Bishop Hill Scrutinising the models
May 19, 2011 Climate: Models There has been something of a flurry of posts around the sceptic blogosphere about climate models and I wonder if this may continue to be a theme in coming weeks after James Hansen's recent admission that climate models are getting ocean heat uptake and the mixing of heat in the ocean wildly wrong.
This story is covered in layman's terms here by Anthony Cox and David Stockwell:
The Earth’s energy balance is the most important measure of anthropogenic global warming [AGW] because it shows whether energy is leaving or accumulating.
Among 52 dense pages of science, Hansen reports on two experiments from the last eight years that call for major revisions to the GCMs.
This is definitely a "read the whole thing" kind of article.
Put alongside the poor performance of the models against observations in recent years do we really have a watertight scientific case that demands a policy response?
Reader Comments (34)
So even James Hansen admits that ocean heat update is slowing and sea level rise is not accelerating. Wow.
But the penny has still not dropped, so another fiddle is added to get the right answer.
'But Hansen does not question the magnitude of warming from CO2 despite recent measurements showing CO2 heating may be saturated. Indeed, the need to increase the cooling of aerosols comes about because of the large positive forcing from CO2 is held constant.'
James Hansen is currently in NZ to provide a lecture tour promoting the evils of CO2, and specifically the evils of mining lignite.
The tour has a bit of a rock star quality. I doubt if there is room for discussion of GCMs
There is nothing too surprising in this article, except, perhaps, that it was written and published. Hansen is trying to accommodate his theories (understandings) to available observations which is a difficult task for even the most simple theories: nature has a way of introducing factors.
But the science is settled. It is blasphemous to attempt to revise it.
We don't deny the evidence; we simply argue that we haven't yet seen any evidence other than model output which, in the absence of supporting observations, is not evidence.
One cheer for Hansen for at last admitting the models are inefficient at their job, but -- as andyscrase points out -- he is still campaigning hard against CO2. Isn't there a cure for schizophrenia?
Bishop, sorry but this is off topic and I have nowhere else to post it:
I was awake for a while last night about 0300hrs listening to BBC1. Every Thursday morning at this time, the BBC broadcaster has an appointment with an Australian, a Dr Karl who answers questions on medical and scientific matters. He seems to know a lot about a lot of things and I often wonder how fast his fingers might be on Google He is an avowed alarmist and in fact, last week at the same time he passed some comments about Climate Change on the alarmist side. (Yes I do get some sleep during the night, seeing as you asked!)
Last night, this Dr Karl and Naomi (I think this is the spelling) Oreskes and the BBC person doing the interviewing were concentrating solely on Climate Change. The BBC man on at least two occasions referred to "Deniers" very early on in the broadcastand Oreskes passed a comment that the world oceans were rising fast again.
At this stage I switched channels to RTE 1 radio as I was so incensed and drifted off.
This is so typical of a Biased BBC as we all know. There was no balance to the interview as there was no one in opposition and the use, twice of the word denier is reprehensible. Perhaps in the UK you can downlaod a podcast of this programme to check out the complete recording as I can't do this here in Dublin.
I thought that the use of the word "denier" was pretty well banned on the BBC.
Rgds
Peter Walsh
Bishop,
Follow up to my comment above, I have managed to download the podcast. The programme is "Up All Night" and the podcast is 50 minutes.
PW
Peter
The best place for OT observations/links etc is Unthreaded (see Navigation, top RH corner of the page). And it's a very interesting read, too. Recommended.
BBD thanks for that.
PW
Not at all. I should also have said that BH sometimes follows up comments there with a blog post... it's not a dustbin!
Is this what is known as "Trenberth's Travesty"?
Doubtless the models will be tweaked again and the science will "settle" once more.
Jack Savage,
I don't have a problem with the models being "tweaked". The biggest problem, it seems to me, is that the models haven't been tweaked enough in the past to the extent that they are virtually useless as research tools because they do not model enough (how many's enough?) of the variables in the climate system.
Whether they ever can is another question, given that climate/weather is by nature chaotic.
If they can be programmed to account more accurately for ocean heat uptake or for the cosmic rays/clouds theory or for any other possible variable that has a reasonably decent chance of being right then go ahead and tweak.
Given modern computing power I don't understand why all these variables haven't been tried in combination to see which best replicates historical climate trends. I would have thought there was still enough scope for research funding even if the current belief system was proved false. I can understand why polticians and eco-activists like the present paradigm but why the scientists?
This news article Anthony got published in the Newcastle Herald may interest you all.
http://landshape.org/enm/files/2011/05/NH-ScienceSaysNoTo-ClimateComput.pdf
I wonder if these aerosols are the same ones that are claimed (but have never been demonstrated) to have caused the 1940 - 1975 global cool period?
Such a lot depends on something that has neither been observed nor quantified to a satisfactory degree (RIP Glory).
My understanding is that even the anthropogenic origin of the various species of stratospheric aerosol is speculative. Do we have proof that industrial sulphate aerosols are offsetting CO2 forcing? Or is this Hansen's inference from the observations of the reduced rate of increase in OHC?
These quotes are from a newly released historical perspective on the Hot Spot. It shows that in order perserve a beautiful theory from ugly facts it is necessary to create as much doubt as you can about reality.
Can anyone recommend an article/blog post that summarizes the whole issue of what the Bish refers to as
?RichieRich
Try this for starters. http://rossmckitrick.weebly.com/model-testing.html
@Mac at 11:39'
'The uncertainty of both models and observations is currently wide enough, and the agreement in trends close enough, to support a finding of no fundamental discrepancy between the observations and model estimates throughout the tropospheric column. '
I cannot believe that anyone could write that sentence with a straight face.
It has got to be an award winner.
How about
'The observations and the models are both rubbish, and could be just about anything, but our wild guesses for both agree beautifully?'
That's precisely what Santer et al. have been trying for a long time, stretching the error bars of both observations and models beyond their elasticity limit, until they barely touch each other and then claiming victory.
McKitrick and company have shown that that is not the case.
At bottom, Hansen has only his faith in the "runaway greenhouse" effect; that is what his entire career is based upon. This explains why he is perfectly willing to admit that the models have failed at predicting the temperature trend over the last 10 years or more, thus figuratively chewing off a leg to free himself from having to admit there is NO greenhouse effect as he claims. He doesn't even realize he has chewed off a leg, because he thinks he can put over the idea that the missing piece in the models is the effect of aerosols -- and note, specifically anthropogenic aerosols at that. So it is also fundamental to his mind, that it is all mankind's fault. He is now driven to shoot his mouth off wildly at every turn, and at the innocent public, like one of those deranged men who find themselves failed and stalled by the world, and end up on violent killing sprees. But he has only his failed hypothesis to blame, and he should have known it as failed nearly 20 years ago, as I have pointed out on my site. He cannot face his own incompetence, especially with all our scientific institutions and the media behind him. So the scientific test, which should have been clear long ago, has now become a moral test for everyone, and 97% (of climate scientists anyway) are happily failing it.
Richie Rich
Lucia's Blackboard
C3 Headlines has multiple climate model articles regarding their failures.
http://www.c3headlines.com/climate-models/
Tony and David end with "As the admission by James Hansen aptly demonstrates, experiment trumps theory, and the science supporting AGW is still subject to very large revisions and is not settled."
The ABC gets an incoming blog "Secondly, its ('settled') an expression scientists simply don't use, nor do scientists think in the manner that the science of anything is 'settled', whether it be the the science of gravity, light, evolution or AGW. I don't know what Dr Stockwell's science credentials are, but if he really is a scientist I'm surprised he would put his name to a piece that uses this expression as it is used at the end of the piece."
Then to ABC closes off discussion. Like in music, nice to end on a high note, even if it's a bummer.
I can assure you that David and Tony do not live by the credo of "settled" science. Their criticism of it has been twisted here and that is not cricket. Nor is it cricket for the ABC to close off argument with an insidious blogger- in-denial claiming that scientists do not use the concept of "settled", when the word is given star rating by the IPCC, Sir Stern, Sir Beddington, Lord Ox, the BBC and the like.
Richie Rich
Following up on BH's comment, this particular Lucia post looks at the three ground based temperature data sets and all analyses show less warming than the IPCC's predicted 0.2 C per decade.
http://rankexploits.com/musings/2011/which-observation-is-the-outlier/
I came across an aspect of model performance yesterday which surprised me. I followed up a comment at Climate Audit which linked to a post from a couple of years ago at Lucia's blog, "The Blackboard". That post is here... http://rankexploits.com/musings/2009/fact-6a-model-simulations-dont-match-average-surface-temperature-of-the-earth/
The post shows that, while all the models' calculations for surface temperature anomalies agree well (to fractions of a degree, surprisingly well given their error bars), their estimates of the actual temperature vary much more widely, ranging over about 4 degrees.
It seems as though the models have been tuned to closely match recent anomalies but in so doing they are generally unable to get near the actual surface temperature.
This begs the question of what would happen to the calculated temperature anomalies if, instead, the models were tuned to match the actual surface temperature. One can only speculate...
BTW, if this is well known or has been superseded by subsequent developments, I apologise for mentioning it. It was new to me: I'd naively assumed that the models' anomalies were worked out by first calculating the actual temperature and then subtracting an average. Still, at least I am seldom disappointed in the abilities of climate scientists to surprise me with their unconventional methods
@ Simon Anthony
Yet modellers insist that it is the trend and not the actual value what counts. I personally think that a model that is unable to replicate the initial conditions is useless.
Here you have some more charts:
The one you mention from Lucia about temperatures:
http://rankexploits.com/musings/wp-content/uploads/2009/03/temperatures_absolute.jpg
A similar one about precipitation, which I did with similar data:
http://a.imageshack.us/img14/6786/gcmmodobspcp.png
Then the regional bias is even worse, up to 4K and 400% in specific humidity
this is from John and Soden:
http://imageshack.us/m/145/1171/modelbiasestq.png
Who claim that there is a "robust" agreement in the bias
and "Thus, the robustness of temperature and water
vapor responses in the models suggests that their climate
sensitivity is not affected by mean state biases in temperature
and humidity fields."
on which I beg to disagree. It is the equivalent of saying "it does not matter what is the actual concentration of CO2 in the model, the robustness of the response indicates that the trend in response to CO2 increase is reliable...." I beg your pardon?
@Patagon
The discrepancy in the precipitation values is even more striking than the temperatures. So it seems that models can be adjusted to match trends in temp and precipitation but then they don't match the actual values of either temp and precipitation, and the discrepancies are far greater than the supposed trends.
Now if I were to make a model of "something", with some adjustable parameters, I'd adjust the parameters so that the values of "something" matched measured values. Then I might see whether the trends in the values of "something" over time also matched. I'd do this because the trend is a small perturbation of the dominant effect and, if my model isn't able to account for the dominant effect, I'd have no confidence that it could account for small perturbations.
It seems that climate modellers have instead done things the other way around, believing, or at any rate claiming, that they can accurately describe small perturbations yet aren't concerned to get the dominant effect right. To put it gently, this is counterintuitive.
The obvious question is just what happens to the trends when the models are adjusted to match actual temp and precipitation. It may be that when such adjustments are made, the trends from the different models still roughly agree; on the other hand they may turn out to disagree not only in the magnitude but also the sign of the trend.
Have you come across any work on this?
Geoff, 'the science is settled' or 'not settled' is a slogan overloaded with meaning. Of course the obvious rejoinder is the science is never settled, but the slogan suggests there are major issues to be addressed, as shown by this issue, that could lead to the net warming in models being revised by half. Right now I cant think of another way of communicating that 'the uncertainties are greater than being portrayed, and sufficiently great as to need resolution before making significant policy changes" more simply.
Aerosols are very strange,
the things they do to climate change.
They warm things when we need the heat
Or cool things, which is pretty neat.
And if you want things just the same
Aerosols are what's to blame.
Oh let us never, ever, doubt
What nobody is sure about.
JF
Thank you Mr Belloc....
An official reply! Game on. http://www.abc.net.au/unleashed/2722486.html
Bish, the answer to the final question you finished this post with is NO BLOODY WAY!! Forgive my use of an epithet, but I revert to speaking in Antipodean at times of great excitement. :-)
DS
http://www.abc.net.au/unleashed/2722486.html
Quote, Geoff Davis, "The measured variations of temperature and atmospheric carbon dioxide during the ice ages accuratelyconstrain the so-called climate sensitivity to 3 degrees per doubling of carbon dioxide, plus or minus about 0.3 degrees."
Now that needs to be challenged because a paper by Chylek, P., and U. Lohmann, 2008, using Vostok data from the last ice age determined climate sensitivity to produce warming of between 1.3 and 2.3 degrees due to doubling of atmospheric concentration of CO2.
Mac
You mean: "Aerosol radiative forcing and climate sensitivity deduced from the Last Glacial Maximum to Holocene transition" I assume. This is of course a study heavily reliant on estimated aerosol loadings ;-)
It's quite possibly also incorrect in its conclusions although I leave that for people to decide for themselves.
See Hargreaves and Annan's comment on the paper in Climate of the Past here (small pdf).