So that they are not as visible? I assume the SpaceX engineers have considered this option.
1$\begingroup$ somewhat different (more specific and narrowly-defined) but related: Is it feasible to paint SpaceX Starlink satellites black so as not to frequently saturate the CCDs of the Large Synoptic Survey Telescope? $\endgroup$– uhohNov 12, 2019 at 16:08
1$\begingroup$ This would cause problems with the thermal conditions of the satellite, so I guess blackening a satellite would require some (much?) work on dealing with keeping temperatures in reasonable bounds. $\endgroup$– Dohn JoeNov 13, 2019 at 9:00
$\begingroup$ I think this should be a broader question, e.g., "What parts of any satellite could be made nonreflective" . The ISS is easy to see; Many geostationary television satellites are quite bright as well. $\endgroup$– Carl WitthoftNov 13, 2019 at 14:08
$\begingroup$ @CarlWitthoft something similar to that had already been asked: "Is it feasible to paint SpaceX Starlink satellites black so as not to frequently saturate the CCDs of the Large Synoptic Survey Telescope? Or would that be to hot without adding a separate thermal radiator of some kind?" $\endgroup$– uhohDec 8, 2019 at 1:31
They have said that they will mitigate the reflectivity of the satellites in future batches and that the 1.0 launch, Nov 11 they did not complete that for this batch.
In the December 2019 launch they plan to test darkening one of the satellites to see if it meets their needs.
(chief operating officer Gwynne) Shotwell said the next batch has one satellite “where we put a coating on the bottom.” She noted that this is just an experiment and could not predict if it will work. “We’re do trial and error to figure out the best way to get this done,” said Shotwell.
[...] The coating that is being applied to one of the satellites in the third batch of Starlinks is just the first step toward finding a permanent solution as more satellites get deployed. Shotwell said the company plans to launch batches of 60 satellites every two to three weeks over the next year to build the constellation that by mid 2020 will be ready to provide global coverage.
[...] The experimental coating that would make the satellite less reflective could affect its performance, so that is something that will be examined, said Shotwell. “It definitely changes the performance of the satellite, thermally. It’ll be some trial and error but we’ll fix it.”
As if addressing the "I assume the SpaceX engineers have considered this option." part of the question (and not everyone will believe this):
Shotwell admitted that nobody in the company anticipated the problem when the satellites were first designed.
“No one thought of this,” she said. “We didn’t think of it. The astronomy community didn’t think of it.”
Additionally, Jonathon of JSR (Jonathon's Space Report) has a great site for tracking Starlink with images of the Starlinks satellites attitude while changing orbits, vs in use.
He has a Shark Fin image, during operational use.
And ascent configuration.
It seems that the Open Book mode is much more reflective to Earth based observers, and the satellites take about 3 months to get to their proper orbit. But once they are in their operational orbit, and rotated to Shark Fin configuration, it seems like it could be less reflective.
2$\begingroup$ I'll give you the 'unsourced' part (but SpaceX tends to provide the most crappy public statements possible in the form of Twitter comments), but the answer absolutely does address both "can" and "should". You're being overly critical. $\endgroup$– HobbesNov 13, 2019 at 9:23
1$\begingroup$ @Hobbes respectfully disagree. "mitigate" could be anything from 0.5% to 99.5% reduction. $\endgroup$ Nov 13, 2019 at 14:10
2$\begingroup$ @uhoh Added a link I found. $\endgroup$– geoffcDec 8, 2019 at 0:04
$\begingroup$ I've added some quotes that address the question to avoid the possibility of link-only-ness. Thanks for the follow-up! $\endgroup$– uhohDec 8, 2019 at 0:20
$\begingroup$ Did nobody in the Astronomy community think 12,000 new satellites in LEO might be a problem? $\endgroup$– uhohDec 9, 2019 at 0:48