I have spent a good part of my life pointing telescopes upward and trying to persuade people to look up too. So it takes something fairly extraordinary to make me genuinely uneasy about what's coming for the night sky. This does it.

A new paper, just accepted for publication in Astrophysical Journal Letters, has put hard numbers on a plan that's been rattling around the astronomy community for a while now - a Californian company called Reflect Orbital wants to put giant mirrors in orbit and beam sunlight down to Earth after dark. Their pilot satellite, Eärendil-1, carries an 18 by 18 metre reflector at roughly 600 kilometres altitude and is designed to throw a 2.5 kilometre patch of daylight onto a chosen spot on the ground, whenever a customer wants it. Solar farms working through the night, disaster zones lit for rescue crews, that's the pitch. It's not a one-off. Eärendil-1 is the test article for a proposed constellation of around 50,000 of these things, the larger production versions stretching to 54 by 54 metres apiece.

A photographic mosaic of the full Moon, assembled from over a thousand images taken by NASA's Lunar Reconnaissance Orbiter, the benchmark researchers used to measure Eärendil-1's brightness against, and which a single 54-metre mirror would outshine roughly four magnitudes over (Credit : NASA/Lunar Reconnaissance Orbiter/Goddard Space Flight Center)

The team behind the new paper, led by researchers including Gáspár Bakos, did the atmospheric physics properly. They modelled how sunlight bounced off a single mirror would scatter through the air, both the direct beam and the glow it throws off into the surrounding sky through Rayleigh and aerosol scattering. The results are stark. Anyone standing in the beam of one 54 metre satellite would see it as a point of light roughly magnitude -16.7, about four magnitudes brighter than a full moon. Step outside the beam and it doesn't let you off lightly either: from 14 kilometres away, the scattered glow alone would outshine full moonlight across most of the sky, and you'd still notice it as brighter than moonlight from 34 kilometres out. The diffuse background would resemble dusk shortly after sunset, bright enough to wash out every star but the very brightest. Run 400 of these mirrors on the same patch at once, which the company's long-term plans essentially call for, and the glow becomes obvious from 80 kilometres away.

This isn't really a story about one satellite. It's a story about who gets to decide what the sky looks like. Eärendil-1 already cleared the FCC in July, over more than 1,800 public comments and formal objections from the American Astronomical Society and dark sky groups, on the reasoning that light pollution and astronomical interference simply aren't the FCC's problem to regulate. Nobody else has claimed that job either.

A clear, dark night sky over rural Pennsylvania, one of the darkest remaining regions in the eastern United States, the kind of view increasingly rare, and precisely what's at stake as orbital light sources multiply overhead (Credit : Ken Slade)

I think about the people who'll never get a say in this. The amateur astronomer setting up a scope in a garden for the first time, hoping for a clean view of Saturn. The migrating birds and nocturnal animals whose entire behaviour is calibrated to darkness. The people who moved to rural locations just for the piece and quiet and yes, the dark night-time environment. The handful of dark sky places left in Britain, mine included, that people drive for hours to reach precisely because the sky there still does what skies are supposed to do. A single test satellite is not the end of the night sky. But it is a fairly clear preview of what "sunlight on demand" actually costs everyone who isn't buying it.

Source : Atmospheric Light Pollution by Proposed Reflect Orbital Space Mirrors