Potential Mechanism for Starlink Flares

From: Robert Herrick via Seesat-l <seesat-l_at_satobs.org>
Date: Sun, 26 Apr 2020 19:35:07 -0700
I believe that the Starlink flares are not due to the motion of the
specular reflection of the sun across the observer as the satellite moves
along its orbit, as that would produce flares lasting 8-9 seconds rather
than those observed to last one second or less.
I propose, however, that the flares are a result of moving the satellite in
elevation, i.e., in the sun-satellite plane, to optimize solar power input
for the system including battery charging. During that time of each orbit
when the satellites are illuminated by the sun; the solar elevation at the
satellites varies between -20 and +42 degrees, a range of 62 degrees. To
maximize output from the solar panel the satellite must re-position the
plane of the solar panel, in elevation, to approximately face the sun. This
re-pointing is not done continuously but in steps along the orbit at
predetermined times / locations. This re-positioning, moves the surface
responsible for the flares sufficiently quickly to produce the observed
short duration flares from the Starlink satellites. For a 1 second duration
flare only requires a satellite rotation rate of 9 milliradians / second to
steer the 0.5 deg. width of the sun across an observer; for other positions
of the sun the flare duration would be even shorter or if he observer is
more off axis the flare would be less bright.
Relative to the observer, there is an angular component that also must be
satisfied, in the plane perpendicular to the  sun-satellite vector in order
to move the reflection out of the sun-satellite plane as most observations
have been reported. The approximate difference in azimuth between the sun
and the observer at the satellite was approximately 140 degrees.
The range of solar elevation angles listed were for the 550 km orbit plane
that included satellites 45096 & 45080 that I used. My calculations were
for 1 April 2020, the observer location was Buenos Aires Argentina (see
Ruben Iianza's earlier post) and I used TLEs with an epoch of 91.00001157
for both satellites.
As always comments welcome!
Seesat-l mailing list
Received on Sun Apr 26 2020 - 21:36:31 UTC

This archive was generated by hypermail 2.3.0 : Mon Apr 27 2020 - 02:36:31 UTC