Are CFCs Really the Villains?

Their relationship with ozone ...

A reader of my last post, asked to see a plot that shows the relationship between changes in the atmospheric concentration of CFCs and ozone. I guess he wanted to be convinced that CFCs are the cause of the problem.

The answer is yes.

Here’s my attempt to answer his specific question.

Note the reversed ozone minimum y-axis (blue). Satellite ozone data were incomplete in 1995, so those plot lines have been interpolated.

The plot’s essentially an update of the one shown in Joe Farman’s landmark paper back in 1985 that heralded the discovery of the Antarctic ozone hole. It showed that the increasingly large reductions in springtime ozone over their observation site at Halley Bay were in inverse proportion to the increasing concentrations of CFCs in the atmosphere over the mid 1970s to the mid 1980s.

My plot extends that work by making use of the broader geographical coverage available from satellite ozone data, and covering a period of about 50 years from the 1970s to 2020.

In charting the changes in atmospheric chlorine (shown as the black line), I didn’t restrict myself to just the effects of CFC-11. I also included CFC-12, which has a slightly larger contribution to ozone loss, and has an even longer atmospheric lifetime. Data sets for both are freely available from NOAA. Each molecule of CFC-11 (i.e., CCl3F) contains three chlorine atoms, and each molecule of CFC-12 (i.e., CCl2F2) contains two, so the axis labelled “chlorine from CFCs” is the weighted sum of those concentrations, (i.e., 3 x CFC-11 + 2 x CFC-12). Several other CFCs and their replacements also contribute to the problem, but these are the major two. The chart shows that the amount of chlorine available from these two gases doubled over the last two decades of the 20th century before starting to slowly decline.

To show the effects of those changes, I used satellite-measured ozone data that are freely available from NASA to chart the year to year variability in both the size (in red) and depth (in blue) of the Antarctic ozone hole. The axes for these are at the right, with a reversed scale for the ozone minimum, which represents the depth of the hole. The changes are far from trivial. In the last two decades of the 20th century, the area of the springtime ozone hole area increased from zero to more than 25 million square kilometres (that’s 2.5 times larger than the USA), and the minimum ozone column amount was at times reduced from its undisturbed value of around 200 DU down to less than 100 DU. More than half of the ozone was being destroyed. Since the turn of the century there seems to have been a slight improvement.

As you can see when the curves are overlaid in the same plot, there are strong correlations (with a correlation R-value coefficients of about 0.9 in each case) between the amount of chlorine from CFCs and these two metrics for ozone depletion. No doubt the correlations would have been stronger if the chlorine (and bromine) concentrations of those other more minor ozone depleting gases had been included, but you get the picture. They could also have been slightly stronger if I’d made allowances for the lag between these concentrations of CFCs measured in the atmosphere, and the subsequent effect on chlorine in the stratosphere. That lag can be several years.

You get similar correlations if you consider the mean size and depth of the ozone hole over the period they are worst (rather than the extremes), as shown below.

Despite the strong anti-correlation between atmospheric chlorine and the severity of the ozone hole, there are significant year-to-year departures due to year-to-year differences in wind patterns, which influence the air’s exposure to sunlight, and therefore its temperatures. As mentioned in an earlier post, very low air temperatures are a prerequisite for the rapid ozone depletion that occurs in polar regions. For example, in 2019 there was less ozone depletion in the disturbed wind flow, but in 2020 the ozone hole was much deeper and more persistent because of the stable circumpolar wind patterns.

It would be interesting to see a similar plot for the global mass of atmospheric ozone versus atmospheric chlorine. But that’s a bit more work. Perhaps a task for one of my colleagues? Or perhaps they can give me a lucrative contract to do it myself 😊? The overall changes would be much smaller, but the signal-to-noise ratio might be improved because wind effects will be smaller (though volcanic influences will be larger).

Of course correlation doesn’t prove causation, but it looks pretty convincing to me. When chlorine increases, ozone is destroyed; and when chlorine decreases, ozone recovers. And the observed relationship so far is pretty much in line with results from model calculations. The models can therefore be used with confidence to predict future changes in ozone, as shown in the answer to Question 20 of the United Nations Ozone Secretariat’s “Twenty Questions and Answers” (provided we know the real production rate of the ozone damaging chemicals).

Aside: More on the Historical Perspective

Joe Farman’s original plot, redrawn below from his 1985 paper in Nature, shows that there was a threshold effect. Large ozone reductions began only after 1970 when the concentrations of chlorine in the atmosphere passed a threshold value of about 1 ppb (or 1000 ppt). Note the reversed y-axis scale for the CFCs (CFC-11, and CC-12, labelled here as F11 and F12).

---

Thanks for listening. Previous posts on the intersection between Ozone, UV, Climate, and Health can be found at my UV & You area on Substack. Click below to subscribe for occasional free updates.