Re: KH alert and "Do we need a TLE format change"

From: Jonathan McDowell (
Date: Tue Feb 01 2005 - 18:06:43 EST

  • Next message: Allen Thomson: "Re: oig data"

    > Note that I used a fake USSTRATCOM no. to avoid overwriting
    > current elsets. I strongly feel that we should use another
    > code than U for amateur TLEs, after the catalog #.
    > Additionally, it would be very useful, in archived TLEs
    > especially, to use a third code for guesstimates and
    > search orbits.
    I entirely agree. Note that there are two concepts here:
    - source of elements:  STRATCOM, "amateurs", ... (perhaps Russian Forces someday?)
    - quality of elements: observed, predicted, inferred from published data, WAG, ...
    I think these concepts should be kept separate (i.e. don't use the
    'U' field for both of them).
    If such codes were adopted I would rework 
    my public element archive to use them. In the recent update
    to my site, there are a significant number of TLEs that
    are guesstimates I've made; I use a third line for comments
    to indicate that, and sometimes I've used a comment line
    with such a code 'G' for guess, etc, but in general: caveat
    It would also be nice to mark old NORAD elsets that are
    clearly bogus due to one error or another.
    In my own code I have idenfified the following cases:
    S  Shape (mean motion, ecc and inc correct, but orbit angles unknown)
    G  Guess  (attempted reconstruction of probable orbit)
    E  Error (Spurious data)
    EM Error - Mis-ID (good data, wrong object)
    EW Error - Wrong (just downright bad)
    EI Error - Inaccurate 
    ED Error - Doubtful  (might be right, but looks suspicious)
    EP Predicted orbit only
    EC Error - Corrupted data (obvious typo or computer problem)
    In addition we should add codes for:
       Independent (or Amateur) - data from non-government observers     
       Spacetrack  - data may be subject to restrictions
    Alternatively, if you want to get really serious,
    we could allocate numeric codes for sources of elsets,
    a bit like the minor planet folks allocate codes for observatories:
       001 NORAD via Old OIG
       002 SpaceTrack
       003 Ted Molczan
       004 Bjorn
       005 ... etc, etc...
    since this allows credit (and blame) to be tracked.
    But we might have to go to a 3-line format at that point.
     - Jonathan McDowell
    Subscribe/Unsubscribe info, Frequently Asked Questions, SeeSat-L archive:

    This archive was generated by hypermail 2b29 : Tue Feb 01 2005 - 18:11:21 EST