Eclipse 2026 Success Criteria

Success criteria — what "the study succeeded" means, decidably

Every criterion below is pinned to a sealed, fingerprint-checked Form and decided by the T+1 resolution on measured public data. Nothing here is judged by narrative; each tier passes or fails in the same rendered tables the world can read now.

Tier 0 — Pre-registration integrity (testable every day; already green)

All Forms re-verify from disk, and both prediction Forms re-derive byte-for-byte from the NASA Besselian elements. Any drift before the resolution runs — constants, roster, coordinates, elements — falsifies the pre-registration itself. This is the criterion that gives the others their meaning.

Form	Kind	Class	Closure	Terminal	Fingerprint	Stamped
2024 benchmark	projection	shadow-projection	7/7	CALORIE	89afd825011f060e	2026-07-16T16:01:47Z
2026 prediction	prediction-2026-08-12	shadow-projection	PREREGISTERED	CALORIE	c3e6841a206b8058	2026-07-16T16:02:00Z
2026 prediction extensions	prediction-extensions-2026-08-12	shadow-projection	PREREGISTERED	CALORIE	b106d2926b1bf2cf	2026-07-16T18:54:39Z
historical universality	historical-universality	shadow-projection	16/20	CURE	cdf43e80f06f10a3	2026-08-18T16:42:21Z

Loaded and fingerprint-checked from ~/.gaiaftcl/franklin/invariants/ at render time.

Tier 1 — The base prediction (Form c3e6841a206b8058)

ID	Criterion	Decided by
S1	All five totality stations — Reykjavik, Látrabjarg, A Coruña, León, Zaragoza — PROJECT under the frozen law: depletion ≥ 300‰ of obscuration inside their ≥80% gate windows	resolution crucible on measured vTEC, when the archive posts (~2026-08-30)
S2	Both declared controls — Norwich CT (13.2%), Millstone Hill (15.9%) — seal dead cat (their gate can never open, by geometry alone)	same crucible

CALORIE iff 5/5 projections and 0 false positives.

Tier 2 — The shear-born extensions (Form b106d2926b1bf2cf)

The criteria no depletion-magnitude model even states:

ID	Criterion	Decided by
S3 — traveling signature	Depletion minima in the pinned order Látrabjarg → +192 s → Reykjavik → +2372 s → A Coruña → +53 s → León → +32 s → Zaragoza, each deepest bin riding its own max-obscuration second within [−300, +1200] s	eclipse-shear on measured 2026 vTEC
S4 — confinement (storm-proof)	Excess in-window depletion (gate-window median − ±2 h flank median) ≥ 50,189 ppm at every totality station, < 50,189 ppm at both controls	same analysis
S5 — quantified nulls	No gate-scale signature at the controls, with pre-stated sub-threshold expectations: Norwich 40,790–73,923 ppm · Millstone 49,036–88,868 ppm (below quiet variance at 300-s bins)	same analysis

Tier 3 — Cross-regime universality (reported, not sealed)

The 2026 couplings land inside the measured historical band 309–560‰ — extending the one-frozen-law result from three eclipses and a full solar cycle (2017 SC24-declining · 2023 annular · 2024 SC25-max) to a fourth eclipse and the hardest geometry transfer: high latitude, +25.8° Sun, near sunset over Europe.

The storm contingency (pre-registered; never triggered — eclipse day measured quiet, max Kp 2.667)

If SWPC est-Kp ≥ 5 during 15:30–20:00 UT on eclipse day: S1–S4 remain decidable — that is exactly what the shear bought — while S5 weakens with a storm annotation. A G-storm cannot un-total an eclipse, and it cannot fake confinement.

Failure semantics (LAW 3)

Any miss seals CURE, raw, in these same tables — never renamed. A high-latitude coupling break or a lag-window violation would itself be a finding of the study, published identically to a pass.

---

Why reading the shear changes the game

1. From magnitude to signature

Every conventional eclipse-ionosphere result is a scalar claim — "TEC dropped 20–60%" — and the Gannon corpus proves scalars are forgeable: May 11, 2024 reproduced eclipse-grade depletions at all seven benchmark stations with zero shadow anywhere, and the frozen magnitude law false-positived 7/7 (Model shear). Reading the shear replaced how much with what shape: confined to the gate minutes, riding each station's own maximum with the measured bottomside lag. A storm can fake the amount. Only the Moon writes the shape.

2. Storm-immune eclipse science

The worst operational risk of any solar-max eclipse campaign is a coincident geomagnetic storm — historically, that day's data is discarded as contaminated. The confinement discriminant stays decidable through a G5: a storm floods the ±2 h flanks (measured at 665,000–724,000 ppm on May 11) and pins its minima to window edges; a shadow cannot do either. At SC25 maximum, the elevated storm probability flips from study-killer to strongest possible test — S4 passing through a storm would be the headline result.

3. The failures became the instrument

Standard practice filters model failures out as bad data. This study ingested them on purpose, measured exactly how they fail — flank flooding, edge-pinned minima, no lag relation to the geometry — and that failure structure became the second-generation discriminant, then a sealed falsifiable prediction 27 days ahead. Data sheared of the models revealed what the models could not have imagined; the recorded CURE taught the next CALORIE.

4. A general template

Wherever a computable forcing meets a noisy medium, the same read applies: pair the forcing's own timetable with the response, demand confinement + traveling coherence against that timetable, and use the adversarial corpus (the "storms" of that domain) as the shear. Launch- and eruption-driven TIDs, flare-vs-storm thermospheric response, any detector whose adversary mimics magnitude but not structure. And if S3+S4 hold when the pinned archive grades August 12, the ionosphere will have been shown to carry a readable, per-second copy of the umbra's ground track that survives storm contamination — the Moon as a calibrated signal generator, photochemistry separated from electrodynamics at 300-second resolution, from public raw data on one machine.