RE: Modeling Starlink magnitudes in a simulation

From: Ted Molczan via Seesat-l <>
Date: Tue, 2 Jun 2020 12:12:06 -0400
Max Hartshorn Wrote:

> What I'd like to do is create a *responsible* and *non alarmist*
> visualization of what the night sky *could* look like when 12,000 Starlink
> satellites are up and running. As far as I know such a simulation does not
> exist.

I vaguely recall seeing or hearing about something similar, but I could be wrong.

> I've been looking through the archives here and I've seen a range of
> apparent magnitude observations, anywhere from 2 to 9. Additionally
> Starlink is iterating the design of their satellites to minimize
> reflectivity.
> The threshold that's most interesting to me is the 6.5 threshold for naked
> eye observation. If the satellites (particularly those in the lower orbital
> shells) are below that threshold, they could potentially impact how the
> general public views the night sky.
> Is there any agreed upon estimate for the apparent magnitude of the
> Starlink satellites currently in orbit? Do we just not know yet? Are there
> any responsible methods for modeling / estimating the magnitude of the
> remaining satellites?

SeeSat-L subscribers have reported hundreds of Starlink magnitude observations. In April, I posted a preliminary
analysis of a subset of this data:

The graph on page 2 of the pdf in the above post gives a good idea of the magnitude of regular Starlinks at the
operational altitude. The magnitudes have been normalized to 1000 km range based on the inverse-square law, which is the
convention for analysis of LEO satellites. 

Eyeballing the graph, it is evident that the mean 1000 km magnitude was around 6, but with some variation, which is
typical of satellites. Most of the observations were within +/- 1 magnitude of the mean. This means that actual
brightness can be expected to be up to 1 magnitude brighter or fainter than predicted. For a realistic simulation, it
would desirable to show this variation, if practical. 

I have been working on statistical analyses that might be helpful in that regard once completed. I can imagine using a
random number generator to approximate the variation in magnitude, based on the mean and variance. The devil will be in
the details. Does the +/- x mag variation tend to occur primarily within a pass, or between different passes? Do
satellites in a given plane tend experience the same variation at the same time?

Ted Molczan

Seesat-l mailing list
Received on Tue Jun 02 2020 - 11:13:23 UTC

This archive was generated by hypermail 2.3.0 : Tue Jun 02 2020 - 16:13:23 UTC