by Judith Curry
Seeking once again to clarify the problems in communicating the IPCC climate change attribution statements.
Context
The immediate motivation for this post is a tweet from Gavin Schmidt that he is #stillwaiting for a response to his critique of my 50-50 essay [link]. Well this post is a response to only one point that he raises (some of the rest of his points seem pretty incoherent to me), but it is an issue that has been used by Schmidt to discredit my arguments about attribution.
The main conclusion of the IPCC AR4 was:
“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.”
In my Uncertainty Monster paper, I criticized the IPCC statement for the ambiguity of the word ‘most‘. Nowhere in the AR4 report is this clarified, but Hegerl et al., in their response to the uncertainty monster paper, state
The likelihood describes the assessed probability that ‘most’, i.e. more than 50%, of the warming is due to the increase in greenhouse gases. This statement has a clear meaning and an associated uncertainty, although explicitly listing ‘>50%’ in the text to ensure that no misunderstandings are possible could be helpful in future work.
Well, here is how the AR5 states it:
It is extremely likely that more than half of the observed increase in global average surface temperature from 1951 to 2010 was caused by the anthropogenic increase in greenhouse gas concentrations and other anthropogenic forcings together. The best estimate of the human induced contribution to warming is similar to the observed warming over this period.
Hmm . . . the AR5 didn’t use ‘>50%‘, but rather elected to use ‘more than half‘.
One problem with the IPCC’s attribution statement illustrated by this statement from a document Lost in Translation: Closing the Gap Between Climate Science and National Security Policy, published by the Center for a New American Security (which I also cited in the Uncertainty Monster paper):
“For the past 20 years, scientists have been content to ask simply whether most of the observed warming was caused by human activities. But is the percentage closer to 51 percent or to 99 percent? This question has not generated a great deal of discussion within the scientific community, perhaps because it is not critical to further progress in understanding the climate system. In the policy arena, however, this question is asked often and largely goes unanswered.”
This statement was written in response to the AR4 statement; the AR5 statement has arguably added some precision with its words The best estimate of the human induced contribution to warming is similar to the observed warming over this period.
Help from the dictionary
For reference:
Most: greatest in amount or degree; the majority of
Half: one of two equal or approximately parts of a divisible whole, as an object, or unit of measure or time; a part of a whole equal to the remainder.
Legal definition of more than half: majority
Majority: A majority is a subset of a set consisting of more than half of the set’s elements. This can be compared to a plurality, which is a subset larger than any other subset considered; i.e. a plurality is not necessarily a majority as the largest subset considered may consist of less than half the set’s elements.
I did not find a specific definition for ‘greater than 50%‘, so lets look at these definitions:
Percent: out of each hundred; per hundred; one part in a hundred.
Percentage: In mathematics, a percentage is a number or ratio expressed as a fraction of 100. Percentages are used to express how large or small one quantity is relative to another quantity. While percentage values are often between 0 and 100 there is no restriction and one may, for example, refer to 111% or −35%.
And while we’re at it, we need one more definition:
Predominant: present as the strongest or main element. Synonyms: main, most important, foremost, key, paramount
Gavin’s critique
I started my 50-50 essay with this:
Pick one:
a) Warming since 1950 is predominantly (more than 50%) caused by humans.
b) Warming since 1950 is predominantly caused by natural processes.
Gavin states in his critique:
Here Judith makes the same mistake that I commented on in my 2012 post – assuming that a statement about where the bulk of the pdf lies is a statement about where it’s mean is and that it must be cut off at some value (whether it is 99% or 100%). Neither of those things follow. I will gloss over the completely unnecessary confusion of the meaning of the word ‘most’ (again thoroughly discussed in 2012). I will also not get into policy implications since the question itself is purely a scientific one.
To understand the critique in Gavin’s 2nd paragraph above, it is instructive to look at John Nielsen-Gammon’s essay Your Logic Escapes Me, which is discussed further in my post The logic(?) of the IPCC’s attribution statement:
It can be a bit misleading to express this in terms of percentages. When most people see percentages, they imagine small positive numbers that collectively add up to 100%. However, different agents of climate change can have positive (example, increasing greenhouse gases) and negative (increasing aerosols) contributions. Pick a random time interval, and natural variability is just as likely to make a negative contribution as a positive one.
Curry’s assumption that the IPCC means ‘most’ in this context to cover a range of 51-95% is flat-out wrong. Here’s the full quote from the Summary for Policymakers: “It is extremely likely that more than half of the observed increase in global average surface temperature from 1951 to 2010 was caused by the anthropogenic increase in greenhouse gas concentrations and other anthropogenic forcings together. The best estimate of the human-induced contribution to warming is similar to the observed warming over this period.” They explicitly state that their best estimate for the human-induced contribution is about 100%, which is outside the range that Curry assumes they mean!
If you’re going to impute a range for them, at least have their most likely value somewhere near the midpoint of the range. For the sake of argument, a reasonable range would be 51-135%.
The figure that Gavin and JN-G refer to is:
The probability density function for the fraction of warming attributable to human activity (derived from Fig. 10.5 in IPCC AR5). The bulk of the probability is far to the right of the “50%” line, and the peak is around 110%.
Semantics
Until this exchange, it never occurred to me that the IPCC’s attribution statement was attempting to convey AGW attribution that was possibly outside the range of 0 to 100% (and apparently it didn’t occur to the Center for a New Security, either). ‘Most’ used in the AR4 wouldn’t necessarily preclude an interpretation of AGW attribution that was greater than 100%, but it in the common understanding of the word ‘most’, most people would interpret this to be some number that did not exceed 100%.
However, the use of ‘more than half’ in the AR5 attribution statement, to infer the the possibility of AGW attribution that exceeded 100%, violates any conceivable understanding of the word ‘half’. If you are interpreting ‘percent’ as something between 0 and 100%, then ‘more than half’ is equivalent to ‘greater than 50%’. However, if you interpret ‘>50%’ to allow for numbers >100%, then ‘>50%’ is not equivalent to more than half.
So what did the IPCC intend by its statement ‘more than half’? If we take Gavin’s word for it, then the IPCC has attempted to communicate this with very poor semantics.
If the IPCC does really mean ‘more than half’, but limited not to exceed 100%, then it is appropriate to view anthropogenically forced climate change and natural climate change as two parts of a divisible whole. Therefore there is absolutely nothing wrong or illogical about my statement:
Pick one:
a) Warming since 1950 is predominantly (more than 50%) caused by humans.
b) Warming since 1950 is predominantly caused by natural processes.
Using ‘>50%’, as suggested by Hegerl et avoids the specific problems of using ‘more than half’, but again the common understanding of percentage is to expect the values to relate to a divisible whole of 100%.
The AR5 further states:
The best estimate of the human induced contribution to warming is similar to the observed warming over this period.
I would infer that the IPCC’s best estimate is that human induced contribution is close to 100% for the period since 1950, which effectively implies that low values of sensitivity and the ‘pause’ are irrelevant.
This issue of semantics, while it may seem arcane, is an important one, and illustrates the main source of my disagreement with J N-G on this issue (there are of course additional disagreements with Gavin).
Attribution
It seems that I need to take on the logic of ‘fingerprinting’ as a method for attribution. I note here that Science of Doom is beginning to address this very messy issue:
- Natural Variability and Chaos — Three — Attribution and Fingerprints
- Natural Variability and Chaos — Seven — Attribution and Fingerprints or Shadows?
The ‘magic’ of fingerprinting is described by Gavin:
Judith’s argument misstates how forcing fingerprints from GCMs are used in attribution studies.
No, I didn’t misstate this, I simply don’t buy the IPCC’s fingerprinting (I actually think the AR4 approach makes more sense than the AR5).
Notably, they are scaled to get the best fit to the observations (along with the other terms). If the models all had sensitivities of either 1ºC or 6ºC, the attribution to anthropogenic changes would be the same as long as the pattern of change was robust. What would change would be the scaling – less than one would imply a better fit with a lower sensitivity (or smaller forcing), and vice versa (see figure 10.4).
So . . . even a minuscule sensitivity to CO2 – say TCR = 0.1C, would not change the attribution argument. Hard to imagine such a small sensitivity would not change the size of the green and the orange bars in the figure below:
(From AR5, Fig 10.5)
Necessary (but not sufficient) for a credible fingerprinting attribution is to understand the fingerprints associated with natural internal variability on multidecadal and longer timescales, which is essentially ignored. In addition to sensitivity being irrelevant, the ‘pause’ since 1998 and the cooling period 1940-1975 seem irrelevant to the fingerprinting. A substantial contribution for multi-decadal and longer internal variability has been essentially defined out of existence (it appears the AR5 ‘forgot’ to do the detection step.)
Stay tuned for a future post ‘Muddy fingerprints.’ Not exactly sure when I will get to this tho.
Bottom line: the climate attribution problem needs to be reframed. Attempting to discredit my arguments over semantics reflects tilting at windmills, with the root cause being very unclear statements made by the IPCC in their main conclusion statement on attribution.
Filed under: Attribution, Communication, IPCC
