Apportioning firework burning and routine source contributions to PM2.5 in a Chinese megacity using two PMF–machine learning frameworks
Environmental Research · 2026-11 · 126121
DOI: 10.1016/j.envres.2026.126121 ↗Kurz gefasst
Fachartikel in der Zeitschrift Environmental Research, 2026-11. Verfasst von Jiang-Yao Tan und 8 weiteren Autor:innen. Beteiligte Einrichtung: Shenzhen Key Laboratory for Urban Habitat Environmental Science and Technology, School of Environment and Energy, Peking University Shenzhen Graduate…. Die Arbeit nennt 67 Literaturangaben. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
The firework burning (FB) during the Chinese Spring Festival can drastically elevate PM2.5 concentrations and rapidly deteriorate local air quality, highlighting the need for timely quantification of its contributions to PM2.5. However, as a short-term episodic source, FB poses substantial challenges to the accurate and rapid quantification of its contributions to PM2.5. In this study, two source apportionment frameworks integrating the Positive Matrix Factorization (PMF) model with machine learning (ML) were developed to quantify the contributions of FB and routine sources to PM2.5. The frameworks were applied to hourly measurements data of PM2.5 chemical constituents collected in Shenzhen, China, from January to March during 2023-2025. Twelve sources were identified, including 11 routine sources (e.g., secondary sulfate, secondary nitrate, vehicle emissions) and an episodic source of FB. Although the FB contributed only 4.7% to the total PM2.5 mass over the entire study period, its share surged to 33% during firework display episodes and even exceeded 50% during peak burning hours. Two machine-learning-based approaches, namely the direct and indirect source apportionment methods, showed good fidelity to the PMF-resolved results and consistently identified FB as the dominant PM2.5 source during firework display episodes, with estimated contributions comparable to those from the PMF model. The proposed PMF-ML framework captured episodic pollution signals while quantifying routine sources, with its ML component serving as a surrogate alternative to PMF.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Shenzhen Key Laboratory for Urban Habitat Environmental Science and Technology, School of Environment and Energy, Peking University Shenzhen Graduate School, Shenzhen, 518055, China.
- Shenzhen Key Laboratory for Urban Habitat Environmental Science and Technology, School of Environment and Energy, Peking University Shenzhen Graduate School, Shenzhen, 518055, China. Electronic address: pengxing@pku.edu…
- Department of Mechanical and Manufacturing Engineering, University of Calgary, Calgary, Canada.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Tan, J. Y., Peng, X., Zhang, Z. Y., Li, J. H., Yu, W. J., Liu, Y. F., Huang, X. F., Du, K., & He, L. Y. (2026). Apportioning firework burning and routine source contributions to PM2.5 in a Chinese megacity using two PMF–machine learning frameworks. Environmental Research, 126121. https://doi.org/10.1016/j.envres.2026.126121
Vancouver
Tan JY, Peng X, Zhang ZY, Li JH, Yu WJ, Liu YF, et al. Apportioning firework burning and routine source contributions to PM2.5 in a Chinese megacity using two PMF–machine learning frameworks. Environmental Research. 2026:126121. doi:10.1016/j.envres.2026.126121
BibTeX
@article{Tan2026,
author = {Tan, Jiang-Yao and Peng, Xing and Zhang, Zi-Yi and Li, Jun-Hong and Yu, Wen-Jie and Liu, Yu-Fei and Huang, Xiao-Feng and Du, Ke and He, Ling-Yan},
title = {Apportioning firework burning and routine source contributions to PM2.5 in a Chinese megacity using two PMF–machine learning frameworks},
journal = {Environmental Research},
year = {2026},
pages = {126121},
doi = {10.1016/j.envres.2026.126121}
}
EUVIMED