Beatriz Villarroel

Villarroel states she personally believes that tic-tac craft, non-human biologics, and interdimensional-being claims are true, based on trusting unnamed 'really credible people' who discuss crash retrievals, while stating that no such evidence is directly accessible to her as a scientist.

I believe it's true... there's I believe that what is being said there is most I mean I believe it's true

The hypothesis that massive stars can collapse directly into a black hole without a visible supernova motivated Villarroel's original 2016 search for vanishing stars.

there's a a class of uh like um uh massive stars that are believed that they might be able to collapse into a black hole without emitting any supernova

Villarroel states that when she and two bachelor's students first tested the vanishing-star search on about 1% of a US Naval Observatory sample in 2016, the full network/project did not yet exist; it was formally organized with two senior scientists in 2017.

in 2016 I had together with two bachelor students tried to do a small search in a like 1% of the sample... but then they said okay now we need to do the whole sky

VASCO's automated pipeline analyzed about 600 million objects across the northern hemisphere comparing 1950s Palomar plates to modern surveys and found zero genuinely vanishing stars.

we actually went through something like uh 600 million objects... and we found zero vanishing stars

Our check Confirmed: Villarroel's VASCO project, using an automated cross-matching pipeline (associated with Enrique Solano) compared roughly 600 million objects from the USNO B1.0 catalog against Pan-STARRS DR1, flagging about 150,000 candidates lacking a modern counterpart (about 106,000 of them in the northern hemisphere). Despite this large-scale search, the project's original goal of finding a genuinely vanishing star (one that disappears with no trace and no plausible mundane cause) was not achieved; no such object was confirmed.

Sources astro.uu.se, arxiv.org

While searching for vanishing stars, VASCO found one plate image containing nine star-like transient objects appearing and vanishing simultaneously within a small patch of sky.

one of these images had nine objects appearing and vanishing at the same time but in a small image

Across large samples, VASCO transients show a statistically narrower, sharper brightness profile than real stars of matched magnitude, consistent with a very short flash rather than a steady point source.

if you take a big sample of these transients and you compare to a big sample of the stars you see that they are slightly... sharper and more narrowed

Villarroel's team pointed a large optical telescope in the Canary Islands, using a colleague's rapid-light-burst detector technology, at transient locations to test for active laser/artificial-signaling activity, and found nothing, which she says takes that hypothesis 'totally off the table.'

we actually even use the largest telescope in the world, optical telescope in the world in Caner Islands, pointed at those things and we found nothing there. So that hypothesis is is totally off the table

VASCO transients show a statistically significant deficit specifically within the geometric shadow of the Earth (42,000 km altitude, ~8.7 degree radius), which Villarroel calls her strongest single piece of evidence because atmospheric scattering of cosmic rays could not track that precise geometry.

the earth shadow def the visit I think is really beautiful to work with and uh I mean we see a big deficit of this transis in the earth shadow

VASCO transients are found about 68% more often within plus or minus one day of a nuclear weapons test, and this correlation gets stronger, not weaker, after a machine-learning pass removes suspected plate defects; the correlation is strongest one day before the test date.

like 68% more transient within plus minus one day actually the new paper that we have now the machine learning paper led by Steven Brule

A researcher (name transcribed 'Evo Busco') posted a paper about a week before this interview showing that a significant fraction of VASCO transients carry the same optical aberrations (coma-shaped distortions from imperfect telescope optics) as real stars on the same plates, indicating the light genuinely passed through the telescope rather than being a plate/emulsion defect.

he actually essentially shows that some of these transients have to be real... he also shows that they come from light passing through the optics

VASCO's main candidate sample totals 107,000 transients, mixed with plate defects, bubbles, hairs, and cosmic rays, but a significant fraction pass through the telescope optics and specifically vanish in the Earth's shadow.

among 107,000 uh candidates there's a lot of dirt there... but the important part is that a significant fraction of this population... pass through are produced by light that pass through the telescope and they are vanishing in the earth's shadow

A retired nuclear engineer (name transcribed 'Kevin Khn') independently reanalyzed the VASCO sample and found an anti-correlation between transients and geomagnetic storm activity, which Villarroel says contradicts the cosmic-ray explanation.

another independent scientist a nuclear engineer... called Kevin Khn. He played around with the sample and discovers an anti-correlation with geomagnetic storm activity

Villarroel says Loeb's cosmic-ray argument originally predicted a positive correlation between transients and solar activity, but after she pointed out the data showed the opposite, he revised his essay to argue for an anti-correlation caused by coronal mass ejections, without having disclosed that revision when they last spoke.

he was expecting a correlation and then I pointed it out in his medium essay and then he changed the essay and then he expected suddenly an anti-correlation

A peer-reviewed paper by a team at Edinburgh (transcribed 'Hambly and Bear') examined nine VASCO transients and concluded they were most likely photographic emulsion flaws, based on their profiles being narrower than typical stars.

there's a second paper by this Hambbley and Bear uh team as I understand at Edinburgh... peer-reviewed... they looked at nine of the transients and they said... they were attributable to what are called emulsion flaws

Villarroel says a microscope re-examination of the same nine transients that Hambly and Bear attributed to emulsion flaws still shows eight or nine of them as round and sharper than typical defects, so the narrowness argument alone cannot separate real short flashes from plate defects without population-level statistics.

if you look at them with microscope, you again this like this nine or eight out of the nine transients are looking like they're round. They're slightly sharper and rounder

Villarroel's leading interpretation of the combined evidence is that there is something flat and reflective in orbit around the Earth, i.e., the transients are most likely artificial rather than natural.

I think there is something very flat and very reflective in orbit around the earth... I find it hard not to attribute it to something artificial

A researcher (name transcribed 'Brian Doerty') found that restricting the sample to sunlit transients strengthens the nuclear-test correlation, and that the transients avoid the ecliptic plane, unlike asteroids; a new, unpublished VASCO paper finds the transients clustering around the celestial equator with size and feature statistics matching modern-day satellite/space-debris detections (about 1 per square degree per hour historically vs. 1.7 today).

if you actually focus on the sunlit transients and remove the others, the nuclear correlation gets even stronger... he also sees that they are avoiding the ecliptic

A preprint by Waters et al. claims that about 91% of VASCO's reported transient features belong to a data set that was never properly distinguished from catalogued astronomical objects, due to skipped scan-artifact-removal steps.

Waters estimates 91% of those features belong to a set that wasn't distinguished from catalog object

Villarroel rebuts Waters et al. by saying they analyzed the wrong dataset: a 5,000-transient sample from Solano et al. 2022 that was specially filtered to remove anything appearing in any catalog across the electromagnetic spectrum, which creates a large gap in sky coverage that by itself produces an apparent excess in the earth-shadow region; she also says that dataset lacks observation timestamps and, despite being called 'cleaner,' is not actually cleaner, just 20 times smaller.

they used a sample of 5,000 uh transients reported in Solano at all 2022... a big hole... their sample isn't cleaner. is just 20 times smaller

Our check Partially corroborated: a critique paper (Watters et al., referenced in the field as arguing the nuclear-test correlation is fully explained by the Palomar telescope's observation schedule) does exist, and the VASCO dossier separately documents general 'garbage-in-garbage-out' and selection-bias objections to the 2025 statistical papers, consistent with the flavor of Villarroel's pushback described. However, the specific technical details in this claim (that the disputed dataset is Solano et al. 2022's ~5,400-source catalog, that it is 20 times smaller and lacks timestamps, and that its filtering specifically creates the earth-shadow gap) could not be independently confirmed within the research budget; searches found the existence of the debate but not a direct primary-source statement matching these exact figures.

Sources arxiv.org, arxiv.org

Villarroel's team tested Waters et al.'s claim that the nuclear-test correlation is an artifact of not normalizing for the number of telescope observation nights, and found the correlation persists and gets stronger after correcting for it.

we have tested that and addressed it in the commentary that's not even if you uh normalize it correctly you still have a actually a correlation it stays and It even gets stronger

Villarroel says she and colleagues started a 'European Crash Retrieval Initiative' that solicits tips from the public about crashed craft.

We even we even have this European crash retrieval initiative that we started where we have been uh requesting the public to give us uh tips for uh crashed saucers

Villarroel states that about 97% of all UFO reports are misidentifications, but roughly 3% cannot be explained even with full access to the available data.

even if most of it, like 97% of all the UFO reports are just like misidentifications, you still have some 3% that can't be explained

Villarroel cites the Nimitz 2004 encounter and the 'Washington flap' as highly credible, still-unexplained historical UAP cases.

you still have cases like the Nimits case that are highly fascinating. Uh you still have the Washington flap that you can't just explain away just like that

Villarroel says the most evidentially compelling UAP cases get classified once they become 'really good,' which prevents scientists like her from accessing the underlying data.

The problem again with all these really interesting highquality cases is that it's classified. Once it's something is really good, it gets classified and we scientists can't access it

Villarroel says that objects producing the VASCO transient signature are still visible in the sky today, with thousands detectable by the naked eye every hour, as her answer to why such objects would not still be present.

I see lots of these things in the sky still. I mean there are papers that are reporting the same kind of flashes... today when you look at the sky you see thousands and thousands of these kind of transients

Villarroel states she has experienced 'a number of incidents' related to her research and has some form of 'support' regarding safety concerns, without giving further detail.

I don't want to explain all the details, but there have been a number of incidents and uh it's um it's let's say I do have support for it