How Starlink and Satellite Constellations Are Threatening Darkness for Astronomers

In January 2020, astronomers at the Cerro Tololo Inter-American Observatory in Chile opened an archival image taken by the Dark Energy Camera and found 19 parallel white streaks cutting across a single 332-second exposure. Every one of them was a Starlink satellite. That image became the photograph that made the problem impossible to dismiss.

SpaceX had launched its first batch of 60 Starlink satellites in May 2019. Within months, observers around the world were reporting bright trains of moving lights crossing the sky in formation, visible to the naked eye, photographed by amateurs, and landing in the data of professional telescopes that had no mechanism to simply edit them out. By mid-2025, SpaceX had placed more than 6,700 Starlink satellites in orbit. Amazon's Project Kuiper, OneWeb, and China's Guowang constellation are adding thousands more. The International Astronomical Union estimates that by the early 2030s, the total number of active satellites in low Earth orbit could exceed 100,000.

The damage a satellite trail does to an astronomical image depends on the telescope, the exposure length, and what the telescope was hunting. For wide-field surveys, the kind designed to catch near-Earth asteroids, supernovae, or gravitational wave counterparts, a single bright streak can blind the detector in that column for the entire exposure. The Vera C. Rubin Observatory in Chile, which will conduct the Legacy Survey of Space and Time starting around 2025, is expected to have up to 30 percent of its twilight images affected by satellite trails at current constellation sizes, according to modelling published by the observatory's own team.

Radio astronomy faces a different version of the same problem. Satellite internet constellations transmit in frequency bands that overlap with those used by radio telescopes. The Square Kilometre Array, being built partly in South Africa and partly in Western Australia, has raised formal concerns about interference from broadband satellite signals leaking into protected radio-astronomy bands. The National Radio Astronomy Observatory in the United States filed detailed technical objections with the Federal Communications Commission over SpaceX's shell 2 deployment, citing measured interference levels that exceeded the ITU's recommended thresholds by orders of magnitude.

India has a direct stake. The Indian Institute of Astrophysics operates the Vainu Bappu Observatory at Kavalur in Tamil Nadu and the Indian Astronomical Observatory at Hanle in Ladakh, one of the world's highest-altitude optical observatories at 4,500 metres. Both sites are already dealing with increased sky brightness from satellite trails and ground-based light pollution. The Hanle Dark Sky Reserve, declared in 2022 around the observatory, was designed to protect the site from artificial light. Satellite trails are a different category of intrusion: no local ordinance can stop them.

ISRO itself is navigating the same orbital environment. As India's own satellite fleet grows, Chandrayaan-3 reached the Moon's south pole in August 2023, and Gaganyaan crewed missions are in preparation, the crowding of low Earth orbit creates collision-avoidance calculations that grow more complex with every new constellation shell. The 2009 collision between Iridium 33 and Cosmos 2251 produced roughly 2,000 trackable debris fragments. A collision inside a mega-constellation would be geometrically worse.

SpaceX responded to early complaints by developing a sunshade called VisorSat, which reduced the brightness of Starlink satellites by about a magnitude. The company later switched to a dielectric mirror film on newer satellites that deflects sunlight away from Earth-facing angles. These measures help, but they do not eliminate the problem. Independent measurements by a team at the University of Astrophysics in Leiden found that even the darkened satellites remain two to three times brighter than the limit the IAU considers acceptable for protecting optical astronomy.

The IAU established a Centre for the Protection of the Dark and Quiet Sky from Satellite Constellation Interference, known as CPS, in 2022, co-hosted by the NOIRLab and the SKA Observatory. The CPS coordinates technical working groups on optical interference, radio interference, and regulatory strategy. Their most concrete tool so far is SatChecker, a software system that allows observatories to predict when a satellite will cross a telescope's field of view and schedule exposures around the conflict. It is a workaround, not a fix.

Regulatory pressure is mounting. The United States National Science Foundation has called for environmental reviews of large constellation approvals. The United Nations Committee on the Peaceful Uses of Outer Space has discussed satellite brightness guidelines, though it has no enforcement authority. The ITU allocates radio-frequency spectrum and orbital slots, but its processes were designed for an era when a single satellite licence represented a single spacecraft, not a fleet of thousands filed under one application.

There is a number that does not appear in most regulatory filings: the fraction of humanity that has never seen the Milky Way. A 2016 study published in Science Advances by Fabio Falchi and colleagues calculated that 80 percent of the world's population lives under light-polluted skies. In India, the figure is higher in most urban corridors. The night sky that a child in Jaisalmer or Spiti can still see, the actual band of the galaxy, not a glow, is already a rarity. Satellite constellations are not the cause of that loss, but they are the next layer being added to it, and unlike a streetlamp, they cannot be switched off by a municipality.

The stars that ancient Indian astronomers used to develop the Panchangam calendar, the nakshatras that structured Vedic timekeeping, the sky that Aryabhata calculated from in the 5th century, that sky is still mostly there, technically. The photons still arrive. What is changing is the noise floor around them, the number of moving human objects now competing for the same dark field. Whether the regulatory frameworks catch up before the observing window closes is a question the next decade will answer, not this one.