Is is possible to replicate elliptical orbits around other planets from JPL Horizons Orbital Elements?

Asked by Jon Seamans on 2018-12-20

Version: 0.15.1
I'm trying to create NASA spacecraft orbits around other planets using JPL HORIZONS orbital elements data. I'd be happy with a reasonable accuracy over a limited, narrow window of time. Stellarium appears to use a subset of HORIZONS 12 elements, and the terminology is different. Ultimately, I'd like to recreate a single Juno perijove over a period of 30 minutes.
I think this is an aggressive task for Stellarium. Is this even possible?

An example I'm trying to replicate is Neptune's moon, Laomedeia, which Stellarium uses the "ell_orbit" coordinate function. I look up HORIZONS' orbital elements at a certain date/time and attempt to pull the pertinent ones to substitute for Stellarium's equivalent orbital elements. Long story short, I can get within millions of km which, over months, oscillate millions of km. I see HORIZONS predicts different element data sets depending on coordinate system reference plane. The choices are: 1) Ecliptic and mean equinox of reference epoch, 2) Body mean equator (e.g. Neptune) and node of date, and 3) Earth mean equator and equinox of reference epoch. I don't know what Stellarium uses.

The answer might be it can't be done, but any information would be helpful.

Thanks!

Question information

Language:
English Edit question
Status:
Solved
For:
Stellarium Edit question
Assignee:
No assignee Edit question
Solved by:
Alexander Wolf
Solved:
2018-12-27
Last query:
2018-12-27
Last reply:
2018-12-20
Best Alexander Wolf (alexwolf) said : #1

Please see our old wiki: http://stellarium.sourceforge.net/wiki/index.php/JPL_HORIZONS

I think this will be helpful for you.

gzotti (georg-zotti) said : #2

Thank you Alex, I did not even know this page!

Jon, take this decade-old page as first guide. Use ell_orbit function for moons, but note that distances are given in km, not AU. Use the current User Guide for more details.

Jon Seamans (silvy5) said : #3

Yes, that link provided the necessary conversions to equate HORIZONS orbital parameters to Stellarium's parameters. Given the dynamics of the perijove orbit, I was pleasantly surprised the Stellarium could simulate pretty well 30 minutes of Juno's closest approach to Jupiter (perijove 16, 10/29/18 21:06 UT to 21:37 UT). However, Stellarium's time was 8h 13m early.

Thank you Alexander and Georg for your help. You might be interested in the results comparison below.

At four times over the 30-minute interval, I compared many observables predicted for Juno:
1) RA / Dec of Jupiter,
2) RA / Dec of Io,
3) RA / Dec of Ganymede,
4) Angular size of Jupiter,
5) Distance to center of Jupiter, and
5) The angular separation of Io from Jupiter's limb. (From this I arrived at HORIZONS error)

To summarize:
- Stellarium's position (RA/Dec) accuracies were <2°.
- Stellarium's angular sizes were ≤ 2° (Note: At perijove, HORIZONS predicted 144° and Stellarium predicted 142.9°)
- Over 30 minutes, the HORZIONS distance to Jupiter's center ranged from about 74,541 km to 104,805km. Stellarium's predicted distance was off by about 100km at perijove to about 3000km at the furthest distance.
- For Io's separation from the limb, HORIZONS predicted an altitude range of -26° to + 21°. Stellarium's predicted range was -25° to +19. The maximum error was <2° while <1° was more typical.

Io's angular separation was of particular interest because Juno took an image (https://www.missionjuno.swri.edu/news/io_rising) of Io and the limb at 21:26 UT. From this picture, the specific accuracy of HORIZONS prediction was determined. The picture reasonably images Io and given the easily predicted angular size of Io, the actual limb separation was determined to be 5.7°. At 21:26 UT, HORIZONS prediction was 3.3°, or about a 2.5° error.

Except for the 8+ hour timing error, Stellarium's prediction error appears to be in the same ballpark as HORIZONS. Though I have only the one self-check prediction against the Io/Jupiter image. I don't know the actual RA/Dec coordinates of any object viewed from Juno.

gzotti (georg-zotti) said : #4

Interesting application, thanks for sharing!