Go outside after dark in an average mid-latitude town, look up, and you are under more artificial objects than you might think. A new analysis reported by Space.com on November 17 estimates that about 710 low-Earth-orbit satellites sit above the horizon of a typical UK town at any given moment. For anyone stargazing from a city, that is a larger number than the stars they can actually see.

The headline needs some unpacking, because "above the horizon" and "visible" mean different things. The comparison also rests on two very different kinds of evidence. The town-level count comes from an industry analysis by Accu Components that has not been peer-reviewed. The longer-range warning comes from a peer-reviewed study by European Southern Observatory astronomer Olivier Hainaut, accepted for publication in Astronomy and Astrophysics. Both say the night sky is changing quickly, but they should be read differently.

What the new count actually measured

The Accu Components analysis used US Space Force tracking data, distributed through CelesTrak, for two dates: October 12 and October 18. Across those snapshots it worked with 15,833 active satellites in low Earth orbit and calculated how many would be above the horizon of a given location, assuming nothing like buildings or hills blocks the view.

For an average UK town, the answer was about 710 at any moment. That figure counts every satellite overhead, lit or not. A satellite is only visible from the ground when sunlight reaches it while the observer is in darkness, so the more useful number is smaller. Of those 710, the analysis estimates about 105 are sunlit and roughly as bright as stars.

Timing matters a lot. In the first hour after dark, when the Sun is just below the observer's horizon but still lights satellites high above, about 300 are sunlit. Around midnight, that drops to fewer than 20. For a casual observer this is the most practical takeaway: early evening is when satellites are most likely to show up, and the middle of the night is much quieter.

Why mid-latitudes get the crowd

Location matters too. The analysis found the equator is quieter, with about 490 satellites overhead at a time compared with roughly 710 over the UK. The reason is orbital geometry. Most Starlink satellites orbit at about 53 degrees to the equator. According to Accu Components, that means the satellites bunch up over the mid-latitudes, including most of Europe and the United States.

The stars on the other side of the comparison

The "outnumber" claim depends on how many stars you can see, and that number is hit hard by light pollution. According to the Space.com report, an observer in London or New York can see only a few dozen stars. In a small town with darker skies, the count rises to a few hundred.

Advertisement

That is where the comparison comes from. Under bright urban skies, the roughly 710 satellites overhead clearly outnumber the visible stars, even though most of those satellites cannot be seen at that moment. In a small town with darker skies, the roughly 300 sunlit satellites in the first hour after dark are in the same range as the few hundred stars visible. Neither comparison means the sky looks like a swarm. It does mean that satellites are now a structural part of the night sky rather than an occasional novelty.

The peer-reviewed warning: where this could go

The more serious alarm comes from ESO. In a press release dated July 1, 2026, the observatory described Hainaut's study, which modelled what the sky would look like if current proposals were built. ESO puts today's population at more than 14,000 satellites in orbit, most of them Starlink. The proposals Hainaut considered add up to more than 1.7 million satellites, which the study says would have "devastating consequences for astronomy."

Advertisement

The largest single proposal in that total is SpaceX's plan for one million more satellites to host space-based data centres. According to ESO, if that constellation were built, hundreds of its satellites would be visible for much of each night, and at times several thousand. That is similar to the number of stars visible to the naked eye under good conditions. The study also covers other proposed systems, including E-Space's Cinnamon constellation, China's CTC-1 and CTC-2, and orbital mirrors from Reflect Orbital, which are designed to reflect sunlight to the ground.

Hainaut's recommendation is specific: no more than 100,000 satellites in total, and all of them faint enough to stay below naked-eye visibility. Open Access Government's summary puts that threshold at magnitude 7, the threshold of naked-eye visibility. The same coverage notes that the satellite population has grown rapidly since 2019 to more than 14,000 active spacecraft, driven mainly by Starlink, and that the study quantifies both sky brightening and direct interference at professional observatories.

Why It Matters

The two analyses tell a consistent story at different scales. The Accu Components count, with its caveats, describes the sky right now: roughly 15,800 active low-orbit satellites already add up to hundreds overhead at mid-latitudes and dozens to hundreds of sunlit ones in the hours after dusk. The ESO study describes what happens if the industry's stated plans go ahead: a population about a hundred times larger, with naked-eye satellites at times rivalling the stars themselves.

For amateur observers, the effect is already clear: satellite trails in long-exposure photos, moving points of light in early-evening views, and a sky that no longer looks the same as it did a decade ago. For professional astronomy, the stakes are higher. The ESO study's language is not the usual cautious academic tone. It calls the outcome devastating and proposes a hard cap. Whether regulators or operators pay attention is a separate question, but the peer-reviewed numbers now exist for anyone making that decision.

Readers should keep the evidence levels separate. The town-by-town counts are a useful illustration from a company analysis, not a vetted scientific result. The mega-constellation projections and the 100,000-satellite recommendation are the peer-reviewed part. Together they suggest that "satellites outnumber the stars" is already true in some places today, and could become true almost everywhere if current proposals are built.

Sources