H2-087 SCIENTIFIC INTENT
========================

Goal
----
Make a falsifiable prediction with the current Frostyverse precession closure
rather than tuning another closure against a known target.

The FV lane is evaluated and SHA-256 fingerprinted BEFORE the GR/reference ruler
opens.  The frozen effective model is the current best local black-box chain:

  existing mechanically supported stable response (~2D)
  + H-019/H-021/H-057 six-support release/capture multiplicity (~4D)

H2-087 does not claim that the final D <-> one release-mode microscopic bridge
has already been derived.

External inputs frozen for this test
------------------------------------
1566 Icarus:
  a = 1.078 au
  e = 0.827
  Source = Sun
  JPL radar-planning reference:
  https://echo.jpl.nasa.gov/asteroids/Icarus/Icarus_planning.html
  JPL notes that the low perihelion makes Icarus useful for relativity tests.

Europa:
  a = 671100 km
  e = 0.009
  Source GM (Jupiter) = 126686531.9 km^3/s^2
  Europa GM = 3202.71210 km^3/s^2
  JPL satellite mean-elements / physical-parameters references:
  https://ssd.jpl.nasa.gov/sats/elem/
  https://ssd.jpl.nasa.gov/sats/phys_par/

Moon:
  a = 384400 km
  e = 0.0554
  Source GM (Earth) = 398600.436 km^3/s^2
  Moon GM = 4902.800 km^3/s^2
  JPL references:
  https://ssd.jpl.nasa.gov/sats/elem/
  https://ssd.jpl.nasa.gov/sats/phys_par/

Sun GM:
  1.32712440041279419e11 km^3/s^2
  https://ssd.jpl.nasa.gov/astro_par.html

Why physical c appears before the GR ruler
------------------------------------------
The inherited FV equations use c=1 units.  For a chosen numerical length label,
the program converts a real source GM into the dimensionless compactness
GM/(c^2 L).  This is a unit conversion / source-strength input, not a GR
perihelion target.  The GR perihelion formula is not evaluated until after the
FV prediction fingerprint is sealed.

Numerical controls
------------------
Icarus has high eccentricity and reaches much deeper relative to its semimajor
axis than the original Mercury-centered test family.  H2-087 uses a compact,
fixed-total-loading numerical source so Icarus stays outside the source while
preserving the exterior source strength.  A second coordinate normalization is
run with the same physical GM/(c^2 r) to detect coordinate-label artifacts.

Europa has e ~= 0.009, even closer to circular than most prior cases.  Its
precession extraction is therefore deliberately stress-tested by repeating the
same Jupiter/Europa-a setup at e=0.02, 0.05, and 0.10 WITHOUT using GR.  If the
coordinate-normalization spread collapses as e increases, the actual-Europa
number is classified as conditioning-limited rather than new physics.

Moon mass bookkeeping
----------------------
The inherited FV orbit lane is effectively test-particle-like, so SOURCE_ONLY is
primary.  TOTAL_TWO_BODY is reported separately because Moon/Earth mass ratio is
not negligible.  This diagnostic does not silently modify the FV force law.

Observational caveat
--------------------
JPL mean apsis periods are about 1.394 yr for Europa and 5.997 yr for the Moon.
Those raw apsidal motions are dominated by Jupiter's oblateness / satellite
perturbations and by solar/three-body / Earth-figure effects, respectively.
They are context only, not the isolated central-source relativistic observable.

Icarus is the clean external prediction target.  Published radar/astrometry
literature commonly quotes a relativistic perihelion-shift scale of about
10 arcsec/century for Icarus.  H2-087 does not use that number until after the
FV prediction is sealed and does not retune afterward.
