Palomar transients form hotspots—if the orbital model is right

Stephen Bruehl, Beatriz Villarroel, and colleagues analyzed 107,875 candidate transient sources found on first-generation Palomar Observatory Sky Survey plates taken before Sputnik. A machine-learning model selected a subset judged more likely to represent real optical events. The authors then projected those candidates onto Earth under an assumed geosynchronous-orbit interpretation and applied two clustering methods: fixed five-degree grid cells produced 16 statistically significant hotspots, while a separate two-step cluster analysis produced six distinct groups.

Several clusters from the two methods overlap near the Pacific west of southern Mexico and Central America, the southern Gulf of Mexico, and the Southwest around Sedona and White Sands. The authors also report geographic differences for transients associated with Pacific and Nevada nuclear-test dates. Those results are unreviewed and model-dependent: no plate flash was ranged, tracked from a second site, or recovered, and projecting a sky position onto Earth does not prove a geosynchronous object existed there. Related Palomar statistics remain contested in the literature, so independent access to the catalog, selection pipeline, null models, and observation-schedule controls is essential.

Why it matters

It turns an archival anomaly into a geographic prediction that can be reproduced or challenged, while making clear how much of the result depends on selection and an inferred orbit rather than direct tracking.

What the source establishes

Verified new preprint and reported clustering outputs. Developing and disputed: the candidate selection, hotspot significance, orbital interpretation, nuclear-test association, and physical nature of every transient. No artificial or non-human object has been established.