; Any line starting with ';' is treated as a comment. This file
; is read by 'eph2tle' (see 'eph2tle.cpp') as boilerplate to be put
; at the start of all TLE files.
# The following TLEs (Two-Line Elements) were generated by fitting
# a numerically-integrated ephemeris, based on the orbit given below,
# to the SDP4/SGP4 models. Each TLE is good for the interval between
# its epoch and the epoch of the subsequent TLE (usually one day).
# The maximum deviation between the numerically integrated ephemerides
# and the TLE-based ephemeris, in kilometers, is given for each TLE
# as the "worst residual".
# This fitting is done using the 'eph2tle' code packaged with Find_Orb.
# The code has details on how the TLEs were computed :
# Be warned: the error given is that between the numerically
# integrated ephems and the TLE model. There will usually be at
# least some additional difference between the ephems and the actual
# motion of the satellite, especially if the ephemerides are based
# on a short arc. If the following orbital elements say something
# like "[based on] 10 observations 2015 Jan. 1-13", and the TLEs
# are for 2015 May, the extrapolation may be tenuous at best. If
# your application requires a specific level of accuracy, you may
# want to contact me; I can give you some idea as to the fit of the
# underlying data (and possibly TLEs based on newer data).
# For certain objects, I use an alternative model in which most of
# the traditional TLE data is replaced by six signed 8-digit hex numbers.
# These six numbers are a state vector, numerically integrated using
# a simplified force model. This is necessary for objects at L2 or L1,
# and sometimes for objects in high orbits where the SDP4/SGP4 model
# fails. I'll be posting source code for this. But be aware: it's my
# own extension, doesn't work with other code, and is therefore used
# only when necessary.
# Note that for some cases, the fit to SDP4/SGP4 fails, and the
# "worst residual" is thousands or even hundreds of thousands of kilometers.
# (This usually happens near perigee for highly eccentric objects.) In
# such cases, we pretty much just have to throw up our hands and say
# that we can't use TLEs. (Which is why I developed the above state
# vector version.)
# Also, if you use a TLE outside the "correct" time range (between
# its epoch and that of the next epoch), all bets are off. It usually
# seems to be a decent fit, at least for a while, but I don't
# recommend doing it.