How inaccurate are predictions made months in advance?

CmdrJaycee (
Thu, 23 Apr 1998 16:27:11 EDT

I was tempted to ask this directly of some of the very helpful people on this
list I've corresponded with in the past, but since I'm sure others might be
interested in the answer, I decided to submit this to the whole list instead.

As an amateur astronomer, I volunteer at our local park system's public
stargazing programs. Unfortunately, being a governmental body, they need to
schedule their programs months in advance.  So when I'm asked by the park
staff if I know of anything interesting going on in the sky, say, next August
or September, as I was this morning, I don't usually check any of the
satellite prediction tools I have simply because I know they may not be
accurate that far down the road.  But it finally dawned on me that I really
had little understanding of just what that error margin might be.

For example, if, hypothetically, I came up with a good pass for Mir on August
31, one that might be a nice highlight for a public program, should I go ahead
and have them schedule a program for that night (even though the precise time
might be off a bit)?  Or is the error margin so wide that Mir might not have a
pass at all that evening?  Are we talking of error margins of minutes or hours
(or even days) in most cases for low earth orbit satellites?

What about Iridium flares?  If I see a nice -7 flare predicted for mid
September by Mike McCants' Iridflar, is there any chance that prediction might
still be accurate for "sometime" that evening (if not for the exact "minute")?

Is it generally true that the higher up a satellite is, the more predictable
its pass times would be months in advance (I'm thinking here of EGP, for

I just want to make sure I haven't over-reacted to the cautions about using
*old* TLEs, that I and maybe others are missing opportunities for public
programs by being overly concerned that predictions made so far in advance are
totally unreliable.  Just how unreliable are they?  As long as something will
still be visible on a given "evening," that's all I really need to know at
this point; obvioulsy the precise timing can be refined as the date gets

As usually, any help along these lines if very much appreciated.

- Jim Cook
Germantown, MD
39.2N, 77.3W