Earth 13 min read

Where Does Space Begin? The Kármán Line at 100 km

The atmosphere has no edge, so space begins wherever you draw the line. Why the Kármán line sits at 100 km when von Kármán's own figure was 84 km.

World of Physics Editorial October 10, 2026

karman lineatmospherespace boundaryaltitudeorbital velocityarmstrong limitexosphereastronaut wingsspace lawthermosphereaerodynamics

Table of Contents

Nobody Can Point at the Edge

Look up on a clear night and the sky does not end anywhere. It just gets thinner. That is not a poetic observation, it is the whole problem.

If you ask where space begins, you are asking a question that physics declines to answer. There is no altitude at which air density drops to zero, no shell, no surface, no discontinuity in any measurable quantity. The atmosphere falls off roughly exponentially, losing about half its density every 5 to 7 kilometres in the lower layers, and it keeps doing something like that forever. In 2019 a team working with the SWAN instrument on ESA’s SOHO spacecraft reported that Earth’s hydrogen geocorona extends to at least 100 Earth radii, about 630,000 kilometres. The Moon, at 384,000 kilometres, orbits inside Earth’s atmosphere. The density out there is around 0.2 hydrogen atoms per cubic centimetre, which is vacuum by any practical standard, but it is not nothing.

So every boundary you have ever heard of is a human decision. The interesting part is what physics each decision is trying to encode.

What von Kármán Actually Calculated

Theodore von Kármán’s argument is the best of them, because it is about a genuine crossover rather than a threshold.

An aircraft holds itself up with lift, and lift goes as L = ½ρv²SC_L. The density ρ is in there linearly, so as you climb and the air thins, you must fly faster to generate the same lift. Keep climbing and the required speed keeps rising. Meanwhile there is another speed in the problem: the orbital velocity v = √(GM/r), about 7.9 kilometres per second near Earth. If you are moving that fast, you do not need wings, because the curvature of your own trajectory already matches the curvature of the planet.

Those two speeds converge. Below the crossing, you fly. Above it, you orbit. The altitude where the speed needed for aerodynamic lift exceeds orbital velocity is the point where aeronautics hands over to astronautics, and that is von Kármán’s line.

Work it through with the US Standard Atmosphere and a lift coefficient of 1, and the answer depends on how heavily loaded the wing is. At 32 kilograms per square metre the crossover lands at 83.7 kilometres, which is where the familiar figure of 84 km comes from. Load the wing like an airliner, at 600 kg/m², and the crossover drops to about 64 kilometres, because a heavy aircraft needs denser air and therefore has to stay lower.

Why 100 km? Because It Is a Round Number

Here is the part that surprises people who assume standards are derived rather than chosen.

Von Kármán never published the calculation. The 84-kilometre figure reaches us mainly through Andrew Haley’s later account of his reasoning. And when the Federation Aeronautique Internationale, which certifies aviation and spaceflight records, needed a single unambiguous altitude, it did not pick 84 km. It picked 100 km: safely above anything an aeroplane could do, easy to remember, and a tidy number in kilometres.

That is the actual provenance of the most famous boundary in spaceflight. It is a rounding, in a particular unit, of an unpublished estimate. Had the standard been set in a country using statute miles, 50 miles would have been the obvious round number instead, and the Kármán line would sit at 80.47 km.

Fifty Miles, Eighty Kilometres, and McDowell’s Satellites

The United States never adopted 100 km. During the X-15 programme the Air Force set 50 statute miles, 80.47 km, as the altitude qualifying a pilot for astronaut wings, and the FAA later used the same figure for commercial astronauts. Thirteen X-15 flights crossed that line. Only two crossed 100 km.

In 2018 Jonathan McDowell went looking for a boundary the universe would recognise. Writing in Acta Astronautica, he worked through the perigees of roughly 43,000 catalogued objects and asked where an orbit stops being an orbit. He found satellites completing full revolutions with perigees well below 100 km, down into the 80s, and concluded that 80 km is the physically defensible number. Usefully, it also happens to coincide with the US convention and with the mesopause, the temperature minimum near 85 km that is as close to a natural atmospheric ceiling as anything gets.

So in the 80 to 100 km band you can be an astronaut and not an astronaut at once, depending on who counts.

The Boundaries That Are Real

None of this means the vertical structure of the atmosphere is featureless. It is full of genuine physical transitions, they just do not line up with anybody’s legal definition.

The troposphere, holding about 80% of the atmosphere’s mass, ends somewhere between 8 and 18 km depending on latitude, and the temperature stops falling there. The stratosphere warms with height because ozone absorbs ultraviolet, and tops out near 50 km. The mesosphere cools again to the coldest temperatures on the planet, around 180 K at the mesopause. Then the thermosphere heats up violently, to 1,000 K or more, which sounds alarming until you remember that temperature in a near-vacuum describes the speed of individual particles and not how much heat they can deliver.

The limits that matter for a human body arrive long before any space boundary. Above about 3 km you need supplemental oxygen. Above roughly 12 km oxygen alone stops working and you need it delivered under positive pressure. At 19.1 km, the Armstrong limit, ambient pressure falls to 6.3 kPa, the vapour pressure of water at 37 °C, and your saliva boils. The sky stops being blue somewhere above 30 km as there is no longer enough air to scatter sunlight. The aurora glows between about 90 and 400 km, which is to say that the prettiest part of the atmosphere is above every proposed edge of it.

Where Orbits Stop Being Orbits

The practical boundary, the one engineers actually care about, is drag.

Below about 160 km an object completes a few revolutions at most before the atmosphere takes it. At 400 km the International Space Station still runs into enough air to lose altitude continuously, which is why it needs reboosting several times a year, more often when solar activity puffs the thermosphere outward. Above 700 km drag becomes negligible on human timescales, which is exactly why debris at those altitudes is a century-scale problem.

None of these numbers is a boundary either. They are a smooth gradient of consequences on one exponential density profile. Even escape velocity, the cleanest threshold in orbital mechanics, is a speed rather than a height.

Why the Argument Has Stakes

This would be pedantry if nothing depended on it, but three things do.

Sovereignty is the big one. Under the Chicago Convention a state has complete and exclusive sovereignty over the airspace above its territory, with no stated upper limit, while the 1967 Outer Space Treaty declares space free for all and subject to no national appropriation. Those two regimes have to meet somewhere, and the treaty does not say where. The UN Committee on the Peaceful Uses of Outer Space has been discussing it since 1966 and has agreed on nothing, mostly because any line creates winners and losers in overflight rights.

Second, it decides who is an astronaut. When Virgin Galactic’s SpaceShipTwo reached 82.7 km in December 2018 and 89.9 km in February 2019, its pilots qualified under the US standard but not under the FAI’s. Blue Origin’s New Shepard, which crosses 100 km, made a public point of the difference.

Third, the whole suborbital industry sits in the disputed band, which means the marketing claim, the regulatory category and the record book can all disagree about the same flight.

The Honest Answer

Space does not begin. The atmosphere thins, the physics of flight gives way to the physics of orbits, the sky darkens, your blood boils, and the air gets sparse enough to count atom by atom — all at different altitudes.

The Kármán line is a good answer to a badly posed question. What von Kármán identified was real: a crossover between two regimes, computable from air density and orbital velocity, which lands around 84 km for a lightly loaded wing. What we built on top of it was a round number. Both facts are worth knowing, and only one of them is physics.

Sources and Further Reading

Frequently Asked Questions

Where does space begin?

There is no physical boundary, so the answer is a matter of convention and the conventions disagree. The most widely cited line is the Kármán line at 100 kilometres, adopted by the Federation Aeronautique Internationale, the body that certifies aviation and astronautics records. The United States uses 50 statute miles, which is 80.47 kilometres, as the altitude above which a pilot earns astronaut wings, a standard inherited from the X-15 programme in the 1960s. Jonathan McDowell argued in 2018 for 80 kilometres on the evidence of real satellite trajectories. The United Nations Committee on the Peaceful Uses of Outer Space has debated the question since 1966 without agreeing on anything. Physically, the atmosphere simply thins out: there is no altitude at which the density drops to zero, and Earth's hydrogen geocorona has been measured out to roughly 630,000 kilometres, well past the Moon.

What is the Kármán line?

The Kármán line is the altitude, conventionally 100 kilometres, at which the atmosphere becomes too thin for aerodynamic flight and spaceflight must take over. Theodore von Kármán reasoned that lift scales with air density and the square of airspeed, so the thinner the air the faster a vehicle must fly to hold itself up. Above a certain altitude the speed required for aerodynamic lift exceeds the orbital velocity of about 7.9 kilometres per second, at which point wings are pointless because centrifugal effects are already supporting the vehicle. That is the crossover from aeronautics to astronautics. The exact altitude depends on the aircraft: for a wing loading of about 32 kilograms per square metre and a lift coefficient near 1, the US Standard Atmosphere puts the crossover at 84 kilometres. For a heavy airliner it falls below 70 kilometres.

Why is the Kármán line at 100 km and not 84 km?

Because 100 is a round number in kilometres. Von Kármán's own rough calculation landed near 84 kilometres, which is 52 statute miles, and the figure survives mainly through Andrew Haley's later write-up because von Kármán never published the derivation himself. When the Federation Aeronautique Internationale needed a fixed, unambiguous altitude for records, 84 kilometres was awkward and model-dependent, while 100 kilometres was memorable, comfortably above any aircraft, and tidy in SI units. The physics does not single out 100 kilometres at all. In his 2018 reanalysis in Acta Astronautica, Jonathan McDowell examined roughly 43,000 catalogued satellites and found objects completing full revolutions with perigees well below 100 kilometres, which led him to propose 80 kilometres as the physically defensible boundary instead.

What is the Armstrong limit?

The Armstrong limit is the altitude at which the ambient air pressure falls to about 6.3 kilopascals, roughly 6% of sea-level pressure, which is the saturated vapour pressure of water at human body temperature. In the US Standard Atmosphere that happens at about 19.1 kilometres, and the figure is usually quoted as 18.9 to 19.35 kilometres depending on conditions. Above it, exposed body fluids including the water in saliva, tears and the lungs boil at 37 degrees Celsius. This is the altitude at which breathing pure oxygen stops being enough: below it a pressure-fed oxygen mask can keep you conscious, above it you need a sealed pressure suit or a pressurised cabin, because there is no partial pressure gradient left to drive oxygen into the blood. It is named after Harry Armstrong, the US Air Force physician who identified it in the 1930s.