avalon-vanguard/star-map
2
0
Fork 0
Code Issues Pull Requests 1 Actions Packages Projects Releases Wiki Activity

Add Charon, the moons of Uranus, Saturn's other large moons and the four dwarf planets past Pluto's table

Browse Source 
The solar system stopped at 18 bodies: Pluto without Charon, Uranus without a moon, Saturn with
Titan alone, no dwarf planet but Pluto (audit #22). bodies.json now holds 38: the eight planets,
the five IAU dwarf planets, and every moon in JPL's mean-element table more than 100 km in mean
radius. New: Ceres, Eris, Haumea, Makemake; Mimas, Enceladus, Tethys, Dione, Rhea, Hyperion,
Iapetus, Phoebe; Miranda, Ariel, Umbriel, Titania, Oberon; Nereid, Proteus; Charon. Search finds
each by name (it indexes bodies.json), each has a body page, and the Sun's system draws them.

Where they come from
- Moons: the same archived JPL satellite table as the others. Uranus's and Pluto's are given
  against the planet's equator, with the IAU WGCCRE 2015 poles: Pluto's as the IAU gives it
  (132.993, -6.163), Uranus's at the end the table measures inclinations from (77.311, 15.175)
  with its nodes counted 180 degrees on, from the IAU pole's crossing; read without that offset
  every Uranian moon was 180 degrees from Horizons at every date from 1980 to 2100.
- Two rows are corrected where they disagree with JPL's own ephemeris and the reason is known.
  Pluto's section prints epoch 2000 Jan 1.0; JPL's current table gives Charon's as 2000-01-01.5,
  and at 1.0 Charon was 27.8-28.2 degrees from Horizons at every date, half a day of its motion.
  Phoebe's mean motion gives 548.02 days where its Horizons page and the current table give
  550.30 (the table's own note says its source misstated retrograde moons' mean motions); on the
  row's figure Phoebe was 24.6 degrees out by 2025 and 100 by 2075.
- Dwarf planets: JPL SBDB osculating heliocentric elements with their epoch (2026 Jun 9), carried
  at their own n. Against Horizons (heliocentric, 1950-2300; the clock only runs forward from now):
  Ceres 0.02 degrees in 2025, 1.9 in 2050, 4.0 in 2075, 5.3 in 2100, 11.6 in 2200 (Jupiter pulls
  on it and nothing here carries that); Eris within 0.06 to 2100 and 0.5 to 2300; Haumea within
  0.35 to 2100; Makemake within 0.25 to 2100 and 1.7 by 2200.
- Size and spin: Horizons pages for the moons (Charon 606 km, Miranda 235.7 as the mean of its
  three axes). The SBDB for Ceres (469.7 km, 9.074 h) and for the other three's spins (Eris 25.9 h,
  Haumea 3.915 h, Makemake 22.83 h). Neither source nor the WGCCRE 2015 report has a radius for
  Eris, Haumea or Makemake, so each carries its stellar-occultation measurement: Eris 1163 km
  (Sicardy et al. 2011), Makemake 715 (Brown 2013, the mean of 1434 x 1434 x 1422 km), and
  Haumea 797.6, the radius of a sphere of its volume: it is triaxial, 1161 x 852 x 513 km
  (Ortiz et al. 2017), and is drawn as that sphere.
- Rotation uses the branch's model. Every moon is locked except three: Hyperion's page says
  "Chaotic" and Nereid's gives no spin, so both are left still; Phoebe turns in 9.274 h.
- Charon carries massRatio 0.12205, the GM ratio of the two Horizons pages (106.10 / 869.326), so
  it and Pluto are drawn round their barycentre 2 131 km from Pluto's centre.

Validators (tools/etl/build.ts, on the real catalogue; full npm run etl passes)
- Offsets from Horizons on 2025-01-01, new bodies: dwarf planets at most 0.016 degrees (Ceres),
  under the 0.25 ceiling; moons Dione 0.009, Ariel 0.058, Rhea 0.070, Charon 0.111, Oberon 0.142,
  Titania 0.185, Umbriel 0.219, Proteus 0.245, Enceladus 0.309, Phoebe 0.984, Miranda 1.162,
  Tethys 2.042, under the 2.5 ceiling, which is unchanged.
- Four moons get their own ceiling, each just above its worst offset at twelve dates from 1980
  to 2100 and each named with its reason: Mimas 46 (measured up to 44.7: its resonance with
  Tethys swings its longitude 44 degrees either way over 70.8 years, which the table has no
  column for), Hyperion 21 (20.2; held in resonance by Titan, and the row's eccentricity 0.0232
  is under a quarter of the current table's 0.105), Iapetus 11 (10.1; the row sits 9.4 degrees
  behind Horizons at its own epoch and keeps that, with its plane within 0.07 degrees and its
  period within 0.001 per cent), Nereid 3 (2.6 in 2025; eccentricity 0.75).
- New checks: every body has a radius over 0 (Charon's would have been 0 before the page
  parser learnt its form); a freely spinning moon is not locked; a moon with a mass ratio puts
  the barycentre outside its planet; there are 5 dwarf planets.
- Negative controls, each a full npm run etl on the real catalogue refused with the named
  message: Uranus's node offset removed (Miranda 172.50 degrees), Charon at the printed epoch
  (28.08), Phoebe on the row's mean motion (24.61), Charon's radius unread (no radius),
  free spinners locked (Hyperion), mass ratio inverted (barycentre 17 460 km out).

Measured in the running app (port 4311): the Sun's system has 38 members ("13 + 25 moons");
Charon comes back to within 0.0004 degrees of where it started after 6.38723 days and is 179.98
degrees round after half that; Pluto is 2 130.6 km from the barycentre and Charon 17 456.8,
exactly opposite; Saturn's moons in order of distance now: Mimas 185 617 km, Enceladus 238 042,
Tethys 294 648, Dione 376 805, Rhea 526 964, Titan 1 231 389, Hyperion 1 470 453, Iapetus
3 637 059, Phoebe 11 740 900. At the arrival framing the dwarf planets are held at the 3 px
floor and the moons at 1.5 px, half their planet's drawn radius, the scene's existing rule.
Searching Charon, Enceladus, Ceres, Titania, Makemake and Phoebe each finds the body; the body
pages show Charon 6.39 d and 606 km, Titania 8.71 d, Ceres 4.6 yr and 470 km, Haumea 283 yr and
798 km, Hyperion 21.3 d, each with its orbit source. Long tasks on entering: see the previous
commit.

The Sun's note now says the four dwarf planets are on the SBDB's osculating elements. Holding
Eris's orbit, the arrival framing widens: 192 AU of range on a 1600 x 1000 window, under the
200 AU ceiling.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
This commit is contained in:
Senrokai
2026-09-24 21:54:56 +02:00
co-authored by Claude Opus 5.5
parent de34ffff43
commit 1d42be2ad5
5 changed files with 709 additions and 33 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tools/etl/build.ts
+51 -10
View File
@@ -8,7 +8,7 @@ import { ExoplanetRecord } from '../../src/app/shared/models/exoplanet.model';
import { StarRecord, SUN_STAR_ID } from '../../src/app/shared/models/star.model';
import { fetchDeepSky } from './fetchDeepSky';
import { fetchExoplanets } from './fetchExoplanets';
import { fetchSolarSystem } from './fetchSolarSystem';
import { fetchSolarSystem, FREELY_SPINNING_MOONS } from './fetchSolarSystem';
import { BYTES_PER_STAR_META, BYTES_PER_STAR_POSITION, decodeStarCatalog, encodeStarCatalog } from '../../src/app/shared/models/star-catalog';
import { fetchStars } from './fetchStars';
import { describeSources } from './sources/registry';
@@ -145,6 +145,23 @@ function validateMerge(stars: StarRecord[]): void {
*/
const MAX_PLANET_OFFSET_DEG = 0.25;
const MAX_MOON_OFFSET_DEG = 2.5;
const KM_PER_AU = 149597870.7;
/**
* The moons whose table row cannot come within that, each for a reason no mean ellipse carries,
* with a ceiling just above its worst offset from Horizons at twelve dates from 1980 to 2100:
*
* - Mimas, 44.7 degrees: its resonance with Tethys swings its mean longitude 44 degrees either
* way over 70.8 years, and the table has no column for it (Tethys, on the other end, swings 2).
* - Hyperion, 20.2: held in a 4:3 resonance by Titan; the row's eccentricity, 0.0232, is less than
* a quarter of the 0.105 JPL's current table gives.
* - Iapetus, 10.1: the row sits 9.4 degrees behind Horizons at its own epoch, 2000 Jan 1.5, and
* keeps that offset; its plane agrees with Horizons' to 0.07 degrees and its period to 0.001 per
* cent, so the fault is in the row's longitude, which this has no second source to correct.
* - Nereid, 2.6: an eccentricity of 0.75, the largest here, which a mean ellipse follows least
* well: under 0.9 degrees in every year measured but 2025 and 2030 (2.6 and 2.3) and 2100 (1.7).
*/
const MOON_OFFSET_CEILINGS_DEG: Record<string, number> = { mimas: 46, hyperion: 21, iapetus: 11, nereid: 3 };
function angleBetweenDeg(a: { x: number; y: number; z: number }, b: { x: number; y: number; z: number }): number {
const cosine = (a.x * b.x + a.y * b.y + a.z * b.z) / (Math.hypot(a.x, a.y, a.z) * Math.hypot(b.x, b.y, b.z));
@@ -169,28 +186,52 @@ function validateBodies(bodies: BodyRecord[], horizonsOrbits: Map<string, Orbita
const truth = eclipticToEquatorial(positionAtEpoch(horizons!));
const mean = positionAtEpoch(meanElementsAt(body.orbit, body.rates, horizons!.epochJd));
const offset = angleBetweenDeg(body.laplacePole ? laplacePlaneToEquatorial(mean, body.laplacePole) : eclipticToEquatorial(mean), truth);
const ceiling = body.kind === 'moon' ? MAX_MOON_OFFSET_DEG : MAX_PLANET_OFFSET_DEG;
const ceiling = body.kind === 'moon' ? (MOON_OFFSET_CEILINGS_DEG[body.id] ?? MAX_MOON_OFFSET_DEG) : MAX_PLANET_OFFSET_DEG;
assertCondition(
offset <= ceiling,
`${body.name}'s mean elements put it ${offset.toFixed(2)} degrees from where Horizons has it (at most ${ceiling} expected) — the elements were read wrongly.`
);
offsets.push(`${body.id} ${offset.toFixed(3)}`);
// A radius of 0 is what a page whose radius no pattern reads comes out as — Charon's did.
assertCondition(body.radiusKm > 0, `Body ${body.id} has no radius; its page states it in a form the ETL does not read.`);
if (body.kind === 'moon') {
assertCondition(!!body.parentBodyId && ids.has(body.parentBodyId), `Moon ${body.id} has no valid parentBodyId.`);
// Every moon here is tidally locked: its day is its orbit, from the same mean motion that
// carries it round, or its face turns away from its planet: the Kepler period of the
// osculating orbit this used to take would turn the Moon's five degrees an orbit.
const parent = bodies.find((candidate) => candidate.id === body.parentBodyId);
assertCondition(parent !== undefined, `Moon ${body.id} has no valid parentBodyId.`);
const orbitHours = (360 / body.rates.meanMotionDegPerDay) * 24;
assertCondition(
body.rotationPeriodHours !== undefined && Math.abs(body.rotationPeriodHours - orbitHours) <= orbitHours * 1e-9,
`Moon ${body.id} turns once in ${body.rotationPeriodHours} hours but goes round in ${orbitHours} — it will not keep one face to its planet.`
);
if (FREELY_SPINNING_MOONS.has(body.id)) {
// Hyperion tumbles and Nereid's page gives no spin, so they have none; Phoebe turns in
// 9.27 hours against a 550-day orbit. A lock here would be the rule below misapplied.
assertCondition(
body.rotationPeriodHours === undefined || Math.abs(body.rotationPeriodHours - orbitHours) > orbitHours * 0.1,
`Moon ${body.id} does not keep one face to its planet, yet turns once in ${body.rotationPeriodHours} hours against an orbit of ${orbitHours}.`
);
} else {
// Every other moon here is tidally locked: its day is its orbit, from the same mean motion
// that carries it round, or its face turns away from its planet: the Kepler period of the
// osculating orbit this used to take would turn the Moon's five degrees an orbit.
assertCondition(
body.rotationPeriodHours !== undefined && Math.abs(body.rotationPeriodHours - orbitHours) <= orbitHours * 1e-9,
`Moon ${body.id} turns once in ${body.rotationPeriodHours} hours but goes round in ${orbitHours} — it will not keep one face to its planet.`
);
}
if (body.massRatio !== undefined) {
// The pair's barycentre, which the planet's elements place, must lie outside the planet —
// that is why the two are drawn going round it — and nearer the planet than the moon.
const offsetKm = (body.orbit.semiMajorAxisAu * KM_PER_AU * body.massRatio) / (1 + body.massRatio);
assertCondition(
body.massRatio > 0 && body.massRatio < 1 && offsetKm > parent!.radiusKm,
`${body.name}'s mass ratio ${body.massRatio} puts its barycentre ${offsetKm.toFixed(0)} km from ${parent!.name}'s centre, which is not between its surface, ${parent!.radiusKm} km out, and the moon.`
);
}
}
}
const planetCount = bodies.filter((body) => body.kind === 'planet').length;
assertCondition(planetCount === 8, `Expected 8 planets, found ${planetCount}.`);
const dwarfCount = bodies.filter((body) => body.kind === 'dwarf').length;
assertCondition(dwarfCount === 5, `Expected the IAU's 5 dwarf planets, found ${dwarfCount}.`);
console.log(` mean elements against Horizons, degrees: ${offsets.join(', ')}.`);
}