Published October 17-20, 2025 in Scientific Reports (a Nature journal) and Publications of the Astronomical Society of the Pacific. Lead researchers: Dr. Beatriz Villarroel (Nordic Institute for Theoretical Physics, Stockholm) and Dr. Stephen Bruehl (Vanderbilt University Medical Center). Study: 2,718 days of Palomar Observatory Sky Survey (POSS-I) data, 1949-1957 — all images captured BEFORE Sputnik 1 (October 1957), eliminating any human-made satellite as explanation. Method: Automated analysis of glass photographic plates identified 107,875 transient star-like objects that appeared in one 50-minute exposure and were absent from all subsequent surveys. KEY FINDINGS: (1) Transients were 45% more likely on dates within ±1 day of above-ground nuclear weapons tests (statistical significance: p=0.008). (2) On days with no nuclear test: transients appeared on 11% of days. On the day after a nuclear test: 19% of days. (3) 8.5% increase in transient count for every additional UAP report on that date (p=0.015). (4) On days with BOTH a nuclear test AND UAP sightings — more than twice as many transients as baseline. (5) Shadow test: significant deficit of transients within Earth’s umbra — consistent with solar reflections off orbiting objects, not photographic plate defects. (6) Last transient over a nuclear testing site: March 17, 1956 — one day after the Soviet “Joe 21” test. No transients over nuclear test sites after that date despite continued testing. Villarroel: “From what I see, I cannot find any other consistent explanation than that we are looking at something artificial.” Bruehl: “There’s no easy explanation for what these transients are and why they show up at nuclear tests.” This is the first statistically significant peer-reviewed evidence that something — artificial in nature — was monitoring nuclear weapons testing from orbit BEFORE humanity put anything in orbit. It directly corroborates Malmstrom 1967, Northern Tier 1975, Woomera 1954, and AARO’s 18 nuclear site UAP cases.