Hi Peter,

I can't blame MPCORB.DAT for this error. (Well, not accurately,
anyway.)

The problem is that the elements are for the epoch 2010 July 23,
and perturbations between then and now are ignored by Guide. Usually,
this isn't a very big deal (except for people doing asteroid occultation
work, who need things down to the last milliarcsecond of precision).
99% of Guide users will never notice. But 2003 UV11 is a very near-earth
object. Any perturbations are "magnified" by the object being so close
to us. A thousand-kilometer error would be negligible for a main-belt
asteroid two AU from us; it suddenly becomes much more noticeable when
the object is only .01 AU away.

MPC keeps all elements on a "standard" epoch. In this case,
2010 July 23 = JD 2455400.5. The epoch is always an even multiple of
200 days, plus .5 days. They usually keep it within 100 days of the
present... today is 2010 Oct 31 = JD 2455500.5, so you are seeing
the maximum possible error!

JPL chose the epoch JD 2455502.5 = 2010-Nov-02. You wouldn't see
much error at all with those elements.

The ideal solution to this problem would be for Guide to do the
numerical integration so that perturbations would be included. There
are some problems with this, mostly involving speed. Another excellent
solution would be for MPCORB to have an epoch that changes more often
than once every 200 days. That would both decrease the sort of error
you're seeing now, _and_ mean than if Guide did the numerical integration,
it would usually be for a few days instead of up to 100 days.

One _can_ create an MPCORB.DAT with a more current epoch. (Or with
any desired epoch, useful in case you want to run things backward or
forward a hundred years or so.) The tool required is described at

http://home.gwi.net/~pluto/integrat.htm

It's a bit of a hassle to do this, and I don't expect most people
to bother. That's one reason having Guide do the integration itself
still matters.

-- Bill