Differentiation of primary and secondary marine organic aerosol with machine learning

Chen, Baihua; Lei, Lu; Chevassus, Emmanuel; Xu, Wei; Zhen, Ling; Zhong, Haobin; Wang, Lin; Lin, Chunshui; Huang, Ru-Jin; Ceburnis, Darius; O'Dowd, Colin; Ovadnevaite, Jurgita

doi:https://doi.org/10.5194/acp-25-14205-2025

Articles | Volume 25, issue 21

https://doi.org/10.5194/acp-25-14205-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/acp-25-14205-2025

© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 25, issue 21

Research article

|

30 Oct 2025

Research article |

| 30 Oct 2025

Differentiation of primary and secondary marine organic aerosol with machine learning

Baihua Chen, Lu Lei, Emmanuel Chevassus, Wei Xu, Ling Zhen, Haobin Zhong, Lin Wang, Chunshui Lin, Ru-Jin Huang, Darius Ceburnis, Colin O'Dowd, and Jurgita Ovadnevaite

Supplement

https://doi.org/10.5194/acp-25-14205-2025-supplement

Data sets

Cloud condensation nuclei and hygroscopic growth measurement at Mace Head from 2009 to 2010 W. Xu et al. https://doi.org/10.17632/3dx6pnx869.1

Download

Article (3461 KB)
Full-text XML

Short summary

This study uses machine learning to separate marine primary organic aerosol (POA) and secondary organic aerosol (SOA) from 1 decade of high-resolution data. POA averages 51 % of marine organic aerosols annually, peaking at 63 % in summer. A support vector regression model, validated via fuzzy clustering and Monte Carlo simulations, identifies seasonal patterns of POA linked to biological activity. We found diverse impacts of marine POA and SOA on the aerosol hygroscopicity and mixing state.