This arXiv preprint (not yet peer-reviewed) presents an independent replication of two unusual findings first reported in 2025 papers by Bruehl and Villarroel and Villarroel et al. The study examines transient sources on Palomar Observatory Sky Survey (POSS-I) photographic plates taken between 1949 and 1958. Transients are defined as objects appearing on a single exposure but absent from modern digital sky surveys and adjacent plates.

Methodology: The author used the exact original dataset provided by the prior researchers. Statistical approaches included chi-square contingency analysis on the full catalog of roughly 108,500 transients, negative binomial regression controlling for precipitation, lunar illumination, and cloud cover, and a 10,000-iteration permutation test to assess temporal specificity. For the second finding, transients were classified by position relative to Earth's umbral shadow cone at geosynchronous orbit altitude (approximately 36,000 km). Sample size is large but derived from historical analog plates, which introduces inherent limitations such as variable image quality and potential selection biases in plate scanning.

Key results: The analysis reproduced a relative risk of 1.45 (p=0.011) for transients during periods of atmospheric nuclear weapons tests, with the incidence rate ratio rising to 3.98 when limited to sunlit conditions. The Earth-shadow deficit was also confirmed: only 0.46% of transients appeared in the shadow cone versus a geometric expectation of about 1.4%. All detected transients predate Sputnik 1 in 1957.

What original coverage missed: Early reports focused heavily on the statistical surprise but underplayed the historical context of overlapping Cold War nuclear testing campaigns (such as Operations Greenhouse and Ivy in 1951-1952) with POSS-I observations. They also gave limited attention to the sunlit-only strengthening of the signal, which suggests a possible illumination-dependent phenomenon at high altitude. This replication strengthens reproducibility claims but still cannot prove causation; controls for weather reduce but do not eliminate the possibility of undetected confounders.

Synthesizing with related work: When viewed alongside Villarroel’s earlier searches for vanishing stars using similar plates (e.g., arXiv:1906.00980 on century-long variability surveys) and declassified atmospheric nuclear effects studies from the Los Alamos archives, a broader pattern emerges. The shadow deficit implies sources or reflections originating beyond low-Earth orbit yet before human satellite launches. This raises questions about unknown upper-atmospheric or near-geosynchronous optical phenomena triggered or amplified by nuclear events, perhaps analogous to but distinct from later high-altitude nuclear explosions like Starfish Prime (1962).

Genuine analysis: The validated correlations highlight an opportunity for new anomaly-detection methods using archival astronomical data as an unintended monitor of historical events. While extraordinary claims require extraordinary evidence, the independent confirmation addresses a key reproducibility crisis in transient astronomy. Limitations remain significant: the work relies on a single dataset, uses pre-digital technology, and describes correlation rather than mechanism. Future digital surveys of additional plate archives could test whether these patterns hold beyond POSS-I. The findings invite interdisciplinary collaboration between astronomers, atmospheric physicists, and historians of science to explore whether nuclear detonations produced previously unrecognized optical side-effects detectable from ground-based telescopes.