EUVIMEDEUVIMED entdecken
Crossref

Driving factors of chlorinated OPEs pollution in industrial soils: an empirical study based on machine learning and a novel petrochemical source

Chi Zhu, Lu Cao, Enze Pei, Xin Huang, Shixiang Dai, Liang Ding, Bo Wang, Changsheng Qu

Asian Journal of Water, Environment and Pollution · 01.07.2026 · 23 (4) · 026200139

DOI: 10.36922/ajwep026200139 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Asian Journal of Water, Environment and Pollution, online erschienen am 01.07.2026. Verfasst von Chi Zhu und 7 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Chlorinated organophosphate esters (Cl-OPEs), a major class of organophosphate ester pollutants, are ubiquitously detected in industrial soils, yet quantitatively apportioning these contaminants remains a critical challenge. Traditional machine learning-based source apportionment approaches are often undermined by the spatial clustering of monitoring data, a phenomenon known as geographical bias. To address this limitation, this study first identified geographical bias through exploratory data diagnostics and subsequently constructed a robust random forest-based machine learning prediction framework that excludes geographical features to eliminate the confounding effect of spatial bias. The optimized model exhibited excellent reliability in predicting soil Cl-OPEs concentrations (test set R2 = 0.90). Model interpretability analysis via feature importance assessment confirmed E-waste processing (site function category_E-waste processing) as the primary driver of Cl-OPEs pollution. To validate the model and explore overlooked sources, we also conducted an empirical case study of the petrochemical industry. Comparative analysis not only confirmed E-waste processing areas as the primary source but also identified the petrochemical industry as a previously underestimated yet significant source of emissions, with its coil Cl-OPEs levels markedly exceeding those of other traditional industrial areas. Overall, this study proposes a novel, geographical-bias-free machine learning source apportionment method and provides new preliminary empirical evidence supporting pollution control of Cl-OPEs.

Zitieren

APA

Zhu, C., Cao, L., Pei, E., Huang, X., Dai, S., Ding, L., Wang, B., & Qu, C. (2026). Driving factors of chlorinated OPEs pollution in industrial soils: an empirical study based on machine learning and a novel petrochemical source. Asian Journal of Water, Environment and Pollution, 23(4), 026200139. https://doi.org/10.36922/ajwep026200139

Vancouver

Zhu C, Cao L, Pei E, Huang X, Dai S, Ding L, et al. Driving factors of chlorinated OPEs pollution in industrial soils: an empirical study based on machine learning and a novel petrochemical source. Asian Journal of Water, Environment and Pollution. 2026;23(4):026200139. doi:10.36922/ajwep026200139

BibTeX
@article{Zhu2026,
  author = {Zhu, Chi and Cao, Lu and Pei, Enze and Huang, Xin and Dai, Shixiang and Ding, Liang and Wang, Bo and Qu, Changsheng},
  title = {Driving factors of chlorinated OPEs pollution in industrial soils: an empirical study based on machine learning and a novel petrochemical source},
  journal = {Asian Journal of Water, Environment and Pollution},
  year = {2026},
  volume = {23},
  number = {4},
  pages = {026200139},
  doi = {10.36922/ajwep026200139}
}
Kostenlos starten →