The occurrence, distribution, and risks of PFAS at AFFF-impacted sites in Finland

Reinikainen, Jussi; Perkola, Noora; Äystö, Lauri; Sorvari, Jaana

doi:10.1016/j.scitotenv.2022.154237

Cited by 28 publications

(8 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To validate the KNN classification model, data were scraped from four publications with readily available classification data ( n = 192). ,− Many publications measure fish concentrations of PFASs; however, most do not provide labeled sources that would make them amenable to this validation analysis. While this is a data set of limited size, it captures data from geographically diverse regions where PFAS sources are known.…”

Section: Resultsmentioning

confidence: 99%

“…Labeled source data were obtained from five external publications for the external validation (Section S.1). ,− …”

Section: Methodsmentioning

confidence: 99%

“…All data cleaning, analysis, and modeling were conducted with Python 3.9.0. Data cleaning consisted of importing and cleaning (i.e., organization and replacing null/nondetect (ND) values) the machine-readable data from Langberg et al and the other independent sources. ,− All data were checked for consistency prior to ML modeling. Thorough preliminary analysis of the data as well as data subsets are available in Langberg et al; primary findings included higher PFAS levels in fish liver than fish muscle, distinct geographic trends, and distinct distributions of PFCAs and PFSAs based on principal component analysis (PCA) of non-normalized concentration data.…”

Section: Methodsmentioning

confidence: 99%

“…We also discuss our approach to limit model overfitting, address censored data, and optimize calibration parameters. We then apply the trained classifiers to external validation data ,− ( n = 192) from geographically diverse regions to assess the generalizability of source classification using fish data. We conclude by discussing the limitations of using fish data for PFAS source tracking in surface water and areas for future research.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Integration of Per- and Polyfluoroalkyl Substance (PFAS) Fingerprints in Fish with Machine Learning for PFAS Source Tracking in Surface Water

Stults

Higgins

Helbling

2023

Environ. Sci. Technol. Lett.

View full text Add to dashboard Cite

Per- and polyfluoroalkyl substances (PFASs) are a class of environmental contaminants that originate from various sources. The unique chemical fingerprints associated with many commercial products and industrial applications make PFASs ideal candidates for machine learning (ML)-assisted environmental forensics. Here, we propose a novel use of PFAS fingerprints in fish tissue from surface water systems to classify exposure from multiple sources of PFASs using a proof-of-concept demonstration. Three supervised ML classification techniques (k-nearest neighbors (KNN), decision trees, support vector machines) implementing two predictive features are used to classify literature-reported PFAS fingerprints in fish (n = 1057). The importance of additional predictive features was explored using brute force optimization of a multifeature KNN algorithm. The multiclass classification considered exposure to aqueous film-forming foam-impacted water, paper industry wastewater, diffuse sources, or PFASs undergoing long-range transport. The optimized classifiers demonstrated 85%–94% classification accuracy for this first known multiclass classification of PFASs for environmental forensics. The optimized classifiers also demonstrated 79%–92% classification accuracy with a set of independent external validation data (n = 192). Our results demonstrate that PFAS fingerprints in fish tissue may be an effective means of PFAS source tracking in surface water systems. The source code is provided for guidance on best practices for ML-assisted environmental forensics.

show abstract

Section: Resultsmentioning

confidence: 99%

“…Labeled source data were obtained from five external publications for the external validation (Section S.1). ,− …”

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Integration of Per- and Polyfluoroalkyl Substance (PFAS) Fingerprints in Fish with Machine Learning for PFAS Source Tracking in Surface Water

Stults

Higgins

Helbling

2023

Environ. Sci. Technol. Lett.

View full text Add to dashboard Cite

show abstract

“…The extensive application of aqueous film-forming foams (AFFF) containing per- and polyfluoroalkyl substances (PFAS) for extinguishing hydrocarbon-based fires, along with other surfactants and solvents, has led to their widespread distribution in a variety of environmental media. − AFFF solutions can infiltrate into the vadose zone and become a major reservoir for PFAS release and thus serve as a long-term source of groundwater contamination. − PFAS-contaminated aquifers that also serve as a drinking water source can result in a primary PFAS exposure route for humans. − …”

Section: Introductionmentioning

confidence: 99%

Influence of Tension-Driven Flow on the Transport of AFFF in Unsaturated Media

Vahedian,

Silva,

Šimůnek

et al. 2024

ACS EST Water

View full text Add to dashboard Cite

Per- and polyfluoroalkyl substances, and other chemicals in aqueous fire-fighting foam (AFFF), accumulate at air–water interfaces in the vadose zone, reducing interfacial tension and potentially causing tension-driven flow (TDF). This study investigates the importance of TDF on AFFF lateral spreading in the vadose zone using bench-scale column experiments and numerical modeling. Understanding AFFF spreading beyond an original spill footprint would enhance site conceptual models, wherein understanding the contaminant source zone is critical. Our experiments demonstrate that AFFF exhibits noteworthy TDF, characterized by substantial lateral water flow and AFFF spreading, causing variations in moisture content, which alters capillary pressure and hydraulic conductivity. The modified HYDRUS model accurately simulated experiments and was then used to simulate site-scale AFFF source zone dynamics. The numerical investigation demonstrated that TDF could lead to a significant increase in the surface footprint of an AFFF source zone, where wetter conditions yield a 6-fold increase compared to an initial spill, while drier soils exhibit a 4-fold expansion. The footprint expansion occurs over decades, which is highly relevant for our conceptual understanding of both historic and relatively new AFFF spills. This study highlights that the lateral flow of AFFF due to TDF is an important transport mechanism that controls the AFFF source zone.

show abstract

Per‐ and Polyfluoroalkyl Substances (PFAS) Migration from Water to Soil–Plant Systems, Health Risks, and Implications for Remediation

2023

View full text Add to dashboard Cite

The occurrence, distribution, and risks of PFAS at AFFF-impacted sites in Finland

Cited by 28 publications

References 43 publications

Integration of Per- and Polyfluoroalkyl Substance (PFAS) Fingerprints in Fish with Machine Learning for PFAS Source Tracking in Surface Water

Integration of Per- and Polyfluoroalkyl Substance (PFAS) Fingerprints in Fish with Machine Learning for PFAS Source Tracking in Surface Water

Influence of Tension-Driven Flow on the Transport of AFFF in Unsaturated Media

Per‐ and Polyfluoroalkyl Substances (PFAS) Migration from Water to Soil–Plant Systems, Health Risks, and Implications for Remediation

Contact Info

Product

Resources

About