Antarctic ozone hole

Why does ozone loss peak over Antarctica each spring when the chemicals responsible were released elsewhere?

Nobody manufactured refrigerants in Antarctica. The chlorofluorocarbons that caused the damage came out of factories, air conditioners and aerosol cans in the industrial north. Yet the loss appeared over the emptiest continent on Earth, and it appeared in a narrow window — September and October — not in the depth of the polar winter and not in summer.

That combination is the interesting thing. A pollutant spread evenly over the globe produced an effect concentrated in one place, at one time of year. So the chemical cannot be the whole explanation. Something about where and when must be supplying the rest.

Start with why the chemical reaches the stratosphere at all. CFCs are prized precisely for being inert: they do not burn, react with little, and are not scrubbed out by rain. A gas that nothing in the lower atmosphere can destroy survives long enough to be mixed everywhere and to drift upward, and only above the ozone layer does short-wave ultraviolet finally break it apart, releasing chlorine. So the release point is irrelevant by the time the chlorine exists — the geography of the hole is already not the geography of the emissions.

Now the puzzle sharpens. If chlorine is spread through the whole stratosphere, why is the loss not spread too? Because most of that chlorine is not doing anything. It sits bound up in two comparatively unreactive molecules — hydrogen chloride and chlorine nitrate — which chemists call reservoirs. Chlorine locked in a reservoir does not attack ozone. So the real question becomes: what unlocks it?

Here is the step that surprised the chemists of the time. The unlocking does not happen in the gas at all; it happens on surfaces. In the Antarctic winter a ring of fierce westerly winds — the polar vortex — closes around the continent and seals its air off from the rest of the world. Sealed in darkness, that air gets extraordinarily cold, below about -78 degrees Celsius, and at those temperatures thin clouds of nitric acid and water particles form in a stratosphere normally far too dry for cloud. On the surfaces of those particles the two reservoirs react with each other, and chlorine comes off as a form that sunlight can split with ease. The nitric acid stays behind in the particle — and if the particles sink, that nitrogen is removed altogether, which matters because nitrogen oxides are what would otherwise re-lock the chlorine.

Through the whole dark winter, then, the atmosphere has been quietly loading a spring: the chlorine is primed and the brake is off, but nothing happens, because the reaction that destroys ozone needs photons. Then the sun comes back. In September the returning light splits the primed chlorine into free atoms, and each atom enters a catalytic cycle — it destroys ozone, is regenerated, and destroys again, thousands of times over before anything finally retires it. That is why loss is not merely fast but disproportionate to the amount of chlorine present. The cycles that dominate here are peculiar to these conditions: one requires two chlorine monoxide molecules to meet and pair up, which only happens when their concentration is high, and another pairs chlorine with bromine. Both need cold, and both need light. Cold without light gives you a primed vortex and no loss. Light without cold gives you the ordinary stratosphere everywhere else.

So the answer to "why spring" is that spring is the only time both conditions hold at once. And "why Antarctica" is that only the southern vortex is reliably cold and tight enough to prime the chemistry — the Arctic's is warmer and more easily disturbed by the land beneath it, which is why the north shows sporadic bad years rather than an annual hole.

Late in spring the vortex weakens and breaks up, ozone-rich air floods in from mid-latitudes, and the hole closes. It was never a hole in the sense of a puncture; it is a seasonal deficit in the total ozone overhead.

Two refinements are worth holding onto. First, "chlorine destroys ozone" is true but incomplete in a way that misleads: the availability of chlorine is not the switch. The switch is meteorological — how cold the vortex got, how long it stayed sealed, how much nitrogen sank out. That is why the hole's depth and area swing so widely from year to year while the chlorine burden changes only slowly.

Second, and this is where honesty is required about evidence: that same variability made the trend genuinely hard to see. The British Antarctic Survey found the decline in ground measurements at Halley Bay in 1985; satellites had been recording it too, but the low values had been treated as instrument error. Distinguishing a slow chemical trend from ordinary polar weather took a record decades long, and it still does — a year with a sudden stratospheric warming that splits the vortex early produces a small hole for reasons that have nothing to do with recovery, and 2019 was such a year. Claims that the hole is healing rest on looking at a consistent point in the season across many years and controlling for those dynamics, not on any single season.

The strongest test of the mechanism was not statistical at all. If the chemistry above is right, then inside the vortex the reactive form of chlorine must be abundant exactly where ozone is missing. Aircraft flying across the vortex edge in 1987 measured precisely that anticorrelation — chlorine monoxide rising as ozone fell, across the same boundary. Had ozone been depleted where reactive chlorine was not elevated, or had loss appeared in the midwinter dark, the account would have been refuted.

Recovery is correspondingly unglamorous. Emissions largely stopped after the Montreal Protocol, but CFC molecules already aloft have lifetimes of many decades, so the chlorine burden falls slowly. Assessments put a return to earlier ozone levels around the middle of this century — I would not quote a single year, because the estimate moves with each assessment and with assumptions about compliance.

Every term the collection defines is gathered in the glossary.