Prediction methods

From: via Seesat-l <>
Date: Fri, 29 Aug 2014 13:08:11 +0100
Hello folks - I would like to put forward my ideas about the prediction of
US secret satellites when they approach decay.
Last night I observed 14041B using elements of age 12.6.  I might have got
newer elements but that cannot be done with US secrets.  I used what I call
"model-B" predictions.  When I saw the satellite I was undecided as to whether
it was early or late.  On reducing the positions, OBSREDUCE told me the
satellite was 297.3 seconds early.  My predictions had anticipated this.
The key to "model-B" predictions lies in the way elements are updated from
the issue date to the time when you observe the satellite.  I define an
element which I call L = remaining life in orbit, measured in days.  I then
use calculus starting from the differential equation dL/dt = -1 to calculate
the mean anomaly of the orbit at prediction-time.
It is not always as successful as for me last night - the gain depends on the
quality of the 2-line element set used and the way in which L is expressed
in terms of the other orbital elements.

        On another point, I would like to extend the range of TLEs in my
TLEDUMP in order to help identifystrays.  Does someone have a master file
they would be willing to share with me ?

Regards from David Hopkins COSPAR 2414, Bournemouth, England.

Seesat-l mailing list
Received on Fri Aug 29 2014 - 07:08:47 UTC

This archive was generated by hypermail 2.2.0 : Fri Aug 29 2014 - 12:08:47 UTC