Re: Solar panels (1080) and brightness variation with phase (1081)

Mike McCants (mike@comshare.com)
Mon, 13 Nov 1995 12:13:39 -0600

Walter Nissen's sentences appear after >'s.

>I point to the fact that observers (or at least, this 
>observer) have very seldom observed very bright glints from active 
>satellites.

I agree with you.

I estimated the brightness of RadarSat last night at magnitude 6.5.
This pass was about 40 degrees up in the west.  I watched it for a
minute and I didn't see any variation.  I suspect the solar panels
are shadowing the payload to some extent.  The 95 59B rocket was
only 3 minutes ahead of the payload.  I estimated it was 1 magnitude
brighter and it gets 1/2 magnitude for being farther away, so its
intrinsic magnitude was 1.5 magnitudes brighter.  This seems to me to
be larger than what would be expected from the physical sizes.

>As an example, HST frequently glints at 
>observers for a few seconds, once per pass, by a few magnitudes, 
>presumably off the very long tube.  I would freely admit that these could 
>be solar-panel glints, but ...

I wouldn't.  :-)  I think that if I saw a solar panel reflection from the
HST, I would be temporarily blinded.  :-)  But it goes nearly directly
overhead for me.  I would guess that a good solar panel reflection would
give it an intrinsic magnitude of about -5.  I have never seen such a
brightness from it.  I have seen some fairly bright flashes from the
tumbling HST Solar Array, 90 37C.  But it is my understanding that it
is not very flat at present.

We could ask John Broman if he can find out how close the HST solar
panels are normally kept to pointing directly at the sun.  I'll bet
"within 5 degrees" is a good answer.  :-)
 
>In looking at the C* 1933 family of payloads, ...
>I have advanced the hypothesis that these result from 
>solar-panel glints on a rapidly rotating satellite which has very recently 
>lost attitude control.

I agree with you.

>SeaSat, long inactive, is known for very bright glints, presumably off its 
>long boom,

I diaagree.  I think SeaSat is giving off solar panel glints.  I have seen
it 5 or 6 magnitudes brighter than predicted with quite unfavorable
phase angles.  But for this to be true, the solar panels would have
to be more or less flat relative to the Earth's surface.
 
>There are a few active, or presumably active, satellites which do flash 
>rapidly, such as COBE, USA 32, USA 81, and some others.  I don't pretend 
>to any theory about them, but I don't think I'm seeing solar-panel glints. 

I have never seen "short" glints from CoBE.  I do think USA 32 and USA 81
are very similar to EGP in that they are giving off glints from relatively
large flat spots.  I suspect they are covered with solar panels.
I think their "attitude control" is spin stabilized since the brightest
glints always occur with a particular relationship to the sun.  I often
see naked eye flashes from them for a little while.  And that "little
while" always has a particular relation to the sun.

>I used to think of 
>UME 1 as a fainter version of EGP, but haven't seen wild flashing behavior 
>from it in a long time.  Is that consistent with others obervations of 
>UME 1? 

I haven't observed UME 1 or 2 in a long time.  There must be a picture
of one of them in the Caprara book.  Unlike EGP, the observers usually
come up with a reliable period of about 3 seconds.  So they aren't
"varying wildly".  :-)
 
Note that Ted also gives elements for UME 2, but he (in agreement with
NORAD) gives the name as UME 2 r:

UME 2 r          0.8  0.9  0.0  8.5
1 10675U 78018  B 95308.83997860 -.00000043  00000-0  11093-4 0  4449
2 10675  69.3534 100.5414 0161023  12.9932 347.5216 13.44004325869192

(I bet Ted changes it by next week.  :-)

>Mir ...
>I would readily believe that some of its brightest passes achieve that 
>status because of solar-panel glints. 

I rarely observe Mir, so I rarely see bright glints.  But the brighest
I remember was very near full phase.

I guess another way to look at this is to ask the question:

If, on the average, an object is this big and this bright, then what
kind of increase in brightness would we expect from a flat spot with
10% reflectivity of size X square meters?  If the answer for different
X's turns out to be 2 magnitudes or 5 magnitudes, then that gives us some
indication of the required size when we actually observe such a brightness
increase.  If Mir goes up to magnitude -5 at a range of 1000Km, it must
be a rather large flat spot.

I compute that a 1 square meter flat spot reflecting with 10% efficiency
would give a magnitude of about -5 at 1000Km.

I guess that large solar panels are usually not "flat" to high accuracy,
so a large solar panel would be likely to give a -5 brightness over
a significant part of the Earth, but unlikely to give a brightness
much greater than that.

--- From Walter's post 1081:

>Real satellites, manned and otherwise, seem 
>to be boxier, shinier, and more shadowy.  Thus, the variance of brightness 
>with phase tends to be more pronounced than for the theoretical model. 
>Thus when the phase angle is unfavorable and the object is near the Sun, 
>the computed magnitude may be brighter than the actual observed magnitude. 

What I interpret this to mean is that "on the average", a non-spherical
object will be fainter than the spherical assumption when the phase angle
is unfavorable.  I agree.

>However, if you 
>get a specular glint off a long, shiny tube or a nice, near-specular glint 
>off a white surface, the opposite can occur. 

Yes, I agree, a non-spherical object may be brighter than the spherical
assumption when the attitude of the object is "favorable".
 
>Can anyone confirm these various speculations? 

Maybe the Air Force with a 48 inch, 3-axis, high-resolution tracking
system.  But they don't usually tell us.  :-)

Or perhaps you would like to hold a "vote".  I vote "I agree".  :-)
 
Cheers, 
       Mike McCants        mike@comshare.com