0X

Scientific efforts — the Galileo Project and the instrumentation problem

A handful of credentialed scientists have stopped arguing about old gun-camera footage and started building instruments — and the most useful thing they have produced so far is a rigorous null result.

What the Galileo Project actually is
  • DOCUMENTED Founded July 2021 by Avi Loeb at Harvard. Loeb is the Frank B. Baird Jr. Professor of Science at Harvard, former chair of its astronomy department (2011–2020), and director of the Institute for Theory and Computation. Whatever one concludes about his UAP conclusions, his standing as an astrophysicist is not in question — this is not a fringe operator.
  • DOCUMENTED Three declared research branches. Per Loeb's own overview paper: (1) build multi-sensor ground observatories to obtain high-resolution data on unidentified aerial objects; (2) search for and characterise ʻOumuamua-like interstellar objects; (3) search for possible non-human satellites in Earth orbit. Explicitly framed as a search for physical objects, in contrast to SETI's search for electromagnetic signals.
  • DOCUMENTED Privately funded, not a Harvard institutional program. The project runs on private donations and is hosted at Harvard rather than funded by it. Co-founder Frank Laukien — CEO of Bruker Corporation, a scientific-instrument manufacturer — is a named co-author on the project's methodology paper. Bruker instrumentation appears in the project's later materials analysis. That is a real institutional interest and readers should hold it in view.
  • DOCUMENTED The methodology is defensible, and that is the point. Calibrated instruments, known error budgets, pre-registered analysis pipelines, published negative results. This is how you would design a UAP study if you wanted it to survive peer review. It is a categorically different activity from re-litigating 1947.

A. Loeb, "Overview of the Galileo Project", Journal of Astronomical Instrumentation 12 (2023) — preprint arXiv:2209.02479. W. A. Watters, A. Loeb, F. Laukien et al., "The Scientific Investigation of Unidentified Aerial Phenomena (UAP) Using Multimodal Ground-Based Observatories", Journal of Astronomical Instrumentation 12(1), 2340006 (2023).

The hardware — what is actually deployed
  • DOCUMENTED First observatory: the roof of the Harvard College Observatory, Cambridge, Massachusetts. A second site operates in Colorado. Each is a self-contained sensor pod, informally called a "Dalek" for its shape.
  • DOCUMENTED Eight uncooled long-wave infrared FLIR Boson 640 cameras arranged as an all-sky array, running continuously. Supplemented by visible-band cameras, radio spectrum monitoring, audio, and a geomagnetic variometer.
  • DOCUMENTED The calibration trick is the clever part. The array is calibrated against ADS-B transponder broadcasts from commercial aircraft — objects of known position, altitude and velocity crossing the field of view thousands of times a day. This converts pixel motion into real kinematics with a stated error budget, which is precisely what military gun-camera footage cannot do.
  • DOCUMENTED Magnetometry added in 2025. A geomagnetic variometer station was commissioned at the Colorado observatory to look for magnetic-field disturbances coincident with optical detections — a cross-check that any single-sensor artifact would fail.

L. Dominé, A. Biswas, R. Cloete, A. Delacroix et al., "Commissioning an All-Sky Infrared Camera Array for Detection of Airborne Objects", Sensors 25(3), 783 (2025). F. Vervelidou, A. Delacroix, L. Domine, E. Kelderman, S. Little, A. Loeb, E. Masson, W. A. Watters, A. White, "The deployment of a geomagnetic variometer station as auxiliary instrumentation for the study of Unidentified Aerial Phenomena" (preprint, July 2025).

The first real result: half a million objects, zero anomalies
  • DOCUMENTED Five months of continuous all-sky IR observation produced ~500,000 reconstructed object trajectories. Roughly 16% were flagged as statistical outliers on a trajectory-sinuosity metric. After manual review, 144 trajectories remained ambiguous — described by the authors as "likely mundane objects" that could not be resolved with the available data.
  • DOCUMENTED Zero confirmed anomalies. The paper reports no object exhibiting performance inconsistent with known aircraft, birds, balloons, insects or instrument artifacts. It sets a 95%-confidence upper limit of 18,271 outliers over the five-month interval.
  • DOCUMENTED This is the most scientifically valuable thing the field has produced in years — and it is a negative. A published upper limit is a real constraint. It gives the next experiment something to beat. Loeb's group published it anyway, which is the behaviour of people doing science rather than advocacy.
StageCountWhat it means
Trajectories reconstructed (5 months)~500,000Aircraft, birds, insects, balloons, debris
Flagged as sinuosity outliers~16%Automated first pass, deliberately over-inclusive
Ambiguous after manual review144Insufficient data to classify — not "anomalous"
Confirmed anomalous0Nothing exceeded known-object performance
95% CL upper limit on outliers18,271A publishable constraint for future work

← swipe table →

Dominé et al., Sensors 25(3), 783 (2025).

Common misconception

"Harvard is studying UFOs" and "a Harvard study found alien material" are both wrong. The Galileo Project is a privately funded program directed by a Harvard professor and physically hosted on Harvard property. Harvard University has not endorsed its conclusions, and several of its most contested claims — the IM1 spherules in particular — are disputed by other academics in the peer-reviewed literature. Institutional address is not institutional endorsement.

IM1 / CNEOS 2014-01-08 — the claim chain, link by link
  • DOCUMENTED The event. A ~0.45 m bolide entered the atmosphere near Manus Island, Papua New Guinea on 8 January 2014 at 17:05:34 UT, logged in NASA's CNEOS fireball catalogue from US government sensor data.
  • DISPUTED The interstellar claim. Amir Siraj and Avi Loeb argued in 2019 (arXiv:1904.07224, later published in ApJ) that the catalogued velocity — around 60 km/s — implies an unbound hyperbolic orbit, i.e. an object from outside the Solar System. Against: the CNEOS catalogue publishes no uncertainties. Jérémie Vaubaillon argued the anomaly is better explained as measurement error (WGN, Journal of the IMO 50(5):140–143, Nov 2022), and Peter Brown and Jiří Borovička made a detailed case in The Astrophysical Journal (Aug 2023) that uncharacterised velocity errors in the US sensor data undermine the interstellar inference.
  • DOCUMENTED The USSPACECOM memo (April 2022). A memo from US Space Command stated the velocity estimate was "sufficiently accurate to indicate an interstellar trajectory." The memo genuinely exists — this is not folklore. But note exactly what it is: an assertion of accuracy by an office that will not release the error bars, because the sensor performance is classified. Scientists outside the classification boundary cannot check it, and several have said so plainly. A government letter is a document, not a measurement.
  • DOCUMENTED The 2023 expedition. In June 2023 the Galileo Project ran a magnetic-sled survey of the seafloor near Manus Island along the calculated IM1 path, on a reported budget of roughly $1.5M, with press coverage identifying the principal funder as entrepreneur Charles Hoskinson. Approximately 850 sub-millimetre metallic spherules (0.1–1.3 mm) were recovered.

CNEOS Fireball and Bolide Data, NASA/JPL. Siraj & Loeb, arXiv:1904.07224; published as "A Meteor of Apparent Interstellar Origin in the CNEOS Fireball Catalog," ApJ 939, 53 (2022). US Space Command memorandum, April 2022. Vaubaillon, "Hyperbolic meteors: is CNEOS 2014-01-08 interstellar?", WGN J. IMO 50(5), 2022. Brown & Borovička, "On the Proposed Interstellar Origin of the USG 20140108 Fireball," ApJ, 2023.

The spherules and the "BeLaU" composition claim
  • DOCUMENTED The paper exists and passed peer review. Loeb, S. B. Jacobsen, R. Tagle and ~27 co-authors, "Chemical Classification of Spherules Recovered From the Pacific Ocean Site of the CNEOS 2014-01-08 (IM1) Bolide," accepted in Chemical Geology (2024; preprint arXiv:2401.09882). Analysis by micro-XRF, electron probe microanalysis and mass spectrometry.
  • DISPUTED The headline finding. About 10% of the spherules form a class the authors label "BeLaU," enriched in beryllium, lanthanum and uranium by up to three orders of magnitude relative to CI chondrites. Loeb's group interprets this as a "highly differentiated, extremely evolved composition of an unknown source," possibly the magma ocean of a planet outside the Solar System. Against: every step of that inference has been contested in print — see the next card.
  • UNSUPPORTED "Alien technology" framing. No published analysis, including Loeb's own, claims the spherules are manufactured. Popular coverage that translated "unusual composition" into "alien technology" went well beyond anything the paper asserts. Even if the BeLaU spherules were extrasolar natural material, that would be a landmark geochemical result and nothing more.
Important distinction

Four separate claims are routinely collapsed into one: (1) the bolide was interstellar; (2) the search was in the right place; (3) the spherules came from that bolide; (4) the spherules are extrasolar. Each requires independent support. Claim 1 is disputed, claim 2 has been directly challenged by seismic and infrasound reanalysis, claim 3 depends on claim 2, and claim 4 is disputed on geochemistry. A chain is only as strong as its weakest link, and here every link is contested.

Loeb et al., "Chemical Classification of Spherules Recovered From the Pacific Ocean Site of the CNEOS 2014-01-08 (IM1) Bolide," Chemical Geology (2024), arXiv:2401.09882.

The criticism — including the seismometer that was a truck
  • DOCUMENTED The seismic signal used to localise the impact was spurious. Benjamin Fernando (Johns Hopkins), Pierrick Mialle, Göran Ekström, Constantinos Charalambous, Steve Desch, Alan Jackson and Eleanor Sansom re-analysed the Manus Island station data in Geophysical Journal International (2024; preprint arXiv:2403.03966). They concluded that "both previously-reported seismic signals are spurious" — the signal attributed to the fireball is consistent with ground vibration from vehicle traffic on a nearby road, not an atmospheric event.
  • DOCUMENTED Infrasound puts the fireball ~170 km away from where they searched. The same paper used acoustic data from the CTBTO infrasound network to relocate the event roughly 170 km from the reported position — well outside the surveyed seafloor corridor. Their conclusion: the recovered material is "almost certainly unrelated to it, and is likely of more mundane (non-interstellar) origin."
  • DOCUMENTED Coal ash reproduces the anomalous chemistry. Patricio A. Gallardo, "Anthropogenic Coal Ash as a Contaminant in a Micro-meteoritic Underwater Search," Research Notes of the AAS 7(10), Oct 2023. Comparing the reported Be/La/U levels against public coal-composition databases, he found "concentrations are within expectation for all elements" for coal fly ash — a century of coal-burning ships dumped exactly this material along Pacific shipping lanes.
  • DISPUTED Iron isotopes. Steve Desch and Alan Jackson (Arizona State) argued in a November 2023 critique that the iron isotopic ratios in the spherules indicate Solar System origin at greater than 99.995% probability. Desch followed with "Be,La,U-rich spherules as microtektites of terrestrial laterites: What goes up must come down" (submitted to Meteoritics & Planetary Science, March 2024), proposing they are Australasian microtektites derived from terrestrial lateritic rock. Loeb's group replies in a 2025 RNAAS note (Hyung, Levy, Cook, Jacobsen, Loeb, Squire, Farkas) directly comparing BeLaU spherules with Australasian tektites and microtektites and reporting distinct elemental patterns.
  • DOCUMENTED Loeb contested the relocation. "Peak-Brightness Localization of the CNEOS 2014-01-08 (IM1) Fireball" (arXiv:2403.07696, March 2024) argues the search corridor was correct. The exchange remains open; the dispute has not been closed by an independent third-party laboratory analysis of the recovered material.
CritiqueWhoWhere / whenLoeb group's response
Velocity anomaly is measurement errorVaubaillonWGN J. IMO 50(5), Nov 2022Cites the 2022 USSPACECOM accuracy memo
CNEOS velocity errors uncharacterisedBrown & BorovičkaApJ, Aug 2023Same; error bars remain classified
Spherule chemistry matches coal fly ashGallardoRNAAS 7(10), Oct 2023Argues BeLaU pattern is distinct from ash
Fe isotopes → Solar System originDesch & JacksonPreprint critique, Nov 2023Contested in Chemical Geology paper
Spherules are terrestrial laterite microtektitesDeschMAPS submission, Mar 2024Hyung et al., RNAAS, 2025 comparison
Seismic signal was road traffic; site off by ~170 kmFernando et al. (7 authors)Geophys. J. Int., 2024arXiv:2403.07696 defends localisation

← swipe table →

Fernando et al., "Seismic and acoustic signals from the 2014 'Interstellar Meteor'," Geophysical Journal International (2024), arXiv:2403.03966. Gallardo, "Anthropogenic Coal Ash as a Contaminant in a Micro-meteoritic Underwater Search," RNAAS 7(10) (2023). Desch & Jackson, critique of arXiv:2308.15623 (Nov 2023). Desch, "Be,La,U-rich spherules as microtektites of terrestrial laterites," arXiv preprint (Mar 2024). Hyung et al., RNAAS 9(10) (2025). Loeb, "Peak-Brightness Localization of the CNEOS 2014-01-08 (IM1) Fireball," arXiv:2403.07696.

Being fair to Loeb — and noting the stakes on both sides
  • DOCUMENTED What he gets right. Loeb's central methodological argument — that the correct response to poor UAP data is to build better instruments rather than to argue about bad footage — is simply correct, and almost nobody else with his credentials was willing to say it out loud. The Galileo Project's published null result is a genuine contribution. Searching the seafloor for a meteorite is a normal, if expensive, thing to do.
  • DOCUMENTED Where he has a stake. Loeb has authored two trade books on the subject — Extraterrestrial (2021) and Interstellar (2023) — publishes a high-volume stream of public essays, and raises private money on the strength of the search. A hypothesis that generates headlines is worth more to a privately funded program than one that does not. This does not make him wrong; it does mean his priors should not be taken as neutral.
  • DOCUMENTED The 3I/ATLAS episode is instructive. After the third interstellar object was discovered on 1 July 2025, Loeb co-authored "Is the Interstellar Object 3I/ATLAS Alien Technology?" with Adam Hibberd and Adam Crowl (July 2025). The paper appeared in the first issue of a brand-new, obscure journal rather than a mainstream planetary-science venue. Subsequent observations by JWST, Hubble and the VLT characterised 3I/ATLAS as an unusually CO₂-rich comet with a normal coma. The pattern — dramatic technological hypothesis first, mundane resolution later — is one readers should weight when the next claim arrives.
  • DOCUMENTED The skeptics are not disinterested either. Desch and Jackson have been publicly combative, and the exchange has at points been more heated than the evidence warrants. Prominent debunkers in this field frequently have books, media platforms or subscription revenue tied to the debunking. Apply the same discount in both directions: check the argument, not the temperament.
  • DOCUMENTED The instrument-vendor overlap is real. Galileo Project co-founder Frank Laukien heads Bruker, whose analytical instruments feature in the materials work; a Bruker-affiliated scientist is a co-author on the spherule paper. This is common in geochemistry and is not evidence of misconduct — but it is a disclosure a reader is entitled to have.

Hibberd, Crowl & Loeb, "Is the Interstellar Object 3I/ATLAS Alien Technology?" (July 2025). Cordiner et al., "JWST detection of a carbon dioxide dominated gas coma surrounding interstellar object 3I/ATLAS", ApJL (2025). Loeb, Extraterrestrial (Houghton Mifflin Harcourt, 2021); Interstellar (2023).

UAPx — the field expedition that published its own null
  • DOCUMENTED Who they are. UAPx pairs Navy veterans associated with the 2004 Nimitz incident (including Gary Voorhis and Kevin Day) with academic physicists — principally Matthew Szydagis of the University at Albany, SUNY, whose day job is on dark-matter detection experiments.
  • DOCUMENTED The expedition and the paper. A July 2021 field campaign off Catalina Island, California deployed visible-light, infrared and ionising-radiation sensors and collected 600+ hours of footage. Published as M. Szydagis, K. H. Knuth, B. W. Kugielsky, C. Levy, "Initial Results From the First Field Expedition of UAPx to Study Unidentified Anomalous Phenomena," Progress in Aerospace Sciences 156, 101099 (2025).
  • DOCUMENTED The result was one unresolved item, honestly reported. After ruling out mundane explanations for the bulk of candidates, a single residual remained: a dark spot in visible/near-IR possibly coincident with an ionising-radiation count, in the early morning of 16 July 2021. The authors do not claim it is anomalous. Much of the paper is methodological — explicit quantitative thresholds set in advance to constrain confirmation bias.
  • DOCUMENTED Note the media stake. The same expedition was filmed for the commercial documentary A Tear in the Sky (2022). A film needs a result; a paper does not. The paper is the artefact to judge, and it is considerably more restrained than the film.

Szydagis, Knuth, Kugielsky & Levy, "Initial Results From the First Field Expedition of UAPx to Study Unidentified Anomalous Phenomena," Progress in Aerospace Sciences 156, 101099 (2025); preprint arXiv (Dec 2023, rev. June 2025).

The Sol Foundation and the Scientific Coalition for UAP Studies
  • DOCUMENTED Sol Foundation. Co-founded in 2023 by Garry Nolan — Rachford and Carlota A. Harris Professor of Pathology at Stanford — and anthropologist Peter Skafish, with an inaugural symposium at Stanford in November 2023 and annual conferences since. Its emphasis is as much on policy, governance and social-science implications as on physical evidence, which distinguishes it from the Galileo Project's instrument-first posture.
  • TESTIMONY Nolan's anomalous-materials claims. Nolan has stated in talks and interviews that he was approached to analyse materials and medical injuries associated with alleged UAP encounters. His main peer-reviewed output here — Nolan, Vallée, Jiang & Lemke, "Improved instrumental techniques, including isotopic analysis, applicable to the characterization of unusual materials with potential relevance to aerospace forensics," Progress in Aerospace Sciences (2022) — describes methods and reports that most samples examined were ordinary, with anomalies attributable to manufacturing processes. The stronger claims Nolan makes verbally are not matched by published data.
  • RESOLVED The Atacama skeleton is the honest precedent. Nolan led the genomic analysis of "Ata," a small mummified specimen from Chile widely promoted as alien. His team's 2018 Genome Research paper identified it as a human female fetus or neonate with multiple skeletal mutations. He applied the method, got a mundane answer, and published it — which is exactly the behaviour that makes his other work worth taking seriously.
  • DOCUMENTED Nolan's commercial stake. He has founded multiple biotech companies — Rigel, Nodality, BINA Technology (acquired by Roche), Apprise and Akoya Biosciences. Publicly, his UAP profile and his scientific-entrepreneurial profile reinforce one another. Note it and move on.
  • DOCUMENTED SCU. The Scientific Coalition for UAP Studies is a donor-funded US 501(c)(3), formed in 2017, running annual conferences and publishing case reports and analyses openly (via Zenodo). Its best-known outputs are analyses of the 2004 Nimitz encounter and the 2013 Aguadilla, Puerto Rico infrared video, plus Knuth, Powell & Reali, "Estimating Flight Characteristics of Anomalous Unidentified Aerial Vehicles," Entropy 21(10), 939 (2019).
  • DISPUTED The SCU kinematics papers. For: SCU argues that reported accelerations and speeds, if the sensor data are taken at face value, exceed known aircraft performance. Against: the Entropy analysis derives extreme accelerations from witness recollection and uncalibrated video of unknown range — inputs whose uncertainties dominate the output, making the derived g-forces essentially unconstrained. Separately, analyst Mick West has argued the Aguadilla object's motion is consistent with a pair of drifting lanterns; SCU disputes this. Neither side has produced a decisive measurement, because the source data were never calibrated.
  • DOCUMENTED SCU is also critical of the government. In May 2026 it issued a public statement that the first tranche of the PURSUE government file release "falls short of scientific transparency" — a useful reminder that "pro-UAP" organisations are not uniformly credulous about official disclosures.

Nolan, Vallée, Jiang & Lemke, "Improved instrumental techniques, including isotopic analysis, applicable to the characterization of unusual materials with potential relevance to aerospace forensics," Progress in Aerospace Sciences (2022). Bhattacharya, Nolan et al., "Whole-genome sequencing of Atacama skeleton shows novel mutations linked with dysplasia," Genome Research (2018), on the Atacama specimen. Knuth, Powell & Reali, "Estimating Flight Characteristics of Anomalous Unidentified Aerial Vehicles," Entropy 21(10), 939 (2019). SCU publications archive, explorescu.org / Zenodo.

The astronomers' question: five billion cameras, and the pictures are still terrible

This is the strongest single argument in the entire field, and it cuts against the extraordinary hypotheses. If something structured and physical is regularly present in Earth's atmosphere, the observing capacity now pointed at the sky is enormous — and the evidence has not improved.

  • DOCUMENTED All-sky surveys are now industrial in scale. The Vera C. Rubin Observatory achieved first light on 23 June 2025 and began full LSST survey operations in mid-2026: an image every ~40 seconds, ~18,000 deg² of southern sky revisited every few nights, and up to 10 million transient alerts per night issued within 60 seconds. Nothing like this has ever existed before.
  • DOCUMENTED But the caveat is real and cuts both ways. LSST is optimised for faint, effectively-infinite-distance astronomical sources. Near-field fast movers saturate, trail across the frame and are systematically rejected by the pipeline as artifacts. Absence of UAP in LSST alerts is weak evidence, because LSST is built to throw exactly that class of object away. The honest statement is that the instrument is not well-matched to the question — which is precisely the gap the Galileo Project was designed to fill.
  • DOCUMENTED Satellite constellations now manufacture sightings. Several thousand Starlink satellites — the largest constellation ever flown — plus competitors, routinely produce "trains" of lights, deorbit fireballs and specular flares that generate large clusters of public reports. A meaningful and growing fraction of new UAP reports have this prosaic origin, and reporting volume no longer tracks phenomenon rate.
  • DOCUMENTED Why phone footage does not help. The optical physics is unforgiving: tiny sensors, short focal lengths, aggressive computational denoising and sharpening, rolling shutters that smear fast motion, and autofocus that hunts on point sources at night. These produce lens flare, out-of-focus bokeh shapes (often triangular or hexagonal, matching the aperture blades), and rolling-shutter distortion — artifacts that look anomalous and are not. More cameras of this kind do not converge on better data; they converge on more artifacts.
  • DOCUMENTED The one dedicated instrument that has run long enough to publish reports zero anomalies. Half a million tracked objects, 144 unresolved, none anomalous. That is currently the best-quality empirical answer anyone has to the question, and it constrains the extraordinary hypotheses more than any single video ever will.
Observing capacityScaleWhy it still doesn't settle the question
Rubin Observatory / LSST~10M alerts/night, 18,000 deg²Pipeline rejects near-field trailed objects by design
Smartphone camerasBillions worldwideShort focal length + computational processing = artifacts, not resolution
Satellite constellationsThousands of objectsGenerate sightings rather than resolve them
Military IR/radar sensorsContinuous, wide-areaOptimised for other targets; artifacts are common; data mostly classified
Galileo Project observatories2 sites, all-sky IR + magnetometerPurpose-built and calibrated — and reports zero anomalies so far

← swipe table →

Vera C. Rubin Observatory / LSST public survey documentation, NSF–DOE, 2025–2026; Rubin Observatory First Look (23 June 2025). Dominé et al., Sensors 25(3), 783 (2025).

Common misconception

"Scientists refuse to study this." They do study it — the problem is what they find. The Galileo Project, UAPx and SCU have all now published in peer-reviewed venues, and the two efforts that ran calibrated instruments for extended periods reported, respectively, zero confirmed anomalies out of half a million tracked objects, and one unresolved ambiguous event out of 600 hours. The obstacle is not institutional cowardice. It is that when you point good instruments at the sky and analyse the output honestly, the anomalies mostly stop being anomalous — and the field's most persuasive proponents are the ones publishing that inconvenient result.

Important caution

Nothing in this section should be read as settling whether unexplained aerial objects exist. It addresses a narrower question: whether the specific scientific claims made about them so far have held up. Unexplained means unexplained. A residual of 144 unclassifiable IR tracks is not evidence of anything beyond the limits of an eight-camera array — and neither is it a licence to fill the gap with a hypothesis the data cannot carry.

Documents & sources — go read them yourself

Links verified August 2026. Unlinked items in this section are ones for which no stable public URL could be confirmed.

Plates — the documentary record
The Very Large Array
The Very Large ArrayThe Karl G. Jansky Very Large Array in New Mexico, the model for what calibrated, publicly archived observational data looks like. The contrast with UAP data collection is the central methodological complaint of the scientific critics.National Radio Astronomy Observatory / NSF · Public domain · source
1I/'Oumuamua (Artist's Impression)
1I/'Oumuamua (Artist's Impression)An artist's impression of the first confirmed interstellar object to pass through the solar system, whose unexplained non-gravitational acceleration prompted competing natural and artificial hypotheses. It is the clearest test case for how anomaly claims should be argued in the literature.ESO / M. Kornmesser · CC BY 4.0 — attribution required: 'ESA/Hubble, NASA, ESO, M. Kornmesser, CC BY 4.0' · source

Images are public domain or freely licensed; each carries its credit and licence, and links to its source record. Served from this site rather than hotlinked.

UFO.jelia.nyc — The Record Room. Last surveyed 8 August 2026.

Every exhibit carries a stamp for the kind of evidence behind it. Documents are named and linked so you can read them yourself. Where qualified people disagree, both positions are shown. Where something has been explained, the explanation stays on display — including for cases once presented as unexplainable.

Surveyed and kept by the house of jelia.nyc.