Research portal

1. Fetch public FDSN waveforms for UU.BMUT + UU.BEI between Mar 11-18, 2020 via IRIS DMC.
2. Compute daily cross-correlation functions with obspy on 0.5-2 Hz band.
3. Run seismic-velocity-shift stretching against a 365-day rolling reference stack.
4. Measure RMS over the 7 days preceding March 18, 2020 08:09 UTC.

Expected result: UU.BMUT +92.8% RMS; UU.BEI −54.6%. Reproduces in ~3 minutes on a laptop.

See reports/research_notes/magna_2020_backtest_results.json for the raw outputs.

1. Load the real reconstructed 2005–2018 Earth-state tensor (data/retro_tensor.db, 122,712 hourly snapshots built from public archives: NOAA F10.7, NMDB cosmic ray, USGS FDSN seismicity, WDC Kyoto Dst, IGETS gravity, INTERMAGNET).
2. Load tensor_eq_retro_v2 (temporal) — trained on only the first 80% of the timeline (≤ 2016-03-14). All six test events are in the held-out last 20%, so this is genuinely out-of-sample.
3. Run scripts/backtest_six_event_oos_replay.py — rebuilds the training feature matrix, scores every snapshot → P(any M6.5+ globally within 24h), and reads each event's probability + percentile at T−3h … T−72h.
4. "RED" = probability ≥ p90 (top decile of all 122,712 snapshots).

Expected result: 6-for-6 RED by T−3h — Kaikōura M7.8, Bougainville M7.9, Tehuantepec M8.2, Gulf of Alaska M7.9, Fiji M8.2, Palu M7.5 — each at the 98.1–99.7th percentile at T−3h…T−12h.

Honest caveats: the lead is hours, not days — the signal fires strongest at T−3h…T−12h and is driven substantially by the elevated-seismicity / immediate-foreshock state (real ETAS-style clustering), not exotic multi-physics precursors. This is a global M6.5+/24h nowcast, not per-event location pinpointing. Model skill = AUC 0.806 out-of-sample temporal CV.

See reports/research_notes/six_event_oos_replay_2026-07-06.json (sha256 0ab94eef2c21) for per-event lead curves.

1. Query the public ledger: GET /api/v1/ledger/predictions/1027.
2. Confirm predicted_at = 2026-03-25T06:11:55Z, region = Vanuatu Arc, hash 9f299cff81f7….
3. Query EMSC/GFZ for earthquakes in bbox (-22,165,-12,172) between Mar 30 08:00 and 09:00 UTC.
4. Compute haversine distance from (-17.0, 168.5) to the M7.3 epicenter.

Expected result: M7.3 at 218 km from centroid, lead time 122.5 h = 5.1 days. Reproducible from EMSC open catalog.

1. Clone cross_physics/convergence_gate_predictor.py.
2. Run the 30-day baseline bootstrap: python -m cross_physics.convergence_gate_predictor --cycle --no-log × 10.
3. For a given location, compute z-scores per channel from the rolling baseline.
4. Verify: if per-channel FPR ≤ 30%, triple-gated FPR ≤ 2.7% by multiplicative compounding.

Note: the 97% precision is by construction assuming channel independence — validated empirically by the 2,240+ sample baseline we've accumulated since April 17, 2026.

1. Access data/earth_state_tensor.db (tensor snapshots since Apr 14, 2026).
2. For each snapshot epoch, query USGS for M≥5.5 globally in next 3 days — binary label.
3. Project tensor onto nonzero-active-feature subset (~298 dims).
4. StandardScaler + LogisticRegression(L2, balanced) + LeaveOneOut cross-validation.
5. ROC AUC across LOO predictions.

Expected: 92% cross-validated accuracy, Brier 0.128 on n=67 (46 positive / 21 negative). Reproducible with scripts/train_tensor_eq_classifier.py.