Transients in the Palomar Observatory Sky Survey may be associated with nuclear testing and reports of unidentified anomalous phenomena

Brief

Bruehl and Villarroel (2025) merged a 107,875-object POSS-I transient catalogue, dates of 124 above-ground nuclear weapons tests, and daily UAP counts from the UFOCAT database into a single 2,718-day dataset spanning November 1949 to April 1957, more than five months before Sputnik. Transients, defined as point sources absent from all subsequent survey images and lacking counterparts in Gaia DR3 or PanSTARRS DR1 within 5 arcsec, occurred on 310 days (11.4%); 15.6% of nuclear-window days carried at least one transient versus 10.8% of control days (RR=1.45, 95% CI 1.10–1.90, p=0.008), with the strongest single-day signal appearing one day after a test. A generalized linear model with negative-binomial specification found an 8.5% increase in transient count per additional UAP report (Beta=0.081, SE=0.006, p<0.001), with nuclear-window and UAP effects additive: dates meeting both criteria averaged 58.4 transients versus 20.0 for dates meeting neither.

Metadata

Category

Phenomenon

Venue

Scientific Reports

Type

Peer-reviewed

Year

2025

Authors

Stephen Bruehl, Beatriz Villarroel

DOI

10.1038/s41598-025-21620-3

Access

Open access

Length

1.5 M

Programs

VASCO, Project Blue Book

Instruments

Palomar Observatory Sky Survey POSS-I, PanSTARRS DR1

Data sources

POSS-I DSS Plate Finder (STScI), UFOCAT (CUFOS), Gaia DR3, PanSTARRS DR1, U.S. DOE/NNSS nuclear test registry, Wikipedia Soviet nuclear test list, CHRC4 Veterans British nuclear test list

Tags

UAP-statistics, UAP-nuclear, photographic-transients, SETI, technosignature, VASCO

Key points

• Transients were 45% more likely (RR=1.45, 95% CI 1.10–1.90, p=0.008) on dates within a ±1-day nuclear-test window compared to all other dates.p.2
• The only temporally-resolved window reaching significance was the day after a nuclear test: 18.5% of those days recorded a transient versus 11.0% of control days (p=0.010, RR=1.68).p.2
• On the 310 days with at least one transient, Spearman's rho=0.138 (p=0.015) linked daily transient count to daily independent UAP report count.p.4
• GLM negative-binomial analysis: Beta=0.081, SE=0.006, p<0.001, Exp(B)=1.085, an 8.5% increase in transients per additional UAP report.p.4
• Nuclear-window and UAP effects are additive: estimated marginal means are 20.0 (neither), 40.6 (one condition), and 58.4 transients (both), with all pairwise differences significant at p<0.001 and non-overlapping 95% CIs.p.5
• No transient was observed within a nuclear-test window after March 17, 1956, despite 38 additional above-ground nuclear tests in the subsequent 13 months of the study period.p.5
• UAP reports were significantly higher within nuclear-test windows (5% trimmed mean 3.68) than outside (3.31; Mann-Whitney U=447,057, p=0.008), a relationship the authors describe as not previously reported in the peer-reviewed literature.p.4
• The 107,875-transient dataset spans all POSS-I E Red plates from the DSS archive; objects were excluded if any counterpart existed in PanSTARRS DR1 or Gaia DR3 within 5 arcsec.p.7

Verbatim

• “Thus, a transient was 45% more likely to be observed on dates within a nuclear test window (day of test + /- 1 day) compared to dates outside of a nuclear test window.”

  p.2

• “Transients were observed on 18.5% of days that were 1 day following a nuclear test, whereas transients were noted on only 11.0% of days not meeting this criterion.”

  p.2

• “Other analyses examining the full sample indicated that for every additional UAP reported on a given date, there was an 8.5% increase in number of transients observed on that date.”

  p.6

• “Our findings do not definitively indicate what transients are nor do they necessarily imply causal associations.”

  p.6

• “The possibility that some transients may represent UAP events in orbit captured on photographic plates prior to the launch of the first artificial satellite cannot be ruled out.”

  p.6

• “To our knowledge, this statistical association has not previously been reported in the peer-reviewed literature, although it is consistent with anecdotal reports of such associations”

  p.4

Most interesting

• 88.5% of the 2,718 study days recorded zero transients; the distribution was so right-skewed (skewness=10.35, variance=28,938.64) that the median daily transient count across the full dataset was zero.
• The POSS-I plates from July 19 and July 27, 1952, both dates of the Washington D.C. radar-visual UAP event, each contain multiple bright transients that are absent from a blue-band image taken immediately afterward and from all subsequent surveys.
• All transients appear as discrete point sources rather than streaks despite the 50-minute exposure time, which the authors argue rules out sustained atmospheric phenomena as an explanation, since such effects would have to remain localized in one position for approximately 24 hours to produce a point source.
• A prior study (cited as reference 10) found UAP activity at U.S. nuclear production facilities peaked through 1952, then dropped sharply in 1953 and stayed low through 1975, the same mid-1950s inflection at which the transient-nuclear association in this dataset also weakens.
• The UFOCAT database recorded at least one UAP report on 89.3% of the 2,718 study days, making binary UAP presence statistically near-useless as a predictor and requiring continuous count-based models for meaningful analysis.
• Nuclear fallout contamination of photographic plates, an alternative explanation the authors consider, produces diffuse fogged spots, not discrete point sources with stellar point spread functions, which the authors use to argue against plate-contamination as the cause of the transient-nuclear association.