# Re: Raduga 27 is accelerating again

From: Bjoern Gimle (b_gimle@algonet.se)
Date: Mon Feb 21 2000 - 23:01:21 PST

• Next message: Dr.Brian Hunter: "Re: Cosmos 2369R"

```>The flashes are only visible during a period of approximately 80
>minutes, with the period of visibility starting approximately 12 minutes
>earlier each day;

>The flashes are intially visible at approx mag+10, brighten to mag+8 in
>the middle of the zone of flash visibility before fading away again.
>
>Question: Can these data be used to determine the rotation axis of
>satellite & is there anyone on the list who knows how to make the
>calculations?
>
...
>Additionally, the flashes observed in December & January might
>consititute a third series of flashes, given the period of visibility of
>the current "evening" flash series.
>
I'm a beginner in this field, but I think that I know it, and that I have
computed the rotation of ETS-6.
1. We need to compute the directions to the satellite (and the Sun) at the
times of best flashes, to establish the normal of reflecting surface(s).
This was easy in the case of Superbird-A, with a sharp transition between
two one-power events, and nearly simultaneous observations across the USA
(in different directions), and for ETS-6, with a 60-100 second period of
one-power flashes.
The duration of Raduga 27 bright events causes several degrees uncertainty,
and the faintness (compared to a direct reflection from a large solar panel)
means that the actual orientation of the panel is several degrees away from
computed position, in ANY direction.

2. We need observations from different locations, or a long period while the
Sun position changes, or from different times of night, to create a
substantial arc of surface normal positions. In the second case we also have
to assume/hope/determine that the arc was not caused by precession of the
axis.

3. We need sufficient accuracy and number of observations to determine that
each observation is caused by the same (or determine by which) surface. For
ETS-6 I found three arcs, with a standard deviation of about 1.5 degrees,
and a separation of 13 and 9 degrees.

4. We need sufficient accuracy to determine the curvature of the arcs, to
compute the axis location.
In the ETS-6 case, they seemed quite straight, but diverging (!), but
least-squares approximation found the axis to be 33,46,55 degrees from the
three bands, with a large uncertainty.

-- bjorn.gimle@tietotech.se (office)                         --
-- b_gimle@algonet.se (home)  http://www.algonet.se/~b_gimle --
-- COSPAR 5919, MALMA,    59.2615 N, 18.6206 E, 33 m         --
-- COSPAR 5918, HAMMARBY, 59.2985 N, 18.1045 E, 44 m         --