Re: [find_orb] Find_Orb - Before Christ MPCORB elements ?

Bill Gray Oct 17 7:32 AM

Hi Alessandro,

You can't get MPCORB elements before the year 0 or after
the year 6199.

To save valuable bytes, the epoch is stored in the form
CYYMD (century, year, month, day). The century, month,
and day are stored in "mutant hex": 0-9 have their usual
meaning, A=10, B=11, ... Z=35, a=36, ... z=61. So we're
limited to 62 centuries, 62 months/year, and 62 days/month.

We're also limited to integer days. If you ask Find_Orb
for elements for (say) 2013 October 17.2, the epoch will be
given as K13AH (2013 October 17.0), but the actual elements
will be osculating at 17.2 (i.e., will replicate the position
and velocity of the object at that time).

Add in the fact that the MPCORB format handles only
heliocentric orbits (no provision to state that "these orbital
elements are relative to Saturn") and elliptical orbits,
and that the elements lose precision badly for long-period
orbits, and that you can't say anything about the sigmas or
covariance matrix, and you can see why I see a need for a
new orbital elements format.

Anyway. None of this addresses your immediate problem...

I'd suggest using the 8-line format, which works for dates
at least as far back as -4400. At present, Find_Orb regards
dates before JD 1 = -4712 Jan 1 (Julian calendar) as errors.
I'll investigate, but I don't think it would be particularly
straightforward to get beyond that limit. (Though now that we
have DE-431 spanning the years -13000 to +17000, it _would_
be nice to be able to investigate longer-range behavior in
some cases...)

If, for some reason, you are locked into the MPCORB
format, let me know. There's a workaround involving a
slightly modified format that might help (though you'd have
to modify your software to accept that modified format.)

-- Bill

On 10/15/2013 04:19 PM, alessandro odasso wrote:
>
>
> I am using the development version of August 18th, 2013
>
> I have downloaded the DE-431 ephemeris file that seems to be properly recognized.
>
> I am trying to play in the past.
>
> _1st test_
> I press the ephemeris button, and I choose a date 99 years B.C.
> So I write 99 with a menus sign:
>
> -99-Jan-01
>
> I choose MPCORB elements and that's fine. After very little, I get the result.
>
>
> _2nd test_
> I press the ephemeris button, and I choose a date 100 years B.C.
> So I write 100 with a menus sign:
>
> -100-Jan-01
>
> I choose MPCORB elements and I get a strange result where the output seems truncated, only Mag and 0.15 Slope.
> If I switch to 8-line elements, all seems fine again.
>
> Is the limitation with MPCORB format inherent in the format itself?
> What is the limit for "packed dates" associated to the Epoch?
>
> Thanks.
>
> Best wishes,
> Alessandro Odasso
>
>
>
>
>