Novel insights from multivariate analysis of metadata from a large PFAS remedial investigation dataset

The extensive use of per- and polyfluoroalkyl substances (PFAS) in aqueous film-forming foam (AFFF) has resulted in widespread environmental contamination. This study applies Principal Component Analysis (PCA) and Hierarchical Clustering Analysis (HCA) to approximately 21,000 PFAS records collected from 29 sites in the continental United States to identify compositional patterns, source signatures, and spatial trends across groundwater, porewater, surface water, soil, and sediment at AFFF source areas and downgradient plumes . A novel concept termed the estimated PFAS vadose zone dilution attenuation factor (the ratio between PFAS porewater and groundwater concentration) is defined herein and negatively correlated with perfluorinated chain length, reflecting sorption and mobility differences. PCA revealed clustering by chain length and media type: short-chain PFAS dominated aqueous samples, while long-chain PFAS were prevalent in soil and sediment. PFHxS was most abundant in groundwater and porewater; PFOS dominated surface water and solids. The maximum PFAS soil concentrations are generally limited to the upper 1 m in arid climates and deeper in sub-humid regions. Site-specific PFAS fingerprints were associated with hydroclimatic variables. PCA results suggest diffuse contamination patterns, likely due to historical AFFF application. These findings support refinement of the site-specific knowledge framework, inform risk assessments, and guide targeted sampling and remediation strategies at PFAS-impacted sites.