Observation

Right Ascension and Declination

I have added to the object hover text RA/DEC, which can be used for observation. 

Compare the RA/DEC in the two images above. The top image is from the orrery.  I plotted Mars on August 26, 2025 23:57 and hovered over the Mars object. The bottom image is a screenshot of the JPL Horizons ephemeris for Mars at approximately the same time, 23:58. 

You will see that the difference between the two readings is less than one second of arc for both RA and DEC. In parentheses, you can see the precision of the JPL data, which was assessed by the DE440/441 ephemeris model. For Mars, this is extremely precise, +/- 0.002 seconds of arc. Please note, these are estimates of precision, meaning consistency between the data, not accuracy. 

The hover text displays with a lot of information about Mars at this point in time, including RA/DEC. It notes that the RA/DEC is for the "apparent" rather than the ICRF reference frame. ICRF means, International Celestial Reference Frame. It's the current standard celestial coordinate system adopted by the International Astronomical Union (IAU) in 1998. On the other hand, the apparent coordinates are coordinates as an observer would see the object, fetched from JPL Horizons. 

"Apparent" in JPL Horizons RA/Dec refers to coordinates that include several corrections for what an observer actually sees from Earth at a specific moment:

1. Precession and Nutation

   • Earth's axis wobbles like a spinning top over 26,000 years (precession)

   • Plus smaller wobbles on ~18.6 year cycles (nutation)

   • Shifts coordinates from the J2000/ICRF reference to the "of-date" frame

2. Aberration of light

   • Earth's orbital motion (~30 km/s) causes apparent position shifts

   • Like rain appearing to slant when you're driving

   • Can shift positions by up to 20 arcseconds

3. Light-time

   • Where the object appears based on when light left it

   • Not where it is "now" but where we see it

   • Mercury's light takes ~8 minutes, so we see where it was 8 minutes ago

4. Gravitational deflection (for high precision)

   • Sun's gravity bends light paths slightly

   • Most noticeable for objects appearing near the Sun

Contrast with other coordinate types:

Astrometric (ICRF): True direction in space without aberration Geometric: Instantaneous position without light-time delay
Apparent: What you'd measure with a telescope right now

For our Mars example, "apparent" means these are the actual coordinates you'd use to point a telescope on August 26, 2025 23:58, including all the effects of Earth's motion and orientation at that instant.

Precision:

Not all objects will receive the same precision estimate. Below is the precision selection priority:

Selection Priority (in order):

1. JPL Horizons 3-sigma (when available in ephemerides query)

   • Real-time observational uncertainties

   • Fetched dynamically from JPL Horizons

   • Typically shows "n.a." for well-known objects like planets

   • Most useful for: asteroids, comets, newly discovered objects

   • Example: Eris returning actual values

2. JPL DE440/441 

   • Pre-computed precision based on ephemeris model quality

   • Represents how well we know the orbit, not current observation error

   • Always available for ~300 major objects

   • Example: Mars = ±0.002″

3. Type-based typical values (fallback)

   • Generic estimates based on object class

   • Used when no data available from above sources

   • Example: major planets = ±1″

Key Differences:

JPL Horizons 3-sigma:

• What it is: Current observational uncertainty

• Changes: Yes, improves with new observations

• Best for: Objects with evolving orbits or limited data

• Limitation: Often unavailable for well-established objects, such as Mars

DE440/441 values:

• What it is: Model precision limits based on all historical data

• Changes: Only with new ephemeris releases (every ~5-10 years)

• Best for: Planets, moons, well-studied asteroids

• Limitation: Static values between ephemeris updates

Typical uncertainties:

• What it is: Conservative estimates by object category

• Changes: No, hardcoded

• Best for: New discoveries, objects not in catalogs

• Limitation: May overestimate uncertainty for known objects

The system tries real-time data first (most current), then uses model-based values (most reliable), and finally falls back to estimates (most conservative).