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)
more here: http://www.drroyspencer.com/2013/10/maybe-that-ipcc-95-certainty-was-correct-after-all/
sophocles says:
October 14, 2013 at 4:16 pm
This projecton possesses 3 different types of quirks. Quirks have several states that define them. As others have noted, the projecton in question has very strong odd spin. This remarkable spin makes it appear to have a TRUTH quirk, but skillful examination reveals it clearly possesses the SHAM quirk. Further evidence indicates projecton has an excited IN quirk. However, this excited state is expected to decay soon, leaving the projection with an OUT quirk after emitting an amazing amount of electromagnetic energy. Finally, its third quirk rapidly alternates state between FLIM and FLAM, effectively remaining in a hybrid dual state. These quirks clearly place this projecton among the Bogons, and this one may in fact be ubiquitous Gore’s Bogon, a very massive particle suspected to mediate several processes of decay, release of energy, and increase in entropy. In effect, Bogons lead to the depletion of stored forms of energy producing heat with little work. Bogons also make it difficult to replenish stored energy. Their effects are very costly, produce instability, and progressive increase in the mass of governmentium.
Canman: “Were they tuned to that pre-98 “W” or did they predict it?”
I’d assume they were tuned on it. Everything thereafter being what happens when you get off the far end of a Taylor series.
Assume that’s not the case: Then the models are actually damn good, within various limitations. But necessarily they require real measurables plugged in for them to reach damn good. Or they wouldn’t fit nice back then and go ‘derp’ in the recent now and beyond. As such, they cannot predict climate not because they are broken as such. But because they rely on an unpredictable measurable. For example, the recent postings about failing to predict the AMO.
One case leaves hope open. Both cases toss out any idea that there’s predictive validity in them.
Canman: Forgot to mention. As far as I know the models are all open ended equations rather than a notion of deflection and tension from a baseline. eg. Grey body estimates from insolation. So in that latter case with unpredictable measureables, there’s every reason to expect the values to diverge to an infinity of either sign than to settle down towards a physically known limit or baseline.
Well not to worry Dr Roy, it seems that past numbers are always too high, and GISS and others seem to like to “adjust” the past observed and recorded data downwards, to keep it in general agreement with present day model requirements.
So pay no attention to all that high data observed in the year 2028, because by 2078, all that bloated stuff will be suitably corrected down to then prevailing terraflop computer models.
I don’t winkle, titter, mutter, tinkle, bleat, inkle, or dweet, so no need to keep repeating any of those, for me to click on.
I have discovered that more and more “sites” no longer have e-mail response capabilities, but have little blue birds, and other juvenilia, for people who can’t write a complete English language sentence.
I’m quite a fan of atto-second phenomena; which is about the amount of time, I’m prepared to devote to web sites that lack open e-mail response capabilities. I’m actually an enthusiastic student of archeo-physics, which is the study of all the really interesting things, that happened in the first 10^-43 seconds after the big bang. After that, it gets kind of boring, which is why I like to have e-mail to keep me busy.
@ur momisugly Werner. You make a good point. I would have thought that Hadcrut 4 is suspect and its use a poor choice. Perhaps Roy Spencer is in a position where he needs to make some concessions to whatever the prevailing political correctness is. Perhaps he hasn’t compared hadcrut 3 and 4. Perhaps he feels that he can better make his point against the warmists by using their own (flawed) data sets.
So 3 out of 90 models did a good job. Interesting then to find out what exactly made them stand out and what they did right (or not, in case they were right for the wrong reasons).
I can’t see The Pause in the observational data. Why?
Werner, by putting the sampling at 60, you have created graphs that are unrecognisable. What is your justification?
Ah, I see now, we are talking about a 5 year running mean. Sorry for asking the obvious.
Gcapologist:
This is the reply I promised to provide to your post addressed to me at October 14, 2013 at 5:18 pm. It said in total
And I note that at October 14, 2013 at 5:34 pm you provided this addendum
Taking your latter point first, we cannot “advance the conversation” with “the powers that be” because they are fulfilling an agenda and are only conversing with those they employ to justify the agenda. I explained this in another WUWT thread, and this link jumps to that explanation
http://wattsupwiththat.com/2013/10/12/tail-wagging-the-dog-ipcc-to-rework-ar5-to-be-consistent-with-the-spm/#comment-1445687
Hence, if we want to influence “the powers that be” then we need to mobilise public opinion behind the truth of the climate models; i.e.
the climate models have the same demonstrated forecasting skill as the casting of chicken bones for determining the future.
The amounts of CO2 and aerosols in the atmosphere vary from year to year. A climate model provides an output for a specific time by being input with the start temperatures and the amounts of CO2 and aerosols in the atmosphere at that time. And the model is run for time increments until a series of runs provides ‘projections’ through the present and for future times. Aerosols wash out of the atmosphere within days so their concentrations are input as being greatest near the sources of their emissions. CO2 is modeled as being ‘well mixed’ in the atmosphere.
It is assumed that the forcing potentials of CO2 and aerosols are constants. Indeed, as my first post to you explained, the models are ‘tuned’ using these assumptions and would ‘run hot’ without that tuning.
It is an interesting question as to whether those forcing potentials are constants in reality. Of course, the radiative properties of their molecules are constants but that does not mean their forcing potentials are constants in the real atmosphere where responses to radiative changes (i.e. feedbacks) may vary depending on circumstances.
I hope these answers are clear and what you wanted from me.
Richard
@dermot o logical
You should look at least at one whole solar cycle (11 years)
and then look at the linear trends. See here:
http://www.woodfortrees.org/plot/hadcrut4gl/from:1987/to:2014/plot/hadcrut4gl/from:2002/to:2014/trend/plot/hadcrut3gl/from:1987/to:2014/plot/hadcrut3gl/from:2002/to:2014/trend/plot/rss/from:1987/to:2014/plot/rss/from:2002/to:2014/trend/plot/hadsst2gl/from:1987/to:2014/plot/hadsst2gl/from:2002/to:2014/trend/plot/hadcrut4gl/from:1987/to:2002/trend/plot/hadcrut3gl/from:1987/to:2002/trend/plot/hadsst2gl/from:1987/to:2002/trend/plot/rss/from:1987/to:2002/trend
So, there is no “pause”
It has started cooling, globally
The Beatles knew a thing or two about GW (TM)
http://youtu.be/Bj1AesMfIf8
And that despite most of the time, there was additional warming from the PDO warm half cycle.
Around 2030, after the cold half cycle, temperatures may then end well below 1/3 of model predictions, equaling a sensitivity of below 1 deg.
Chris Schoneveld:
Your post at October 15, 2013 at 12:18 am says
Of dear! So many mistaken assumptions in so few words.
Clearly, you did not read my above post at October 14, 2013 at 4:21 pm which said
So, I spell out the matter as follows.
There are 90 models which are each different so it would be surprising if some did not provide an output something like the reality of the past decade.
But that does NOT indicate that the “3 out of 90 models did a good job”.
And it does NOT indicate that there was anything “they did right”.
And it does NOT indicate “they were right” for any reason (be it right or wrong).
Therefore, the fact that the output of those 3 out of 90 models coincidentally approximated reality over the last decade does NOT suggest they are likely to approximate reality over the next decade.
This is a link to the wicki explanation of the Texas Sharp Shooter fallacy
http://en.wikipedia.org/wiki/Texas_sharpshooter_fallacy
Rejecting all except 3 of the models is drawing the target around the remaining 3 models after the shot was fired.
Richard
Richard,
No I didn’t read your post. They do often make a lot of sense but now and then they have a strong element of pettifoggery, like in this case.
You write:
“But that does NOT indicate that the “3 out of 90 models did a good job”.
And it does NOT indicate that there was anything “they did right”.
And it does NOT indicate “they were right” for any reason (be it right or wrong).”
You make a big deal of a cursory question (you are a bit of a selfrighteous p…k, I have noted, as if you are the moderator or the CEO on this site. I am on my guard when your posts start with “Friends”, I am not your friend). I am not saying or indicating that what they do is intrinsically right. I didn’t add for nothing that they may well be right (meaning that the outcome is more or less in line with observations, because that’s what I meant with “right” or “with a good job”) for the wrong reasons.
I am just curious to know which input parameters have made these 3 models end up close to what was observed in terms of surface temperatures; indeed in what way do they differ from the other 87 definitely wrong models. Just by handwaving and assuming, as you do, that these models “coincidentally approximated reality” does not satisfy my curiosity. It is as simple as that.
The graph itself is an interesting image, sort of like a rising column of hot air…
Richardscourtney,
No I didn’t read your post. They do often make a lot of sense but now and then they have a strong element of pettifoggery, like in this case.
You write:
“But that does NOT indicate that the “3 out of 90 models did a good job”.
And it does NOT indicate that there was anything “they did right”.
And it does NOT indicate “they were right” for any reason (be it right or wrong).”
You make a big deal of a cursory question (you have a tendency to be selfrighteous, I have noted. I am on my guard when your posts start with the pretentious “Friends”. Yes guru, we are listening to your wise words). I am not saying or indicating that what they do is intrinsically right. I didn’t add for nothing that they may well be right (meaning that the outcome is more or less in line with observations, because that’s what I meant with “right” or “with a good job”) for the wrong reasons.
I am just curious to know which input parameters have made these 3 models end up close to what was observed in terms of surface temperatures; indeed in what way do they differ from the other 87 definitely wrong models. Just by handwaving and assuming, as you do, that these models “coincidentally approximated reality” does not satisfy my curiosity. It is as simple as that.
87 is 96.6% (97%) of 90. Why does 97% keep popping up in the AGW hoax?
– – – – – – – –
Roy,
Reality.
What of the intellectual leadership of the IPCC Bureau’s ideology whose concept of science allows the models to be counter-observational and still be valid? I suggest some at first just ghostly turned a whiter shade of pale***.
*** Procol Harum
John
http://wattsupwiththat.com/2013/10/08/the-taxonomy-of-climate-opinion/#comment-1441009
Steven Mosher says:
“Model answers fall within the range established by observations.”
Steve, did you really say that?
I suggest, with due respect, that the climate models cited by the IPCC are crap (please see Engineering Handbook for technical definition of “crap”).
Regards, Allan
Here is the evidence:
http://wattsupwiththat.com/2013/09/28/models-fail-land-versus-sea-surface-warming-rates/#comment-1432696
Condo be worse. The Global Warmistas could have gone into medical research, in which case there would be mountains of dead people who received incorrect treatments based on their theoretical understanding of science.
All they have really done is torqued public policies and squandered something just south of $2Trillion of public money on useless gree schemes, scams, cons and gifts.
That money would have purchased a lot of research for a cancer cure, better schools, upgrading infrastructure and helping developing nations. You know, useful stuff.
But nooooooooooooo, the Warmistas decided they were smarter than everyone else, smarter than Mother Nature and that their kindergarten level theories were infallible.
Quite the legacy they will leave, a legacy of blind adherence to their quasi Religous Eco Greenie ideas, of eliminating conflicting data, of operating personal smear campaigns to cover up their scientific malfeasance.
History will be very unkind to The Team et al. Mikey Mann could be the poster boy for courses taught to Freshmen science students about what not to do, to be aware of pride getting between scientific process and personal political and religous beliefs.
I would suggest the Warmista problem be summarized in the future as Mann Falling Sky Syndrome.
richardscourtney says:
October 15, 2013 at 2:23 am
Therefore, the fact that the output of those 3 out of 90 models coincidentally approximated reality over the last decade does NOT suggest they are likely to approximate reality over the next decade.
============
the results do indicate however, that the 87 other models should be rejected. whether these 3 remaining models were right by accident is quite possible, given that there were 90 “guesses” to start with. You would expect some to be right just by accident.
However, one cannot assume they were simply right by accident, even if it is likely. It would be of interest to know the assumptions in these models – to see how they were different – if at all – from the other models.
The most telling result of the models is that a single model, if run multiple times with very small changes to the inputs/assumptions, gives very different results. That each model itself generates spaghetti, and what we are seeing is simply the averages for each of the 90 models.
However, you cannot average all the possible throws of a pair of dice and arrive at what will actually happen. You get a number between 2 and 12. The average, 7, might be the most likely result, but when you add up all the possibilities, 7 is less likely than something other than 7. Of the 36 possible results, only 6 will yield a 7 (the average). 30 other times you get something other than the average.
This is the fallacy of modelling the future. Contrary to Einstein’s famous quote, God does play dice with the universe. Quantum mechanics has established this at a fundamental level, and until we develop some new theory to replace quantum mechanics, this limits our ability to predict the future at a fundamental level.
Climate modellers are trying to tell us that quantum mechanics does not apply to climate, that the future is given by an average, but this is not what the models themselves are saying. The models show quite clearly that for a given scenario, a near infinite number of futures are possible. And while the average might be the most likely, something other than the average is much more likely.
Hawking covers this in a lecture:
Thus it seems Einstein was doubly wrong when he said, God does not play dice. Not only does God definitely play dice, but He sometimes confuses us by throwing them where they can’t be seen….Thus, the future of the universe is not completely determined by the laws of science, and its present state, as Laplace thought. God still has a few tricks up his sleeve.
http://www.hawking.org.uk/does-god-play-dice.html
richardscourtenay
Never heard of the Texas Sharpshooter fallacy, but you are quite right to discount the wider significance of 2-3 reasonable fits out of 90 tries. Its a familiar problem which arises in many applications of statistics, sometimes called ‘data-mining’ or other names. Researchers carry out a number of experiments or whatever, looking for the result they want. When they succeed on eg. the 90th try, they attach the same statistical significance to the outcome as if it had come from a single experiment. Of course, the probability of heads in 90 spins of a coin is way greater than the probability of a head on one spin. So success on 2-3 out of 90 is most unlikely to be of any significance without a lot of other evidence.
Chris Schoneveld, it seems to me that two or three models have not been shown by the data to be wrong. Whatever some sharpshooter in Texas might think, discarding the models that have been shown to be wrong, and examining the remaining models to see if they are also wrong, seems simply an obvious thing to do.
Throwing out all models because most have been shown to be wrong is exactly like throwing the toys out of the pram.