Hi Dave,

A late reply to:

"...Does anyone know how far off Guide8 will be using October 15, 2008
MPC data with 'Use MPCORB' activated? Before each imaging run I grab the
latest files at, ftp://cfa-ftp.harvard.edu/pub/MPCORB/ , but due to
server problems at Harvard those files have not been updated in a week."

Puzzling out asteroid ephemeris uncertainties is a really nasty problem,
and there's no quick answer to your question such as "positions will be
off by three arcminutes" or anything like that. The error will vary from
object to object, and there will be two sources of that error.

First, because you're using week-old elements, observations made
during that week won't be included in the computation of those elements.
That may not matter much, if (say) the object has been well-observed
for decades, and the new observations just extend the arc from 40 years
to 40.02 years. If you're observing objects of this ilk, you need not
update 'mpcorb' very often.

If the object was observed over the course of a few nights last month
or last year, though, then the new observations suddenly become really
important: they will greatly refine the accuracy of the orbit.

If a lot of time goes by and you've not updated 'mpcorb.dat', that
sort of problem will become more common. And obviously, you'll be missing
any objects found since your last 'mpcorb.dat' update.

So how much should you worry about this? The best clue is to look at
'more info' for the object in question. If it has years of data leading
up close to the present, you're probably perfectly okay. If there is
little data for it, and you think it's bright enough that somebody might
have gotten more data since you downloaded 'mpcorb.dat', then maybe the
current 'mpcorb.dat' would give you a better orbit.

The second source of error isn't going to affect you at all in the
current case: 'mpcorb.dat' elements are given with a particular epoch,
one that changes every 100 days. The current elements have an epoch
of Julian Day 2454800.5, or 2008 Nov 30. Perturbations between that
date and the current date aren't included. So if your orbital elements
for an object have epoch 2008 Nov 30, and you're looking at the object
on 2008 Oct 30, and there was a close planetary encounter in the 31
days in between, the position could be badly off.

The key phrase is "if there was a close planetary encounter". Such
encounters are rare. However, suppose you got your latest 'mpcorb.dat'
two years ago. In that case, instead of just ignoring 31 days of
perturbations, you're ignoring about 730. Over that time, perturbations
will pile up and you'll see more significant errors. (Though even then,
visual observers of most bright asteroids will just start to notice the
difference.)

-- Bill