Re: Starlink brightness observations of Russell Eberst posted by Ted Molczan

From: Thierry Legault via Seesat-l <>
Date: Fri, 28 Feb 2020 11:29:43 +0100

I have taken new videos Wednesday evening from 
Versailles, France, between 19:00 and 19:45 UTC.
The satellites are all from Nov 11th 2019 launch: 
#1056, #1054, #1062, #1053, #1017, #1067, #1058, #1943.
Altitude range at culmination is 630 km to 555 km 
(altitude above horizon 62 to 88).

Careful examination of videos and comparison with 
many stars in the field show very consistent 
magnitude for all units: 4.2 plus or minus 0.2 at 
culmination. Brightness is fairly stable during a passage, without any flare.

This is consistent with the video I have taken on 
Feb 22nd of several units (launched Jan 7th) 
where I identified two populations, mag 2.5 for 
the brightest (including Darksat) and 4 to 4.5 for the dimmest:

So far, I got much brighter magnitudes than the 
ones measured by Russel, but I don't understand why.

Note: the videos are taken with Sony A7S with 
Atomos Shogun in 4K low compression ProResHQ 800mbit/s.


At 15:33 26/02/2020, Anthony Mallama via Seesat-l wrote:
>Thanks for sharing Russell Eberst's magnitude data. I found that the
>average 1,000 km magnitude for the satellites at the final 550 km altitude
>is 6.8. When that is adjusted to 550 km it becomes magnitude 5.6.
>Your plot shows a moderately weak correlation between brightness and phase
>angle (again for the satellites at the final 550 km altitude). So, these
>data do not indicate a very pronounced backward or forward scattering of
>sunlight - more like a diffuse scattering. Additional data at high and low
>phase angles would provide a more definitive result though.
>Since Starlink satellites themselves are shaped like a table top, I would
>expect to see a fairly strong dependency of distance-corrected brightness
>on elevation above the horizon. The satellites are seen face-on (presumably
>brighter) when overhead but edge-on (fainter) when near the horizon, if I
>understand their orientation correctly.
>There could also be a relationship between brightness and the azimuth
>difference between satellite and sun - due to scattering. (Similar to phase
>angle though not quite the same.) Altogether, the brightness model might be
>fairly complex.
>Question: If the satellite elevation and azimuth turn out to be important,
>can those quantities be determined for past observations? I ask because
>positions from TLEs become inaccurate pretty quickly. So, I'm wondering
>whether az and el should be recorded at the time of observation.
>In my opinion, the brightness of Starlink satellites is an important issue
>(ask any astronomer) and a major way in which we observers can contribute
>to our field of study. Anyone who wants to make visual brightness estimates
>can find instructions for stars at The same method can be
>applied to satellites.
>Lastly, I can send you a spreadsheet with my first 20 magnitude
>observations if you want to include them in your analysis. They're not in
>the format you're using though as they include az and el.
>Best regards,
>Tony Mallama
>Seesat-l mailing list

Thierry Legault

Seesat-l mailing list
Received on Fri Feb 28 2020 - 04:31:31 UTC

This archive was generated by hypermail 2.3.0 : Fri Feb 28 2020 - 10:31:31 UTC