The atmosphere and the layers of the sky

Above the curved horizon of the Earth, a thin luminous band of orange, then pink and blue, separates the dark clouds from the black of space. — Sunset over the Indian Ocean seen from the International Space Station: viewed edge-on, the atmosphere shows its layers, orange at the bottom (troposphere), pink and white above (stratosphere), then blue up to the black of space. ISS Expedition 23 crew, Public domain, source

A man wanted to reassure the public. He compared his measurements for the year with those from twenty years earlier, expecting them to be identical.

They were not.

That is how the hole in the ozone layer was discovered.

The station

At Halley, in Antarctica, the amount of ozone above the station has been measured since 1956.¹

Nobody was looking for anything in particular. It was a matter of monitoring: reading, recording, filing. Year after year, with a ground-based instrument.²

This kind of work is regarded as boring. Yet what it produced is worth remembering.

How it began, in his own words

Jon Shanklin himself recounts the starting point.

An open day was approaching, at which members of the public would come and ask questions. At the time, there was a great deal of concern about aerosols and supersonic aircraft, which were suspected of damaging the ozone layer.

So he took the year’s readings and compared them with those from twenty years earlier, expecting to find them identical — so that he could tell visitors there was no cause for concern.³

They were not identical.

So he went back through the whole backlog of data, to check whether that year was an exception.

It was not.

The real thickness: the skin of an apple?

This comparison is very often made: “If the Earth were an apple, the atmosphere would be as thick as its skin.”

Before using a vivid metaphor, one must check that the ratio is accurate.

The Earth’s mean radius is about 6,371 kilometres.
99% of the total mass of the air is concentrated in the first 30 kilometres.
This dense core represents 0.47% of the Earth’s radius.

For a standard apple 7.5 centimetres in diameter (37.5 millimetres in radius), 0.47% corresponds to a film 0.17 millimetre thick. That is exactly the average thickness of the peel.

The metaphor is scientifically accurate, provided one is talking about the breathable mass and not the tenuous traces of air that extend up to several thousand kilometres into the exosphere.

The architecture of the sky: the layers of the atmosphere

The atmosphere is not a homogeneous block. It is structured in superimposed layers, defined mainly by the variations in its temperature as one goes higher.

Troposphere (from 0 to 8–15 km) — This is the layer in which we live. It contains about 80% of the total mass of the air and almost all the water vapour. The temperature falls with altitude (down to −55 °C). This is where all weather phenomena take place (clouds, rain, storms).
Stratosphere (from 12–15 to 50 km) — Here the trend reverses: the temperature rises again to about 0 °C. Why? Because the ozone layer is located there (between 15 and 35 km) and absorbs the Sun’s ultraviolet rays, converting this energy into heat.
Mesosphere (from 50 to 85 km) — The temperature plunges again, reaching the coldest places on our planet (down to −90 °C). This is the layer in which most meteors disintegrate, forming “shooting stars”.
Thermosphere (from 85 to 600 km) — The temperature soars here (sometimes exceeding 1,500 °C), heated directly by solar X-rays and UV rays. But the density of the air is so low that an ordinary thermometer would measure nothing at all: the heat is not transmitted by contact. This is the realm of the polar auroras and of the International Space Station.
Exosphere (beyond 600 km) — The blurred boundary with space. Atoms gradually escape from it into the interplanetary void.⁴

Where does space begin?

At an altitude of a hundred kilometres: that is the usual answer. This is the Kármán line.

This number is a rounded figure, and its history is worth telling.

The original calculation did not give a hundred

Theodore von Kármán had sought the altitude above which an aircraft would have to fly so fast to support itself on the air that it would be in orbit anyway. In other words: the altitude at which flying ceases to make sense.⁵

His calculation gave about 84 kilometres.

A hundred is rounder, more convenient, easier to memorise. A hundred is what stuck.

A delightful detail: the line bears the name of a man who had calculated something else. What he established was an altitude of aerodynamic transition — not an absolute boundary of space.

Not everyone agrees

In 2005, the American space agency recognised as astronauts the pilots who had climbed above 80 kilometres.⁶

An article published in 2018 in an astronautics journal revisits the problem and proposes setting the limit at 80 kilometres.⁷

Its argument is worth knowing, because it rests not on convenience but on two measurable things:

By calculating the altitude at which a satellite can no longer stay up, one finds a value that remains within ten kilometres of 80, whatever the solar and atmospheric conditions. The author writes that this altitude lies between 70 and 90 kilometres, and that a hundred is always too high.
Moreover, 80 kilometres corresponds to something real: the mesopause, the boundary between two layers of the atmosphere — the one discussed above.⁸

In other words, one round number is not being replaced by another round number. A convenient number is being replaced by a number that corresponds to a phenomenon.

The international federation that certifies aeronautical records has been examining this change since 2018. As of 2026, it has still not formally adopted it.

A very concrete consequence

Five pilots of an experimental rocket plane (the X-15) flew between 80 and 100 kilometres in the 1960s.

Depending on the definition adopted, they went into space — or not.

The author of the article puts it bluntly: if a hundred kilometres is used instead of eighty, these five men lose their status as space travellers.⁹

It was not until 2005 that NASA awarded astronaut wings to three civilian X-15 pilots, two of whom were already dead.

It is not the flights, the men or the altitude that change. It is a line drawn by humans, at a place they chose.

What to take away

There is no physical boundary between the atmosphere and space. The air thins out gradually, without ever stopping abruptly — traces of it remain well beyond a hundred kilometres.

Any limit is therefore a convention. This is not a flaw: one is needed, for treaties, records and space law.

But one must be aware that it is one. To read “he went into space” is also to read “according to the definition that was chosen”.

What they found

From the late 1970s onwards, a systematic decline in ozone in the austral spring.

In 1984, the layer above Halley was only about two thirds of its thickness in previous decades — falls of about thirty per cent in October.¹⁰

The paper was published on 16 May 1985.¹¹

Satellites then confirmed it: the phenomenon covers the entire continent and often reaches twenty-four million square kilometres. By around 1992, the area affected was equivalent to that of North America.¹²

A nuance that keeps the story from being too neat

One could tell this as a confirmed prediction: chemists had announced as early as 1974 that certain industrial gases would destroy ozone, and this was observed.

That is not what happened.

What was observed was not what had been predicted. The magnitude, the time of onset, the region concerned and the seasonal character were unexpected in every respect. No study had foreseen a sensitivity to chlorine in the lower Antarctic atmosphere — nobody was looking there.¹³

So it was not a risky prediction that succeeded.

It was long-term monitoring that found something other than what was being looked for.

It is the same lesson as the eclipses recorded by the astronomers of Babylon nearly three thousand years ago, which today serve to measure the slowing of the Earth: data collected for one reason serve another, decades or centuries later.

What followed

The Vienna Convention in 1985. Then the Montreal Protocol in 1987, which organises the phase-out of the production of the substances concerned (chlorofluorocarbons, or CFCs).¹⁴

A 2015 analysis, carried out by the firm Deloitte for British research, estimates that without this discovery, the agreement could have been delayed by five to ten years. The United Nations Environment Programme estimates that thanks to the agreement, up to two million cases of skin cancer will be avoided each year by 2030.¹⁵

⚠ These two figures are estimates of what did not happen. They are given with their author and their date, and are not treated as direct measurements.

What came next — the measured recovery, and what did not work — is told in our page on protecting the planet.

The same gas, and its opposite

A distinction must be added, because it changes everything: ozone is good or bad depending on its altitude.

High up, in the stratosphere, it forms the shield that stops ultraviolet rays. This is the one discussed above.
Down low, at ground level, it is a corrosive pollutant, dangerous for the respiratory tract and for vegetation.

So it is not only the distant air that is at stake: the air we breathe is too. In 1952, London’s Great Smog, a fog caused by coal, killed thousands of people: 3,000 to 4,000 according to the long-accepted toll, about 12,000 if the following months are also counted. It was the event that triggered modern health regulations: the British Clean Air Act of 1956, then its American counterpart, broadened in 1970, which sets thresholds for six major pollutants — sulphur dioxide, particulate matter, carbon monoxide, ground-level ozone, nitrogen dioxide and lead.¹⁶

Note 1

Measurements by the Dobson ozone spectrophotometer at the British Antarctic Survey's (BAS) Halley research station form a record of Antarctic total column ozone that dates back to 1956.

Zhang, Lily N., Solomon, Susan, Stone, Kane A., Shanklin, Jonathan D. et al., “On the Use of Satellite Observations to Fill Gaps in the Halley Station Total Ozone Record”, Atmospheric Chemistry and Physics, vol. 21, 2021, p. 9829-9838