Analysis and Prediction of Factors Influencing Carbon Emissions in Guangdong Province Based on STIRPAT Interpretability Analysis
DOI:
https://doi.org/10.54097/2bksv019Keywords:
STIRPAT model, Carbon-emission drivers, Ridge regression, IPSO-BP neural network.Abstract
This study employs an extended STIRPAT model and SHAP interpretability analysis to identify and quantify the key socioeconomic drivers of carbon emissions in Guangdong Province, China. Ridge regression is applied to mitigate multicollinearity among six variables: population, affluence, industrial structure, openness, technology, and energy mix. Results indicate that population size exerts the strongest positive effect (+2.31 % per 1 % increase), followed by industrial structure (+2.1 %), whereas technology (−0.31 %), openness (−0.07 %), and cleaner energy structure (−0.18 %) significantly curb emissions. A high-performing back-propagation neural network optimized by improved particle swarm optimization (IPSO-BP) is then trained to predict future emissions, with SHAP values revealing the marginal contribution of each driver under different policy scenarios. The integrated framework offers transparent, evidence-based guidance for Guangdong to achieve its carbon-peaking and carbon-neutrality targets and provides a replicable template for other rapidly developing regions.
References
[1] Dong Wei, Che Jinxing. Modelling China's Industrial Carbon Emissions Forecast Based on the Kupman Theory [J]. Journal of Nanchang Institute of Engineering, 2025, 44(01): 107-112.
[2] Liu Fengying, Zhong Mingte, Zhu Bifeng. Construction of a Prediction Framework for Carbon Emission Intensity in University Campuses Based on Energy Consumption Drivers [J]. Green Building, 2025, 17(02): 8-14.
[3] Yao Xiaojian, Wu Yuyuan. Analysis of Carbon Emission Influencing Factorsand Scenario Prediction for the Guangdong-Hong Kong-Macao Greater Bay Area [J/OL]. Resources and Industries, 1-17 [2025-04-18]. https://doi.org/10.13776/j.cnki.resourcesindustries.20250217.002.
[4] Guan Minjie, Yuan Yanhong, Ran Muxi, Wang Zi. Influencing Factors and Peak Prediction of Energy Carbon Emissions in Shanxi Province Based on the STIRPAT Model [J]. China Coal, 2021, 47(09): 48-55.
[5] Feng Qisheng, Gao Xinhua. The Role of Excel in Leave-One-Out Cross-Validation Experimental Teaching [J]. Experimental Science and Technology, 2015,13(02): 49-51+65.
[6] Chen Jiaxin, Deng Ao. Analysis of Factors Affecting Carbon Emissions: A Case Study of Guangdong Province [J]. China Environmental Protection Industry, 2025, (07): 59-62+72. DOI: CNKI:SUN:ZHBY.0.2025-07-018.
[7] Kong Xiangbo. Analysis of Operational Efficiency and Influencing Factors of China's Coastal Ports Considering Carbon Emissions [J]. Journal of Inner Mongolia University for Nationalities (Natural Science Edition), 2025, 40(04): 62-69. DOI: 10.14045/j.cnki.15-1220.2025.04.010.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Yiting Qiu, Jingmin Lan, Xiaolu Zou, Yun Huang, Ying Tan, Dang Gan, Jiaxin Huang, Jiaqi Zhang, Zhihan Wang, Junpeng Yuan, Hang Zeng, Yiling Zhang

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.







