The new Starlink V2 satellites incorporate improvements to the chassis brightness through dielectric mirrors, off-pointing solar arrays, and black paint on exposed components. For the general case in which the reflectivities are initially unknown, we simulate LSST operations and repeated photometry of every satellite in simulated model constellations. We derive a brightness model of the Starlink V2 satellite and study the simulated apparent brightness as a function of the satellite position relative to the observer and the sun. We find that the V2 Starlink satellites appear brightest at two distinct positions in the sky: when oriented toward the sun at low elevations where light is specularly reflected, and nearly overhead where the satellite is closest to the observer. A simulation of Starlink V2 satellites at 550 km height distributed across a series of Walker constellations with varying inclinations was analyzed to study the impact on the LSST observations. Some bright satellites will be visible in LSST observations. For every thousand V1.5 Starlink satellites imaged by LSST in the first hour of a summer night, we find 1.2 of them will appear brighter than 7 AB magnitude. By comparison, for every thousand V2 Starlink satellites observed, we find only 0.93 of them will appear this bright. The off-pointed solar array and reduced diffuse reflection of the chassis mitigate the brightness. Finally, we simulate lowering this Walker constellation to 350km. Only 0.56 V2 Starlink satellites per thousand brighter than 7 AB magnitude will be observed in the first hour at this height. This is a 40% reduction in number of bright satellites entering the focal plane compared to the constellation at 550km height. We find that a combination of factors yield an apparent surface brightness of these satellites for LSST operations only 5% brighter than at 550km orbit.
Kandula et al. (Mon,) studied this question.