Skip to content

Still Hiding the Decline

November 29, 2009

Even in their Nov 24, 2009 statement, the University of East Anglia failed to come clean about the amount of decline that was hidden. The graphic in their statement continued to “hide the decline” in the Briffa reconstruction by deleting adverse results in the last part of the 20th century. This is what Gavin Schmidt characterizes as a “good thing to do”.

First here is the Nov 2009 diagram offered up by UEA:

Figure 1. Resized UEA version of Nov 2009, supposedly “showing the decline”. Original here ,

Here’s what UEA appears to have done in the above diagram. While they’ve used the actual Briffa reconstruction after 1960 in making their smooth, even now, they deleted values after 1960 so that the full measure of the decline of the Briffa reconstruction is hidden. Deleted values are shown in magenta. Source code is in first comment.

Figure 2. Emulation of UEA Nov 2009, using all the Briffa reconstruction.

89 Comments leave one →
  1. stevemcintyre permalink*
    November 29, 2009 1:08 pm


    # archive is truncated in 1960:


    #2. J98 has reference 1961-1990
    #note that there is another version at

    test[test== -9.999]=NA
    count= apply(!,1,sum)

    #3. MBH : reference 1902-1980
    MBH99<-read.table(url) ;#this goes to 1980

    #4. CRU instrumental: 1961-1990 reference
    # use old version to 1997 in Briffa archive extended
    Briffa[Briffa< -900]=NA
    Briffa= ts(Briffa,start=1000)
    tsp(CRU) # 1850 1999 #but starts 1871 and ends 1997
    delta # -0.118922
    #used to get MBH values with 1961-1990 reference: compare to 0.12 mentioned in Climategate letters

    #get updated version of CRU to update 1998 and 1999 values
    D=read.table(loc) #dim(D) #158 12 #start 1850
    tsp(cru) # 1850 2009

    # update 1998-1999 values with 1998 values
    CRU[(1998:1999)-1849]= rep(cru[(1998)-1849],2)

    #Fig 2.21 Caption
    #The horizontal zero line denotes the 1961 to 1990 reference
    #period mean temperature. All series were smoothed with a 40-year Hamming-weights lowpass filter, with boundary constraints
    # imposed by padding the series with its mean values during the first and last 25 years.
    #this is a low-pass filter
    source(" ;) #get filter.combine.pad function
    hamming.filter<-function(N) {
    hamming.filter<-0.54 – 0.46 *w
    f=function(x) filter.combine.pad(x,a=hamming.filter(40),M=25)[,2]

    ## WMO Figure at CRU

    X=ts.union(MBH=MBH99+delta,J2001,briffa=gate,CRU=cru ) #collate
    Y=data.frame(X); year=c(time(X))
    sapply(Y, function(x) range(year [!]) )
    # MBH J2001 briffa CRU
    # [1,] 1000 1000 1402 1850
    # [2,] 1980 1991 1994 2009

    smoothb= ts(apply(Y,2,f),start=1000)

    xlim0=c(1000,2000) #xlim0=c(1900,2000)

    plot( c(time(smoothb)),smoothb[,1],col=col.ipcc,lwd=2,bg="beige",xlim=xlim0,xaxs="i",ylim=ylim0,yaxs="i",type="n",axes=FALSE,xlab="",ylab="deg C (1961-1990)")
    usr 1960
    points( c(time(smoothb))[temp],smoothb[temp,"briffa"],pch=19,cex=.7,col=”magenta”)

  2. bender permalink
    November 29, 2009 2:20 pm

    But there’s a trend up at then end, you see? The last 5 data points? Hidden in plain sight.

  3. Hoi Polloi permalink
    November 29, 2009 2:47 pm

    If you rotate the graph 90 degrees anticlockwise you’ll get a nice hockey stick on Briffa 2000…

  4. bender permalink
    November 29, 2009 2:53 pm

    That would be in the prorot.m function? :)

  5. Gary permalink
    November 29, 2009 2:57 pm

    And yet, Phil Jones says in the UEA statement, “We are, and have always been, scrupulous in ensuring that our science publications are robust and honest.”

  6. Rhoda R permalink
    November 29, 2009 3:03 pm

    “We are, and have always been, scrupulous in ensuring that our science publications are robust and honest.”

    For certain values of ‘honest’ and ‘robust’.

  7. jacob permalink
    November 29, 2009 3:30 pm

    I don’t understand why the Jones 1998 and Mann 1999 series no longer break the X axis after 1980 or so as they do in the UEA press release linked to in the “Replicating the Trick Diagram” post. Could someone explain the difference between this and previous versions of the graph?


  8. fulltilt permalink
    November 29, 2009 3:32 pm

    @Phil Jones, “We are, and have always been, scrupulous in ensuring that our science publications are robust and honest.”

    it’s not what you say but what you do.

  9. george hanson permalink
    November 29, 2009 4:00 pm

    CRU dumped the climate change data! Read it hear

  10. November 29, 2009 4:02 pm

    The World Meteorological Organization’s “Statement on the Status of the Climate in 1999″ report is here:

    The graph in which Jones’ email indicates he used “Mike’s trick” to “hide the decline” is apparently the one on the front cover of that WMO report.

    If you look at the graph on the report’s cover, the decline for Briffa’s data stops at about minus 0.1 degree (not minus 0.2 as indicated in the recent UAE graph), at which point it begins going up all the way with the other two lines in the graph to about 0.3 degree. (If you enlarge the page in the pdf file, you can see all three are up there.)

    So, was something hidden?

    The text hints at how the graph could be drawn in the way it was.

    In the report, the front cover image is explained. Here is the first sentence:
    “Front cover: Northern Hemisphere temperatures were reconstructed for the past 1000 years (up to 1999) using palaeoclimatic records (tree rings, corals, ice cores, lake sediments, etc.), along with historical and long instrumental records.”

    And on page 4 is this description:
    “Millennium Temperatures
    Our knowledge of pre-20th century temperature variations in the Northern Hemisphere has increased dramatically in recent years. The availability of natural archives of past temperature such as tree rings, banded corals, ice cores and lake sediments, in addition to historical and long instrumental records, has enabled NH temperature variations to be reconstructed for the last 1000 years at an annual resolution. It is not yet possible to do the same for the Southern Hemisphere due to the lack of adequate palaeoclimatic records. Despite their different emphases on annual or extended summer seasonal temperatures and their different geographical biases, all the reconstructions (shown on the front cover as 50-year smoothed differences from the 1961–1990 normal) indicate that against the background of the millenium as a whole, the 20th century was unusually warm.”

    In both, the graph is described as being based in part on “instrumental records.”

    It seems obvious that the instrumental records were somehow grafted onto the proxy data (probably for all three lines), rather than being shown as a separate line in the graph.

    Is that hiding something?

    It surely is omitting any mention or illustration of the significant divergence between proxies and instruments in some part of the 20th century.

    The result is possibly misleading for anyone who isn’t aware of this divergence and the full extent of it.

    In such reports, swaying public opinion seems the paramount purpose; and glossing over the divergence serves that purpose.

  11. Antonio San permalink
    November 29, 2009 4:05 pm

    you must be aware of this new data page at

    Is this a smoke screen or serious?

  12. CBDenver permalink
    November 29, 2009 4:10 pm

    snip – nothing to do with thread

  13. Mac Lorry permalink
    November 29, 2009 4:24 pm

    I believe the key is for a sustained and focused call for transparency in the scientific underpinnings of AGW.

    snip – policy

  14. November 29, 2009 4:33 pm

    Not just hiding the decline, hiding the entire record. Not trying to high jack this post, but I completed a reasonable analysis of some of the CRU data graphs and, in my opinion, they show there as been little to no significant warming.

    Wish I had real data instead of eyeballing PDF charts. Comments welcomed at my site.

  15. Hoi Polloi permalink
    November 29, 2009 4:34 pm

    ““We are, and have always been, scrupulous in ensuring that our science publications are robust and honest.”

    In this case robusted…

  16. EdeF permalink
    November 29, 2009 4:43 pm

    Wow! Figure 2 above really shows some huge divergence between Briffa
    and either Mann or Jones for most of the 17th century and the later
    half of the 20th century. I didn’t realize the Briffa plunge was so
    dramatic, reminds me of the cliff divers at Acapulco. There is also
    a full 0.7 deg C divergence between Briffa and the instument data
    in the recent years. Note that the MWP is much closer to 1960-1991
    values than Briffa is to either Mann or Jones in the two time periods
    previously noted. With this level of uncertainty, not sure if one can
    positivly state that the present temperatures are significantly higher
    than the MWP.

  17. November 29, 2009 8:07 pm

    Perhaps the category should be -things that aren’t temperature.

  18. Norbert permalink
    November 29, 2009 8:34 pm

    Why does Figure 2 refer to “Briffa 2000″, when UEA refers to Briffa 1999 ?

    Steve: There wasn’t any relevant Briffa 1999.

  19. November 29, 2009 8:45 pm

    I wrote a small utility for folks like me that are running these scripts using the R GUI in Windows. It simply fixes the directional quotes into the standard quotes so that you do not have to fix them all by hand (the issue is described in the R Faq For Windows). You can download the utility here:

    At the main blog, this isn’t a problem.

  20. Norbert permalink
    November 29, 2009 9:59 pm

    It seems difficult to find out which table, in which column, contains the un-hidden data. Which URL and column was used for the mangenta part?

  21. Norbert permalink
    November 29, 2009 10:36 pm

    Is it correct to say that the instrumental data is (more) authoritative, while the tree ring data attempts to be an approximation? If so, then then the un-hidden data (magenta) does in fact illustrate that the Briffa series does not correlate with the actual temperatures (and also not with the data series from Mann and Jones) after 1960. It start diverging from the “real” temperature values even before 1960.

    So it seems that the deleted data really isn’t valid.

    Another point is that you seem to accuse UEA of still not showing the previously hidden data. But they haven’t claimed to do so, in that statement. What they did was more important: they showed the graph with the tree ring data and the instrumental data separate, which allows one to see how the combine graph came about. UEA refer to the articles which discussed why the data of Briffa, after 1960, was not included in the graph. That’s probably what they meant when they said “hidden in plain site”. So I don’t really understand why you are so harsh with UEA, except for the one point that you wanted to see the data. But now that you found it, it apparently only confirms what UEA has said about that data in the first place.

    It seems that except for the Briffa data after 1960 (or maybe even after 1950), the tree ring data corresponds well to the instrumental data.

  22. Rev. Dr. E. Buzz Miller permalink
    November 29, 2009 11:37 pm

    I merely have an undergraduate degree from a Big Ten school, but if I pulled this kind of infantile stuff “Oh, I promise I’ll put ALL of the stuff in!” on my final project for my degree, I would have been in trouble…

    I am pretty sure I would have been in trouble, but if the academy is so polluted with this relative truthiness, perhaps I would have been graded higher for playing childish games, if I had the right political attitude.

    This is ridiculous.

  23. Bryan H. permalink
    November 30, 2009 12:26 am

    With the emails and data coming out all but proving what I’d always thought was the case I’ve started to do more research into all the analysis that you guys have been doing here at CA. What I’m curious is if anybody has taken the time to load the raw instrument data and run control and stability tests on them? If so do you have links to those articles/blog posts?

    So far I see a lot of correlation work being done between various proxies and instrument data but even on the surface the instrument data look remarkably boring (normal and in control). I’ve been tempted to download chunks of the instrument dataset into minitab myself and see what comes out.

  24. John A permalink
    November 30, 2009 12:43 am

    Someone please explain this one.

    From the CRU code file osborn-tree6/ , used to prepare a graph purported to be of Northern Hemisphere temperatures and reconstructions.

    ; Apply a VERY ARTIFICAL correction for decline!!
    valadj=[0.,0.,0.,0.,0.,-0.1,-0.25,-0.3,0.,- 0.1,0.3,0.8,1.2,1.7,2.5,2.6,2.6,$
    2.6,2.6,2.6]*0.75 ; fudge factor
    if n_elements(yrloc) ne n_elements(valadj) then message,’Oooops!’

    … It flattens a period of warm temperatures in the 1940s 1930s — see those negative coefficients? Then, later on, it applies a positive multiplier so you get a nice dramatic hockey stick at the end of the century.

    Tell me this isn’t blatent data manipulation.

  25. Norbert permalink
    November 30, 2009 1:11 am

    Steve points out in “New!! Data from the Decline” :

    “While the decline was shown in Briffa et al 1998 and Briffa 2000, it was not shown in the IPCC 2001 graph” [...]

    So apparently it was decided, in accordance with published articles, to exclude the post-960 data, yet that data was shown in the referenced literature by Briffa. What then, if I may ask, is the problem ? The data appears to have been deleted from certain files simply because it was decided not to use that (apparently invalid) data.

  26. Scott Gibson permalink
    November 30, 2009 1:25 am


    You don’t see any problem with deleting divergent data and splicing in something else that confirms your theory? And presenting it to the public as if it is all of one piece?

    There are two obvious problems:

    1) If the data diverges from temperature now, how do we know it didn’t in the past?

    2) Wouldn’t it be a little more honest to keep separate types of data separate, even if you show them on the same plot? For example, a temperature curve, plotted as a separate line from the dendro. Then someone just looking at the pictures (as for example, those watching Al Gore’s “An Inconvenient Truth”), would have a better chance of evaluating the plots.

  27. Patrik permalink
    November 30, 2009 1:46 am

    I’ve worked in the computer graphics industry for 20 years. Rule no 1: Never delete anything. You never know when You’re going to need it.

  28. J. Peden permalink
    November 30, 2009 2:12 am


    So it seems that the deleted data really isn’t valid.

    The problem is that the rest of the same-stock data is, therefore, not valid as a temperature proxy. It happens to track the instrumental data for a while, then it surely doesn’t, and that’s all we know. We have no idea what it does previous to instrumental data.

  29. Grant permalink
    November 30, 2009 2:57 am

    I think the decline, or more precisely the divergence, raises two important issues.
    If we accept that tree ring data correlates with temperature, then the divergence needs to be explained. The obvious first hypothesis is that either the trees are “deniers” and are conspiring to grow more slowly and so deliberately distorting the data, or alternatively, the instumental record is inflating recent temperatures. Alternatively there may be other unknown factors which render the whole method unhelpful.
    Secondly, as I understand it, the instrumental record also calibrates the earlier temperatures. The divergence gap then has the effect of depressing all earlier inferred temperatures.

  30. Norbert permalink
    November 30, 2009 3:38 am

    Scott, there surely is a problem with using only part of data, and that is apparently why that has been discussed in published articles. I wish all scientific articles were available for free.

    The graph with separate curves for tree ring data and instrumental data seems more interesting, so I’d partially agree with you. However, the combined graph is probably a better summary. I’m sure that in preparing the graphs, many other simplifications had to be made. They appear to be the culmination of a huge amount of research, rather than just a plot of simple data points (which is what they look like).

    J.Peden, I think it would be difficult to make a judgement about that without being familiar with the scientific literature, specifically the articles mentioned above. Their content would have to be part of this discussion. Surely everyone agrees that the situation is far from ideal. The topic has a large amount of implicit problems, but so far I don’t see the scandal.

  31. Djbiggs permalink
    November 30, 2009 4:08 am


    I would hope that the culmination of a huge amount of research, would be just that: a simple plot of data points, though ones which are both accurate and precise. As my research (or lack there of) could estimate global temperature to +- 30 degrees over most time scales, I wouldn’t be making policy based on them however.

    The scandal for myself (a scientist) is in part, that the large amount of implicit problems are not being addressed within the normal mechanisms of scientific review.

  32. Dave permalink
    November 30, 2009 6:15 am

    Norbert says: “It seems that except for the Briffa data after 1960 (or maybe even after 1950), the tree ring data corresponds well to the instrumental data.”

    IOW, trees are good thermometers…. except when they’re not.

  33. Geoff Sherrington permalink
    November 30, 2009 6:51 am

    2009 November 30 Bryan H. permalink

    ” … if anybody has taken the time to load the raw instrument data and run control and stability tests on them?”

    The problem is that we no longer know how to find the raw instrument data. CRU state that they threw out a lot. Some countries have adjusted the raw data and seem reluctant to release it, if they still have it. Some agencies like CRU and GISS have at times further adjusted the “adjusted” raw data.

    snip : overeditorializing

  34. Michael Smith permalink
    November 30, 2009 9:36 am

    Norbert wrote:

    The topic has a large amount of implicit problems, but so far I don’t see the scandal.

    The scandal is that these guys see nothing wrong with deliberately misleading the public by publishing a graph that purports to show how tree ring data reveals unprecedented 20th century warming — a graph which can only be created by using instrument data in place of the inconveniently “diverging” tree ring data.

    The fact that the truth about the tree ring data — the “divergence problem”– is discussed in the scientific literature is hardly an excuse for what Jones did. An accurate and honest treatment of the data in one venue is not a license for misrepresenting it in another.

  35. Scott Gibson permalink
    November 30, 2009 10:18 am


    The reason a combined graph doesn’t work because they are plotting two things that may not be related to each other. In order to plot them together, you have to assume they are all real temperature anomalies. In my experience, I need a lot better correlation before I would accept that assumption.

    I’ll give you an example from the financial world. Lets say you follow a gold-mining stock, and note that it tends to diverge with the price of gold in the later part of the graph. We both know you wouldn’t replace the value of the stock with the price of gold (anomalies). Obviously, there are other factors that affect the price of gold mining stocks besides the value of gold; for example cash flow issues, ore deposit depletion, etc.

  36. DaveJR permalink
    November 30, 2009 11:21 am

    Norbert wrote: “Scott, there surely is a problem with using only part of data, and that is apparently why that has been discussed in published articles. I wish all scientific articles were available for free.”

    It’s been discussed, yes, but what do you call discussion without a shred of evidence to support any of it? Wishful thinking? The fact of the matter is that dendroclimatologists simply do not know what causes divergence and therefore it is wrong to present these chronologies as authoritative, never mind attempt to cover up data that blows huge holes in the assumption that we can accurately determine past temperatures using them.

  37. Norbert permalink
    November 30, 2009 12:59 pm

    It sounds like neither of you have actually read those articles. Without further knowledge, I would generally assume that such a graph represents a summary of the presented research, rather than the proof. I have some experience with graphs in other areas of research, and a graph by itself can “never” be taken as proof: it is always necessary to scientifically examine the underlying assumptions and methodology.

    I read elsewhere (on that the IPCC report’s info for those graphs did mention that both tree ring data and instrumental data were used. At this point, your disagreement with the graph is a scientific disagreement with the underlying research, and would need to be discussed scientifically.

  38. November 30, 2009 1:06 pm

    John A and Mikajah,

    If I add those artificial corrections to northern hemisphere tree-ring data, we get this, which is what appeared for tree rings on the WHO report. It appears, therefore, that those artificial corrections were applied to the tree-ring data before some kind of smoothing and averaging to obtain the graph. Can that be true? That would be too brazen, wouldn’t it? Perhaps these corrections just a side-show as a way of determining the size of the divergence?

  39. November 30, 2009 1:10 pm

    Isn’t Mann 1999 the original “Hockey Stick” that was produced with inappropriate data (bristlecone pine rings) and code that generated a hockey stick even when fed Red Noise?

    If so, how can they possibly use this?

    Maybe I’m missing something obvious.

  40. Craig Loehle permalink
    November 30, 2009 1:15 pm

    Something that people are missing, is that to make the splice (even the overlay) of instrumental data valid, you have to assume a precise matchup of proxy and instrumental at some point. It seems they did this by eye to make the graph look best. There is no documentation I have seen that this was done in some justified manner.

  41. Scott Gibson permalink
    November 30, 2009 1:37 pm


    Do you accept the premise that the tree ring data as presented is directly equivalent to temperature anomalies? If so, that is where we disagree.

  42. Scott Gibson permalink
    November 30, 2009 1:43 pm


    Yes. If they had chosen a different point to “splice” the datasets, they would have had a very different graph. If prior to 1960 closer to where the divergence began, they would have had a graph that no one would have taken seriously. If much later, they don’t have the hockey stick.

  43. Rob Herron permalink
    November 30, 2009 2:38 pm

    It would be great if you could comment on the importance of the 100 year temperature data to current global warming claims by the IPCC. It would seem to have been the principal argument for global warming but as I haven’t read the IPCC reports in depth maybe there is lots of other support for this claim
    The Hadley/CRU generated claim for global warming can no longer be supported after the emails and their destruction of the base data, however many of us “sceptics” are not sure how critical their analysis was to the whole IPCC global warming argument.
    It would help all of us if you could comment on this, I haven’t seen any other blog that addresses this specifically.

  44. Mark H. permalink
    November 30, 2009 3:39 pm

    Hey, where’s the instrumented data line?

    This must be McIntyre’s Trick, don’t rub it in?

  45. Norbert permalink
    November 30, 2009 7:30 pm

    Scott, just a quick reply before I get home: Personally, I take tree ring data as an indication rather than as proof. The graph also shows that the three data series differ even amongst each other. However, climate research has been dealing with this question for a very long time. The recent update to the IPCC report acknowledges that there is less than 100% scientific certainty by saying that waiting for scientific certainty could mean that we miss crucial tipping points without even knowing it. (Of course the latter statement isn’t just related to the use of tree ring data).

  46. Mark_DJ permalink
    November 30, 2009 8:45 pm

    After digging out a true gem of an email from the CRU stash, in which Michael Mann stated in no uncertain terms that the temperature plots from the GRIP and Dye 3 icecores showed “significant differences” for the MWP (despite the relevant graphs in reality being virtually identical), I’ve had to come to the conclusion that his and UEA’s little clique are so thoroughly self-deluded on the whole subject of the millenial temperature record that they just can’t help themselves any longer. They’ll do whatever it takes to keep on believing their “20th century is the hottest ever” mantra, and have the rest of the world believe it too, no matter how unscientific or generally dodgy that course of action might need to be. Hacking off a few decades of data from a series here or there so as to make a graph look more convincing would have to seem trivial, to anyone of that mindset.

  47. Ian permalink
    November 30, 2009 10:44 pm

    Kind of on topic (kind of not…): from Warwick Hughes site:

    CRU quietly revising 8 months of 2009 global temperature data
    November 29th, 2009 by Warwick Hughes
    njcons has just told me CRU have just “deleted” – removed 8 months of their CRUT3 monthly anomalies for 2009. Unannounced that I have seen. I checked at about 6am Canberra time on 30 Nov.
    Sure enough downloading the CRUTEM3 Global mean I see only Jan and Feb for 2009;
    2009 0.561 0.713 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.637
    2009 27 26 0 0 0 0 0 0 0 0 0 0
    HadCRUT3 is the same, 8 months edited out – as is HadSST2 – so the issue is through land and sea data.
    However, nobody has told the Hadley Centre…..

  48. Ian permalink
    November 30, 2009 10:58 pm

    No – sorry. Comment on the site indicated the data was still available on the MetOffice site. Speculating that it is the “Emergency Server” issue at the CRU.

    1.Bob Koss Says:
    November 29th, 2009 at 5:35 pm
    The gridded anomalies thru October are still available here.

  49. Jan permalink
    December 1, 2009 12:20 am

    I wish I knew more science and I curse my lazy teenage high school brain for my current state of ignorance. I’ve been plodding through this stuff for over a week now and am only just beginning to get past the acronyms and the terms you all use with such ease.

    One thing I do know is people and the emails told me enough about the made-men of climate science to start digging. People confident in their expertise do not hide their light, or their data/code, so to speak, under a bushel. They willingly share it, almost to the point of obnoxiousness, as a shining beacon for others to follow.

    In that spirit I thank you and others for shedding some light for me on what is a difficult subject. I am, indeed, learning.

  50. Norbert permalink
    December 1, 2009 3:03 am

    Question: At which theoretical point of a future instrumental temperature development would you agree that there really is (or would be) a warming problem? (Aside from the question of what causes it.)

  51. DaveJR permalink
    December 1, 2009 3:45 am

    Norbert wrote: “At which theoretical point of a future instrumental temperature development would you agree that there really is (or would be) a warming problem? (Aside from the question of what causes it.)”

    Suppose all of current temperature trends are down to natural forces. Would you consider natural warming to represent “a problem” now, or possibly some time in the future?

  52. gober permalink
    December 1, 2009 4:08 am


    I have had difficulty running the code in R.

    I’m a relatively basic R user: not advanced enough to debug your code. However I do run the code you post from time to time. (Actually it was Climate Audit that got me onto R in the first place. Thank you for that: I have found it incredibly useful for other things.)

    When the code gets to the line:
    ” CRU=window(Briffa[,"CRU"],start=1850)”

    I get the error message:
    “Error in NextMethod(“[“) : subscript out of bounds”

    Is there some library I should have loaded? From my very inexperienced look, is “CRU” a header in the Briffa table? The only header that looks like that is “CRU99″. Does that mean the data has changed since you prepared the code?

    I’m sorry – this will be the least of your concerns, but since you’ve gone to the effort of posting the code, I wanted to ask.

    Steve: I’m sorry for the buggy code that I posted. I’ve obviously got slightly tweaked things in my console. I didn’t shut down my computer and run from scratch. I should have. At some point, I’ll shut down my computer and re-run without things in my console.

  53. Norbert permalink
    December 1, 2009 4:34 am

    DaveJR: Not sure about the ice melting we already have, I see it more as a future problem. Does it answer your question? Will mine be answered as well? Is there some consensus here around this blog, is Steve’s position on this question known?

  54. J. Peden permalink
    December 1, 2009 5:03 am


    J.Peden, I think it would be difficult to make a judgement about that [use of Briffa's tree ring data as Global average temp. proxies] without being familiar with the scientific literature, specifically the articles mentioned above.

    No, it’s cut and dried just the way it is: Briffa’s tree rings simply cannot be used as temperature proxies to reconstruct pre-instrumental Global temperature averages – because they haven’t even been able to do it in the relatively short span of instrumental records, unless maybe the instrumental temps. are wrong and they should have been following the tree rings downward when the divergence occured, which is actually a real possibility the way things have been looking now for some years in regard to the methods of Climate “science”.

    Note that the scientists who are presumeably familiar with the literature have no solution to the “divergence” problem themselves, other than to try to hide it.


  55. DaveJR permalink
    December 1, 2009 6:52 am

    Norbert wrote: “Not sure about the ice melting we already have, I see it more as a future problem. Does it answer your question?”
    Not exactly. If climate change was entirely natural, how far would you want to go to prevent it, as opposed to adapt to it?
    How warm is too warm depends, IMO, on a) how much is man-made compared to natural (what changes need to be made to produce x degC change in temperature?), and b) how far we’re prepared to go to artificially alter climate if natural forces are the most dominant.
    My personal opinion is that the effects of CO2 have been vastly overstated compared to other factors (a sum of natural and other man-made sources) because it’s politically convenient just to focus on CO2 and it’s politically convenient for some scientists to maximise this view. Therefore, it is unclear to me whether massive efforts to reduce CO2 would a) have any discernable effect while b) impacting on the quality of life, and potentially technological advancement, of everyone.
    I much prefer the thorough approach by Pielke Sr and can sympathise with his exasperation at the IPCC for taking such a politically convenient view of climate change. Only by identifying the various effects can we best determine what policy or combination of policies to take. I’m afraid I don’t subscribe to the idea that making huge efforts to reduce CO2 (beyond increasing efficiency) is a good thing regardles of whether the predictions of CO2 on climate are right or wrong ie I don’t believe the ends can only justify the means. This is a view opposed by at least one of the Climategate correspondents.

  56. PaulM permalink
    December 1, 2009 7:50 am

    Norbert, Steve is not renowned for speculating about what might happen re future temperatures. He also discourages general global warming discussions – please try to stick to the specific topic. Splicing the two graphs together is misleading, as the two things are quite different, and it should never be done. Even Mann agrees with that.

    snip – pls don’t go a bridge too far

  57. December 1, 2009 8:46 am

    Glad my last URL-laden post didn’t make it through moderation. I’ve now done a much better piece, “Hiding the Decline” (click), for newcomers as well as regulars; it takes cognizance of Steve’s work here, and zooms in to enlarge the crucial splice details of the two UEA charts, their origins and their differences.

    Please tell me if I’ve got any of it wrong. I’ve worked hard to get it out as I think clear visuals are important to show enough people EXACTLY what “hide the decline” is about, and how strange stuff is still continuing, but my science may not be perfect.

  58. December 1, 2009 9:29 am

    Is it possible that I am only one who is finding it difficult to navigate this mirror site? In fact, I have to admit to being lost, which is unfortunate and more than a bit frustrating at this exciting time. OK, I admit to being a dummy when it comes to understanding the structure of sites, but I thought that CA had been quite easy to cope with.
    Is there a primer on site navigation readily available?


    (P.S. WUWT gives me much greater problems :-(( Oh dear!)

    Steve: This was thrown up quickly because CA crashed. Stop complaining.

  59. Jeffrey Davis permalink
    December 1, 2009 11:17 am

    What is the significance of a modern proxy study? Why wouldn’t you just look at the instrumentation?

    Calibration of the validity of the “proxy”

  60. December 1, 2009 12:30 pm

    Very nice work Lucy and you are right – crisp, clear, correct graphics are crucial in explaining the issues.

  61. Shallow Climate permalink
    December 1, 2009 1:26 pm

    Re Lucy Skywalker (above): You asked for feedback in case you got any of it wrong. (By the way, I like your site and applaud your efforts–thank you!) You mention about the two black lines (summer and winter temperatures) plotted separately, and that the two crisscross (which is crazy). Actually, it is the temperature ANOMALIES that are being plotted, not the actual temperatures (note y-axis label), so there is, prima facie, nothing crazy about the two black lines crisscrossing. Thanks again for a great visual.

  62. December 1, 2009 2:43 pm

    Re my posting of yesterday.

    I wasn’t complaining – far from it, I love what’s going on here – but I was getting lost. I understand why, now that you explain that the mirror is a quick solution to a difficult technical problem.

    Thanks, Robin

  63. Norbert permalink
    December 1, 2009 3:21 pm

    PaulM, my question wasn’t aimed at a general discussion, and only partly at the future. The basic intention of the question is to find out what the criteria fro establishing warming would be. To phrase this question in terms of present data: When interpreting the available data (either both tree data etc and instrumental records, or only instrumental records) in a way that would be considered “correct” here on this site, then how far away is that data (or the part of it which is considered valid) of indicating a global warming period? This would give me a better understanding of what the data, according to this site, is really showing. Or perhaps I should phrase it like this:

    How wrong is this graph considered to be, how far away from what it should be? And more importantly: how far away would a corrected graph be from showing global warming? Would have to remove tree data completely? And what would it be like then? Or is this thread, and this site, only about finding errors?

    BTW, as I said above, I personally also prefer a graph with separates lines, but I find the combined graph acceptable as a simplification for a summary. Certainly I wouldn’t consider it a scandal. For a graph which allows a more critical examination of the underlying data, there are many other available graphs, for example by Mann, which also show more data series than just these three.

  64. hengav permalink
    December 1, 2009 4:03 pm

    Craig Loehle says:
    “There is no documentation I have seen that this was done in some justified manner.”

    In the context of the thread’s graph, ending in 2000, smoothed and referenced to the 1960′s, remember that 1998 was the largest spike in the global instrument record. It was caused by an El Nino and totally natural. Is there any way of adding the modern (to 2009) portion of either your reconstructions, or simply the land/sat records to show where the graph goes after 2000?

    Justified manner… If 1998 was not a GHG/man made phenomenon then you would be justified in removing it as a choice as to where to splice the temperature record to the reconstruction. I would love to read your thoughts about THAT choice.

  65. December 1, 2009 5:28 pm

    Thanks a lot for the code. As there are a few bugs in it and the debugging can be quite time consuming I want to share my edited R-file (all data files are included, sorry for the messy plot part – I’m an absolute R-beginner).

  66. December 1, 2009 7:02 pm

    Shallow Climate, thanks for that correction. Actually, far from resolving anything it deepens the mystery. I’ve altered my text accordingly. Also put in a shot from the UEA page.

  67. December 1, 2009 9:29 pm

    Folks, in all this focusing on the search for correlations per se, don’t forget basic theory: CO2 absorbs IR, and so must be a warming stimulus. Sure, we can argue over how much effect, what other things complicate it, but that’s that. As for the misconduct: first, it was a clique and not all AGW supporters all over, and second: you’ve heard of prosecutors/police “framing a guilty man”? They worry if they can get a conviction, so they “improve” the evidence. That doesn’t make the man innocent. Note, it isn’t likely old Svante Arrhenius was being sneaky way back in 1896.

    Steve: I do not want people to debate the “big picture” in one paragraph bites. Every thread then looks the same.

  68. Shallow Climate permalink
    December 1, 2009 9:29 pm

    Re Lucy Skywalker (just above): You respond to me with more grace than I had true humility in responding to you. I find the points you raise (at your linked site) to be fascinating, and I appreciate your thoroughness and devotion. Again, thanks for what it is, IMO, a great site.

  69. Scott Gibson permalink
    December 1, 2009 11:34 pm


    I am a geologist, so I naturally have a different view than most. The Earth has seen much warmer periods than the present, and much colder. Sea levels have been much higher than the present, and much lower. Species have come and gone, with no concern to how they felt about it. The Earth has had no problem surviving these changes, and I’m sure that man can adapt to any climate shifts that occur.

    With reference to what kind of graphs and changes would convince me; I don’t know. I do know I have no confidence in much of what I’ve seen, and I could write a book on why. Suffice it to say that tree rings, varves, and every other type of proxy data we have, seem to depend on factors other than temperature. Can someone convince me that a certain dendro line (whether from ring width, wood density, isotopic fractionation measures, etc) correlates to temperature? Of course. However, I haven’t seen one yet, though I could be convinced if I saw the methodology behind the data and I trusted the sources. Another problem is that even our instrument temperature data is questionable, and it doesn’t help that the slopes of regression lines are reported to the hundredth of a degree, when the original data is probably +/- a degree.

    As a skeptic, I am willing to look at all points of view. However, after all my experience viewing hundreds of lines of evidence, I don’t see anything unusual in today’s climate. We are much closer to the coldest period in Earth’s history than the warmest, even if you only include the Phanerozoic. We are barely out of glaciation, after all!

  70. Martin Hoppe permalink
    December 2, 2009 12:54 am

    I would agree, merging the thermometer temperatures on the same graphwith the proxies is not the correct thing to do. Actually, it is =- snip – in this case to do so. As the thermometer temps diverge severely from the proxies post 1960, there obviously is something unaccounted for in the temperature data. The three proxies tend to agree with each other, it is the thermometer data that is the odd man out. Best to look for reasons why the thermometer data may be out of whack. (heat islands or changes in way temperatures were taken as time progessed or ??) Since there is something quite amiss between the proxies and the thermometers post 1960, to replace proxy data with thermometer data is closer to =snip – than good science.

  71. Norbert permalink
    December 2, 2009 2:30 am

    Scott, which time scales are you referring to? Very long ones, like Ice Age and Dinosaur age?

  72. Hugo M permalink
    December 2, 2009 10:28 am

    Steve, as a test for an R installation, I tried to run your script after having fixed some WordPress related issues (quotation marks and so on). But it seems as if you posted a wrong version. I fixed the obvious “CRU” vs “CRU99″ issue, but R also halts when it encountered the gated portion of britta:

    > X=ts.union(MBH=MBH99+delta,J2001,briffa=briffa[,"gate"],CRU=cru ) #collate
    Error in is.vector(X) : object ‘briffa’ not found

    Obviously, this could also due to an outdated R version, since it didn’t say it couldn’t find jones :-)

    Steve: I closed down my R-session and will re-run the script and tidy. My abject apologies for not doing so before.

  73. Scott Gibson permalink
    December 2, 2009 10:45 am


    Both long and short, geologically speaking.

    The geologic record shows quite clearly that something cooled the Earth during the Pleistocene, causing the ice ages. Prior to that, we didn’t see these repeated large scale continental glaciations. In fact, Greenland is still covered with ice, and there is evidence it had less ice just 700 (?) years ago, which may or may not indicate cooling since then.

    Geologists generally regard the current time as an interglacial, and the sense is that we will see another glaciation soon (geologically speaking), assuming the trends continue.

    As for the Mesozoic (dinosaur age), rocks state quite clearly that sea levels were generally much higher than currently, and it is likely that we had no or little icecaps at the poles for much of this time. Given that, it is probable that temperatures were almost always warmer than currently.

  74. Hugo M permalink
    December 2, 2009 10:51 am

    Well, debugging your script becomes quite amusing. Replacing the small “briffa” by a great “Briffa”, I again get a sybilline comment:

    > X=ts.union(MBH=MBH99+delta,J2001,briffa=Briffa[,"gate"],CRU=cru ) #collate
    Error in NextMethod(“[“) : subscript out of bounds

  75. December 2, 2009 1:38 pm

    @Hugo M: I think this should be just “briffa=gate” (gate from line 11). I think I managed to get all bugs out.

  76. stevemcintyre permalink*
    December 2, 2009 1:51 pm

    Steve: I re-ran with a fresh console and fixed the two small bugs in the online script noted above. Scripts is now clean. The edits are not relevant to the output (as I’d dealt with them on the console) but I try to make these turnkey and apologize for not checking it on a fresh console before placing it online. Too hasty.

  77. Norbert permalink
    December 2, 2009 2:06 pm

    Scott, I didn’t have the impression that the question is whether planet earth would survive. The question is whether we would want any drastic changes. But this leads off-topic. Perhaps there is some other thread where we can discuss this?

  78. Norbert permalink
    December 2, 2009 2:20 pm

    snip – editorially there is no purpose in one paragraph discussion of the big picture without working off an article. Hold this debate for now.

  79. James permalink
    December 2, 2009 3:58 pm

    (off topic)

    Norbert (please excuse me for jumping in, Scott) – I’m also a geologist and a skeptic. As for your queries concerning whether we would want drastic changes, this is irrelevant. The earth has always been (and will continue to be, at least until the Sun eats it) in a state of flux. Our planet has run the gamut from molten planetoid (according to currently accepted theory) to frozen ice cube (I hear the Wooly Mammoths loved it). What you’re asking is about human preference — and that’s the crux of the climate change argument. Geo-engineering aside (as I think it’s a horrible idea), human beings are certainly not in the driver’s seat regarding the planet’s thermostat.

    If you truly wish to know more about the particulars of the debate, may I humbly suggest searching Bishop Hill’s Blog (

    As others have stated, ClimateAudit is concerned with the statistical analysis (and validity) of climatological studies.

  80. Scott Gibson permalink
    December 2, 2009 7:10 pm


    Yes we are off topic, and thanks to Steve for allowing that. I’ll just answer your last question, because I have to run.

    As I said, I am a geologist although I have worked with tree ring geochemistry, so my area of expertise is in geology not meteorology. The point I’m trying to make is, that regardless of whether the global temperature is going up or down at this moment, we are much closer to another ice age than to catasrophic warming. I don’t believe the data we have proves or disproves global warming. I don’t know whether it is warming now or not.

  81. gober permalink
    December 2, 2009 7:16 pm

    Hi Steve,

    I got all the way through to the last two lines, at which point “usr 1960″ generated the message:

    ‘Error: unexpected numeric constant in “usr 1960″‘

    and the “points( c(time(smoothb))…….” line generated:

    ‘Error in points(c(time(smoothb))[temp], smoothb[temp, "briffa"], pch = 19, :
    object ‘temp’ not found’

    Also, from comparing your code with Michael’s (above), it looks like yours is plotting only one out of four time series: only one “plot( c(time(smoothb))…” line.

    Again, I’m sure this is the least of your worries, but I assume you wouldn’t have bothered to post it if you didn’t want it to be entirely reproducible.

  82. Mike Hollinshead permalink
    December 3, 2009 7:55 am

    re: AJ Strat raw CRU data analysis

    AJ Strat provides a CRU map showing the range of variation in measurement.

    It is significant that the widest variation is seen in Northern high latitudes where claimed global warming has been most significant. It is all in the >4 degree C category.

  83. airfoilmod permalink
    December 3, 2009 8:58 pm

    Delete if too vanilla/plebeian. My Weather/Climate memory has retreating Ice at Poles a result of not only temp. but Drought, which can be caused by declining GMT. Less energy at Equator to evaporate water, weaker winds to distribute said vapor, less water transported to Poles to fall as precip, to then augment ice after seasonal melt, etc. What memory fails is the relationship and scale, GMT/precip/and related trends. Anyone? I’m calling cooling, but my degree is old.

  84. nanny_govt_sucks permalink
    December 5, 2009 1:18 am

    Did anyone catch this from Dr. Mann:,0,3182494.story

    “Mann said last week the ”trick” was not a deception, but a clever way to combine the best data from the last 1,000 years until 1960 with actual temperature readings from 1961 to 1999. He says ”decline” refers to the decline in the reliability of tree rings to reflect temperatures after 1960.”

    … and how is that “reliability” assessed? By comparing to temperatures, right?

  85. Eben permalink
    December 5, 2009 9:39 am

    Seeking an interview with you Mr. McIntyre!!?? I am a producer/host on KPFK 90.7fm in Los Angeles and have always questioned the science behind human induced global warming.

    Please email me at your earliest convenience!! And thank you so much for all that you do!

    Eben Rey
    KPFK 90.7fm


  1. Climategate Round-Up #4 | GORE LIED
  2. Playing hide and seek behind the trees « Watts Up With That?
  3. Climategate: Der Trick mit dem Hockey Stick « So Halt
  4. Climategate Round-Up #4 « The Daily Bayonet

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

%d bloggers like this: