PAGES2K South America

A commenter observed that the forthcoming PAGES2K received over 50 pages of review comments from one reviewer. One wonders what he had to say about the PAGES2K South American network which has some very odd characteristics.

Here is a list of proxies with a couple of interesting features highlighted.

soamer network

First, note that the “proxy” network includes four instrumental records, which seems to be peeking at the answer if the “skill” of the early portion of the reconstruction is in any way assessed on the ability of the network (including instrumental) to estimate instrumental temperature.

Second, one-third of the tree ring series are inverted. Is this an ex ante relationship or mere ex post correlation? We’ll find out eventually, I guess.

Finally, the two longest series are from Quelccaya, a site that we’ve regularly discussed. In the PAGES reconstruction, the d18O values are said to have been inverted. If this is how they actually used the data, as opposed to an error in the SI, it will be a surprise. A quick reverse engineering check indicates that the d18O orientation is inverted (A multiple correlation of recon against proxies in the 857-963 period.)

The PAGES2K Quelccaya version is different from any other Thompson version (as usual.) The graphic below compares the PAGES version (ending in 1995) with the PNAS version (archived in 2006) and the most recent (2013) version. The PAGES version has a sharp downtick in the late 1980s that was not reported in the PNAS version (ending in 1997) or in the 2013 version, though earlier aspects of the graphic cohere. Where did the new version come from? With Thompson, these inconsistencies are the rule, rather than the exception. If Thompson’s data is to be used by IPCC, every damn sample and measurement should be archived so that there is an audit trail.

quelccaya versions

32 Comments

  1. HaroldW
    Posted Apr 20, 2013 at 3:59 PM | Permalink

    For the tree proxies, it seems improbable that CAN 4, at (39 S, 71.08 W), would physically differ significantly from CAN 31, at (39.03 S, 71.05 W), yet one is inverted and the other isn’t. I’m betting on an auto-flipper.

  2. John Ritson
    Posted Apr 20, 2013 at 4:54 PM | Permalink

    At least the instrumentals had a positive correlation.

    Steve: this is funnier than you think. Several of the instrumental “proxies” in MBH98 had negative correlations to the overall temperature PC1 and were flipped in the reconstruction.

  3. Posted Apr 20, 2013 at 5:22 PM | Permalink

    “every damn”

    Ouch doc!! Your Id is showing.

  4. Posted Apr 20, 2013 at 5:32 PM | Permalink

    I don’t want to be the pile on crowd but — Deservedly so.

  5. H.M.
    Posted Apr 20, 2013 at 6:16 PM | Permalink

    About half an hour ago I received an RSS through Blogger of a post in Real Climate about PAGES 2k, a guest post actually, by Darrell Kaufman. I took some minutes before looking it up, but by the time I arrived the post (if it ever existed) was no more.
    The short description in the RSS I received was as follows:
    “Guest commentary by Darrell Kaufman (N. Arizona U.) In a major step forward in proxy data synthesis, the PAst Global Changes (PAGES) 2k Consortium has just a suite of continental scale reconstructions of temperature for the past two millennia in Nature Geoscience. More information about the study and its implications are available at the FAQ”
    Post under revision?

  6. batheswithwhales
    Posted Apr 20, 2013 at 6:36 PM | Permalink

    Here is the RC post. I was there earlier and just shuffled back in the browser:

    The PAGES-2k synthesis
    Filed under: Climate Science IPCC Paleoclimate — group @ 20 April 2013

    Guest commentary by Darrell Kaufman (N. Arizona U.)

    In a major step forward in proxy data synthesis, the PAst Global Changes (PAGES) 2k Consortium has just published a suite of continental scale reconstructions of temperature for the past two millennia in Nature Geoscience. More information about the study and its implications are available at the FAQ on the PAGES website and the datasets themselves are available at NOAA Paleoclimate.

    The main conclusion of the study is that the most coherent feature in nearly all of the regional temperature reconstructions is a long-term cooling trend, which ended late in the 19th century, and which was followed by a warming trend in the 20th C. The 20th century in the reconstructions ranks as the warmest or nearly the warmest century in all regions except Antarctica. During the last 30-year period in the reconstructions (1971-2000 CE), the average reconstructed temperature among all of the regions was likely higher than anytime in nearly 1400 years. Interestingly, temperatures did not fluctuate uniformly among all regions at multi-decadal to centennial scales. For example, there were no globally synchronous multi-decadal warm or cold intervals that define a worldwide Medieval Warm Period or Little Ice Age. Cool 30-year periods between the years 830 and 1910 CE were particularly pronounced during times of weak solar activity and strong tropical volcanic eruptions and especially if both phenomena often occurred simultaneously.

    Figure: Thirty-year mean relative temperatures for the seven PAGES 2k continental-scale regions arranged vertically from north to south.

    The origin of the ‘PAGES 2k Network‘ and its activities can be found here and consists of nearly 80 individual collaborators. The Consortium’s collection of local expertise and proxy records was transformed into a synthesis by a smaller team of lead authors, but the large author list recognizes that the expertise of the wider team was essential in increasing the range of data used and interpreting it.

    In addition to the background available at the FAQ, I think it is important to also highlight some aspects of the analytical procedures behind the study and the vital contributions of three young co-authors.

    The benefit of the ‘regions-up’ approach embodied in the PAGES-2k consortium is that it made it easy to take advantage of local expertise and include a large amount of new data that would have been more difficult to assemble for a centralized global reconstruction. However, being decentralized, the groups in different regions opted for different methodologies for building their default reconstructions. While justifiable, this does raise a question about the impact different methodologies would have. To address this, the synthesis team (ably led by Nicholas McKay) applied three particular reconstruction methods to all of the regions, as well as looking at the basic area-averaged and weighted composites. He further analyzed the site-level records individually and without many of the assumptions that underlie the regional temperature reconstructions. These results show that the long-term cooling trend and recent warming are dominant features of the dataset however you analyze it. There is a sizable fraction of the records that do not conform to the continental averages, highlighting the spatial variability and/or the noise level in specific proxies.

    One of the new procedures used to reconstruct temperature is an approach developed by Sami Hanhijärvi (U. Helsinki), which was also recently applied to the North Atlantic region. The method (PaiCo) relies on pairwise comparisons to arrive at a time series that integrates records with differing temporal resolutions and relaxes assumptions about the relation between the proxy series and temperature. Hanhijärvi applied this procedure to the proxy data from each of the continental-scale regions and found that reconstructions using different approaches are similar and generally support the primary conclusions of the study.

    Regions where this study helps clarify the temperature history are mainly in the Southern Hemisphere. We include new and updated temperature reconstructions from Antarctica, Australasia and South America. The proxy records from these three regions come from many sources, ranging from glacier ice to trees and from lake sediment to corals. Raphael Neukom (Swiss Federal Research Institute WSL and University of Bern) played a key role in the analyses across the Southern Hemisphere. He used principal components regression (Australasia), a scaled composite (Antarctica), and an integration of these two approaches (South America) to create the time series of annual temperature change.

    Inevitably, assembling such a large and diverse dataset involves many judgement calls. The PAGES-2k consortium has tried to assess the impact of these structural decisions by using multiple methods, but we hope that this synthesis is really just the start of a more detailed analysis of regional temperature trends and we welcome constructive suggestions for improvements.

    • Chad Jessup
      Posted Apr 20, 2013 at 8:33 PM | Permalink

      Anytime a study claims that the most recent high temperatures exceeded the high temperatures of the well documented Medieval Warm Period, I gringe, because there are solid reasons why certain crops would grow at higher latitudes during the MWP than today and why certain fish that the Vikings caught by Greenland were there then but not today.

      Still using tree rings as a proxy for temperature! In this area, tree rings correlate well with precipitation but not with temperature.

    • Jean S
      Posted Apr 21, 2013 at 10:52 AM | Permalink

      Re: batheswithwhales (Apr 20 18:36),

      Regions where this study helps clarify the temperature history are mainly in the Southern Hemisphere. We include new and updated temperature reconstructions from Antarctica, Australasia and South America. The proxy records from these three regions come from many sources, ranging from glacier ice to trees and from lake sediment to corals. Raphael Neukom (Swiss Federal Research Institute WSL and University of Bern) played a key role in the analyses across the Southern Hemisphere. He used principal components regression (Australasia), a scaled composite (Antarctica), and an integration of these two approaches (South America) to create the time series of annual temperature change.

      I’m stunned. Seems like they are actually proud of these reconstructions?!? South America is the one this post is about, Australasia is the living dead Gergis et al., and Antarctica is the worst reconstruction I’ve seen (and I’ve seen a lot of them). I really don’t know what to think of these people.

      In the PAGES2k Antarctica they first standardise nine (out of eleven; “Ind22/B4” and “James Ross Is” are not used; don’t know why) to zero mean unit variance over the maximum overlap. Then they form the recon as the correlation (to Steig’s “intrumental”; each proxy using maximum overlap) based weighted mean. Then they rescale the recon to match Steig’s “instrumental” over the period 1961-1991, which they call the “calibration period”! “Verification” is done by dividing the “calibration period” in half, and doing the last step for the other half, and calculating the RE from the other half! With this method they obtain negative average verification RE (as they do for the other two recons, see below)! In other words, simple mean outperforms their reconstruction. Since the advocator (Fritts) of the RE statistic has said that “any positive value indicates there is some information in the reconstructions”, I conclude that apparently there is no information in PAGES2K Antarctica reconstruction.

      But that not all folks! Instead of reporting that reconstruction, it seems they used another in the actual paper (at least the above reconstruction is not the one listed in the PAGES2K database of the regional reconstructions). The reason for this is rather apparent when you plot the above reconstruction. Instead, they calculated West (six proxies; now including the James Ross Is omitted in the full reconstruction) and East (four proxies!) Antarctica reconstructions with the same method outlined above. The East Ant reconstruction has the “magnificent” calibration RE of 0.028 (after AD 1232; before it’s negative), which has to be the lowest calibration RE ever reported on any reconstruction. Then these two reconstruction are combined by “areal weighting” (=25% West, 75% East) to a new reconstruction, which is then somehow again rescaled (although already in temperature units; I couldn’t figure out exact rescaling details, but visually these two match) to obtain the reconstruction apparently used in the paper.

      They are also calculating (but not reporting) r2 values. I haven’t checked those yet, but I believe the famous low r2 record, which many of us seriously thought to be unbeatable, set by Mann is now seriously treathened.

      I really don’t know what to say when this is type of crap (excuse my French) passes an extremely tough review process except I’m shocked. Truly.

      If someone wants to check the above claims, the Antarctica “mole hole” is here:
      ftp://ftp.ncdc.noaa.gov/pub/data/paleo/pages2k/auxiliary-information/

      Edit: I striked out the wrong information. They are indeed using the full recon but they have cut it (2-166AD with high values) to match the length of the combined reconstruction!

      • Skiphil
        Posted Apr 21, 2013 at 11:05 AM | Permalink

        In light of all the controversies in climate science one would think that the PAGES and IPCC processes would be reformed and refined by now, to produce much better work. Are these people simply this brazen and shameless, or are they really incapable of understanding and addressing criticism?

      • Steve McIntyre
        Posted Apr 21, 2013 at 1:55 PM | Permalink

        # Mostly extracted from library of code from R. Neukom

        # Make the proxy composite ouf of the scaled proxies weighting them with their individual cor with the target

        So they’re not just screening by correlation, but going full MBH and weighting by correlation. My, my. Since they’re using Neukom’s code, I presume that’s what they did in the Gergis zombie as well.

        • Jean S
          Posted Apr 21, 2013 at 3:03 PM | Permalink

          Re: Steve McIntyre (Apr 21 13:55),
          yes, and notice from Kaufman’s RC piece who’s responsible for the South American recon, where the ~AD1200 values barely match to those of the modern values. Hanhijärvi’s PaiCo results are completely different.

      • pottereaton
        Posted Apr 21, 2013 at 3:26 PM | Permalink

        This gives new meaning to the phrase: “It’s worse than we thought.”

        • JEM
          Posted Apr 21, 2013 at 4:10 PM | Permalink

          I’m reminded of Wellington’s remark: They came on in the same old way, and we shot them down in the same old way.

    • Bob Koss
      Posted Apr 21, 2013 at 4:22 PM | Permalink

      RC has reinstated the post by Kaufman. Evidently around mid-day today. First comment 21 Apr 2013 at 12:20 PM.

      I don’t see any textual or graphical changes.

  7. batheswithwhales
    Posted Apr 20, 2013 at 6:44 PM | Permalink

    There was a figure with a temperature grid as well which didn’t come along in my copy of the post.

    I put it up on my website if anyone is interested:

    http://klimatilsynet.wordpress.com/

  8. EdeF
    Posted Apr 20, 2013 at 8:41 PM | Permalink

    Would be interesting to know if tree ring that increase and decrease are from the same species, in roughly the same area? Reminds me of Salzer and friends circa 2006, the bristlecones just below the tree line ramped up during the instrumental period, and Methusulah Grove, down the hill just a bit in the heart of the forest……….went down.

  9. Posted Apr 20, 2013 at 8:46 PM | Permalink

    It’s more than just coincidental
    That the lines now marked “instrumental”
    Seemed to be filter-tripped
    For some there were flipped
    Such treatment, to me, seems ungentle

    ===|==============/ Keith DeHavelle

  10. JCM
    Posted Apr 20, 2013 at 10:38 PM | Permalink

    “There is now enough evidence to say what many have long thought: that any claim coming from an observational study is most likely to be wrong in the sense that it will not replicate if tested rigorously,” wrote S. Stanley Young and Alan Karr of the U.S. National Institute of Statistical Sciences in a recent critique of the trend.
    http://news.nationalpost.com/2013/04/19/science-vs-math/

  11. Geoff Sherrington
    Posted Apr 20, 2013 at 11:22 PM | Permalink

    “but we hope that this synthesis is really just the start of a more detailed analysis of regional temperature trends and we welcome constructive suggestions for improvements”

    What business does work in progress have, getting mixed up in IPCC submissions, however late they are?

    First you do your work, then you replicate it, over and over, then you review it, then (if it passes) you release it in final form with final conclusions to allow outside analysis.

    BTW: I have quite detailed local knowledge of the Australian region, but not the Australasian region selected by the Gergis/Pages groups. My first reaction to the first Gergis pre-release was “Where is the Australian content?” Those examples which do come from Australia, which I have studied, are not good enough to go to making global decisions. Each has substantial unsolved problems, such as an absence of long term, good quality, climate relevant instrumental temperature data against which to calibrate.

    If you can’t sort your instrumental temperatures (and I have tried several times), there is little chance to sort proxy temperatures. A wiser set of investigators would have quickly concluded that the path to progress with their data set was arid and strewn with boulders and a wise group would possibly have halted. Instead, we get this hodge-podge of uncertain, method-selective and incomplete work that draws the opening quote.

    This is not science.

  12. Jeff Norman
    Posted Apr 21, 2013 at 6:16 AM | Permalink

    Steve,

    Is the Quelccaya 2013 graph off scale in 1941/42?

    Some of the spikes seem to coincide with el Nino years suggesting it may well be a proxy for the source of the precipitation.

  13. JFB
    Posted Apr 21, 2013 at 7:39 AM | Permalink

    Please, can I get the location of Campinas (Brazil) instrumental data location ?

  14. mt
    Posted Apr 21, 2013 at 9:54 AM | Permalink

    It looks like the PAGES version may be an updated version of this. Comparing q83summ.txt plot by eye looks the same, although the PAGES data you have plotted seems shifted by one year relative to the 2013 data, the archived q83summ matches 2013 better.

    Steve: Nope. The PAGES version goes to 1985 and its the additional decade that caught my eye. q83sum ends in 1983-84. Plus I’m obviously familiar with it.

    • mt
      Posted Apr 22, 2013 at 7:54 AM | Permalink

      I saved the south american proxy sheet from the supplemental spreadsheet as a csv file:

      q83sum = read.table(“q83summ.txt”, skip=26, nrows=1241);
      SAm = read.csv(“SAmProxies.csv”,stringsAsFactors=FALSE);
      pdata=as.numeric(rev(SAm[seq(3,1243),3])); # 1984..744
      cor(q83sum[seq(1,100),6], pdata[seq(2,101)])
      [1] 0.9996512
      cor(q83sum[seq(1,1000),6], pdata[seq(2,1001)])
      [1] 0.993666

      PAGES seems to be q83summ shifted one year older and with 12 years of new data.

      Steve: It’s the new data that caught my eye. Why does it differ so much from the data published by Thompson in PNAS 2006 also covering the newer period?

      • mt
        Posted Apr 22, 2013 at 10:43 AM | Permalink

        That, I cannot answer. However, from the OSU Quelccaya page

        Recently the Quelccaya ice cap has experienced an accelerating rate of retreat as the snowline has risen in response to a marked warming. The d18O record from a core drilled at the summit in 1976 is compared with that from one drilled in 1991 and reveals that the isotopic seasonal signal is no longer preserved and that the isotopic average is enriched by 2o/oo.

        There’s a figure on that page that shows a loss of yearly signal between the two cores, which would seem to infer that the new data may not be directly comparable to the old.

        Separately, there’s this finding in the Thompson 2013 Supplemental

        It is clear that on multi-decadal time scales there is no significant relationship between stable isotopes and precipitation (net accumulation) on the Quelccaya ice cap.

        So it’s curious that the reconstruction would use both d18O and accumulation as temperature proxies.

  15. Matt Skaggs
    Posted Apr 21, 2013 at 10:18 AM | Permalink

    Everyone knows that south of the equator, trees grow upside down, so all the measurements have to be flipped. Geez, do I have to explain everything?

  16. P. pasten
    Posted May 1, 2013 at 9:35 PM | Permalink

    Don´t bother about Quelccaya and check Laguna Aculeo, the YAD061 of the South American PAGES2K reconstruction.

4 Trackbacks

  1. […] has been looking at the resurfacing of some Hockey Stick science papers, and the fun continues here and […]

  2. […] ClimateAudit has been reporting on it for a number of days. See Steve McIntyre’s posts here, here and especially here. WattsUpWithThat has discussed the paper here and here. SkepticalScience […]

  3. […] ClimateAudit has been reporting on it for a number of days. See Steve McIntyre’s posts here, here and especially here. WattsUpWithThat has discussed the paper here and here. SkepticalScience […]

  4. […] reconstruction used in IPCC AR5. I previously discussed the PAGES2K South American reconstruction here. pointing out that it had used the Quelccaya O18 and accumulation data upside-down to the […]