Determining orbits from a single location?

Bruno Tilgner (Bruno_Tilgner@compuserve.com)
Sat, 29 Aug 1998 18:03:07 -0400

Marco Han wrote:

>I am playing with the idea of building some soft-/hardware combination t=
hat
>would allow for automatic orbit determination using multiple observation=
s
>with a small telescope and a CCD camera.
>
>Is that feasible?

This is a very challenging project. Some of it is feasible, some is not.
I believe the "automatic" orbit determination is not feasible, unless
you set up your mini-USSPACECOM, preferably with a radar installation
in your backyard, for which you would need a PTT licence.

=46rom what you write you would not attempt an initial orbit determinatio=
n
but rather the refinement of orbits already known. The mathematics of
of this "differential orbit improvement" technique are documented in
several books, Vallado (mentioned in another reply to your mail) being
one of them. I could recommend a few more.

Leaving theory aside for the moment, I see considerable practical
difficulties of measuring the position of a satellite at a known time.
Off-the-shelf CCD video cameras are way too insensitve. You would
only record the very brigthest satellites, say magnitude 3 or brighter.

Position determination by optical methods is only feasible against
a field of known background stars, unless you invest into a telescope
mount which gives you arcsecond pointing accuracy. It can be done, but
be prepared to spend a few million dollars or euros.

On the other hand, if you know an orbit and want to measure at what
time a satellite is at a given location (in terms of right ascension
and declination) this is a lot easier. The main difficulty is precise
synchronisation of your clock. In Europe this is best done with the
time signals of DCF77. There are relatively inexpensive commercial
products on the market which synchronise a PC clock via the serial
interface. This would give you 1/18 sec accuracy, the theoretical
limit for PC's.

As a practical exercise to get a feeling for what you are up to,
I would recommend you start with a sensitivity analysis. Take the
TLEs of any LEO object, Mir being the most popular example, vary
the orbital elements, in particular the number of revolutions per
day, and see which effect this has on the apparent trajectory on
the sky. This would give you a rough idea of how errors in position
determination and time propagate into the orbital elements.

If you are then still determined to pursue your project I would
be pleased to discuss further details via private e-mail.


Bruno Tilgner
48.85N 2.2E UTC+2
Bruno_Tilgner@compuserve.com
 =






 =