[Carbon Peak Prediction in Fujian Province Based on Combined STIRPAT and CNN-LSTM Models]

Carbon peaking is of great significance for China to achieve the "dual carbon" target goal and promote the green transformation of the economy and society. Based on the improved STIRPAT model, to analyze the main factors affecting carbon emissions in Fujian Province, we set up three scenarios and predicted the carbon emissions in Fujian Province from 2022 to 2035 using the hybrid CNN-LSTM neural network model. The results showed that ① Population, GDP per capita, and industrial structure positively drove carbon emissions in Fujian Province, while energy intensity, energy structure, and foreign trade degree negatively drove them. ② The baseline scenario achieved peak carbon in 2033 with a peak value of 361.107 9 Mt； the low-carbon scenario and the optimization scenario could reach the peak one year earlier, and the peak value decreased to a different extent, with 333.028 4 Mt and 301.748 3 Mt, respectively. ③ Comparing the optimization scenario and the low carbon scenario, adjusting the industrial and energy structure could control the reduction in the peak carbon value in Fujian Province by 10.37%, and accelerating the promotion of the optimization and transformation of the energy and industrial structure is the key to unlocking the constraints between carbon emissions and economic development. Finally, combined with the current policy planning and development status of Fujian Province, we put forward suggestions for low-carbon development from the perspectives of energy emission reduction, industrial structure, and institutional system.