Horizons Ephemeris System - precision?

[hjw]

Hello,

can anybody tell me, what's wrong with Horizons data (or my view of the universe)?

I just queried again the coords for Messenger, starting at 2004-08-03 07:15.
- they put the spacecraft on the wrong side of earth
- the distance is much too big (about1500km, I expected nearer to surface)
You can also see this with Grant's new xyz-Data.

My first guess was, that the position of Earth in celestia is wrong, so I queried
the data of Earth (geocenter, object=399) and made a quick check... but Earth
in celestia is at the correct position (give or take a view km's).

To be sure I wrote a perl script to compare Horizon's Earth data with
Horizon's Messenger data (step-size 5m, calculating the distances). After
subtracting the radius of Earth I got nearly the same values as celestia
shows (at 07:15 celestia shows 1461km, my calculation 1438km).

So... celestia is OK.

Are Horizons data not exact so short after launch???
I guesstimate that the initial position of Messenger as given by Horizon
is about 15000km off...

Do I expect too much precision or am I missing something? Any opinions?

MfG - HJW

[granthutchison]

I think the Horizons team just aren't that interested in knowing where Messenger was at launch ... the point of the effort is to show where it's going to be. I'd guess they've made no particular effort to generate the details of acceleration during launch and Earth escape (and you can't blame them - it would be a lot of effort for no benefit). So what you're seeing is (I think) a spurious backward extension of the departure orbit once it became ballistic. I did consider just suppressing that segment of the data in my Messenger xyz.

Grant

[hjw]

Obviously you're right, Grant.
I just hoped that the calculations that are needed for the
launch window and are used to check if the craft is on track
during the initial segment of the flight (calculated BEFORE launch
with HIGH precision) are available via Horizons.

But... wishfull thinking...

Thanks
MfG - HJW