Every detection receives a Watchtower Threat Index score and a tier 1–5 classification. Method is open. Source data is public. Highest-scoring events surface here first.
Why this matters: LA-basin traffic volume is ~10× Kern's. A single regional ranking buries Kern signals under LA noise. The county lens scores each event against the airspace it actually occurred in.
Each county learns its own baseline. An aircraft at 800 ft over Kern is anomalous because Kern's normal is high-altitude transit — even if the same altitude would be unremarkable over the LA basin.
| County | Samples (48h) | Unique aircraft | Median altitude | 10th %ile altitude | Median speed | Night % |
|---|---|---|---|---|---|---|
| Kern | 9,525 | 1,790 | 12,575 ft | 1,900 ft | 219.2 kts | 34.7% |
| Tulare | 2,726 | 847 | 14,300 ft | 2,050 ft | 253.3 kts | 35.9% |
| Kings | 1,408 | 751 | 29,000 ft | 3,150 ft | 397.6 kts | 41.1% |
| Fresno | 4,772 | 1,530 | 12,412.5 ft | 1,125 ft | 253.9 kts | 37.1% |
| San Bernardino | 10,481 | 2,572 | 11,600 ft | 2,100 ft | 287.8 kts | 35.8% |
| Other | 40,070 | 10,058 | 16,657.3 ft | 3,412.1 ft | 265.1 kts | 38.2% |
Baselines refresh on every page load from the last 48 hours of detections. Same math for every county.
Each score is a weighted sum of five public-data components: altitude (vs. CFR floor), temporal pattern (vs. 48-hour baseline), convergence (multi-aircraft proximity), shell-network linkage (public corporate filings), and repeat frequency. The weights below ship with every row — no hidden parameters.
The Watchtower Threat Index (WTI) ranks flight events by statistical abnormality and regulatory risk — not intent.
WTI is a signal for journalists, attorneys, and regulators. A high score means a detection deviates sharply from the 48-hour baseline and/or implicates a public regulation. It is not, and never will be, an accusation against any individual.
Weights ship with every row under components.weights — fully reproducible.
Worked example — WTI = Σ(component × weight)
altitude 100 × 0.35 = 35.00
temporal 100 × 0.20 = 20.00
convergence 26 × 0.12 = 3.12
shell 100 × 0.08 = 8.00
repeat 100 × 0.25 = 25.00
──────
WTI = 91.12Every row carries the method version so a third party can reproduce the score even if the method later evolves.
| Tier | Meaning | Recommended action |
|---|---|---|
| T4 | Critical statistical abnormality | Flag for legal / regulatory review |
| T3 | High abnormality | Flag for analyst review |
| T2 | Elevated | Monitor |
| T1 | Moderate | No action — context only |
| T0 | Low | Baseline |
Convergence captures the pattern that two or more aircraft in the same airspace band, at the same time, on coordinated tracks, is not coincidence — it is operational coordination. The metric weighs four factors:
Many top-ranked events share the exact same WTI (for example, multiple rows at 91.12). This is expected: when aircraft share the same altitude deviation, timing band, and repeat pattern, the weighted sum collapses to the same value. Identical scores are a feature of the method, not a bug — they make repeat behavior visible at a glance.
WTI is a statistical abnormality score computed from public ADS-B and FAA registry data. It is a signal for review by humans with legal authority — nothing more.
Source: anomaly_events (quiet-math). WTI = anomaly_score × 100; tier bands 25 / 50 / 75. Every row carries the method version so a third party can reproduce the score.