Physical Heliacal Visibility Phase 4 Jones Input-Authority Checkpoint

Date: 2026-07-31 Status: input-authority gate closed; MYSTIC pilot authorized; runtime model not admitted Candidate identifier: jones_paranal_scattered_moonlight_2013_v1

Outcome

The Jones candidate's solar, lunar-reflectance, and aerosol phase-function inputs now have separate, machine-checkable truth classifications:

• the solar spectrum is independently reconstructable over the first candidate's 380-780 nm domain;
• the ROLO lunar-reflectance coefficients are independently reconstructable inside the 1.55-97 degree empirical phase domain; and
• ESO's selected mie_m15s1.dat aerosol phase function is an exact, checksum-locked source-owned input, but its transformation from the published particle parameters is not reconstructable from the identified public sources.

That final result is not hidden or treated as a model failure. The exact external ESO aerosol table may govern an offline independent radiative-transfer pilot, with its receipt and source-owned role visible. The pilot may claim independent radiative transfer, not independent aerosol microphysics. No runtime model, API, data pack, dependency, network path, or legacy behavior is admitted by this checkpoint.

Solar Irradiance Authority

The official ESO solspec_ext.dat member is locked at:

• bytes: 141021;
• SHA-256: 0f75353a72cfd7f1b314652f47aa9edff33434be5154bdfd590e2b5b2926933d;
• rows: 1741; and
• wavelength domain: 0.1195-30 micrometres.

Its first 1,467 rows were independently replayed against STScI CALSPEC sun_reference_stis_002.fits, using the explicit conversions from angstroms and erg s-1 cm-2 angstrom-1 to micrometres and W m-2 micrometre-1. The complete comparison produced:

• maximum wavelength delta: 5.48828125002121e-06 micrometres;
• maximum absolute flux delta: 0.0004998779297125111 W m-2 micrometre-1; and
• maximum relative flux delta: 4.4698819115193814e-06.

The 325 ESO samples inside the candidate response domain span 0.38061-0.77921 micrometres and have numeric signature SHA-256 fcdb33a16166f8ee3e9f894371f3d79e14efddb063753104d547897071be9024. The NMSU Colina/Bohlin/Castelli reference is retained as a separate bibliographic and low-wavelength crosscheck; it is not misrepresented as a complete replacement for the STIS file because its public text contains a blank flux interval.

The external STIS and NMSU bytes are not committed to the repository.

Lunar Reflectance Authority

Kieffer and Stone's published ROLO coefficient table is independently bound to the ESO moonalbedo.dat member:

• member bytes: 3365;
• member SHA-256: 86b9f9860fabb283de6659aabee8959186dc03e9d081aaa7c2761c2869ff16cc;
• four constant coefficients, numeric SHA-256: 6eb466a6c1d8fb4bca3f0dda99c4865a58a5c54e0c25fb782fc8358a2fe54e65;
• 32 wavelength rows from 350.0 to 2383.6 nm; and
• Table 4 numeric SHA-256: 620b1ca086edda0221a1db7461d69602479c5c584aa403843084786b5608278e.

The first candidate does not add the omitted libration terms. The paper's 1.55-97 degree empirical phase domain remains a hard boundary; cases outside it are typed not_evaluable instead of extrapolated.

Aerosol Reconstruction Falsification

The selected ESO phase table is locked at:

• bytes: 78231;
• SHA-256: dba01f9b49ddf9a547bccc7eaca013bec1e4b1d8e081ec5ec4dd284ea7ec425e;
• 40 wavelengths from 0.3 to 30 micrometres; and
• 181 scattering angles from 0 to 180 degrees at 1-degree spacing.

At 0.55 micrometres, direct integration of the table gives a half-solid-angle normalization of 1.0010108296057996 and an asymmetry parameter of 0.680602583549528.

The Jones 2013 nucleation, accumulation, coarse, and stratospheric particle parameters were replayed with the identified Oxford Earth Observation Data Group log-normal Mie implementation. The Oxford archive is locked at:

• bytes: 24542; and
• SHA-256: a026c570b2d39988e597fb7ce5b7bc3e451f650a8d6013670e68af5343cf9561.

That replay produced an asymmetry parameter of 0.595268389115, along with a large systematic phase-function mismatch; for example, its zero-degree value was about 103.904 versus ESO's 27.53286. The mismatch survives the identified integration bounds, trapezoidal quadrature, and documented log-normal parameter semantics. The public paper, official package, and identified Mie source do not disclose the extra m15s1 transformation or smoothing needed to reproduce the table.

Moira therefore records source_owned_checksum_locked_not_reconstructable. It does not invent a smoothing law, substitute the independently calculated curve for ESO's selected table, or claim the source-owned table is an independent oracle.

Pilot and Distribution Boundary

The next gate may use mie_m15s1.dat only as an explicit operator-supplied external input to a separately receipted libRadtran 2.0.6 MYSTIC pilot. The ESO source and data bytes remain outside the repository. Whether a generated artifact based on that input may be distributed requires a separate release disposition before packaging; this checkpoint makes no legal conclusion.

Acceptance thresholds remain unset until the pilot measures repeatability, Monte Carlo convergence, and interpolation behavior. Production admission, engine implementation, and public-surface work remain closed.

Machine-Checkable Files

• scripts/visibility_reference_lab/phase4_jones_input_authority_spec.json
• scripts/audit_visibility_phase4_jones_inputs.py
• tests/artifacts/visibility_reference_lab/phase4_jones_input_authority_checkpoint_2026-07-31.json
• tests/unit/test_visibility_phase4_jones_input_authority.py

Reproduce the checkpoint with explicitly supplied external files:

.\.venv\Scripts\python.exe scripts\audit_visibility_phase4_jones_inputs.py `
  C:\path\to\SM-01.tar.gz `
  --stis-solar C:\path\to\sun_reference_stis_002.fits `
  --nmsu-solar C:\path\to\solarspectra.txt `
  --eodg-archive C:\path\to\eodg_mie.tar.gz

The auditor reads and hashes the supplied files in place. It never downloads, extracts to disk, compiles, executes, or redistributes them.

Gate State

Closed by this checkpoint:

• independent solar-spectrum identity inside the candidate domain;
• primary-table ROLO coefficient identity and phase limit;
• exact ESO aerosol-table structure and numeric invariants;
• falsification of the claimed public aerosol reconstruction path;
• source-owned versus independent truth labels; and
• permission to proceed with a bounded external-input MYSTIC pilot.

Still open:

• frozen pilot geometry matrix and generation receipt;
• pilot execution, convergence, and repeatability;
• measured numerical and interpolation threshold freeze;
• independently generated immutable spectral artifact;
• independent in-domain geometry holdouts;
• release/distribution disposition for any derived artifact;
• engine implementation and typed component receipt;
• public-surface parity; and
• Phase 4 closure.