The Night Sky
If Everything Gets Approved
Two FCC applications are pending that would together place over one million bright objects in Earth's orbit. Astronomers have run the simulations. The headline finding: for large portions of the night and year, more satellites would be visible than stars.
The Science Behind the Headlines
The "more satellites than stars" finding comes from simulations published in early 2026 by astronomers Samantha Lawler (University of Regina) and Aaron Boley (University of British Columbia), who modelled SpaceX's proposed orbital data centre constellation using the orbital parameters from the FCC filing. They calibrated their model against real Starlink satellite brightness observations, scaling for the larger solar arrays required by SpaceX's data centre satellites.
Their finding: for large portions of the night and year, throughout the world, the number of simultaneously visible SpaceX satellites would exceed the total number of stars visible to the naked eye in a natural dark sky. At certain times of year at mid-latitudes, tens of thousands of satellites would be visible simultaneously — a moving grid of artificial lights overlaid on the sky.
Reflect Orbital's 50,000 mirrors add a qualitatively different problem. Each beam would create a fast-moving intense spot 4× brighter than the full moon sweeping across the landscape. These are not faint background points — they are active, directed illumination events affecting specific ground areas for minutes at a time. The Royal Astronomical Society calculated that the full 50,000-mirror constellation would make the entire night sky 3–4× brighter than its natural state.
Seven Consequences — From Concrete to Civilisational
1. Professional astronomy crippled. The Very Large Telescope would lose approximately 10% of data per image from satellite trails — rising to 30% for certain observation types. NASA's SPHEREx mission could have up to 96% of images affected. Essentially every Hubble Space Telescope observation would contain at least one satellite trail. New telescopes like the Vera C. Rubin Observatory — designed to survey the entire sky repeatedly — would face continuous contamination.
2. Radio astronomy affected. SpaceX's satellites will also use Ka-band radio frequencies for telemetry. Radio observatories, which already manage carefully around Starlink's radio emissions, would face dramatically increased interference. The Square Kilometre Array — the most sensitive radio telescope ever built — would be severely affected.
3. Wildlife impacts unshieldable. Migratory birds navigate by stars. Sea turtles use the sky to orient toward the ocean. Nocturnal pollinators use light cues for activity timing. Ground-level light pollution can be managed with shielding and local ordinance. Orbital light — moving, unpredictable, international — cannot be managed at the ecosystem level.
4. Human circadian disruption. The mirror beams in particular — 4× full moon intensity over 5km ground footprints — would constitute meaningful light exposure for anyone in their path. A person asleep in a room with thin curtains would experience sudden near-moonlight-level illumination. Repeatedly. Unpredictably. The satellite does not know what time zone it is passing over.
5. The dark sky as heritage — permanently altered. Every major religion and culture has a relationship with the night sky. Navigation, calendar, mythology, poetry, science — all have their roots in humans looking up at a dark sky and seeing the stars. Generations born after full deployment of these constellations would grow up under a permanently altered sky, never knowing what their ancestors saw. This is not recoverable — once the infrastructure is in orbit, removing it is a project of comparable scale to deploying it.
6. Cascade risk. At one million satellites, the risk of Kessler Syndrome — the runaway debris cascade from satellite collisions — increases dramatically. A chain reaction that fills low Earth orbit with debris would block access to space for potentially centuries, affecting GPS, weather satellites, and communications infrastructure that modern civilisation depends on.
7. Precedent without governance. If these constellations are approved, they establish the precedent that any commercial operator can alter the night sky for everyone on Earth without international consent. China, Europe, and other actors are all developing orbital data centre and space energy programmes. The regulatory framework for managing this collectively does not yet exist.
The Reality Check
No. The simulations model what happens if everything is approved and deployed at full stated scale. Neither application has been approved. A million-satellite deployment is economically and logistically unprecedented — Ars Technica estimates the launch cost alone would exceed one trillion dollars, roughly SpaceX's entire IPO valuation. Reflect Orbital's full 50,000-mirror constellation would require sustained investment and deployment over many years.
The more realistic near-term scenario is: Eärendil-1 launches as a demonstration satellite, Reflect Orbital attempts to scale commercially over years, and the SpaceX data centre constellation begins at a much smaller scale than one million. The night sky impact of a single Eärendil-1 — one mirror — is negligible. The impact of 50,000 is severe. The path between those two points runs through economics, regulation, and public response.
What astronomers argue — and what this page is intended to communicate — is that the time to establish governance frameworks, brightness standards, and international coordination is before deployment, not after. Once orbital infrastructure at this scale exists, the leverage to regulate it largely disappears.