Re: Question on Elset Corrections
=?ISO8859-1?Q?Bj=F6rn_Gimle?= (b_gimle@algonet.se)
Mon, 15 Mar 1999 10:01:31 +0100 (CET)
Addition to my previous reply:
Ron Lee wrote:
...
>iterative process..changing each one until the residual is the
>lowest?
>
ELCOR builds a system of equations and solves for all selected
elements at once. It treats all errors as positional errors, so
if the time error is a minute or more, the satellite would be
in a different part of the sky, even below horizon, where the
sensitivity to the deltas is much smaller. This can cause ELCOR
to diverge.
FITELEM helps you by splitting the error into cross-track and
along-track, but you must iterate each element manually, which
I find tedious.
I haven't tried George Lewis' IOD yet.
To try to avoid the ELCOR trap, you should correct MM/B*
manually, before letting ELCOR repeat that, and include other
elements only when a good solution has been found for these.
If you have fudged the elements, or there was a maneuvre in
between, you may also have to let ELCOR compute the quasi-
anomaly, corresponding to a change in epoch/node time.
-- bjorn@tt-tech.se (office) b_gimle@algonet.se (home) --
-- 59.2237N, 18.2286E, 44 m http://www.algonet.se/~b_gimle --
-- 59.298 N, 18.104 E, 55 m from 1999-01-25 --
-- SeeSat-L / Visual Satellite Observer Home Page found at --
-- http://www.satellite.eu.org/satintro.html --