Bishop Hill Von Storch on the models again
Aug 23, 2013 Climate: MWP Climate: Models Climate: Zorita Climate: von Storch In a post at Klimazwiebel, Von Storch and Zorita have expanded onf the views they put forward in the discusion paper I mentioned the other day.
We want here to set straight some misinterpretations that may have arisen in the blogosphere, e.g. Bishophill, and may also have been present in the review processes by Nature as well.
The main result is that climate models run under realistic scenarios (for the recent past) have some difficulty in simulating the observed trends of the last 15 years, and that are not able to simulate a continuing trend of the observed magnitude for a total of 20 years or more. This main result does not imply that the anthropogenic greenhouse gases have not been the most important cause for the warming observed during the second half of the 20th century. That greenhouse gases have been responsible for, at least, part or even most of the observed warming, is not only based on the results of climate simulations, but can be derived from basic physical principles, and thus it is not really debated. It is important to stress that there is to date no realistic alternative explanation for the warming observed in the last 50 years. The effect of greenhouse gases is not only in the trend in global mean near-surface temperature, but has been also identified in the spatial pattern of the observed warming and in other variables, such as stratospheric temperature, sea-level pressure and others.
Although von Storch and Zorita talk of misinterpretations, I'm not sure there is any great difference between what they say here and what I concluded last week. At the time I said that the models are falsified - they run too hot. A model that had a slower rate of warming would not be. So a claim that part of the observed warming is anthropogenic is still scientifically tenable. How big a part is manmade is, given the failure of the models, anyone's guess.
I'm not, however, convinced that the only plausible explanation for the warming of the last 50 years is greenhouse gases. As I mentioned the other day, the IPCC looks set to conclude that there was a Medieval Warm Period. Last time I heard Rob Wilson discuss the matter, he said that the climate models couldn't even get the MWP in the right historical position, let alone explain its apparently large magnitude.
So once again I find myself returning to the point I make so often. The unknown unknowns are a big problem. Scientists would do themselves a favour if they recognised it.
[Comments will be tightly edited for relevance and tone]
Reader Comments (89)
"and thus it is not really debated"
Well, maybe it ought to be. Maybe some clever person ought to work out a better way of finding whether this is happening or not. Like an experiment or some real-world observations, because once you find out that a model, indeed every model, is wrong, you can't really make that your prime evidence.
In any sensible logical approach you'd look at the problem from a contrary stance, you'd wonder whether your assumptions were mistaken. You'd look for a fingerprint to check theory against reality.
You would not say, for public consumption, "and thus it is not really debated".
Rhoda: I agree it ought to be debated. But what?
Emphasis added. At least part. That could be 2%. Which also ought to be debated. But isn't it time that Von Storch and Zorita agreed expliclty with Andrew:
Instead of using vague language that sounds as if it's saying something worthy of debate but the moment you poke it it collapses into a heap of nothing.
derived from basic physical principles, and thus it is not really debated: a rhetorical sentence, almost void of information
there is to date no realistic alternative explanation for the warming observed in the last 50 years: von Storch, or von Daniken?
The effect of greenhouse gases [...] has been also identified in the spatial pattern of the observed warming: Arctic amplification, I presume? What else?
and in other variables, such as stratospheric temperature, sea-level pressure and others.: we are back to the "tropospheric hotspot" argument. Is Von Storch really claiming this other stuff is _unique_ to GHG warming?
Perhaps he is. And so we'd finally have something to falsify GHG theory against.
If scientists cannot produce a computer model which replicates temperature fluctuations, especially the location of the peaks and troughs, over the last 2000 years, then they cannot predict the future. I cannot think of any other area of modelling where such differences with observed data would be acceptable.
What exactly are the "basic physical principles" that he refers to when all around there are people — some within the "climate science community", some with other equally serious qualifications that imply a knowledge of the laws of physics — who dispute in whole or in part that CO2 (because that is what "greenhouse gases" means in this context) has the ability to do what the models are demanding of it.
It appears to me that he is going back, in different words, to the warmist argument that "it must be CO2 because we can't think of anything else".
And we, inevitably, come back to the sceptic argument that "it's CO2 because you've told the models it's CO2 and the feedback is positive because you've told the models it's positive".
With increasing evidence from the real world that the models are (read my lips) w-r-o-n-g, at least over the last decade+, rhoda has it spot on when she says that the logical, scientific approach is to step back and try to view the problem from the angle that "it looks as if we have this wrong. Why is that?"
so von Storch's position is:
we can't explain the current pause and our models based on the CO2 theory, are inadequate. However, we can't think of anything else so it must be CO2 that is driving the climate!
Good job we had real scientists in the last century or Quantum Theory might not have been developed!
Of course it can be accepted that greenhouse gases have been partly responsible for the late 20th century warming. No one disputes this. But, without assuming positive feedback enhancing the water vapour content, the degree of GHG warming will be small. And, as sceptics are tired of saying, "what caused the MWP?" And what caused earlier recorded warm periods?
As a retired experimentalist I find it difficult to accept the veracity of the climate models. They seem plain wrong, being unable to predict the current GMST hiatus. I would suggest that they are getting close to being falsified.
Falsified models failing to disprove the hypothesis is evidence that the hypothesis is unfalsifable.
From the quote:
This demonstrates there are two hypotheses to consider. The first is a basic principles more CO2 = more greenhouse gases = warmer. The basic 1.2 degrees C from doubling CO2 levels. This does not bring with it much cause for concern nor the opportunity for major policy decisions and political power.
The second is the alarmist hypothesis that has the net feedbacks from higher CO2 and warming being positive, leveraging that 1.2 degrees C per doubling of CO2 to potentially much higher levels. This provides ample scope for major policy decisions and political power.
The former has not been falsified yet. The latter, it would appear, has. This ought to have significant implications for policy decisions but I doubt it will.
From the quote:
What caused earlier warming could also have caused recent warming. Given the uncertainties in knowledge I find it hard to swallow that the warming in the first half of the last century and earlier is understood well enough to discount those same factors causing later warming.
Yes I agree with commenters. Von S can bang on all he like about basic physical principles, but they do not trump basic failures of reasoning.
I'm still waiting for the derivation from basic physical principles.
“It is important to stress that there is to date no realistic alternative explanation [than greenhouse gases] for the warming observed in the last 50 years.”
Why only 50 years? If you look at a plot of hadcrut3 against the natural variation of the Atlantic Multidecadal Oscillation since 1880, the close correlation over the entire period is striking, see http://www.woodfortrees.org/plot/hadcrut3vgl/from:1880/plot/esrl-amo/from:1880.
Why couldn’t the same natural warming effect of 1918 to 1945 have been the cause of the warming in the 80s and 90s, rather than atmospheric CO2 as claimed by the IPCC? The suspicion is that the IPCC discounted man-made CO2 as the predominant cause of the 1918 to 1945 warming period not for valid scientific reasons but simply because if they hadn’t, they would have had difficulty explaining the 1945 to 1979 cooling in the face of ongoing CO2 forcing, just as they are now having difficulty explaining the unanticipated temperature standstill of the last 16 years in the face of rising atmospheric CO2 levels. As it is, their explanation for the cooling from 1945 to 1979 smacks of pulling a rabbit out of the hat, namely that it was due, uniquely for that period, to atmospheric aerosols.
As I understand the theory:
1) Light comes into the atmosphere.
2) Some is absorbed by the greenhouse gases in the atmosphere.
3) This energy is released again thus allowing the Earth's temperature to tend to equillibrium by a black box of processes (clouds, air flows, ocean flows, unknown flows...).
4) Increasing the greenhouses gases will increase the point that the Earth's temperature tends towards (the equillibrium).
But as the black box is unknown (even Trenberth acknowledges that heat can get into the deep ocean) then the affect of greenhouse gases is not known from basic physical principles.
I'm reminded of the old chestnut about science 'proving' that bumblebees can't fly. It's actually an important teaching point, I think - if one starts with certain assumptions about aerodynamics etc and does the calculations for a bumblebee, one finds that the bumblebee ought not to be able to fly; a very important result, for it shows that the assumptions one started with do not hold for bumblebees, and other effects must come into play for them.
Similarly, we heard for years that 'it must be CO2 because we can't think of anything else' (I recall watching a climate scientist say that, pretty much word-for-word, on Newsnight around the time of Climategate, though I can't find it on YouTube). When reality started to diverge from the models, suddenly they could think of other factors - the deep oceans, aerosols, whatever - to explain the pause. Clearly, they now need to go back a re-examine the original assumptions built in to the models.
I've always felt that the models weren't very useful for global temperature predictions. They may provide very interesting insights into regional effects etc, but if you build a climate model based on a finite set of known physical principles, you can surely be pretty confident of its large-scale, average results before you even run it - increased CO2 -> increased temps, etc. If relatively simple treatments of the radiative physics show that increased GHGs trap heat, then it would be surprising indeed if supercomputer scale models based on the same physics showed a different result. It's the fact that reality is behaving differently that is really interesting.
How about this: from basic physical principles it is highly likely that the temperature over the last 150 years was warmer than it would otherwise have been without human CO2 emissions. The current temperature and future temperature should also be warmer than they would otherwise be. Other human effects may increase (urbanisation) or decrease (aerosols) temperature. As we cannot currently model the climate over extended periods with a useful degree of accuracy we cannot state whether these human effects are beneficial or detrimental.
I think that stands unless there is some effect from increasing CO2 that is so negative it negates the modest warming from the "basic physical properties".
Omnologos has the critical part pegged:
but has been also identified in the spatial pattern of the observed warming and in other variables, such as stratospheric temperature, sea-level pressure and others.
All of the supposed greenhouse gas fingerprints are either absent or seriously questionable (you can't claim polar warming as a fingerprint when only one pole is warming). It is the absence of these that drives the biggest nail in the CO2-induced warming theory since the actual temperatures are always going to be affected by many variables and so would not be expected to follow any kind of smooth change.
That a scientist (I use the term advisedly) like Von Storch uses the "we can't think of anything else, therefore it must be CO2" argument is pathetic - he really should hang his head in shame.
Raise the tone please.
From basic physical principles:
Cooling of the Earth's surface is multimodal, with non-radiative (evaporation and convection) fluxes dominating in average. The atmosphere, on the other hand, can lose the (predominantly non-radiatively) gained energy from the surface only by IR radiation to space and only the so-called GHGs (and clouds) can do that. The bulk of the atmosphere (N2 and O2) cannot radiate significantly and therefore insulate the atmosphere and the surface. N2 and O2 are the real GHGs and the radiatively active gases (and clouds) are the 'coolants'.
http://science-edu.larc.nasa.gov/EDDOCS/images/Erb/components2.gif
'That greenhouse gases have been responsible for, at least, part or even most of the observed warming, is not only based on the results of climate simulations, but can be derived from basic physical principles, and thus it is not really debated. '
Those same simulation which fail because in reality the 'basic physical principles , are both far from basic and well know. Basic science this stuff , if you do not know what the cause and effect relationship is, you cannot make statements about cause and effect worth a dam.'
'That greenhouse gases have been responsible for, at least, part or even most of the observed warming, is not only based on the results of climate simulations, but can be derived from basic physical principles, and thus it is not really debated. '
Those same simulation which fail because in reality the 'basic physical principles , are both far from basic and well know. Basic science this stuff , if you do not know what the cause and effect relationship is, you cannot make statements about cause and effect worth a dam.'
"That greenhouse gases have been responsible for, at least, part or even most of the observed warming, is not only based on the results of climate simulations, but can be derived from basic physical principles, and thus it is not really debated."
"Not really debated". Oh, really?
"Can be derived from basic physical principles". The black-body model frequently presented is obvious a gross over-simplification and is no more than a plausibility argument.
At Judith Curry's blog there was discussion about this and her conclusion was that between the grossly simplified model and huge computer simulations there is nothing. For example, a graduate-physics level analysis.
Martin's conclusion: Neither "basic physical principles" nor computer models provide evidence that CO2 caused the recent (now halted) warming.
Isn't this a circular argument.
1) The attribution case (derived from models) shows CO2 drives climate.
2) Test attribution case by testing models outside training sets - using the last 15 years of data.
3) Models diverge significantly from observations
4) The attribution case -> go to 1.
Just more arm waving, now that their pals started to lean on them.
Yes, sounds to me that Von Storch had a phone call from the Team.
I don't understand exactly what you are saying here. Are you saying there is a cut off - a model that runs too far from reality is falsified? But how far is too far? What is the cut off? How is it evaluated? Unless you are saying that the model must match exactly - that the cut off is zero. But in that case there would never have been any unfalsified models in the history of science, and even if there were we would not be able to know it because we could never measure with infinite precision.
Chris Long: 'it must be CO2 because we can't think of anything else'. That was said on telly by Bob Watson (ex-IPCC chair etc) when he was chief scientific advisor to Defra. I wrote to Defra complaining about an advisor spouting such nonsense, but I got no response from them.
Martin A: It sounds exactly like that. But the guy has some remnants of a scientific conscience. Hence the 'at least part' I highlighted earlier. So that when the catastrophe-predicting/grant-milking phase is finally over, the globally averaged temperature anomaly is trundling along up or down, in a very unscary way, and extreme events are no worse and even less fatal than they were ... when almost the whole building has come down, Von Storch can say "but I did say it might only be part".
It's moral courage meets infinite regress. Forgive me if I can't manage a standing ovation.
Anybody caring to discuss how Von Storch's "It is important to stress that there is to date no realistic alternative explanation for the warming observed in the last 50 years" relates to this slightly older way of arguing?
At least they are narrowing their argument to its realistic value: a few parts per million, at most.
Anyway, why do they need that they have to stress their adherence to orthodoxy so vehemently? There is nothing wrong with questioning everything in science.
A direct conclusion from their paper and subsequent explanations is that models are worthless in predicting future scenarios, therefore all the hype about Cagw should vanish.
Finally, BH says
Many of us here are scientists, even in fields somehow related to climatology, I am sure we don't have any problem in recognising the very large imbalance between known and unknown.
"It is important to stress that there is to date no realistic alternative explanation for the warming observed in the last 50 years."
And 1910 to 1940 will be ignored because climate "scientists" have hard time saying "natural variability" out loud.
Von Storch has some difficulty with the models but states;
"It is important to stress that there is to date no realistic alternative explanation for the warming observed in the last 50 years."
Change 'alternative' to 'model' and we are all, including Von Storch, in agreement. (Although 'warming' would be better as 'temperatures', but then he would say that, wouldn't he).
omnologos
Many thanks for that link.
The only reason I don't laugh at the concept is that I am well aware that that thinking would have been considered fairly mainstream in the late sixteenth/early seventeenth century. I doubt if you would find many believers today (outside the more rabid strict creationists) who believe that God micro-manages the universe in that way.
But since we now know that the events referred to were "natural" even if a bit extreme why should we suddenly suspend our critical faculties 400 years later and assume that equally natural and arguably less extreme events are the direct result of an increase in an essential atmospheric trace gas — which is also (apparently) a punishment for man's sinfulness?
I think of the two I'd prefer the God version!
Bish, you say this: "I'm not, however, convinced that the only plausible explanation for the warming of the last 50 years is greenhouse gases."
Absolutely.
It would be completely implausible that CO2 (and black carbon, and methane, and tropospheric ozone) would have no or very little effect on rising temps. I agree with von Storch and Zorita on this, there are basic physical principles involved.
I feel equally strongly that man made emissions are not the exclusive or near exclusive cause of the warming, there has been too much natural variability in the recent past when CO2 had barely begun to creep up.
My take, similar to that of Judith Curry, is that about half, give or take, of the warming in the last half century is due to CO2 and its warming cousins. This guesstimate is based on the new studies on lowered climate sensitivities, and on the relationship of actual temperature trends vs. modeled. Here is Curry's take, in an NPR interview:
http://www.capradio.org/news/npr/story?storyid=213894792
This is sort of like Occam's razor -- the simplest, clearest explanation make the most sense. Do you believe, on the basis of additional complex models, like the ones that are failing to predict actual temperatures, that the warming has somehow escaped all our notice, and is hiding in the deep sea, undetected by all the Argo and other sensors as it made its way down? Or are we seeing something like the reverse of the rapid ramp up of temps from about 1915 through 1940, roughly? In other words, can we believe our lying eyes?
Mike - they had no realistic alternative explanation, therefore they presumed it was God's punishment for sinners.
We haven't changed a bit.
omnologos
Isn't a bit ironic that those of us who actually believe in a God have indeed changed and it's the new religion of climate science that appears to have reverted to the superstitions of the 17th century?
From Von Storch we have:
"It is important to stress that there is to date no realistic alternative explanation for the warming observed in the last 50 years."
The fact that we have no alternative explanation should not prevent us from recognizing that our existing explanation, rising CO2 plus large positive feedbacks, has been falsified or proved unworkable. Von Storch is saying that CO2 is such an appealing explanation that he cannot surrender it even though it is false or unworkable.
What is needed is some humility. Fortunately, Judith Curry shows the way when she says that we do not know what the explanation is.
Very interesting point Mike. I'd only add that in changing, believers, in the UK at least, have been going back to more radical traditions within church history, such as the early Celtic church. How the devout scholars of Iona saved civilisation on the decline and fall of Rome and all that. There's much still to learn in our time. And we still need to change.
To omnologos, Mike, Richard and everyone,
Belief in God, as God is defined in Christianity, is a very powerful source of humility. What we have seen in mainstream climate science is belief in one's self or one's group that is truly unshakable. Yet humility is essential to scientists. Feynman always taught that the scientist must publish all of his results that conflict with his hypotheses along with the results that support his hypotheses.
Theo: Thank you. I've always felt this deep affinity with Feynman, who was convinced he didn't believe. But the humility thing he got. Someone once said that God raises up those kind of people. Happily I don't have any of the answers or my resultant pride would rule me out of all the fun. :)
"other variables, such as .. sea-level pressure"
Has it changed more than usual then? News to me...
Richard,
Very well said. You are a solid member of the Feynman camp. I have learned a great deal from your excellent posts at this site and elsewhere. I am honored and very pleased that you responded to me. Because this is the Bishop's site, I will refrain from additional statements that might have implications for theology.
The basic mechanism(s) associated with radiative energy transport in a homogeneous mixture of gaseous interactive media are well understood. This characterization of the 'first principles' understanding of the basic mechanisms, however, is incomplete and deeply misleading when the physical phenomena and processes of interest are the earth's climate system.
The hypothesis relative to earth's climate system includes also an equally important concept that is seldom mentioned in the bumper-sticker / press-release version of climate change. This concept is an equilibrium state for the radiative energy transport at the top of the atmosphere (TOA). The incoming SW radiative energy is always equal to the out-going LW radiative energy.
The radiative energy budget at the TOA cannot be in a state of radiative equilibrium. There will always be time periods for which the incoming energy is greater than the outgoing and time periods for which the out-going is greater than the incoming. This behavior is not due solely to 'feedbacks', but instead is a function of the daily, seasonal and yearly cycles of the earth's climate system.
What might be meant, but is also seldom clearly stated, is that eventually at some unspecified future time there will be extensive time periods for which the incoming energy will be greater than the outgoing and that the oscillations will occur at an energy level that is greater than that obtained when less radiatively interactive gases were present in the atmosphere. And further that a radiative-energy-transport equilibrium state will be obtained at this higher energy level. And additionally further that for long time periods following getting to a new state that the periods of out-going energy will not ever be sufficient to reduce the energy content of the system to a level less than the new state for long time periods.
However, the systems internal to the earth's climate system are not at equilibrium, and it is impossible for these to ever be at equilibrium. The states of these systems also influence the radiative-energy transport, and other means of energy transport and storage that effect the energy budget of the earth's climate system. So long as these systems are not at equilibrium, the energy budget at the TOA will reflect the changes that are occurring due to the lack of internal equilibrium. And so long as the gradients of driving potentials remain the same before and after a change in the state level the interactions between the sub-systems will remain the same.
None of this reflects the interactions of humans, other than changing the concentration of CO2 in the atmosphere, that can also change the radiative energy budget; radiatively interactive solids as aerosols for example.
The earth's climate system presents a multi-physics, multi-scale (both temporal and spatial) problem. The mathematical description of some of the critically important small-scale phenomena and processes also presents its own difficulties due to the inherent complexity of these. If they are critically important, they must be, without exception, accurately described in the continuous equation domain, that correctness carefully / consistent carried over into the discrete-approximation domain, and the solution of the discrete approximations accurately resolve the important small-scale phenomena and processes, and all this correctly utilized in the application domain by qualified users.
At the present time, several of the critically important small-scale process associated with radiative-energy exchange are not at all described at anything close to first principles. Instead these are described by parameterizations, some of which are based on empirical data obtained when the system was in a state nearby to the states of interest. However others are of a more nearly ad hoc nature. If we stick to only the radiative-energy transport effects of the phases of water, all of the related phenomena and processes are described by parameterizations, especially all aspects of the all-important clouds. And let's not get into solids, including ice, aerosols. Vertically convective motions of clouds, and the associated radiative-energy-transport aspects, are all parameterizations, if I correctly understand.
The Earth is a water planet, after all.
Feedback is not the problem. Lack of the assumed equilibrium is the problem.
"Comments will be tightly edited for relevance and tone": on the whole, Bish, I'd rather you did it sober.
Chris Long
Be wary of the bumble bee flight analogy disproving modelling per se. The problem was that the behaviour of fluids like air at a macro scale is significantly different to at insect sized modelling. To a bumblebee, flight is like swimming through molasses while those effects are insignificant to a jumbo jet.
But overall, there are serious faults in a model if it can't hindcast. "No realistic alternative" is really meaning we don't know.
Why is nobody claiming that low solar activity in the 21st century is allowing extra cosmic rays through and causing the slowdown in warming?
EM, because we don't all think we know. And that's why we tend to be suspicious of those who claim to be sure when the evidence presented seems to fall short of certainty.
"...have some difficulty in simulating the observed trends of the last 15 years, and that are not able to simulate a continuing trend of the observed magnitude..."
They know its a flat-line but just can't bring themselves to say it.
All models are wrong. It's doubtful that any are actually useful!
rhoda
We've been round this merrygoround before. You make ignorance a virtue. All you can tell the policymakers is that you believe nothing is happening, without evidence to back your hypothesis.
There is considerable evidence for AGW, which (out of respect for Mr. Montford's ban on radiative physics) cannot be discussed here. The policy makers have chosen to accept that evidence and act on it.
If you want them to accept your view, you need evidence to support it. Vague statements about doubt, unsubstaniated comments on uncertainty and a refusal to accept empirical data because it is not 100% proven (an impossibility in science) are not enough. To be convincing you need to show better evidence than the climate scientists, instead of the vague waffling and unsubstantiated standard tropes which are the usual fare of spin-sceptic propoganda.
They know its a flat-line but just can't bring themselves to say it.
Aug 23, 2013 at 10:00 PM | ssat
This is one of the unsubstatiated standard spin-sceptic tropes I referred to in my reply to rhoda.
Look at the table at the end of this page.
http://www.metoffice.gov.uk/research/monitoring/climate/surface-temperature
Note that the warmest year on record, by all four datasets, is 2010. 2005 is second and 1998 is third.
The rate of change has slowed, but not stopped.
Perhaps the Bish should ban anybody who mentions measurements of any kind without the faintest idea about uncertainties
omnologos
In every paper reporting climate measurements you will find error bars or their equivalent included, to show the uncertainties. This is standard scientific practice.
If you claim that these are too small, then go to the raw data, which nowadays is normally made available to other researchers, and calculate your own values. Dont tell us the uncertainties are too small. Show us.