Fire From the Sky · evidence ledger

Evidence

All 41 evidence records in this case's ledger. Every record carries an explicit direction, keeps what the source states separate from what we infer, and links its source and the claims it bears on. Supporting and undermining evidence get the same structure and the same seriousness.

Supports22 Undermines14 Qualifies2 Context3

Supports

22 records · strongest first

Records that favor a rung of the featured hypothesis.

Supports· strongYDIH-E001

The GISP2 Greenland platinum anomaly

what the source states: High-resolution (2.5–4.6 year) ICP-MS time series of GISP2 ice identified a platinum anomaly near the Younger Dryas onset — per the 2026 corrected characterization: spanning ~21 years, average ~30 ppt, single-sample peak ~80 ppt — with rising Pt over ~14 years and a fall over ~7, which the authors interpreted as consistent with atmospheric deposition of Pt-rich dust from an unusual, Ir-deficient, possibly extraterrestrial source whose identity could not be established with certainty.
what we infer: The best-attested single observation in the case: an independent (non-CRG) laboratory, a premier archive, a real geochemical event. It is evidential grade one of four — an elemental anomaly with no identified carrier — and everything beyond that grade is interpretation.
limitations: The anomaly's source and carrier remain unidentified; a contemporaneous PNAS comment (Boslough 2013) proposed a non-cataclysmic meteorite entry as sufficient.
~30 ppt over ~21 years is a subtle influx, not by itself a catastrophe signature.

Petaev, M. I. (2013)AI-verified citationYDIH-C001

Supports· strongYDIH-E014

Haynes 2008: the black mats documented — as wetland deposits

what the source states: A survey of terminal-Pleistocene sites documented widespread dark, organic-rich "black mat" horizons overlying the Rancholabrean surface — frequently draping the last in-place megafaunal remains and Clovis occupations — attributing the mats to elevated water tables, spring discharge, and organic accumulation during the moist Younger Dryas interval.
what we infer: Supports the observation claim precisely because it is the careful version of it: the mats are real, striking, and stratigraphically associated with the transition — and their documenter's own account of what they are (wetland facies) is the strongest reason they cannot serve as an impact marker.
limitations: Black mats occur at many other times and places; the association with the YD onset is environmental, not event-diagnostic.

Haynes, C. V., Jr. (2008)AI-verified citationYDIH-C006

Supports· strongYDIH-E024

Hiawatha dated: 58 million years — excluded as a YDIH candidate

what the source states: 40Ar/39Ar dating of impact-heated sand and U-Pb dating of zircon in impact-melt pebbles, sampled from glaciofluvial sediment draining the crater, converged on an impact age of 57.99 ± 0.54 million years (Late Paleocene).
what we infer: Clean, dual-method, direct dating that removes the only proposed crater from the case — supporting the no-crater claim and, in the process, demonstrating what decisive evidence in this field actually looks like. Retractions cut both ways and so do datings: proponents dropped Hiawatha; the hypothesis's airburst wing never needed it.
limitations: Dates the one candidate structure; cannot exclude undiscovered structures, though none with credible ~12.8-ka dating has been proposed.

Kenny, G. G. (2022)AI-verified citationYDIH-C009

Supports· strongYDIH-E025

Faith & Surovell 2009: a geologically concentrated extinction

what the source states: Analysis of last-appearance radiocarbon data for North American megafaunal genera found the record statistically consistent with a synchronous extinction concentrated in the terminal Pleistocene (assessed interval ~12,000–10,000 radiocarbon years BP, roughly 13.8–11.4 ka calibrated), rather than a long staggered decline.
what we infer: Supports the extinction observation while setting its resolution honestly: "synchronous" at the scale of this analysis means a window of two millennia. The paper is regularly miscited as dating the extinction to the impact horizon; it cannot and does not.
limitations: Last-appearance dates systematically underestimate true extinction times (Signor-Lipps effect); species-level chronologies remain sparse.

Faith, J. T. (2009)AI-verified citationYDIH-C008

Supports· strongYDIH-E028

AMOC weakening as the Younger Dryas's proximate mechanism

what the source states: A review of paleoceanographic evidence concludes that multiple independent proxies support reduced Atlantic Meridional Overturning Circulation during the Younger Dryas, consistent with the cold reversal's North Atlantic pattern, while noting ongoing debate over the initiating freshwater forcing's routing and magnitude.
what we infer: The common ground both camps stand on — and the reason the impact is evidentially optional: the mechanism that actually produced the cold is well supported and has candidate initiators that require nothing from the sky.
limitations: The initiating trigger remains genuinely open; 'known processes suffice' is a plausibility judgment, not a demonstration.

Lynch-Stieglitz, J. (2017)AI-verified citationYDIH-C013

Supports· strongYDIH-E034

The Baffin Bay retraction notice, in the editors' own words

what the source states: The PLOS One Editors retracted the article on 2026-02-11, citing: incorrect citations "raising concerns about undisclosed use of AI tools"; multiple citations that do not support the associated statements; chronology problems including missing foraminifera dating information, unaccounted 5-cm sampling uncertainty, and a calibration method not recommended for the site's latitude and period — with recommended calibration producing ages "outside the Younger Dryas"; misidentification of claimed cosmic "spherule fragments" that consulted experts identified as marine foraminifera; sampling insufficient to assess long-term particle variability, including an out-of-boundary peak equal in magnitude to the claimed boundary peak; and unclear data attribution. One author agreed with the retraction; eight disagreed; one was neutral; the rest did not respond or could not be reached.
what we infer: The most consequential of the 2026 editorial actions, because its defect list reads as a checklist of this case's methodological themes: calibration choices that manufacture synchrony, particle identification by morphology, and no background sampling. It documents error — not fabrication — and the recorded author dissent is honest provenance, not exoneration.
limitations: Retraction is evidence about a paper, not about the Pleistocene: it removes claimed support without demonstrating the event's absence.

The PLOS One Editors (2026)AI-verified citationYDIH-C031

Supports· strongYDIH-E035

The shocked-quartz retraction notice, in the editors' own words

what the source states: The PLOS One Editors retracted the article on 2026-02-11 after expert consultation, citing: an age model that did not include all available published radiocarbon samples, with some dates untraceable to source studies and included ages ranging "centuries before and after the Younger Dryas onset"; and incomplete core sampling that precludes ruling out comparable proxy spikes at other depths, prevents any background comparison, and — absent reported age models — leaves open that the claimed spike reflects sedimentation-rate change rather than an event. Eight authors disagreed with the retraction; the rest did not respond or could not be reached.
what we infer: The second 2026 editorial action, with a defect list matching the chronology and sampling critiques the independent literature had been making since 2009 — the same failure modes, a decade later, in the hypothesis's newest flagship evidence. As with Baffin Bay: error and method, not fabrication.
limitations: The underlying claim (shock-metamorphosed quartz at the boundary) could in principle be re-established by properly sampled, properly dated future work.

The PLOS One Editors (2026)AI-verified citationYDIH-C031

Supports· strongYDIH-E039

Green et al. 2025: an independent re-dating confirms the decades-long lag

what the source states: Comparing the GISP2 platinum series with annually resolved NGRIP oxygen isotopes, Greenland and Antarctic volcanic sulphate and a Spanish stalagmite record on the GICC05 timescale, the authors place the platinum peak at 12,822 yr BP and the cooling's onset about 45 years earlier. They write that their timing work "was conducted independently and in parallel with these works, without any knowledge of their results," and conclude: "Our results confirm the conclusions of these other works, that the Pt spike occurs several decades after the onset of the YDE, so whatever caused the Pt spike occurred too late to be a causal mechanism of the YDE."
what we infer: Four analyses (Svensson 2020, Abbott 2021, Holliday 2023, and this one) now agree on the lag, the last carried out without sight of the others. The case's best-attested single observation therefore cannot be the trigger's signature: if an impact cooled the North Atlantic, the Greenland platinum is not its fallout, or the fallout came decades after the cooling started. A multi-year debris-stream encounter of the kind the Taurid scenario describes would not by itself remove the problem, because the platinum still arrives after the cooling had begun.
limitations: Same platinum data and same GICC05 chronology as the earlier analyses: the agreement shows the transfer is reproducible, not that a second core shows the same lag. The authors' own first recommendation is to replicate the platinum signal in another deep Greenland core.
The authors are advocates of a rival (volcanic) trigger; the timing result does not depend on that view, but the paper's framing does.

Green, C. E. (2025)AI-verified citationYDIH-C507

Supports· strongYDIH-E041

Abbott et al. 2021: a cluster of large eruptions just before the cooling

what the source states: Using synchronised sulphate and sulphur records from Greenland and Antarctic ice cores for 13,200 to 12,800 years before 1950, the authors write that their reconstruction "highlights a ~110-year cluster of four major bipolar volcanic signals alongside several smaller events just prior to the YD inception," that "The cumulative Northern Hemisphere aerosol burden and radiative forcing from this cluster exceeds the most volcanically active periods during the Common Era," and that this "highlights the need to consider stratospheric sulphur injections and their radiative forcing in future analyses and climate model experiments."
what we infer: The volcanic candidate's ice-core basis: the eruptions are measured in both hemispheres' ice and sit at the right moment, unlike the Greenland platinum. That makes volcanic forcing a trigger candidate with an observed, correctly timed forcing behind it — something neither the meltwater account nor the impact account has in the same ice. It is a timed forcing, not a demonstrated cause.
limitations: Shows the forcing occurred; whether it could tip the deglacial climate into the Younger Dryas is left to model experiments the paper calls for, not demonstrated.
Comparable eruptions occur elsewhere in the glacial record: Green et al. 2025, citing a 2022 catalogue of 1,113 eruptions in Greenland cores, rank the eruptions at 12,978 and ~12,870 yr BP 16th and 17th by sulphur released. Coincidence in time is not causation.
The eruptions' sources are unidentified; the paper's tentative Laacher See candidates sit a century before the onset.

Abbott, P. M. (2021)AI-verified citationYDIH-C508 YDIH-C502

Supports· moderateYDIH-E002

North American sedimentary platinum enrichments at eleven sites

what the source states: Fire-assay and ICP-MS analyses at 11 sites in six U.S. states reported Pt abundances in the boundary layer averaging 6.0 ppb (range 0.3–65.6) against above/below background averaging 0.3 ppb — roughly 12× — with all sites showing peaks near strata dated to the Younger Dryas onset.
what we infer: Real measurements that extend the Greenland observation onto the continent — made by a proponent-affiliated team, sampled around pre-identified horizons, and now formally decoupled (by the authors' own 2026 correction) from the extraterrestrial-source conclusion the paper originally drew.
limitations: Single research group with CRG-affiliated authors; no independent-laboratory replication of these specific measurements.
Sampling targeted the putative boundary; without continuous background sampling the peaks' extraordinariness is unquantified.

Moore, C. R. (2017)AI-verified citationYDIH-C002

Supports· moderateYDIH-E005

Firestone et al. 2007: the founding boundary-assemblage claim

what the source states: At Clovis-age sites across North America, the authors reported a thin sediment layer containing magnetic grains with elevated iridium, magnetic microspherules, charcoal, soot, carbon spherules, glass-like carbon with nanodiamonds, and fullerenes with extraterrestrial helium — proposing that one or more large, low-density extraterrestrial objects exploded over northern North America ~12.9 ka, destabilizing the ice sheet, triggering the cooling, and contributing to megafaunal extinction and Clovis disruption.
what we infer: The paper that created the field: a maximal claim resting on a multi-proxy assemblage. Nearly every proxy on its list was subsequently challenged on replication, identification, or diagnosticity — several (fullerenes, extraterrestrial helium) quietly dropped out of even the proponent literature.
limitations: Proxy identifications made without standardized criteria; subsequent independent work could not reproduce key peaks (see the replication records).
The proposed physical scenario (a low-density object fragmenting over the ice sheet) has been substantially revised by proponents since.

Firestone, R. B. (2007)AI-verified citationYDIH-C004 YDIH-C020

Supports· moderateYDIH-E006

Wittke et al. 2013: ten million tonnes of spherules on four continents

what the source states: Analyses of boundary sediments at 18 sites on four continents reported abundance peaks in silica- and iron-rich microspherules whose geochemistry and morphology the authors interpreted as melt products of a cosmic impact event, estimating ~10 million tonnes of spherules deposited across ~50 million km².
what we infer: The proponents' widest observational claim — the hypothesis's footprint at maximum extent. Its force depends entirely on identification criteria and boundary dating that independent workers dispute; the tonnage estimate compounds every upstream uncertainty.
limitations: Proponent-team study; sites and horizons selected and identified by the same investigators whose earlier results were under challenge.
No blinded identification; the 2016 blind test later found no standard criteria exist for these objects.

Wittke, J. H. (2013)AI-verified citationYDIH-C004

Supports· moderateYDIH-E008

LeCompte et al. 2012: the proponents' methodological rebuttal

what the source states: Re-examination of sites common to the Firestone and Surovell studies reported that Surovell's protocol deviated from the original in ways the authors argue explain the null result (including size-fraction and magnetic-separation differences), and reported successful recovery of boundary microspherules at shared sites using the original protocol.
what we infer: The strongest proponent answer in the replication war: a concrete, testable diagnosis of why an independent team found nothing. What it is not: a blinded or criteria-locked demonstration — it recovers the phenomenon using the protocol whose validity is the thing in question.
limitations: Lead author later a Comet Research Group co-founder; 'independent evaluation' in the title overstates the distance from the original team.
The subsequent blind test (2016), designed to transcend exactly this stalemate, again found little at the boundary.

LeCompte, M. A. (2012)AI-verified citationYDIH-C004

Supports· moderateYDIH-E010

Kennett et al. 2009: nanodiamonds reported in the boundary layer

what the source states: The authors reported nanodiamonds (including cubic and hexagonal forms) within the Younger Dryas boundary sediment layer at multiple North American locations, interpreted as consistent with a cosmic impact event at ~12.9 ka.
what we infer: The proxy that gave the hypothesis its most exotic-sounding physical evidence. Its evidential weight rests on identification calls (diamond versus graphene-family aggregates; hexagonal versus cubic) that independent microscopy subsequently rejected — the dispute below is not about whether the sediments were analyzed but about what was in them.
limitations: A one-page report; the detailed identification evidence lives in supplements and follow-ups whose interpretations remain contested.

Kennett, D. J. (2009)AI-verified citationYDIH-C003

Supports· moderateYDIH-E012

Bunch et al. 2012: very high-temperature melt products

what the source states: Meltglass and related phases from boundary-age strata at sites in the Carolinas, Pennsylvania, and Venezuela were reported with mineralogy interpreted to require formation temperatures of roughly 1,730 to >2,200 °C, which the authors attributed to cosmic airbursts/impacts.
what we infer: Interesting mineralogy carrying an attribution it cannot support alone: extreme temperature narrows but does not name the heat source, and the materials' depositional age, in-situ context, and exclusion of mundane analogues (fulgurites, historical fires, industrial contamination) remain the unmet burden.
limitations: Proponent-team study, largely unreplicated independently.
Temperature estimates depend on phase interpretations that have not had blinded confirmation.

Bunch, T. E. (2012)AI-verified citationYDIH-C005

Supports· moderateYDIH-E017

Wolbach et al. 2018: the impact-winter compilation

what the source states: A two-part compilation of ice-core combustion aerosols and lake/marine charcoal and soot proxies reported synchronous peaks at the Younger Dryas onset, interpreted as an extraordinary biomass-burning episode — on the order of ~9–10% of Earth's biomass burned — followed by an impact winter.
what we infer: The proponents' strongest consequence claim, and the one whose statistics were most directly challenged: the "extraordinary" verdict depends on proxy selection and exceptionalism tests that independent reanalysis could not reproduce (next record). The burned-fraction figure should not be quoted without that context.
limitations: Proxy selection and statistical methods formally disputed by Holliday et al. 2020; the authors replied in the same journal.

Wolbach, W. S. (2018)AI-verified citationYDIH-C007

Supports· moderateYDIH-E020

Kennett et al. 2015: the Bayesian synchrony model

what the source states: Bayesian chronological modeling of radiocarbon dates from boundary sites on four continents concluded the data are consistent with a synchronous Younger Dryas boundary age of 12,835–12,735 cal BP.
what we infer: The proponents' formal answer to the chronology audit. "Consistent with synchrony" is weaker than "demonstrates synchrony": the model's force depends on which dates and boundary assignments enter it, which is exactly what critics contest — and the paper carries a 2024 correction (content not reviewed here).
limitations: Date inclusion/exclusion rules and boundary assignments were not independent of the event hypothesis (the circularity objection).
2024 correction recorded by Crossref; unreviewed.

Kennett, J. P. (2015)AI-verified citationYDIH-C010

Supports· moderateYDIH-E022

Napier 2010: the fragmented-comet / Taurid scenario

what the source states: A dynamical model proposing that Earth encountered dense debris from the fragmentation of a formerly giant (roughly 50–100 km) short-period comet — whose residue survives as the Taurid Complex — with such encounters capable of producing multiple dispersed airbursts and surface impacts within hours, around the Younger Dryas onset epoch.
what we infer: The scenario that modernized the hypothesis: it explains why no single crater need exist and gives the event an astronomical genealogy. It is a possibility argument — nothing in it shows the encounter happened, and no recovered material ties any YD deposit to the Taurids.
limitations: Giant-comet fragmentation histories are model-dependent; the 12.8-ka crossing is posited, not derived from observation.

Napier, W. M. (2010)AI-verified citationYDIH-C012

Supports· moderateYDIH-E038

Svensson et al. 2020: the platinum spike falls after the cooling began

what the source states: Placing the GISP2 platinum data on the synchronised Greenland chronology, the paper states: "The Pt spike occurs about 45 years after the volcanic quadruplet" — the "four large acidity spikes" by which the paper synchronizes the Younger Dryas onset between the hemispheres — that is, after Greenland cooling had begun and before it reached its minimum (its supplementary Fig. S1b).
what we infer: The first published placement of the Greenland platinum on the GICC05 timescale used for the cooling itself. If the onset and the spike are both read from the same synchronised ice, the platinum arrives decades into the cooling — so whatever delivered it cannot be what started the cooling, whatever its source.
limitations: A one-sentence aside in a chronology paper; the analysis itself is in the supplement, not set out in the main text.
The offset depends on transferring the Petaev et al. depth series onto GICC05; it is a re-dating of one existing measurement, not a new measurement.
Says nothing about the platinum's source, and nothing about the sediment platinum in North America.

Svensson, A. (2020)AI-verified citationYDIH-C507

Supports· weakYDIH-E013

Abu Hureyra: melt claims extended to a Syrian archaeological site

what the source states: Meltglass associated with the Younger Dryas onset at Abu Hureyra, Syria was reported with phases interpreted to require melting above ~2,200 °C, which the authors attributed to a low-altitude cosmic airburst near the settlement.
what we infer: The geographically boldest melt claim, at a site whose tell stratigraphy is archaeologically complex. Carried at weak strength: the paper received a November 2025 correction (content not reviewed for this case), and none of its identifications have blinded independent confirmation.
limitations: 2025 correction of unreviewed content; treat specifics provisionally.
Anthropogenic high-temperature contexts (settlement fires, later occupation) are exactly the analogues hardest to exclude at a tell site.

Moore, A. M. T. (2020)AI-verified citationYDIH-C005

Supports· weakYDIH-E031

Sweatman 2021: the proponent synthesis

what the source states: A review of the impact evidence argued that the accumulated proxy record — spherules, platinum, nanodiamonds, melt products, burning signals across many sites — collectively supports the Younger Dryas impact hypothesis strongly, and that the critical literature has mischaracterized the state of the evidence.
what we infer: The strongest peer-reviewed statement of the proponent case, carried at weak strength because it is a synthesis of the same primary studies whose replication and identification problems this case documents — aggregation does not launder non-independent, disputed lines into strong evidence.
limitations: Its consensus characterization drew a published methodological comment (Jorgeson et al. 2022) and author response.
Single author, closely identified with the proponent research program.

Sweatman, M. B. (2021)AI-verified citationYDIH-C020

Supports· weakYDIH-E033

The 2024 exchange: proponents' rejection and the critics' rebuttal

what the source states: Sweatman, Powell, and West published a detailed response arguing the 2023 review misrepresented the evidence, applied asymmetric standards, and did not refute the hypothesis; Holliday and colleagues answered with a same-journal rebuttal maintaining their conclusions.
what we infer: Carried as (weak) support because a formal, journal-published defense is part of the hypothesis's live standing — and as a marker that "debated in the literature" is true while "evenly divided" is not. The exchange is argumentative on both sides; neither installment contains new primary evidence.
limitations: Two of three response authors are closely associated with the Comet Research Group.
Rebutted in the same volume; the review's authors did not concede any principal point.

Sweatman, M. B. (2024)AI-verified citationYDIH-C020

Undermines

14 records · strongest first

Records that count against a rung — shown with the same structure and seriousness.

Undermines· strongYDIH-E004

Hall's Cave: osmium isotopes point to volcanism, not a cosmic source

what the source states: Highly-siderophile-element abundances and 187Os/188Os ratios in Hall's Cave (Texas) sediments spanning the Younger Dryas showed that the reported geochemical anomalies there are most consistent with distant volcanic gas emissions, not with an extraterrestrial impactor.
what we infer: The one time source-discriminating isotope geochemistry has been applied to a claimed YD platinum-group anomaly, it returned a terrestrial answer. This does not make every anomaly volcanic; it demonstrates that "platinum peak = impact" is not a valid inference rule — which is the diagnosticity claim's core problem.
limitations: One site; explicitly does not resolve the origin of the Greenland anomaly or other localities.

Sun, N. (2020)AI-verified citationYDIH-C011

Undermines· strongYDIH-E007

Surovell et al. 2009: seven-site independent replication failure

what the source states: An independent seven-site study — including two sites from the original Firestone sample — attempted to reproduce the reported boundary peaks in magnetic grains and magnetic microspherules and did not find spherule enrichment at the Younger Dryas boundary at any site.
what we infer: The first serious outside test of the founding observation, and it failed cleanly. The proponents' answer (LeCompte et al., carried below) is that the protocol differed in ways that lose spherules — a live methodological dispute rather than a settled verdict, but the burden it creates is real: the signature proxy appears only under proponent protocols.
limitations: LeCompte et al. 2012 documented concrete protocol deviations (size-fraction handling, magnetic separation) that could plausibly cause real spherules to be missed.

Surovell, T. A. (2009)AI-verified citationYDIH-C004

Undermines· strongYDIH-E009

The 2016 blind test: little at the boundary, more above it

what the source states: Two laboratories analyzed identical splits from a stratigraphic sequence, blinded to the target layer: the study reported very low or absent Younger Dryas boundary microspherule and nanodiamond abundance, a substantially larger microspherule peak in a younger (post-YD) sample, and the absence of standardized criteria for identifying magnetic microspherules.
what we infer: The single best-designed test in the literature, and the pattern it found — boundary unremarkable, bigger peak elsewhere — is what background-plus-selective-emphasis looks like. One principal site limits its geographic reach; nothing limits its methodological lesson.
limitations: Essentially a single-site test; proponents dispute its sample handling and argue its sites were not their strongest.

Holliday, V. T. (2016)AI-verified citationYDIH-C004

Undermines· strongYDIH-E011

Daulton et al. 2010: no nanodiamonds found; lonsdaleite identification disputed

what the source states: Transmission-electron-microscopy and electron-diffraction re-examination found no nanodiamonds in the tested Younger Dryas sediments and concluded that particles previously identified as impact-diagnostic hexagonal diamond (lonsdaleite) were misidentified — consistent instead with aggregates of graphene/graphane-family sheets.
what we infer: A direct, specialist-level challenge to the identification itself, aimed at the diagnostically crucial form. Later proponent reports at other sites keep the existence question open; the shock-diagnostic lonsdaleite call has never been independently confirmed.
limitations: Tested specific sediments; proponents argue other sites and preparations still show diamonds, so the negative does not automatically generalize.

Daulton, T. L. (2010)AI-verified citationYDIH-C003

Undermines· strongYDIH-E016

Marlon et al. 2009: no continent-wide fire pulse at the onset

what the source states: A synthesis of sedimentary charcoal records across North America found biomass burning increasing gradually from glacial times through the deglaciation, with fire regimes responding to abrupt climate changes — and no continent-wide fire signal specific to the Younger Dryas onset.
what we infer: The unprejudiced baseline: a database built by the fire-history community for its own purposes, examined across this interval, shows climate-tracking fire — not a hemispheric conflagration. Any extraordinary-burning claim has to explain why the reference synthesis cannot see it.
limitations: Charcoal databases have taphonomic and chronological smoothing limits; a very short pulse could in principle be underrepresented.

Marlon, J. R. (2009)AI-verified citationYDIH-C007

Undermines· strongYDIH-E019

Meltzer et al. 2014: 3 of 29 claimed boundary sites securely dated

what the source states: A site-by-site chronological audit of 29 localities claimed to contain the Younger Dryas boundary layer concluded that only 3 had radiocarbon control adequate to place their proxy horizon at ~12,800 cal BP; at the rest, the dating was too imprecise, too sparse, or stratigraphically disconnected from the sampled horizon to support an isochron.
what we infer: The synchroneity claim's foundation examined stone by stone: most of the "synchronous" sites turn out to be synchronous only if one assumes the conclusion when assigning the boundary. Apparent global simultaneity built from loose local chronologies is the specific failure mode the 2026 retractions later exhibited.
limitations: Kennett et al.'s 2015 Bayesian reanalysis answered with a formal synchrony model; the dispute over date selection and boundary assignment continues.

Meltzer, D. J. (2014)AI-verified citationYDIH-C010

Undermines· strongYDIH-E026

Buchanan et al. 2008: no demographic crash in the radiocarbon record

what the source states: Analysis of ~1,500 archaeological radiocarbon dates from across North America found no evidence of the human population decline at 12.9 ka predicted by the extraterrestrial-impact hypothesis; date frequencies show continuity through the boundary.
what we infer: A direct test of a distinctive prediction, published within a year of the hypothesis — and negative. Radiocarbon-frequency demography is an imperfect proxy, but the predicted signal was a continent-wide catastrophe, which is exactly the size of signal the method should see.
limitations: Radiocarbon date frequencies conflate population size with site preservation and research intensity.

Buchanan, B. (2008)AI-verified citationYDIH-C023

Undermines· strongYDIH-E027

Holliday & Meltzer 2010: technological transition, not termination

what the source states: A review of the Paleoindian archaeological record concluded it provides no evidence of a continent-wide catastrophe at 12.9 ka: Clovis lithic technology gives way to Folsom and regional successors in a pattern of normal cultural transition, with no occupation gap, and "Clovis" itself denotes a technocomplex rather than a bounded population.
what we infer: The archaeological specialists' verdict on the hypothesis's most checkable human claim. The category correction matters as much as the data: a projectile-point style cannot "collapse" the way a civilization can, and the case treats conflation of the two as a red flag wherever it appears.
limitations: Regional archaeological records vary in resolution; absence of a visible gap is weaker in poorly sampled regions.

Holliday, V. T. (2010)AI-verified citationYDIH-C023

Undermines· strongYDIH-E032

Holliday et al. 2023: the comprehensive refutation

what the source states: A thirteen-author cross-disciplinary review concluded the hypothesis fails across every examined domain — impact physics, geology, geochemistry, chronology, ecology, and archaeology — finding no self-consistent physical and stratigraphic case for a Younger Dryas cosmic event: proxy peaks non-reproducible or non-diagnostic, chronologies inadequate for synchrony, proposed impact scenarios physically inconsistent, and the claimed consequences explained by conventional processes.
what we infer: The current center of gravity of independent expert assessment. Its force comes from breadth (every evidence line addressed in one place) rather than from new data; the proponents' 2024 "Rejection" and the authors' further "Rebuttal" (both carried in the sources) leave the exchange formally open and substantively where it stood.
limitations: Authored substantially by the hypothesis's long-standing critics — thoroughness and partisanship are both real here.
Answered by Sweatman, Powell & West 2024 (ESR 258: 104960); the authors rebutted (ESR 258: 104961).

Holliday, V. T. (2023)AI-verified citationYDIH-C020

Undermines· moderateYDIH-E015

Pigati et al. 2012: 'impact markers' accumulate in ordinary wetlands

what the source states: Measurements in desert-wetland sequences of multiple ages found that candidate impact markers (including magnetic spherules) accumulate in such settings through ordinary processes, concentrated by the wetland environment, and concluded these materials and black mats are not uniquely diagnostic of an impact horizon.
what we infer: A background-rate result: the materials the hypothesis counts as event markers occur, and concentrate, without any event. Any boundary claim must therefore beat the wetland background — which returns the burden to sampling designs that measure background at all.
limitations: Proponents dispute whether the wetland-accumulated materials are equivalent to their best boundary markers.

Pigati, J. S. (2012)AI-verified citationYDIH-C011

Undermines· moderateYDIH-E018

Holliday et al. 2020: the burning compilation challenged

what the source states: A formal discussion of the Wolbach compilation challenged its data selection, its treatment of chronological uncertainty, the claimed statistical exceptionalism of the onset-age peaks, and the reproducibility of the global-burning interpretation, concluding the evidence does not support an extraordinary synchronous episode.
what we infer: Two of the authors built the global charcoal database the compilation drew on — this is the data's own custodians disputing the use made of it. With Wolbach's reply, the honest status of the fire claim is "disputed, with the independent synthesis against it," not "refuted."
limitations: Answered point-by-point by Wolbach et al. in the same volume (10.1086/706265); the exchange remains open.

Holliday, V. T. (2020)AI-verified citationYDIH-C007

Undermines· moderateYDIH-E029

Page-Ladson 2026: a volcanic rival for the trigger and the geochemistry

what the source states: Highly-siderophile-element and osmium-isotope geochemistry at Page-Ladson, Florida reported volcanic signatures in Younger Dryas onset sediments and proposed volcanic forcing as a contributing trigger for the cooling.
what we infer: A second site where discriminating geochemistry favors volcanism over impact, and — more importantly for the attribution — a rival exotic trigger with actual source evidence behind it. Carried at moderate strength: it is one very recent study and should itself be treated as an emerging hypothesis, not a new consensus.
limitations: Published April 2026; needs replication, independent age synchronization, and climate-model testing.
A volcanic trigger for the YD would be a rival hypothesis, not a confirmation of the conventional meltwater account.

Nana Yobo, L. (2026)AI-verified citationYDIH-C011 YDIH-C021

Undermines· moderateYDIH-E030

Pinter et al. 2011: the first requiem

what the source states: A cross-disciplinary critical review concluded that many of the hypothesis's original evidence lines were non-reproducible or non-diagnostic — with some early markers (fullerenes, extraterrestrial helium) effectively abandoned — and that the remaining evidence did not support a Younger Dryas impact.
what we infer: The field's first systematic negative verdict. Its title proved premature as sociology — the literature grew for fifteen more years — which is itself a datum about how contested hypotheses persist; the substantive criticisms largely stand and were extended by the 2023 review.
limitations: Predates the platinum line, the Bayesian chronology, and everything since 2011.

Pinter, N. (2011)AI-verified citationYDIH-C020

Undermines· weakYDIH-E040

Green et al. 2025: the platinum's chemistry and duration read as volcanic, not as one impact

what the source states: The authors find that elevated platinum spans 12,833–12,819 yr BP and write that this "suggests a persistent yet variable flux of Pt lasting about 14 years, rather than a short-lived event like a single explosive eruption or a bolide impact." Comparing elemental ratios with meteorites, magmas, impact sediments and volcanic products, they report that "The most closely matching geochemical signature is that of volcanic gas condensates," and conclude: "Although we cannot rule out the impact of an unknown type of meteorite, the simplest explanation is that the Pt spike instead probably reflects a relatively small Icelandic fissure eruption." Their new measurements of Laacher See tephra find it platinum-poor, and they conclude the results "very strongly suggest that the LSE was not responsible for the GISP2 Pt anomaly."
what we infer: A terrestrial source for the Greenland platinum is now a published, specified proposal rather than a bare possibility, and the one named volcanic candidate that had been suggested (Laacher See) is ruled out by direct measurement. Carried as weak: the volcanic reading rests on which reference materials the spike's ratios sit nearest, and the spike resembles nothing well; it narrows the impact reading of the platinum without discriminating the source the way osmium isotopes would.
limitations: The authors also write that "we cannot rule out post-depositional diffusion of Pt within the ice to produce the Pt spike profile" — which weakens the duration argument they rest on the 14-year spread.
The ice-core spike's Pt/Al values are far from every reference group, volcanic condensates included; nearest is not matching.
No isotopic source tracer was measured; the decisive measurement (osmium isotopes on the same ice) remains undone.
Their 14-year span counts the elevated interval differently from the ~21-year span in the 2026 correction's characterization; both describe the same measurements.

Green, C. E. (2025)AI-verified citationYDIH-C011

Qualifies

2 records · strongest first

Records that narrow, complicate, or bound what another record can carry.

Qualifies· strongYDIH-E003

The May 2026 correction: extraterrestrial-source conclusion withdrawn

what the source states: The formal correction replaces the original's description of the Greenland anomaly as "the most compelling evidence to date for a catastrophic extraterrestrial event" with source-neutral language, corrects the Petaev values (~21-year span, ~30 ppt average, ~80 ppt peak), adds the Boslough comment and Petaev reply as references, and rewrites the conclusion: the North American enrichments are "compatible with interpretations involving a continentally distributed influx of Pt-rich material" but "do not independently resolve the ultimate source of the Greenland anomaly or the specific mechanism responsible."
what we infer: An official, author-signed example of the distinction this whole case turns on: the observation stands; the attribution is withdrawn. Every secondary summary that still cites Moore et al. 2017 as platinum proof of an impact is now citing language the authors themselves no longer stand behind.
limitations: A correction is not a retraction: the measurements themselves are unchallenged by this action.

Moore, C. R. (2026)AI-verified citationYDIH-C001 YDIH-C002 YDIH-C011

Qualifies· moderateYDIH-E037

The joint-inference problem: eight links, each contested

what the source states: The research report's methodological synthesis: a convincing YDIH case requires a chain of independent propositions to all hold — anomaly exists → correctly identified → extraordinary against background → securely dated → synchronous across sites → extraterrestrial in origin → energetically sufficient → causally responsible for climate, extinction, and cultural change — and evidence for an early link does not validate later ones. No proposed marker has achieved the combination of specificity, reproducibility, chronology, and causal linkage across the claimed footprint.
what we infer: The case's structural summary: most public YDIH argument jumps from grade one (an anomaly exists) to grade four (an impact did it). The ladder in this case is built to make that jump visible — and every disputed link is separately recorded above.
limitations: An AI-generated synthesis judgment, imported as the framing it is rather than as primary evidence.

AI deep-research agent (commissioned by Eugene) (2026)Verified — held in project libraryYDIH-C020

Context

3 records · strongest first

Background that informs the case without pushing either way.

Context· moderateYDIH-E021

The Younger Dryas onset independently dated to ~12,870 ± 30 BP

what the source states: High-resolution speleothem and ice-core chronologies place the Younger Dryas onset at approximately 12,870 ± 30 years before present, with structure suggesting the transition unfolded over decades.
what we infer: The target any synchrony claim must hit — and a reminder that the familiar "12,800 years ago" is a rounded convention, not an exact timestamp. A proxy layer synchronous at 12,785 and a climate onset at 12,870 are only "the same event" within generous uncertainties.
limitations: Different archives sit on different age scales; cross-archive comparison at decadal precision remains genuinely hard.

Cheng, H. (2020)AI-verified citationYDIH-C010

Context· moderateYDIH-E023

Hiawatha discovered: a 31-km crater with an initially open age

what the source states: Airborne radar sounding revealed a ~31-km-wide impact crater beneath Hiawatha Glacier in northwest Greenland; the discovery paper argued the crater could be as old as the Pleistocene's start or as young as its end, explicitly leaving a Younger Dryas age possible.
what we infer: Why the crater mattered to this case: for four years it was the candidate smoking gun under the ice sheet the original hypothesis had pointed at. Recorded as context because its relevance was extinguished by the direct dating that followed.
limitations: The discovery paper's age discussion was explicitly provisional, pending direct dating.

Kjær, K. H. (2018)AI-verified citationYDIH-C009

Context· weakYDIH-E036

The Comet Research Group's public posture

what the source states: The proponent organization's website (fetched 2026-08-22) states that "12,800 years ago, a fragmented comet detonated over Earth's northern hemisphere. The evidence is undeniable: a global burn layer, mass extinctions, and the near-erasure of human civilization," and solicits donations for Greenland ice-core and Gulf of St. Lawrence crater expeditions.
what we infer: Recorded so readers can calibrate sources: the advocacy organization's public certainty ("undeniable") exceeds anything in the peer-reviewed literature — including its own members' papers, two of which were retracted and one corrected in 2026. Funding and organizational overlap are disclosed facts to weigh, not misconduct; but "independent replication" claims should be checked against this network.
limitations: Website language changes over time; specifics should be archived before being cited.

Comet Research Group (2026)AI-verified citationYDIH-C020