Wednesday, August 6, 2008

Global Climate Change

In the last issue of Skeptic magazine there are a series of articles on global warming. The most controversial is that of Dr. Patrick Frank, a chemist working for the Stanford Linear Accelerator Center (funded by US Dept of Energy). He contends that the models of global warming are all wrong, and conclude that the observed climate changes are not for certain due to man's activity. The trend though, in real world observations, are incontrovertible. Global temperature has gone up, concomitant with increased levels of CO2 emissions from industry.

The problem with Dr. Frank's conclusions:

Dear Editor (Skeptic):

I have read and re-read Dr. Frank’s article on the accuracy of global climate models (GCM) and most of it leaves me in the dust. I’m familiar enough with statistics to understand the point he is making about the accuracy of the models and will concede that he may have a point here. I won’t quibble about his analysis of the models and their accuracy – he likely has done his homework well enough to get this part right. I’ll leave any criticism of his analysis of the models to those who have skills and expertise in this area.

Where I will take issue with Frank is in his concluding remarks where he caricatures the skeptics of his conclusions about global warming in this way:

‘Some may decide to believe anyway, “We can’t prove it”, they might say, “but the correlation of CO2 with temperature is there (they’re both rising after all), and so the causality is there, too, even if we can’t prove it yet.”’

Frank is right in suggesting that some of us have noticed that man-made CO2 production has seen a precipitous increase since 1850. Since 1850 the amount of carbon we have put into the atmosphere per year has risen from insignificant amounts, through to two billion metric tons in 1950, to 8 billion tons in the latter part of the 20th century to present day. Furthermore, direct measurements of CO2 in the atmosphere show an increase from 312 ppm in 1958 to 375 ppm in 2005 (this is the so-called Keeling curve). Measurements of global mean temperature show an increase of 0.60 C ± 0.2o C in 20th century. (see These are the correlations in the data that give some of us pause.

Frank points out that it is an error in logic to go from correlation to causation – no problem here (we’re all taught this in Stats 101). But isn’t there another alternative in our application of scientific principles than jumping from correlation to causation? Those of us who notice the correlation just might suggest that we continue with the investigation. We might argue that the continued use of fossil fuels isn’t sustainable (we’ll simply run out) and we may be concerned that all of the pollution generated by fossil fuels may be a causal factor in some of the ecological events of recent decades (the bleaching of coral just one of many). If that were the case, couldn’t we act rationally and try an experiment that would meet the gold standard of science?

In psychology, the behaviourists taught us a really good technique to test a causal inference called the A-B-A design. In phase A of the experiment we measure behaviour at baseline. In phase B we institute the treatment and note any changes in behaviour. Finally, in the second phase A we withdraw the treatment and note whether the behaviour returns to baseline. That observed pattern leads to pretty strong statement of causality, but to be on the safe side we institute the treatment phase again and run another B. What is practical in the climate change scenario is to do an AB design. We already have the baseline data as noted above. Now we do the B part – run the treatment; namely reduce CO2 (William Calvin has some excellent suggestions on how this might be done in his book Global Fever). If things get better, and it will take time to get the data – these things take decades given the size of events we are talking about, a tentative hypothesis can be made about man’s involvement in global warming. If the results are negative – no effect on climate change, we have at least found cleaner energy that is less polluting and all those other toxins and particulates will have made our air and water safer.

To just simply deny that there is a possibility that climate science has some of the picture right and to continue as usual seems counterproductive and not very scientifically minded – we have the chance to conduct a wonderful experiment in living more sustainably; what could be more scientific than that?

Barry Cull

A better approach:

William Calvin in Global Fever points out that the direct and careful observations by Keeling at Mauna Loa in Hawaii have shown a steady rise in mean global temperatures since 1958. This isn't a "model", it's the real thing.

Treating the planet as a "patient" with a high fever, Dr. Calvin offers some viable strategies to first halt carbon emissions from tracking higher and second to begin decreasing man-made CO2. The cure isn't going to be easy, and the initial costs may be high, but Calvin shows that it is doable. Having just finished the book, of which the Skeptic article is just an excerpt, I suggest that Calvin has a better grasp of the subject matter, and has covered more of the findings than Frank has. My money is with Calvin in this debate.


No comments: