NUFORC UFOs
98.0% of events in this dataset were triggered by prior events. Only 2.0% are independent (1,556 of 79,638).
Pattern Survival
| Pattern | Status |
|---|---|
| 9pm peak | Survives |
| Western US concentration | Survives |
| Summer peak | Stripped |
| Weekend effect | Stripped |
| Exponential growth | Stripped |
| 2014 spike | Stripped |
Content ComparisonComparing what independent reporters describe vs what triggered reporters describe. Significant differences suggest the two groups are seeing different things.
66 of 116 features significantly different (FDR < 0.05)
Independent Reports Enriched For
- •Disks (+14.5%)
- •Cigars (+3.5%)
- •Silver coloring
- •"Observed" language
- •Longer narratives
- •Emotion markers (3.3× more)
Triggered Reports Enriched For
- •Lights (+14.6%)
- •"Bright" (+9.4%)
- •Orange coloring (+7.0%)
- •Fireballs (+5.4%)
- •"Fast" descriptor
Cultural Template ScoreHow closely report descriptions match popular culture depictions. Lower = more original observations. Higher = more influenced by movies and media.
Independent: 0.102 | Triggered: 0.131
Temporal StabilityWhether the content patterns stay consistent across different time periods — stable patterns are more likely to be real, not fads.
4 of 5 top shapes in independent reports remain stable across decades.
Data Quality Notes
- •79% of reported times round to nearest 5 minutes
- •18% of reports filed more than 1 year after sighting
- •July 4th has 4.5x average daily sightings
Calendar Distribution
Calendar distribution biases in NUFORC reporting
- •9pm hourly peak (Rayleigh R = 0.552)
- •Summer sightings 2x winter
- •Observer availability explains 65.7% of temporal pattern
Reporting Lag
Time between sighting and report submission
- •Median reporting lag: 39 days
- •Spikes at 7, 14, 30, 60, and 90 days (round-number heaping)
Instrument Changes
- •1995: NUFORC website launch (step change in report volume)
- •2004: Report form redesign (changed data fields collected)
- •2012: Starlink satellite train era begins (contamination source)
If You're Using This Dataset
Known Issues
- •Starlink contamination post-2012
- •1995 website launch step change
- •2004 form redesign
Key Insight
98% of UFO reports are people reporting because other people reported. The 2% that survive represent genuine observations that cannot be explained by social contagion.
Downloads
Processed data exports for this dataset. Includes Hawkes-thinned independent events, decomposition results, and curiosity scanner reports.
Also available on GitHub
Citation
Suggested citation
Brothers, B. (2026). NUFORC UFOs Hawkes Process Decomposition: 2.0% Independent Events from 79,638 Total. Project Aletheia. https://projectaletheia.org/datasets/nuforc