search

OSSOS / SurveySimulator

Simulate detecting TNOs, nearly as fun as really detecting TNOs.
Other
9 stars 7 forks source link

Are the coordinates from JPL in Ecliptic or Equatorial coordinates? #18

Open ijiraq opened 1 year ago

ijiraq commented 1 year ago

https://github.com/OSSOS/SurveySimulator/blob/3a2b24a57a85056d3a3c559ef025bf7b7429f45c/fortran/F95/xvutils.f95#L223

BaryXV returns ecliptic coordinates. If JPL coordinates are an ecliptic reference frame, then the pos_b values from BaryXV doesn't need to be rotated when used with those coordinates, but if the JPL values are in equatorial, then they do need to be rotated.

JPL allows the user to SELECT ecliptic or equatorial. This could be found by scanning the file for the word Eclitic. Should this be added to the read_jpl_csv subroutine?

jmpetit commented 1 year ago

Hi JJ,

I've moved to using the DE440 ephemerides to provide the position of the Earth, Sun and everything else. I also now have included the real position of the observatories, which are defined in Equatorial coordinates w.r.t. geocenter. And pointings are also equatorial. I know the orbits of the objects are given in ecliptic. So there will always be the need to move between ecliptic and equatorial. In the new version I've commited and pushed in new branch JPLephem, I've discarded BaryXV to use DE440 which gives coordinates in equatorial. So the question is solved in this new branch. I've run the new version and it succeeds in all 3 tests for F95. I haven't updated F77 yet, and am planning to abandon it straight. I just have to check that I can easily use the F95 module with my Python codes. I mean, I use some functions that are defined in the surveysubs library and need to check that the API is still valid from F77 to F95.