2010-09-26 00:15:35Blog post on probabilities and whether the IPCC language is alarmist
James Wight


Reading about the way people perceive probabilities gave me an idea for a blog post. Depending on when it is published it could tie in with Graham’s rebuttal of “IPCC is alarmist” (and also the ASC post, come to think of it). Now I've finished writing it, though, I'm wondering if my point is too trivial to write a whole post on it. Let me know what you think.

Title: Do the IPCC use alarmist language?

Graham Wayne has recently written rebuttals to “The IPCC consensus is phony” and “IPCC is alarmist”. But, you might say, that’s only half the story – do the IPCC present their conclusions in an alarmist way? There are many different ways you might look at this, but one of the more important ones is how the IPCC present probabilities (or “likelihoods”).

Thinking about probability does not come intuitively to the human mind. Our assessment of a risk often depends on how the probability is presented.

Suppose you are about to get on a plane and a reliable source tells you that there is a 1% chance that the plane will crash during your flight. Do you still want to get on the plane? I’m guessing you’d be having second thoughts about it.

What if the probability of a crash is 1 in 20? 1 in 10? 1 in 3? You’d probably run away screaming.

I’ll get to the point of all this shortly, but please bear with me and consider the following quote from the IPCC’s Fourth Assessment Report (AR4):

“It is very unlikely that [Atlantic Ocean circulation] will undergo a large abrupt transition during the 21st century.” [Source]

Are you alarmed yet? Is this an example of the IPCC using alarmist language in reporting its conclusions?

To answer this question, you first have to understand what the IPCC are trying to say. In the introduction to the AR4 Synthesis Report, there is a detailed description of the treatment of uncertainty in IPCC reports, and I don’t think the public appreciates just how un-alarmist it is. A 1% chance scarcely rates a mention: anything with such a low probability is described as “exceptionally unlikely”. A probability of 1 in 20 is considered to be “extremely unlikely”; 1 in 10 is “very unlikely”, and even 1 in 3 is still “unlikely”. Conversely, 2 in 3 is “likely”, 9 in 10 is “very likely”, 19 in 20 is “extremely likely”, and 99% is “virtually certain”.

So if you asked the IPCC to do a report on your plane trip, and the probability of a crash was smaller than 1 in 10, about half a decade later they’d get back to you with something like: “It is very unlikely that this plane will crash.” (Except that it would probably be a lot wordier than that.)

And when the IPCC says an abrupt transition in Atlantic Ocean circulation is “very unlikely”, they mean the same thing: the chance is less than 1 in 10. Yet you’re probably not running away screaming.

Most of the IPCC’s main conclusions are given a high degree of likelihood. Probably the most quoted sentence from the entire AR4 is:

“Most of the observed increase in global average temperatures since the mid-20th century is very likely due to the observed increase in anthropogenic greenhouse gas concentrations.” [Source]

Translation: the likelihood that humanity’s greenhouse gas emissions are causing global warming is greater than 9 in 10.

Another important conclusion (though not particularly new to the AR4) is this:

“[T]he equilibrium global mean [surface air temperature] warming for a doubling of atmospheric carbon dioxide (CO2), or ‘equilibrium climate sensitivity’, is likely to lie in the range 2°C to 4.5°C, with a most likely value of about 3°C. Equilibrium climate sensitivity is very likely larger than 1.5°C.” [Source]

Translation: the chances are 2 out of 3 that a doubling of atmospheric CO2 will warm the planet by between 2 and 4.5 degrees; 9 out of 10 that it will be more than 1.5 degrees.

One more IPCC quote:

“It is very likely that hot extremes, heat waves and heavy precipitation events will continue to become more frequent.” [Source]

That is, the chances of more extreme weather are higher than 9 in 10. If you’re thinking that wilder weather is not exactly as serious as a plane crash, then consider that over 20 million people have been affected by the 2010 Pakistan floods. This sort of extreme weather event will become more frequent with global warming. Do we, does humanity, really want to get on this plane?

The IPCC are not alarmist in their conclusions, and they are no more alarmist in the way they report their conclusions.

2010-10-01 17:50:28Bump
James Wight

This thread is languishing.
2010-10-13 21:23:37Sorry for the delayed response
John Cook

I like what you've done here, James, translating the sometimes opaque language of the IPCC into terms the average person can understand. If you add this as a blog post via Author Admin, let me know and I'll go live with it.
2010-10-14 00:35:45Okay, I've done that
James Wight

I have just added it to Author Admin.
2010-11-07 07:08:29alarmist language


I like this very much, and think something like it needs to be posted, or even published in some major media. Almost nobody I meet understands odds at all, and trying to explain trend lines and scatter graphs nearly impossible.


i'm agonna give ie a green thumbs up.