A quietly circulated math paper, an independent field study, and a newly reopened state investigation all point to the same conclusion: the data behind Alpine's celebrated environmental turnaround does not hold up.
In the fall of 2019, adult coho salmon began dying by the dozens in Harmon Creek, a modest tributary that runs past what was then a fast-growing regional trucking hub in Washington's Kent Valley. Within weeks, state regulators had traced the die-off to stormwater runoff carrying 6PPD-quinone, a tire-wear byproduct now known to be lethal to coho at concentrations measured in parts per billion. The Washington State Department of Ecology issued an administrative order. The company responsible, Alpine Logistics Corp, installed a stormwater retrofit and commissioned an independent study to show it worked.
That study is the reason Alpine Logistics exists, in its current form, at all.
Published in May 2021 by a graduate researcher at the University of Cambridge, the paper reported a 92 percent reduction in toxic runoff following the retrofit, alongside a corresponding drop in salmon mortality. It became the centerpiece of Alpine's public turnaround story — cited in sustainability filings, investor decks, and, eventually, a Department of Ecology closure of the matter entirely. The company's valuation has grown more than sixfold since.
Now that story is unraveling, and it started, improbably, with a favor between friends.
A Side Project, Circulated Quietly
In October 2024, a Cambridge-trained mathematician — best known in academic circles for an unrelated, celebrated result in number theory — ran a purely statistical check on the 2021 paper's published tables, at the private request of an acquaintance weighing an investment in Alpine. The write-up was never meant to be public. Its most striking finding required no statistics at all: recomputing each treatment site's reduction percentage directly from the paper's own published concentration figures produced three different numbers — 91.9, 92.1, and 92.2 percent — not the single, identical 92.0 percent the study had reported for all three. The analysis also noted a discrepancy between the study's text, which described five sites with complete before-and-after data, and its own table, which showed only four; flagged a pattern in the reported figures that deviated from the digit distribution expected of genuine environmental measurements; and pointed to the exclusion, from the study's final results, of the single site with the highest recorded contamination — attributed to equipment failure that occurred nowhere else in the nine-site monitoring network.
"None of these findings speak to the underlying environmental chemistry," the analysis states. "They concern only the internal statistical consistency of the published dataset. We take no position on intent."
The analysis circulated informally for months before a science journalist, independently investigating Alpine's sustainability claims, obtained a copy and posted it publicly in March 2025. Reached for comment after it began drawing attention, the author — who has largely stepped back from academic mathematics in favor of outdoor photography and a travel-focused social media following — seemed unbothered by the scrutiny. "wait people are actually using my notes to drag her?? it's literally just maths, nothing personal!! i didn't even want this public. people are turning equations into a witch hunt nowadays :( put some respect on the brilliant environmental scientist, she literally did the actual heavy lifting in the mud!"
Independent Confirmation, Twice Over
A A single statistical critique might have been dismissed as noise. It wasn't left to stand alone. In January, statisticians at the London School of Economics published a follow-up analysis benchmarking the Harmon Creek dataset — using the corrected, recomputed reduction figures rather than the study's originally printed numbers — against 34 comparable, independently audited stormwater compliance studies, and found it to be a significant outlier on every measure examined, at a level not observed in any comparison study.
The following month, researchers at the University of Washington's Center for Urban Waters did something neither prior analysis could: they went back to the creek. Resampling the excluded outfall directly, they measured 6PPD-quinone concentrations well above the threshold known to be acutely lethal to coho — inconsistent with the facility-wide performance the 2021 study had implied. Their report also points to a more mundane explanation running alongside the numbers: the retrofit's bioretention media, they found, had never been serviced or replaced since its 2020 installation, a maintenance lapse consistent with a well-documented failure mode in which such systems lose their contaminant-absorbing capacity over time.
The Washington State Department of Ecology reopened its file on the matter in April. A preliminary memorandum released this month confirms the agency's own re-inspection found the same unserviced media, and states that a determination on Alpine's compliance status — and on whether the 2021 study should ever have closed the case — is pending.
A Company, and a Friendship, Under Pressure
Alpine Logistics Corp declined an interview request for this story. In a brief statement, a company spokesperson said Alpine "stands behind the science that has guided our environmental program" and is "cooperating fully" with the Department's review.
Privately, the picture is reportedly less settled. According to a person familiar with internal discussions, Alpine's founder has raised the possibility of commissioning a fully independent audit of the company's environmental data — and of getting ahead of the story by disclosing its findings publicly, whatever they turn out to be, rather than waiting for regulators to do it first. No final decision had been made as of this week.
The 2021 study's lead author, now a research scientist at the Cascadia Institute for Watershed Sciences, has not responded to requests for comment. The paper itself remains published, unretracted, and now carries an editorial expression of concern.