular economy advantage", where per-unit-area carbon emissions steadily rise in economically advantageous zones. Despite high model accuracy, uncertainties persist due to variations in regional economic activities and the limitations of satellite-based emission proxies. Forecasts suggest elevated emission levels in major cities will continue, while changes in other areas remain relatively minimal. Consequently, achieving a low-carbon economy in Southeast Asia requires a top-down approach, emphasizing infrastructure enhancement, resource and energy optimization, and fostering a sustainable, circular socio-economic system.