Kola versus Yamal

A news release on a new tree ring study here (h/t Anthony Watts) reported a reconstruction maxing out in the mid-20th century, with the characteristic late 20th century divergence problem. Their results contrast with CRU’s notorious Yamal chronology:

Following the summer temperature reconstruction on the Kola Peninsula, the researchers compared their results with similar tree-ring studies from Swedish Lapland and from the Yamal and Taimyr Peninsulas in Russian Siberia, which had been published in Holocene in 2002. The reconstructed summer temperatures of the last four centuries from Lapland and the Kola and Taimyr Peninsulas are similar in that all three data series display a temperature peak in the middle of the twentieth century, followed by a cooling of one or two degrees. Only the data series from the Yamal Peninsula differed, reaching its peak later, around 1990. What stands out in the data from the Kola Peninsula is that the highest temperatures were found in the period around 1935 and 1955, and that by 1990 the curve had fallen to the 1870 level, which corresponds to the start of the Industrial Age. Since 1990, however, temperatures have increased again evidently.

Although the reconstruction declined since mid-20th century, the sub-headline reads: “New data indicate rapid temperature rise in the coldest region of mainland Europe”.


  1. Salamano
    Posted Jul 29, 2010 at 4:24 PM | Permalink

    Couldn’t someone attempt to instantly solve this divergence by simply declaring that certain tree-ring proxies are valid…but only up until the CO2 concentrations are anthroprogenically forced above 320PPM or so, and then state that all-bets-are-off as far as future rings retaining climate sensitivity?

    I’m not sure how someone would arrive at such a conclusion empirically, but simply stating so in the right journal may be enough.

    • mpaul
      Posted Jul 29, 2010 at 5:03 PM | Permalink

      Or alterantively, “certain tree ring proxies are valid, but only up until Mickey Mantle hit his 400th home run”.

    • LearDog
      Posted Jul 29, 2010 at 5:20 PM | Permalink

      Or ….

      One could instantly solve this divergence problem by simply declaring that ALL tree-ring proxies are invalid for the purposes of reconstructing past climates – because they don’t even work during the period of we have been observing them since the dark art of Tree Ring Phrenology was created?

      A corollary is – then all-bets-are-off as far as “unprecedented warming” caused by Mann.

      And if memory serves, Briffa tried your suggestion anyway (it must be excess CO2 causing the divergence because I can’t think of anything else (without putting my life’s work at risk), there – discussed (whew! that was close!)? (snip away)

      • bobdenton
        Posted Jul 30, 2010 at 1:39 AM | Permalink

        This study appears to maintain that the tree ring growth correlates with temperature throughout the record, including in the second half of the 20th century. In other words, there is no divergence problem, the decline in tree ring growth simply reflecting a decline in temperature at the site of growth. The decline in local temperature is attributed to solar effects with which it also correlates.

        It’s not clear how the reconstruction was calibrated to local instrumental records (haven’t seen the paper itself), that should be the interesting part.

    • Posted Jul 29, 2010 at 8:50 PM | Permalink

      I’m guessing that you’re joking, because doing what you suggest is more like an imperial decree, and nothing to do with reason or logic.

      • Carbon-based life form
        Posted Jul 29, 2010 at 9:19 PM | Permalink

        Re: Jeff Alberts (Jul 29 20:50),

        • Richard T. Fowler
          Posted Jul 30, 2010 at 7:52 AM | Permalink

          This is a form of juvenile sarcasm meaning “obviously”. Carbon agrees that Salamano is joking. Having read some of Salamano’s other recent comments, it seems reasonably clear to me as well that Salamano is joking, though I did have to reflect on it for a moment.


  2. Bernie
    Posted Jul 29, 2010 at 4:43 PM | Permalink

    Have you managed to explore Salzer et al (2009) any further beyond the “trick” of changing the base period to create an imaginary similarity between strip bark and non-strip bark BCPs?

  3. BillyBob
    Posted Jul 29, 2010 at 5:45 PM | Permalink

    I firmly believe the 1930’s were the hottest decade — as in max not mean. It doesn’t suprise me that 1935 would show up in the proxy (if the proxy is actually a temperature proxy). I think higher mean temperatures are caused by UHI. (My favorite proxy for max is state temperature maximums — 25 out of 50 came in the 1930’s)

  4. Pat Frank
    Posted Jul 29, 2010 at 9:17 PM | Permalink

    Here’s the title, authors, and full abstract. Figure 7 in the paper, giving the reconstructed temperatures since 1600 through 1999, purportedly shows that 1930-1955 was as warm as now.

    Regional Summer Temperature Reconstruction in the Khibiny Low Mountains (Kola Peninsula, NW Russia) by Means of Tree-ring Width during the Last Four Centuries

    Yu. M. Kononov, M. Friedrich, and T. Boettger

    Arctic, Antarctic, and Alpine Research Vol. 41, No. 4, 2009 pp. 460-468

    Abstract: This study presents a new pine (Pinus sylvestris L.) ring-width chronology and a summer temperature reconstruction for the last 400 years from the Khibiny Low Mountains (Kola Peninsula, NW Russia). Pine trees from sites at the altitudinal timberline of Khibiny Mountains show pronounced climatic signals in tree-ring width. We found a strong positive correlation with summer temperature of July–August (r = 0.58). The reconstruction shows lower summer temperatures from A.D. 1630 to 1840, a subsequent warming up to the mid-20th century and a cooling trend afterwards. According to our data, a temperature increase is observed during the past decade. The good coherence of multi-decadal to secular trends of our reconstruction and series of observed solar activity indicate that solar activity may have been one major driving factor of past climate on Kola Peninsula.

  5. Rob Wilson
    Posted Jul 30, 2010 at 2:05 AM | Permalink

    Dear Steve et al.
    I think Kola would be better compared to Scandinavian TR proxy series.

    Here is a spatial correlation for the July-August season (mean temperatures) for the 1×1 degree grid for Yamal again Scandinavia and Russia.



    • AndyL
      Posted Jul 31, 2010 at 5:33 PM | Permalink

      Re: Rob Wilson (Jul 30 02:05),

      Rob Wilson’s comment looks interesting, so I’m surprised there have been no responses. I’m bumping it before it gets hidden by the CA-Assistant setup

      • bobdenton
        Posted Aug 1, 2010 at 3:47 PM | Permalink

        The paper to compares to both Scandanavian and Russian series. They appear to note similarities and differences rather than to suggest a correlation.

  6. Mike Mustermann
    Posted Jul 30, 2010 at 4:45 AM | Permalink

    Dear Sirs,
    isn’t there also the problem, that Mann et al. – in the way of IPCC 2007 – still argue on a hemispheric MWP and not on a global?

    I found a new interview with an Austrian researcher Dr. Böhm on http://mittelalterlichewarmperiode.blogspot.com (in German) in which Dr. Böhm argues for a globally located MWP.

    Wishes, Mike

  7. stereo
    Posted Jul 30, 2010 at 8:17 AM | Permalink

    Their results contrast with CRU’s notorious Yamal chronology

    I thought you didn’t approve of exaggeration.

    As for Steve’s claims of being first about the temperature measurement problem during WWII, did he ever publish it? Does he really expect scientists to trawl blogs?

    Steve: I don’t “expect” scientists to “trawl blogs”.

    In the case of proxy commentary, I think that specialists in the field would generally benefit from reading my commentary on their articles and defending their work against criticism. Most scientists don’t get much feedback on their work; many articles don’t seem to get read past their abstracts.

    In the case of the WWII thing, I wouldn’t have “expected” the scientists in question to have taken note of the blog post. Nonetheless, I think that I had correctly observed the incongruity of how the SST records had been adjusted and was entitled to point this out at the time. I did so without animus.

    However, sometimes scientists do ‘trawl” blogs. We’ve seen incidents where websites have been changed almost in real time following commentary here e.g. the laughable incident of Gavin and the Mystery Man, where Schmidt was trawling Climate Audit and pretending that some unknown Mystery Man was doing so.

    If scientists do in fact draw on materials from a blog, I think that they should cite the blog (and do so with an accurate page reference. My daughter’s university class had procedures for citing websites and I see no reason why such procedures are outside the capability of climate scientists if the situation arises.

    • Posted Jul 30, 2010 at 10:25 AM | Permalink

      My daughter’s university class had procedures for citing websites and I see no reason why such procedures are outside the capability of climate scientists if the situation arises.

      So did my Community College English course. It’s an absurdly simple thing.

  8. John Blake
    Posted Jul 30, 2010 at 8:27 AM | Permalink

    Since Yamal data materially “diverges” from comparable sites’, why not send a team or two there to fact-check and verify results? Or has every dendrochronological specimen suddenly met with rare but invariably fatal Russian Larch disease?

  9. Hu McCulloch
    Posted Jul 30, 2010 at 8:38 AM | Permalink

    From the press release —

    What is conspicuous about the new data is that the reconstructed minimum temperatures coincide exactly with times of low solar activity. The researchers therefore assume that in the past, solar activity was a significant factor contributing to summer temperature fluctuations in the Arctic.

    Perhaps this means that the trees are responding directly to solar activity — ie via photosynthesis or some other mechanism — rather than indirectly through temperature, which is correlated, but only loosely, with solar activity.

    Unfortunately, the press release doesn’t say just how much warming since 1990 is inferred. Are we back to the 1957 inferred peak or still short?

    • Posted Jul 30, 2010 at 8:59 AM | Permalink

      I have a graphic up. Maybe it’ll help you,

      • Posted Jul 30, 2010 at 9:00 AM | Permalink


        • Hu McCulloch
          Posted Jul 30, 2010 at 9:41 AM | Permalink

          Thanks — the graph shows 2001 as still well below 1957. Carbon shmioxide!

        • Kenneth Fritsch
          Posted Jul 30, 2010 at 11:17 AM | Permalink

          “Thanks — the graph shows 2001 as still well below 1957. Carbon shmioxide!”

          Or the graphs (and the underlying reconstructions) show nothing.

        • Bob Koss
          Posted Jul 30, 2010 at 12:16 PM | Permalink

          Re: pgosselin (Jul 30 09:00), I posted this at your site. Figured I’d also put it here.

          Certainly, something isn’t copacetic with the graph.

          I did a pixel level comparison and 2001 doesn’t even reach 13C. The center of the graphed line ends somewhere between 2005-2006. The line itself is about 3 years in width.

        • Hu McCulloch
          Posted Jul 30, 2010 at 1:11 PM | Permalink

          Re copacetic:

        • Bob Koss
          Posted Jul 30, 2010 at 1:40 PM | Permalink

          Re: Hu McCulloch (Jul 30 13:11), Wow! A whole page dedicated to the provenance of copacetic?

          I first used it circa 1955. Most people seemed to understand its usage back then. I tapered off usage in the 1980s when people started giving me quizzical looks. 🙂 Just thought I’d dust it off one more time as a substitute for satisfactory.

        • Hu McCulloch
          Posted Jul 30, 2010 at 1:17 PM | Permalink

          The detail pgosselin posts at
          shows the detail much better.

          Clearly the graph goes to mid-decade, but whether or not this is an error, it still falls far short of 1957.

          I don’t now if they accounted for CO2 fertilization in their calibration equation, but if they did not, the corrected reconstruction would be even lower at the end. (Assuming there is any correlation with temperature at all with or without this correction.)

  10. Espen
    Posted Jul 30, 2010 at 8:42 AM | Permalink

    Please note that the Kola peninsula is in the west of Russia, i.e. close to the Barents Sea and thus influenced by the ups and downs of the North Atlantic and Arctic SST. Also on Greenland (e.g. Nuuk: http://data.giss.nasa.gov/cgi-bin/gistemp/gistemp_station.py?id=431042500000&data_set=1&num_neighbors=1 ) temperatures were cooling from the 30s-40s warm period until 1990, and have since not quite reached the levels of back then. The big question is why these two warming episodes in the north Atlantic happened, and mainstream AGW theory wants us to believe that the latter episode is mostly due to AGW, while the former is left largely unexplained. But it’s not at all obvious to me that one needs to invoke external “forcings” (be it solar activity or CO2) as an explanation, the variations may simply be generated by the oceans themselves and their interactions with cloud cover.

    • Posted Jul 30, 2010 at 10:22 AM | Permalink

      I have to agree. When multiple oceans are “in-synch” with each other ,warm or cool, in their decadal oscillations, I think you’re going to get these sorts of upticks/downticks. I think they could even cause changes in equilibrium if heat is stored longer than “normal”, only to be released to the atmosphere in fits and starts.

  11. Posted Jul 30, 2010 at 8:57 AM | Permalink

    The graphic they use in the German press release goes only until 2001.
    This morning the German newspapers were all hollering about rapid warming since 1990, implying over the last 20 years.

  12. Posted Jul 30, 2010 at 10:42 AM | Permalink

    I just want to say that I spoke directly to the press spokesman Tilo Arnhold, and he told me the dataset ended at 2001. I asked him twice.
    I don’t know why the graphic seems to show the curve going well beyond 2001. Perhaps it’s a graphic crafted for the media. It’s misleading.

    • P Solar
      Posted Aug 1, 2010 at 10:58 AM | Permalink

      >> Perhaps it’s a graphic crafted for the media. It’s misleading.

      Maybe someone grafted on the recent thermometer temperature record. ( I wish I was joking! )

      Seriously, this is worth following up.

      Either they don’t know the length of the data set (but your source seems clear and firm on that) or someone tampered with the graph to give the impression this is a 20 year trend and not just 10 years which would seem more likely.

      You say the graph you posted on your blog came from the press release, could you link that?

      If the source is directly attributable to UZF or Hohenhiem maybe you could ask Tilo Arnhold for a clarification.

  13. TAC
    Posted Jul 30, 2010 at 11:59 AM | Permalink

    Knowing what we know, does anyone take the tree rings seriously as indicating more than the growth of a particular tree? Perhaps there is something to be learned about climate from the dendro record, but I would like to see some evidence.

  14. EdeF
    Posted Jul 31, 2010 at 12:31 PM | Permalink

    I noticed this reference to the above report, it is a recent study by
    Hughes, Salzer et al concerning (once again) bristlecone pines at high
    elevations in California and Nevada. Apparently, BCPs at the upper treeline
    level have been showing “unprecedented” growth in the last 50 years. They
    also find no difference in tree ring width in this time between strip bark
    and whole bark trees. Sheep Mt. in the White Mts is the main area of interest
    along with two sites in Nevada. What is really interesting however is that
    at elevations below the upper tree line the growth is fairly flat; the
    plot at Cottonwood looks like the typical NH curve, that is, rise in the
    first half of the 20th century, then cooling dip to 1970s, then a steep
    rise into the 90s, then downward trend from there. My question is, why do
    they attribute more importance to the very upper treeline trees, which show
    a giant growth spurt, than to all of the other high altitude trees, which don’t? Isn’t it really important to understand what is causing this huge divergence in tree ring widths from trees in the same area, undergoing about the same type of environmental conditions? I am also concerned about
    the abundant use of adjectives in a science report.


  15. P Solar
    Posted Aug 1, 2010 at 11:19 AM | Permalink

    My question is, why do
    they attribute more importance to the very upper treeline trees, which show
    a giant growth spurt, than to all of the other high altitude trees, which don’t?

    It certainly makes a fairly convincing reason NOT to use such atypical sites to create temperature proxies without a full and detailed understanding of the reasons for this anomalous growth pattern.

    Since that is clearly lacking publishing anything claiming to be a temperature proxy seems very unscientific. (Although it may be a good means to attract grant money for next year!)

    Trees on or near the tree line are by definition in stressed conditions. This will make them highly sensitive to certain changes in their environment.

    That increased sensitivity will increase the experimental uncertainty on everything from the physical measurement of the ring widths, the assumed adjustment for age drying of the wood to the projection of width to temperature.


    • P Solar
      Posted Aug 1, 2010 at 11:22 AM | Permalink

      Just clarifying , those comments apply to the PNAS article : a reference of the subject here, not the kola article. Although, much of it applies equally.



  16. scientist
    Posted Aug 1, 2010 at 1:03 PM | Permalink

    37 posts (so far) speculating on the implications of the referenced study, but all based on the press release and the abstract, rather than the paper itself. I understand the welcoming appreciation for “another paper that backs up our side”. But would suggest actually going to the paper and reading it and seeing how much it supports the ideas we hope it does.

    For instance,

    A. Divergence: If the rings were higher mid-century than recently, does that correspond to local temperature doing the same thing (or not)? I don’t know, but a curious objective person would want to check, vice just fitting it into the metanarrative as a “good study” unlike the “bad study” of Yamal.

    B. How does this study correspond to the other two “good studies” or the “bad study” of Yamal? Are there reasonable explanations in terms of geography and/or biology to explain why the studies differ?

    Note: I’m very OPEN to the possibility that Yamal was a “bad study”…and that if we keep doing more and more studies in Northern Europe, we’ll get more and more stakes through the heart (of Yamal). But we should just really check it out and make sure that this simple story is the best explanation. IOW, I “want” it to be true. but I know enough, to actually “pressure test” what I want and see if it holds up. If it survives “pressure testing” than it makes our case stronger. This approach of being self-critical is something that both sides need more of.

    Steve: the measurement data doesn’t seem to be archived. This makes it hard to analyse.

    • scientist
      Posted Aug 1, 2010 at 8:13 PM | Permalink

      That’s unfortunate. I still suggest that there is way more detail in the paper than in press releases and that you might change some perceptions from the limited info in the press release by reading the paper. Even if not, your hypotheses about the implications of the Kola study will be strengthened from the paper reading.

      If you are not paying for a comprehensive literature package or using a university library, your ability to understand the field where you are doing work, will be highly negatively impacted.

      Reading a paper instead of the press realease is like looking at the 10K. Yes, it is not as good as digging into the sales ledgers, but it’s far superior to relying on an earnings report. And takes a modicum of effort. Just go to SEC Edgar for the 10K. Go to the Toronto library for the paper. Basic stuff every grad student does.

      P.s. My comment on geography was actually made without having read Rob’s. It’s just an example about how to think curiously about things. (And I’m not asserting that Kola is not dirrectly analagous to Yalta…just something to look at!)

      Steve: articles are generally considered in exhaustive detail at this site, often with examination of original data if it is not withheld. I notice that you are replying here to your own post – so I presume that you are offering this sanctimonious advice to yourself.

      • scientist
        Posted Aug 1, 2010 at 8:32 PM | Permalink

        I replied to your comment in the bold, that you type into my post.

        Steve: fair enough. Why would you assume that I hadn’t obtained a copy of the article? I’m working on other things right now, but will look at it in more detail in due course.

        • scientist
          Posted Aug 1, 2010 at 8:58 PM | Permalink

          Because your post did not seem to indicate anything beyond the abstract and press release. Did not seem to be based on a good reading of the paper (for instance, not discussing limitations of the study, geography comparisons, etc.)

          Also the headpost puts the press release front and center, vice the paper front and center, as the object of interest. Why not give the citation for the paper?

          And none of your commenters give the impression of having read the paper either (they’re not digging into the content). That was my simple point.

          Looking forward to your insightful discussion of the paper (even if you don’t have the data) when you get time.

          P.s. This is getting silly for both of us. No hard feelings. You may have the last word, bold-edited, Voice of God style into “my rectangle”, like Gavin at RC. 😉

3 Trackbacks

  1. By Top Posts — WordPress.com on Jul 30, 2010 at 7:12 PM

    […] Kola versus Yamal A news release on a new tree ring study here (h/t Anthony Watts) reported a reconstruction maxing out in the mid-20th […] […]

  2. […] Climate Audit [–>] […]

  3. […] had mentioned this new dendro paper to Steve McIntyre, who wrote a short note about it while pointing out […]

%d bloggers like this: