Satellite megaconstellations

Satellite megaconstellations are networks of hundreds to tens of thousands of small spacecraft flying in low Earth orbit, linked to beam internet service across the planet. The largest, SpaceX’s Starlink, passed 10,000 active satellites in 2026, and it is no longer alone. What began as a single company’s project has become a race among governments and corporations to fill low orbit with hardware, and the consequences for astronomy, orbital safety, and the atmosphere are only starting to come into focus.

This page covers what a megaconstellation is, who is building them, and the two problems that follow from putting so many objects overhead: what it does to the night sky, and what happens when those satellites come back down.

What a megaconstellation is

The word describes scale. A traditional communications satellite sat in geostationary orbit about 36,000 kilometers up, a single large spacecraft covering a wide area with noticeable signal lag. A megaconstellation flips that design. It uses thousands of small, cheap satellites flying low, often between 300 and 600 kilometers, close enough to cut the delay that made older satellite internet frustrating to use.

Flying low forces the numbers up. A satellite at 550 kilometers sees only a small patch of ground and moves across the sky in minutes, so covering the whole planet without gaps takes a dense mesh of them handing off connections as they pass. That is why these systems are counted in the thousands rather than the dozens. The tradeoff is that low-orbit satellites feel atmospheric drag and last only about five years before they fall, which means the operator has to keep launching replacements indefinitely to hold the network together.

Who is building them

SpaceX set the pace and still dominates. As of July 2025, Starlink operated about 7,788 active satellites. By the middle of 2026 that figure had grown past 10,700, drawn from more than 11,500 launched over the program’s life. Starlink by itself accounts for more than half of every working satellite in orbit around Earth. In January 2026 the U.S. Federal Communications Commission cleared SpaceX to add 7,500 more second-generation satellites, lifting its approved total to 15,000.

The competitors are further behind but moving. Amazon’s Project Kuiper, planned at 3,236 satellites, launched its first operational batch of 27 in April 2025 and faces an FCC requirement to orbit half its fleet by July 30, 2026. Eutelsat OneWeb, an earlier entrant that went through bankruptcy and restructuring, holds a smaller constellation of roughly 650 satellites in a higher orbit, serving enterprise and government customers rather than consumers.

China has entered with state backing and ambitions to match SpaceX. The Guowang constellation, run by a state-owned enterprise, began launching in December 2024 and had around 145 satellites in orbit by 2026, with plans for roughly 13,000. A second Chinese network, Qianfan, also called Thousand Sails, launched its first satellites in August 2024 and aims for about 15,000, though its deployment paused for months in 2025 after some satellites suffered thruster and gyroscope failures before resuming in 2026. Between the two, China has authorized plans for tens of thousands of satellites, turning what was one company’s domain into a contest between nations.

Add the announced plans together and the trajectory is clear. Low Earth orbit, which held a few thousand active satellites total before Starlink, is on course to hold tens of thousands within a decade.

The impact on orbit and astronomy

Astronomers raised the alarm early, and the data has caught up with their worry. Satellites are visible because they reflect sunlight, and to a telescope a bright satellite crossing the field of view leaves a streak that ruins part of the exposure. The problem scales directly with the number of objects overhead.

A 2025 study published in Monthly Notices of the Royal Astronomical Society Letters measured constellation satellites against the brightness limits the International Astronomical Union set to protect research. Most of the satellites surveyed, across Starlink, OneWeb, and other fleets, exceeded the magnitude 7 threshold the IAU uses for professional observatories, and many are bright enough to cross the magnitude 6 line for naked-eye visibility. In plain terms, the objects are brighter than the sky can absorb without harm to serious astronomy.

The response has been partly technical and partly institutional. SpaceX has tried darkening coatings and sunshades to dim its satellites, with real but incomplete results, and the dimming is offset when the company flies satellites at lower altitudes where they appear brighter. The IAU created a body called the Centre for the Protection of the Dark and Quiet Sky from Satellite Constellation Interference to track the problem and push for standards. No binding international rule governs how bright a satellite is allowed to be, so mitigation depends on operators choosing to cooperate.

Beyond brightness sits the harder question of orbital traffic. Tens of thousands of satellites in a shared band of low orbit means constant collision-avoidance maneuvers and a rising risk that one bad collision could scatter debris into the paths of others. That crowding connects directly to the last piece of the story, because the way these satellites leave orbit is now a subject of its own.

The debris and reentry angle

The same low orbit that makes megaconstellations work also guarantees they fall apart on a schedule. Starlink satellites are designed to deorbit and burn up after about five years, and with more than 10,000 of them aloft, the retirements have become a steady stream. Astrophysicist Jonathan McDowell, who tracks the fleet, reports that one to two Starlink satellites reentered the atmosphere each day through 2025, a rate he expects to climb toward five per day as the constellation matures and its first generation ages out.

Burning up on reentry was long treated as the clean way to dispose of a satellite. Recent research questions that assumption. A 2024 study by José Ferreira and colleagues at the University of Southern California, published in Geophysical Research Letters, modeled what happens as a satellite vaporizes. A typical 250-kilogram satellite generates roughly 30 kilograms of aluminum oxide nanoparticles as it burns. Aluminum oxide is a catalyst for the chlorine reactions that destroy stratospheric ozone, and the particles can linger for decades, drifting down over as long as 30 years to the altitude where most of Earth’s ozone sits.

The totals are climbing with the launch rate. The study estimated that satellite reentries put about 17 metric tons of aluminum oxide into the upper atmosphere in 2022, and projected that figure could reach roughly 360 metric tons per year if planned constellations deploy as announced. Independent atmospheric measurements support the concern: the U.S. National Oceanic and Atmospheric Administration has found metals including aluminum, lithium, and copper in stratospheric particles at concentrations well above what natural space dust would leave, traced to spacecraft and rocket debris burning on reentry. The science on the eventual effect is unsettled, and researchers are careful to say the full impact is not yet understood. The scale of what is being injected, though, is measurable and growing.

There is a grim symmetry here. The industry solved the old problem of dead satellites lingering as space junk by designing them to burn up on the way down. That solution is now trading an orbital debris problem for an atmospheric one, moving the waste from space into the air we breathe under. Neither problem is settled, and both grow with every launch.

For a closer look at what happens when a spacecraft comes down, see our explainer on a satellite falling to Earth. For the wider picture of crowding and collision risk in low orbit, visit our space debris hub.