The Case of the Impossible Satellites

by Brian Dunning
June 23, 2026

Sponsor an episode!

In 2025, science news outlets all exploded with what sounded like the story of the century: artificial satellites had been discovered orbiting the Earth in the early 1950s, at least five years before the launch of Sputnik I. Artificial satellites before the existence of artificial satellites could mean only one thing, claimed the authors of the paper: aliens! The evidence was found on photos taken of the night sky at the Mount Palomar Observatory in Southern California between 1952 and 1955. It manifested as what the authors referred to as “transients,” dots that looked just like stars, but had disappeared when the same patch of sky was photographed six days later. Their best explanation was alien spacecraft sitting in geosynchronous orbits, and then leaving.

Moreover, the authors noted that the dates of the transients’ appearances were all shortly after major nuclear tests, and they proposed that these tests may have attracted the aliens to Earth to see what was going on.

It is no exaggeration to say that the popular science media ate this up with a spoon. Some of the headlines:

"Could these mysterious flashes of light in 1950s photos be UFOs? Some researchers think so"

"Mysterious transient objects in sky linked to nuclear testing and unidentified anomalous phenomena"

"Did aliens spy on our nuclear tests? Study finds signs of UFOs near US sites in 1950s"

"Did aliens watch 1950s nuclear tests? 'Maybe,' studies say"

"A Mysterious UFO Sighting Went Unsolved for 70 Years — Now 'Rigorous' Tests May Confirm It, Scientists Say"

"Did Science Just Confirm 1950s UFO Visits?"

Sounds pretty convincing, doesn’t it? Well, it turns out it’s easier to convince a headline editor than it is actual astronomers. Here’s what really happened.

A guy named Stephen Bruehl, who is a clinical psychologist and professor of anesthesiology at the Vanderbilt School of Medicine — and also, apparently, a big fan of UFOs — was at a UFO conference in 2024 and heard a remote talk about these transients by astronomer Beatriz Villarroel, a longtime UFOlogist and board member at the The Sol Foundation, which is dedicated to promoting the belief that aliens visit the Earth.

Villarroel was a longtime advocate of this claim, having established the VASCO Project years before to study the Palomar transients, and also advocating for the retrieval of crashed UFO debris in Europe, among other alien-related pursuits.

Bruehl reached out to Villarroel, and the two of them agreed to collaborate on two papers advocating their shared belief that the transients were caused by light reflecting off of alien spacecraft in geostationary orbits which had been attracted to the Earth by nuclear testing. Out of more than 107,000 transients that had been detected in the Palomar plates by automated methods previously, Bruehl and Villarroel focused on just nine that they considered the best proof of their hypothesis. The papers were “Transients in the Palomar Observatory Sky Survey (POSS-I) may be associated with nuclear testing and reports of unidentified anomalous phenomena” published in Scientific Reports, and “Aligned, Multiple-transient Events in the First Palomar Sky Survey” published in Publications of the Astronomical Society of the Pacific, both in October 2025.

Well, the popular science media may have been enamored, but the astronomy community was not. The papers were met with essentially zero support, and with massive criticism. Of course the so-called “transients” had always been well known to any astronomers who had worked with photos taken on old glass plates, as there are any number of things that can cause them. Quite a few papers were written refuting Bruehl and Villarroel, but perhaps the most impactful was written actually before their papers were published. In 2023, astronomers Nigel Hambly and Adam Blair of the Royal Astronomical Society published “On the nature of apparent transient sources on the National Geographic Society — Palomar Observatory Sky Survey glass copy plates” which advocated for the most obvious and common explanation: that the transients were simply spurious artifacts of the photographic emulsion, a familiar problem with these old glass plates. Their paper was largely in response to Villarroel’s longtime, ongoing assertions. She had been on her soapbox annoying the astronomy community for a long time, since well before she co-authored with Bruehl when it finally went viral to the mass media audience.

Hambly and Blair identified five key reasons that the transients were not actual objects photographed, but were flaws in the plates:

• All nine of the transients were far too sharp to have represented points of light that filtered through the atmosphere, which inevitably adds fuzziness. A perfectly crisp dot is consistent only with a flaw in the emulsion, and never with an object on the other side of the atmosphere.
  
  
• Hambly and Blair trained a random-forest machine-learning algorithm on known stars, known galaxies, and known emulsion artifacts from outside the image area. All nine transients were rejected as artifacts, with probabilities ranging from 62% to 99%.
  
  
• While Bruehl and Villarroel picked nine dots that were present on the first image but gone on an overlapping image from six days later, Hambly and Blair decided to also go the other direction. They found more than 100 dots that were present on the later image but were not on the first, all with brightness, size, shape, and sharpness comparable to the nine.
  
  
• Villarroel responded to the criticism that these were scanning artifacts by pointing out that they appeared not only on the 1992-1994 scans, but also on a 2001 set of SuperCOSMOS scans made at the Royal Observatory in Edinburgh from a set of the second-generation glass plates (more about those in a moment). This eliminates the possibility that the transients were flaws introduced in the scanning process. Hambly and Blair acknowledge this, but since all the scans were made from second-generation glass plates, it says nothing of the possibility that the flaws were introduced in either the first or second generation photographic replications.
  
  
• It just so happens that the SuperCOSMOS scans were made at Hambly and Blair’s own institution, scanned from their own set of the glass plates; so they were able to directly access the plates and examine the dots microscopically. It was something of a case of Bruehl and Villarroel bringing a knife to a gunfight. Hambly and Blair catalogued a number of obvious emulsion blemishes of various kinds both inside and outside the image area, and found that the nine were perfectly consistent with emulsion holes — tiny specks that did not quite get covered with photographic emulsion.

Hambly and Blair’s work uncovered the one really, really, really obvious problem with Bruehl and Villarroel’s papers, and that’s that they were based on examinations of publicly available digital scans of copies of copies of the original photographic plates, each generation of which was subject to the introduction of new flaws. Extraordinary and exotic explanations involving alien spaceships were proposed — without anyone actually going back and seeing if these transients appeared on the original plates!

Had this work been done by serious astronomers aware of what peer review entails, and not by UFO seekers hungry to support their preferred conclusion, there’s no way anyone would attach their name to such an extraordinary suggestion as these papers make without actually going and looking at the original photographic plates (the ones Edinburgh’s copies were made from), and making absolutely sure that what they see in the scans actually matches what was photographed.

Now I don’t want to leave you with the impression that that’s an easy thing that anyone can just go and do — if it was, I’d have done it myself before writing this episode. But in the course of my initial research, I learned that everyone who should have examined the original glass plates — but didn’t — gets a pass, and I’ll now explain why.

Let’s have a closer look at this glass plate situation. The original observations that Bruehl and Villarroel identified were one in 1952, three in 1954, and one in 1955, depicting nine transients among them. They were taken during a major photographic survey of the northern sky conducted from 1949 to 1958 using the 48-inch Samuel Oschin Schmidt telescope at California's Palomar Observatory, called the Palomar Sky Survey, or POSS-I.

The survey’s images were captured on glass photographic plates, 355 mm square and only 1 mm thick. They were so terrifyingly thin and fragile because each needed to be vacuum-clamped into a curved shape in order for the Schmidt telescope’s image to be properly focused across the entire plate. Each image consisted of one red-sensitive plate and one blue-sensitive plate, which allowed the images to be studied in color. These two plates were exposed one after the other, with each exposure time ranging from 10 to 50 minutes, depending on atmospheric conditions. The plates were then photographically developed in an onsite darkroom, swishing them around by hand in chemical baths — a process which, obviously, introduces numerous opportunities for dust, debris, and flaws; so it was controlled to the highest possible standards.

These originals, which were negatives since that’s how photography works, were far too valuable and fragile to be touched or shipped or shown around. Instead, first-generation positives were created by contact-printing onto new, stronger glass plates using ultraviolet light in a vacuum chamber. These positives were then used the same way to create about 100 second-generation negatives. Some of these second-generation negatives were then shipped to top universities and research institutions; then hundreds of other copy sets were made onto glossy paper from others of the second-generation negatives, and these were sent to everyone else requesting them.

The digital scans used by modern researchers were made between 1992 and 1994 at the Space Telescope Science Institute in Baltimore, Maryland from a set of the second-generation negative glass plates from the 1950s. When this was completed, the scans were made available on a boxed set of 102 CD-ROMs.

Here is what survives today. The original survey consisted of 1,872 images. All of those original thin photographic plates still exist, but they are locked away safely in secure and climate-controlled archives; you’re not getting your grubby paws on those no matter who you are. Two sets of the first-generation positives were made; one was locked away in a deep storage vault as the master backup, the other was kept in Palomar’s workroom and was used to make all the second-generation negatives. Approximately 100 sets of those glass negatives were made for distribution as the premium “Glass Sky Atlas,” of which about 3/4 are believed to still survive; and some 200-300 paper sets were printed, of which an unknown number survive in various states of completion and condition.

If you’d like to view the digital scans yourself, fear not, you don’t need to figure out how to cobble together a CD-ROM drive. You can do an online search of The STScI Digitized Sky Survey hosted by the Space Telescope Science Institute; just make sure you select POSS1-Red or POSS1-Blue to be sure you’re accessing the 1949-1958 survey.

If there was a really important science reason to take a look at either the original negatives or the archived set of positives, access would certainly be arranged. But no important reason existed in this case. A guy who likes UFOs co-wrote a paper with another UFOlogist speculating that aliens were probably attracted by our nuclear tests so they buzzed around the Earth in their UFOs. That’s an extraordinary claim supported by no evidence at all; and moreover, the dots on the scans were very easily explainable by the kinds of anomalies that these types of photographic plates are known to have. So Bruehl and Villarroel, lacking any compelling evidence of anything unusual, would no way have been granted access to the originals.

The astronomers who criticized the Bruehl-Villarroel paper do not bear any burden of proof, so they also have no genuinely important reason to put those originals at risk by pulling them out of storage, nor would refuting a claim so irrational that it doesn’t really need refuting be considered an important enough reason to grant access.

And so we lay the case of the impossible satellites to rest. It’s not that we found anything to disprove the claim; rather it’s simply that the evidence put forth was of too low quality to warrant any effort. Perhaps Villarroel’s efforts to unearth crashed UFO debris in Europe will yield more worthy evidence. Until that time, we remain on the null hypothesis: that flaws in the photographic emulsion are exactly what they appear to be.

By Brian Dunning
Follow @BrianDunning