Artemis II — Geocentric RA/Dec & Sky Track (J2000)

Timeline range

Geocentric sky track — RA / Dec J2000 (colour = time: blue→yellow)

RA increases left→right here but is conventionally plotted right→left; hover for values. EI = Entry Interface (~120 km altitude, atmospheric entry boundary).

Topocentric ICRF RA / Dec data table

Notes & Caveats

Source data

The input is a CCSDS OEM v2.0 (Orbit Ephemeris Message) file produced by NASA JSC FDO. It contains Cartesian state vectors (position & velocity in km and km/s) in the EME2000 reference frame (Earth Mean Equator & Equinox of J2000.0), with timestamps in UTC. The file covers the Artemis II mission from shortly after launch through Entry Interface (EI).

RA / Dec — Geocentric columns

Because the OEM frame is EME2000 — whose X-axis points toward the J2000 vernal equinox and Z-axis toward the J2000 mean celestial pole — converting to geocentric RA/Dec (epoch & equinox J2000.0) is a pure rectangular-to-spherical transformation:

RA  = atan2(Y, X)   → 0–360°
Dec = asin(Z / |r|) → ±90°

These coordinates are geocentric (origin at Earth’s centre of mass) in the ICRF/J2000 reference frame. EME2000 and ICRF agree to ~0.01″.

Topo RA / Dec — Topocentric ICRF columns (telescope-ready)

The Topo columns give the direction to the spacecraft as seen from the selected observer location (Sky View tab), expressed in ICRF coordinates. These are suitable for direct input to a telescope control system (TCS) that accepts J2000/ICRF mean place.

For OEM data: the observer’s geocentric position in the EME2000/ICRF frame is computed by:

1. Geodetic (lat, lon, height) → ECEF using the WGS84 ellipsoid
2. ECEF → true-of-date equatorial using GAST (Greenwich Apparent Sidereal Time)
3. True-of-date → mean-of-date using the inverse IAU 1980 nutation matrix (4 dominant terms: 18.6-yr lunar node, semi-annual, monthly, fortnightly)
4. Mean-of-date → J2000 using the inverse Lieske (1976) precession matrix

The topocentric vector is then: r_topo = r_spacecraft − r_observer, and RA/Dec are computed from this vector. Accuracy is ~0.5″ (dominated by the truncated nutation model).

For JPL Horizons data: the observer’s geodetic coordinates are passed directly to the Horizons API (CENTER='coord@399'), which computes rigorous topocentric astrometric ICRF coordinates using the full IAU nutation/precession models.

What the TCS does with these coordinates: A TCS (e.g. the AAT) accepts ICRF/J2000 mean place and internally applies precession, nutation, annual aberration, and atmospheric refraction to compute the apparent place and drive the telescope. These corrections are distance-independent, so the TCS handles them correctly. The only distance-dependent correction is topocentric parallax, which we compute here.

At Artemis II distances (7,000–413,000 km), the topocentric parallax can reach up to 3.2° at closest approach. The Topo dRA and Topo dDec columns give the non-sidereal tracking rates needed by the TCS, in arcsec/s on the sky (RA rate includes the cos(Dec) factor).

Proper motion

The RA and Dec rates of change (dα/dt, dδ/dt) are computed by finite differencing between consecutive ephemeris points. The total angular rate is the great-circle angular separation between consecutive points divided by the time step. The RA component is multiplied by cos(Dec) to give the true angular rate on the sky rather than the coordinate rate. Near EI, the rapidly shrinking range causes proper motion to spike to hundreds of arcsec/s.

Estimated visual magnitude

The capsule is modelled as a Lambertian sphere with:

• Radius: 2.5 m (Orion crew module is ~5.02 m diameter)
• Geometric albedo: 0.15 (rough estimate for metallic/MLI spacecraft surfaces)

The formula is:

m = −26.74 − 2.5 log₁₀( p · (r/d)² · Φ(α) )

where Φ(α) = (1/π)[sinα + (π−α)cosα] is the Lambertian phase function and α is the Sun–Object–Observer phase angle.

Caveats:

• No eclipse/shadow modelling. The spacecraft will periodically enter Earth’s shadow (especially during early parking orbits, ~30–40 min per ~1.5 hr orbit) and will be invisible during those intervals. The magnitude shown is an upper bound on brightness.
• No lunar occultation. Near closest lunar approach the Moon could briefly occult the spacecraft as seen from Earth, but the geometry is narrow and short-lived.
• Simplified shape. The real Orion vehicle (crew module + service module + solar arrays) is not a sphere. Specular glints from solar panels or metallic surfaces could make it transiently brighter by several magnitudes; conversely, an unfavourable orientation could make it fainter. The Lambertian sphere is a reasonable average.
• Albedo is uncertain. Spacecraft albedos depend on surface materials and degradation; 0.15 is a reasonable mid-range value but could easily be 0.10–0.30.
• Geocentric, not topocentric. The observer is assumed at Earth’s centre. For a real ground observer, the range (and hence magnitude) differs slightly, especially at close range.

Sky view (Alt/Az)

The sky view converts geocentric J2000 RA/Dec to local Alt/Az using GMST and the observer’s latitude and longitude. Note: the Alt/Az display uses geocentric coordinates and does not apply topocentric parallax, so the plotted position may differ from the true apparent position by up to ~3° at closest approach. The topocentric ICRF columns in the table are the authoritative telescope-pointing coordinates.

• Sun position: Low-precision formula from Meeus, accurate to ~0.01°.
• Moon position: Truncated Meeus Chapter 47 (largest 27 longitude + 18 latitude terms), accurate to ~0.3°. Good enough for sky placement but not for precise occultation predictions.
• Moon phase: Computed from the difference in ecliptic longitude between Moon and Sun. Illumination fraction = (1 − cos(elongation)) / 2. The terminator is drawn as an ellipse.
• Stars: A catalogue of ~45 bright stars (to mag ~2.5) with J2000 coordinates. No proper motion corrections are applied to the stars (negligible for 26-year extrapolation at this display scale for most stars).
• Atmospheric refraction is not modelled; objects near the horizon may appear ~0.5° higher than shown.

Time system

All times in the OEM are UTC. AEST (Australian Eastern Standard Time) is UTC+10, applied as a simple offset. No leap-second table is used — JavaScript Date handles UTC internally. The difference between UTC and TT (~69 seconds in 2026) is ignored; it affects the GMST calculation at the ~0.01° level, which is negligible for this display.

Currently loaded trajectory

Load a newer OEM file

NASA publishes updated trajectory files at the link below as the mission progresses. Download the latest .zip, extract the .asc file, then drop it here.

NASA Artemis II real-time tracking & ephemeris download page

Fetch latest from JPL Horizons

Pulls current trajectory directly from JPL Horizons (NAIF ID −1024, geocentric J2000 astrometric). May be more up-to-date than the embedded OEM. Note: visual magnitude is not available from Horizons (no XYZ → no phase angle); all other columns work normally.

⬇ Fetch from JPL Horizons

Uses allorigins.win to relay the request (required to bypass browser CORS restrictions). Your query is just JPL's public API — no private data is sent.