Estados Unidos / órbita terrestre baja.Skylab · light flashes and reddish object (NASA, PURSUE)

Skylab, launched on May 14, 1973, was the United States' first space station. Between 1973 and 1974 three successive crews occupied it, and document NASA-UAP-D7 —declassified on May 8, 2026 in the first PURSUE release— gathers excerpts from the technical debriefings of all three in which the astronauts describe observations of unidentified phenomena. They are three distinct kinds of observation, and it is worth treating them separately because their nature differs.

The first is the light flashes. In the Skylab 2 debriefing (the first crew, June 30, 1973), science pilot Joseph Kerwin reported: 'We saw light flashes. I think all of us saw them. I saw them most often when I was in the sack at night with my eyes closed but awake. They tended to wax and wane in frequency.' An astronaut seeing flashes with his eyes closed is the unmistakable signature of the 'light flashes' or cosmic phosphene phenomenon: high-energy charged particles —galactic cosmic rays— that cross the eyeball and interact with the retina and optic nerve, producing the perception of light with no external source. The effect, proposed in the 1950s and confirmed during Apollo, is stronger outside Earth's magnetic field; in low orbit Skylab was still exposed to an appreciable flux, especially when crossing the South Atlantic Anomaly.

The second observation is the most striking and, at the same time, the one the witnesses themselves resolved. In the Skylab 3 debriefing (October 4, 1973), science pilot Owen Garriott described an object the crew tracked for about five to ten minutes roughly a week before splashdown: 'This bright reddish object was out there and we tracked it... it was obviously a satellite in a very similar orbit to our own.' In another passage he added that Jack Lousma first noticed 'a rather large red star' out the wardroom window, much brighter than Jupiter, with a reddish hue, and well above the horizon. The description —an object in a similar orbit, followed visually for several minutes— fits a sunlit satellite or rocket stage, and the crew itself identified it as such.

The third, in the Skylab 4 debriefing (February 22, 1974), came from commander Gerald Carr: 'On occasion we saw some lights flashing outside with very definite motion relative to ours. We presumed that they were other pieces of Skylab, or possibly other satellites.' Again, the witnesses themselves offer the prosaic explanation: the station's own orbital debris or satellite traffic, both abundant in low orbit and easy to mistake for anomalous objects when they reflect sunlight against the black background.

The document, then, presents not an unresolved enigma but an honest record: three classes of observation that the crews themselves interpreted, at the time, as a known physiological phenomenon (the flashes), a satellite (the reddish object) and debris or satellites (the flashing lights). Its inclusion in PURSUE follows the program's logic of publishing everything internally cataloged as UAP, not from the material defying explanation.

Skylab's value in the corpus is not evidential but methodological and one of coverage, and it is worth saying plainly: it is a case whose prosaic explanation is robust and, for the most part, was offered by the astronauts themselves at the time. The flashes with eyes closed are the cosmic phosphene phenomenon, documented since Apollo; the reddish object the Skylab 3 crew tracked for minutes was identified by them as 'obviously a satellite'; and the Skylab 4 flashing lights were attributed by the commander to debris or satellites. Hence the posterior concentrates its mass in 'mundano_natural' and the file enters as tier B with low probability: it is not an enigma, it is a record.

That said, its inclusion serves three legitimate functions. First, it completes the Skylab node in the corpus's thread of astronaut observations, alongside Gemini, Apollo 12 and Apollo 17, showing that the same repertoire of phenomena —phosphenes, particles, debris, satellites— recurs mission after mission with a stable explanation. Second, it is a calibration check: a corpus that recorded only the spectacular cases and discarded the well-explained ones would be biased; including Skylab, and openly declaring its three observations prosaic, preserves the discipline of treating evidence by its real strength rather than its allure. Third, it illustrates a feature of the PURSUE system itself: the program publishes everything once internally cataloged as UAP, so the holdings mix, without ranking, genuinely anomalous material with resolved observations; knowing how to tell them apart is precisely the analytical task. Skylab is, in that sense, the useful counterexample: the declassified document that, read carefully, explains itself.