NOTE: This has been running two weeks at the top of WUWT, discussion has slowed, so I’m placing it back in regular que. – Anthony
UPDATES:
Statistician William Briggs weighs in here
Eduardo Zorita weighs in here
Anonymous blogger “Deep Climate” weighs in with what he/she calls a “deeply flawed study” here
After a week of being “preoccupied” Real Climate finally breaks radio silence here. It appears to be a prelude to a dismissal with a “wave of the hand”
Supplementary Info now available: All data and code used in this paper are available at the Annals of Applied Statistics supplementary materials website:
http://www.imstat.org/aoas/supplements/default.htm
=========================================
Sticky Wicket – phrase, meaning: “A difficult situation”.
Oh, my. There is a new and important study on temperature proxy reconstructions (McShane and Wyner 2010) submitted into the Annals of Applied Statistics and is listed to be published in the next issue. According to Steve McIntyre, this is one of the “top statistical journals”. This paper is a direct and serious rebuttal to the proxy reconstructions of Mann. It seems watertight on the surface, because instead of trying to attack the proxy data quality issues, they assumed the proxy data was accurate for their purpose, then created a bayesian backcast method. Then, using the proxy data, they demonstrate it fails to reproduce the sharp 20th century uptick.
Now, there’s a new look to the familiar “hockey stick”.
Before:
After:
Not only are the results stunning, but the paper is highly readable, written in a sensible style that most laymen can absorb, even if they don’t understand some of the finer points of bayesian and loess filters, or principal components. Not only that, this paper is a confirmation of McIntyre and McKitrick’s work, with a strong nod to Wegman. I highly recommend reading this and distributing this story widely.
Here’s the submitted paper:
(PDF, 2.5 MB. Backup download available here: McShane and Wyner 2010 )
It states in its abstract:
We find that the proxies do not predict temperature significantly better than random series generated independently of temperature. Furthermore, various model specifications that perform similarly at predicting temperature produce extremely different historical backcasts. Finally, the proxies seem unable to forecast the high levels of and sharp run-up in temperature in the 1990s either in-sample or from contiguous holdout blocks, thus casting doubt on their ability to predict such phenomena if in fact they occurred several hundred years ago.
Here are some excerpts from the paper (emphasis in paragraphs mine):
This one shows that M&M hit the mark, because it is independent validation:
In other words, our model performs better when using highly autocorrelated
noise rather than proxies to ”predict” temperature. The real proxies are less predictive than our ”fake” data. While the Lasso generated reconstructions using the proxies are highly statistically significant compared to simple null models, they do not achieve statistical significance against sophisticated null models.
We are not the first to observe this effect. It was shown, in McIntyre
and McKitrick (2005a,c), that random sequences with complex local dependence
structures can predict temperatures. Their approach has been
roundly dismissed in the climate science literature:
To generate ”random” noise series, MM05c apply the full autoregressive structure of the real world proxy series. In this way, they in fact train their stochastic engine with significant (if not dominant) low frequency climate signal rather than purely non-climatic noise and its persistence. [Emphasis in original]
Ammann and Wahl (2007)
…
On the power of the proxy data to actually detect climate change:
This is disturbing: if a model cannot predict the occurrence of a sharp run-up in an out-of-sample block which is contiguous with the insample training set, then it seems highly unlikely that it has power to detect such levels or run-ups in the more distant past. It is even more discouraging when one recalls Figure 15: the model cannot capture the sharp run-up even in-sample. In sum, these results suggest that the ninety-three sequences that comprise the 1,000 year old proxy record simply lack power to detect a sharp increase in temperature. See Footnote 12
Footnote 12:
On the other hand, perhaps our model is unable to detect the high level of and sharp run-up in recent temperatures because anthropogenic factors have, for example, caused a regime change in the relation between temperatures and proxies. While this is certainly a consistent line of reasoning, it is also fraught with peril for, once one admits the possibility of regime changes in the instrumental period, it raises the question of whether such changes exist elsewhere over the past 1,000 years. Furthermore, it implies that up to half of the already short instrumental record is corrupted by anthropogenic factors, thus undermining paleoclimatology as a statistical enterprise.
…
We plot the in-sample portion of this backcast (1850-1998 AD) in Figure 15. Not surprisingly, the model tracks CRU reasonably well because it is in-sample. However, despite the fact that the backcast is both in-sample and initialized with the high true temperatures from 1999 AD and 2000 AD, it still cannot capture either the high level of or the sharp run-up in temperatures of the 1990s. It is substantially biased low. That the model cannot capture run-up even in-sample does not portend well for its ability
to capture similar levels and run-ups if they exist out-of-sample.
…
Conclusion.
Research on multi-proxy temperature reconstructions of the earth’s temperature is now entering its second decade. While the literature is large, there has been very little collaboration with universitylevel, professional statisticians (Wegman et al., 2006; Wegman, 2006). Our paper is an effort to apply some modern statistical methods to these problems. While our results agree with the climate scientists findings in some
respects, our methods of estimating model uncertainty and accuracy are in sharp disagreement.
On the one hand, we conclude unequivocally that the evidence for a ”long-handled” hockey stick (where the shaft of the hockey stick extends to the year 1000 AD) is lacking in the data. The fundamental problem is that there is a limited amount of proxy data which dates back to 1000 AD; what is available is weakly predictive of global annual temperature. Our backcasting methods, which track quite closely the methods applied most recently in Mann (2008) to the same data, are unable to catch the sharp run up in temperatures recorded in the 1990s, even in-sample.
As can be seen in Figure 15, our estimate of the run up in temperature in the 1990s has
a much smaller slope than the actual temperature series. Furthermore, the lower frame of Figure 18 clearly reveals that the proxy model is not at all able to track the high gradient segment. Consequently, the long flat handle of the hockey stick is best understood to be a feature of regression and less a reflection of our knowledge of the truth. Nevertheless, the temperatures of the last few decades have been relatively warm compared to many of the thousand year temperature curves sampled from the posterior distribution of our model.
Our main contribution is our efforts to seriously grapple with the uncertainty involved in paleoclimatological reconstructions. Regression of high dimensional time series is always a complex problem with many traps. In our case, the particular challenges include (i) a short sequence of training data, (ii) more predictors than observations, (iii) a very weak signal, and (iv) response and predictor variables which are both strongly autocorrelated.
The final point is particularly troublesome: since the data is not easily modeled by a simple autoregressive process it follows that the number of truly independent observations (i.e., the effective sample size) may be just too small for accurate reconstruction.
Climate scientists have greatly underestimated the uncertainty of proxy based reconstructions and hence have been overconfident in their models. We have shown that time dependence in the temperature series is sufficiently strong to permit complex sequences of random numbers to forecast out-of-sample reasonably well fairly frequently (see, for example, Figure 9). Furthermore, even proxy based models with approximately the same amount of reconstructive skill (Figures 11,12, and 13), produce strikingly dissimilar historical backcasts: some of these look like hockey sticks but most do not (Figure 14).
Natural climate variability is not well understood and is probably quite large. It is not clear that the proxies currently used to predict temperature are even predictive of it at the scale of several decades let alone over many centuries. Nonetheless, paleoclimatoligical reconstructions constitute only one source of evidence in the AGW debate. Our work stands entirely on the shoulders of those environmental scientists who labored untold years to assemble the vast network of natural proxies. Although we assume the reliability of their data for our purposes here, there still remains a considerable number of outstanding questions that can only be answered with a free and open inquiry and a great deal of replication.
===============================================================
Commenters on WUWT report that Tamino and Romm are deleting comments even mentioning this paper on their blog comment forum. Their refusal to even acknowledge it tells you it has squarely hit the target, and the fat lady has sung – loudly.
(h/t to WUWT reader “thechuckr”)
—————
Gail Combs,
I enjoyed your post and agree with your critique of the mainstream US educational philosophy of the majority of the 20th century. Dewey was a part of the so-called American Pragmatist philosophical movement. I agree it had its root in the Kant to Hegel stream. The American Pragmatist movement had a high content borrowed directly from then current European thinking . . .
I have been a long term admirer of Marie Montessori’s educational methods (k through 12).
Also, I have been a long term admirer of the greater merits of private versus public education including collegiate level. My problem with US public education is the politicization, the group think base of its techniques and the dumbing down of the higher to the lower performers in a student body. Private I find is, overall, more individualistic, varied and has better focus on achievement. NOTE: I was educated mostly in public institutions.
John
So Mann’s sharp uptick is refuted by this paper but it still shows a significant warming trend for the last 150+ years does it not? So the climate has been warming for 150 years.
Scipio says:
“So the climate has been warming for 150 years.”
Is that in dispute? The LIA ended about that time.
Smokey says:
“Is that in dispute?”
I don’t know for sure, as I’m certainly no expert in this field, but I don’t believe it is and that most people agree that it has been. What I am a bit perplexed about is that Mann, while fudging his numbers by reducing the size of his data set until he obtained that data he needed to support his ideas about AGW, gave us the infamous hockey stick. Yet I look at the 2 graphs and while Mann’s severe uptick has been smoothed out by the work of McShane and Wyner their graph too shows(?) a somewhat significant warming trend over the last 30(+/-) years.
The warming over the last 200 years (as shown by the graphs) coincides with the beginning of the burning of large volumes of fossil fuels. This may just be coincidental or perhaps not, I don’t study this type of science.
I personally don’t believe in AGW, but I sometimes wonder.
[snip – thanks for that, but I’m not going to let that confusion over being the first running a press release versus the spotting actual event take over this thread, since the two aren’t related. If you wish, you can repost on the source thread where it is relevant. I’d move it, but wordpress.com has no move feature – Anthony]
Atomic Hairdryer:
Thankyou for your lucid post at August 18, 2010 at 6:21 am .
You state the issues exactly.
Each of your points is important and I choose to comment on only one of them. You say:
“Is there a 900 year cycle, or are people confusing correlation with causation, as has happened with CO2 and temperature.”
The test of this is prediction. If the ~900 year cycle exists then during this century the global temperature will start to fall towards the value it had in the LIA whether or not atmospheric CO2 concentration continues to rise. (Please see my post at August 17, 2010 at 2:52 am).
Some may say this is too long a time to wait, but science does not set such limits. For example, Halley made a prediction for the return of ‘his’ comet that proved correct long after his death. And the proof of the present AGW ‘projections’ for the next 50 and 100 years will not be seen for several decades to come.
Even then, “one swallow does not make a Spring” so a single correct prediction based on the assumption of a ~900 year oscillation could be a result of mere chance so – although it would provide confidence to the assumption – it would not be a proof that there is a ~900 year oscillation. Hence, the importance of your comment concerning statistical evaluation.
However, the apparent existence of the ~900 year cycle does provide a reason to seek a plausible and falsifiable physical explanation for such a cycle.
Richard
Scipio, the amount of CO2 poroduced 200 years ago was inconsequential. The temp rise then, as it was, was not due to CO2.
This is old romanticist nonsense about the evil Dark Satanic Mills
Scipio
Please look at the Hockey stick at the head of this paper from, around 1600 onwards .
Now look at the worlds oldest actual instrumental records here;
http://i47.tinypic.com/2zgt4ly.jpg
http://i45.tinypic.com/125rs3m.jpg
The distinct downwards trend on the hockey stick blade is not shown by the records-in fact the blade of the stick seems to be pointing in the wrong direction, down instead of up.
CET from 1660 is backed up by various other records which show the latter stages of the early 18th Century warming, such as this one from Uppsala.
http://chiefio.wordpress.com/2009/10/09/how-long-is-a-long-temperature-history/
We are fortunate with this particular record- from our friend Arrhenius’s home town- to have the botanical garden records as well. These take us back to around 1695. Around 1710 the custodians start to plant outside some quite exotic plants-together with mulberries.
The temperature rise throughout the record can be readily sen in this larger graph of CET.
http://homepage.ntlworld.com/jdrake/Questioning_Climate/_sgg/m2_1.htm
The instrumental record is backed up by high quality contemporary observational records. Anyone browsing the diary of Samuel Pepys for January 1660/61-the year the Royal Society was established- would read;
“It is strange what weather we have had all this winter; no cold at all; but the ways are dusty, and the flyes fly up and down, and the rose-bushes are full of leaves, such a time of the year as was never known
So we can see this warm year was unusual and the cold trend can be traced further back before the records, indeed if we look further back, before the English Civil War, we know that the coldest part of this second phase of the LIA occurred in the early part of the 17th Century, so we can actually trace that rise from around 1601, which some say was the coldest year in our history.
So temperatures have been rising -albeit in notable fits and starts- throughout most of the instrumental records with the coldest decade around 1680 and a notable and remarkable rise commencing in 1698 for three decades that exceeds anything seen in the modern warm period.
It would appear that the Giss records -which start at in 1880- merely ‘plug’ into this well documented, gently warming, centuries long trend as a continuation of it -not the start.
http://data.giss.nasa.gov/gistemp/graphs/
Our international institutions appear to have inexplicably forgotten their climate history and not be aware that, far from being ‘unprecedented,’ the apparent cyclical nature of our climate explains the current temperature trends rather nicely.
Historic instrumental temperature records can be found here on my web site together with a variety of related articles.
http://climatereason.com/LittleIceAgeThermometers/
tonyb
stephen richards says:
August 18, 2010 at 5:21 am
“Micheal said: In essence you are correct but the English language is the problem, I think… You must also read this paper with tongue in cheek. Ie look for the hidden sarcasm.”
Many thanks, Stephen. Good to see I was essentially on the nail and I am grateful for your refinement of what the paper means. I played it dead straight, but I did pick up on some likely and rather delicious, if understated, digs! 🙂
James Sexton says:
August 18, 2010 at 6:53 am
“Michael Larkin says:
August 18, 2010 at 1:46 am
That’s pretty much the way I read it, too.”
I enjoyed your expurgation of one particular slap. That might be one I wasn’t overly sure was such… maybe, I thought, they were just being polite to molify the AGW people, i.e to sugar the pill somewhat.
Thank heavens for this great blog where I can check my understanding of a paper knowing I won’t be gunned down if I don’t toe some party line or other.
Smokey says:
“That is exactly what MBH critics have been saying: Mann carefully selected proxies that reflected very local conditions [if they reflect anything at all]. But he foisted off his erroneous results as proof of anthropogenic global warming.
Seriously, get a copy of A.W. Montford’s The Hockey Stick Illusion. ”
Right, and somehow you think Mann was wrong to do that, however, McIntyre disagrees not with Mann using PCA (principle component analysis), but with which kind of PCA to use.
Since tree growth can be considered to be a function of more than one variable, at least temperature and rainfall to name two, it seems to me to be appropriate to use PCA to determine which proxies best reflect the temperature to growth function rather than an average of all available proxies.
And it’s a global conspiracy, seriously?
And Davil L said:
“But Mann, not a statistician, applied the statistics properly to the problem?”
So, are you saying one has to be a statistician to use statistics?
Anyway, Mann’s reconstuction agrees better with past sea-levels for the past 1000 years than McShane and Wyner.
Sea level was flat from 1000 to 1850 or so, if McShane and Wyner are correct we should have seen a measurable decrease in sea level over that time period.
barry:
Your post at August 18, 2010 at 6:56 am seems to be an attempt to deflect discussion of this thread into an ‘Angels On A Pin’ debate.
The important issue is that you choose to deny all the historical and archaelogical evidence for the existence of the MWP. But you provide no reason why anybody should agree with your denial.
Furthermore, you have ignored my point that not all the globe warmed – some regions cooled – over the twentieth century but this does not disprove the globe warmed over the twentieth century. Similarly, the finding of a few places that were cooling at the time of the MWP – when most of the Earth’s surface warmed to hotter than now – does not disprove the MWP. Your argument is ‘cherry picking’ of an extreme kind.
The fact is that history, archaelogy and paleo data from locations around the globe agree that most of the globe warmed to be hotter than now during the MWP.
The M&W paper throws doubt on the MBH reconstruction but until some similar analyses are conducted on the hundreds of other paleo reconstructions then their evidence of the MWP remains.
Therefore, I shall ignore all else in your post except your concluding point and question to me; viz.
“In the first decade of the previous two cooling phases (1880 – 1890, and 1940 – 1950), the temperature trend was downwards. This is not the case for the alleged cooling phase 2000 to present.
If you contend that a few more years will make a cooling trend apparent, could you give an estimate (nominate a year in the future) when you think this will be demonstrated. When might be the earliest date we could expect a downward trend from 2000 to emerge?”
There is no reason to expect that this cooling phase must be at the same rate as the previous two because it is a combination of at least two apparent cycles (i.e. the ~900 and the ~60 year cycles). They would only be the same if the phase of the ~900 year cycle were the same in each case, and this does not seem to be so: i.e. the present cooling period seems to have a less negative trend than the last one which seems to have a less negative trend than the one before that.
Richard
Re: Barry and duckster regardning the MWP
Do you believe that the farming done by “vikings” on Greenland during an extended period of time was the result of a sustained local climate variation?
If so what might that kind of variation consist of? A changing gulf stream seems unlikely since both northern Sweden and Norway were inhabited and farmed.
I myself find it very hard that you would in this manner be able to insulate a specific part of the globe for such a long time.
Duckster:
At August 18, 2010 at 7:19 am you assert:
“There simply is no system for establishing peak MWP without temperature proxies. And remember, you can’t use these any more (if you have accepted this paper), but you are willing to fall back on historical records and archeological evidence.”
No!
Proxies of various kinds from individual sites have not been disproved by the M&W paper. And the historical and araeological evidence exists.
Please read my last reply to Barry.
Richard
Pogo says “we have met the enemy and he is us.” Statisticians are not the enemy. They are needed in the very beginning of a reseach project to help frame the question and strategize data collection so that a meaningful answer can be obtained. Statistics at the end of the research project help frame the answer as to what can be said from the data. Statistics in general are straight forward as long as one follows the rules. When combining dis-similar data sets, such as proxy data, there are a lot more rules. One needs some high level statisticians to be sure all the rules have been followed for one to be able to say what one wants to say. It seems that MacIntyre & McKintric in 2004 told Mann that he had not followed the rules and he shouldn’t say what he wanted and ultimately did say about the data. In March of this year, VS on Bart Verheggen’s blog developed the arguement that the entire insturment temperature time series data fell within the boundaries of natural variation. McShane & Wyner demonstrate that the entire last 1000 year proxy data set falls within the statistical limits of natural variation. To me, that is strike three. Its time to listen; hard at times after all the blood, sweat & tears.
Henry@Fred Haynie
thanks Fred!
What do you think of the Easterbrook paper?
http://wattsupwiththat.com/2008/12/29/don-easterbrooks-agu-paper-on-potential-global-cooling/
tonyb, thanks a lot. Your post and links helps to clear some of the fog for me.
bob says: August 18, 2010 at 12:18 pm
“And Davil L said: ‘But Mann, not a statistician, applied the statistics properly to the problem?’
So, are you saying one has to be a statistician to use statistics?”
Not at all. I don’t have a degree in statistics but I use them all the time. However I take the time to learn and use them correctly and seek the advice of statisticians when it really matters. But I wouldn’t prefer my statistics over a professional statistician’s. Your comment seemed to prefer to believe Mann, the amateur statistician who clearly didn’t follow conventional rules of statistics, over professional statisticians.
I tried posting the link to the fantastic take on the paper by Briggs
(http://wmbriggs.com/blog/?p=2773) on Taminos blog but he was
impressively fast in deleting it =)
REPLY: you and others shouldn’t be bashfull about trying again. – Anthony
NeilT says:
August 18, 2010 at 8:44 am
lattituide, you should go back and read the document. That figure shows that they stopped using the proxy at 1968 and then used the statistical method to try and predict the warming from 1969 to 2000. At the same time they put in the real recorded figures (the black line).
The point I was making is that the whole thing is rubbish. The scientists don’t use the proxies to predict the future.
Did you read the part of the paper that explained how paleoclimate proxies work. The proxy data are not temperatures but measurements of various minutial elements assorted artifacts of the past(trees, ice, sediments,etc,). To arrive at a temperature projection a selection of the data is “trained” against know empirical observations from the same time period. A model or algorithm is constructed which, when a proxy datapoint is entered, outputs a temp which matches the empirical observation. Then to verify the model, not to predict the future, the model is applied to forward proxy data points to see if they continue to match the empirical observations. From what I’ve seen, mostly they don’t. Which is probably the main reason Mann spliced actual temp records on to his proxy in his “neat trick”. The point is that, if they can’t match the record going forward, there is little reason to suspect that they can reconstruct the temperature record of the past with any level of certainty.
If you google “McShane and Wyner 2010” you get get 5,500 hits, if you do the same search on google news you get 2 reports: Prisonplanet.com and Dellingpole writing in the Daily Telegraph.
Scare stories sell newpapers, solid debate like is taking place on this thread is not considered newsworthy. There is a bias in the media for the theories of CAGW and a bias against correcting them afterwards.
REPLY: you and others shouldn’t be bashfull about trying again. – Anthony
Bashful just isn’t our style =)
I think it speaks volumes that they are choosing not to respond and the commentary around the net from skilled and insightful people with advanced statistical skills in no way invalidates the “end of the hockey stick” title.
NeilT says:
August 18, 2010 at 8:44 am
lattituide, you should go back and read the document. That figure shows that they stopped using the proxy at 1968 and then used the statistical method to try and predict the warming from 1969 to 2000. At the same time they put in the real recorded figures (the black line).
=======================================================
Neil, I’m looking for an explanation for the following. There has to be a reason, other than an obvious reason. Would there be something written into the programing, numbers, etc
It’s not dumb luck that takes a program that’s all over the place, that random numbers give the same or better results, then all of a sudden at the end, it pulls it’s act together.
What would cause it to be “better than chance” “to predict the final thirty-year block”?
“”“[T]he proxy record has some ability to predict the final thirty-year block, where temperatures have increased most significantly, better than chance would suggest.”
tonyb says:
August 18, 2010 at 12:04 pm
========================
Tony, thank you
Scipio says: ‘I personally don’t believe in AGW, but I sometimes wonder.’
Here is the latest paper from Erl Happ, which has convinced me beyond doubt that a benign trace gas has nothing to do with climate change. ENSO Rules.
http://climatechange1.wordpress.com/
Scipio says:
August 18, 2010 at 11:47 am
“…Yet I look at the 2 graphs and while Mann’s severe uptick has been smoothed out by the work of McShane and Wyner their graph too shows(?) a somewhat significant warming trend over the last 30(+/-) years……”
Scipio, I would highly encourage you to view the graphs while you read the study. It will help greatly to keep the graphs contextually clear.