Impact of PM 2.5 Pollution on Solar Photovoltaic Power Generation in Hebei Province, China

Ankun Hu, Zexia Duan (), Yichi Zhang, Zifan Huang, Tianbo Ji and Xuanhua Yin
Additional contact information
Ankun Hu: School of Electrical Engineering and Automation, Nantong University, Nantong 226019, China
Zexia Duan: School of Electrical Engineering and Automation, Nantong University, Nantong 226019, China
Yichi Zhang: Hangzhou Qiantang District Bureau of Meteorology, Hangzhou 311225, China
Zifan Huang: School of Electrical Engineering and Automation, Nantong University, Nantong 226019, China
Tianbo Ji: School of Transportation and Civil Engineering, Nantong University, Nantong 226019, China
Xuanhua Yin: School of Computer Science, University of Sydney, Sydney, NSW 2008, Australia

Energies, 2025, vol. 18, issue 15, 1-26

Abstract: Atmospheric aerosols significantly impact solar photovoltaic (PV) energy generation through their effects on surface solar radiation. This study quantifies the impact of PM 2.5 pollution on PV power output using observational data from 10 stations across Hebei Province, China (2018–2019). Our analysis reveals that elevated PM 2.5 concentrations substantially attenuate solar irradiance, resulting in PV power losses reaching up to a 48.2% reduction in PV power output during severe pollution episodes. To capture these complex aerosol–radiation–PV interactions, we developed and compared the following six machine learning models: Support Vector Regression, Random Forest, Decision Tree, K-Nearest Neighbors, AdaBoost, and Backpropagation Neural Network. The inclusion of PM 2.5 as a predictor variable systematically enhanced model performance across all algorithms. To further optimize prediction accuracy, we implemented a stacking ensemble framework that integrates multiple base learners through meta-learning. The optimal stacking configuration achieved superior performance (MAE = 0.479 MW, indicating an average prediction error of 479 kilowatts; R 2 = 0.967, reflecting that 96.7% of the variance in power output is explained by the model), demonstrating robust predictive capability under diverse atmospheric conditions. These findings underscore the importance of aerosol–radiation interactions in PV forecasting and provide crucial insights for grid management in pollution-affected regions.

Keywords: PM 2.5 pollution; photovoltaic power output; machine learning; stacking framework (search for similar items in EconPapers)
JEL-codes: Q Q0 Q4 Q40 Q41 Q42 Q43 Q47 Q48 Q49 (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/1996-1073/18/15/4195/pdf (application/pdf)
https://www.mdpi.com/1996-1073/18/15/4195/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jeners:v:18:y:2025:i:15:p:4195-:d:1719566

Access Statistics for this article

Energies is currently edited by Ms. Agatha Cao

More articles in Energies from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().