A guest post by Roger Pielke, Jr., University of Colorado
The IPCC offers a number of statements expressing its confidence in the likelihood of various claims based on very explicit guidance that it prepared for conveying uncertainties to its readers. These statements are the subject of much confusion and debate. This post discusses the IPCC statements on attribution of increasing global temperatures to various causes, as reported in the IPCC Fourth Assessment Report from 2007.
First, let’s consider three possibilities (which are actually quite close to those presented here by my father on this blog not long ago).
A) Natural forces alone account for the observed warming
B) Natural and human forces together account for the observed warming
C) Human forces alone account for the observed warming
Not only are these three possibilities mutually exclusive, meaning that only one of them can be true, but logically one of them must be true, since they cover the range of possibilities. Thus if I were to ask you to express your confidence in the likelihood of each statement being correct, to be logically sound, then your expressed probabilities would need to add to 100%. (We could further refine this analysis by distinguishing natural forces into natural forcings and internal variability, and human forces into greenhouse gases and other forcings, but for the most part I’ll keep things simple.)
What does the IPCC (WG1 TS p. 60) say about these three logical possibilities for attribution? Unfortunately, the IPCC was not particularly systematic in its approach, but it did say a few things that are directly relevant and this post seeks to sort through these statements.
For A) the IPCC provided a statement that:
“Attribution studies show that it is very likely that these natural forcing factors alone cannot account for the observed warming”
With “very likely” meaning greater than 90% but less than 95% (the IPCC uses the term “extremely likely” to represent greater than 95% but less than 99%, and “virtually certain” to mean greater than 99%), this statement is logically equivalent to saying:
It is very likely that either B) or C) is true.
Thus the balance of the likelihood – 5% to 10% — must then fall on A). If the IPCC wanted to say something stronger about the likelihood of natural forcing factors alone accounting for the observed warming, it had a number of terms at the 95% or even 99% thresholds that it chose not to use.
The IPCC’s statement is thus also logically equivalent to saying:
It is very unlikely that these natural forcing factors alone can account for the observed warming.
In other words, the IPCC conveyed that there is between a 5% and 10% chance that the observed warming is due to natural factors alone. This is the logical implication of their statement.
Let’s consider B) next. The IPCC states:
“It is very likely that anthropogenic greenhouse gas increases caused most of the observed increase in global average temperatures since the mid-20th century.”
This statement has a number of qualifications, but it is predicated on the balance between greenhouse gas and other causes of the observed temperature increase. I take “most” to mean a majority of, which is certainly more than half but is not “all.” The IPCC could have said “nearly all” or “vast majority” or some other terminology, but chose “most.” Presumably if the IPCC had meant “all” it would have said “all” rather than “most.” So it is safe to say that this statement refers to the balance of causes between greenhouse gases and other possible causes of the observed temperature increase.
So this statement means that the IPCC is 90% to 95% confident that greenhouse gas increases (a human cause) resulted in a majority (e.g., 51% to 99%) of the temperature increase. The logical implication is that there is a 5% to 10% chance that the greenhouse gas increases have resulted in exactly half or a minority of the observed temperature increase. The other forces at work could of course be human or natural.
Interestingly, the IPCC does not appear to say anything explicitly in its likelihood findings about the possibility that the temperature increase is all human caused. I would have expected to see a statement like:
It is likely that all of the observed temperature increase is due to greenhouse gas emissions
But I can find nothing close, and presumably this is the case because separating out the exact signal remains a topic of ongoing research.
So here are the conclusions that the logical structure of the likelihood findings of the IPCC lead to:
- The IPCC finds between a 5% and 10% likelihood that the observed warming is due to natural forces alone.
- The IPCC finds a 5% to 10% chance that greenhouse gases are responsible for a minority of the temperature increase, leaving other human and natural forces to explain the balance.
Now let me say that these results are surprising to me. Given the venom that characterizes debates over climate change, and the assumption by many that C) above is the only acceptable view, I wonder if the IPCC simply constructed logically unsound statements, or if in fact uncertainties are as large as they imply. None of this changes my views on any of the policy options on mitigation or adaptation that I have advocated, as 90% to 95% likelihoods are plenty strong. But they do raise questions about strong claims to certainty and the demonization of skeptical points of view.