Under the green and low-carbon development goal of "carbon peaking and carbon neutrality" in China, cyclical analysis and accurate prediction of carbon emissions have important practical implications. This paper takes the carbon emissions in Zhejiang Province as the research object, variational modal decomposition method is introduced to extract multi-scale information, LASSO algorithm and GM (1, N) model is combined for carbon emissions forecasting in Zhejiang Province. Firstly, variable mode decomposition method is used to decompose the carbon emissions in Zhejiang Province and analyze the cyclicality fluctuations of its historical data. Secondly, the key influencing factors of carbon emissions is selected by LASSO algorithm. Finally, according to the 14th Five-Year Plan with the actual development path, assuming normal, low-carbon, and inertia development scenarios, the carbon emissions of Zhejiang Province from 2020 to 2030 is predicted based on the CM (1, N) model, overcoming the limitations of traditional prediction models in handling non-linear and small sample data. The results indicate that the proportion of the third industry in GDP, the number of private cars, the total fixed asset investment in the province , the total electricity consumption, R&D intensity , and technology market turnover are the dominant factors of carbon emissions in Zhejiang Province. The carbon emissions under the low-carbon development scenario are expected to peak at 400.28 Mt in 2030, while the carbon emissions under the normal development scenario are 474.23 Mt and the inertia development scenario are 568.77 Mt. Moreover, the carbon peak under the normal and inertia development scenarios is expected to increase after 2030. It is suggested that Zhejiang Province should focus on optimizing its industrial structure, improving energy efficiency, increasing investment in low-carbon research and development, and steadily promote the goal of "carbon peaking".
Close-HONG Jingke,DU Wei,SHAO Jin,LAO Huimin.Carbon emission forecasting in Zhejiang Province based on LASSO algorithm and grey model[J/OL].Energy Environmental Protection:1-10[2024-01-04].https://doi.org/10.20078/j.eep.20240101.