More low climate sensitivity

This paper A lower and more constrained estimate of climate sensitivity using updated observations and detailed radiative forcing time series from Skeie et al is now in open peer review at Earth System Dynamics. They say an ECS of 1.84

See Figure E2 from the paper.

Skeie_ECS_FIGe2

Abstract:

The equilibrium climate sensitivity (ECS) is constrained based on observed near-surface temperature change, changes in ocean heat content (OHC) and detailed radiative forcing (RF) time series from pre-industrial times to 2010 for all main anthropogenic and natural forcing mechanism.

The RF time series are linked to the observations of OHC and temperature change through an energy balance model and a stochastic model, using a Bayesian approach to estimate the ECS and other unknown parameters from the data. For the net anthropogenic RF the posterior mean in 2010 is 2.1 W m−2 with a 90% credible interval (C.I.) of 1.3 to 2.8 W m−2, excluding present day total aerosol effects (direct + indirect) stronger than −1.7 W m−2. The posterior mean of the ECS is 1.8 °C with 90% C.I. ranging from 0.9 to 3.2 °C which is tighter than most previously published estimates.

We find that using 3 OHC data sets simultaneously substantially narrows the range in ECS, while using only one set and similar time periods can produce comparable results as previously published estimates including the heavy tail in the probability function. The use of additional 10 yr of data for global mean temperature change and ocean heat content data narrow the probability density function of the ECS. In addition when data only until year 2000 is used the estimated mean of ECS is 20% higher. Explicitly accounting for internal variability widens the 90% C.I. for the ECS by 60%, while the mean ECS only becomes slightly higher.

h/t to Bishop Hill

About these ads

54 thoughts on “More low climate sensitivity

  1. In the limit, as they say in calculus, as time tends to infinity (or whenever) ‘climate sensitivity” will tend to 1…

  2. Alas for climatologi$ts, the vast fluid dynamic ocean that dominates the heat cycles known as climate doesn’t have equilibrium in its résumé.

  3. OldWeirdHarold says:
    August 14, 2013 at 9:31 am
    This is the Kubler-Ross stage known as “bargaining”.

    ######

    no its called having more data.
    1. 10 more years of surface data
    2. 10 more years of RF
    3. 10 more years of OHC

    If you understand the formula for estimating ECS, its clear why the estimate is coming down and getting more narrow.

    One theme of skepticism is that we need more data.
    When you get more data your estimates can change. especially when you have 10 more years of OHC which is highly informative ( in the baysian sense)

    [For new readers, clarify your abbreviations, please. Mod]

  4. It seems to me that if ECS is about 1.8C, then we need to hit about 620ppm CO2 in the atmosphere to reach the fabled 2C temperature increase which the IPCC considers the safe upper limit.

    Seems like we probably have a good long while before we hit 620ppm.

  5. I believe it; 1.84 it is, and the very next time that earth’s climate is in equilibrium, we will see that for ourselves.

    Unfortunately, I am planning to not be here, next time earth climate is in equilibrium.

  6. I understand the concept of climate sensitivity but I have not explored how it is calculated. However, I get the impression that it is obtained by running the particular model. Is this right?

    So does this mean that the sensitivity is model specific and with 70 plus models giving any temperature prediction you can imagine, provided it is much higher than reality, there will be 70 plus versions of climate sensitivity to choose from?

    Worse than that, since the models are clearly not fit for purpose, does this mean that discussing climate sensitivity is a waste of breath?

  7. What this still does not tell us is the effect of all the feedbacks in the system, both known and unknown, or even if the apparent CO2 sensitivity is an artifact of another process. At least teh sky is no longer falling faster than terminal velocity.

  8. This is not English–

    Explicitly accounting for internal variability widens the 90% C.I. for the ECS by 60%, while the mean ECS only becomes slightly higher.

    Could someone who reads “climate sensitivity-ease” summarize please?

  9. This is consistent with the observations to date.

    And observations are 1,000 times more valuable than a climate model output or theroritical guesses.

  10. Q: How do climate modelers keep the CAGW gravy train chugging along?
    A: With a tweak-tweak here,
    Fudge-fudge there,
    And a couple of tra-la-las,
    That’s how we graph the Pause away
    In the Climatists’ land of Oz.

  11. Feedbacks ultimately tend to dampening, positive feedbacks are unstable and tend toward not existing. It’s highly unlikely the CS is greater than 1.2C for doubling. It is as likely the CS would go negative as it would go above that (for very long).

  12. Day by Day, basically what they are saying is that if you take into account the variability of the climate system, their best estimate (C I ) gets 60% greater. So the say .9 to 3.2 is their C I, or a total of 2.3C. 60% of that is 1.2C roughly, so the .9C and 3.2, combined would decrease and increase by approx 1.2C, where the total difference would be 3.5 instead of 2.3C. They also say that their mean estimate of 1.8C would not increase greatly, hope that helps. They did not define in the abstract, if the increase is equal on both the upper or lower, typically as we are seeing in the leaked IPCC5, they will put more of the increase on the higher side.

  13. “Day By Day says:
    August 14, 2013 at 10:10 am
    This is not English–

    Explicitly accounting for internal variability widens the 90% C.I. for the ECS by 60%, while the mean ECS only becomes slightly higher.

    Could someone who reads “climate sensitivity-ease” summarize please?”

    It means that if you take into account that climate is variable (huh, whodathought that?) the uncertainty becomes larger (surprise!) but the average sensitivity estimate barely increases.

  14. One basic problem. The OHC values are not much more than wild guesses before 2003. So, while the effort is commendable, the result is questionable. We should have some idea in another 50 years or so when we have at least one full 60 years cycle worth of data.

  15. Thanks to:

    brian says:
    August 14, 2013 at 10:28 am
    and
    tty says:
    August 14, 2013 at 10:29 am

    the combined answers gave me a clue as to what they were saying!

  16. It’s interesting that the lower end of their estimate would require negative feedback. As time goes on estimates will continue to fall. I’m betting on a final number between .5 and 1.5.

  17. Old Weird Harold nails it, this fictitious “constant” of climatology numeracy, is whatever the computer modeller requires to produce the desired output.
    Garbage in Gospel out.
    In the real world its called a fudge or wild ass guess and the mumbo jumbo of climatology numeracy ain’t working.
    So now the team is walking it back hoping to approximate reality and deny their previous BS.
    What is next? They will discover statistical competency?

  18. Driving a late model sedan, hands fiercely gripping the steering wheel, he tuned the radio more finely and stepped on the gas, “I’ll show her,” he smirked silently, “we’re not lost, Carbonville will be just around that bend.” An hour later and deep in the forest, now, miles from anywhere, he glanced at the clock on the dashboard and frowned. “Better slow down,” he muttered, “might be lost.”

    “What’s that? Slowing down, dear?” his wife murmured, “Possibly willing to admit that I was right — that Carbonville is NOW 100 miles in the opposite direction?”

    “No!” he snapped. “I just have more data now. It makes sense to slow down.”

    “Whatever you say, dear.”

  19. Elliot, the climate sensitivity number includes feedback. If it did not it would be the same as the lab tests on CO2, which is 1C for 2xCO2.

  20. Mosher, I take it that your meaning is that they did not have enough data 10 years ago to make accurate predictions – and yet they made predictions. And it follows that they may not have enough data today – and yet they continue to make predictions.

  21. @ Schrodingers Cat

    “I understand the concept of climate sensitivity but I have not explored how it is calculated. However, I get the impression that it is obtained by running the particular model. Is this right?

    So does this mean that the sensitivity is model specific and with 70 plus models giving any temperature prediction you can imagine, provided it is much higher than reality, there will be 70 plus versions of climate sensitivity to choose from?

    Worse than that, since the models are clearly not fit for purpose, does this mean that discussing climate sensitivity is a waste of breath?”

    Basically Yes. Nicely put!

  22. If you look at the observed data that any layman with a science background can understand (pre-industrial temps & CO2 concentration + current temps & CO2 concentration) & calculate sensitivity on that basis, you will come in very close to the P50 of this paper (1.7 to 2.0, depending on datasets used) – so this result is really no surprise at all.

    Of course, as we all know, the CAGW hypothesis has always been more about politics than science so I am sure this latest publication will do nothing to sway the faithful.

  23. So, what is the equation for ECS and what are its dependent variables?

    Can we have a running month by month ECS estimate as the RF, OHC and surface data come in?

  24. Skeie et all wrongly claim their sensitivity estimate is based on

    all main anthropogenic and natural forcing mechanism[s].

    They are still not accounting any indirect solar effects so their sensitivity estimate-range is still far too high. Their abstract is not explicit, but it seems that they are using the same RF estimates as the IPCC, just with a little more data. That means the only solar effect they are accounting is the very slight variation in total irradiation but the paleo-evidence shows a much stronger correlation between solar activity and climate than can be accounted for by TSI variation, giving a strong indication that some kind of solar amplifying mechanism is at work. In the scheme of the IPCC and of these Skeie et al., whatever work that amplifying mechanism is doing is getting misattributed to CO2, which was coincidentally correlated with high solar activity over the 20th century.

    Probably the actual sensitivity is less than one (temperature forcings are damped rather than amplified), but either way, solar effects are likely to outweigh CO2 effects (already indicated by the lack of 21st century warming) and the rational expectation going forward is for cooling, not warming.

  25. Steven Mosher says:
    August 14, 2013 at 9:47 am
    “… If you understand the formula for estimating ECS, its clear why the estimate is coming down and getting more narrow. …”

    If you understand the decline effect described by Jonah Lehrer here (The New Yorker)and here (Wired), and the uses of the funnel plot in meta-analysis, it is even clearer why the estimate is coming down and getting more narrow.

    Although the decline effect is unrelated to the decline of “hide the decline” fame, it is yet another decline that the catastrophic climatists will try to hide.

  26. Steven Mosher says:
    August 14, 2013 at 9:47 am
    OldWeirdHarold says:
    August 14, 2013 at 9:31 am
    This is the Kubler-Ross stage known as “bargaining”.
    ######
    no its called having more data

    ————————————————

    More data has only an effect, because AMO/PDO/Sun have been ignored/undervalued before

  27. Steven Mosher says:
    August 14, 2013 at 9:47 am

    “OldWeirdHarold says:
    August 14, 2013 at 9:31 am
    This is the Kubler-Ross stage known as “bargaining”.

    One theme of skepticism is that we need more data.
    When you get more data your estimates can change.”

    So how have guys like Guy Callandar (1938) ECS 1.67, Lindzen over a decade ago ~1.2 or so and Roy Spencer similar numbers gotten ahead of the curve? Look Steven, if one has accepted CO2 as the overriding principal element and there is a “range” that permits some subjective choice in climate data, this unavoidably results in choosing the warmer alternatives available particularly if there is a motive to move governments to action. If they actually had some evidence that it might be 1.5 to 2.0 they would look where they could alter the weights in their models (and we’re aware that even the data can be coaxed to suit) to make the number an index for disaster (who gives a damn about 1.5 of doubling when we have centuries to mitigate problems if they become evident). It’s just psychology.

    When skeptics were not a major force on the scene – the golden age of monopoly climate science- this resulted in the cats-away-mice-will-play effect. When IPCC was talking, what, 4-7?
    But when peer reviewed science began to question the unassailed hypothesis, after climategate released waffling skeptics from reputable institutions (and emboldened for sure by the comparatively few, untenured or retired skeptics who had been fighting a lopsided battle) and loosened the locked chains on tyrannized journals, ECS, among other things started to change and quickly. When nature joined the fray after the warming hiatus could no longer be decently ignored, then everything was up for grabs. Golden age climate scientists (GACS) were thunderstruck into disarray for a few years, trying to rationalize where the heat was hidden, flinging out how long it was acceptable for the heat to be flat, adjusting the HadCrut (4) up a notch, sayiing that oh we predicted this a long time ago, the Met Office forming a group of their kind to investigate, not their disappointing forecasts, but rather UK’s “disappointing weather” (it had been lousy as it always has been). Poor undergraduates (fresh faces) were herded into getting end-of-the-world studies out for the IPCC AR5 deadline (which had been extended several months) and the sold out jounals compressed the review purpose from a year to a month. These hurried works were mercilessly dismantled in hours by Steve McIntyre – a one man wrecking ball. Irresponsible, dishonest psychologists’, sociologists’, geographers’ (I thought this contemporary of alchemy had ceased to exist!), business school professors’, butchers’, bakers’ and candlestick makers’ climate science articles were now outnumbering golden age climate science papers. Then grudgingly at first they began to resurrect and beef up the buried “climate variability” taboo. Emboldened they decided to push this factor up even more to show how CO2 was still king and we were just lucky that some cycle saved us from certain death. Then the heart of the matter couldn’t be avoided any longer. Most had stopped publishing in their despair and had taken to going to “T” groups for communication therapy and giving oral presentations, some for cash. Institutes started handing out Cracker Jack (TM) prizes and medals to the disheartened GACS folks. Gritting their teeth, the then began to talk openly about the “E” word. Heck, there was a new ECS estimate a week for several months. Plot them on a curve and you could see the asymptote ~ 1.0 or thereabouts.

    Yes, Steven, Old Weird Harold is right, it is a negotiation.

  28. Another lustrum, or perhaps two, and the sensitivity estimates will have declined to the value that physics tells us, 1.23 degrees per doubling of CO2.

  29. Anthony,

    How about a scatter plot:
    x-axis is date of prediction
    y-axis is ECS

    It is the input data to my climatologist model.

  30. With every new report, the trend of climate sensitivity progresses down the slippery slope towards zero. When then, will folk actually take a serious look at the (seemingly unacceptable) possibility that it actually might in fact BE zero, i.e., that CO2 has no effect whatsoever in driving temperature, and that the climate alarmists claim of a greenhouse effect doesn’t actually exist?

    I realise that this notion is sacrilege to some, but scientists should never discount any possibility, but examine it with equal vigour and dispassionate objectivity as any other hypothesis.

  31. @ A. D. Everard — Thank you! And thanks for taking the time to say so.

    The story has a sad ending, the road dead ends at a rock wall before he takes his foot off the gas… . Okay, okay, the air bags kick in and they are okay. But, their model is totaled. BWAH, HA, HA, HA, HAAAAAAAAAAAAAAA!

    ******************************
    “Golden age climate scientists (GACS)” [Gary Pearse at 12:42PM] — LOL. Very nice article, Mr. Pearse, for the Encyclopedia Britannica for the “Science Blunders” section. You covered it all with accuracy and flair.

  32. Thanks Caveman, I liked that read.

    “A significant factor contributing to the decline effect can also be the sample size of the scientific research, since smaller sample size is very likely to give more extreme results, suggesting a significant breakthrough, but also a higher probability of an error. Typical examples of this effect are the opinion polls, where those including a larger number of people are closer to reality than those with a small pool of respondents.This suggestion would not appear to account for the observed decrease over time regardless of sample size. Researcher John Ioannidis offers some explanation. He states that early research is usually small and more prone to highly positive results supporting the original idea, including early confirmatory studies. Later, as larger studies are being made, they often show regression to the mean and a failure to repeat the early exaggerated results”

  33. If you look at the IPCC reports, they rate ALL factors other then co2 as having low consensus and low levels of understanding. Even if the sensitivity claims were 100% accurate no one that claims we KNOW the world will warm because of it should be able to retain the title “scientist”. (of course it isnt warming as claimed anyway LOL) You cant understand one out of a dozen factors in a dynamic moving somewhat self correcting system and pretend you can get valid results running simulations.

    Ive poured over our historical data and see no indication co2 does as claimed. Instead I see co2 following not leading temps and even peaking over levels claimed to trigger unstoppable warming as temps continue to fall, with co2 slowly following. yet today magically co2 will raise temps in decades?

    I WISH co2 could heat as claimed, because interglacial periods seem to be trending cooler and shorter. We would have to go up several degrees to be as warm as the last interglacial period. So pieces like this are meaningless to me. Until we have a decent grasp of natural factors its simply impossible to have assurance of what anthropogenic factors could be doing. Id be shocked if we found humanities activities from land use changes to several other factors were not affecting climate, but calling most of climate science science is a joke. Including these papers lowering climate sensitivity. They literally still use the same assumptions of past work, weighting variables we do not have numbers we KNOW are accurate as others have done and leaving co2 the balance. It simply isnt possible to get correct values in that paradigm.

    Future generations will look at us and laugh the way we might laugh today at the science of the 1850s.

  34. You cannot do controlled experiments on the earth’s atmosphere, so you cannot measure climate sensitivity. Any values arrived at by estimations (not calculations) are little more than guesses. What little emipirical data we have, gives a strong indication that the value of climate sensitivity is indistinguishable from zero (c.f Bennstock et al).

  35. Mosher writes “no its called having more data.
    1. 10 more years of surface data
    2. 10 more years of RF
    3. 10 more years of OHC”

    When you have less data you have greater error margins. But if you have a correct method, you still have an answer within those error margins. The IPCC has said the ECS is between 1.5 and 4.5C and 1.84 is still in that range to be sure. However its getting towards the edge of it and so that puts the method itself into question.

    Or perhaps it puts even more fundamental assumptions about “cause” into question.

  36. I’m pretty vague about the all the maths, but I can tell you that Brisbane is getting warmer. The last few days have been quite pleasant. So Global Warming must be true!

  37. Janice Moore says:
    August 14, 2013 at 2:02 pm

    “Golden age climate scientists (GACS)” [Gary Pearse at 12:42PM] — LOL. Very nice article, Mr. Pearse, for the Encyclopedia Britannica for the “Science Blunders” section. You covered it all with accuracy and flair.

    Janice Moore says:
    August 14, 2013 at 10:48 am

    “…now, miles from anywhere, he glanced at the clock on the dashboard and frowned. “Better slow down,” he muttered, “might be lost.”

    “What’s that? Slowing down, dear?” his wife murmured, “Possibly willing to admit that I was right — that Carbonville is NOW 100 miles in the opposite direction?”

    “No!” he snapped. “I just have more data now. It makes sense to slow down.”

    Janice, I was just saying the same thing as you did above, only you did it much more succinctly. Oh and I loved your Martian GACS chattering gac gac gac. It looks like a mission searching for a new hair dresser to replace the single one they appear to have on Mars.

  38. Well, Mr. Pearse, you are too kind. Thank you. I think you covered a LOT more of the field, though, than my Lucy and Desi skit. So nice that one of the brainy scientists here (YOU, O Pearse Effect genius — no, I’m not being sarcastic) liked it. Your theory sounds highly plausible to me, I never say so because my non-scientist opinion is worth little toward serious evaluation of your hypothesis — I wish some of the guys (and gals) here would give you more thoughtful feedback on it… . Keep pursuing your idea!

    Yeah, “Mars Attacks” is one of the best parodies I’ve ever watched. Lots of great vignettes and truth in it. It even has a happy ending. Warning, though, don’t let children under 12 or so watch it — I thought it would be okay for an 11-year-old boy, but, it gets pretty ugly at times and, even now, thinking about the horror etched on his face, I regret not shutting it off. Well, I survived watching all those old horror movies (“Dracula”, etc) around age 12, alone in the dark rec room (I was having fun!). [if this thread weren’t ebbing to its end, I wouldn’t waste this space chatting away, I hope you know

    SCIENCE, SCIENCE, SCIENCE — there, back on track, heh, heh]

  39. Schrodinger’s Cat says:
    August 14, 2013 at 9:53 am

    “…Worse than that, since the models are clearly not fit for purpose, does this mean that discussing climate sensitivity is a waste of breath?”
    //////////////////////////////

    YES. YES. YES.

    Any discusssions of Climate Sensitivity are disengenuous unless they make it clear that presently we are not in a position to estimate this concept from observational evidence/data, and that any figure proposed is just a speculative guess.

    Climate sensitivity and natural varuiation are instrinsically linked. Climate sensitivity can at best expalin any temperature change whiich temperature change is not itself the result of natural variation.

    Accordingly, before one can begin to meaningfully discuss climate sensitivity one first has to eliminate any and all temperature changes brought about by natural variation. It is only then that one will be left with the signal (if any) of climate sensitivity.

    Accordingly as a matter of logic, we first need to fully know and understand everything there is to know and understand abouut natural variation, before we can start to evaluate climate sensitivity. Accordingly we need to know and identify each consituate forcing which goes to make up natural variation, the direction of that forcing and the upper and lower bounds of each and every constituent forcing.

    Since we do not yet know what natural variation encompasses, we cannot begin to meaningfully discuss climate sensitivity.

    All we know are the most general of matters such as natural variation trumps climate sensitivity (ie., is a stronger force) as is apparent form the 1940s to 1970s cooling (natural variation was stronger than any CO2 forcing brought about by increasing CO2 emissions/levels during that period), and is at least equal to climate sensitivity encompassing the residual CO2 emissions of the last 50 to 70 years (depending upon claimed residency times) plus recent emissions (say post 1995) in view of the temperature stasis these past 16 to 22 years (depending upon data set used).

    Given that there is no first order correlation between CO2 and temperature in the sat record; it being essentially flat between say 1979 and up to and around the 1998 super El Nino, and then essentially flat post that El Nino to date, with only a step change taking place in and around the Super El Nino then unless that El Nino was in some way caused by CO2 (and as far as I know no one alleges that it was caused by CO2 emissions), one would be led to conclude from 33 years worth of sat data that climate sensitivity is so low that it cannot presently be measured with present equipment and the limits of resolution of that equipment.

  40. Jimmy Haigh. says:
    August 14, 2013 at 9:32 am

    AND

    Steven Mosher says:
    August 14, 2013 at 9:47 am
    /////////////////////////

    Jimmy, I think you mean to zero.

    The longer the present temperature stasis continues, the more apparent that climate sensitivity is low and the longer it continues it will begin to approach zero.

    There is reason to consider that it may be rather close to zero, eg the 1940 to 1970s cooling which more than off-set any positive climate sensitivity to the forcing of rising CO2 levels. And the present day stasis, which equals the climate sensitivity to the combined effects of long and medium term CO2 residency of CO2 emissions these past 70 years and the substantial emissions which have taken place these past 20 years with no increase in temperature. To this one must add the paleo record when there are instances of high levels of CO2 and low temperatures and rising CO2 and falling temperatures etc.

    Steven is right on the more recent data and the need for more data. As this comes in, it is becoming an inescapable fact that past estimates (which were no more than guesses) were way too high. If this temperature statsis contines, even more so should temperatures fall during the next 20 years, climate sensitivity will be seen at most to be nominal.

  41. For those interested in Climate Sensitivity and the carbon cycle there was a very good (and clear) debate on this issue in the recent article “Murry Salby responds to critics ” See:

    http://wattsupwiththat.com/2013/08/11/murry-salby-responds-to-critics/#comments

    This debate is well worth a read. The contributions by Bart, Richard Courtney, dbStealey, Allan MacRae, and (for the warmists) by Nick Stokes, Ferdinand Engelbeen and Nyq Only are in particular worthy of note. The debate was very clearly set out and the contributors are to be congratulated for the clarity of their response which makes it particularly easy for the reader to understand and follow the arguments being made. I highly recommend a read of this debate as it is is illuminating.

    In this dabate Bart produced evidence that suggested that CO2 is driven by temperature, and the consensus view that CO2 drives temperature is not supported by the observational data. Now I do not wish to misinform what was conceded, but I gained the impression that it was accepted that temperature drives CO2 but only as far as recent response is concerned and it is a short term effect, not a long term response.

    There was a debate whether there could be a different short and long term response, and whether the long term response is simply an accumulation of the short term response, but leaving that issue to one side, the effect of the concession to Bart’s evidence (see http://www.woodfortrees.org/plot/esrl-co2/derivative/mean:12/from:1979/plot/rss/from:1959/scale:0.2/offset:0.125) would seem to make it very difficult to measure climate sensitivity from short term observational data. So whilst Steven Mosher call for more data and extolls the virtue of the examination of recent data (something that I do not disagree with), it may well be the case that this data will give little insight into climate sensitivity because of the link 9at least in the short term response0 that changes in temperature drive changes in CO2 levels).

  42. Gary Pearse says:
    August 14, 2013 at 12:42 pm
    ……………..
    So how have guys like Guy Callandar (1938) ECS 1.67, Lindzen over a decade ago ~1.2 or so and Roy Spencer similar numbers gotten ahead of the curve? Look Steven,…………

    Gary, yours was a truly fine response to Mosher and a great, humorous summary of the path climate science has travelled. Well done.

  43. UnfrozenCavemanMD says:
    August 14, 2013 at 11:14 am

    Steven Mosher says:
    August 14, 2013 at 9:47 am
    “… If you understand the formula for estimating ECS, its clear why the estimate is coming down and getting more narrow. …”

    If you understand the decline effect described by Jonah Lehrer………

    Thanks! I am still reading the New Yorker’s 5 page article on the issue and I can’t help thinking about climate scientists.

    [The decline effect and the scientific method : The New Yorker
    The Truth Wears Off
    Is there something wrong with the scientific method?
    by Jonah Lehrer December 13, 2010]

    http://www.newyorker.com/reporting/2010/12/13/101213fa_fact_lehrer?currentPage=1

Comments are closed.