[Spice_discussion] Re: how accurate is conics?

[Trent Waddington]

On Fri, Sep 3, 2010 at 12:18 PM, Nat Bachman
<nathaniel.bachman at jpl.nasa.gov> wrote:
>   3) Run your program with elements provided by the JPL Horizons
>      program instead of those from OSCELT. These elements
>      likely have been selected to fit all of the observational data.

Hmm, that's interesting.  To use the elements provided in the object
description I need to know how to convert from the heliocentric
ecliptic reference plane to the inertial reference frame.  Horizons
kinda has support for this in the [E]phemeris option, but I don't
think that's what you're saying.

Another mystery, when I ask Horizons for an ephemeris of state vectors
I get different values to when I ask it for an SPK and then use
spkez_c to retrieve the state vectors at the same epoch.  For example,
2006 RH120 at 2010-Jan-01 gives me:

from spk 72324528.908090 121285770.538049 52886682.629751 -26.346082
13.091303 6.010543
horizons 72326272.589071 121284904.093955 52886284.823933 -26.345897
13.091614 6.010679

Which seems significant.  I don't see how this could be a result of
interpolation.. I generated the SPK file starting at the same epoch.

Trent

full log of the Horizons session: http://quantumg.net/horizons_log1.txt
full source code used: http://quantumg.net/pconics.c
spk file: http://quantumg.net/tmp.bsp