Satellite Identification
Wed, 19 Mar 1997 23:01:56 -0500 (EST)

There have been a number of messages about how to identify a satellite after
seeing it. I'd like to describe my process because it is part way between the
"brute force method (manually check every satellite on file)" and the
"sophisticated satellite track analysis (i.e., reduction of observation)".

I know enough about satellite calculations to do a rough positional
calculation but not enough to write a full predictive program of my own. But,
I decided that I didn't need to write a new program since there are plenty
programs available that make predictions. (I'm using Ted Molczan's
"PREDICT".) So, I wrote a program that would take my observation consisting
of the following:
   - date and time
   - latitude and longitude
   - approximate height of satellite (such as 200 to 500 miles, 700 to 1200
   - approximate inclination of satellite (such as 70 to 100 degrees)
   - direction satellite was heading (ascending, desending, don't know)

Then my program reads all of the orbital elements on file, calculates which
satellites were visible at the time AND meet the criteria, and outputs the
most favorable ones to a data file. Then I use Molczan's program to make the
official calculations for those selected satellites. With some luck (or skill
at making the observations), I can find a satellite that matches. 

I'd guestimate an 80% success rate. I don't know if the failures are due to
poor observations or outdated or unavailable elements. Naturally, this method
requires some experience at judging the orbital height and inclination, but
it may be more lenient than other procedures that require more precise data
points along the path of the satellite.

John Holtz
Pittsburgh PA
40 deg N, 80 deg W