Time (again) and Ganymede

[BrianJ]

Hi,
it was my understanding that the Celestial Body modules in Orbiter use TDB time frame to set their pos/vel, rather than UTC.

So any vessels or celestial bodies added to Orbiter using external data (e.g. JPL Horizons data) should use the TDB time as the epoch.

But...I just grabbed the pos/vel state vectors for Europa Clipper from JPL Horizons, relative to Jupiter, using TDB time frame, for just before the initial Ganymede flyby on arrival at Jupiter.

I was expecting a flyby of Ganymede at ~250km....but WHAM! I smack into the surface at unpleasantly high velocity. Periapsis altitude -415km.

So...I grabbed the pos/vel vectors using UTC time frame instead.
And I get a nice Ganymede flyby at 275km.

The difference between TDB and UTC is ~69s.
Ganymede orbital velocity ~10km/s
Periapsis altitude difference between the TDB vs. UTC data sets is ~690km
Which makes me think it is a TDB vs. UTC issue.

So..
Am I wrong in thinking I should use TDB as epoch when grabbing vectors or elements from JPL Horizons or elsewhere?
Or is it that ganymede.dll (and possibly other Galilean satellites) use UTC? (I suspect this is the case)
Or has there been an mix up between TDB/UTC at JPL Horizons?!

Anyway, I thought it might be of interest to Orbiter developers, or anyone simulating real life missions at Jupiter.

Here are the two data sets from JPL Horizons I used for epoch:

Date MJD  62601.5

TDB

  STATUS Orbiting Jupiter
  RPOS  1154593569.42984  49149450.9101117  427453876.026179
  RVEL -12898.4977350327 -673.39823531841  8594.24079393013

UTC

  STATUS Orbiting Jupiter
  RPOS  1153700990.34268  49102853.3477609  428048404.614901
  RVEL -12903.9378378357 -673.631521003498  8592.22987175082

Cheers,
BrianJ

 

[MontBlanc2012]

I've just noticed this post.

A couple of factoids that might be relevant. Orbiter uses the VSOP87 ephemeris for planetary positions, not a DEXXX ephemeris from JPL. The VSOP87 ephemeris uses the time coordinate known as TT (Terrestrial Time), which is a modern continuation of the former standard ET (Ephemeris Time). TDB (Barycentric Dynamical Time) is a relativistic time scale that is used to describe the motion of solar system bodies from the solar system barycenter. TDB is related to TT through periodic and secular corrections. However, VSOP87 itself does not directly use TDB; it uses TT as its time coordinate. So, picking up time and state-vectors from JPL Horizons and dumping them into Orbiter is not going to be precise.

Moreover, in calculating the position of the Galilean moons, Orbiter uses a rather outdated GalSat ephemeris (originally coded into Fortran). This almost certainly will produce results that differ from JPL Horizons data. Another source of error when using JPL Horizons data with Orbiter.

 

[Boxx]

this is quiet well explained here by ESA:

 

[BrianJ]

Thanks for the clarifications
The different time coordinates always confuse me (not difficult, I know).
I realised that my thinking on the orbital velocity/flyby distance/time coincidence was too simplistic.
And its just comparing different mathematical models, of course.
Anyhow, thanks guys, and have a good Christmas!
Brian

 

[toutatis]

Thanks for the clarifications
The different time coordinates always confuse me (not difficult, I know).
I realised that my thinking on the orbital velocity/flyby distance/time coincidence was too simplistic.
And its just comparing different mathematical models, of course.
Anyhow, thanks guys, and have a good Christmas!
Brian

Need to use SPICE Kernels for celestial bodies in Orbiter... Then things will be okay.