Multi-scenario prediction of carbon emissions in China's six regions: Driving factor investigation and decoupling state analysis

Against the backdrop of global climate change and China's dual-carbon goals, identifying regional carbon-emission (CE) patterns is critical for implementing targeted emission-reduction governance. This study takes 30 provincial-level administrative units in China, which are divided into six geographic regions, and sets six influencing factors to investigate CE characteristics during 2000–2023. The LMDI model is adopted to decompose six driving factors of CE, and the Tapio decoupling model is combined to evaluate the decoupling relationship between CE and economic growth. The STIRPAT-ridge regression model is constructed, and three policy scenarios including baseline (S1), intensified (S2), and low-carbon (S3) are established to predict CE trajectories for 2024–2038 under corresponding scenario assumptions. The results show that economic growth and industrial development are the major positive drivers of CE, while energy intensity, population efficiency and urbanization-related factors exert significant inhibitory effects, with distinct heterogeneity across the six regions. Weak decoupling accounts for more than 50% of all observed decoupling samples in each region, and the six regions shift toward a coupling state as a whole in 2023. Under given scenario assumptions, compared with the S1 baseline scenario, the average CE of Northwest China decreases by 5.41% under the S2 scenario and by 10.41% under the S3 scenario. Jilin and Zhejiang achieve provincial-level carbon peaks in 2023 under scenario assumptions; five provinces reach provincial-level carbon peaks before 2024 under the S3 low-carbon scenario.