# Study 04 — Results (PARTIAL corpus seal)

**Status: PARTIAL_CORPUS_SEAL 2026-08-20T20:10:00Z** — not the full multi-gauge Form. Single-gauge integer-mm onset vs sealed chart minutes for two tsunami positives + Sandy adversary; Tonga water-branch **VOID** until pressure channel pulled.

Charter: [Study 04](Study-04-Tsunami-vs-Storm-Surge) · Ledger: `corpus/study-04/study04_ledger.json`

---

## Frozen integers (this partial seal)

| Symbol | Value |
|---|---|
| `T_exc_min_mm` | 500 |
| `T_lag_tol_min` | 45 |
| Chart Pago (Samoa) | 57 min |
| Chart Nawiliwili (Tohoku) | 456 min |
| Onset | first sample \|level − first-hour median\| ≥ 500 mm |
| Quantization | CO-OPS meters decimal-string → integer mm |

---

## Grades

| Gauge / window | Class | lag_min | chart_min | onset_mm | Verdict |
|---|---|---:|---:|---:|---|
| Pago Pago 2009-09-29 (Samoa M8.1) | tsunami | 24 | 57 | −1255 | **WIN** |
| Nawiliwili 2011-03-11 (Tohoku M9.1) | tsunami | 446 | 456 | +608 | **WIN** |
| The Battery 2012-10-29/31 (Sandy) | adversary | — | — | −503 | **WIN** (no seismic origin clock) |
| Port San Luis 2022-01-15 (Tonga) | dual | −111 | 780 | +510 | **VOID_NEED_PRESSURE** |

**Summary:** tsunami 2/2 WIN · adversary 1/1 WIN · Tonga VOID.

---

## Still OPEN for full LAW FROZEN

- Multi-gauge coherence (≥3 azimuthal quadrants)
- Full corpus rows (Kuril, Maule, Chignik, Noto, …)
- Tonga Lamb branch with co-timed air_pressure
- Detide via CO-OPS `predictions` product (this seal used first-hour median)
- Standing trigger on next M≥7.5 subsea event

> Partial seal is a finding, not a rename of the full Form. Full Form freezes only when the OPEN list closes.