Why you won’t see headlines as climate science enters the doldrums
Guest post by Dr. Robert G. Brown, Physics Department of Duke University (elevated from a comment on this thread: RSS Reaches Santer’s 17 Years)
This (17 years) is a non-event, just as 15 and 16 years were non-events. Non-events do not make headlines. Other non-events of the year are one of the fewest numbers of tornadoes (especially when corrected for under-reporting in the radar-free past) in at least the recent past (if not the remote past), the lowest number of Atlantic hurricanes since I was 2 years old (I’m 58), the continuation of the longest stretch in recorded history without a category 3 or higher hurricane making landfall in the US (in fact, I don’t recall there being a category 3 hurricane in the North Atlantic this year, although one of the ones that spun out far from land might have gotten there for a few hours).
We (the world) didn’t have an unusual number of floods, we don’t seem to have any major droughts going on, total polar ice is unremarkable, arctic ice bottomed out well within the tolerances slowly being established by its absurdly short baseline, antarctic ice set a maximum record (but just barely, hardly newsworthy) in ITS absurdly short baseline, the LTT temperatures were downright boring, and in spite of the absurdly large spikes in GASTA in GISS vs HADCRUT4 on a so-called “temperature anomaly” relative to a GAST baseline nobody can measure to within a whole degree centigrade, neither one of them did more than bounce around in near-neutral, however much the “trend” in GISS is amplified every second or third month by its extra-high endpoint.
The US spent months of the summer setting cold temperature records, but still, aside from making the summer remarkably pleasant in an anecdotal sort of way (the kind you tell your grandchildren when they experience a more extreme weather, “Eh, sonny, I remember the summer of ’13, aye, that was a good one, gentle as a virgin’s kiss outdoors it was…”) it was unremarked on at the time.
Let’s face it. The climate has never been more boring. Even the weather blogs trying to toe the party line and promote public panic — I mean “awareness” — of global warming are reduced to reporting one of GISS’s excessive spikes as being “the fourth warmest September on record” while quietly neglecting the fact that in HADCRUT4, RSS and UAH it was nothing of the sort and while even more quietly neglecting the fact that if one goes back a few months the report might have been that June was the fourth coldest in 20 years. Reduced to reporting a carefully cherry-picked fourth warmest event? Ho hum.
So, good luck in getting any news agency to report reaching 17 years in any or all of the indices — this isn’t news, it is anti-news. It is old. It is boring.
It is also irrelevant. If GASTA (Global Average Surface Temperature Anomaly) stubbornly refuses to rise for five more years, stretching the interval out to 20 to 22 years in a way that nobody can ignore, does this really disprove GW, AGW, or CAGW? It does not. The only thing that will disprove GW or CGW is reaching 2100 without a climate catastrophe and without significantly more warming or with net cooling. A demonstrated total climate sensitivity of zero beats all predictions or argument. The “A”(nthropogenic) part is actually easier to prove or disprove in a contingent sort of way, although it will probably take decades to do so. Contingent because if there is no observed GW at all, AGW seems difficult to prove. But since we are in the part of the periodic climate cycle observed over the last 150 years where the climate remains neutral to cools around an overall warming trend, we might well see neutral to very slow warming even if AGW is correct, if there is an anthropogenic component to the long term trend and oscillation that we can observe but not really explain over the last 150 years.
The one thing the 33 years of satellite measurements and increasingly precise surface temperature measurements have been able to prove is the one thing that the 17 year interval is truly relevant to. The GCMs used to predict CAGW suck. The GCMs in CMIP5 (Coupled Model Intercomparison Project) that contribute to the conclusions of AR5 are almost without exception terrible predictors of the Earth’s actual climate.
This conclusion is unavoidable. Even if they all cannot be rejected at the “95% confidence level”, almost none of them are close to predicting even GASTA alone, let alone RSS/UAH, global rainfall, frequency and violence of storms, etc. As we leave 2013′s hurricane season behind with almost no chance for an Atlantic storm this year, which GCM predicted the paucity of hurricanes and tornadoes over the last few years? Where are the droughts and floods? Which GCMs actually got the temperature distribution right (when they didn’t get the average or average anomaly right, the answer is almost certainly “none of them”)?
We are told “Catastrophic warming is coming, it is just around the corner”. We ask why and without exception we are told “Because the 30 or more GCMs we carefully built in the 1990′s in response to the CAGW threat and normalized with the warming data from the 70′s and 80′s (not to mention Hansen’s initial model report from the late 1980′s) all say so.” We then quite reasonably ask what they predicted for the last 20 years, and of course we can see that they all did indeed predict shockingly rapid warming. We then compare this to what actually happened, which is almost no warming over the last 20 years — a single warming pulse associated with the 1997/1998 ENSO event and then neutral ever since. We note that the warmest of the models that are still included in the CMIP5 data because nobody ever rejects a model just because it doesn’t work are a whopping 0.5 to 0.6C warmer than reality — they are the models with a total sensitivity of 5 or 6 C by 2100, so they have to warm at 0.5C a decade to get there.
This really is shocking. Shockingly bad science, shockingly dishonest political manipulation of policy makers on the part of scientists who participated in the creation of AR5 and permitted their names to give the report its weight.
As I’ve pointed out once and will point out again, by failing to be honest in AR5, by removing words that expressed honest doubt from the earlier draft and redrawing the figure to obscure the GCM failure, the IPCC has now gone far out on a limb that will end the career of many scientists and politicians before AR6 if there is no significant warming by that time. Not only significant warming, but a resumption of some sort of regular upslope to GASTA. Even if there is another ENSO-related burst of warming (which I’m sure is what they are hoping for) if it is only 0.2 C — and it is difficult to imagine that it could be much more given evidence from the past — it will barely suffice to restore the warming trend to 0.1 C/decade give or take a hair, roughly half of the lowest estimates of climate sensitivity. And they run the very real risk of getting to 2020 with GASTA basically the same as it was in 2000.
At that time, the hottest GCMs are going to be almost a full degree C too hot compared to reality. The people who contribute to the IPCC reports aren’t fools — most of them know perfectly well that the high sensitivity models are trash at this point, and they know equally well that it will no longer be possible to conceal this fact even from ignorant politicians by 2020 if there is no statistically significant warming by that time. Because it is an open secret that there was a cover-up that deliberately concealed this, effectively lying to policy makers, there will be a public scandal. Heads will roll.
The only way the IPCC can possibly avoid this as it proceeds is to issue a correction to AR5. Go back in and eliminate the GCMs with absurdly high sensitivity, the ones that obviously fail a hypothesis test when compared to the actual climate record. Personally I would advise eliminating at a much more generous level than 95% — a complete idiot with experience in computational modeling could go into these models and figure out what is wrong, given an additional 16 years of data — simply retune the models until they can manage both the warming of the late 20th century AND the warming hiatus since. Models for which no tuning can reproduce the actual past go into the dustbin, period — ones that can manage it will all have a vastly lowered climate sensitivity and will produce a much larger fraction of warming from “natural” variability, and less from CO_2. Finally, insist that all models use common numbers for things like CO_2 and aerosol contributions instead of individually tuning the largely cancelling contributions to reproduce an interpolated temperature change.
I’m guessing that over half of the participating models will simply go away at this point. They can then reconstruct figure 1.4 in the SPM, note the good news that even though the remaining models will all still predict more warming than actually occurred the warming that they project by 2100 will be between 0.5 and 1.5 C, not 2.5 C or more. This is almost precisely in line with what was observed in the 19th and 20th century without CO_2, and will grant a far larger role to natural variability (and hence a smaller one to CO_2).
Why should they do this, even though it is near-suicide to do it at this point? Because it is sure thing suicide not to do it. Because it is the right thing to do. Because they have a queasy feeling in their tum-tums every time they look at figure 1.4 in the AR5 SPM and realize that the dent that they made in the car isn’t going to go away and Dad is going to be even more pissed when he finds out if they lie about it. After all, everybody knows that the worst models in CMIP5 are wrong at this point. The people that wrote the models and ran the models, they know that their models are broken at this point. It’s not like the failure of a model is difficult to detect or something.
If it were “just science”, all of this would have been happening in the literature for some time anyway. People would jump all over models that fail, because in the usual realm of science there is little money on the line and because trial and error and try try again is the normal order of business and what keeps you getting paid. Not so in climate science. Here it is all political. Hundreds of billions of dollars and the directed energy of the entire global civilization ride on the numbers. Here there is a real risk of congressional hearings where a flinty-eyed committee chair grills you by showing you GCM curves selected from figure 1.4 of the AR5 SPM and asks you “Sir, at what point was it obvious to you that this curve was not a good predictor of the future climate?” Because if the answer was “2012″ — and given the REMOVED TEXT from the earlier draft of AR5 everybody knows that it was 2012 at the latest — that’s contempt of congress right there, given that AR5 directs billions of dollars in federal research money and hundreds of billions of dollars of subsidies and misdirected governmental energy at all levels from federal to state to local to personal.
We pay, pay, and pay again in the form of taxes, higher energy prices, neglect of competing services and goals — and what we pay pales to nothing compared to the terrible price paid by the third world for the amelioration of hypothetical CAGW. Millions of people die every year from respiratory diseases alone brought about because they are still cooking on animal dung and charcoal because coal burning power plants are now “unclean” and have artificially inflated price tags at every level.
If CAGW is a true hypothesis, then maybe — just maybe — it is worth sacrificing all of these people, most of them children under five, on the altar to expiate our carbon sins. But given this sort of ongoing catastrophe, this ongoing moral price we pay on the basis of the “projections” of the GCMs, how great is the obligation of the scientists who wrote AR5 towards “mere honesty”, to put down not their own beliefs but to put down the objective support for their beliefs given the data?
For some time the data has been sufficient to prove that the tools that claim the biggest, scariest AGW are simply incorrect, broken, in error, failed. Yet their predictions are still included in AR5 because without them, the “catastrophe” disappears and we are forced to rebalance the cost of gradual accommodation of the warming while continuing to civilize and raise the standard of living of the third world against the ongoing catastrophe of adopting measures that everybody knows will not prevent the catastrophe anyway (if the extreme models are correct) at the cost of a hundred million or more lives and unspeakable poverty, disease, and human misery perpetuated for decades along the way.
Related articles
- If climate data were a stock, now would be the time to SELL (wattsupwiththat.com)
- A Sea Change for Climate Science? (wattsupwiththat.com)
- US Tornado Count So Low That It’s Invaded The Legend… (wattsupwiththat.com)
Shockingly bad science? I can’t see how willfully bad science can be considered shocking. What is shocking is the quality of the story. Have you heard about the one with a mysterious threat, a hope of redemption, ascension of the righteous and punishment of the wicked? The current one features CO2, Windmills, Michael Mann and Big Oil. Social democracy not only has an ongoing financial crisis, it also has shockingly bad mythology.
Always good to read anything by rgbatduke.
There is however, one cloud in his silver lining. I don’t believe ignorant policy makers will be holding cAGW scientists to account at some future date. I don’t believe the policy makers are ignorant at all (except in the most general sense of the word). Policy makers are, imo, believers in the ideology that cAGW brings, adherents of the higher degrees of control that can be leveraged from it, and enamoured by the promises of endless tax receipts.
These policy makers will do their best to support the dodgy science of the climate scientists long after it becomes blindingly obvious to everyone. Lest anyone forgets, the current UK minister for transport claimed that the olympic games came in on budget, but conveniently forgot to mention that said budget was raised 3 times (from 2.5bn to 10bn). Are these the policy makers who will be holding feet to the fire?
***
Pamela Gray says:
November 4, 2013 at 1:41 pm
***
Then you’ve got Assawoman Bay:
http://www.stateparks.com/assawoman_bay.html
evGiven the complexity of the system a model that got within 50% of the right number would be stunning.
Note, this is separate from the question of whether one should use models to set policy.
That said we set defense policy of the nation using models that were much less accurate.
Steven, this was a very reasonable reply and I agree. As I’ve said many time before in different threads, predicting the climate is one of the most difficult tasks humans have undertaken — one is solving at least two coupled navier-stokes systems (arguably coupled to a third magnetohydrodynamic one on the surface of the sun) on a rotating, tilted, magnetized oblate spheroid with a highly inhomogeneous surface in a moderately eccentric orbit around an irregularly variable star. Both fluids (the atmosphere and the ocean) have complex chemistry, complex biology (in the case of the ocean), variable density and complex transport and stratification. The atmosphere is critically saturated with water vapor over a substantial part of the surface daily and precipitates out into clouds at many heights and rain a many locations, causing rapid and contrary changes in both albedo and greenhouse warming, and with an entire cycle of latent heat transport both vertically and laterally. The ocean (and continental sized areas of the land) are both annually and permanently covered with ice and snow with a high albedo. The whole system pulses and has known named structures of atmospheric flow that have (at least) decadal periods — humans haven’t had the tools to verify any such structures with periods longer than decades — even though none of the visible drivers or dynamics in the GCMs are capable of producing coherence on the time scale of a single decade or reproducing the pattern of the shortest period such oscillation (that nevertheless seems to have a major impact on the time development of the climate).
It is a hard problem and it is really unsurprising that we haven’t yet solved it. In the ordinary course of science, this wouldn’t be a big deal — one expects to work on absurdly difficult problems for a lifetime and still maybe not solve it. Look at the search for the Higgs. Look at the discovery of high temperature superconductivity. People spent careers looking for both before either one was found (the former still arguably a bit tentatively, for all that they needed to give Higgs the prize now if they wanted to manage it in his lifetime at all). Looking for the Higgs boson is child’s play compared to accurately solving the equations for the time evolution of the climate. It could take 50 years to even start to get it right — we’re still working on having sufficiently good and accurate instrumentation and a long enough base of sufficiently precise and globe-spanning measurements made with that instrumentation to have what is needed to initialize such a predictive computation. I repeat — if we cannot compute GAST to better than plus or minus a degree C or more — so we are reduced to saying it is “around” 15 C — then claims for accurate computations of GASTA outside of the satellite era are IMO egregious and the estimates of global warming at all become at least highly variable and uncertain.
But my article wasn’t about patting the creators of the GCMs on the back for hard work reasonably well done, because the GCMs have not been used for the purpose of seeing if they get the climate right and then correcting them until they do when they don’t. They’ve been used for the sole purpose of predicting a global catastrophe. They continue to be used for that purpose in spite of the fact that everybody who is actually in the game knows perfectly well that they are unsuited for that purpose at this time. And AR5 was, I think you have to admit, a shameful cover up of that fact. The spaghetti graphs in figure 1.4 of the SPM has no business including models that are currently sitting 0.5C or more above the actual present temperature and you know it.
This would be perfectly obvious if one took every CIMP5 model and plotted it one at a time against the actual data. Some of the models wouldn’t be too bad — they would look plausible. They would also predict far, far less than the “average” climate sensitivity of the whole set of models. Some of the models would be absolutely terrible and have no business being included in the figure at all. Some models would be somewhere in between — almost certainly wrong, but not to the point where one can “reject them with 95% confidence”. Since one has many shots, to avoid data dredging in claims of correlation/correspondence one really has to tighten the criteria for rejection, and at the very least those models should be reweighted to have less probable predictive power than the ones in good correspondence.
But if one did anything like this, the models would no longer predict catastrophe, at least not catastrophic warming.
This, in turn, would completely alter the political landscape. We as a species have to solve cost-benefit problems all of the time. We are trying to opimize our use of global resources to the advantage of an emerging global civilization. Like all decision making processes, where there is garbage in, one gets garbage out. Climate science has been manipulated into a Big Lie, or more properly a series of Big Lies.
* Catastrophe is definite and inevitable. Note that this is true even if we alter our use of carbon based fuels — it is “too late” according to the high sensitivity models and the most shrill of the catastrophists.
* Since warming has apparently all but ceased, CAGW has been quietly transformed to CACC — GW has turned into “climate change” so that we can blame humans and claim catastrophe for GASTA neutral weather events like droughts, tornadoes, hurricanes, floods. Since events of this sort happen literally all of the time, somewhere, the claims that any given event is related to anthopogenic activity cannot be refuted. Sandy was our fault. I’m certain that the storm that just hit England was our fault. If there is a large tornado, to Al Gore it is clear proof that it is our fault. Surely large tornadoes never happened before we got around to increasing the CO_2 content of the atmosphere.
Halfway decent scientists who take the time to actually look at the data know that this is piffle, but how in the world is the average citizen going to figure this out, especially when supposedly reputable scientists stand mute when Gore makes his absurd claims in internationally promoted events. The end, apparently, justifies the means even if the means is turning a blind eye to the evisceration of science on national television. Where are climate scientists who are willing to address not their political certainty but their scientific doubt, not the “97% consensus” but the disagreement of some the models with the data at confidence levels even greater than 97% even though those models are still included in summary reports without any caveat — removing the caveat that was once there — for what one can only imagine is the sole purpose of lying to the policy makers, making what are literally the highest possible claims for future warming.
Where is Bayes’ theorem when you need it? Where are actual statisticians pointing out that the mean behavior of 30 models, many of them obviously failed, is not a statistical predictor of the future to which one can assign confidence intervals? Where is simple honesty?
Not in AR5. Definitely not in the only document of AR5 that matters, the summary for policy makers. Because once you shoot down (let’s say half of) the models in CIMP5, the entire document changes radically. Indeed, the credibility of the IPCC and all prior ARs crumbles.
So damn skippy, GCMs are trying to solve an enormously difficult problem. They aren’t doing a completely terrible job given the difficulty of the problem, but they are decades away from being adequate to control the public use of hundreds of billions of dollars at the cost of millions of lives a year, right now, not in 50 or 60 or 80 years. Ignoring the increasing deviation between the models and the actual GASTA record is sticking one’s head firmly into the sand and praying that the future will not simply lop it right off. If AR5 were honest, it would explain this clearly to the policy makers — after throwing out all of the models that are obviously too inaccurate at this point (whatever their supposed virtues) to be used to guide public policy.
Note very well that I hold no position on what the actual climate sensitivity is. I don’t think we know. It isn’t enough to assert that “physics” predicts CO_2 linked warming — all of the GCMs are supposedly physics based, and yet look at the spread of model results! Look at the deviation (of most) from reality! One can easily build a complex model using valid physical theory and get the wrong answer — especially when studying nonlinear systems with complex, chaotic dynamics, especially when those models are accidentally initialized during a period of strong natural warming under the misapprehension that most of that warming was caused by CO_2.
This is the sad thing. We can even see why the models are getting it largely wrong. It’s just that the political cost of fixing the models so that they hindcast the last 16 plus years (which without any question would eliminate most of the predicted warming, drastically dropping climate sensitivity and which might require the addition of omitted physics) is too great to bear. Careers would end. Heads would roll. An “oops” recantation of claims that have caused the diversion of perhaps a trillion dollars and that have indirectly killed tens of millions of people already in the process would go down in history as one of the most shameful moments in the history of science, and — if and when it happens — will affect the credibility of every scientist on the planet because the promoters of CAGW have taken care to connect all of us to their claims of catastrophe and disaster. 97% of us agree. We will therefore all bear the guilt and shame, if only for our silence, if those claims turn out to be false.
The only possible defense against this is simple honesty. AR5 needs to issue a correction — effectively rewriting its SPM after doing a serious statistical review of the GCMs that are allowed to contribute to its conclusions. How can doing this not be good science? You know and I know that it should be done, that honesty demands it. Tell me that you cannot pick a few models in the collection that just shouldn’t be there because they are absurdly too warm and get lots of other things wrong as well. How can it possibly be wrong to only make statements that are justified by the actual theory of statistics instead of pulling phrases like “high confidence” or “medium confidence” out of one’s ass when talking about future “projections” of warming or disaster? How about removing altogether the assertion that any human alive has any idea what fraction of the 20th century warming in either half was anthropogenic and what was natural? How about removing the term “unprecedented” from the report altogether (since it is incorrect or unknown almost every single place it is used, and it is used only to increase the sense of alarm rather than present the scientific facts and hypotheses)?
rgb
thallstd says:
November 4, 2013 at 7:17 pm
Thank you, this is a great reference. Indeed, interpreting experimental data is not a straightforward process. And sciences mainly based on statistical interpretation of measurements, like those described in the article plus climate science, can easily become a casualty of erroneous data interpretation. The referenced article demonstrates that some scientists are aware, “self-doubt” (in a good way), and try to eliminate glitches. The result of this “self-doubt” is positive – rejecting inadequate theories, proposing more accurate theories, etc. However, there is a different phenomenon (that I believe is dominating current academic science): some scientist (like Mann, Hansen, Caldeira) propose a hypothesis, formulate theoretical concept, collect data supporting this concept and form a clique that suppress any attempts of either verification, or finding the range of applicability of their model, or publishing any theories that differ from theirs. What are the reasons for such behavior? How widely spread is such behavior? – We learn from Lee Smolin (read “Troubles with Physics…”) that, similarly to climate science, a string theory clique dominates and suppresses theoretical astrophysics. Is it a consequence of the way of how research funding is set?
John Whitman says:
November 5, 2013 at 3:19 am
Robert G. Brown said,
“…
The only thing that will disprove GW or CGW is reaching 2100 without a climate catastrophe and without significantly more warming or with net cooling. A demonstrated total climate sensitivity of zero beats all predictions or argument. …”
– – – – – – – –
Robert G. Brown,
Thought about your post quite a lot.
A future disproof of GW or AGW or CAGW based solely on more observations versus a disproof right now based solely on past observations (on all timescales).
Before we disprove something, don’t the proponents have to prove it first? What would define and prove that anthropogenic CO2 emissions are having an observable, measureable, effect on the overall temperature of the atmosphere?
There is no need to disprove GW – it has been happening recently (since end of LIA) just as GC has happened many times in the past.
I believe, if Dr. Brown is specifically talking about proving or disproving the models, then his assertions seem correct.
rgbatduke says:
November 5, 2013 at 6:13 am
… How about removing the term “unprecedented” from the report altogether (since it is incorrect or unknown almost every single place it is used, and it is used only to increase the sense of alarm rather than present the scientific facts and hypotheses)?
rgb
Interesting that you would say that.
I would suggest that if you changed the term “unprecedented” to the phrase “not unprecedented” in every incident where it was used, it would still be just as correct (or incorrect or unknown).
rgbatduke comment on November 5, 2013 at 6:13 am, should also be elevated to a Guest Post. And maybe introduced into the Congressional Record.
rgbatduke says:
November 5, 2013 at 6:13 am writes
@ur momisugly@@ur momisugly@@ur momisugly
We will therefore all bear the guilt and shame, if only for our silence, if those claims turn out to be false.
The only possible defense against this is simple honesty.
@ur momisugly@@ur momisugly@@ur momisugly
Once again, you have written a clear statement of the problem. But, with these sentences, you are, I believe, neglecting the realities of what the situation is in the current scientific establishment. Not all of us have been silent. But the scientists who really matter have been silent. And they remain silent.
The cure you propose of “simple honesty”, here on November 2013, is pie in the sky daydreaming. It is NEVER going to happen unless and until the learned scientific societies, led by the Royal Society and the American Physical Society decide that they WANT to be honest. They don’t.
We have the ridiculous situation where the [Astronomer] Royal, and former President of the Royal Society, Lord Rees. can state, publicly, scientific nonsense. http://theconversation.com/astronomer-royal-on-science-environment-and-the-future-18162
I quote “Doubling of CO2 in itself just causes 1.2 degrees warming. But the effect can be amplified by associated changes in water vapour and clouds.”
I would be quite happy to debate Lord Rees and show that he is talking scientific nonsense. But this will never happen.
The scientific establishment is locked into the hoax of CAGW, and is not going to change unless a miracle happens.
Wow, you both got it wrong. CMIP5 (Coupled Model Intercomparison Project) is correct, CIMP5 is incorrect. Right now there is a mixture of both in the head post. I know, it’s a minor nit, but easy to fix.
[“Now” are you happy? 8<) Mod]
All good points except for this one: “Because it is an open secret that there was a cover-up that deliberately concealed this, effectively lying to policy makers, there will be a public scandal. Heads will roll.”
That would be true if we lived under a reality based government, but we don’t. The government we have is more like that depicted in Orwell’s 1984. Government decides what reality is. If government says 2+2=5 then it does. Anyone who disagrees will be tortured until he does agree. In fact he will be tortured until he actually believes 2+2=5.
Sadly a large portion of the public is so dumb they simply accept whatever hogwash the government tells them. They lack the intellectual ability to distinguish fact from hogwash. Even sadder yet, they vote.
To put the current climate alarmism into perspective, let’s look at some longer time scales:
click1
click2
click3
click4
click5
There is nothing either unusual or unprecedented in the current climate, and the Null Hypothesis has never been falsified.
Even those GCMs that can be retuned to incorporate the pause are not fit for popupose for long term projections/forecasts/estimates/predictions (or whatever other words the AGW crowd uses to describe GCM outputs). The most powerful supercomputers do not permit sufficiently small grid cells to accurately model convection cells. Hence they all underestimate rainfall, as a result of which the positive water vapor feedback is overstated and cloud feedback misstated. And AR5 WG1 chapter 7 final draft said this problem was not remedial in the next decade, if ever.
There may be other approaches (cloud superparameterization, tropical rescaling) that could solve the observational problems. But merely selecting the best of a bad bunch is not the way to go.
rgbatduke said:
Robert: Four, immensely powerful words: I felt the incredible emotion behind them. They should be engraved in stone and hung above the doors to every CAGW/CACC ‘scientist’s’ lab.
Clipe above at 1237 4/11/13 shows the piece from Mr FOIA from climategate 2.
Does any blogger/ administrator know what became of Climategate2.
Just curious.
– – – – – – – – –
JohnWho,
Appreciate your thoughts in your comment to me.
When any claims of GW or AGW or CAGW are made then anyone can decide whether to work on disproving them. While intellectual onus of proof of claims may be on those making claims, it doesn’t restrict skeptic analysis by others that leads toward a state of disproving the claims; that is, skeptics can choose to work on disproving claims even though the claims were never even attempted to be ‘proven’ only asserted boldly as true.
As to disproving GW or AGW or CAGW versus disproving the current GCM models, I think they were treated separately in Brown’s lead post.
In my view, disproving the current GCMs does not disprove GW or AGW or CAGW. But I think disproving GW or AGW or CAGW does disprove the current GCMs.
John
I find it strange that some people are willing to accept models that do not work!
That they do not match history, that have no predictive ability what so ever would cause me to sack anyone who used a model output for any reason other than to demonstrate poor programming/logic/theory etc.
The models are wrong.
No model output should be publically discussed unless its hindcasting ability is extremely good.
One could tolerate a model that forecast history well, with a few exceptional events not currently handled well on the basis that the odd error is being addressed.
But here, we have the output of known, failing models, being published, used, acted upon and some people say they a ‘not bad because it is a difficult job to do’!!!!!!!!!!!!!!
You can perform statistical tests on model outputs to see how close they correlate with observations but first we need model outputs that appear in the right ball park before we do any heavy duty analysis.
Lets hope commentators here do not have jobs that involve safety critical work such as flight control systems, where you need to understand the system be for you touch a keyboard……
rgbatduke says:
November 5, 2013 at 6:13 am
This excellent post is as strong as your original post. Thanks again for your brilliant work in behalf of science and humanity.
Hear, hear.
Including 97% of the world’s scientific societies. (200 in total, according to one warmist comment I read.) The well-organized, well-funded warmist assertion machine made sure to make all those organizations toe the line. When the warm turns, those groups are going to have some ‘splaining to do, because ex-warmist journalists will toss the hot potato of blame into their laps.
rgbatduke says:
November 5, 2013 at 6:13 am
“…Careers would end. Heads would roll…”
Dear rgb, Thank you for starting such a needed conversation. It would have been great if the clique science lead to unpleasant consequences. The reality is different: those tenured will retain tenure, the centers opened as a result of recently initiated NSF project “The Network for Sustainable Climate Risk Management (SCRIM)” (such as Penn State’s CLIMA or opened this summer Center on Weather Risk Solutions… also at Penn State;) will remain opened and spend tens of millions of taxpayer dollars, … Dr. Mann and Dr. Caldeira will continue their careers and keep on contributing to the publication rejections, blocking new faculty hire who does not belong to the clique, and occasionally dining with Bono or Gates and delivering talks. Similar situation is in the astrophysics where the String Theory clique took over (Lee Smolin “The Trouble With Physics…”). I know from experience that similar situation is in the nonlinear optics where “Nonlinear Schrodinger’s Equation and Soliton” clique usurped the power. You write: “…will affect the credibility of every scientist on the planet because the promoters of CAGW…” Not only because of them. It seems that all academia is in trouble state. What is the cause of it?!
taxed said:
l don’t think we should be so quick to welcome boring weather.
Because its when the weather does become stable with little in the way of change over the longer term, is just when you do get extremes in climate.
lts what causes deserts to form, and l also think its what causes ice ages to form when the weather gets locked into a certain pattern
========
That is a very insightful comment. Droughts and heat waves happen when the weather gets locked into a certain pattern, for example. How well do climate models predict such patterns in the distant future when weather models can’t even do it into the near future. And the developers of weather models do try to get them right. Pity they don’t get the big funding that climate models do.
</itypo: not CIMP5 but CMIP5
CMIP = “Coupled Model Intercomparison Project”
http://cmip-pcmdi.llnl.gov/cmip5/
Yes, big oops. I think I was confusing it in my mind with ChIMP5, the Coupled humaniform Monkey Intercomparison Project. Or it was a sustained, pre-morning-coffee dyslexic moment. Or early signs of alzheimer’s. Or too much tasty beer over the last few days. Thanks for correcting all of this Anthony.
I really need to finish my coffee and THEN type replies. Thanks also to those that have identified (correctly) Global Average Surface Temperature (Anomaly), that is GAST or GASTA. I just get tired of typing it all out and since this isn’t a formal journal (hell, I was just typing a reply on another thread) I didn’t bother to define my acronyms as a lot of regulars know what they stand for.
One more comment regarding replies above. It was pointed out that one cannot falsify the existing GCMs — and one could have stopped right there, as this is sufficient to demonstrate that they are not scientific hypotheses that should be used for any purpose whatsoever — by merely noting that the warmest ones are 0.5 C or more warmer than the present, diverging by this much over only 16 years. After all, we cannot say that natural variability hasn’t cleverly arranged itself to precisely cancel all of the warming these models represent and that once the natural track of the model returns to — um — some imagined “normal”, the warming will come roaring back on us all at once.
You are, of course, quite right. We cannot say that this is not the case. As you will note, I have carefully not taken a stance on whether or not the CAGW hypothesis is or is not correct, as the correctness of a hypothesis is subject to both verification and falsification. If I hypothesize “the sun will appear to rise tomorrow at my location on the planet Earth”, I cannot say that this is or is not certainly the case, all I can do is make my best guess as to the probabilities and then see how things turn out. Tomorrow morning the hypothesis will without any doubt become certainly true or certainly false, verified or falsified by direct observation.
At this point I have a fair bit of evidence — I personally think a compelling amount of evidence — that the hypothesis is true, so I think it is rather probably true that the sun will rise tomorrow. I can support an argument for it that includes things like astronomical observations, the laws of physics (including the law of conservation of angular momentum and the observation that the earth has a huge angular momentum influenced by only tiny, tiny net torque), an argument for the permanance of observed things like suns and planets (conservation of mass energy), the high probability that the sun will neither explode nor go out in the meantime, and some rather unprovable assumptions concerning my experience of sunrise being a real experience and not a simulation being carefully prepared by an intelligent agent, an “evil genius” bent on deceiving me, perhaps because I’m a power unit in The Matrix. At the end of the day (literally) I expect the sun to rise tomorrow far more firmly than I expect to be alive to see it.
Let’s see what sort of similar argument could be raised for a GCM that has produced a 16 year warm-side, systematic deviation of 0.6 C from the near neutral actual behavior (of GASTA) over that same period. If we assume that the deviation is caused by natural variability, that is — first order, simplest linearized hypothesis — equivalent to stating that it is somehow likely that the actual temperature of the planet would have fallen by this amount over the same time frame if it were not for the compensating warming behavior produced by increased CO_2. Now let’s look at the entire thermometric climate record, and see how many 16 year intervals we can count where GASTA varied by 0.6 C. I get zero, how about you? If we allow for some 10 of those intervals (in HADCRUT4, at least), then it seems that based on the sample we have the odds of having such a rapid change in temperatures is pretty small. If we extend this using the increasingly imprecise proxy GASTA data over the last 1000 years, 2000 years, 20000 years, 5 million years, etc (where we have serious resolution difficulties at the end of this, of course) it looks like the probability of such a rapid temperature change is nearly zero. If we look at HADCRUT4 — currently my “favorite” representation of GASTA if only because they publish an error estimate and have some sort of thermometric number all the way back to the mid-19th century — the linear trend over the entire graph:
http://www.woodfortrees.org/plot/hadcrut4gl/from:1800/to:2013/plot/hadcrut4gl/trend
is less than 0.8 C over the entire curve! If you carefully cherrypick start and end dates you can find a couple of intervals where the temperature went up 0.6 C over 30 years (at half this rate) — but sadly one occurred between 1910 and 1940 before CO_2 was an issue is almost indistinguishable from the second one between 1970-something and 2000-something.
This seems implausible to me. Of course it could be true — no denying it, the world could have managed a state that (without all the extra CO_2, which supposedly accounts for over half of the warming post 1955 according to AR5 — and we could have reached an orbital state that would have started the next ice age in 1980 or 1990, so that global temperatures should really really dropped by perhaps 0.8 C from 1980 to 2013, and would have if it weren’t for all of the anthropogenic CO_2. It does make it a bit difficult to explain the nearly identical warming from 1900 to 1950 vs 1950 to 2000, with a clearly defined oscillation of temperature around the OLS trend (green line) in this figure — quite a coincidence, that. It’s almost as if nothing really changed in the pattern of global climate in 1950 with the advent of anthropogenic CO_2 beyond a slight exaggeration of a preexisting trend we do not understand and cannot retroactively or otherwise explain.
In the end, though, it is quite clear that the actual temperature record post the 1997/1998 ENSO that is associated with the last observed meaningful warming is flat, and it is equally clear that flat temperatures do not support the null hypothesis that any of the models with excessive warming are still correct. They merely may not be sufficient to reject those models at arbitrarily high confidence levels but no one sane will look at them and go “look how well this model is doing” outside of the range of the actual data used to tune the model!
Predictive modeling is one of my professional games. In the actual business, we have a name for a model that systematically deviates from observation the instant one applies it to a trial set outside of the training set. We call it a “failed model”. Not because it is certain to be a failed model, but because if you try to convince a client to put actual money down based on your model they will decline. In fact, if you were forced to put your own money down on your own model, you would (if you had any sense) decline. You would without any possible question doubt your own model, especially given the secular trends in the data that are clearly visible across HADCRUT4. You would probably conclude that your model is quite naive, and overfitting a linear trend on the training set that is coincidentally excessively steep.
So no, I cannot agree with your rationale for keeping the obviously failed models in CMIP5 and moving them all over to AR5’s SPM without any comment like “Figure 1.4 includes models that are unlikely to be correct at this point because they exhibit warming in excess of what any reasonable natural variation could cancel, and spend zero time even touching the actual temperature curve over the last 16 years.” Or without a comment that the “mean” of the models in CMIP5 is a “meaningless” quantity as far as the actual theory of statistics, having precisely zero predictive power compared to the actual temperature, meaningless variance, no possible way to connect the mean and variance to a probability of future warming.
Just because one can imagine an argument that would permit a failed model to still be correct is not an excuse for keeping it. Sure, it could still be correct, and as we wait for the future to unfold we’ll no doubt discover whether or not it is. At the moment however, the data strongly suggest that it is not correct. In no possible universe could you look at one of the failed models vs actual curve and tell me “wow, what a great, accurate, precise model”.
So why, exactly, are we basing AR5’s SPM on all of the models in CMIP5 instead of the actual behavior of the actual temperature? What exactly does the “I” in this acronym stand for? Intercomparison of the models themselves as a standard for the believability of the models, or the more proper comparison of each model, one at a time, to the actual climate?
How exactly can you validate a model with another model?
rgb
I know from experience that similar situation is in the nonlinear optics where “Nonlinear Schrodinger’s Equation and Soliton” clique usurped the power.
Yeah, but that’s because solitons are way cool, right? And when you say “usurped the power” it doesn’t mean anything like it does in climate science. One could still get funded and published outside of this — at least I did, where my own work involved microscopic coupled Langevin descriptions of linear optics that reproduced macroscopically nonlinear descriptions with a model where one could track in microscopic detail the disorder and feedback for things like bistability and photon echos. Then I got sidetracked for a decade plus into (dynamic) critical phenomena, which is also quite interesting. Physics can be clique-y and “wars” between competing points of view and a certain amount of gatekeeping have been known to occur, but since journal referees are usually not “tightly” in the field, they don’t really prevent good work from continuing or cause a field to truly become monolithic. And in the end, nobody dies if solitons beat out master equations for a while in nonlinear optics.
rgb
rgbatduke, you ask “How exactly can you validate a model with another model?”
You cannot. Many of us have known this for years, but few people took any notice of us. Now we have someone of your standing who is starting to ask the right questions. You seem to find the answers surprising. We skeptics have been listening to this scientific nonsense for years.
What I hope is that when more scientists like yourself realise just how bad the science is that the IPCC has been presenting for years, that this realization will creep higher up the pecking order, until someone who really matters starts asking these questions, and finds there are no answers.
The bad science in the AR5 seems to be slowly percolating through the scientific community. Let us hope that this continues, until something actually happens to lay bare the hoax of CAGW.
“…since journal referees are usually not “tightly” in the field, they don’t really prevent good work from continuing or cause a field to truly become monolithic. And in the end, nobody dies if solitons beat out master equations for a while in nonlinear optics.”
I agree, nobody dies, just science. In 1961 Leo Szilard (who is practically unknown these days to most of the physics students) published a science fiction book “The Voice of the Dolphins”. One of the short stories in this book was about intentional retardation of science – “The Mark Gable Foundation”. He didn’t anticipate that situation described in his story will come true and further evolve into clique science and clique academia.