Turn Eris once in 15.77 days, locked to Dysnomia, not in the 25.9 hours the SBDB flags as unreliable

Eris took its day from the SBDB's rot_per, 25.9 hours, whose own note reads "Result based on less
than full coverage, so that the period may be wrong by 30 percent or so" (Roe et al. 2008). Eris
is locked to Dysnomia: its light curve repeats every 15.771 +/- 0.008 days (Bernstein et al. 2023,
PSJ 4, 115), Dysnomia's 15.78590-day orbit (Holler et al. 2021; Szakáts et al. 2023, A&A 669, L3).
It was drawn turning 14.6 times too fast.

Eris's BodySpec now carries that day, 378.504 hours, cited as its radius already cites Sicardy et
al., and a spec's measured day comes before its source's. build.ts checks that Eris's day is
Dysnomia's orbit within 0.2 per cent. In the running app Eris turns 5.707 degrees in six hours, as
15.771 days gives; on 25.9 hours it turned 83.4.

Makemake's SBDB period, 22.83 hours, carries the same flag; it is Hromakina et al. 2019's own
result and nothing later overturns it, so it is kept.

Control: Eris on the SBDB's period fails the ETL: "Eris turns once in 1.079 days".

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>

tools/etl/fetchSolarSystem.ts

@@ -31,6 +31,8 @@ interface BodySpec {
   sbdb?: string;
   /** A measured mean radius, in km, for a body neither Horizons nor the SBDB gives one for. */
   radiusKm?: number;
+  /** A measured sidereal day, in hours, where a later measurement overturns the one its source gives. */
+  rotationPeriodHours?: number;
   /** A moon that does not keep one face to its planet: its page's own spin, or none, is kept. */
   spinsFreely?: boolean;
   /** A moon heavy enough to move its planet round their barycentre visibly; see `BodyRecord.massRatio`. */
@@ -87,7 +89,12 @@ const BODY_SPECS: BodySpec[] = [
   // 2017 (Nature 550, 219); drawn as a sphere, at the radius of the sphere of the same volume.
   // Makemake: 1434 km across its equator and 1422 across its projected pole, Brown 2013 (ApJ 767,
   // L7); the same mean.
-  { id: 'eris', name: 'Eris', kind: 'dwarf', horizonsCommand: '136199;', center: '500@10', sbdb: 'Eris', radiusKm: 1163 },
+  //
+  // Eris's day is not the SBDB's 25.9 hours, a light curve of partial coverage (Roe et al. 2008) the
+  // SBDB itself flags as "may be wrong by 30 percent or so": it turns once in 15.771 +/- 0.008 days
+  // (Bernstein et al. 2023, PSJ 4, 115), locked to Dysnomia's 15.786-day orbit (Szakáts et al.
+  // 2023, A&A 669, L3). On the SBDB's figure it turned 14.6 times too fast.
+  { id: 'eris', name: 'Eris', kind: 'dwarf', horizonsCommand: '136199;', center: '500@10', sbdb: 'Eris', radiusKm: 1163, rotationPeriodHours: 15.771 * 24 },
   { id: 'haumea', name: 'Haumea', kind: 'dwarf', horizonsCommand: '136108;', center: '500@10', sbdb: 'Haumea', radiusKm: 797.6 },
   { id: 'makemake', name: 'Makemake', kind: 'dwarf', horizonsCommand: '136472;', center: '500@10', sbdb: 'Makemake', radiusKm: 715 },
   { id: 'moon', name: 'Moon', kind: 'moon', horizonsCommand: '301', center: '500@399', parentBodyId: 'earth' },
@@ -206,9 +213,7 @@ export async function fetchSolarSystem(): Promise<{ bodies: BodyRecord[]; horizo
     // without W by.
     const rotationPeriodHours = result.tidallyLocked || (spec.kind === 'moon' && !spec.spinsFreely)
       ? (360 / mean.rates.meanMotionDegPerDay) * HOURS_PER_DAY
-      : smallBody
-        ? smallBody.rotationPeriodHours
-        : result.rotationPeriodHours;
+      : (spec.rotationPeriodHours ?? (smallBody ? smallBody.rotationPeriodHours : result.rotationPeriodHours));
     const parentGm = spec.barycentric && spec.parentBodyId ? gmById.get(spec.parentBodyId) : undefined;
     if (spec.barycentric && (result.gmKm3PerS2 === undefined || parentGm === undefined)) {
       throw new Error(`${spec.name} and its planet need a GM each to place their barycentre.`);