Commentary from Nature Climate Change, by John C. Fyfe, Nathan P. Gillett, & Francis W. Zwiers
Recent observed global warming is significantly less than that simulated by climate models. This difference might be explained by some combination of errors in external forcing, model response and internal climate variability.
Global mean surface temperature over the past 20 years (1993–2012) rose at a rate of 0.14 ± 0.06 °C per decade (95% confidence interval)1. This rate of warming is significantly slower than that simulated by the climate models participating in Phase 5 of the Coupled Model Intercomparison Project (CMIP5). To illustrate this, we considered trends in global mean surface temperature computed from 117 simulations of the climate by 37 CMIP5
models (see Supplementary Information).
These models generally simulate natural variability — including that associated
with the El Niño–Southern Oscillation and explosive volcanic eruptions — as
well as estimate the combined response of climate to changes in greenhouse gas
concentrations, aerosol abundance (of sulphate, black carbon and organic carbon,
for example), ozone concentrations (tropospheric and stratospheric), land
use (for example, deforestation) and solar variability. By averaging simulated
temperatures only at locations where corresponding observations exist, we find
an average simulated rise in global mean surface temperature of 0.30 ± 0.02 °C
per decade (using 95% confidence intervals on the model average). The
observed rate of warming given above is less than half of this simulated rate, and
only a few simulations provide warming trends within the range of observational
uncertainty (Fig. 1a).
Figure 1 | Trends in global mean surface temperature. a, 1993–2012. b, 1998–2012. Histograms of observed trends (red hatching) are from 100 reconstructions of the HadCRUT4 dataset1. Histograms of model trends (grey bars) are based on 117 simulations of the models, and black curves are smoothed versions of the model trends. The ranges of observed trends reflect observational uncertainty, whereas the ranges of model trends reflect forcing uncertainty, as well as differences in individual model responses to external forcings and uncertainty arising from internal climate variability.
The inconsistency between observed and simulated global warming is even more
striking for temperature trends computed over the past fifteen years (1998–2012).
For this period, the observed trend of 0.05 ± 0.08 °C per decade is more than four
times smaller than the average simulated trend of 0.21 ± 0.03 °C per decade (Fig. 1b).
It is worth noting that the observed trend over this period — not significantly
different from zero — suggests a temporary ‘hiatus’ in global warming. The
divergence between observed and CMIP5-simulated global warming begins in the
early 1990s, as can be seen when comparing observed and simulated running trends
from 1970–2012 (Fig. 2a and 2b for 20-year and 15-year running trends, respectively).
The evidence, therefore, indicates that the current generation of climate models
(when run as a group, with the CMIP5 prescribed forcings) do not reproduce
the observed global warming over the past 20 years, or the slowdown in global
warming over the past fifteen years.
This interpretation is supported by statistical tests of the null hypothesis that the
observed and model mean trends are equal, assuming that either: (1) the models are
exchangeable with each other (that is, the ‘truth plus error’ view); or (2) the models
are exchangeable with each other and with the observations (see Supplementary
Information).
Brief: http://www.pacificclimate.org/sites/default/files/publications/pcic_science_brief_FGZ.pdf
Paper at NCC: http://www.nature.com/nclimate/journal/v3/n9/full/nclimate1972.html?WT.ec_id=NCLIMATE-201309
- Supplementary Information (241 KB) CMIP5 Models
Theo Goodwin says:
September 5, 2013 at 2:56 pm
Can’t tell if you’re kidding or not, but yes, I do have first hand evidence, or first ear, from the late 1950s.
Also direct documentary evidence, too, from early ’50s or late ’40s DoD jargon. A November 1951 Colliers magazine is quoted on the Net as saying: “Two long-time Pentagon stand-bys are fly-speckers and nit-pickers. The first of these nouns refers to people whose sole occupation seems to be studying papers in the hope of finding flaws in the writing, rather than making any effort to improve the thought or meaning; nit-pickers are those who quarrel with trivialities of expression and meaning, but who usually end up without making concrete or justified suggestions for improvement.”
You could check at a big library or buy the Nov 3, 10, 17 or 24, 1951 issues of Colliers on Amazon or eBay.
RACookPE1978:
You conclude your post at September 5, 2013 at 3:19 pm asking
http://wattsupwiththat.com/2013/09/05/statistical-proof-of-the-pause-overestimated-global-warming-over-the-past-20-years/#comment-1408995
Obviously, rgb will make whatever answer he wants. This is my ‘two pence’.
The problem is the ‘Texas sharpshooter fallacy’.
The Texas sharpshooter fires a scatter-gun at a wall, then paints a target around the middle of the impacts on the wall, and points to the target as evidence he is a good shot.
The models which failed to make an accurate forecast need to be rejected or amended because they are known to lack forecasting skill.
But removing the models which missed the target of an accurate prediction does not – of itself – demonstrate that the remaining models have forecasting skill: the models which seem to have made an accurate forecast may only have done that by chance (removing the ‘failed’ models is ‘painting the target’ after the gun was fired).
Therefore, and importantly, the remaining models may not accurately forecast the next 20 years.
There is an infinite number of possible futures. A model must emulate the dominant mechanisms of the modelled system if it is to be capable of agreement with the future that will eventuate. And each model is unique (e.g. each incorporates a unique value of climate sensitivity). Therefore, at most only one of them emulates the Earth’s climate system.
Hence, the outputs of the models cannot be averaged because average wrong is wrong.
Furthermore, there is no reason to suppose a model can forecast if it cannot hindecast, but an ability to hindecast does not indicate an ability to forecast. This is because there are many ways a model can be ‘tuned’ to match the past, and none of those ways may make the model capable of an accurate forecast.
Therefore, a model has no demonstrated forecast skill until it has made a series of successful forecasts.
Richard
BBould:
re your suggestion to me at September 5, 2013 at 3:55 pm.
Yes, having read the abstract I really do want to read that paper. I will do the search in the morning. It is now past midnight here. And, of course, I will reply to you when I have read it.
Richard
BBould:
What is the title and who is the author, please?
Richard
I see that issues of Collier’s are on line:
http://www.unz.org/Pub/Colliers-1951nov03
Wordola’s first recorded use is 1954:
http://www.wordola.com/wusage/nitpicking/f1950-t1959.html
A physics-based equation, using only one external forcing, calculates average global temperature anomalies since before 1900 with R2 = 0.9. The equation is at http://climatechange90.blogspot.com/2013/05/natural-climate-change-has-been.html . Everything not explicitly considered must find room in that unexplained 10%.
Richardscourtney: The Bayesian Treatment of
Auxiliary Hypotheses
Michael Strevens
British Journal for the Philosophy of Science, 52, 515–537, 2001
Copyright British Society for the Philosophy of Science
Theo Goodwin: Thank you very much!
milodonharlani says:
September 5, 2013 at 12:02 pm
> PS: I eschewed inserting the term nit-wit into the above copy.
Ah. I thought there was something witless about that previous comment.
milodonharlani says:
September 5, 2013 at 4:02 pm
Thanks. You are a class act.
=====================================================================
Perhaps nitlamps were invented to help nitwits see the light?
“hiatus”, the fall-back defense? Like luke-warmism, agreeing cedes the unspoken assumptions, which are comprehensively false.
Good to see a little etymological sparring in amongst the science.
Now, if only we could all distinguish between “its” and it’s”……..
@richard Barraclough – knowing the difference between those 2 won me a IT contract at the State Library. 😉
BBould:
I have now downloaded the paper
Strevens M, ‘The Bayesian Treatment of Auxiliary Hypotheses’, British Journal for the Philosophy of Science, 52, 515–537, 2001
At September 5, 2013 at 11:29 am you suggested to me
I have made a cursory study of the paper and will continue to give it much more thought. However, I am writing now to say that I do not think the paper is relevant to the discussion in this thread.
Firstly, I was surprised that I was unaware of a paper published 12 years ago which had the importance you suspected. My initial impression is that it does not have that importance.
Secondly, as a general rule, the importance of a paper is inversely related to its length. This paper is 42 pages long. My first reading of the paper suggests that it obeys that general rule.
The purpose of the paper seems to be to express a personal reaction of Michael Strevens to the work of Newstein. I do not know what if any personal or professional interactions Strevens has had with her, but he makes some personal remarks; e.g.
Section 2.4, page12
The subject of the paper is an attempt to quantify to what degree evidence refutes a theory.
In the 1950s Pierre Duhem cogently demonstrated that a scientific hypothesis is not directly refuted by evidence. This is because the evidence also represents additional hypotheses concerning how the evidence was produced and observed.
Duhem’s argument is plain when stated; e.g. if is assumed a long-jumper broke the world record, then the measurement to assess that assumption assumes the tape measure was accurate.
The Quine-Duhem thesis expands on that seminal work of Duhem.
There is always more than one assumption concerning the evidence (e.g. in the long-jump illustration, in addition to assumptions about the tape measure there are assumptions about how it was used). And there is a central hypothesis (e.g. the long-jump measurement provided a correct indication). In essence, the Quine-Duhem thesis says there is no way to determine how an individual assumption affects the importance of the indication provided by the evidence.
Hence, it cannot be known to what degree a piece of evidence refutes a theory because the acceptance of the evidence is adoption of unquantified assumptions.
This, of course, is undeniably true and it affords a get-out to pseudoscientists. Indeed, it has been used by climastrologists (e.g. unmeasured heat must be in deep ocean where it cannot be measured). As you imply, it could also be used as a get-out to falsification of climate models (i.e. the models are right so the evidence must be wrong).
Avoidance of such get-outs requires clear recognition of what is – and what is not – being assessed. An example of this need for clarity is stated by my post in this thread at September 5, 2013 at 3:20 am
http://wattsupwiththat.com/2013/09/05/statistical-proof-of-the-pause-overestimated-global-warming-over-the-past-20-years/#comment-1408432
That post is directly pertinent to the subject of Srevens’ paper because it argues that the uncertainties in the data are a separate issue from whether the climate models emulate the data.
Strevens’ paper claims it is possible to assign individual assessments to the assumptions included in a piece of evidence. In his Introduction on page 1 he writes
Those are bold claims which the paper fails to fulfill.
This failure seems to be because those claims are not the true purpose of the paper which says in Section 2.4, page 14
Indeed, a Bayesian would reply that. And would not see a need to refute Newstein.
In a peer review of the paper I would discuss the purported refutation, but that does not seem to be needed here. That is because, as the paper admits, the refutation is pointless. It admits in Section 5, page 26
In other words, Strevens’ admits his analysis only affords a solution to one limited type of assessment and is not generally applicable.
I hope this brief and cursory reply is sufficient answer to your request.
Richard
@ur momisugly Kadaka: I applaud you for doing the experiment, however …
In your first experiment you used enough energy to boil the water [1], yet only warmed it 3C. Yes, you falsified the proposition, but you rather proved the point – imo.
[1] if you’d put your roughly 0.5kg of water in a 1600 Watt kettle for 5 minutes it assuredly would’ve boiled.
Dan Pangburn:
I see nobody has answered your question in your post at September 5, 2013 at 4:30 pm which says in total
The model is merely an exercise in curve fitting. As the link says
This matches the data because the ‘effects’ are tuned to obtain a fit with the anomaly.
Hence, the model demonstrates that those ‘effects’ can be made to match the anomaly, but it does not demonstrate there are not other variables which may be similarly tuned to obtain a match with the anomaly.
The model matches the form of the anomaly. But, importantly, it only explains the opinion of its constructor: it does NOT explain anything about climate behaviour. Therefore, it does not have a residual of “10%” of climate behaviour which is unexplained.
The model – as every model – represents the understanding of its constructor. But the model has no demonstrated predictive skill and, in that sense, it is similar to the GCMs.
Richard
Sleepalot said on September 6, 2013 at 3:16 am:
If all of the air molecules exiting the heat gun impacted the water, and all energy gained from passage through the heat gun was transferred to the water resulting in the temperature of the air that bounced off the water surface being no greater than room temperature, you might have a point.
Except mere hot air blowing on a surface is far less efficient than the direct heating of an electric kettle, where the heating element may be immersed in the water with some designs. All that energy is not transferred, all the air molecules did not hit the surface. As the heat gun was agitating the water surface, there was likely energy lost as latent heat due to vaporization, to a degree far in excess of that of an electric kettle. Etc.
So, since a process that is many times more efficient could have delivered enough energy to boil the water in that time, and the process used that was far less efficient only warmed the water a few degrees, what is shown is… A more efficient heating method could have heated the water faster. And that’s about it for your comparison.
richardscourtney says: September 5, 2013 at 1:03 am
Friends:
The paper is reported to say
It is worth noting that the observed trend over this period — not significantly different from zero — suggests a temporary ‘hiatus’ in global warming.
NO! That is an unjustifiable assumption tantamount to a lie.
Peer reviewed should have required that it be corrected to say something like:
It is worth noting that the observed trend over this period — not significantly different from zero — indicates a cessation of global warming. It remains to be seen when and if warming will resume or will be replaced by cooling.
Richard
________
Hello Richard,
I agree with your above assessment. Furthermore:
In several recent papers, we are witnessing an undignified scramble by the warmist establishment to spin the story one more time. It is just more warmist nonsense, espoused by people who have ABSOLUTELY NO PREDICTIVE TRACK RECORD. I suggest that one’s predictive track record is perhaps the only objective measure of scientific competence.
In 2002, we wrote with confidence:
“Climate science does not support the theory of catastrophic human-made global warming – the alleged warming crisis does not exist.”
http://www.apegga.org/Members/Publications/peggs/WEB11_02/kyoto_pt.htm
The above statement was based on strong evidence available at that time that the Sensitivity of Earth Temperature to increased atmospheric CO2 was not significant and was vastly over-estimated by the climate models cited by the IPCC.
The term “temporary warming hiatus” implies that warming will resume. I submit that it will not, and Earth is entering a natural cooling period. I wrote this global cooling prediction in an article, also published in 2002.
The above global cooling prediction was based on strong evidence available at that time that global warming and cooling cycles were primarily natural in origin and Earth was nearing the end of a natural warming cycle and about to enter a natural cooling cycle. These natural cycles are somewhat irregular and the timing of our prediction (cooling to start by 2020-2030) may be a bit late – global cooling may have already begun, although we will likely only know this with certainty in hindsight.
We do know that SC24 is a dud and similar periods of solar inactivity (e.g. the Dalton and Maunder Minimums) have coincided with severe global cooling and major population declines due to famine in Northern countries. I suggest that IF this imminent global cooling is severe, and we are woefully unprepared due to global warming nonsense, the price society pays for our lack of preparedness will be much more grave.
Warmist nonsense has resulted in the squandering of over a trillion dollars of scarce global resources, mostly on inefficient and ineffective “green energy” schemes.
In 2002 we wrote with confidence:
“The ultimate agenda of pro-Kyoto advocates is to eliminate fossil fuels, but this would result in a catastrophic shortfall in global energy supply – the wasteful, inefficient energy solutions proposed by Kyoto advocates simply cannot replace fossil fuels.”
The policy makers of Europe, Ontario and California could have benefitted from this advice – instead, they severely damaged their economies by foolishly adopting worthless green energy schemes and are now having to reverse these decisions, due to soaring energy prices.
Another point – the satellite temperature record suggests a probable warming bias in the surface temperature record of about 0.2 C since 1979, or about 0.07C per decade, so one should regard the alleged surface temperature warming trends as of being questionable accuracy.
The global warming camp has much to answer for. They have promoted false alarm and have profited from it. They have squandered significant global resources. They have caused us to focus our attentions on a non-crisis – global warming – and thus have caused us to ignore a much greater potential threat, probable imminent global cooling. They have acted like imbecilic thugs, and have caused several eminent scientists to be dismissed from their academic positions.
At a minimum, I suggest that these thuggish university dismissals should be reversed without delay, with suitable apologies.
Foolish “green energy” schemes and the lavish subsidies that make them attractive should cease immediately.
I also suggest that serious study of probable global cooling and its possible mitigation, if it is severe, should be commenced without delay.
Regards, Allan
Allan MacRae:
Thankyou for your post addressed to me at September 6, 2013 at 6:18 am
http://wattsupwiththat.com/2013/09/05/statistical-proof-of-the-pause-overestimated-global-warming-over-the-past-20-years/#comment-1409424
It says
Hmmmm.
Well, if you are talking about the taking of empirical measurements then, no, I don’t see how that can be true.
But if you are talking about theoretical modeling then your assertion must be true. And it goes to the nub of this thread.
Richard
Global warming climatology is notable for the absence of the statistical populations underlying its models. The absence of these populations wounds this discipline. Casualties from this wound include probability theory, information theory, mathematical statistics and logic.
In their paper, Fyfe et al show how conclusions may be drawn from a global temperature time series despite this seemingly unsurmountable barrier. One makes a bunch of assumptions and buries these assumptions!
Friends:
Please resist temptation to answer the post from Terry Oldberg at September 6, 2013 at 6:44 am.
You know he is wrong.
I know he is wrong.
And he knows he is wrong because on previous IPCC threads he has been unable to define the “the statistical populations” he claims are “absent”.
Any attempt to engage with him is like entering Alice’s rabbit hole. And it completely destroys a thread.
If anybody doubts the need for my request I suggest that – as a recent example – they peruse the recent thread at
http://wattsupwiththat.com/2013/08/31/wuwt-hot-sheet-for-saturday-august-31st-2013/
Richard
John F. Hultquist says: September 5, 2013 at 2:03 pm
“No one — however smart, however well-educated, however experienced — is the suppository of all wisdom.”
– Tony Abbott
Disagree: I respectfully submit that the IPCC is the suppository of all wisdom.
A pondering about the pause:
During La Nina/La Nada conditions, when there are fewer clouds overhead but more wind, the ocean surface is roughened up, which leads to a less warm surface due to top layers mixing as well as shoving warmer top water away. That’s not of interest to me in the pause. What is of interest is the amount of warming that happens below the surface due to SWIR penetration under these conditions. If the skies are not under “clear sky” conditions during these recharge periods, we should see less warming of the water below the surface. Eventually, the needle goes to the positive side of the ENSO dial (IE El Nado or El Modoki), and the surface calms down to the point that this now less warmed water again sits on top, If these conditions continue, we should see a stable pause in subsequent land temperatures. However, if the swing back to La Nina/La Nada gets even less defined with more clouds, and the oceans get less and less recharged due to equatorial cloud cover, we could even see a stepping down process in subsequent land temperatures.
So then the question is, what data do we have on subsurface recharging warming during non El Nino conditions over this time period?
Maybe we need more definition and descriptive names for these ENSO periods, such as: La Nina, La Modiki, La Nada, Neutral, El Nado, El Modiki, and El Nino. It seems to me that the goodies in the pause could be found in the waters of La Modiki, La Nada and Neutral.