Reality wins, it seems. Dr Roy Spencer writes:
As seen in the following graphic, over the period of the satellite record (1979-2012), both the surface and satellite observations produce linear temperature trends which are below 87 of the 90 climate models used in the comparison.
![CMIP5-90-models-global-Tsfc-vs-obs[1]](http://wattsupwiththat.files.wordpress.com/2013/10/cmip5-90-models-global-tsfc-vs-obs1.jpg?quality=83)
more here: http://www.drroyspencer.com/2013/10/maybe-that-ipcc-95-certainty-was-correct-after-all/
Chris Schoneveld:
I did read your post at October 15, 2013 at 4:40 am.
It says ou don’t have a clue what you are talking about and try to hide it with ad hom. and bluster.
To quote a narcassist, it is as simple as that.
Richard
ferd berple and steveta_uk:
re your posts at October 15, 2013 at 7:12 am and October 15, 2013 at 7:26 am, respectively.
Yes, the 87 models that did not emulate the last decade are known to be wrong. But it does not follow from this that the other three models have any merit. If you really think examination of the models will reveal why they have and have not matched the last decade then there is no more reason to retain the three than to throw out the 87. Either the examination may reveal something so keep and examine them all, or it won’t so junk them all.
Please read the post from passingstatistican at October 15, 2013 at 7:15 am. It explains the matter and this link jumps to it
http://wattsupwiththat.com/2013/10/14/90-climate-model-projectons-versus-reality/#comment-1448619
Richard
Richard, I don’t think neither Ferd nor I suggested that the 2 or 3 models are right. Simply that these 2 or 3 have not been shown to be wrong.
As I implied before, rejecting them solely on the grounds that other models are wrong is foolish.
As RJB would state. The really clever work now would be to bin all the model that are failures and investigate why the two at the bottom have matched reality and what they have assumed compared with the failed ones.
That’s the Potential published paper work What have the model winners assumed that the model failures haven’t.
As rGb does in fact state, the right thing to do is to go through the 90 models, apply a hypothesis test to each model independently, and reject all of the models whose entire envelope of variation is outside the actual climate, all models whose envelope of variation spends over 95% of its time outside of the actual climate, all models that have autocorrelation or variance that is egregiously incorrect compared to the autocorrelation or variance of the actual climate. Eyeballing the data, one could eliminate at least 2/3 of the models in the figure above on the basis of these general rejection criteria for HADCRUT, a lot more on the basis of UAH LTT. The ones that survive the cut would, of course, have a MUCH lower (and still high-biased) mean, corresponding to a vastly reduced climate sensitivity.
We are not finished, in other words, with the plunge in climate sensitivity that is underway. Almost all of the 90 models above clearly have too much positive feedback as is evident BOTH from the growing divergence from the actual climate AND from the fact that their variation is both too broad and too rapid compared to the data. The models are not correctly reproducing the “forces” that maintain approximate climate stability given the “inertia” of the climate relative to all forcings, in other words. To borrow words from the IPCC’s own report it is very likely that the effective restoring forces have “spring constants” in some mean field treatment that are too large, have a positive feedback bias on top of this, and underrepresent the heat capacity of the system. And that’s just from a glance at the data. A systematic analysis of the data (one model at a time) would reveal much more, per model, and might even provide one with enough information to indicate how to “fix” the models to better represent reality without breaking their physics content.
rgb
Before drawing any conclusions from this chart, I would ask the following questions:
1. Which RCP scenario is being fed into the models?
2. How closely does the chosen scenario resemble the actual observed concentrations to date?
3. What are the OLS trends for each of the model projections and the observations along with their 95% confidence ranges (after correcting for auto correlation)?
4. To what degree do the confidence ranges overlap?
5. How robust is the analysis to different start years?
We’re dealing with an alien worldview that places theory over empirical data and beliefs/values over results.
Robert Brown:
Thankyou for your post at October 15, 2013 at 8:01 am. This link jumps to it.
http://wattsupwiththat.com/2013/10/14/90-climate-model-projectons-versus-reality/#comment-1448653
I agree, and I am especially grateful for your writing
That is precisely what I meant when I wrote at October 15, 2013 at 7:53 am
Richard
steveta_uk:
Thankyou for your post to me at October 15, 2013 at 7:58 am.
I think we may be talking at cross purposes. Please see my post addressed to Robert Brown at October 15, 2013 at 8:14 am. This link jumps to it
http://wattsupwiththat.com/2013/10/14/90-climate-model-projectons-versus-reality/#comment-1448663
Also, please read his post (I link to it from my post to him) because his post expands on exactly what I was saying, so your reading his post may unblock the impasse between you and me.
Richard
CMIP5 is a modeler’s smorgasbord of every parameter under the Sun, including the Sun, and is primarily a couple-able and tweakable menu of models with high resolution such that micro-climate regions can be targeted for experiment. I assume that a super computer somewhere has this ready to go with a long waiting list of groups wanting to use it. With this computer design, just about everything can be tweaked, including the component code for each individual parameter. So it is possible (note I did not say probable) that someone may have come close to a fairly good model of how the real climate works. Do we have the name of the two that shadowed observations? Here is a link to a useful page on CMIP5. Scroll down to find the link of all the “official” model names using CMIP5.
http://cmip-pcmdi.llnl.gov/cmip5/guide_to_cmip5.html
Thinking out loud: Note how many groups using CMIP5 are from China. Are they attempting to use CMIP5 to refute the contention that they are the ones pumping anthropogenic CO2 into the air or that it is harmful if they are? CMIP5 certainly can be used by skeptics (you don’t have to couple the AGW-CO2 model component into your selection but you can also dial down greenhouse gas affects) and I wonder who among the list are bent that way. China would be a country that would parse out its study results only if it benefits them, given that its researchers are shackled to government control to a far greater degree than elsewhere. And if it behooves them to continue a ruse, they would do so. If someone were to demonstrate that the world will not burn up, we would have no reason to ship our manufacturing to China if we can do it right here at home (yeh, I know, gotta have cheap labor too). Developing countries willing to manufacture our stuff would be shooting themselves in the foot if they were to show there is no runaway anthropogenic CO2 affect.
richardscourtney says:
October 15, 2013 at 8:14 am
“Robert Brown:
Thankyou for your post at October 15, 2013 at 8:01 am. This link jumps to it.
http://wattsupwiththat.com/2013/10/14/90-climate-model-projectons-versus-reality/#comment-1448653
I agree,…”
I have been hoping that someone like Robert Brown or Richard Courtney would take a crack at models that DO better match the temperature record. The total failure of all models based only on a CO2 cause has been known for some time, and although CAGW proponents have tried to shoe horn their averages into the envelope, they, too are increasingly unhappy about the performance. I know the skeptics’ only job is to attempt to falsify the science, but because of the cast-in-stone stance of the “proponents”, who are shaving the ECS down a bit and beginning to appeal to aerosols, ENSO, AMO/PDO and even the much denigrated quiet sun to shore up the central CO2 cause, it might be time for someone outside the proponent fraternity to finally do the job for them. I think that otherwise, the obvious open door to better models as a face-saving or renovation of crumbling reputations will invite the brighter ones remaining among them to step through and throw the rest of them under the bus. It would be nice to have a skeptic finally do this and leave them to their phlogiston.
One step before the modelling, however, would be to do them on UAH temps or Hadcrut 3 at least. Hadcrut 4 is simply jacked up <0.2C in the more recent record to try to nip the bottom of the model ensemble and to shorten the period of the flat temps after UEA's P. Jones had remarked on the 15 year hiatus a few years ago. Ultimately we have to reconstruct temps using only well-sited thermometers because if you get the physics right in the new model, it will be judged not fully adequate because of the fiddled higher trend of the temperature record.
The problem with most data sets (and models based on it) is that they are not properly balanced:
1) balanced sh/nh i.e amount weather stations SH = amount weather stations NH
2) balanced by latitude i.e. sum must be zero or close to zero
3) balanced 70% @sea and 30% in-land
4) all continents included
look at global minima to rubbish AGW (minima are not pushing up the average temps)
look at global maxima if you want to predict the future (maxima are an independent proxy for energy coming through the atmosphere)
if you follow these simple procedures you will or you should get the same results as I got
http://blogs.24.com/henryp/2013/02/21/henrys-pool-tables-on-global-warmingcooling/
http://blogs.24.com/henryp/2012/10/02/best-sine-wave-fit-for-the-drop-in-global-maximum-temperatures/
From my understanding: 1. Modelers get to decide what to use to “drive” the coupled models. (models are initially driven with input data and then let run on their own). The data input driver can be temperature observations from the past as well as input forcings such as SST or CO2 ghg forcing scenarios. You can even put in white noise false data in place of real observations. 2. You then get to decide how you will “tune” the models to match your particular set of observations before setting the computer to “run”. 3. The models are then allowed to run through several “years” worth of a myriad of coupled model calculations designed to mimic how the Earth’s various systems respond to your set of input drivers which you then compare to new observations.
Obviously it takes a while to be able to publish your results. You want the model setup you put together to match observations so I am guessing you need at least 5 years worth of new observations to compare with your modeled output. This is probably why researchers started using the term “projected scenarios”. They could publish right away because scenarios don’t have to be compared to new observations. However, the risk of being a flash in the pan is great. 5 years later your published research could be trashed by mother nature.
Given that so many CMIP5 models (and more importantly, they way they were driven) were indeed trashed by mother nature, a new set of model components and runs will be put together for AR6 in the unending attempt to live on research grants strictly used to study human-forced anthropogenic warming. If I had a booky, I would bet the farm that most of the next generation will not match new observations either. The gravy train will continue until the last voting person standing decides to keep a hand on the old wallet.
I went looking for the next set of model construction parameters for AR6 the IPCC is looking for. This is something they have done in the past. But could only find this:
http://www.ipcc.ch/apps/eventmanager/documents/5/030920131000-INF_1_p37.pdf
It is a synopsis of how a selection of countries responded to questions about the future of IPCC. Several said stop with the AR reports. You are done with the research. Now tell us how to force people to change.
Kenya was a standout and said to work on the models, bringing up the fact that the regional Kenyan model included a sea-ice coupling parameter. Yes, they stated the obvious. There is no sea-ice in Kenya.
I wonder what it would like like if the models had the CO2 knob set to 400ppm today and used that as the trajectory. My guess is that reality would be below them all.
What other enterprise can consistently make projections that are 200% wrong and have any credibility let alone suggest $trillions be spent on their highly inaccurate projections of doom?
It’s time the adults returned to the room.
This shows that Mother Nature has been wrong. The modelers should sue Mother Nature in a world court. Lots of luck getting Mother Nature to pay damages.
The ENSO forecasts look very much like the above graph. The predictions are almost all above 0 for this time but the actual numbers dropped from 0.00 to -0.28 over the last two weeks. Is this a coincidence?
Based on the trend of the graphic, reality meets observation about 5 kilometers underground.
The problem is now apparent: the IPCC hired geophysicists instead of climatologists by mistake.
– – – – – – –
Robert Brown,
I am pleased you started in this thread after you were mentioned by JustAnotherPoster on October 14, 2013 at 2:15 pm.
The IPCC itself recognizes, you pointed out in the above quote, the models used in their report were rather limited in their capability for correctly reproducing future climate behavior.
To me this means they consider the models as still a development work in progress. It implies serious efforts are still needed to overcome what they do incorrectly.
If that is a reasonable statement of the IPCC’s view of the models included in their report then the certainty of the future calculated by the models in the SPM do appear overstated as many have pointed out. What occurs to me is the IPCC can just say, in the face of criticism, something like (my words) => ‘we are being reasonably pre-cautious on the safe side in showing more future warming until, in the indefinite future, we finally get the models right.
Just looking for were the IPCC’s CAGW hockey puck is going to be come January 2014.
Your comment?
PERSONAL REQUEST. => rgb, what is the status of your book ? You have mentioned in previous comments over the past year or so that you are working on (IIRC) a book on epistemic subjects.
John
f that is a reasonable statement of the IPCC’s view of the models included in their report then the certainty of the future calculated by the models in the SPM do appear overstated as many have pointed out. What occurs to me is the IPCC can just say, in the face of criticism, something like (my words) => ‘we are being reasonably pre-cautious on the safe side in showing more future warming until, in the indefinite future, we finally get the models right.
Just looking for were the IPCC’s CAGW hockey puck is going to be come January 2014.
Your comment?
PERSONAL REQUEST. => rgb, what is the status of your book ? You have mentioned in previous comments over the past year or so that you are working on (IIRC) a book on epistemic subjects.
Yes, the IPCC could indeed say something like this. If the authors of its reports wanted to be brought before congress and charged with contempt of congress as the preferable and civilized alternative to being attacked by an angry mob armed with pitchforks and torches.
This would be basically saying “We’ve been lying to you from the beginning, but it is for your own good, maybe, because we could have turned out to be right.”
At times like these, I like to trot out a few lines from Feynman’s Cargo Cult address:
Details that could throw doubt on your interpretation must be given, if you know them. You must do the best you can–if you know anything at all wrong, or possibly wrong–to explain it. If you make a theory, for example, and advertise it, or put it out, then you must also put down all the facts that disagree with it, as well as those that agree with it. There is also a more subtle problem. When you have put a lot of ideas together to make an elaborate theory, you want to make sure, when explaining what it fits, that those things it fits are not just the things that gave you the idea for the theory; but that the finished theory makes something else come out right, in addition.
The removal of the lines clearly stating reasonable doubt from AR5’s SPM — is that the mark of good, honest science? Is failing to point out that the GCMs’ GASTA predictions alone are already in poor agreement with facts, let alone all the other parts of this quintessentially complex theory that don’t fit, the mark of good, honest science?
I would like to add something that’s not essential to the science, but something I kind of believe, which is that you should not fool the layman when you’re talking as a scientist. I am not trying to tell you what to do about cheating on your wife, or fooling your girlfriend, or something like that, when you’re not trying to be a scientist, but just trying to be an ordinary human being. We’ll leave those problems up to you and your rabbi. I’m talking about a specific, extra type of integrity that is not lying, but bending over backwards to show how you’re maybe wrong, that you ought to have when acting as a scientist. And this is our responsibility as scientists, certainly to other scientists, and I think to laymen.
For example, I was a little surprised when I was talking to a friend who was going to go on the radio. He does work on cosmology and astronomy, and he wondered how he would explain what the applications of his work were. “Well,” I said, “there aren’t any.” He said, “Yes, but then we won’t get support for more research of this kind.” I think that’s kind of dishonest. If you’re representing yourself as a scientist, then you should explain to the layman what you’re doing– and if they don’t support you under those circumstances, then that’s their decision.
I would think that the same principle would apply to people who claim that their research is going to “save the world” to guarantee the continuation of what has grown to become one of the world’s fattest funding trees — provided, of course, that your proposed work is looking into anthropogenic global warming (that is, provided that you’ve already begged the question that AGW exists). Is the vast research infrastructure that has been built to study the climate and predict its future capable of surviving a “never mind, sorry, we got it wrong, there probably won’t be any catastrophic AGW after all” moment? Is it capable of the scientific honesty required to commit public seppuku, to literally spill its guts in expiation of the hundreds of billions of dollars misspent and the millions of lives being lost per year all due to the artificial inflation of carbon based energy prices?
Even if it were, will it be given the chance? For a scientist you are right — saying “I was wrong” is a part of honest science. For a politician who supported the incorrect scientific conclusion and wasted our hard earned money and quite possibly contributed to the recent depression and near-collapse of the Euro, there are no second chances. Expect the tail to wag the dog, because the tail is in control of everything from funding streams to an entire network of media devoted to controlling public opinion and perception. Why do you think that they rewrote AR5’s SPM, the same way that they rewrote AR4’s SPM, after the actual scientists were done with it? Because if the SPM honestly stated the uncertainties, the IPCC would never have been more than a tiny, nearly irrelevant UN structure devoted to predicting and ameliorating things like the southeast asian monsoon, and the world’s poorest people would have far cheaper energy. Even the energy companies benefit from the panic that has been created. It has “forced” them to raise their prices, and their profits are margins on those prices. They don’t lose money because of CAGW, they make it!
One example of the principle is this: If you’ve made up your mind to test a theory, or you want to explain some idea, you should always decide to publish it whichever way it comes out. If we only publish results of a certain kind, we can make the argument look good. We must publish BOTH kinds of results.
I say that’s also important in giving certain types of government advice. Supposing a senator asked you for advice about whether drilling a hole should be done in his state; and you decide it would be better in some other state. If you don’t publish such a result, it seems to me you’re not giving scientific advice. You’re being used. If your answer happens to come out in the direction the government or the politicians like, they can use it as an argument in their favor; if it comes out the other way, they don’t publish at all. That’s not giving scientific advice.
Where is the evidence that the people running the GCMs have ever “tested their theories”? When I glance at figure 1.4 of AR5’s SPM, can I pick out model results that nobody sane would consider not to have been falsified by the actual data? I can, easily. There are model results at the very top of the spaghetti envelope that are never anywhere close to the data. Why are they still there in the first place, contributing to the “meaningless mean” of all of the model results? Instead of openly acknowledging that these models, at least, have failed and throwing them out, they are included for the sole reason that they lift the meaningless mean of many GCMs, indeed, lift it a LOT as outliers.
A lowered mean would be in better agreement with observation (and still would be meaningless as the average of many models is not a predictor of anything other than the average of many models according to the theory of statistics) but it would weaken all of the political arguments for expensive and pointless measures such as “Carbon Taxes” that bring great profit to selected individuals and will not, even according to their promoters, solve the climate problem by ameliorating CO_2 in the foreseeable future.
We’ve learned from experience that the truth will come out. Other experimenters will repeat your experiment and find out whether you were wrong or right. Nature’s phenomena will agree or they’ll disagree with your theory. And, although you may gain some temporary fame and excitement, you will not gain a good reputation as a scientist if you haven’t tried to be very careful in this kind of work. And it’s this type of integrity, this kind of care not to fool yourself, that is missing to a large extent in much of the research in cargo cult science.
What more can one say? AR5 has now “officially” bet the farm on its SPM. Everybody knows that the draft openly acknowledged the fact that the models are not working and contained a now-infamous figure that allowed any non-technical reader to see this for themselves. Everybody knows that this acknowledgement was removed in the official release, and that the figure in 1.4 was replaced by a figure that fairly obviously obscured the obvious conclusion — shifting and renormalizing the axes so that the data divergence was less obvious, replacing colored ranges with a plate full of incomprehensible spaghetti so that one can see that some colored strands spend some of their time as low as the actual climate.
At this point they are at the absolute mercy of Nature. In two years, in five years, in ten years, either Nature will cause GASTA to shoot back up by 0.5C or so all at once so that it starts to correspond with the GCM predictions, or it won’t. If it doesn’t — worst case for them, if GASTA remains constant or actually descends (and there are some halfway decent reasons to think that it might well descend even without the use of GCMs at all, and they are not unaware of this and there are signs that the climate community is starting to break ranks on this) then they are done. The temporary fame and excitement that brought Michael Mann to the foreground as the cover story of many books will be replaced by ignominy, congressional investigations, and yes, pitchforks and torches and now they cannot back out of the latter because the changes in AR5’s SPM will be damning proof that climate science has been good old fashioned cargo cult science for two decades now, benefitting nobody but the high priests and politicians leading the cult.
IMO this is unfortunate. Not all climate science has been dishonest. The actual scientific reports from the working groups have been a lot more open about uncertainties (although they too have suffered from political rewriting after the fact to eliminate some of this before the reports were allowed to go public). And I’m certain that a lot of research has been done in the best of faith. But when one is funded to do research on and report on how CAGW is going to affect the migratory behavior of species, you aren’t going to return an answer of “it isn’t” or an answer qualified by “IF AGW turns out to be a correct hypothesis”, you’re going to return an answer of “here are the expected effects given an assumed warming of X”. Bayes might as well never have lived.
Finally, as regards my book Axioms, it is still being written, unfortunately. I’ve finished maybe half of it (and am pretty happy with that half) but the second half is the “messy” part of analyzing things like religion and ethics and I tend to rant too much and write too long every time I dig into it. I’m also insanely busy, and Axioms is just one of a dozen things on the back burner as I’m teaching a large class in physics, trying to fix up and improve my textbooks, get a startup company to take off so I can earn enough wealth in the process to be able to do whatever I like for the rest of my professional career, and get kids through college and launched. But it is near and dear to my heart. You can always go and grab the last image I uploaded before I quit working on it at:
http://www.phy.duke.edu/~rgb/axioms.pdf
This part does a fair job of working through elementary axiomatic metaphysics to where one has a defensibly “best” basis for epistemology and ontology, for a worldview, but one that is flexible enough to accommodate both some personal choice in what to believe and to accommodate the imperfect and incomplete and constantly changing description of “probable best belief” concerning propositions about the real world.
Enjoy, at least so far.
rgb
Anthony:
I have deliberately addressed this to you using the correct spelling of your first name in hope of ensuring that you read it.
I write to request that you ask Robert Brown to ‘tidy’ the parts of his post at October 16, 2013 at 9:42 am which pertain to GCMs and Feynman’s ‘Cargo Cult Science’ with a view to the resulting version being a Guest Essay for WUWT. This link is to his post to which I refer
http://wattsupwiththat.com/2013/10/14/90-climate-model-projectons-versus-reality/#comment-1449916
It really is superb stuff and is far, far too good to be lost as a post near the end of a long thread. In my opinion, it deserves to be a ‘sticky’.
Richard
– – – – – – – – –
rgbatduke,
Thank you for your reply. It is going to become a classic WUWT inline comment, I think.
When Feynman was alive I never met him personally or saw his talks live. My loss.
I will look at what you’ve written so far of ‘Axioms’.
Over the last year or so in comments you several times have recommended Thompson Jaynes’ book ‘Probability Theory: The Logic of Science’. I have located a copy at a local university library to which I have access . . . that will be slow reading . . . I expect a lot of heavy statistical lifting in it even with my engineering focused statistics education. : )
John
Over the last year or so in comments you several times have recommended Thompson Jaynes’ book ‘Probability Theory: The Logic of Science’. I have located a copy at a local university library to which I have access . . . that will be slow reading . . . I expect a lot of heavy statistical lifting in it even with my engineering focused statistics education. : )
You can find a copy online for free here:
http://omega.albany.edu:8008/JaynesBook.html
Less some editing added posthumously. This book was knocking around for well over a decade before Jaynes died; I’ve had a privately circulated copy since maybe the late 80s or early 90s. You should also grab the online PDF of his “Mobil Lecture” — I can’t really give the link because they haven’t set it up with its own page, but it can be found on the WUSTL site with his publications on it. It is the original base for the book.
You should also purchase Richard Cox’s monograph on the law of probable inference. Cox is arguably the originator of this in a 1940-something paper that narrowly preceded Shannon’s paper on information theory that turned out to derive the same thing a different way. The Cox axioms are the basis of probability theory AS the logic of science and by inheritance, what it is reasonable to believe in an entire (reasonably) internally consistent (but unprovable) worldview, and they are very simple axioms indeed. You’ll have no trouble reading either the Cox monograph, the Mobil lectures, or at least the first few chapters of Jaynes’ book derived from the Mobil lectures. They are accessible to anyone with a bit of skill withe algebra and knowledge of the IDEAS of probability theory.
There is more reading out there — George Boole’s book on on thought in the 19th century brilliantly anticipates Cox/Jaynes without the axiomatic foundation, and John Maynard Keynes put down a lot of the axiomatic foundation in his book on probability theory without realizing its broader implications in physics and epistemology. Overall, this work completely replaces logical positivism, Popper’s falsificationism, and other efforts by e.g. the Munich school to come up with a sound basis for a theory of knowledge. It perfectly interpolates the extremes — one cannot generally empirically “prove” an assertion, but positive evidence has some weight, one cannot generally emperically “falsify” an assertion, but negative evidence has some de facto weight. Best of all, there is a lot of reason to think that this is how our brains actually function at the neural level, as a book by David MacKay makes rather brilliantly clear.
As I said, enjoy.
rgb
– – – – – – – –
rgbatduke,
I think there is now no doubt you are a university educator with all the assigned homework opportunities to me! : )
Especially thanks for the free online link to Jaynes’ book. It’s much more convenient than hardcover.
John