UAP Pattern Recognition Study 1945–1975: US Military Atomic Warfare Complex

A 2022 white paper published through the Scientific Coalition for UAP Studies (SCU) by a team led by historian-researcher Larry Hancock. The paper tests whether UAP reports cluster around the US Atomic Warfare Complex over the 1945–1975 window. The dataset comprises approximately 1,500 reports drawn from Project Blue Book, the FBI Vault, the NICAP archive, and contemporary newspaper indices.

It is the foundational document of the modern peer-reviewed statistical-pattern UAP literature. The 2024 SCU follow-up (Hancock et al. 'UAP Activity Pattern Study') and the 2025 Limina peer-reviewed extension (Grosvenor et al. 'UAP Indications Analysis') both build directly on its methodology.

The 'UAP-clusters-around-nuclear-sites' claim had, before Hancock 2022, lived almost entirely in oral history — Robert Hastings' 'UFOs and Nukes' book (2008), the Salas Malmstrom testimony, the surviving Bentwaters narrative. Hancock's contribution was to make the claim *testable*: to compile the underlying dataset, define the controls, and run the comparison against a population-weighted baseline.

That move took the claim out of the oral-tradition register and put it into the statistical register. Whether the conclusion holds up under scrutiny is a separate question — and the Grosvenor 2025 paper is the answer to that question — but the methodological shift is the contribution.

The dataset was compiled from Blue Book (the largest single source), the FBI Vault's UFO holdings, the NICAP case archive, and contemporary newspaper records. Each case was geocoded to its reported location and coded for date, time, witness profile, and reported physical attributes.

The comparison: case density per geographic cell was compared against a baseline derived from contemporaneous civilian population, military airspace, and known reporting-infrastructure footprints. Cells overlapping AWC facilities (Los Alamos, Sandia, Hanford, Oak Ridge, Pantex, Rocky Flats, Savannah River, the Pacific Proving Grounds, NTS, plus the operational ICBM/SLBM bases) were flagged separately.

UAP report density clusters around AWC sites at densities significantly higher than baseline population-weighted expectations across the 1945–1975 window. The clustering is most concentrated in the 1945–1955 window (the early-Cold-War weapons-design and -testing era) and 1965–1975 (the Minuteman operational deployment era).

The strongest single-site signal in the original Hancock paper sits at the Oak Ridge complex over 1948–1952. The Hanford site is a close second over 1949–1952. Among ICBM bases, Malmstrom AFB produces the strongest signal over 1966–1968.

The paper is careful about interpretive limits: statistical clustering does not specify a mechanism. The data establish that something happened more often near AWC sites than baseline. They do not establish what.

Hancock 2022 was the first piece of UAP pattern-recognition work that the academic statistical literature could engage on its own methodological terms. The Grosvenor 2025 Limina paper is the formal peer-reviewed extension; it refines the methodology, adds additional controls, and finds the pattern robust.

The 2025 Brühl/Villarroel Palomar Sky Survey paper in Scientific Reports — methodologically completely different — produces an adjacent finding (photographic transients in the 1950s sky survey correlate with nuclear-test dates). The three independent results together make the underlying claim difficult to dismiss as reporting-bias artefact.

AARO's Historical Record Report Volume I (2024) treats the nuclear-site pattern as part of the historical baseline but does not characterise its underlying nature. The institutional gap between the peer-reviewed literature and AARO's analytical posture is, as of mid-2026, widening.