Leveraging machine learning and ten-cities observations to disentangle emission and meteorological controls on secondary inorganic aerosols in China
Environmental Pollution · 2026
DOI: 10.1016/j.envpol.2026.128190 ↗Kurz gefasst
Fachartikel in der Zeitschrift Environmental Pollution, 2026. Verfasst von Hao Xiao und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, 200240, China; Shanghai Yangtze River Delta Eco-Environmental Change and …. Hauptthemen laut MEDLINE: Machine Learning, Air Pollutants, Environmental Monitoring, Air Pollution. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 71 Literaturangaben. Gefördert von: China Postdoctoral Science Foundation, National Natural Science Foundation of China.
Abstract (Originalsprache)
Secondary inorganic aerosols ((NH4)2SO4 and NH4NO3) play a crucial role in air pollution and climate change, yet the differences in the key drivers and dominant mechanisms governing their formation under China's complex regional backgrounds remain unclear. Here, we conducted synchronous winter PM2.5 observations across ten representative Chinese cities. By integrating machine learning methods, we systematically analyzed the spatial patterns of (NH4)2SO4 and NH4NO3 pollution and identified important predictors associated with their concentrations. This study reveals the differential formation mechanisms of (NH4)2SO4 and NH4NO3 in China: In non-heating areas, persistently high relative humidity significantly promotes the formation of NH4NO3 through aqueous-phase processes; whereas in cities relying on coal combustion heating, elevated SO2 concentrations not only exacerbate (NH4)2SO4 pollution but also consume ammonia, thereby suppressing NH4NO3 formation in some "NH3-poor" cities. Random forest models were developed to assess the associations between (NH4)2SO4 and NH4NO3 with gaseous pollutants and meteorological parameters, respectively. The analysis revealed that the marked increase in concentrations of both species during the transition from clean to haze conditions is primarily associated with atmospheric stagnation and aqueous-phase chemical reactions under high humidity. Sensitivity simulations further indicate that if precursor emissions in non-heating regions were comparable to those in heating regions, the former would likely experience more severe air pollution; conversely, an increase in humidity in heating regions would more readily trigger more severe pollution episodes. These findings underscore the necessity for region-specific air pollution control strategies.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, 200240, China; Shanghai Yangtze River Delta Eco-Environmental Change and Management Observation and Research Station, Ministry of Science and T…
- School of Agriculture and Biology, Shanghai Jiao Tong University, Shanghai, 200240, China. Electronic address: yinmeiju@sjtu.edu.cn.
Förderung
- China Postdoctoral Science Foundation
- National Natural Science Foundation of China
Zitieren
APA
Xiao, H., Yin, M. J., Xiao, H. W., Xu, Y., & Xiao, H. Y. (2026). Leveraging machine learning and ten-cities observations to disentangle emission and meteorological controls on secondary inorganic aerosols in China. Environmental Pollution. https://doi.org/10.1016/j.envpol.2026.128190
Vancouver
Xiao H, Yin MJ, Xiao HW, Xu Y, Xiao HY. Leveraging machine learning and ten-cities observations to disentangle emission and meteorological controls on secondary inorganic aerosols in China. Environmental Pollution. 2026. doi:10.1016/j.envpol.2026.128190
BibTeX
@article{Xiao2026,
author = {Xiao, Hao and Yin, Mei-Ju and Xiao, Hong-Wei and Xu, Yu and Xiao, Hua-Yun},
title = {Leveraging machine learning and ten-cities observations to disentangle emission and meteorological controls on secondary inorganic aerosols in China},
journal = {Environmental Pollution},
year = {2026},
doi = {10.1016/j.envpol.2026.128190}
}
EUVIMED