Encounter 22 · 22 of 42 · 16 May 1985
The hole they did not expect
In 1985 three names on a Nature page reported a spring collapse of ozone over Halley. The instrument had been reading that sky since the International Geophysical Year. The models had not advertised this shape.
Halley Research Station, Antarctica
What didn’t work
Waiting for a satellite picture before trusting the ground.
What they knew
- The Dobson numbers for Antarctic spring were far below the 1957–73 envelope.
- The instrument had been reading this sky for sixteen seasons of normals.
- Molina and Rowland had already, in 1974, shown that chlorofluorocarbons could destroy ozone. The shape of the damage was not the textbook gradient.
What they could not know
- Whether the Dobson was wrong.
- Whether anyone would believe a ground station against a satellite.
- What the spring numbers would do next year.
The situation
Halley, on the Brunt Ice Shelf, had run a Dobson spectrophotometer on Antarctic ozone since the International Geophysical Year, 1956–57. Years of October means, then a slide. Jonathan Shanklin’s office included checking and reducing that record. On 10 October 1983 he wrote NASA Wallops reporting about 200 Dobson units at Halley, asking whether a satellite could confirm. It was too early. Spring 1983 TOMS data were not yet processed. They were not willing to publish a glitch.
The thing is a number that refused to be smoothed. Joseph Farman, Brian Gardiner and Jonathan Shanklin were reading a Dobson spectrophotometer at Halley. Spring ozone over Antarctica had collapsed so far the reading looked like an instrument fault. They checked the instrument. They kept the number.
The ground, then the page.
- 1956–57Dobson spectrophotometer at Halley Bay, International Geophysical Year. October means begin.
- 10 October 1983Shanklin writes NASA Wallops: about 200 Dobson units at Halley. Can a satellite confirm. Too early. Spring 1983 TOMS not yet processed.
- 16 May 1985Nature 315: large losses of total ozone in Antarctica, seasonal ClOₓ/NOₓ. October ozone down about 30 percent against the old mean. Antarctic spring only.
1985
The page
Three names. A ground station. A number that did not look like the model. No picture from space in the paper.
Later
The filter
Nimbus-7 TOMS had been seeing the same low values. The algorithm’s standard profiles ran about 200–650 Dobson units. Readings below that range were flagged as bad instrument, not sky.P. K. Bhartia, later account of TOMS processing. NASA maps after the Nature paper. The Montreal Protocol is 1987. None of that is on the 1985 page.
They are not given a treaty. NASA is not said to have missed it on purpose. The filter was a calibration range.
What was at stake
If the Dobson is wrong, they have just told the world the sky is broken.
The next move
J. C. Farman, B. G. Gardiner and J. D. Shanklin published in Nature on 16 May 1985: large spring losses over Antarctica, and a pointer at chlorine chemistry. They did not title it a hole.
What they had
- A spectrophotometer
- Sixteen seasons of normals
- A wintering station
- A 1974 theory that did not predict this shape
What happened
Three names on a Nature page. Halley numbers against the old mean. No satellite map in the paper.
Three names on a page. The sky still unnamed by a satellite.
Private. No one else will see it.