Thursday
Mar062014
by 
Bishop Hill
Bishop Hill Orlowski on the GWPF report
Mar 6, 2014 Climate: sensitivity Andrew Orlowski in the Register has an interview with Nic Lewis here. I liked this bit:
You've spent a lifetime modelling - surely [models are] not completely useless"
Lewis: Models are extremely useful but better at somethings than others. They're pretty good at atmospheric circulation. But when it comes to ECS there's really no reason to think they’re going to be accurate.
Reader Comments (33)
Great stuff. It's almost as if dissenters from climate orthodoxy have talked to each other beforehand.
Of course, that would never do. Working together? That could be really dangerous. :)
Great to see Nic agrees that the models are pretty good at atmospheric circulation! And this is indeed an important point -what do we really care about - changes in atmospheric circulation (which affect the weather brought to where you live - like shifts in the jet stream bringing more storms to the UK), or changes in some number like global mean warming which doesn't actually affect anybody directly?
The concept of ECS was originally a way of systematically comparing different climate models. High or low ECS in itself says little about what the changes in climate actually mean for people, ecosystems etc.
Aye, that's why the Met Office forecast a drier than average winter.....
Atmospheric circulation is basic fluid mechanics (only approximately solvable). It's all those pesky parameterisations (assumptions to you and me) to get the desired, but wrong answer.
Richard Betts
You are losing this reader. Are we to stop caring about golbal warming altogether? Is that your point? Beause it doesn't affect anybody directly? And are we to start start worrying about an increase in stormy weather - just so long as we worry about something. I have looked (in vain) for observational evidence of an increase in stormy weather. Where is it? This AGW thing isn't supposed to be restricted to the UK in February, you know.
Richard, climate always changes, and there is demonstrable net benefit from warming. Be glad.
=============
I have more confidence in Nic Lewis than I have in the majority of climate scientists. However, he has to work with the basic science and that seems to involve masses of assumptions and so-called experimental work based on yet more models.
I still don't trust climate science.
Before we can trust climate science, it must make successful predictions (this is a necessary but not a sufficient condition).
The title of Andrew Orlowski spells it out neatly: Global Warming IS REAL, argues sceptic mathematician - it just isn't THERMAGEDDON
And Nic Lewis makes a specifically telling point concerning what he calls, “a mystery regulator”
“In modern times it (CO2 levels) has varied with the glacial cycle between 180 and 280 parts per million, and is currently at around 400 parts per million. However it has been much higher than the current level before; a mystery regulator prevents "runaway global warming"
If the Earth's climate system was that sensitive, it would tipped over into runaway warming many times in its long history, and it hasn't.
I tend to be sceptical of the theory that human activities are the main drivers of climate change. However, I think we should be wary of dismissing claims that cases of "extreme weather" have been increasing. In Britain within my lifetime we have experienced the coldest winter for 200 years (1963), the Great Storm of 1987, the longest drought on record (1976), the coldest December since 1659 (in 2010), and the wettest winter on record (this year).
The winter of 1963 - the coldest for more than 200 years
http://www.metoffice.gov.uk/education/teens/case-studies/severe-winters
Great Storm of 1987
http://www.metoffice.gov.uk/education/teens/case-studies/great-storm
Heatwave Britain: What it was like during record-breaking 1976 scorcher
http://www.independent.co.uk/news/uk/home-news/heatwave-britain-what-it-was-like-during-recordbreaking-1976-scorcher-8713469.html
December 2010 update: Second coldest since 1659
http://www.bbc.co.uk/blogs/paulhudson/2011/01/december-2010-update-second-co.shtml
UK weather: Winter wettest ever, says Met Office
http://www.bbc.co.uk/news/uk-26278995
Now, I know that there have been many other cases of "extreme weather" in Britain in previous centuries. Tony B has unearthed numerous examples and posted articles on his work in this and other blogs. However, have there been any other 50 year periods with as many examples as the past 50 or so? Perhaps there have been. However this is a question that ought to be studied by orthodox climatologists as well as sceptics. In my opinion this is a much more important problem to tackle than crystal ball gazing, i.e. producing reports based on flawed models.
"Are we to stop caring about golbal warming altogether? Is that your point? Beause it doesn't affect anybody directly?"
YES!
I only wish you weren't asking rhetorically, and that you were humble and circumspect about your own ability to measure climate change in a meaningful way.
I have said before, climate sensitivity calculations seem to be the one bit of climate science that even sceptics seem to believe in. Until I read Schrodinger's Cat's (6:26) comment, I assumed I was perhaps the only person in the Universe who assumed it was as flaky as the rest of CS.
I suppose I should read up on it sometime. But it's hard to find the motivation to spend time reading about something that is no doubt, as Schrodinger's Cat said, based on the same old assumptions and so-called experimental work based on yet more models.
It's a quandry. If climate sensitivity calculations suggest that AGW is nothing to worry about, should we rejoice and wait for the Great Delusion to peter out? Or should we point out that climate sensitivity calculations are probably as much rubbish as anything else coming from CS?
...some number like global mean warming which doesn't actually affect anybody directly?
The concept of ECS was originally a way of systematically comparing different climate models. High or low ECS in itself says little about what the changes in climate actually mean for people, ecosystems etc.
Mar 6, 2014 at 5:24 PM Richard Betts
Richard, I have clear recollections of Met Office publications explaining how increased average temperatures of a few degrees would lead to dire consequences. I think I remember a 'My CLimate and Me' video explaining that with graphics showing probability distributions of temperature.
Here is the sort of thing the Met Office was scaremongering...
I get the feeling there is some Met Office backpedalling in progress.
Martin A: I agree, as I said on another thread ECS was set at 3C by Charney in 1979 and we have consistently been told that an increase in global temperatures of 3C would be disastrous. Now Richard is telling us that the ECS was only meant to compare climate models and isn't that important. I long ago came to the conclusion that I couldn't understand the clisci definition of feedbacks. I couldn't see, and still can't, how once positive feedbacks of any sort kicked in they didn't go into uncontrolled mode as was assumed on Venus I believe. But now I'm told they never mattered anyway as ECS was merely a way of distinguishing between models. Is that what I'm being told?
CS may be flaky, but you have to give its practitioners some credit they never tire of moving goalposts.
Warming has been taking place since the LIA so it is not surprising that we are currently at the high end of the modern temperature dataset. It is not surprising that sea levels have also been rising throughout this period. In recent years, the rate of rise has slowed down.
Violent storms, hurricanes and typhoons have all decreased in frequency in recent years.
We had rapid warming during the period 1975 to 1998. All warming stopped 17 years ago.
I don't really see much to fuel the alarmism that we have been witnessing. Unless, of course, you believe the models rather than reality.
"They're pretty good at atmospheric circulation"
AIRFIX
"High or low ECS in itself says little about what the changes in climate actually mean for people"
Let's see... a decade+ of Climate Science saying the opposite and this what we get now?
I guess I should say Thank You.
Andrew
Is this a claim that the models accurately reproduce changes in atmospheric circulation in response to global warming?
I was under the impression that the regional output of climate models was even more risible than their carefully fudge-factored global averages.
(Bish: the link in the post is to p2 of the article btw)
Good post by lapogus this morning (on an earlier thread) reminding us of some quotes
http://www.bishop-hill.net/blog/2014/3/6/ipcc-hides-the-good-news.html#comment20803332
If I may summarize:
No crisis, no paycheck. It is left to the student to deduce the possible outcomes and their probabilities. Hint: give the Politics of Warming function a very high forcing value.
The article states that forcing is 2.4 w/sq.m of which 0.5 goes into the ocean ( I assume the 0.5 degree C into the ocean is a typo.)
I fail to understand how the forcing from down welling IR radiation, which does not penetrate more than a few mms of the water surface, warms the deep ocean. I thought the IR warmed the surface producing more evaporation and the increased water vapour exacerbated the effect of CO2 in causing atmos. warming.
Trenberth's explanation of the 'pause' seems to be very shaky and clearly after the event.Has anyone explained the mechanism by which the oceans are heated by rising CO2 ?
osseo , martin a
climate sensitivity indicates the correlation between rise in co2 levels and rise in temperature, in order to use climate sensitivity to predict future temperatures rising co2 levels have to cause rising temperatures.
If rising co2 levels do not cause rising temperatures you might was well study correlation between temperature and FTSE, you never know it might work better.
Murray Salby and others have shown net emissions of co2 are highly correlated with temperatures, which makes it hard if not impossible to see how co2 levels are affecting temperature to any noticeable extent
Anyone can look at woodfortrees and see temperatures have been flat for last 17 or so years whilst co2 continues to rise.
http://www.woodfortrees.org/plot/hadcrut4gl/from:1960/plot/esrl-co2/scale:0.01/offset:-3.2
I don't understand why anyone discusses climate sensitivity at all.
The Earth makes a perfect replica of itself. No model can match it for representational accuracy. Nothing added and nothing taken away, or simplified, or approximated, or abstracted.
In contrast to some typed-in change in the input data, and the running some computer model, the Earth has had real co2 put into its actual atmosphere. The result of this unintended experiment has been the refutation of a theory. The rest is not, but should have been, an embarrassed silence.
No causation without correlation.
Can someone explain Nic Lewis' answer to Orlowski's first question? I get the idea that some of the heating "trapped" by CO2 goes into the ocean, but that the majority is re-radiated to space.
I don't understand the calculation. First he says that 0.4 - 0.6 W/m^2 goes into the ocean, then he says that the heating produced by 0.75 - 0.8 W/m^2 is enough to nullify 3/4 of the warming we've had.I don't understand how he goes from 0.4-0.6 W/m^2 to 0.75-0.8W/m^2.
Is it covered in the long version of the Oversensitive report?
Regarding models, Nic Lewis states:
His assessment is very much along the lines of the views of Freeman Dyson:
I note Richard Betts is VERY fast at jumping in on these recent threads and maintaining the Met Office party line (and then introduces the "these are not the droids you are looking for" trick to deflect the argument elsewhere). The Met Office have previously claimed that their weather models and the climate change models are the same. I am sure they are. The implication is that because they are so great at weather prediction therefore the climate models are validated. This conclusion does not follow and I think this is a longstanding misdirect by the Met Office and Richard Betts is hiding behind this fallacy.
The models clearly capture the general circulation pretty well - Nic Lewis agrees with this statement, Freeman Dyson agrees with this statement and so do I. However, this ability of the models to simulate atmospheric (and perhaps ocean) circulation very well is because we understand the large scale circulation and the earth's rotation, land and seafloor topographies, general heat distribution, insolation etc very well. This is the big, everyday picture. But this success is then used to imply, without any justification, that ergo the models are also good at predicting global warming/climate change. This is clearly untrue. A model that's good at simulating the circulation system will tell you what might happen if global warming of X occurs, but in no way can a circulation model confirm that global warming of X will actually occur - the GCM's would be quasi-steady state unless an external forcing (eg CO2; aerosols) is introduced. The putative physics of CO2 global warming is simply not required to make a valid global circulation model of the atmosphere and oceans capable of showing the current main circulation behaviour. And this is the logical fallacy of the argument.
Consider the Met Office statement from Julia Slingo (in one of the post-ClimateGate hearings, I think) that the Met Office uses the same models for weather and climate prediction. Well, so what? The effect of CO2 global warming over the short event horizon of a weather prediction is completely irrelevent. The effect is way too small to influence weather models, even seasonal predictions. (Its argued, eg by Lindzen, that the effect is so small that we cannot even detect it with confidence in the 60 years of global temperature data from the 1950's onwards). Conversely, to model CO2 global warming predictions you don't need a GCM, you only need a GCM to predict local changes given that CO2 global warming has occured - Richard Betts makes this point in his post, but avoids the more important point that this has no bearing on the validity of the CO2 global warming part, only on the modelling of the consequence assuming the global warming has taken place.
But the constant repeating of the fallacy I describe is used as a smokescreen to imply that because the GCM part is good, therefore the global warming predictions are good.
I think this is the crux of the case and it is a serious misdirect by the Met Office in its public statements. I think it has mislead MP's and parliament by not coming clean about this issue and clearly separating the circulation part from the CO2 warming part and pointing out the huge difference in the confidence we have on each. GCM's can be validated with current observational data, CO2 global warming predictions are not validated by current (or historical) observational (consider the hiatus and the scrabbling for explanations, for example, and the point that a 20 year hiatus never occurs in the model output). By blending the two together as one, it is possible to give the impression that because GCM's are validated by observations, ergo so are CO2 global warming predictions.
It is a non-sequitor I am afraid.
In the meantime - did the Met Office forecast the current - er - PLEASANT weather..?
We need to be told - preumably settled high pressure conditions are as sure a sign of climate change as 'more frequent storms'...
rogue - The units in that paragraph are horribly mangled. Let me take a crack at it.
---- a mystery regulator prevents "runaway global warming"
Thus Spake Zarathustra;: "Eeee lass, t'is a bit warm on't planet. Oupen oop window will thee. Aye that'll be aboot reet".
HaroldW
Thank you - that makes more sense.
OK - trying to put this into practical terms.
Transient Climate Response (<1.5C) takes 70 years, Equilibrium Climate Response (1.75C) takes centuries (200 - 300 years).
Increase from per-industrial CO2 to present gives us 0.8C increase already, with another 0.25C to equilibrium in the pipeline.
An arbitrary "SAFE" limit is an additional 2C of warming. Richard Tol has proposed that CO2 is a net benefit up to this limit. Which suggest that a further doubling of CO2 (400ppm to 800ppm) would get us to that additional 2C (0.25C pipeline + 1.75C from an additional doubling).
So, at 2ppm increase in CO2 / year it would take 200 years to go from 400ppm to 800ppm, and then 200 years after that we would feel the full 2C increase.
Is that what is being said here?
Rogue: "So, at 2ppm increase in CO2 / year it would take 200 years to go from 400ppm to 800ppm"
Since the CO2 increase is more like quadratic or exponential, if unchanged it will actually take less than 100 years to get to 800 ppm, Here is what the best-fit exponential to the Mauna Loa data shows (through Sept 2013):
ppm year
400 2015
500 2047
600 2069
700 2085
800 2097
The model is y=y0 +exp((t-t0)/tau)
where y0 is the pre-industrial CO2 concentration (257 ppm), t is the number of years starting with March 1958 with the first Mauna Loa observations), t0 is the number of years before 1958 when the CO2 began to rise, (t0=-248 years, which turns out to be about 1710, near the development of the steam engine), and tau is the e-folding time of 61 years. (Doubling time of the anthropogenic CO2 is thus about 42 years.) The model fits about 650 consecutive months of the Mauna Loa record (seasonally detrended) to within 1% each and every month. A quadratic model fits the data even slightly better but the parameters are not as nicely interpretable.
The assumption here is of continued economic growth using fossil fuel energy. Considering Chinese plans for coal-powered energy, it seems a safe assumption for at least the next couple of decades. I tried to see if the data suggest any slowing down of CO2 increase recently due to worries about global warming, movement to renewable energy, use of fracking to reduce dependence on coal, etc.. If so, the doubling time should be increasing. In fact, it was increasing, but by a negligible amount, from about 41 years in 2005 to 42 years now.
Lance W - thank you, a definite improvement on my linear model.
I generally take the anthrropogenic source of increasing CO2 as a given.
Murry Salby provided an interesting question mark as to whether temperature (Henry's Law) and soil moisture provided a better explanation.
Today I came across a set of papers by Prof Gosta Pettersson of Lund University that uses a similar model that challenges the Bern model carbon cycle. Starting with the historical data series of C14 produced from atmospheric bomb tests in the 1960s he derives a relaxation time for CO2 of 14 years (Papers 1 & 2), which he matches to the Keeling curve in Paper 4.
The implication is that anthropogenic CO2 does not have a centuries long persistence in the atmosphere, and that if/when we move beyond fossil fuels CO2 levels would return quickly to pre-industrial levels.
Is it time to formally question the Bern carbon cycle model?
http://www.false-alarm.net/
http://daedalearth.files.wordpress.com/2013/09/paper1.pdf
http://daedalearth.files.wordpress.com/2013/09/paper2.pdf
http://www.false-alarm.net/wp-content/uploads/2014/02/paper4.pdf
Of course the climate models are not any good at atmospheric circulation either unless the region is highly restricted and the boundaries and interior points are heavily seeded with real results. Such models are of limited value - but make pretty and also pretty misleading pictures in the IPCC report.
Sad of course to see the continuing logic fail about some models being good therefore the rest are probably acceptable. They are not! Alas these scope-restricted but comparitively "good" models are not the ones that the IPCC relies upon and hence are not used to drive the current, ruinous energy policy.