Re: Visible Satellite Observations...

Torben Noerup Nielsen (torben@net.Hawaii.Edu)
Mon, 29 Nov 1999 21:12:25 -1000

>While eclipsed, their radiation is mainly in the infrared. So you would
>need an infrared telescope on a high mountaintop.

I can deal with the high mountaintop. A real infrared telescope is a good 
deal harder due to the cost of the infrared arrays.

>If you want to do daytime observations, you need a radar installation.
>For nighttime observations photographic recordings (perhaps a modern
>version of the Baker-Nunn camera) will give you accurate positions if
>you can record the surrounding starfield, time marks and if you have
>the necessary astrometric correction software (the easiest part).

Well, that's what I was wondering about. According to the pictures at, it is indeed possible to get visual images during 
the daytime. I just haven't gotten enough information to know if I can get 
detection with a single image which is what I need since I want to record 
the position accurately.

>Since an initial point won't help you much (you need at least a point in
>3 dimensions plus a velocity vector, also in 3 D), you might find it
>useful to do simultaneous observations from a different location. This
>will allow position determination by triangulation. The time of the
>observations must be known with an accuracy of 0.01 sec or preferably
>better if you want to achieve an accuracy of +/- 50 m (100 m bounding box).

Getting the time right is not a big problem. A fairly straightforward GPS 
system will give me 1 ms and I usually keep my computer synchronized to GPS 
time (I have a couple of receivers around). And no, a single observation 
will of course not do; there are too many degrees of freedom to deal with. 
But multiple position observations *will* do even without triangulation. It 
is fairly straightforward to derive a first order orbit from a couple of 
positions. And if I have a lot of position measurements, I can treat it as 
a numerical optimization problem. I do not need to know the orbit in true 
real time; within an hour or so is good enough for my purposes.

>Your software should not only model the Earth's gravitational field,
>it must also provide a much better model than NORAD's for the density
>of the atmosphere. It must cater for daily and seasonal variations as
>well as for changes in solar activity.

That I'm aware of and we can deal with that. We use a model that 
automatically updates for changes in solar activity and the atmospheric 
effects we can deal with pretty well too.

>All in all, you are in for a mini-USSPACECOM tracking station (or two)
>which I suspect is in the multi-million $ range as far as investment
>is concerned. Operating costs, mainly for a technician, come on top.
>A challenging project.

Well, I'm hoping this can be done somewhat cheaper than that. But I do 
understand that it is going to be expensive. The investment is the major 
part, the operations are easier to deal with as I have the people.

Thanks, Torben

Unsubscribe from SeeSat-L by sending a message with 'unsubscribe'
in the SUBJECT to