Eclipse 2026 Blind Spots

Blind spots — five stories the shear tells that magnitude models will always miss

Every story below is read directly from ledger rows in the shear-discriminant table — not argued, measured. Each names a structural blindness: a class of event a depletion-magnitude model glosses over not by bad luck but by construction, because a scalar cannot carry shape.

---

Story 1 — Eugene, Oregon 2023: the eclipse the magnitude model threw away

The frozen magnitude law sealed 2023-10-14-eugene-or dead cat — inside its narrow annular gate window (only ~18 minutes at ≥80% ring coverage) the depletion never reached the 300‰ coupling line, so the scalar test said no eclipse signature.

The shear row says otherwise: window 131,566 ppm · flank 0 ppm · excess 131,566 ppm — deepest bin at 16:27:30 UT, riding its own max-obscuration second by +517 s, squarely in the measured bottomside-lag window. A textbook-shaped, perfectly confined, correctly traveling eclipse signature — just a weak one. It clears the confinement threshold (50,189 ppm) by 2.6×.

The blindness: magnitude thresholds discard weak-but-perfectly-shaped signals. Any survey that filters events by "depletion ≥ X%" silently deletes every marginal-geometry eclipse response — annular edges, path edges, low-Sun crossings — from the record. The shape-read recovers them.

Story 2 — May 11, 2024: the phantom eclipse every scalar pipeline records

On May 11 the Gannon negative phase depressed TEC 67–73% below quiet baselines at all seven benchmark stations — deeper than any real eclipse in the corpus. A magnitude model doesn't just miss here; it hallucinates: our own frozen law false-positived 7/7, "detecting" an eclipse at every station under a cloudless, unobscured Sun.

The shear rows: window 667,827–733,664 ppm — and flanks 665,528–724,580 ppm. The excess collapses to 2,204–42,836 ppm, under threshold everywhere. The depletion is real; the shape is a flood, not a shadow.

The blindness: a scalar can say that TEC dropped, never why. Every magnitude-threshold eclipse study running near solar maximum is one storm away from recording an event that never happened — and cannot know it from its own data.

Story 3 — Mazatlán, May 11: the inversion no average survives

Same storm, same day — but Mazatlán's row reads window 471,758 · flank 539,170 · excess −67,412 ppm. The gate window was less depleted than its flanks: the storm's negative phase has latitude and local-time structure that happened to spare those exact minutes.

The blindness: magnitude models average that structure away. A negative excess is physically meaningful — it says "whatever did this was structured against the shadow's timetable" — and no scalar summary can even represent the sign.

Story 4 — The seasonal ghost of September 17

San Antonio and Corpus Christi, control day 2023-09-17: month-scale equinox TEC decline against an October baseline faked a 27–28% "depletion" — and the magnitude law false-positived both (the two residual control falsifiers in the sealed historical CURE). The shear rows show the tell: deepest bins at −706 s and −646 s — before the would-be maximum, pinned to the drift, with excesses (64,247 and 35,020) that only exist because the flanks drift too.

The lag gate rejects both: seasonal drift doesn't ride the Moon's timetable.

The blindness: every magnitude-vs-baseline method is hostage to its baseline's calendar. Season, solar rotation, and secular solar-cycle trend all masquerade as depletion. The traveling-lag clause is calendar-proof because its clock is the umbra, not the archive.

Story 5 — Carbondale, twice: one ground, one law, two suns

The only station under both the 2017 and 2024 totality paths. Same grid cells, same instrument network, one full solar cycle apart: coupling 488‰ (2017, SC24 declining) vs 342‰ (2024, SC25 maximum) — a 30% modulation of the eclipse response by the solar-cycle background, on identical ground.

The blindness: a fixed magnitude threshold calibrated in one cycle silently mis-calibrates in the next — too strict at maximum, too loose at minimum. The corpus measures the modulation instead of hiding it, and the confinement discriminant is threshold-stable across both regimes because flanks and window scale together.

---

The live table these stories are read from

The full confinement table the stories are read from (click to expand — generated live from the ledger)

Eclipse	Kind	Station	Day	Window depletion	Flank depletion	EXCESS	Deepest bin (UT)	Δ from own max
2017-08-21	control	carbondale	2017-07-25	0 ppm	0 ppm	0 ppm	18:07:30	-835 s ✗
2017-08-21	control	carbondale	2017-08-14	0 ppm	0 ppm	0 ppm	18:07:30	-835 s ✗
2017-08-21	control	carbondale	2017-08-28	42553 ppm	77669 ppm	-35116 ppm	18:12:30	-535 s ✗
2017-08-21	control	casper-wy	2017-07-25	0 ppm	0 ppm	0 ppm	17:27:30	-981 s ✗
2017-08-21	control	casper-wy	2017-08-14	0 ppm	0 ppm	0 ppm	17:27:30	-981 s ✗
2017-08-21	control	casper-wy	2017-08-28	64516 ppm	75949 ppm	-11433 ppm	17:52:30	519 s ✓
2017-08-21	control	columbia-sc	2017-07-25	0 ppm	0 ppm	0 ppm	18:27:30	-936 s ✗
2017-08-21	control	columbia-sc	2017-08-14	0 ppm	0 ppm	0 ppm	18:27:30	-936 s ✗
2017-08-21	control	columbia-sc	2017-08-28	100917 ppm	99099 ppm	1818 ppm	18:47:30	264 s ✓
2017-08-21	control	nashville	2017-07-25	0 ppm	0 ppm	0 ppm	18:12:30	-954 s ✗
2017-08-21	control	nashville	2017-08-14	0 ppm	0 ppm	0 ppm	18:12:30	-954 s ✗
2017-08-21	control	nashville	2017-08-28	31578 ppm	67415 ppm	-35837 ppm	18:27:30	-54 s ✓
2017-08-21	control	salem-or	2017-07-25	34090 ppm	0 ppm	34090 ppm	17:32:30	852 s ✓
2017-08-21	control	salem-or	2017-08-14	0 ppm	0 ppm	0 ppm	17:07:30	-648 s ✗
2017-08-21	control	salem-or	2017-08-28	0 ppm	31250 ppm	-31250 ppm	17:07:30	-648 s ✗
2017-08-21	eclipse	carbondale	2017-08-21	477395 ppm	162019 ppm	315376 ppm	18:27:30	365 s ✓
2017-08-21	eclipse	casper-wy	2017-08-21	388617 ppm	86941 ppm	301676 ppm	17:57:30	819 s ✓
2017-08-21	eclipse	columbia-sc	2017-08-21	498899 ppm	237474 ppm	261425 ppm	18:47:30	264 s ✓
2017-08-21	eclipse	nashville	2017-08-21	437216 ppm	208971 ppm	228245 ppm	18:27:30	-54 s ✓
2017-08-21	eclipse	salem-or	2017-08-21	384069 ppm	191000 ppm	193069 ppm	17:27:30	552 s ✓
2023-10-14	control	albuquerque	2023-09-17	200344 ppm	183114 ppm	17230 ppm	16:37:30	35 s ✓
2023-10-14	control	albuquerque	2023-09-30	0 ppm	0 ppm	0 ppm	16:27:30	-565 s ✗
2023-10-14	control	albuquerque	2023-10-02	32451 ppm	17336 ppm	15115 ppm	16:27:30	-565 s ✗
2023-10-14	control	albuquerque	2023-10-07	0 ppm	0 ppm	0 ppm	16:27:30	-565 s ✗
2023-10-14	control	corpus-christi	2023-09-17	258198 ppm	223178 ppm	35020 ppm	16:47:30	-646 s ✗
2023-10-14	control	corpus-christi	2023-09-30	0 ppm	0 ppm	0 ppm	16:47:30	-646 s ✗
2023-10-14	control	corpus-christi	2023-10-02	54107 ppm	0 ppm	54107 ppm	16:47:30	-646 s ✗
2023-10-14	control	corpus-christi	2023-10-07	0 ppm	0 ppm	0 ppm	16:47:30	-646 s ✗
2023-10-14	control	eugene-or	2023-09-17	71380 ppm	80460 ppm	-9080 ppm	16:17:30	-83 s ✓
2023-10-14	control	eugene-or	2023-09-30	0 ppm	0 ppm	0 ppm	16:12:30	-383 s ✗
2023-10-14	control	eugene-or	2023-10-02	33316 ppm	31168 ppm	2148 ppm	16:12:30	-383 s ✗
2023-10-14	control	eugene-or	2023-10-07	0 ppm	0 ppm	0 ppm	16:12:30	-383 s ✗
2023-10-14	control	san-antonio	2023-09-17	265203 ppm	200956 ppm	64247 ppm	16:42:30	-706 s ✗
2023-10-14	control	san-antonio	2023-09-30	0 ppm	0 ppm	0 ppm	16:42:30	-706 s ✗
2023-10-14	control	san-antonio	2023-10-02	20180 ppm	4829 ppm	15351 ppm	16:47:30	-406 s ✗
2023-10-14	control	san-antonio	2023-10-07	0 ppm	0 ppm	0 ppm	16:42:30	-706 s ✗
2023-10-14	eclipse	albuquerque	2023-10-14	322446 ppm	152017 ppm	170429 ppm	16:47:30	635 s ✓
2023-10-14	eclipse	corpus-christi	2023-10-14	348812 ppm	223811 ppm	125001 ppm	17:12:30	854 s ✓
2023-10-14	eclipse	eugene-or	2023-10-14	131566 ppm	0 ppm	131566 ppm	16:27:30	517 s ✓
2023-10-14	eclipse	san-antonio	2023-10-14	368565 ppm	234830 ppm	133735 ppm	16:57:30	194 s ✓
2024-04-08	control	carbondale	2024-03-12	28804 ppm	32136 ppm	-3332 ppm	18:47:30	-829 s ✗
2024-04-08	control	carbondale	2024-04-01	0 ppm	0 ppm	0 ppm	18:47:30	-829 s ✗
2024-04-08	control	carbondale	2024-04-15	0 ppm	0 ppm	0 ppm	18:47:30	-829 s ✗
2024-04-08	control	cleveland	2024-03-12	31284 ppm	30168 ppm	1116 ppm	19:02:30	-788 s ✗
2024-04-08	control	cleveland	2024-04-01	0 ppm	0 ppm	0 ppm	19:02:30	-788 s ✗
2024-04-08	control	cleveland	2024-04-15	0 ppm	0 ppm	0 ppm	19:02:30	-788 s ✗
2024-04-08	control	dallas	2024-03-12	73593 ppm	55656 ppm	17937 ppm	18:47:30	294 s ✓
2024-04-08	control	dallas	2024-04-01	0 ppm	0 ppm	0 ppm	18:27:30	-906 s ✗
2024-04-08	control	dallas	2024-04-15	0 ppm	0 ppm	0 ppm	18:27:30	-906 s ✗
2024-04-08	control	indianapolis	2024-03-12	54195 ppm	47430 ppm	6765 ppm	19:22:30	873 s ✓
2024-04-08	control	indianapolis	2024-04-01	0 ppm	0 ppm	0 ppm	18:52:30	-927 s ✗
2024-04-08	control	indianapolis	2024-04-15	0 ppm	0 ppm	0 ppm	18:52:30	-927 s ✗
2024-04-08	control	mazatlan	2024-03-12	100549 ppm	65170 ppm	35379 ppm	18:02:30	-427 s ✗
2024-04-08	control	mazatlan	2024-04-01	0 ppm	0 ppm	0 ppm	17:52:30	-1027 s ✗
2024-04-08	control	mazatlan	2024-04-15	0 ppm	0 ppm	0 ppm	17:52:30	-1027 s ✗
2024-04-08	control	millstone	2024-03-12	0 ppm	0 ppm	0 ppm	19:17:30	-709 s ✗
2024-04-08	control	millstone	2024-04-01	11608 ppm	16518 ppm	-4910 ppm	19:32:30	191 s ✓
2024-04-08	control	millstone	2024-04-15	0 ppm	0 ppm	0 ppm	19:17:30	-709 s ✗
2024-04-08	control	norwich	2024-03-12	0 ppm	0 ppm	0 ppm	19:17:30	-648 s ✗
2024-04-08	control	norwich	2024-04-01	19407 ppm	0 ppm	19407 ppm	19:32:30	252 s ✓
2024-04-08	control	norwich	2024-04-15	0 ppm	0 ppm	0 ppm	19:17:30	-648 s ✗
2024-04-08	eclipse	carbondale	2024-04-08	313114 ppm	255571 ppm	57543 ppm	19:17:30	971 s ✓
2024-04-08	eclipse	cleveland	2024-04-08	400669 ppm	330523 ppm	70146 ppm	19:27:30	712 s ✓
2024-04-08	eclipse	dallas	2024-04-08	280267 ppm	194617 ppm	85650 ppm	18:57:30	894 s ✓
2024-04-08	eclipse	indianapolis	2024-04-08	354940 ppm	276144 ppm	78796 ppm	19:22:30	873 s ✓
2024-04-08	eclipse	mazatlan	2024-04-08	272936 ppm	174356 ppm	98580 ppm	18:27:30	1073 s ✓
2024-04-08	eclipse	millstone	2024-04-08	474572 ppm	380621 ppm	93951 ppm	19:42:30	791 s ✓
2024-04-08	eclipse	norwich	2024-04-08	477974 ppm	380160 ppm	97814 ppm	19:37:30	552 s ✓
2024-04-08	shear-storm	carbondale	2024-05-10	0 ppm	0 ppm	0 ppm	18:47:30	-829 s ✗
2024-04-08	shear-storm	carbondale	2024-05-11	731039 ppm	723259 ppm	7780 ppm	18:47:30	-829 s ✗
2024-04-08	shear-storm	cleveland	2024-05-10	0 ppm	0 ppm	0 ppm	19:02:30	-788 s ✗
2024-04-08	shear-storm	cleveland	2024-05-11	709141 ppm	705195 ppm	3946 ppm	19:02:30	-788 s ✗
2024-04-08	shear-storm	dallas	2024-05-10	0 ppm	0 ppm	0 ppm	18:27:30	-906 s ✗
2024-04-08	shear-storm	dallas	2024-05-11	728586 ppm	685750 ppm	42836 ppm	18:42:30	-6 s ✓
2024-04-08	shear-storm	indianapolis	2024-05-10	0 ppm	0 ppm	0 ppm	18:52:30	-927 s ✗
2024-04-08	shear-storm	indianapolis	2024-05-11	733664 ppm	724580 ppm	9084 ppm	18:52:30	-927 s ✗
2024-04-08	shear-storm	mazatlan	2024-05-10	0 ppm	0 ppm	0 ppm	17:52:30	-1027 s ✗
2024-04-08	shear-storm	mazatlan	2024-05-11	471758 ppm	539170 ppm	-67412 ppm	17:52:30	-1027 s ✗
2024-04-08	shear-storm	millstone	2024-05-10	0 ppm	49585 ppm	-49585 ppm	19:17:30	-709 s ✗
2024-04-08	shear-storm	millstone	2024-05-11	667827 ppm	665528 ppm	2299 ppm	19:17:30	-709 s ✗
2024-04-08	shear-storm	norwich	2024-05-10	0 ppm	0 ppm	0 ppm	19:17:30	-648 s ✗
2024-04-08	shear-storm	norwich	2024-05-11	687000 ppm	684796 ppm	2204 ppm	19:17:30	-648 s ✗
2026-08-12	control	coruna	2026-07-16	91635 ppm	57229 ppm	34406 ppm	18:17:30	-648 s ✗
2026-08-12	control	coruna	2026-08-05	0 ppm	0 ppm	0 ppm	18:17:30	-648 s ✗
2026-08-12	control	latrabjarg	2026-07-16	31700 ppm	95949 ppm	-64249 ppm	17:32:30	-784 s ✗
2026-08-12	control	latrabjarg	2026-08-05	0 ppm	0 ppm	0 ppm	17:47:30	116 s ✓
2026-08-12	control	leon	2026-07-16	71830 ppm	43561 ppm	28269 ppm	18:27:30	-101 s ✓
2026-08-12	control	leon	2026-08-05	0 ppm	0 ppm	0 ppm	18:17:30	-701 s ✗
2026-08-12	control	reykjavik	2026-07-16	62519 ppm	106097 ppm	-43578 ppm	17:37:30	-676 s ✗
2026-08-12	control	reykjavik	2026-08-05	0 ppm	0 ppm	0 ppm	17:37:30	-676 s ✗
2026-08-12	eclipse	coruna	2026-08-12	113004 ppm	0 ppm	113004 ppm	18:32:30	252 s ✓
2026-08-12	eclipse	latrabjarg	2026-08-12	293964 ppm	100157 ppm	193807 ppm	17:52:30	416 s ✓
2026-08-12	eclipse	leon	2026-08-12	15304 ppm	0 ppm	15304 ppm	18:37:30	499 s ✓
2026-08-12	eclipse	reykjavik	2026-08-12	197572 ppm	85768 ppm	111804 ppm	17:42:30	-376 s ✗

Measured two-clause separation (lag gate [-300, +1200] s): min eclipse excess 15304 ppm · eclipse lag failures 1 · max lag-passing non-eclipse excess 42836 ppm (shear-storm dallas 2024-05-11) — no clean separation (recorded raw)

Excess = median depletion inside the gate window − median over the ±2 h flanks (exact integers, ledger-direct). A shadow is confined and rides its own maximum; a storm floods the flanks and pins to window edges. Generated live from the V252 ledger.

The principle

A magnitude is one number; blindness is what one number cannot hold. Shape — confinement to the forcing's own minutes, coherence with the forcing's own clock — is what the shear read adds, and each story above is a class of event that scalar methods will keep glossing over forever, not until better tuning: the weak-but-true signal (1), the strong-but-false one (2), the structured impostor (3), the drifting calendar (4), and the moving calibration (5). The prediction extensions carry these lessons forward as sealed, falsifiable claims for August 12.