# Sundog Atlas — Lit-Pass Memo (citation spine + gap map)

> **STATUS: STUB / SCAFFOLD opened 2026-06-05.** The formal reading is Phase 0.5 of
> `docs/SUNDOG_V_ATLAS.md` and is a **hard precondition** for any atlas claim or public surface. This
> file names the citation spine to ground against and the tracks to cover; entries marked
> *(to-verify)* are NOT yet read — they are the targets of the pass, not banked facts. House model:
> `docs/P_V_NP_LITPASS_MEMO.md`.

## 0. Why this lane needs a lit-pass at all

The atlas invents **no** physics. Its apparatus, its geometry, and its catastrophe-theoretic skeleton
are entirely borrowed; the Sundog contribution is the *synthesis* (forward-generated complete atlas as
a **classified bifurcation diagram** + the determining-shadow-set thesis embodiment + the directed
invisible-halo search). A synthesis lane stands or falls on attribution discipline and on not
re-deriving (or misattributing) known results. The public workbench already has a flagged
above-the-fold attribution gap (`docs/calibration/PHASE11_OUTREACH_SYNTHETIC_MEMO.md` W2); this memo
is the instrument that closes it before any tier-3 outreach.

## 1. Tracks (what the pass must settle)

- **Track A — Halo geometry & simulation (the body).** The ray-optics theory of halos and the
  Monte-Carlo simulators. Establish exactly what is canonical, what is each author's, and what the
  apparatus (HaloSim) already computes, so Sundog claims nothing already standard as novel.
- **Track B — Catastrophe / caustic optics (the skeleton). ✅ READ 2026-06-07 →
  `ATLAS_PHASE8_STRATA.md`.** **BANKED FINDINGS (prior-art resolution):** (1) **Tape 1980**
  (JOSA 70:1175) established caustic = Jacobian-kernel (fold level); **Tape & Können 1999** (Appl. Opt.
  38:1552, full text read) is pure *parameterization* — no caustic/catastrophe/bifurcation classification
  (it even dissolves the UTA/LTA merge into "one halo"). (2) **Berry 1994** (Appl. Opt. 33:4563) is the
  decisive prior art: it *raised* catastrophe-typing of ice halos, recorded "some halos do show cusps"
  (Tape), flagged the halo map as **NON-GRADIENT** (Thom A_n/D_n taxonomy does not transfer wholesale),
  **predicted no umbilics (D₄)**, and left the **swallowtail (A₄) as an open question**. (3) **Verdict:
  the systematic *stratification* is genuine bounded-novelty (taking up Berry's question, addressing
  non-gradient); NOT "first catastrophe classification" of halos.** Whitney's theorem rescues the
  fold/cusp (A₂/A₃) labels (generic for any 2-D map); higher strata need the jet-determinacy check, not
  corank alone. The 29.7° merge = A₃ Whitney cusp; no D₄ found (confirms Berry).
- **Track C — The two-component wall taxonomy.** Confirm the literature's treatment of (A) caustic
  catastrophes vs (B) ray-admissibility / TIR boundaries, and whether the *union* is named as the
  organizing structure of the complete atlas anywhere.
- **Track D — Determinacy / inverse problem.** Prior art on the halo inverse problem (recovering
  crystal populations from a display) and on what a display *does* and *does not* determine — the
  determining-shadow-set framing in optics terms.
- **Track E — Crystal size from diffraction / Shadow 2 (Phase 8.5). ✅ READ 2026-06-07 →
  `docs/atlas/S2_LITPASS_E_G.md`.** Prior art on reading particle size off diffraction structure:
  coronae (pure forward diffraction, ring radius ∝ λ/a — the *existence proof* and calibration anchor);
  rainbow supernumeraries → drop size (Airy theory). **BANKED FINDING:** the ice-halo *fine-fringe*
  analogue is essentially **NEGATIVE** — the 22° halo edge is a *step* singularity whose diffraction
  function has **zero-contrast shoulders, no maxima** (Berry, Appl. Opt. 33:4563, 1994 — "rarely if ever
  seen"); parhelia fringes have only 0.178 contrast. **The size shadow's legible home is the CORONA**
  (Airy `[2J₁(x)/x]²`, `θ∝λ/a`, monodisperse→aureole washout), NOT refraction-halo fringes. HaloSim
  (ray-optics) cannot generate either — a standalone wave layer is required.
- **Track G — Polarization / handedness / Shadow 3 (Phases 8.6–8.7). ✅ READ 2026-06-07 →
  `docs/atlas/S2_LITPASS_E_G.md`.** Können, *Polarized Light in Nature* — the anchor. **BANKED
  FINDINGS:** (1) **honesty gate confirmed** — ice Ih is achiral (`P6₃/mmc`, optically positive
  uniaxial, `Δn=+0.0014`); halo circular polarization is **ray-path parity × c-axis birefringent
  retardance, NOT molecular chirality**. (2) The minimal **Mueller chain** (Fresnel diattenuator ×
  retarder `δ=2πΔn·L/λ`) gives `V/I ∝ sin(2φ)·sin(δ)` that sign-flips with parity — physically real,
  ~1% magnitude. No Stokes/Mueller ray-tracer exists; must be built standalone. (3) **⚠ `SYNTHESIS`:**
  there is **NO published observation of nonzero Stokes-V in a visible ice halo** (Können measures
  *linear* only), and "*net* circular polarization = population handedness imbalance" is a **novel,
  uncited Sundog framing** — net-V also cancels in mirror-symmetric populations. The handedness leg is a
  **PREDICTED, not observed, observable** and must be framed so. Ice-phase (halo radius) is the robust
  discrete anchor instead.
- **Track F — Attribution & priority map.** A per-claim ledger: for every statement the atlas will
  make, name the source it rests on, and flag any that read as a Sundog priority claim for demotion.

## 2. Citation spine (to verify / expand in the pass — NOT yet read)

**Halo geometry & simulation (Track A):**
- Greenler, R. — *Rainbows, Halos, and Glories* — the classic; pioneering Monte-Carlo halo
  simulation. *(to-verify: edition/year; which halos he first simulated.)*
- Tape, W. — *Atmospheric Halos*; Tape, W. & Moilanen, J. — *Atmospheric Halos and the Search for
  Angle x* — the geometric theory; the lab already cites "Tape AH-CH06" for the 29° merge. *(to-verify:
  the 29° / 32° / 58° statements and their exact framing.)*
- Können, G. P. — *Polarized Light in Nature* — halo polarization & geometry. *(to-verify.)*
- Cowley, L. — Atmospheric Optics (atoptics) — reference site; **HaloSim** is **Cowley & Schroeder, M.**
  *(to-verify: canonical citation form for the apparatus; this is the tool Sundog runs.)*

**Catastrophe / caustic optics (Track B):**
- Thom, R. — *Structural Stability and Morphogenesis* — the elementary catastrophes (A₂ fold, A₃ cusp,
  A₄ swallowtail, D₄ umbilics). *(to-verify.)*
- Arnold, V. I. — caustic / wavefront singularity classification (A_n, D_n, E_n). *(to-verify.)*
- Berry, M. V. & Upstill, C. — "Catastrophe optics: morphologies of caustics and their diffraction
  patterns" (Progress in Optics) — **the** optical-catastrophe reference; the Airy/Pearcey dressing.
  *(to-verify: whether it treats halos specifically.)*
- Nye, J. F. — *Natural Focusing and Fine Structure of Light: Caustics and Wave Dislocations*.
  *(to-verify.)*
- Berry, M. V. — rainbow-as-fold / "Waves and Thom's theorem"; diffraction catastrophes (Airy on the
  fold, Pearcey on the cusp). *(to-verify the cleanest citation.)*
- **Berry, M. V. — "Supernumerary ice-crystal halos?", Appl. Opt. 33(21):4563 (1994). ✅ READ — the
  decisive ice-halo prior art:** non-gradient halo map → Thom taxonomy doesn't transfer wholesale; cusps
  occur; **no umbilics predicted**; swallowtail open. (Also: 22° edge = step, not fold — `S2_LITPASS_E_G`.)
- **Tape, W. — "Analytic foundations of halo theory", JOSA 70(10):1175 (1980). ✅** caustic = image of
  the halo function's singular set = Jacobian kernel (the fold-level machinery; no A_n/D_n typing).
- **Tape, W. & Können, G. P. — "A general setting for halo theory", Appl. Opt. 38(9):1552 (1999). ✅
  READ (full text):** the oriented-wedge atlas (wedge angle × spin-vector) — pure *parameterization*, NO
  caustic/catastrophe/bifurcation classification (dissolves the UTA/LTA merge into "one halo").

## 3. Gap map (what the pass is looking for)

The pass must place each item below into one of: **STANDARD** (already in the literature — credit,
don't claim), **SYNTHESIS** (Sundog's recombination of standard parts — defensible as framing), or
**OPEN** (genuinely unaddressed — candidate contribution, still bounded).

1. "A halo is a fold caustic (minimum deviation)." — expected **STANDARD** (Berry/Nye).
2. "The merge/appear/disappear transitions form the bifurcation set of the halo control space." —
   expected **SYNTHESIS** at minimum; check whether anyone states it as the organizing skeleton of the
   *complete* atlas (vs. per-phenomenon).
3. "Two-component wall taxonomy: caustic catastrophes (A) ∪ ray-admissibility boundaries (B)." — check
   whether the union is named; expected **SYNTHESIS/OPEN**.
4. "The bifurcation set is the determining shadow of the atlas topology (read transitions off it, can't
   rebuild the display)." — the cross-substrate framing; expected **SYNTHESIS** (Sundog vocabulary over
   standard non-invertibility).
5. "Invisible halos live at the higher-codimension strata (swallowtail/umbilic)." — check prior
   predictions of rare halos via higher catastrophes (Können/Tape may have geometric versions);
   expected **STANDARD-or-SYNTHESIS** — the *directed forward search* is the Sundog move.
6. "Crystal size readable off the diffraction dressing of a halo fold." — Track E; **SETTLED
   2026-06-07: STANDARD for coronae/rainbows (Airy size law); NEGATIVE for the 22° ice halo** (a step,
   not a fold — zero-contrast diffraction shoulders, Berry 1994). The size shadow's legible home is the
   corona, not refraction-halo fringes.
7. "Polarization reads a discrete handedness/parity (Stokes V sign) that the intensity cannot, and is
   the *cleanly-invertible* layer of the shadow tower (invertibility tracks the hidden-variable kind:
   continuous magnitudes resist, discrete parities invert)." — Track G + §1.2; the *polarization reads
   handedness* part is expected **STANDARD** (Können); the *invertibility-tracks-kind tower framing*
   and the AB-exact-analog placement are expected **SYNTHESIS** (Sundog vocabulary over standard
   optics) — the genuinely new framing to defend, never the polarization physics. **SETTLED 2026-06-07:
   the linear-polarization halo physics is STANDARD (Können); but *circular*-V handedness in halos is
   UNOBSERVED, and the "net-V = population handedness" diagnostic is SYNTHESIS/novel — a predicted
   observable, not established. Ice-phase (halo radius) is the safer discrete invertible layer.**

## 4. Disposition (precondition for the lane)

- **Attribution requirement (blocking).** Before any tier-3 outreach or public copy: the apparatus
  (Cowley & Schroeder) and the geometry (Greenler / Tape / Können / Cowley) must be credited
  prominently and above-the-fold on the public surfaces (closes
  `PHASE11_OUTREACH_SYNTHETIC_MEMO.md` W2); catastrophe optics (Thom / Arnold / Berry / Nye) credited
  wherever the bifurcation framing appears.
- **No-priority rule.** Every roadmap claim carries its Track-F source; anything that reads as priority
  over the giants is demoted to "we apply / synthesize."
- **Bounded-novelty rule.** The defensible novelty is the *synthesis* (classified-bifurcation-diagram
  organization of the forward-generated atlas + the determinacy framing + the directed invisible-halo
  search), never the underlying physics.

---

*Sundog Research Lab — Atlas lit-pass memo (stub). The reading is Phase 0.5 and gates every outward
claim. We stand on the shoulders of giants; this is where we name them.*