China's carbon emission prediction from the perspective of shared socioeconomic pathways and machine learning

The adverse effects of carbon emissions have garnered significant attention from the international community. Against this backdrop, China has set the dual carbon target and taken concrete actions to achieve it. This study integrates Shared Socioeconomic Pathways with machine-learning methodologies to project China's carbon emissions from 2021 to 2060 under different scenarios. In addition, sensitivity analysis is used to test the robustness of prediction result and analyze the importance of different influencing factors under the optimal carbon emission scenario. The results indicate that: (1) Under SSP1, SSP2, and SSP4 scenarios, carbon emissions peak in 2025 at 10.625, 10.937, and 10.761 billion tons, respectively. For SSP3 and SSP5, the peaks occur around 2030, reaching 11.466 and 12.220 billion tons, respectively. (2) Under SSP1 and SSP4 scenarios, China can attain carbon neutrality as planned. However, under the SSP2 scenario, the achievement of carbon neutrality hinges on the development and application of negative carbon emission technologies. (3) Under the SSP1 scenario, China exhibits the lowest per capita carbon emission peak and cumulative carbon emissions, at 7.36 tons and 210.052 billion tons, respectively. (4) Among the scenarios, SSP1 emerges as the optimal scenario considering carbon peak, carbon neutrality, per capita carbon emissions and cumulative carbon emissions. (5) Under a 10 % fluctuation in influencing factors, the prediction results under SSP1 remain robust. Energy system reform and carbon-negative technologies are essential for China under it. Based on the above analysis, this study proposes the following mainly three aspects to promote the achievement of China's dual carbon goals: improving the efficiency of fossil energy utilization, facilitating the expansion of clean energy generation and the corresponding transformation of the power system, and accelerating innovation and application of negative emissions technologies. • Select five key influencing factors to predict China's carbon emissions under the SSP scenarios. • All scenarios can achieve carbon peak as planned. • SSP1, SSP2 and SSP4 can achieve carbon neutrality. SSP2 is more difficult and relies on negative carbon emission technology. • SSP1 emerges as the most favorable carbon emission scenario. • Transformation of energy and power system and progress of low-carbon technology is the key to achieve China’s “3060” targets.