How Much of Earth Can One Starlink Satellite See?

Last updated: July 10, 2026

Introduction

A single Starlink satellite can see a surprisingly large patch of Earth in the purely geometric sense. At an altitude near 550 km, a satellite’s line of sight reaches roughly 2,557 km from the point directly below it, making a horizon-to-horizon footprint of about 5,100 km across. The surface area inside that theoretical horizon circle is about 20.3 million square kilometers, or about 4.0% of Earth’s surface.

That number is useful, but it is also easy to misuse. The geometric horizon is not the same as the usable broadband service area. A terminal near the edge of that footprint would see the satellite almost exactly on the horizon, where the signal path is longer, atmospheric loss is worse, obstruction risk is high, interference coordination is harder, and the satellite may not be allowed or designed to serve that geometry. If we require a more practical minimum elevation angle, such as 25 degrees above the local horizon, the footprint of the same 550 km satellite shrinks to about 941 km in surface radius and about 2.77 million square kilometers.

This article uses simple spherical-Earth physics to estimate what one Starlink satellite can see. It then separates “can see” from “can serve,” because those are not the same engineering question.

The Simple Picture: A Satellite Above a Spherical Earth

For a first-order calculation, treat Earth as a sphere with radius:

R = 6,371 km

Then put a satellite at altitude h above the surface. Its distance from Earth’s center is:

r = R + h

The point on Earth directly below the satellite is the subsatellite point. The question is how far away, measured along Earth’s surface, the satellite can still have a straight line of sight to the ground before Earth itself blocks the view.

At the geometric horizon, the line from the satellite to the ground just grazes Earth. That grazing line is tangent to the surface. The radius drawn from Earth’s center to the horizon point meets the line of sight at a right angle. This makes a right triangle with one side R and hypotenuse R + h.

The central angle from the subsatellite point to the horizon is:

cos(theta_h) = R / (R + h)

theta_h = arccos(R / (R + h))

The surface distance from the subsatellite point to the horizon is:

s = R * theta_h

where theta_h must be in radians. The visible surface area is the area of a spherical cap:

A = 2 * pi * R^2 * (1 – cos(theta_h))

The fraction of Earth’s total surface area is:

f = A / (4 * pi * R^2) = (1 – cos(theta_h)) / 2

These equations do not know anything about antennas, spectrum, beams, traffic demand, gateways, national licenses, or user terminals. They only answer the clean geometry question: if nothing else mattered, how much of the spherical Earth is above the satellite’s local horizon?

Horizon Footprints for 340 km, 480 km, and 550 km

Starlink has used and proposed multiple low Earth orbit altitude regimes. Public FCC documents for the Gen2 system include authorizations around the 525-535 km range, and later FCC material discusses flexibility involving shells around 340-365 km and 480-485 km. The exact constellation changes over time, so the numbers below are not a live inventory. They are altitude cases for understanding the physics.

Using R = 6,371 km, the pure horizon results are:

340 km altitude

r = 6,371 + 340 = 6,711 km

cos(theta_h) = 6,371 / 6,711 = 0.949337

theta_h = arccos(0.949337) = 18.32 degrees = 0.3197 radians

s = 6,371 * 0.3197 = about 2,037 km

A = 2 * pi * 6,371^2 * (1 – 0.949337) = about 12.9 million km^2

Fraction of Earth = (1 – 0.949337) / 2 = 0.0253, or about 2.53%

So a 340 km satellite has a theoretical horizon footprint roughly 4,074 km across, covering about 12.9 million square kilometers.

480 km altitude

r = 6,371 + 480 = 6,851 km

cos(theta_h) = 6,371 / 6,851 = 0.929937

theta_h = arccos(0.929937) = 21.57 degrees = 0.3766 radians

s = 6,371 * 0.3766 = about 2,399 km

A = about 17.9 million km^2

Fraction of Earth = about 3.50%

At 480 km altitude, the geometric horizon footprint is about 4,798 km across. Compared with 340 km, the satellite is only 140 km higher, but the visible cap grows by about 5 million square kilometers. Horizon geometry is sensitive at low altitudes because the satellite is close to a curved planet.

550 km altitude

r = 6,371 + 550 = 6,921 km

cos(theta_h) = 6,371 / 6,921 = 0.920532

theta_h = arccos(0.920532) = 23.00 degrees = 0.4013 radians

s = 6,371 * 0.4013 = about 2,557 km

A = about 20.3 million km^2

Fraction of Earth = about 3.97%

At 550 km altitude, the geometric footprint is roughly 5,114 km across from horizon edge to horizon edge. That is the number behind the tempting statement that one low Earth orbit broadband satellite can “see” several countries at once. Geometrically, that is true. Operationally, it is incomplete.

Why the Horizon Footprint Is Too Generous

The horizon case assumes the terminal can use a satellite at 0 degrees elevation, exactly along the local horizon. That is rarely a good broadband assumption.

At low elevation, the signal travels through more atmosphere. The path also passes through more rain, cloud water, and near-ground clutter. Buildings, trees, terrain, roof edges, vehicle structures, and ship hardware are more likely to block the view. Antennas also have steering limits and gain patterns. A satellite system must manage interference with other satellites and terrestrial systems, which can limit the lowest useful angles. Regulations and coordination agreements matter too.

For this reason, a practical footprint is usually smaller than the pure horizon footprint. Instead of asking, “Where is the satellite barely visible?” a network engineer asks something closer to, “Where is the satellite high enough in the sky to be useful under the system’s rules?”

Let e be the minimum elevation angle above the local horizon. The central angle to the edge of the practical footprint is:

theta_e = 90 degrees – e – arcsin((R / (R + h)) * cos(e))

This is the same triangle as before, but now the line of sight is not tangent to Earth. The terminal is required to see the satellite at least e degrees above the horizon. When e = 0 degrees, the equation collapses back to the geometric horizon formula.

For illustration, use e = 25 degrees. This is not a claim that every Starlink link everywhere always uses exactly 25 degrees. It is a transparent cutoff for showing how strongly the footprint changes once the horizon edge is excluded.

Practical footprint at 25 degrees minimum elevation

At 340 km altitude:

theta_25 = about 5.64 degrees

surface radius = about 627 km

area = about 1.23 million km^2

fraction of Earth = about 0.242%

At 480 km altitude:

theta_25 = about 7.56 degrees

surface radius = about 841 km

area = about 2.22 million km^2

fraction of Earth = about 0.435%

At 550 km altitude:

theta_25 = about 8.46 degrees

surface radius = about 941 km

area = about 2.77 million km^2

fraction of Earth = about 0.544%

This comparison is the most important result in the article. A 550 km satellite can geometrically see about 20.3 million km^2 down to the horizon, but if the useful link budget and operating rules require the satellite to be at least 25 degrees above the horizon, the footprint is closer to 2.77 million km^2. That is still huge, but it is only about 14% of the pure horizon cap.

The exact practical number changes with the chosen elevation angle. At 550 km, a 10-degree minimum elevation gives a much larger footprint, about 1,664 km in surface radius and 8.65 million km^2 in area. At 35 degrees, the footprint falls to about 674 km in radius and 1.42 million km^2. The physics is continuous, but the business and regulatory system usually has to choose thresholds.

Visible Area Is Not the Same as Capacity

The biggest misconception is to turn the footprint area into a capacity claim. A visible footprint says where a satellite could have line of sight. It does not say how many users can be served, how fast they can download, or whether service is available in a specific neighborhood.

A broadband satellite does not pour a uniform internet signal over the whole visible cap. It uses antennas, beams, frequencies, scheduling, power budgets, and backhaul paths. A modern Starlink satellite uses phased-array technology and, on many satellites, optical inter-satellite links, but the public geometry alone does not reveal the instantaneous number of beams, the bandwidth assigned to each beam, the modulation under weather conditions, the number of active terminals, or the traffic demand in each cell.

Imagine the geometric horizon footprint as the maximum stage on which the satellite could interact with Earth. The actual service footprint is the part of that stage selected by antenna steering, spectrum reuse, network scheduling, and legal authorization. Within that service footprint, capacity is shared.

If a satellite is over the ocean, much of its visible surface may have few or no users. If it is over a dense city with many terminals, the same geometric footprint can contain more demand than the satellite can satisfy at peak speeds. If it is over a country where Starlink lacks authorization, geometry may say “visible” while regulation says “not serviceable.” If the satellite has no direct path through the constellation or to an approved gateway for a particular route, line of sight alone is still not enough.

This is why coverage maps and real user experience do not follow from a single satellite footprint calculation. Starlink capacity depends on the whole constellation: satellite density, orbital shell geometry, gateway placement, spectrum rights, user terminal design, beam scheduling, inter-satellite laser routing, local congestion, obstructions, and the regulatory status of each market.

Altitude Tradeoff: Bigger Footprint or Lower Latency?

Higher low Earth orbit altitude increases the visible footprint. That can reduce the number of satellites required for continuous geometric coverage. However, altitude is not simply “higher is better.”

Lower altitude reduces free-space path length and can reduce the space segment contribution to latency. It can also improve link budget in some cases because the satellite is physically closer to the terminal. Lower altitude may help with end-of-life disposal because atmospheric drag is stronger, although it also means the satellite must fight more drag during operations. The tradeoff is that lower altitude shrinks each satellite’s footprint and usually requires more satellites for the same continuous coverage pattern.

The numbers above show the geometric side of this tradeoff. Moving from 340 km to 550 km increases the pure horizon cap from about 12.9 million km^2 to about 20.3 million km^2. With a 25-degree elevation cutoff, the practical cap grows from about 1.23 million km^2 to about 2.77 million km^2. That is more than double the practical area in this simplified calculation.

But a larger footprint can also mean a larger region of potential demand competing for the satellite’s limited radio resources. A satellite that can see more terminals does not automatically have more total capacity. It may simply have more possible users to schedule.

For a realistic network, altitude must be considered alongside orbital inclination, number of planes, number of satellites per plane, beam design, gateway topology, optical links, spectrum reuse, and the distribution of customers on the ground. The altitude calculation is necessary, but it is only one piece.

What About Continuous Coverage?

One satellite in low Earth orbit moves quickly across the sky. ESA’s general LEO explainer notes that satellites in this region travel at roughly 7.8 km/s and can circle Earth in about 90 minutes, depending on altitude. That means a single Starlink satellite does not park over one region. It rises, crosses the sky, and sets.

Continuous service requires handoffs from one satellite to the next. From the user’s perspective, the terminal and network must keep finding usable satellites above the required elevation angle. From the constellation perspective, many satellites in coordinated orbital planes create overlapping windows of service.

This is also why the phrase “one satellite covers 4% of Earth” can mislead. Even if the 550 km horizon cap is about 4% of Earth’s surface, that cap is moving rapidly. The satellite is not providing fixed, permanent 4% global coverage. It is sweeping a changing line-of-sight region around the planet.

For a rough time scale, a satellite near 550 km has an orbital period of about 95 minutes. Earth rotates underneath the orbit, the satellite moves along its plane, and the ground track shifts. A user does not care about the total cap area as much as whether at least one authorized, uncongested, unobstructed satellite is above the local useful elevation angle at that moment.

A Worked Example: Reading the 550 km Number Correctly

Suppose a Starlink satellite is at 550 km altitude. The clean horizon calculation says:

theta_h = 23.00 degrees

surface radius = 2,557 km

horizon area = 20.3 million km^2

This means a point 2,500 km away along Earth’s surface may still be barely visible in the geometric sense. But “barely visible” is doing a lot of work. Near that edge, the satellite is close to the horizon. A tree line, a hill, a building, or a ship mast could block it. The signal has a long slant path. The antenna geometry is less favorable. The system may avoid serving that link.

Now apply a 25-degree minimum elevation:

theta_25 = 8.46 degrees

surface radius = 941 km

practical illustrative area = 2.77 million km^2

The satellite still covers a region larger than many countries. But the footprint is no longer continental in the casual horizon sense. It is a large moving service opportunity zone, and only part of that zone may be actively served at any instant.

This is the most realistic way to talk about single-satellite coverage: give the horizon number, then immediately explain the elevation-limited number and the network constraints.

Why Ground Stations and Lasers Matter

Line of sight from user to satellite is only the first hop. The user’s traffic must also reach the internet. That can happen through a gateway ground station connected to terrestrial fiber, through optical inter-satellite links to another satellite, or through a combination of routing paths.

If a satellite can see a user but cannot route the traffic efficiently to a gateway or through the satellite mesh, service may be limited or unavailable. Laser links help reduce dependence on nearby ground stations, especially over oceans, polar areas, and regions with sparse terrestrial infrastructure. But lasers do not remove all constraints. The network still has finite satellite capacity, finite spectrum, finite power, routing limits, weather effects on gateway links, and national rules.

This is why the footprint question should be framed as geometry, not as a promise of service. The geometry tells us where links are possible. The network tells us which links are useful. Regulation tells us which links are allowed.

Bottom Line

At 550 km altitude, one Starlink satellite can geometrically see a circular cap of Earth with a surface radius of about 2,557 km, covering about 20.3 million km^2, or about 4.0% of Earth’s surface. At 480 km, the comparable horizon footprint is about 17.9 million km^2. At 340 km, it is about 12.9 million km^2.

For practical broadband realism, the more important number is the elevation-limited footprint. If we require the satellite to sit at least 25 degrees above the local horizon, the 550 km case falls to about 941 km in surface radius and about 2.77 million km^2. At 480 km it is about 841 km and 2.22 million km^2. At 340 km it is about 627 km and 1.23 million km^2.

So the answer depends on the meaning of “see.” If it means pure line of sight to the geometric horizon, one satellite at Starlink-like LEO altitude can see millions to tens of millions of square kilometers. If it means usable broadband service, the footprint is much smaller and depends on minimum elevation, beams, spectrum, terminal view, gateways, inter-satellite links, local load, obstruction, and regulation.

The safe conclusion is not that one satellite “covers” an entire continent with guaranteed capacity. The safe conclusion is that low Earth orbit geometry gives each satellite a large moving field of view, while real Starlink service emerges only when many satellites, many beams, legal spectrum access, backhaul, and user terminals work together.

Sources

FCC, Space Exploration Holdings, LLC, FCC 22-91, Order and Authorization, released December 1, 2022: https://docs.fcc.gov/public/attachments/FCC-22-91A1.pdf
FCC, SpaceX Gen2 Upgrade Applications Partial Grant, DA 26-36, released January 2026: https://docs.fcc.gov/public/attachments/DA-26-36A1.pdf
Starlink, Technology: https://www.starlink.com/technology
NASA Earthdata, Orbits: https://www.earthdata.nasa.gov/learn/earth-observation-data-basics/orbits
ESA, Low Earth orbit: https://www.esa.int/ESA_Multimedia/Images/2020/03/Low_Earth_orbit