Bishop Hill Detection, attribution, disintegration
Feb 20, 2013 Climate: D&A James Annan hit the climate blog headlines the other day, with a post that not only wrote off the possibility of high climate sensitivities but also revealed that one climate scientist had been so sure of the omerta among his colleagues that he openly admitted to lying to promote political action on climate change.
Annan has now written another potentially blockbuster post, which discusses recent publications in the area of detection and attribution (D&A) - the bit of the climate change science that assigns guilt. What Annan reveals is that the studies that have supported the claim of "the majority of recent warming is manmade" are fatally flawed.
I've been fairly critical of the conventional D&A approach in the past, primarily on the grounds that the null hypothesis of no anthropogenic influence is always false a priori (and therefore a failure to detect an anthropogenic influence is always a matter of insufficient data). These recent papers point to another, arguably more terminal, problem. Attribution will inevitably fail as the anthropogenic effect increases!
Read the (relatively technical) reasons why here.
As Annan explains, this is going to leave D&A specialists with a headache to rank alongside the one being ensured by the climate sensitivity team.
It will be interesting to see how the D&A community addresses this problem. Atribution of the observed changes to GHG and other influences was touted as a major step forward when it was first achieved, so it would surely be rather embarassing to lose the ability to do this. It looks a bit like they are trying to just ignore it for now, but that can't really be tenable as a long-term strategy.
Reader Comments (120)
I, too, have wondered if R Betts gains acceptance here in spite of making unrealistic claims for Met Office modelling rather than because of it.
To me, a model is such a simplistic approximation of reality that only those who share a quasi-religious vision can believe in models.
Roy,
Like the prophecies of Nostradamus, they're claimed to be good at predicting things which have already happened, but they're not so hot when it comes to the future.
Quite right, sorry.
Roy -
1st point
There really is no doubt that this is what the Met Office claim validates their models - they state it in several publications.
2nd point.
For a mathematical/computer model of a physical system to have any hope of predicting the future behaviour of a physical system it is necessary that it is a correct representation of the physical effects within the system. (Even then, it is not guaranteed that it can predict the future behaviour with any accuracy - small errors may become magnified until they dominate the picture .)
3rd point.
Checking that the model reproduces the past history of the system being modelled provides no direct verification that the physics have been correctly modelled. As I said earlier, a simple lookup table containing records of the observed past is useless at predicting what will happen in the future, although it can replay past history accurately.
Suppose I had the mistaken but sincere view that the Earth's climate was equivalent to a set of invisible electrical resisances and capacitors, whose natural behaviour is (from standard electrical circuit theory) known to be correctly modelled by a set of exponential fuctions with appropriate parameters. I could twiddle the parameters to give a good fit to known climate history. But because my model in fact bore no relation to the actual physics, it would actually be useless at predicting future climate, even though I had tuned it to replay the past with fidelity.
------------------------
I think this is pretty much an analogy of what has happened with the Met Office. Ten years ago, they were confidently predicting multi-degree global warming. Now they are scratching their heads, trying to figure out what happened, still claiming full confidence in their accuracy of their models (but making less noise about it).
The Met Office goes on about how their models are derived from the actual physics. But they are a bit coy about how effects that are incompletely understood are "parameterized" and incorporated into the overall model.
Martin A. I do believe there was also a very good correlation with pirates. Hard to justify putting it into the model in the first place mind, but hey? Could lead to a new definition of climate sensitivity based on the doubloon of CO2...I'll get my coat.
http://www.forbes.com/sites/erikaandersen/2012/03/23/true-fact-the-lack-of-pirates-is-causing-global-warming/
Don Keiller "@geronimo I agree "we should give them space to backtrack and get the truth out."
And when they have then stamp heavily all over them in the same way as they have done to sceptics.
Reapplication for tenure would be a start."
There's not doubt that many of these people deserve stamping on heavily, but in any given conflict we should have a view of what victory looks like. For me victory would be rolling back the legislation that is going to cause massive damage to our people and our economy. If to gain that victory I have to ignore the malfeasance of a few miscreants, who will, in any case be consigned to the dustbin of history and the science hall of infamy, then so be it. As much as I'd like to see them pay for their arrogance it's ultimate victory that matters, and if we delay that victory by keeping people who would otherwise help us achieve it by crossing the line back into real science through scoffing at them or punishing them then we are doing ourselves no favours.
By the way, your exchange with Julia Slingo on how to do statistical analysis should be on the wall of every bureaucrat that tell's us that 97% of scientists cannot be wrong.
I've often pondered about the epithet "citizen scientists" that these arrogant people apply to those who question their dodgy science which itself is based on dodgy statistical analysis. Does that make the likes of Julia Slingo a "citizen statistician"?
Paul Matthews: Andrews bio - lukewarmer -
I wonder when Andrew put that there..
My twitter bio has been about lukewarm thougts since I started it:
https://twitter.com/BarryJWoods
Barry: Where you lead, others inevitably follow :) But I second your question: when did the Bish change his bio? I'm sure Lukewarmer wasn't how it began a few weeks back. We demand to know!
Martin A: Completely agree, well said.
Alexander K: There are various reasons Richard Betts has been welcomed here, in my view, not least his unfailing courtesy. Except one time I remember he wasn't too kind to AlecM. But that I think is a special case.
Martin A
"The Met Office goes on about how their models are derived from the actual physics. But they are a bit coy about how effects that are incompletely understood are "parameterized" and incorporated into the overall model."
a reminder of the level of scientific understanding from AR4
Table 2.11 from the Fourth Assessment report of the IPCC sets out the uncertainty assessment of forcing agents. Listed below are the agents and the level of scientific understanding
LLGHGs – high
Stratospheric ozone – medium
Tropospheric ozone – medium
Stratospheric water vapour from CH4 – low
Direct aerosol – medium to low
Cloud albedo effect (all aerosols) – low
Surface albedo (land use) – medium to low
Surface albedo (BC aerosol on snow) – low
Persistent linear contrails – low
Solar irradiance – low
Volcanic aerosol – low
Stratospheric water vapour from causes other than CH4 oxidation – very low
Tropospheric water vapour from irrigation – very low
Aviation induced cirrus – very low
Cosmic rays – very low
Other surface effects – very low
This is, of course, the known unknowns. One suspects that the list of the unknown unknowns (which includes how these forcings all interact) is even longer.
And Richard Betts would have us believe that the climate can be modelled. I hope he is nice and warm unlike the millions in this country who have to live in the cold because they cannot afford to keep warm.
"And Richard Betts would have us believe that the climate can be modelled"
In a rare moment of lucidity, before a descent into madness, even the IPCC admitted it was impossible:
"In climate research and modelling, we should recognise that we are dealing
with a coupled non-linear chaotic system, and therefore that the
long-term prediction of future climate states is not possible.” IPCC
From the 3rd IPCC report, Section 14.2.2.2 “The Climate System”, page 774"
Dolphinhead wrote:
Table 2.11 from the Fourth Assessment report of the IPCC sets out the uncertainty assessment of forcing agents. Listed below are the agents and the level of scientific understanding.
In many of the cases listed in that table our understanding is described as low. How on earth could anyone claim that the science of global warming is settled when even many advocates of the theory of CAGW admit that there are many things that they do not properly understand?
It seems to me that climate prediction is a classically ill-conditioned problem.
Yes, given a set of inputs and a set of outputs one can compute a model that relates the two.
However, if the problem is not posed correctly, vary small variations in inputs or outputs will create large changes in the parameters of the model in question. In essence, the model is ill-conditioned.
Hence, given a model, small variations in the future inputs will give large changes in the predicted outputs, which will not be accurate.
Geronimo.
I'm in Thailand too - I live in Pattaya. Give me a shout if you fancy a beer.
Cheers!
(P.S.: Thailand has a lovely warm climate...)
Lots of people here promoting the null hypothesis that the warming trend of the 20th century was not anthropogenic.
What is missing is anyone able to supply a viable alternative to cAGW as an explaination for the warming. Where did all that accumulating energy come from?
EM, would that be the energy that has not been 'accumulating' for a decade and a half? Despite increasing CO2 levels? Please tell us more about this alleged accumulation - location would be a start. Maybe we could send somebody to measure it.
EM. You are definitely not keeping up with the programme. Thre heat is missing and it's a travesty. No one can find the accumulated heat you speak of. Apparently it's hiding in the deep oceans having dodged the 3000 Argo buoys to get there. I swear to God future generations won't believe serious scientists tried this bunkum.
geronimo,
I reckon it's all been a plan to relieve Trofim Lysenko of his status of having the worst anti-science record of all time. In fact, through the abuse of some novel principal components analyses some self-styled Nobel laureates have stumbled across ways of making Lysenko look relatively talented.
It seems to me that the commenters on this thread are not grasping Annan's point. I don't think he is expressing personal doubt about AGW. He is raising a general methodological objection to classical hypothesis testing of a hypothesis that is known not to be perfectly accurate. One way of looking at this problem is as a confusion between statistical significance and practical importance.
At the risk of compounding the confusion, let me take an example from economics to illustrate. One theory in international trade theory is the doctrine of "Purchasing Power Parity," which says that despite floating currencies the relative prices of tradeable goods are the same (correcting for transportation costs) across all countries. So a ton of steel should trade for the same amount of oil everywhere under this theory. When people go out and test this theory with the abundant data available on tradeable commodities over time, it is always refuted at any confidence level you want. But defenders of the theory may say, "Yes we know that these price ratios are not EXACTLY the same everywhere, but they are tolerably close--around 1.x" where x is some number that seems pretty small to the defender of the theory. In other words, we can prove that to a statistical certainty that the theory is not exactly right, but we knew that to start. What we want to know is whether the theory is a good approximation to reality.
The classical hypothesis test is useless for that. You need 1) some sense of the probability distribution of the values and 2) JUDGEMENT about "how close is close enough" for the purpose the estimates are intended for.
@Richard Drake, re chocolate teapots
" . . . statues made of matchsticks." Dylan
Annan has definitely wandered off the reservation, as we say here Out West. Gone rogue. His honesty is refreshing. I'd give good money to hear what Mann, Trenberth and Jones et al are saying about this in private.
Richard Drake, I fully recognise Dr Betts as unfailingly polite, but politeness as a personal quality is irrelevant to the topic.
When I begin scanning a contribution from Alec M on any blog, I tend to switch to the next comment. To me, he is the blog equivalent of the central character in the 'Rhyme of the Ancient Mariner' and as such, always completely OT..
Yes, Geronimo. Aside from the attribution, last I heard, Kevin may still have a detection problem. And seeing as Entropic Man asked:
">The fact is that we can't account for the lack of warming at the moment
> and it is a travesty that we can't. The CERES data published in the
> August BAMS 09 supplement on 2008 shows there should be even more
> warming: but the data are surely wrong. Our observing system is inadequate."
Of course much of that recalcitrant heat may now be further away than Alpha Centauri, so it might take a lot of modelling to get it back in the atmosphere.
When they can produce irrefutable evidence that CO2 is the predominant cause of warming on a global scale, I'll take some notice. Until then, all of the evidence says that warming results in the natural release of CO2 from the environment ... Vostok ice cores amongst others ... this destroys the fundamental pillars of CAGW. That warming comes from Sol, what don't they understand ? ... if you turn the sun off, life stops! CO2 makes bugger all difference other than sustaining flora. And anthropogenic CO2 makes even less difference.
When I begin scanning a contribution from Alec M on any blog, I tend to switch to the next comment.
I read his comment, think "who is this nutcase ranting off topic?" then spot the name and move on.
You have to credit him though. In the finest traditions of climate science, no contrary evidence tio his cause will dissuade him!
Question for the lukewarmers: If you go back in time, you'd see that today's warmists held the exact same position as you do, in the late 1980s and early 1990s. You can look at Stephen Schneider's Global Warming: Are we entering the greenhouse century as an example. The only thing that changed between now and then are the publication of the hockey stick and the fact that temperature continued to rise after 1988, continuously.
No specific credible major scientific advance contributed to the rising confidence in attribution that made yesterday's lukewarmers into warmists. Nothing except the tide of short-term climatic change that seemed to flow, seductively, in their favour.
That continuous rise in temps stopped in 1998. We didn't know it until, say, 2008-2009. The wamies sustained their warmism, in denial and ignorance, for another 10 years. Why should today's lukewarmers, who are a different species, join them?
Suppose, it cools dramatically for another 10 years. The CO2 hypothesis need not be wrong at all. But today's warmies - Al Goore, Mann, Gleick, Tim Worstall, Annan, Michael Tobis, Mosher, etc etc, would all be washed away. Suppose it warms dramatically for the next 20 years. Where would today's lukewarmers be?
@entropic
I give my little dog a nice chewy treat to help him remember that the dog toilet is outside the back door, not on the carpet in the hallway. And to aid his development I 'hide' it for him..he needs to use his eyes and nose to find it.
Perhaps we need to do the same for some climatologists.
Where's the heat? Where's it gone? Where's it hiding?
Is it down the back of the sofa? Or halfway up the stairs?
Or did somebody forget to put it out this time?
Sure as heck wherever this oft-discussed, but never detected 'heat' has gone it ain't anywhere that is affecting the climate in the 'lifeosphere' - at least not for the last fifteen years.
So - if it is there at all...Where The F*** Is It?
Alexander K: There are various reasons Richard Betts has been welcomed here, in my view, not least his unfailing courtesy. Except one time I remember he wasn't too kind to AlecM. But that I think is a special case.
Feb 20, 2013 at 10:02 PM | Richard Drake
Sooooooo, if someone murders you with a sorry and a smile that's alright then ?
Betts is the MO face on the blogs and twitter. He has tried and failed totackle Ryan Maue, Joe Bastardi and here, à l'eglise, he has been given the red carpet (at times). I'm not sure what his game is because IMHO if he comes here to defend models and their accuracy then he is either a fool or a charlatan. If he comes here in total honesty then he is a fool and not a scientist.
Re Dr. Betts
Just to say that on the couple of brief occasions I have been in his company he left me with the strong impression of being an honest and thoughtful man trying to do the best he can with a difficult brief.
I'm sure that he would also agree that he is not the world's biggest extrovert and finds aggressive challenges hard to take. But in a small group - even when surrounded by ubersceptics - he is engaging company and keen to listen and learn as well as teach.
He is one of the few senior climate scientists who have taken the trouble to engage directly with the members of this blog both electronically and in person. I'm reminded that following a meeting and social bash with Steve McIntyre and a host of the rest of us in Pall Mall he opined that the last ten years would have been a lot less fraught if we had all spent more time mingling down the pub. A fine sentiment.
We do ourselves no favours by alienating somebody who is already taking small steps in our direction. There are many others who better deserve our ire.
There is a tendency to want to crush the 'enemy' into dust, and I think Dr. Betts gets the brunt of that when he doesn't offer his complete attention or respond to all-comers when he pops on here.
Lat,
If you notice there is a group, wanting to claim that it is the sceptics who are taking baby steps in their direction.
I've met Richard Betts on the occasion of the McIntyre visit and he seemed to me to be a nice man in an awkward position. I had no intention above to criticise his person or behaviour, just his logic in a specific case.
One of the central planks of the alarmist argument. If you replace "the CO2 bit of the model" with "the parameter in the model which represents what the authors of the model believe to be the effect of CO2 on it", it doesn't look quite such irrefutable evidence of CAGW.
As it appears from the "retrospective forecast" thread that the parameters are in fact derived by tweaking them until the model agrees with the past (and retweaking them when they fail to agree with the present), I doubt that the physical basis for them is very solid.
entropic
Our body varies in temperature during the day. There are human processes that vary body temperature by the time of the month. Imagine you woke up one day, unaware that you are a human, and started measuring your own body temperature. You might interpret you have a 'fever', or that you have 'accumulated energy'. Both interpretations would not be correct. If the hockeysticker didn't fool you that current warming is 'unprecedented' and 'oustide the range of normal variation', what would you have thought?
@ shub
A few years ago I had occasion to carry out an analysis of bank research units' accuracy in forecasting the price of crude oil. I took the published pronouncements of five or six banks over a few years, and compared their price forecast with the price that actually resulted for whatever their forecast period was. I also compared the forecast price with the oil price at the time they made the forecast.
So if someone predicted in 2006 that the price would be $100 in 2008, I compared the oil price of 2006 with their forecast for 2008 and with the actual now-known price for 2008.
The result was a correlation of 0.0% between forecast and outcome, but a correlation of 80% between forecast and price at the time they published the forecast.
In other words, their forecasts are - literally - useless for the future, but are bad rather than useless for finding out what the oil price is today.
I suspect something similar has been going on with climate forecasters. The banks would argue that they used extremely complex and sophisticated models to forecast the price of oil.It is empirically clear from the results that even if this true, the most significant input is today's price of oil.
In the same way I reckon climate forecasters can't help themselves, and the most significant input to their models is what's happening now. Children aren't going to know what snow is, and so on. This might be intellectually defensible but if the results are rubbish then all it says is that this is a fundamentally futile pursuit.
It is also interesting, of course, that while banks demonstrably can't forecast the oil price two or three years ahead, alarmists reckon they can forecast the price of oil so accurately a hundred years ahead that they already know with almost total confidence that we're going to be having a warming crisis in 2100.
The only prediction I'd care to make a hundred years ahead is that the anthropologists of 2100 will find a rich vein for study in current climate psyence.
lest anyone misunderstand this much unloved aquatic mammal I admire Dr Betts' for coming here from time to time and mixing it with the unworthy. A point which I have made on a number of occasions. I just think there is a massive mismatch between his faith in climate models and our current level of scientific understanding as I mention in my post above. And the chances of me meeting him in a pub are slim to say the least.
And picking up on other comments above, it is standard IPCC double-speak to say one thing in the chapters on the science and to say the complete opposite (or reach a conclusion that is unsupported by the science) in its summary for policy makers. Is it any wonder that we hold that august body in contempt. And those that seek to defend the IPCC align themselves with a body that is clearly and demonstrably corrupt.
As my ol granpappy used to say 'Carthago delenda est'
There is of course the possibility that we are plummeting into a new glacial which we are, at least in part, saving ourselves from. Unfortunately, there is no guilt that can be attributed nor grandchildren to suffer, no moral high-ground to be claimed nor powers to be transferred. What a bummer, heh?
Unlike the alchemists of old, who failed for reasons they could not have understood at the time, climate modellers (many of whom seem to be well educated and pleasant people) seem to have perfected the Orwellian concept of doublethink:
"The power of holding two contradictory beliefs in one's mind simultaneously, and accepting both of them... To tell deliberate lies while genuinely believing in them, to forget any fact that has become inconvenient, and then, when it becomes necessary again, to draw it back from oblivion for just as long as it is needed, to deny the existence of objective reality and all the while to take account of the reality which one denies "
When Richard Betts has posted on BH he has served as the Met Office's lightning conductor. Anger that deservedly should have been directed at the Met Office and its propaganda generation has been directed at him.
It's virtually impossible for the people working in a propaganda organisation not to sign up to its own propaganda. Don't underestimate the power of groupthink nor the Upton Sinclair effect: “It is difficult to get a man to understand something, when his salary depends on his not understanding it” .
I don't think it is realistic to expect anyone working on climate models in the Met Office to turn around and agree that they have been worse than valueless.
I agree with Martin A, it seems churlish to criticise the MO chap who turns up here and answers some of our questions when the bosses of the organisation will not do so, and although responsible for the farrago will not defend it against criticism in any forum that doesn't tilt the balance in their favour. Weasels.
Feb 21, 2013 at 12:47 AM | Entropic man
/////////////////////////////////////
May be the same or similar or related sources that we do not presently know of or understand but which brought about the Holocene Optimum, the Minoan, Roman and Medieval Warm Periods.
As soon as one is not so arrogant to presume that one knows everything, it is a possibie to accept that there may be naturally occuring processes that presently are not well known or understood but which could (if more were known about them) fully explain the 20th century warming.
Heck since we do not accurately know the extent and distribution of cloudiness throughout the 20th century, we can't rule out changes in cloudiness resulting in changes in received at surface solar insolation being a full and complete explanation. Infact, we can't even say withn certainty whether it is warmer today than it was in the 1880s or the 1930s. All we know is that throughout the thermoter record, there have been variations (the full extent of which is uncertain) and there seems little, if any, first order correlation with the level of atmospheric CO2.
It's too late, though. We've already driven him away with all the hostility. Justified or not, is it a better place with or without his contribution?
TPYJ,
Is it too late? that would be sad.
"TPYJ,
Is it too late? that would be sad."
----------------------------------
Richard Betts commented on the matter (and Met Office climate model changes) last month at Tallbloke's:
http://tallbloke.wordpress.com/2013/01/05/major-change-in-uk-met-office-global-warming-forecast/
I know I would find it difficult in his position too, but then, none of the science-areas I have been in are being used to justify changes to the living standards of Six Billion+ people at a cost of some Trillions of Dollars.
Well, I exchange emails with him now and again, and in the last one he seemed a bit overwhelmed by some of the reactions. People generally don't hang around where they feel disliked (only me!).
Perhaps when he has a bit more time we can tempt him back.
Years ago I said that climate modelers were trying to keep their toys running on circular tracks, on the ceiling. Perhaps Richard is dodging falling objects instead of fouling objections.
=====================
Perhaps soi sauce requires double distillation.
Is it hostile when we are told that the forecasts validate the models to ask which forecasts and to what level of accuracy? Certainly never gets an answer.
By the way, the Early Bird on this thread has enough stuff to build dozens of nests in the glades of Akademe, were the birds there of the will.
==============
Feb 20, 2013 at 7:01 PM | Roy
Very well said. Time and again they argue that successful hindcasting shows that they have the physics right and, I might add, that they have the physics complete. Total nonsense. Only successful forecasting can validate models.
Say No To Fearmongers
geronimo
michael hart
shrub
The 21st century pause in measured global temperature rise may be due to the extended solar cycle 24 minimum and the weakness of cycle 25.
Alternatively, the heat may still be accumulating, but in areas where we do not have temperature monitoring. The deep ocean below 2000M is a possibility, beyond the depth range of the ARGO buoys.
Support from this comes from the continued rise in sea level.Since the 1997/98 El Nino sea level has risen at an average of 3.2mm/yr. That's 50mm in 16 years.
http://sealevel.colorado.edu/
TBYJ
I have been thoroughly trained, as one from my Rugby-loving nation should be, that it is extremely poor form to play the man and not the ball and so have been careful to recognise Richard Betts' obvious mannerliness and his (sometimes) willingness to engage here. My training in 'focussing on the ball' should never prevent me expressing scepticism of the Met Office models. If Dr Betts feels he is being attacked rather than the models, that is truly a pity but is not the fault of sceptical thinkers who focussed on the argument and not on the defenders of obviously inadequate and unrealistic models and modelling.