Bishop Hill The fall of Forest 2006?
Jun 25, 2012 Climate: Models Nic Lewis is best known to the sceptic blogosphere as one of the co-authors of the O'Donnell et al paper, which found significant flaws in Steig's paper on Antarctic warming. Lewis has just published an extremely important article about Forest et al 2006, one of the key climate sensitivity papers shown in the IPCC's Fourth Assessment Report. Forest et al found a climate sensitivity of 3°C/doubling, rather higher than many other studies. However, although his research has been hampered by the fact that Forest appears to have lost the raw data (!), Lewis has concluded that there must have been a misprocessing of the figures and that the correct figure for climate sensitivity would be only 1°C/doubling.
If I am right, then correct processing of the data used in Forest 2006 would lead to the conclusion that equilibrium climate sensitivity (to a doubling of CO2 in the atmosphere) is close to 1°C, not 3°C, implying that likely future warming has been grossly overestimated by the IPCC.
It is hard to overstate how important this finding is, if correct.
Reader Comments (61)
I see that Dr Chris E Forest "graduated from the University of Wisconsin-Madison in 1990 with a BS in Applied Math, Engineering, and Physics", whereas Nic Lewis graduated from Cambridge University with a degree in Mathematics.
Climate psyentists like to "win" debates with the argument from authority ("I'm a scientist and you're not, so I'm right and you're wrong"). So let's see. Who's more likely to be right on a question of statistics: the guy who went to the 3rd-best mathematical university in the world (or "on the planet", as greenies like to say), or the guy who went to the 96th-best?
http://www.guardian.co.uk/higher-education-network/2011/sep/05/top-100-universities-world-mathematics-2011
Cambridge 3rd; Wisconsin 96th.
Tough call, tough call....
"Dr Forest reports that the raw model data is now lost"
Yeah, yeah.
"Forest appears to have lost the raw data (!)"
It would appear that sloppiness is the normal mode of operation within climate "science", Therefore, it is not surprising they get all upset when people start asking awkward questions.
It's like watching a conjurer ... every last visible (scientific) support for the body has been removed, yet still it remains suspended by some "miracle".
Of course, the real truth is that it never was supported by the visible supports of science .... and so even when all the scientific support is removed, it was never going to fall.
"It is hard to overstate how important this finding is, if correct."
I'm sure the commenters here rise to the challenge
“the raw model data is now lost” If a student made this claim their essay or dissertation would be failed . But as a ‘professional climate scientist’ such basic standards are not required.
Frankly this idea never stops amazing me that the standard seen in this area is lower than the standard expected of any student on a undergraduate course is a blood* joke.
Looks like the cavalry might have finally appeared over the horizon - some real scientists are starting to check this stuff out.
I think I'll ring up Staples Office Supplies and see if shredder sales are up in East Anglia.
If a three [K] falls in the Forest, and there's no data around to bear it, does it make a sound?
(I'll get my coat.)
Read Nic Lewis’s article at ClimateEtc and you’ll see what he’s up against. The journal won’t accept a comment on an article, only a full counter-article. How many journals will accept an article which says “I don’t believe x because the data’s been lost and what I’ve been able to find out about it is contradictory”? That’s not science, it’s investigative journalism.
So as far as the IPCC and the outside world is concerned, Nic Lewis’s research doesn’t exist, and Forest 2006 will continue unchallenged.
I find the “my university was better than yours” argument contemptible when used by warmists, and equally so when used by “our” side.
You can almost hear the shrugging of the shoulders in the Forest camp.
Global warming is dead. It's served its purpose so who cares any more if I've lost my data or even if I made an "error" or two, or several.
As Geoff says, Forest et al is safely in AR4 and can be referenced in future if necessary, but hey, guys, it's all about sustainability now. That global warming stuff is out of date. Didn't you get the email?
J4R,
yes, being lectured by such obviously low calibre people has always rankled with me. From the supposed "top-end" guys like Steig and friends through to the 3rd rate Phil Jones then onwards to clowns like Richard Black and the arty Guardian types.
Of course this isn't a peeing contest and only the facts matter but it seems bizarre for the apocalyptic side to play the authority card when they have so little to back it up.
(Maths degree from no 27)
Foxgoose: 'some real scientists are starting to check this stuff out'
It appears that all the IPCC physics is wrong.
A fall back position for the IPCC is that the existence of a single, simple constant for something called "equilibrium climate sensitivity" is neither a valid nor a genuinely useful concept in the first place.
The sole point of it is to reduce complexity to a quotable number that can go into headlines to scare the ignorant. I wouldn't be surprised to see mention of it quietly disappear if it stops fulfilling this criterion.
michael hart wrote concerning "equilibrium climate sensitivity":
"I wouldn't be surprised to see mention of it quietly disappear "
Indeed, IPCC linked scientists are now pushing TCR (Transient Climate Response) - what the global mean temperature rise would be after a 70 year period during which CO2 levels doubled. Not a good measure, in my view, since it mixes together the equilibrium response and the delay caused by deep ocean heat uptake.
Effective climate sensitivity, being proportional to the reciprocal of the climate feedback or response parameter (how much net outgoing radiation changes with a 1 deg.C rise in global mean termperature, at a given level of forcings) seems to me a much more sensible metric, if one wants to move away from equilibrium climate sensitivity. It is closer to what most climate sensitivity studies really measure than equilibrium climate sensitivity, although the difference between these two measures is often ignored.
That fall back position is compromised because of fraud. Despite the best efforts of Oxburgh, Beddington and King, the public now knows this and that the scientific establishment, including the Royal Society, have been seriously compromised because even in the absence of complicity, politics made them overlook defects in the science obvious to any objective professional.
How could any 'science' claim 'cooling by more reflection from smaller droplets in polluted clouds' exactly offset AGW when it's obvious large-droplet rain clouds reflect most sunlight. Every process engineer who looks at the IPCC Energy Budget picks up the elementary failure to understand heat transfer at the Earth's surface. Yet renowned scientists failed to spot these basic errors: incompetence.
@ geoffchambers
I see your point, but I'm not so sure. The IPCC and Stern are best exploded using the "facts" they themselves put forward, because then there can be no argument about the quality or reliability of the facts. In the same way, it can be useful to debunk the quality of the arguments as well using the same type of argument that warmists use. "My university is better than your university" is patently an empty argument but on the other hand it's the warmists who are dependent on it, and nobody else.
From Hansen & Sato, Climate sensitivity estimated from earth's climate history (draft here):
Difference in global average T between LGM and late Holocene = 4.5 +/-0.5C
Increase in forcings necessary to maintain this temperature difference:
GHG = 3 +/-0.3 W/m2
Surface albedo = 3.5 +/-0.7 W/m2
ECS = 0.75C per W/m2 forcing change = 3 +/-0.5C for 2xCO2 (4W/m2 forcing).
Bang on top of ye olde Charney sensitivity from 1979.
Number of studies finding ECS to be ~1C vs ~3C? Er, not a lot, guv.
I think if a mistake has been made here, it is most likely to be by NL, not Foster.
BBD: you easily explain all warming attributable to GHGs by correcting Sagan's mistaken aerosol optical physics. A second optical process increases average low level cloud albedo by ~10% at present, more in polar atmospheres. Switching it off by biofeedback explains the end of ice ages, the periodic Arctic melt cycle and recent OHC rise, now reversing in the N. Atlantic. Net CO2-(A)GW, is near zero.
The bottom line; 3.5% reduction of cloud albedo accounts for the climate bistability attributable to GHGs.
Correction A second optical process increases albedo for 40% of low level clouds by ~10% at present, a higher proportion of such clouds in polar atmospheres'
BBD: if other climate scientists generally considered that the LGM studies provided reliable evidence of climate sensitivity being 3, as Hansen claims, I think that the IPCC would have featured them strongly in the AR4 WG1 report. It did not do so, instead discussing at length the great uncertainties that exist, particularly in what forcings were at the LGM. Section 9.6.3.2 concluded merely
"Overall, estimates of climate sensitivity from the LGM are broadly consistent withother estimates of climate sensitivity derived, for example, from the instrumental period."
FWIW, IPCC Figure 9.20 included two LGM studies in its estimates of climate sensitivity (5-95% ranges only). One, Annan LGM 05, did not explore the possibility of sensitivity being under 4 deg.C. The other, Schneider LGM 06, shows a 5-95% range of 1.2 to 4.3 deg.C.
I think, along I imagine with most scientists, that instrumental measurements provide a much more reliable way to estimate climate sensitivity than LGM and other paleoclimate proxy-based methods.
spartacus
We went through all this nonsense in some detail when you were Alistair, and then again when you were MDGNN. I'm not doing it yet again in honour of your new name.
Spartacus, BBD
This thread is about Nic Lewis's paper, not any alternative theories
BH Sorry.
It is not important at all, because the carbon dioxide climate sensitivity is in fact neither 3°C/doubling of CO2 nor 1°C/per doubling, but zero. The Venus atmosphere has 96.5% carbon dioxide, over 2400 times the concentration of carbon dioxide in Earth's atmosphere, or over 11 doublings--but the Venus/Earth temperature ratio is physically explained, fully and precisely, at points of equal pressure over the full range of Earth tropospheric pressures, only by the ratio of the two planets' distances from the Sun. The temperature-versus-pressure (TP) curves of the two atmospheres, when corrected only for their difference in distance from the Sun, are right on top of each other; Venus's TP curve is not 11°C higher than Earth's (as it would be if the sensitivity were 1°C/doubling), nor 33°C higher (if it were 3°C/doubling)--within the thick cloud layer of Venus, it is in fact about 5°C LOWER than Earth's TP curve, but above and below that layer there is no difference in the two curves at all. There is no room for a carbon dioxide greenhouse effect at all in this DEFINITIVE observation, and it should have been obvious to any competent atmospheric physicist or climate scientist 20 years ago, when the Venus data was obtained by the Magellan spacecraft. Any and all talk of climate sensitivity to increasing carbon dioxide is simply incompetent, in the context of the clear Venus/Earth temperature comparison. This is the scandal that not even Bishop Hill, nor any of those "skeptics" seeking to put forward a "lukewarm" compromise position on the greenhouse effect, will openly address. Heaven forbid that anyone be openly outed as incompetent, much less that virtually all of the climate "skeptics" should be, along with all of the consensus followers. That would not be polite, and they would all be left sitting in their authoritative positiions, with their mouths open but with nothing to say, and no way to defend themselves against the anger of a suddenly wiser public.
"I lost the data" reminds me of the school boy excuse of "my dog ate my homework". Unfalsifiable of course like all AGW theory. Forest should clear out his desk and leave the data torturing business, as he's now been rumbled it seems. My advice?
"Run Forest, run"
Hi Nic
Thanks for the response. The problem remains: lots of convergence around ~3C as the most likely value for ECS. Foster06 would be an *outlier* at ~1C.
I can't get worked up about this, especially in the light of the empirical calculations eg H&S which are convincing. Hansen's empirical work post-dates AR4, which is doubtless why it didn't feature there :-) Let's see what's in AR5.
The problem with the instrumental record is that the period affected by CO2 forcing is *too short*. So I would disagree that it's the best and only way to constrain modelled estimates. On reflection, I would disagree with the statement that 'most scientists' think that it is inherently superior to a paleoclimate approach based on the LGM/Holocene transition.
BBD:
I've been following the CS work for years now, and most of the observationally-based studies find a CS of around 1 deg C/doubling CO2. This has been discussed in detail at both Climate Audit and and Judith Curry's. The 3 deg C "convergence" is almost exclusively based on model-based studies, and that's what the IPCC reports. It appears to be almost certainly overstated, and the positive-feedback climate models don't make geologic sense.
Peter D. Tillman
Geologist & geochemist
The problem with the instrumental record is that the period affected by CO2 forcing is *too short*?
And yet it is only over the latter half of the last century (and some would say the beginning of the current century?) that we can apparently be so confident that the impact of CO2 can be separately discerned from the impact of other natural factors ...
"Run Forest, run"
Nice one Farley.
In the film (Forest Gump) the leg-irons fell away from the eponymous hero.
Dr Forest may well be experiencing the opposite!
If the field of climatology is not totally populated by a bunch of charlatans it is a remarkably unhappy coincidence that so many of them are so skilled at appearing as if they are.
'I lost the data'
Yeah right . Course you did Chris. Just one of those things. Just like happens to Mikey and Philboy and all your other mates.....what an amazing set of unfortunate coincidences.....we believe every word....course we do.
Am I right in thinking that Nic Lewis' concerns would be admissible into AR5 regardless of whether it was published in a journal or not? (Given the greater acceptance even of quite grey literature this time around...!)
The Irreproducible Panel on Climate Change?
Jokes aside, even if Nic is wrong it would still be concerning that:
- it wasn't possible to find the data and reproduce the results with a reasonable amount of effort
- the 2 data sets used in subsequent studies don't match
"Hide the decline"- the sequel.
BBD
You find Hansen's calculations to be "convincing" and "empirical". I wonder if you would care to comment on his estimate of albedo change (a very critical factor) which just assumes that this is negatively proportional to sea-level^0.8, without a shred of justification. He quite simply ignores vegetation changes, changes in snow cover and changes in cloudiness all of which are known to have been very large between the LGM and the present.
tty
I'm not going to argue about the alleged demerits of H&S' methodology, not least because BH has already cautioned against drifting off topic. However, a misrepresentation is a misrepresentation, and I hope our host will allow a single comment in response.
From the H&S draft paper:
Climate forcing due to surface albedo change is a function mainly of sea level, which implicitly defines ice sheet size. The albedo forcing does not depend sensitively on ice sheet shape or on how many ice sheets the ice volume is divided among (Fig. S4, Hansen et al., 2008). We assume that division of ice between Northern and Southern Hemispheres at the LGM had 75 m sea level in the Antarctic ice sheet and 105 m in Northern Hemisphere ice, as in Hansen et al. (2008), calculating the albedo forcing as a function of sea level with their equation S1. Surface albedo change due to vegetation change is implicitly included in the surface albedo forcing, which is calibrated via global radiation calculations using surface boundary conditions for the LGM and Holocene (Hansen et al., 2008). Because much of the LGM-Holocene vegetation change is inherent with ice sheet area change, ice sheet and vegetation changes are lumped together in assessing surface albedo climate forcing.
And:
It is not important at all, because the carbon dioxide climate sensitivity is in fact neither 3°C/doubling of CO2 nor 1°C/per doubling, but zero. The Venus atmosphere has 96.5% carbon dioxide, over 2400 times the concentration of carbon dioxide in Earth's atmosphere, or over 11 doublings--but the Venus/Earth temperature ratio is physically explained, fully and precisely, at points of equal pressure over the full range of Earth tropospheric pressures, only by the ratio of the two planets' distances from
the Sun.
Jun 25, 2012 at 3:30 PM Harry Dale Huffman
HDH - this is not a trick question on my part. There seems no doubt that the earth has been far colder in the past, even within the existence of mankind. What changed between then and now? The sun's output?
Slightly O/T but I am reminded of the paper by Singh on diet and heart disease where the the original data
were eaten by termites.
Rather lengthy and the termite bite comes near the end:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1181273/
3C versus 1C is a side show. The real issue is whether climate models with hidden assumptions and randomized runs should be used to generate those misnamed "probability density functions" shown. There is virtually no empirical data to support most of the depicted probability curves and Forest does not even have the saving grace of any empirical data behind it. It is simply an expression of the arbitrary sensitivity programmed into the model before it is run.
BBD, from your quoted text: "Indeed, the empirical paleoclimate estimate of climate sensitivity..."
Paleoclimate sensitivity estimates are not empirical. For example, dust decreases from glacial to interglacial, causing warming, and the magnitude of that change is not measured (not empirical). It must be determined with climate models (the same ones used for contemporary sensitivity estimates). Otherwise there is no way to know the warming contributed by factors such as changes in clouds, the jet stream, dust, etc. all of which have different measurements than today's climate.
Eric (skeptic) wrote: "Forest does not even have the saving grace of any empirical data behind it"
That is not fully correct. Although the MIT 2D model used in Forest 2006 may not accurately simulate patterns of temperature changes, including their magnitudes, it does so on the basis of user-adjustable parameters, which there has been some attemtp, at least, to calibrate. The simulated results are then compared with multiple observed temperature changes. The probability densities are intended to reflect how closely the simulated (average of 4 model runs, to reduce noise) temperatures match the observed ones. The peak of the PDF corresponds to the best match.
There are various weaknesses in this method, but IMO it is far superior to simply using multiple runs of full scale (AOGCM) climate modelsas various uncalibrated physical parameters are varied, or simulations by many different AOGCMs. That, I would agree, tell one more about the models than the real world.
Forest or NicL?... I know where I'd put my money.
Thanks for all the work you put in Nic - a hell of a lot more than just cutting and pasting other people's opnions, that's for sure.
Nic, thanks for the correction. Does the calibration to present-day climate mean the model can accurately predict the climate (i.e.weather) regime with CO2 warming? For example, if the jet stream is displaced poleward with warming, would the model capture that or be able to calculate the water vapor feedback (i.e. weather response) to CO2 warming accurately? Would the model calibration match to present temperatures really be equivalent to a probability of sensitivity in the future scenario? That seems like a stretch.
Eric (skeptic): "Does the calibration to present-day climate mean the model can accurately predict the climate (i.e.weather) regime with CO2 warming? "
Sorry, I'm not a climate model expert. But, as I understand it, the most important and difficult calibration involved here (of climate sensitivity) isn't to present-day climate, but against AOGCMs whose climate sensitivity is known (since they have been run for many hundreds of years, to equilibrium).
The entire CAGW paradigm could be summed up by an old Kiwi saying - 'It can't fall over, mate, as there's nothing holding the bloody thing up'.
And in the many years I spent as a classroom teacher, I don't think I ever heard the excuse 'The dog ate my homework' as most teenagers have far more creativity (and credibility) than scientists promoting catastrophist views appear to own. 'Losing my homework' was an occasional first try-on which was usually discarded, after direct questioning, in favour of 'I didn't actually do it, Sir'.
Whay are climate scientists so amateurish ?
It's really nice to see more clever people finding errors and problems in the CAGW litterature.
I am sure mr McIntyre would have gotten around to it eventually, but he couldn't get the time to see to Forest for all the trees.
Why should Dr. Forest give NicL his data when Nic's aim is to find something wrong with it? /sarc
The dog ate Bart Simpson's homework once. Honest!
Edvin,
Shouldn't you be looking for your coat after that one? ;-)
How to make Money in Climate Science
1. find a major unanswered question.
3. question top scientists to see what they will accept as an answer
3. cherry pick data and methods to arrive at that answer
4. re-label this technique “training” – it makes it sound intelligent.
5. publish the result.
The results will seem correct to fellow scientists, especially those at the top, so they wont bother to check the math. Everyone will be impressed you have answered the hard question. More so because you will have proven their best guess correct and made them look good in the process. You will advance in your career in science. Fame and fortune will follow.
If anyone does question the results:
6. stall - play for time - they will likely give-up.
7. if not - lose the data and methods