The Verification Venue · the good news, checked
The Hole We Talked Shut
Is the ozone layer healing? Yes — and the honest version is stranger and better than the headline. It heals because of a chemical loophole in the villain, and a treaty with a rare kind of teeth: the atmosphere itself keeps the books.
Start with the thing the school poster gets almost right. A chlorofluorocarbon drifts up, sunlight cracks a chlorine atom off it, and that atom eats ozone. What the poster misses is the word that matters: catalytically. The chlorine is not used up. It destroys an ozone molecule, hands the wreckage on, and comes back for more — thousands upon thousands of times. Press the button and watch one atom do it. Then find the switch that turns the whole thing off.
Ozone destroyed
0
molecules, by this one loop
Oxygen (O₂) made
0
net 2 O₃ → 3 O₂ each turn
Chlorine used up
0
the catalyst is never spent
RUNNING — at −81 °C the dimer Cl₂O₂ holds together long enough for sunlight to split it.
The Antarctic polar vortex hits ~185–195 K in late winter. Drag toward mid-latitude warmth (~230 K) and watch the loop stall.
Two gates must both be open for ozone to fall: it must be cold enough for the Cl₂O₂ dimer to survive, and there must be sunlight to split it. Antarctic spring is the only place both open at once.
The counter climbs and the "chlorine used up" box stays stuck on zero. That is the whole point, and it is the true thing you can see in ten seconds: ozone loss is catalytic. One atom, not one collision. The polar version — the cycle that actually digs the hole — pairs two chlorine-monoxide molecules into a fragile dimer and runs like this, returning both chlorines at the end:
The switch is temperature
Now the payoff, and the reason the hole is Antarctic, seasonal and springtime rather than a uniform global thinning. The dimer Cl₂O₂ is the load-bearing part, and it is thermally fragile. How much of the active chlorine sits as the dimer — the form that, once sunlight hits it, destroys ozone — is set by a measured equilibrium constant that plunges as the air warms:
Below, the fraction of chlorine tied up as the dimer, computed live at your slider temperature. At −81 °C it is nearly all dimer and the loop screams along. Warm the air by forty degrees and the dimer shakes itself apart faster than sunlight can split it — the fraction collapses toward zero and the cycle stalls, even with the sun full on. That temperature switch, not chlorine abundance, is why the ozone hole lives at the coldest pole in its coldest, sunlit season and nowhere else.
The atmosphere is the audit ledger
Here is the sophisticated shrug: fine, we banned a chemical — that's just policy trivia. It isn't, and the reason is the most interesting part of the whole story. The Montreal Protocol (1987) has teeth because you cannot hide a banned gas: it mixes into the global air, and the global air is measured. The treaty's enforcement mechanism is a concentration curve.
The comeback that got caught by the curve
CFC-11 was banned, and its concentration was falling on schedule — until, around 2013, the decline abruptly slowed by roughly half. No inspection found this. NOAA's monitoring network found it, in the shape of the curve alone (Montzka et al., 2018), and follow-up work traced much of the excess to new, illegal production in eastern China.
Then the story does the thing good-news stories rarely do: it resolves. Enforcement followed, and the curve bent back. Global CFC-11 emissions fell by 18 ± 6 gigagrams a year (about 26%) from 2018 to 2019, back to their earlier trend (Park et al., 2021). A molecule, a treaty and a measurement, closing a loop.
That is the second layer of the good news: not that people are virtuous, but that this one environmental problem happens to sit inside a rare closed feedback loop between molecule, measurement and law. The ledger cannot be bribed.
- 1974 Molina & Rowland publish the CFC–ozone hypothesis in Nature. The Nobel comes later — 1995, not 1974.
- 1985 Farman, Gardiner & Shanklin report October ozone over Halley, Antarctica, has fallen to about two-thirds of earlier-decade values. The hole is discovered.
- 1987 The Montreal Protocol is signed — the phase-out of CFCs begins.
- 1995 Nobel Prize in Chemistry to Crutzen, Molina and Rowland for atmospheric ozone chemistry.
- 2013–18 CFC-11's decline stalls; NOAA catches it from the curve; illegal production is traced and stopped.
- 2020 & 2023 Two of the largest, deepest holes on record — the 2023 hole ranked 12th largest. Recovery is not monotonic.
- ~2040 Near-global (60°S–60°N) ozone projected to return to its 1980 value — if compliance holds.
- ~2045 Arctic springtime ozone, projected return to 1980.
- ~2066 Antarctic springtime ozone, projected return to 1980. This — not "the hole is gone" — is what "healed by 2066" means.
The world we didn't get
What the treaty bought is easiest to see in the simulation of the world where it never happened. Newman et al. (2009) ran the atmosphere forward without the ban:
global ozone versus the 1970s, by 2065 — falling below ~110 Dobson units
surface UV, doubling — and up to 10,000× at the shortest wavelengths
to sunburn in a mid-latitude city like Washington, D.C.
That is the avoided world. (Note the honest boundary: a much-shared "+4 °C of warming" figure is not in the Newman paper — those are the numbers it actually gives.)
The check — every number recomputed in front of you
The dimer fraction is computed live from the measured equilibrium constant Keq(T) = 3.61×10⁻²⁷·exp(8167/T) (von Hobe et al., 2005), at a representative activated-chlorine level of [ClO] ≈ 1.9×10⁹ cm⁻³ (~1 ppbv). The highlighted row is your current slider temperature. The switch is not asserted — it falls out of the arithmetic.
| T (K) | T (°C) | Keq (cm³) | [Cl₂O₂]/[ClO] | dimer fraction | cycle |
|---|
The 50% switch temperature, solved live from the same equation: T½ ≈ … K. And the three cited figures, each re-derived rather than repeated:
The per-turn stoichiometry (2 O₃ → 3 O₂, chlorine returned) and the temperature gate are exact; the on-screen pace is illustrative. Run it yourself: node research/the-hole-that-is-healing/verify.mjs.
What's exactly true, what's idealised, and where the popular story overruns the record
Exactly true
Ozone loss is catalytic: a single chlorine atom destroys on the order of 100,000 (10⁵) ozone molecules before it is locked into a reservoir species (HCl or ClONO₂) and stops — the EPA and NASA figure. The polar cycle's net reaction is 2 O₃ → 3 O₂ with both chlorines handed back. The dimer Cl₂O₂ is thermally fragile, so the ozone-killing cycle is temperature-and-light gated — which is why the hole is Antarctic, seasonal and springtime. The equilibrium constant, the dimer fractions and the 50% switch temperature on this page are computed straight from Keq(T)=3.61×10⁻²⁷·exp(8167/T).
Idealised, and the free choices we named
The counter's pace is illustrative, not a real turnover rate. What is exact is the per-turn stoichiometry and the fact that the catalyst is never consumed. The "~100,000 per atom" is an order-of-magnitude lifetime turnover, not a precise constant or a law — it ends the instant the chlorine is sequestered. We hold the activated chlorine fixed at a representative [ClO] ≈ 1.9×10⁹ cm⁻³ (~1 ppbv at ~50 hPa); the real value varies with altitude and how much reservoir chlorine the polar stratospheric clouds have already activated. We show one cycle (the ClO dimer); the real hole also runs a ClO+BrO cycle, and bromine is far more efficient per atom. The equilibrium constant itself is genuinely debated: the JPL recommendation and the value used here (von Hobe 2005) disagree at the coldest temperatures, and that disagreement is an open problem in the literature — we used one sourced value and named it.
Where the popular story overruns the record
"Healed by 2066" is not "the hole is gone." It means Antarctic springtime total-column ozone is projected to return to its 1980 value around 2066 (Arctic ~2045; near-global ~2040), if Montreal Protocol compliance continues — it does not mean the hole shrinks every year. It doesn't: 2020 and 2023 produced among the largest, deepest holes on record (2023 ranked 12th largest). Year-to-year recovery attribution carries real scientific uncertainty because volcanoes and wildfires perturb the signal — the 2020 Australian wildfires and the January 2022 Hunga Tonga eruption (which injected water vapour into the stratosphere) both worsened later holes.
The Nobel was 1995, not 1974, and it was shared with Paul Crutzen — the 1974 date is the Molina–Rowland Nature paper, not the prize. And the widely-shared "+4 °C of avoided warming" figure is not in the Newman 2009 paper; that paper gives the 67%-by-2065 ozone loss, the UV doubling and the five-minute sunburn — so those are the only Newman numbers this page cites.
The hazard: distinguish, do not sever
The ozone hole and climate change are different problems — CFCs versus CO₂, stratospheric chemistry versus radiative forcing — and it is a real error to blame the ozone hole for global warming. But the tempting overcorrection, "the ozone hole has nothing to do with climate," is also false. CFCs are potent greenhouse gases; so are the HFCs that replaced them, which is why the 2016 Kigali Amendment phases HFCs down and is projected to avoid up to 0.5 °C of warming by 2100. And Hunga Tonga's stratospheric water vapour couples the two. So: distinguish them, but do not sever them.