Hide the Decline: Sciencemag

Science published one of the first spaghetti graphs (in Briffa and Osborn 1999 here) as part of an invited comment on the Mann et al 1000-year reconstruction, then hot off the press with its supposed proof that 1998 was the “warmest year” of the millennium. Jones et al 1999, discussed recently here, contained a different spaghetti graph.

Referring to this figure, Briffa and Osborn stated that none of the reconstructions covering the MWP reach modern warmth, and thus the MBH conclusion must “surely be accepted”.

The temperature histories that extend through the medieval period do indicate general warmth (see the figure), although with different maxima (in the 9th, 10th, and 11th centuries). Clearly none of these reach the levels of warmth seen today [although the confidence ranges (not shown here) approach them]. On the basis of their analysis, Mann et al. conclude that the 20th century is anomalously warm. Even with the very limited data available and the problems associated with interpreting many of them as unambiguous measures of hemispheric temperature change, this conclusion must surely be accepted.

The figure from Briffa and Osborn 1999 is shown below with the original caption. It stated that the Briffa version came from Briffa et al 1998 (Nature) and Briffa et al 1998 (Pr Roy Soc London), “processed to retain low-frequency signals”.


Original Caption: Records of past climate… Comparison of NH temperature reconstructions, all recalibrated with linear regression against the 1881-1960 mean April-September instrumental temperatures averaged over land areas north of 20ºN. All series have been smoothed with a 50-year Gaussian-weighted filter and are anomalies from the 1961-90 mean. Instrumental temperatures (1871-1997) are in black, circum-Arctic temperature proxies [1600-1990, from (2 - Overpeck)] are in yellow, northern NH tree-ring densities [1550-1960, from (3 - Briffa et al 1998(Nature); Briffa et al 1998 (Proc Roy Soc London)), processed to retain low-frequency signals] are in pale blue, NH temperature proxies [1000-1992, from (4 - Jones et al 1998)] are in red, global climate proxies [1000-1980, from (5, 6 - MBH99)] are in purple, and an average of three northern Eurasian tree-ring width chronologies [1-1993, from (10 - Briffa et al 2000)] is in green. Although representing a much more restricted spatial coverage than the other series, the last of these (also processed to maintain low-frequency climate information) is included here because of its extended length and because it suggests relatively cooler summer temperatures (at least across northern Eurasia) before A.D. 1000.

The Briffa and Osborn 1999 version of the Briffa MXD reconstruction doesn’t match the version of Briffa et al 1998 or the subsequent version of Briffa et al 2001, both of which were archived. Oddly enough, it does match (after truncation) a version archived at NCDC in December 1998 in connection with Jones et al 1998 (though not used in that article), where it occurs in the second sheet of an Excel file here. To my knowledge, this particular version of the Briffa reconstruction was not otherwise published. (The Briffa reconstruction seems to have been very fluid in this period, as the versions in IPCC TAR Zero Order Draft and First Order Draft appear to be different again and still unaccounted for.)

The following figure overlays a 50-year Gaussian smooth of the digital version at the Jones et al 1998 archive (mean padding after truncation) with approximate rescaling to match the graphic on an excerpt of the figure in Science, proving that the Science figure derives from this version.


Figure 2. Emulation of Briffa and Osborn 1999 Figure 1. The Briffa version is in light blue and is overprinted (thin black) with the Briffa version in the NCDC Jones et al 1998 archive, with 50-year gaussian smooth after truncation to 1960.

As previously reported, this figure, together with Jones et al 1999, are the first two bites of the poison apple of hide the decline. This is what the figure would have looked like, had all the data been shown.

Figure 3. Briffa and Osborn 1999 Figure 1 Excerpt showing the decline (in red).

One can reasonably wonder whether the key conclusion of Briffa and Osborn 1999 – “[despite] the problems associated with interpreting many of them as unambiguous measures of hemispheric temperature change, this conclusion [MBH] must surely be accepted” – would have stood up if the decline had been shown.

As opposed to the other possible conclusion: the “problems associated with interpreting many of them as unambiguous measures of hemispheric temperature change” remain an unsurmounted obstacle and the reason why the Mann reconstruction goes up so sharply when the Briffa reconstruction based on a very large population of temperature sensitive sites goes down remains unexplained and a critical problem within the field.

Briffa and Osborn 1999 contains a very sly reference to the divergence problem:

A number of tree-ring chronologies have displayed anomalous growth or changed responses to climate forcing on different time scales in very recent decades (3 – Briffa et al 1998 (Nature), 9 – Jacoby and D’Arrigo 1995). Understanding the reasons for these changes is important for understanding the causes and limits on past tree growth. Paradoxically, therefore, more work in the recent period is required to better interpret the early proxies. Few of the proxy series run up to the present, however, and updating these will involve considerable effort.

Climategate scientists were well aware of the importance of figures. Briffa and Osborn knew that the graphic with the deletion of the decline would leave a different impression than one that disclosed the decline. The sly wording of the running text compounds the problem. Yes, there are proxies that need updating, but the MXD data used in the Briffa reconstruction came right up to the early 1990s. Unavailability of data is not the reason why the Briffa reconstruction ends in 1960.

Briffa and Osborn 1999 contain a number of sensible caveats about the Mann and other reconstructions, raising caution about the role of bristlecones ( the “amplitude series relating to the first principal component of a group of high-elevation tree-ring chronologies in the western United States”) in the Mann reconstruction and readers are referred to the original here.

These sensible caveats occasioned a flurry of Climategate correspondence among the Team in April and May 1999 ( see 98. 0924120405.txt; 99. 0924532891.txt; 100. 0924613924.txt; 105. 0925829267.txt; 106. 0926010576.txt; 107. 0926012905.txt; 108. 0926026654.txt; 109. 0926031061.txt; 111. 0926681134.txt ) in which Mann objected vociferously to even these reasonable caveats. Even Bradley was nonplussed by Mann’s conduct. In 99. 0924532891.txt on Apr 19, 1999, entitled CENSORED!!!!!, Bradley observed:

As for thinking that it is “Better that nothing appear, than something unnacceptable to us” …..as though we are the gatekeepers of all that is acceptable in the world of paleoclimatology seems amazingly arrogant.

A few weeks later, Bradley commented on Mann’s effort to smooth the waters: “Excuse me while I puke…”

Needless to say, a few years later, when our criticisms appeared, Bradley adopted the attitude that he criticized here: Better that nothing appear, than something unacceptable to us.

53 Comments

  1. Steve McIntyre
    Posted Mar 17, 2011 at 11:11 AM | Permalink | Reply

    #Briffa Osborn 1999
    [text]
    source(“http://www.climateaudit.info/scripts/utilities.txt”)
    f=function(x) filter.combine.pad (x, truncated.gauss.weights(50)) [,2]

    ##Science image
    #download gif
    library(caTools)
    download.file(“http://www.climateaudit.info/images/briffa/briffa99-science.gif”,”temp.gif”,mode=”wb”)
    gif <- read.gif(“temp.gif”, verbose = TRUE, flip = TRUE)
    detach(package:caTools)

    #save as reshaped
    par(mar=c(0,0,0,0))
    image(gif$image, col = gif$col, main = gif$comment, asp = 1,axes=FALSE)

    library(GDD)
    # GDD(file=”d:/climate/images/2011/climategate/briffa99-science_reshaped.gif”,type=”gif”,w=480,h=480)
    par(mar=c(0,0,0,0))
    image(gif$image, col = gif$col, main = gif$comment, asp = 1,axes=FALSE)
    # dev.off()
    detach(package:GDD)

    ###download data at NOAA

    library(xlsReadWrite)
    download.file(“ftp://ftp.ncdc.noaa.gov/pub/data/paleo/contributions_by_author/jones1998/jonesdata.xls”,”temp.xls”,mode=”wb”)
    A=read.xls(“temp.xls”,sheet=2)
    A=ts(A,start=A[1,1])
    detach(package:xlsReadWrite)

    ##download cropped version
    # cropped in GIMP and saved as png; could have been saved as gif

    library(png)
    download.file(“http://www.climateaudit.info/images/briffa/briffa99-science_cropped.png”,”temp.png”,mode=”wb”)
    img=readPNG(“temp.png”); dim(img) # 158 104 3
    year=1402:1994

    ## Replicate Briffa version in Science image
    par(mar=c(0,0,0,0))
    plot(year,col=”grey80″,xlim=c(1400,2000),ylim=c(-.5,.25),type=”n”,axes=FALSE)
    rasterImage(img , 1400, -.6, 2000, .3)
    year=1400:1994
    y1= f(window(A[,4],start=1550,end=1960))
    y2= f(window(A[,4],start=1550))
    lines( 1.07*(1550:1960)-160, .8* y1-.17,col=1)
    #contrast to decline
    lines( 1.07*(1550:1960)-160, .8* y1-.17,col=1,lwd=3)
    points( 1.07*1960-160, .8*y1[1960-1549] -.17,col=1,pch=19)
    points( 1.07*1994-160, .8*y2[1994-1549] -.17,col=2,pch=19)
    lines( 1.07*(1550:1994)-160, .8*y2 -.17,col=2,lwd=4)

    #save
    # png(“d:/climate/images/2011/climategate/briffa99-science_emulated.png”,h=480,w=480)
    par(mar=c(0,0,0,0))
    plot(year,col=”grey80″,xlim=c(1400,2000),ylim=c(-.5,.25),type=”n”,axes=FALSE)
    rasterImage(img, 1400, -.6, 2000, .3)
    lines( 1.07*(1550:1960)-160, .8* y1-.17,col=1)
    # dev.off()

    ## Variation without trick to hide the decline

    # png(“d:/climate/images/2011/climategate/briffa99-science_notrick.png”,h=480,w=480)
    par(mar=c(0,0,0,0))
    plot(year,col=”grey80″,xlim=c(1400,2000),ylim=c(-.5,.25),type=”n”,axes=FALSE)
    rasterImage(img , 1400, -.6, 2000, .3)
    year=1400:1994
    y1= f(window(A[,4],start=1550,end=1960))
    y2= f(window(A[,4],start=1550))
    lines( 1.07*(1550:1960)-160, .8* y1-.17,col=1)
    #contrast to decline
    lines( 1.07*(1550:1994)-160, .8*y2 -.17,col=2,lwd=4)
    points( 1.07*1960-160, .8*y1[1960-1549] -.17,col=1,pch=19)
    points( 1.07*1994-160, .8*y2[1994-1549]-.17,col=2,pch=19)
    # dev.off()

    [/text]

  2. RickA
    Posted Mar 17, 2011 at 12:07 PM | Permalink | Reply

    Steve:

    Thank you for research this – this is very interesting.

    At least the caption on this graph stated “Instrumental temperatures (1871-1997) are in black”. So while they “hide the decline” by not showing the post 1960 data diving, at least they didn’t try to graft the instrument record onto the proxy record, without disclosure.

    So this version is less “hidden” than the later versions of the graph.

    • Posted Mar 17, 2011 at 12:58 PM | Permalink | Reply

      Yes, but the graph is still trying to hide things. As well as the truncation of the Briffa 1998 data, it uses the standard climate science trick of drawing the instrumental data with an unnecessarily thick line on top of the proxies, to hide the fact that it does not fit well with the proxy lines underneath. No proper scientist would present data in this way. Also they use a small low-resolution gif.

      • DEEBEE
        Posted Mar 17, 2011 at 7:33 PM | Permalink | Reply

        Yes, but yes it was onlymeant for public consumption — you know the ones with not much there there.

  3. Charlie A
    Posted Mar 17, 2011 at 1:26 PM | Permalink | Reply

    My dumb question for the day:

    There is a lot of discussions about Briffa’s proxy data.

    Are there no other sets of modern tree ring data that?
    Do they show the same sort of divergence?

    ————–

    Actually, one more dumb question …..

    Is there anywhere a basic study that shows that tree ring density or tree ring widths have a high correlation with temperature? How about with precipitation?

  4. BillyBob
    Posted Mar 17, 2011 at 1:31 PM | Permalink | Reply

    “an average of three northern Eurasian tree-ring width chronologies [1-1993, from (10 - Briffa et al 2000)] is in green. Although representing a much more restricted spatial coverage than the other series, the last of these (also processed to maintain low-frequency climate information) is included here because of its extended length and because it suggests relatively cooler summer temperatures (at least across northern Eurasia) before A.D. 1000. ”

    And doesn’t it ALSO suggest relatively cooler temperatures AFTER AD 1000 as well?

    It almost is a mirror image with 1000AD as the center point.

  5. Craig Loehle
    Posted Mar 17, 2011 at 1:33 PM | Permalink | Reply

    Briffa states: “The temperature histories that extend through the medieval period do indicate general warmth (see the figure), although with different maxima (in the 9th, 10th, and 11th centuries)”

    but when there is any dating error (which even tree rings can have when there are missing rings or cross-dated trees) or sampling error it is expected that different reconstructions will have peaks that don’t line up perfectly (Loehle, C. 2005. Estimating Climatic Timeseries from Multi-Site Data Afflicted with Dating Error. Mathematical Geology 37:127-140)
    also, note that it is likely that the peak in the MWP was probably more like 900AD which only the green line covers, and many similar graphics combine recons that don’t go back that far and then claim to refute the MWP as being warm.

  6. Nik
    Posted Mar 17, 2011 at 1:36 PM | Permalink | Reply

    The little biology I learned seems to indicate, and Briffa apparently agrees in his comments, that the major question in all this is why do the tree rings change in late 20th century.

    One would have thought that truly inquiring minds would have grasped on that anomaly, higlighted it and then figured it out. Instead we had “hide the decline” reflex actions.

    Nik

  7. Posted Mar 17, 2011 at 1:45 PM | Permalink | Reply

    So much for the Standard Team Defense for “Hide the Decline”: “We only did it once, and anyway it was by special request, for a dumbed-down cover for an obscure report.”
    — my slightly-sarcastic version of the RealClimate line re the now-notorious email.

    Hard to argue that Science and Nature are obscure publications. Perhaps we’ll see a more-competent defense than was mounted in the previous thread?

  8. Paul Penrose
    Posted Mar 17, 2011 at 2:25 PM | Permalink | Reply

    Peter,
    Their backup excuse has always been “the post-1960 data is an outlier and obviously wrong, so we deleted it.” If you complain that it’s not obvious to you that it’s wrong, then they just sniff and say it’s not their job to educate you on such simple matters. This is all inexcusable elitist BS of course, but completely expected now.

  9. Hu McCulloch
    Posted Mar 17, 2011 at 2:52 PM | Permalink | Reply

    If Scimag has any scientific integrity, it will now be retracting this piece for deceptive graphics.

    A further problem is that the accompanying table of correlations with temperature is based only on the period 1881-1960, even though the temperature series runs back to 1871 and the proxies all run past 1960 (including Briffa MXD when its decline is not hidden). The correlations would be much weaker if 1871-1880 were included, moderately weaker if the post-1960 portions of the graphed series were included, and much weaker if the concealed post-1960 portion of the Briffa MXD series were included. However, since Briffa and Osborne do state in the * note that the correlation period is restricted, the reported correlations are arguably merely misleading, rather than deceptive.

    The table of cross-correlations between proxies is deceptive, however, since the dagger note says that they cover the “maximum overlap period” between the proxies, when in fact they are (presumably) based on only the maximum overlap between the unconcealed portion of the series.

  10. Chris S
    Posted Mar 17, 2011 at 2:56 PM | Permalink | Reply

    “I trust that history will give us all proper credit for what we’re doing here”.

    Mann may yet get his wish.

  11. Posted Mar 17, 2011 at 3:04 PM | Permalink | Reply

    “an average of three northern Eurasian tree-ring width chronologies [1-1993, from (10 - Briffa et al 2000)] is in green. Although representing a much more restricted spatial coverage than the other series, the last of these (also processed to maintain low-frequency climate information) is included here because of its extended length and because it suggests relatively cooler summer temperatures (at least across northern Eurasia) before A.D. 1000. ”

    Of course they mean low frequency tree growth information. I’m not convinced that they are the same.

    • GTFrank
      Posted Mar 17, 2011 at 5:21 PM | Permalink | Reply

      Yes, and who’s to say that the “cooler” period before AD1000 was not actually as warm as since 1960 and is just “bad data” as the post 1960 data is claimed to be. I don’t think they addressed this possibility.

  12. Jean S
    Posted Mar 17, 2011 at 3:04 PM | Permalink | Reply

    I think it is worth pointing out that “Mike’s Nature trick” was not used here (MBH9X pointing slightly downwards) nor in Jones et al 1999. So I think it is safe to assume that Jones (“CRU group” in general) was informed of the “trick” some time during the summer/autumn 1999.

  13. andy
    Posted Mar 17, 2011 at 4:32 PM | Permalink | Reply

    I seem to remember Steve saying that a previous graphic he produced was one of the most disconcerting images he had ever seen on climateaudit.
    The graph with the red line must surely be a contender for the title?

  14. Ken Mival
    Posted Mar 17, 2011 at 4:48 PM | Permalink | Reply

    This is symbolic of the whole debate. Whilst the heavy black lines and coloured lines that take off into the stratosphere represent the mainstream ‘consensus’ view that is rammed down our throats, that plunging little red line represents the skeptical alternative view that they (the consensus) wish would just go away… – so they removed it….

  15. BillyBob
    Posted Mar 17, 2011 at 4:58 PM | Permalink | Reply

    What did the proxies do after 1999?

    • Posted Mar 17, 2011 at 8:08 PM | Permalink | Reply

      Doesn’t matter, since they apparently went on strike after 1960.

  16. Posted Mar 17, 2011 at 6:08 PM | Permalink | Reply

    From afar, it is encouraging that Science magazine is now discussing the issue.

    That suggests that the editor and publisher are now feeling the “heat” coming from budget review time and the new Congress.

  17. BillyBob
    Posted Mar 17, 2011 at 8:23 PM | Permalink | Reply

    “and are anomalies from the 1961-90 mean”

    He means the “instrumental 1961-90 mean”.

    Shouldn’t it be the “proxy 1961-90 mean that sets the zero point? Why do none of the proxies seem to be above zero. Some of them should be.

  18. Geoff Sherrington
    Posted Mar 17, 2011 at 9:17 PM | Permalink | Reply

    One is drawn to a conclusion that the graphs above have next to no climatic meaning.

    Since the 1960s, the temperature records accepted as “useful” for calibration of proxies have grown longer, they have been repeatedly adjusted and they have moved in some unforecast directions, like downwards at present in many regions.

    This means that reconstructions incorporating modern temperature data and up-to-present proxy measurements would have a high probability of producing graphs that do not resemble those above, because the calibration factors would likely be different. Also, the statistical expertise added since the 1960s would affect these graphs if analysis was reperformed.

    With each new look at the old data, the public value to the understanding of past climates decreases and the confidence that the public can put on the authors of the graphs diminishes.

    Have we passed the point of no return? Is this now appropriate to be labelled as junk science?

  19. Charlie A
    Posted Mar 17, 2011 at 11:50 PM | Permalink | Reply

    Geoff Sherrington notes: “.. reconstructions incorporating modern temperature data and up-to-present proxy measurements would have a high probability of producing graphs that do not resemble those above, because the calibration factors would likely be different.”

    That would be a good paper for a graduate student to publish. Of course, if he were in the climate science field he runs the risk of being blackballed. But maybe there are somebody in the statistical fields that gets intrigued by the question of what the hockey stick and other proxy curves would look like using up-to-date data, and what they would look like using more reliable statistical analysis.

  20. Alexander K
    Posted Mar 18, 2011 at 9:05 AM | Permalink | Reply

    It occurred to me that if tree rings are such great and accurate thermometers, why did somebody go to great lengths to invent the thermometer. Which gives me a wonderful Heath Robinson-ish vision of an automatic tree coring machine that would do the job. Much easier than re-inventing the wheel!

  21. oneuniverse
    Posted Mar 18, 2011 at 9:58 AM | Permalink | Reply

    “I trust that history will give us all proper credit for what we’re doing here.” (Michael Mann, 0926010576.txt)

  22. Hu McCulloch
    Posted Mar 18, 2011 at 10:01 AM | Permalink | Reply

    Just in case Scimag and Review of Geophysics haven’t been reading CA, I hope Steve will send them comments documenting these deceptive graphics and suggesting that these articles (including the Jones RG article in the preceding post) be retracted.

  23. Mike Jonas
    Posted Mar 18, 2011 at 4:16 PM | Permalink | Reply

    Steve – Many thanks for your fantastic work. But there is one more trick that you may have missed :

    Hu McCulloch touches on it – there is earlier instrument data which is not shown (he says 1871 but Hadley Centre data actually goes back to about 1850). But the trick is worse than that. If you look at the Briffa and Osborn graph above, you will see that the instrument record starts with a dip. That dip has been removed in later versions of the graph, eg. see WG1 fig 2.21 in the TAR : http://www.grida.no/publications/other/ipcc_tar/?src=/climate/ipcc_tar/wg1/069.htm#2322

    Apologies if I have missed you not missing it.

  24. Harold
    Posted Mar 19, 2011 at 7:32 AM | Permalink | Reply

    I should probably make an old fashioned flip book with everything rescaled to the same size, just to watch the graphs dance.

    I remember when I first read AR4, I thought the graphs were difficult to read due to line weight and low resolution, but from what I saw there were obvious periods where things didn’t correlate well. If companies used charts like these to pitch their stock, they’d be in deep deep trouble.

  25. jeff Id
    Posted Mar 19, 2011 at 9:05 PM | Permalink | Reply

    You have to wonder when Judith Curry’s excellent site gets thousands of comments and an absolutely powerful post like this gets 33.

    Again, this should be on the front pages of the garbage known as news these days yet we get nothing. I’m certain that thousands have viewed it, but really when things this obvious are slammed home in the face of blatant [self snip] by [self snip] and not covered, what are we to think?

    It makes me wild enough that I need a blog. Something to let off the pressure.

    Crazy world.

    • kim
      Posted Mar 20, 2011 at 9:47 AM | Permalink | Reply

      jeff, if the world cools and attention continues to be paid, this cluster of decline hiding will explain much to the future historians of science. If the world warms, from CO2 or not, it will be ignored. And whether the world warms or cools is probably out of our hands. Yep, crazy, but real.
      ===============

    • Steve McIntyre
      Posted Mar 20, 2011 at 11:22 AM | Permalink | Reply

      For some reason, traffic has held up much more than comments. My traffic has gone down partly because I’m posting less and partly due to more climate blogs. I have more restrictive comment policies as well. But I still get 10-20,000 hits per day even when there are hardly any comments. So I don’t think that the point is lost.

      • Posted Mar 20, 2011 at 12:58 PM | Permalink | Reply

        I read this piece mainly to get my fill of the crazy-talk. Don’t comment too much because it is not worth it, but I want to be prepared when the climate skeptics come after our work on the much more important topic of oil depletion.

      • Posted Mar 21, 2011 at 9:07 AM | Permalink | Reply

        I visit regularly and read new posts and comments. I seldom comment because my thoughts are almost always already better stated by others. Please don’t take a lack of comments as a lack of interest!

        • MikeH
          Posted Mar 21, 2011 at 9:19 AM | Permalink

          Like Peter said. I’ve been visiting at least once per day for 5 (?) years and remain as interested today as when I first found the site. I have little to contribute technically and keep my non-technical comments to myself (or make my kids listen to them!). Many thanks.

      • Hu McCulloch
        Posted Mar 21, 2011 at 9:59 AM | Permalink | Reply

        These last two posts are very important, as they document “Hide the Decline” in two peer (pal?) reviewed articles by Jones and Briffa well before TAR or WMO.

        I attribute the lack of commentary to the fact that Steve’s graphs already say everything that needs to be said.

      • observa
        Posted Mar 21, 2011 at 5:47 PM | Permalink | Reply

        Many of us come here with much sketchier statistical credentials, but nevertheless can glean the essence of the disputed claims. We may well rely on other sources to fill in the knowledge gaps for us, or better explain the nuances of the finer points of respective arguments in layman’s terms. In my case my field is economics with a background in economic statistics, although for 26 years I’ve been in business and have met men who left school at 15 or 16 who have as good a grasp of economics that any Sandstone can elicit. Likewise there is no such animal as a climatologist, anymore than I and a bunch of clever mates, with the right connections, farnarckling ourselves into the corridors of power could claim to be the world’s leading experts in Humanology.

        Yes I’ve formed a view about AGW from here and other sources as well as my own experience and no-one sums up my overall view better than a confessed warmist in Prof Richard Butler here-

        A refreshing breath of fresh air, after a very tawdry episode for Western academe and the politicisation of science and the scientific method. For that breath of fresh air and no doubt more to come now that the way has been illuminated, I believe we owe a great debt to true men of science like Steve.

  26. Posted Mar 20, 2011 at 3:15 PM | Permalink | Reply

    Notice how an ‘anomaly’ is something that disagrees with the assumed AGW conclusions. In fact, the data that diverged was against the most accurate of all temperature records (1960-now) and it is therefore the most precise.

    In which, of course, we conclude there is little local (never mind global) temperature signal in tree rings.

    What they call and ‘anomaly’ is actually the actual measured relationship between rings and temps – i.e., “none”.

    This is only an ‘anomaly’ to the AGW theory. Any wonder these people hid the data which proved they had no mathematical basis for their cries of gloom and doom?

    • Dave Dardinger
      Posted Mar 20, 2011 at 6:15 PM | Permalink | Reply

      Re: AJStrata (Mar 20 15:15),

      It should be pointed out that the actual reason for the “decline” is that they were primarily based on strip-barked trees, which don’t react to temperature at all, but to the physical process of recovery from strip-barking. In fact most all of the proxies which actually show the hockey-stick shape do so for reasons not particularly related to temperature increases. The only exceptions are the ones which were initially cherry-picked in the selection process and then assumed to be “special” by the team. Once you’ve eliminated bad proxies and cherry-picked ones, you really have nothing like a hockey-stick left.

      Anyone who doesn’t believe the above, I challenge to provide a counter-example or two so we can got to Steve’s discussion of said proxies and see how they hold up to auditing. Remember:
      1. Not flawed on the face. (Or on reasonable analysis of the proxy’s metadata like lack or statistical power or failure to meet reasonable criteria for proxiness.)
      2. Not originally chosen by cherry-picking. (Though if you can show a peer-reviewed analysis showing why said proxy should be a proxometer as opposed to other proxies which aren’t hockey-stick shaped, that will do.)
      3. Has a hockey-stick shape. (Or at least contributes strongly to the hockey-stick shape of a published proxy temperature reconstruction.)

      • Geoff Sherrington
        Posted Mar 20, 2011 at 8:13 PM | Permalink | Reply

        Dave Dardinger
        Sometimes we are criticised for not publicising or publishing more counter examples, or more advanced applications of mathematics to known data. (Though Steve does this with examples like Starbucks of a couple of years ago). However, I do not see much modernisation of past papers by climate scientists nor retraction of those now reasonably known to be inaccurate.

        My personal reason for not recalculating is that any proxy calibration based on temperatures in the instrumented period has to accept the instrumented record as correct. Personally, I do not. If the record is wrong, a new reconstruction of old proxy data will carry the known uncertainties of the temperature record.

        Because proxy reconstructions often involve relatively small geographic areas, the temperature record of importance is the one for that region, not for a globe or a hemisphere.

        Accordingly, for assistance in Australian work, I have been assisting for some months with data entry on a large scale for a new project looking at UHI effects. The kindergarten approach of the Jones/CRU team in the early 1990s was damaging. (The Australia/China/Russia Nature paper is an example). It has a legacy of entrenched myths that need to be demolished. More recently it is emerging that the changeover from mercury thermometers read daily or so, to almost continuous readouts from thermistors/thermocouple devices, has complexities that are little discussed in the literature. Changeover period roughly 1988-8 here.

        What is your suggestion for new constructions based on old data? Do you trust the integrity adequately? Take another look at what an innocent researcher could select for temperature calibration of a proxy from the Darwin region. All of these lines are derived without modification by me from official, public, web or CD data (apart from a rare infill of missing data; also, horizontal lines at the start and end of records denote no data collection and have been given artificial values to aid to easy Excel use).

        There is also stonewalling by officials. This graph from remote Macquarie Island shows essentially no temperature trend for the last 40 years. It was discussed at a BOM seminar in Hobart last year and a BOM Chief invited me to read the paper. Unfortunately, after the event, they refuse to release it.

        • Dave Dardinger
          Posted Mar 20, 2011 at 11:30 PM | Permalink

          Re: Geoff Sherrington (Mar 20 20:13),

          What is your suggestion for new constructions based on old data?

          To be blunt, in every area that the “auditors” have looked at, there’s problems. So we’re way far from being able to make sensible suggestions on what sort of constructions can be made on old faulty data. But the alarmists keep claiming that there are proxies / reconstructions which are valid even if Mann 199x is wrong. But there is no such thing. They simply turn a blind on on the problems. Until they’re willing to actually engage the auditors, no progress can be made except in increasing the realization among the general public that climate science or at least paleo-reconstructions are playing a busted flush. That’s why I keep asking them to sit back down at the table and show all their cards before they try raking in the pot.

  27. Geoff Sherrington
    Posted Mar 20, 2011 at 8:15 PM | Permalink | Reply

    Post above, that’s 1998-98.

  28. Geoff Sherrington
    Posted Mar 20, 2011 at 8:16 PM | Permalink | Reply

    Just checking. It’s 1988-98.

    See the problem of carryover of errors?

  29. Laws of Nature
    Posted Mar 21, 2011 at 7:50 AM | Permalink | Reply

    Hi Steve et all,

    over at RC there are some comments on this

    http://www.realclimate.org/index.php/archives/2011/03/wahl-to-wahl-coverage/comment-page-3/#comments

    Ray Ladbury says in #106 as an answer to my statement in #103, that
    “… there are dozens of reconstructions now, and they all pretty much agree with MBH ’98…as do borehole reconstructions, etc…”

    My question was (and is), which shape they agree with, do they show the decline or the rise after 1960.
    Unfortunately this question was not allowed over there.
    Perhaps Gavin cares to comment on it over here as of why!?
    Nothing ever changes – this question is unanswered, obvious, topic related and as of last week … censored

    All the best,
    LoN

    • Craig Loehle
      Posted Mar 23, 2011 at 3:12 PM | Permalink | Reply

      The “dozens of reconstructions” that agree are mostly by the same crew of scientists using overlapping data, and often with certain key ingredients like Yamal larch or bristlecone pine, and cherry picked data. Furthermore, in this ref:
      Dergachev, V.A., Raspopov, O.M. 2010. Reconstruction of the Earth’s surface temperature based on data of deep boreholes, global warming in the last millenium, and long-term solar cyclicity. Part 1. Experimental data. Geomagnetism and Aeronomy 50:383-392.
      they show that the various spagetti graph recons do NOT agree with one another at much of any time scale. It is only a see what you want to believe thing. If you include the chopped parts the lack of agreement is even worse.

  30. Ramspace
    Posted Mar 21, 2011 at 9:53 AM | Permalink | Reply

    This article forces me to meditate once again upon the folly of relying upon proxies for one part of a graph and instruments for another. Dendrochronology is not in the least a comparable way of measuring climate!

    “Figures 2, 3, and 5 also make it evident that these singularly wide or narrow annual rings, which are considered in dendrochronology to be environmentally sensitive indicators of abrupt envrionmental change and therefore useful for cross-dating purposes, cannot be relied on to be present in every tree, even when the trees have grown in close proximity.” (Lamlom & Savidge, 2006. p. 467)

  31. Posted Mar 22, 2011 at 11:35 AM | Permalink | Reply

    I’ve had several conversations on this matter that all concluded in a very simple manor. I love this blog! I’ve seen different research that doesn’t seem to be manipulated like these figures. In fact I’ve seen a complete 1 degree C average annual change in amplitude of temperature. for the past few years. As far as 1998 is concerned, I was told that turbulence and large changes in temperature are typical and anything prior to the dates above are speculative. There is no telling the temperature in the beginning of the millennium because there was no impeccable way to tell. Therefore, the sample from which we have this data is not large enough to make a conclusion that accounts for large natural variance….Now that’s one argument..and its pretty solid, but remains ambiguous.

  32. Queen1
    Posted Mar 24, 2011 at 3:27 PM | Permalink | Reply

    Hi Steve,
    I am another lurker who visits daily but who is woefully befuddled by statistics and computer languages. I get the general idea of what you say and have certainly learned a lot about statistics just from reading your posts! You guys are the experts; nevertheless, I must say that I have always been a bit puzzled at the claims of accuracy in “reading” temperature to tenths of a degree from tree rings from 1000 years ago. It just seems like there is a lot of room for error and what I would call fuzziness in the numbers. It is not intuitively obvious that a tree ring or a bore hole can tell us what the temperature was 1000 years ago down to the tenth of a degree.

  33. Tim Sorenson
    Posted Mar 21, 2011 at 11:19 AM | Permalink | Reply

    how does one request deletion of spam?

    [RomanM: Not usually necessary. The "janitor" comes through and sweeps up periodically. ;) ]

6 Trackbacks

  1. By Top Posts — WordPress.com on Mar 18, 2011 at 7:12 PM

    [...] Hide the Decline: Sciencemag Science published one of the first spaghetti graphs (in Briffa and Osborn 1999 here) as part of an invited comment on [...] [...]

  2. By Climate Audit on Mar 23, 2011 at 5:57 AM

    [...] that Briffa and Osborn (Science 1999) had not just deleted the post=1960 decline (see also CA here), but had deleted the pre-1550 portion as well – the deletions contributing to an unwarranted [...]

  3. [...] curva de Man (al famoso palo de hockey) está masajeada. Lean a Steve McIntyre y sabrán más: 1, 2 y [...]

  4. [...] that Briffa and Osborn (Science 1999) had not just deleted the post=1960 decline (see also CA here), but had deleted the pre-1550 portion as well – the deletions contributing to an unwarranted [...]

  5. [...] that Briffa and Osborn (Science 1999) had not just deleted the post=1960 decline (see also CA here), but had deleted the pre-1550 portion as well – the deletions contributing to an unwarranted [...]

  6. [...] that Briffa and Osborn (Science 1999) had not just deleted the post=1960 decline (see also CA here), but had deleted the pre-1550 portion as well – the deletions contributing to an unwarranted [...]

Post a Comment

Required fields are marked *

*
*

Follow

Get every new post delivered to your Inbox.

Join 3,184 other followers

%d bloggers like this: