Bishop Hill Paul Nurse on geoengineering
Sep 9, 2011 Climate: Models Royal Society Paul Nurse has a letter in the Guardian (where else?) on the subject of geoengineering. It's less bad than you might expect.
A time may come when mankind will need to consider geoengineering the climate to counteract climatic effects of greenhouse gases. If that time comes, we need to have a good understanding of whether such efforts will work and, just as importantly, whether they will have any negative side effects. Those who oppose such exploratory research on the grounds that we do not know what its effects may be ... are missing a fundamental point of research, which is to allow us to potentially rule out any technology that would have negative effects that outweigh the positive.
Researching stuff probably does little harm, although one can certainly question whether geoengineering research, or indeed any scientific research, should be a priority for government spending at the moment. Outside the ivory tower, times are hard, but it is not obvious that Sir Paul has noticed.
One little snippet of his letter caught my attention:
Geoengineering research can be considered analogous to pharmaceutical research. One would not take a medicine that had not been rigorously tested to make sure that it worked and was safe.
As Delingpole knows to his cost, Sir Paul likes his analogies, but I think that he may have erred this time. Because if we demand that putative medical treatments are thoroughly tested to see if they actually do what they claim, should we not demand the same from the climate models that Sir Paul and so many others believe justify us redesigning our economies on neanderthal lines.
Reader Comments (80)
Is genocide a geoengineering solution?
Just useing Pual Nurse logic here
"...should we not demand the same from the climate models that Sir Paul and so many others believe justify us redesigning our economies on neanderthal lines"
I certainly recall Sir Paul coming out with some faintly absurd remarks on this issue, but do you have a quote that supports this as being his view?
On past performance we can rest assured geoengineering will be placed under such tight controls and red tape, it won't ever be implemented.
As for the analogy climate control is usually framed as public health not personal ...so it'll go ahead like the war on obesity or drugs, not prescription medicine
Holy Mackerel, I have just covered this very important topic over on "Discussion" at Zed's thread on wind power.
It is not the research that is a potnetial issue, it is the directed research.
The very best science and research invariably come from demand pull. A genuine problem or opportunity that unleashes a desire to profit from solution.
The worst science and research invariably comes from supply push. A central planner with purse strings deciding what is or is not a good idea.
Nurse is tilting towards the latter, suggesting there is no issue to be has with prescribing the use and intention of public money for research. Will there is and it is an opportunity cost. Let those people with the vision and industry to pursue new ideas and research make there case without some bureaucracy deciding which flavours are best. Let the money follow the most succeszsful and productive ideas.
Who knows, the very best use of public research funds might be the pursuit of a yet unknown, but revolutionary advance in the production or use of fossil fuels. Nurse would have us decree that such a path of discovery would, should never be supported from the ideologically driven outset.
To quote Will Smith, Nurse comes across as the dumbest smart person I have ever seen.
Geoengineering to counter specifically AGW is akin to the medieval cathederal building. A lot of money wasted just to glorify, appease and get closer to a god of your choosing.
Why in a letter? Why not deliver a keynote speech?
That's the "whys" of the way. Now the "hows". I am assuming any geoengineering solutions will be on a scale reasonably close to building the Great Wall of China, so I can't see, offhand, how you'd find any unexpected consequences in advance of the implementation. He seems stuck on medical analogies, althoughbhe lowered my expectations considerably with the Horizon programme, he seems to have slumped further in my expectations. Too cloistered a life I believe, completely lacking in practical skills.
BH
Of course you should demand the models be tested. We do test them!
Papers evaluating the Met Office Hadley Centre model used (amongst many others) for IPCC AR5 are here and here.
@Richard Betts... As an engineer, and as a testing expert...how can you model a 4 billion year old chaotic system. And how can you test it? And how can you modify it with your "geo-engineering" models? And then test new model?
What is the baseline for your test? Other models? Deep Thought?
I have never seen this adequately answered. How can you test against something you do not understand? And then effectively claim man is 90% the the cause?
Can you explain? Seriously.
If I were to write the test cases for these models I would have real problems....
Geoengineering? Why are people so keen to damage the planet deliberately?
Oh. I know - they hope to make a sh*tload of money doing it.
Wind farms have been introduced without any evidence that they reduce CO2 emissions. Current thinking is that they can increase CO2 emissions, which is good, but that is an awfully expensive way to do it compared to burning coal and gas.
Richard:
I spent years running a team validating thermal-hydraulics computer models and have presented papers on the topic. The validation of climate models is a joke in comparison with what we had to do to justify using the computer models. In the real world of engineering, computer models of the standard of climate models would not be used in connection with any practical application.
A better analogy might be , asking the pharmaceutical industry to produce a drug to combat a desease that we do not have at present.
Interfearing with a system (climate) before we understand it is some kind of insanity. IMHO
Richard Betts @ 9:48 AM:
A recent post from Pielke found here regarding a paper by Mote appears doubly relevant, because it questions the ability of models to provide useful information to policymakers and impact analysts.
Can you comment on this one too please?
@Richard Betts
Having a scan through the first link "The HadGEM2 family..." I dont see examples of we "We waited 10 years and looked back to see how our models did" (I admit I may have missed them though), however I do see examples like:
This shows a graph of lines roughly like sine waves represented ice extent and the model examples are apparently considered good work over a chosen year long period because they are within 20%. I could sketch a similar line freehand for the next year that would be within those limits of testing I think!
Are models run ever started and then later tested against new data that eventually appears over many years? This is one of the questions Michael Crichton asked and I've never seen it demonstrated as happening.
"validation ", now there is another rum old word used in the climate change debate.
What we do know is that GCMs on their own are not confirmation of AGW.
Where is the AGW signature in the atmospheric, oceanic and weather data that is required to validate the models?
Answer there is none!
Pesadia: "A better analogy might be , asking the pharmaceutical industry to produce a drug to combat a desease that we do not have at present."
Reminds me of one by the great Victor Borge:
"My father invented a cure for which there was no disease and unfortunately my mother caught it and died of it."
Sir Paul had better be careful!
I do like this paragraph on GCMs in Wikipedia.
It sums up perfectly the state of climate science. If models don't match the data, the data must be wrong.
One of the reasons why it is very difficult to take climate modelling seriously is to imagine the following scenarios.
Scenario A: The scientists at CERN discover, by physical experiment, a brand-new, but totally unexpected particle or force. This renders string theory useless. It is totally debunked.
What would happen next?
Scenario B: Some climate scientist discovers, by physical experiment, a brand-new, but totally unexpected agent. This renders current climate theory useless. It is totally debunked.
What would happen next?
Despite careers being ruined, reputations being trashed, and funding being diverted I suspect that scenario A would result in whoops of excitement in the scientific community – even amongst those ruined.
In scenario B, however, well, we've already seen how scenario B players response to criticism and new information. The science is settled. The models are correct. There can be no paradigms shift.
Scenario A is science.
Scenario B is something else…
So medicines have to be validated: Let us say we have one patient with a single instance of some symptoms. The symptoms are that their temperature rose by 0.7C. The doctors (who won't share the data) tell us that this has never happened before and that it proves beyond all doubt that the patient is going to die unless he is given an incredibly expensive treatment, which has never been tried before, for which there is no evidence it will work. etc.
Then imagine ... and this is the laughable bit ... that all the other medics are so astounded by their work that they give/recommend these doctors for a Nobel prize.
1. The patient hasn't even been proven ill.
2. There is no causal proof between a perfectly natural phenomena of body temperature and any disease.
3. There is no proof that even if the temperature went up a little more that (as we don't know what "normal is") that it wouldn't do the patient more good than harm.
4. There is proof they constantly hide data, that they have corrupted the peer review system so that only those few "dealing" with the patient can publish.
5. And there is very clear evidence that there is a strong financial motive.
Wouldn't sane, sensible people start to question those doctors?
Well put Mike Haseler.
pesadia
Victor Borge (of blessed memory) claimed his uncle "invented a cure for which there was no disease in those days. Unfortunately he later caught the cure and died ...."
I think "be careful what you wish for, you might get it" holds good!
Geckko's idea of a revolutionary advance in the use of fossil fuels would seem to make more sense than trying to tinker with a climate system that we don't even begin properly to understand. We are sitting on centuries-worth of the stuff and refusing to use it is criminal in the extreme. The only possible reason for not trying to find a "clean" way to use it (always assuming that there is any good reason why we shouldn't continue to use it in a "dirty" way) would be because the Greenies don't want us to use it at all, ever.
Errr ...... have I stumbled on something here?
Roger Longstaff
Victor Borge=Genious of comedy
http://www.youtube.com/watch?v=6bpIbdZhrzA
There is circular reasoning in the current mainstream view on the role of CO2 as a driver of increasing temperatures. That is; that in the absence of consideration of other possibilities, increasing CO2 in a warming world will have a net positive effect in conjunction with its other effects, primarily on water vapour. A net negative effect is discounted because the observed temperature trend is upward. From that position, the sensitivity of the climate to increases in CO2 is calculated and increasing global average temperatures predicted. Continued increases in GAT are taken as confirmation that the hypothesis is correct as are calculations on radiative forcing. In this scenario, the relationship between CO2 and water vapour is assumed. In the absence of a means to test that assumption (I include models), the evidence is circumstantial and the reasoning circular.
Should however, there be other less understood or even unknown factors increasing GAT, then that sensitivity calculation is incorrect and if so could range from its present high (circumstantial) value to one approaching or even passing zero. All of the uncertainty lies, not in the physics of absorption/radiative properties of CO2 for whatever they might be is not the direct driver in changes to GAT, but in water vapour/cloud interaction and by association, albedo.
The evidence for AGW stands up to scrutiny on circumstantial evidence only. Other observed effects of a warming world – ice, sea-level etc. - are not supporting evidence unless or until that which is circumstantial becomes direct. Current widespread support for the AGW hypothesis is argumentum ad verecundiam and calls for actions to mitigate are based only on the precautionary principle.
My apologies for the long post and in particular to those who have said the same thing before.
TS
Thanks - very important question. The answer is yes!
To test projections of radiatively-forced global warming by itself (ie: without being misled by internal variability) you unfortunately need about a couple of decades of forecast and data (Ben Santer's
recent paper is relevant to this).
Early climate models, which used the same basic principles of simulating the Earth's radiation budget and its impacts on temperature, were used in the early 1970s and forecast warming of about 0.6 degrees C by 2000 on the basis of the expected CO2 rise (sorry, paywall!) - the actual warming was about 0.5 degrees C.
More recently we have started to initialise the models with data on the specific current state of the climate (mainly sea surface and deep ocean temperatures) in order to get them started at the right state within natural cycles of variability - this then allows them to be tested against data over a few years. See this work by Smith et al at the Met Office Hadley Centre which published a forecast for the 10 years following 2006, which can be compared to observations.
Any serious consideration of geoengineering whatsoever is proof that the world has gone completely mad.
If a geo engineering approach is implemented, would its advocates be forced to accept responsibility for the first naturally occurring weather catastrophe that happened afterwards?
Geoengineering research can be considered analogous to pharmaceutical research. One would not take a medicine that had not been rigorously tested to make sure that it worked and was safe.
Anyone got a spare Earth like planet for testing ?
Geo-engineering is a mad scientists' paradise.
I think that from today, if everyone in the world started to stir their tea/coffee with their left hand, we would see no further significant rise in temperatures, and pehaps the onset of a decline
You say it does little harm, but it does take funding away from other projects that might actually benefit mankind. In fact you could say this about the entire edifice (I don't know why, but I like to refer to AGW as an "edifice"!).
Jimmy Haigh notes: "Geoengineering? Why are people so keen to damage the planet deliberately?"
Bears repeating, Jimmy.
Here is an interesting leaflet on this topic published a year ago by Natural Environment Research Council:
http://www.nerc.ac.uk/about/consult/geoengineering-dialogue-leaflet.pdf
It seems that the Royal Society were involved.
A couple of points caught my eye:
How controllable is it? Participants did not support
scientists ‘interfering’ with nature’s complex systems unless
detailed assessments of the consequences were carried out
first.
How reversible is it? Scientists should be able to
‘switch off’ a geoengineering project. Participants called
for research to progress in small stages, to reduce the
likelihood of irreversible consequences.
Richard Betts BH Of course you should demand the models be tested. We do test them!
A year or two back, I read on the Met Office website how climate models are tested. It explained very clearly that the Met Office verifies that the models correctly reproduce past climate variations and this confirms their validity.
In pattern recognition engineering, this is called "testing on the training data".
If the models could not even reproduce past climate variations, they would obviously have fallen at the first hurdle. But being able to reproduce the data from which the models were tuned does not confirm the correctness of the physical models, in terms of their being able to predict future effects. (Tuning a model = adjusting the fiddle factors which are included to account for incompletely understood physical effects.)
From my own experience of the difficulties of verifying models of comparatively simple things such as telecommunication system components s with random traffic fluctuations - where detailed comparisons with the real thing can be made and where a correct and precise description of the physical system is available - I find it quite hard to take the idea of validating climate models seriously.
Modelling: placing betts in a casino.
Devilish thoughts, uh...to use Dear Kev's "attribution" tool in order to sue Sir Paul and his toy...
"How reversible is it? Scientists should be able to
‘switch off’ a geoengineering project. Participants called
for research to progress in small stages, to reduce the
likelihood of irreversible consequences."
I should have thought the nature of the systems involved precluded such reversibility. The medical analogy that springs to mind is trying to reverse the ageing process.
OMG this is the single biggest danger to the planet, the arrogance of man thinking he can tinker with the climate and all will be just fine. It is clear they do not understand the climate else their predictions might be even close, please someone stop these fools, they know not what they do.
Both UAH and RSS datasets still have 1998 as the warmist year on record.
2009, 2010 are and 2011 will be lower than 1998. So for this model to be correct 2012, 2013 and 2014 have to be warmer than 1998.
What is the betting on that?
Good research almost always pays off if only seredipitously. Of course most "climate science research" is unworthy of the name.
I thought Paul Nurse erred last time as well.
BREAKING NEWS: Dessler has been forced to rexamine his calculations in his new GRL paper.
So much for peer review.
Mac, it was probably pal review, same Team tactics, same result.
GC
It would appear that Dessler is re-writing his GRL paper as we speak.
Faster tracked paper. Even faster re-drafting.
It would appear that Remote Sensing's peer review process is far more robust than GRL's.
How embarrassing for all concerned - Wagner, Trenberth and of course Dessler.
Mac/golf charley: let's not accuse anybody shall we.
Dessler's work underwent peer review, obviously. Only thing, Dessler's peers are inferior to Spencer.
Mac - do you have a reference please?
nby - see spencer's blog, good bad and ugly post, update #2, stuff underlined.
Seems like Dessler is making several major changes to the paper, after it has been 'pal reviewed' and accepted!
Richard Betts
The Smith, D et al model is interesting, where can I find more up to date information about it?
The magazine article does not have any blog links associated, not even RC, and no comments and the link to the neswblog about the story brings up a 404 error.
I would presume that with such results 4 years ago subsequent runs, tuning with real data would give a much better prediction up to 2020. Is there anywhere that It is published?
If it is currently in use for the business end of MET predictions I will understand if it is not accessable.
Roger Longstaff at 11.06am
Evidently I managed to miss your post while composing mine!
No plagiarism intended, I assure you! "Great minds" and all that ...
If Dessler 2011 is being redrafted, will it have to be peer reviewed again, and is the journal obliged to use the original reviewers, or will they be disregarded in perpetuity by said journal?
If it was pal review......