Evidentiary methods
PFAS model methods, limits and reproducibility
Written for a reviewer looking for overstatement. What the model does not do comes first, every figure carries the unit that makes it checkable, and each artifact is named with its actual access status rather than implied to be downloadable.
Methods and fit statistics: the Rogue River manuscript, which has not been accepted for publication and is not yet public. Data: DOI 10.5281/zenodo.21536787 (open) and DOI 10.5281/zenodo.21096358 (record public, data files not public).
- PFOS only — not all PFAS
- 846 wells → 73 grid cells
- Source term prescribed, not fitted
A methods page that reports only what went well is read as marketing and discounted whole, including the parts that were true. So this one is ordered the way a reviewer reads rather than the way that flatters us: the limits first, then what was measured and at what unit, then what was deliberately left untuned, then what you can actually download.
One figure on this page is easy to carry away wrong, so it is written to resist it: 846 wells, of which the 477 with a detection aggregate to 73 independent model cells, and the second number is the one that bounds what the comparison can support.
FAQ
What does this PFAS transport model not do?
It simulates PFOS only; compound-specific transport with per-analyte isotherms is future work. The two models run separately: no water and no PFAS pass between them, and MODFLOW 6 alone carries PFOS from a prescribed source, on the calibrated steady flow field, through the aquifer to the streambed. The paper's attribution of PFAS in the lower river comes from the river's PFAS composition and does not use the transport model. Transient transport runs on a steady flow field, which omits non-stationary pumping and recharge and is a stated limitation for plume timing. The source cap is a numerical bound, not a measured ceiling, and it sets most of the release.
What was validated, and against what?
The PFOS comparison draws on 846 wells, and the 477 with a detection aggregate to 73 model grid cells, 8 prescribed and 65 predicted; 73 is the honest denominator for how much independent evidence the comparison carries. The simulated groundwater PFOS spans the observed range above the 10 ng/L background, but most of the agreement is the prescribed background itself. At the cell level the model shows no skill: a constant at the observed mean scores lower error than the model. The plume map is not validated. The paper's attribution comes from the river's PFAS composition, which shows a plume-type contribution in the lower river without a quantitative share, and does not use the transport model. A streambed porewater check at 43 sites is the most direct available test of the modelled pathway.
Why is there no single accuracy percentage?
Because there is no skill for a percentage to report. The manuscript compares the model with the observations in log space, cell by cell, and reports a permutation test showing no cell-level skill: re-pairing the simulated values at random does as well or better in about one in five of 10,000 permutations (19.7 %), and a constant at the observed mean scores lower error. It reports the agreement with the observed range as largely the prescribed background, so a headline accuracy figure would present the background as performance, and this page does not offer one.
What was the model tuned to reproduce?
Very little, and that is the argument rather than an apology. The source is set from measured concentrations, capped at 10⁵ ng/L, and not fitted; the cap sets most of the release. There is no calibrated groundwater concentration multiplier. The only quantity fitted to in-stream PFOS is one soil-loading scale of the surface-water model, fitted to the 16 sampled reaches with a detection; a separate joint fit of both models to the same samples identifies no groundwater term. The site study is not calibrated to PFOS magnitude: its source concentration is inverted from measured source soil through an assumed partition coefficient, and no parameter was adjusted to improve the agreement with measured PFOS. A model that was not tuned to its validation target is harder to attack than one with a better fit and more free parameters.
What can a third party actually obtain?
The methods and the well comparison are in the vadose study, accepted for publication in the Journal of Contaminant Hydrology and not yet published. The site well-level dataset is an open Zenodo deposit, DOI 10.5281/zenodo.21536787, holding 35 wells and 4,275 observations. The national soil inventory has a citable DOI, 10.5281/zenodo.21096358, but its data files are not public — the description is public, the data is not, and citing it as downloadable would be wrong. The groundwater lithology inventory is in open discussion as preprint essd-2026-527. What is not archived is a specific run: reproducing one needs the model bundle for that run, which the paper does not contain.
Next steps
Last updated 2026-08-31. Identifiers on this page were resolved rather than assumed; see the note under Table 2.