Celestia Forums

Hmmm, we seem to be doing reasonably OK with the position of Phobos and Deimos, despite the lack of CustomOrbits. Celestia simulates the following image pretty well:
http://panther-observatory.com/gallery/planets/doc/Mars_20030822C11.htm
(South is at the top, and the southern polar cap is rather more prominent in the image than it is in the default Celestia texture.)

Grant

Celestia didn't do bad with the Lunar eclipse the other night either!
I watched it in tandem; Celestia/Actual eclipse!!!

The Moon has a CustomOrbit, involving a lot of mathematics to place it in the correct position - so do a lot of other solar system moons. But the martian satellites are currently modeled by simple ellipses in Celestia, and I confess I expected them to have wandered a little more from accuracy.

Grant

Speaking of CustomOrbits... is it possible to add extra accuracy to the customorbits for a certain time period, (like adding extra points to an xyz file)? I'm thinking of the Galilean moons during galileo spacecraft flybys in particular...

JackHiggins wrote:Speaking of CustomOrbits... is it possible to add extra accuracy to the customorbits for a certain time period, (like adding extra points to an xyz file)? :? I'm thinking of the Galilean moons during galileo spacecraft flybys in particular...

No - they're calculated as a huge polynomial equation. You can improve accuracy overall by adding more terms to the equation (assuming such terms have been calculated and are available), but you can't inject extra accuracy for a particular time period.
From my experience with the Amalthea flyby, I'd say the bulk of the problem is with inaccuracies in Jupiter's position, not the Galileans themselves: a small proportional error in Jupiter's orbit will shift it hundreds of km - much more than the same error in a Jovian satellite. For instance I was able to produce a very good Amalthea flyby geometry by placing a fake Jupiter at its correct xyz position for the flyby (obtained from Horizons), and then just placing Amalthea in orbit around that.
This is no great help to you at present, of course. I did wonder about measuring the error in Celestia at several strategic points during the Galileo mission, and then biasing the whole xyz to compensate. Problem is with measuring the necessary direction of the offset - I can't think of a way to recover that directly from Celestia without actually rummaging in the code and reconstructing the orbit calculation itself.

Grant

Ah- ok so.

Celestia uses the VSOP87 format for jupiter, so I guess this one just isn't accurate enough... We need an optional plugin for DE405 format customOrbits! I wouldn't mind downloading a 10MB file if it stopped galileo crashing through Io & Callisto...!

There are other options though...

Grant,

What about the moon orbital datas for Jupiter, Saturn and the other gas giants? Do they run a similar pattern to those used for Mars?
I.E., Are they a type of default pattern or are they all based on whatever data/coordinates that are currently available?
I guess what I'm trying to ask is; Why are Mars' moons lacking the accuracy of the other SS moons?

Objects with CustomOrbits:

Mercury

Venus

Earth

• Moon
  
  Mars
  
  Jupiter
• Io
• Europa
• Ganymede
• Callisto
  
  Saturn
• Mimas
• Enceladus
• Tethys
• Dione
• Rhea
• Titan
• Hyperion
• Iapetus
  
  Uranus
• Miranda
• Ariel
• Umbriel
• Titania
• Oberon
  
  Neptune
  
  Pluto

Wild guess here, but maybe because the orbital periods of Phobos & Deimos are so small, even tiny errors in a CustomOrbit would eventually add up & become too erroneous, meaning an ssc orbit is the better option?

Grant, you'll know the proper answer?!!

JackHiggins wrote:Grant, you'll know the proper answer?!!

The CustomOrbits are hugely hard work to set up, so I guess Chris has assigned priority to major satellites first, especially those involved in observable transits and eclipses. Charon doesn't need a CustomOrbit, since its movement is so close to being Keplerian anyway (single satellite of a spherical body far from perturbing influences); but CustomOrbits for other bodies are planned, in time.

Grant

Thanks for the info grant and Jack!
I know I didn't properly phrase my question (long night shifts), but you guys seem to have understood my gist!