moira

Physical Heliacal Visibility Phase 4 Observer-Factor Checkpoint

Date: 2026-07-31 Status: doctrine resolved and receipt hardened; Phase 4 remains open

Question Resolved

The roadmap asked for an admitted observer-factor input and valid range. The source review shows that exposing a generic experience or skill slider would misstate Crumey’s model.

In Crumey (2014), the overall field factor F may combine target, medium, laboratory-scaling, detection-practice, and personal effects. It is not an isolated observer parameter. The paper uses F = 2 as a notional illustrative value for naked-eye point-source examples; contextual values discussed elsewhere in the paper do not establish a universal validity range for individual observers.

Blackwell’s underlying threshold normalization is also a repeated-trial detection condition, not a confidence score for one observation.

Admission Decision

The named physical observer protocol keeps a fixed singleton contract:

model:
  crumey_2014_equation_53_fixed_notional_f2_v1
value:
  F = 2.0
mutable:
  false
probabilistic detection claim:
  false

There is no caller-supplied physical observer-factor field. A future calibrated observer protocol would require a new protocol/model identifier, the calibration experiment and receipt, and an independently justified domain. It would not silently mutate this protocol.

The legacy VisibilityPolicy.crumey_field_factor remains a separate existing contract and is not changed by this decision.

Receipt Hardening

PhysicalObserverProtocolReceipt now carries:

• the fixed field-factor model identifier;
• F = 2.0;
• detection_field_factor_mutable = false;
• exact Crumey source identifiers; and
• probabilistic_detection_claimed = false.

The same facts appear in the generic observer_protocol component receipt. The independently existing threshold receipt continues to report the actual factor used, so the observer and threshold receipts must agree.

Compatibility

The new receipt fields are additive defaults appended to the existing dataclass. No calculation, model identifier, policy input, legacy function, or default changes.

Local Verification

The focused spectral suite asserts:

• the physical policy exposes no detection-field-factor input;
• the observer receipt reports the immutable F = 2 contract;
• the threshold receipt used the same factor;
• no probabilistic claim is made; and
• the generic component receipt preserves the same facts.

Focused result after observer-receipt hardening: tests/unit/test_visibility_spectral.py — 55 passed.

The widened heliacal/visibility unit gate passed 464 tests, and the visibility/phase server-route gate passed 24 tests.

Remaining Phase 4 Work

• independently implement and validate the separately versioned Jones scattered-moonlight component;
• use ESO/PALACE outputs only in site-bound comparison fixtures; and
• derive explicit environmental sensitivity envelopes without calling them observer probability or scientific confidence.

Phase 4 is not closed by this checkpoint.

This site is open source. Improve this page.