Analysis of 325,573 sighting reports reveals a phenomenon that is overwhelmingly nocturnal, seasonal, geographically concentrated, and clustered near airports.
The data shows a 9.4x per-hour concentration between 8PM-midnight vs morning hours, peaking sharply at 9PM. This aligns with the transition from civil twilight to astronomical darkness — the window when objects become visible against a darkening sky but witnesses are still outdoors.
The summer surge (32.7%) and July peak correlate with longer daylight hours producing extended dusk windows, increased outdoor activity, and (notably) Independence Day celebrations. The 28.4x growth from 1990-2012 tracks internet adoption, suggesting reporting accessibility — not phenomenon frequency — may drive the increase.
"Light" as the dominant shape (20.6%) raises important questions: are witnesses observing aerial vehicles at distances too great for shape determination, or are many reports of astronomical/atmospheric phenomena misidentified? The persistence of "classic" UFO shapes (disk/cigar/oval/sphere at 20.4%) alongside modern "triangle" reports (9.8%) suggests either evolving phenomenon morphology or evolving cultural expectations.
The geographic concentration in California (13.7% of US) correlates with population density, clear skies, military aviation corridors, and cultural receptivity to reporting. The 81.1% US dominance reflects NUFORC's US-centric reporting infrastructure rather than a genuine geographic distribution of phenomena.
Shape terminology has evolved dramatically. "Disk" dominated the 1950s-70s (23-29%) but fell to 6.5% by 2010. "Triangle" emerged in the 1980s (13%) and persists. "Fireball" surged in the 2010s (12%). This suggests either changing phenomenon morphology, cultural influence on reporting language, or the introduction of new military/civilian aircraft shapes. Notably, "disk" reports have the lowest night/day ratio (2.0x) — suggesting they're more often seen in daylight when shape is discernible, while "lights" (12.8x night bias) dominate after dark.
Behavioral signatures cluster around 5 key traits: hovering (8.1%), pulsation (6.1%), formation flying (6.1%), silence (5.9%), and extreme brightness (2.4%). The 5.9% "silent" rate is significant — conventional aircraft are rarely described as silent. Military response was noted in 2.2% of reports. Zigzag/erratic movement (1.4%) and rapid acceleration (1.2%) align with the "instantaneous acceleration" observable.
Military/nuclear site proximity shows elevated reporting near Edwards AFB (1,604 within 100km), Wright-Patterson (975), and Cape Canaveral (789). Area 51 shows only 33 reports — not because nothing happens there, but because the surrounding 100km is uninhabited desert. The nuclear weapons nexus (Los Alamos: 410, Oak Ridge: 266) adds data points to the nuclear-UAP correlation documented in the archive.
Physical evidence is rare but consistent. Only 1.15% of reports mention bodily effects. Burns dominate (40.4% of physical claims), with abduction/missing time at 0.27%. The 3 implant reports in 80,331 records contrast sharply with media portrayals. Electrical/EM effects (105 reports) provide the most objectively verifiable category — vehicles stalling, compasses spinning, radios failing — consistent with documented cases like Levelland (1957) and Tehran (1976).
The lunar correlation is bimodal, not random. Sightings spike at BOTH new moon (2.11x expected) AND full moon (1.75x expected), with a dip at quarter phases. This reveals two overlapping populations: dark-sky observers (new moon = faint objects visible) and outdoor-active reporters (full moon = more people outside). The shape×time matrix reinforces this — "disk" has only a 2.1x night bias (visible in daylight) while "light" is 8.2x (only visible in darkness).
68.4% within 10km of an airport (peak at 5-10km band with 41.4%), with Boeing Field Seattle leading at 2,955 nearby sightings. Coastal states report 1.75x more per-state than inland, and the 40-45°N latitude band contains 30% of all sightings globally. The 20.2% HDBSCAN noise points — unclustered outliers far from population centers — are the most anomalous reports in the entire dataset.
This analysis documents what was reported — not what was observed. Reporting biases (language, culture, internet access, time of day, season) significantly shape this dataset. The enriched 325K dataset includes lunar data, airport proximity, weather, and HDBSCAN clustering from multiple merged sources. The Valinor agent swarm provides objective, 24/7 camera-based detection to complement subjective human reporting.