More on the Divergence Problem

Two new things on the Divergence Problem. The IPCC First and Second Drafts did not contain a whisper of a mention of the divergence between ring widths and density in the second half of the 20th century, although this is rather an important issue. It came up at the NAS Panel and was completely unresolved in the hearings discussed here, where D’Arrigo was only able to refer to Briffa’s cargo cult explanation of the phenomenon.

In the absence of a substantiated explanation for the decline, we make the assumption that it is likely to be a response to some kind of recent anthropogenic forcing. On the basis of this assumption, the pre-twentieth century part of the reconstructions can be considered to be free from similar events and thus accurately represent past temperature variability.

In their report, the Panel adopted the view of Cook et al on the matter, which I’ve analyzed previously, in which bristlecones and foxtails seduce the Team once more. The final version of IPCC contains a lengthy paragraph on the Divergence Problem excerpted in full below. D’Arrigo, Wilson also have a new article, which I’ve not seen yet, but which is discussed at Pielke Sr here .

It appears that Cook et al 2004 is relied upon in this article. I discussed this article previously as More Cargo Cult. It will be interesting to see whether D’Arrigo, Wilson et al considered the issues raised in this post (not because they are obligated to deal with issues raised at this blog, but because they are obligated to deal with germane issues) .

The IPCC paragraph is as follows. (A question in passing: given that this paragraph was not presented in the Second Draft, what are the “peer review” procedures involved in vetting this? Or is Briffa’s say-so good enough?

All of the large-scale temperature reconstructions discussed in this section, with the exception of the borehole and glacier interpretations, include tree ring data among their predictors so it is pertinent to note several issues associated with them. The construction of ring width and ring density chronologies involves statistical processing designed to remove non-climate trends that could obscure the evidence of climate that they contain. In certain situations, this process may restrict the extent to which a chronology portrays the evidence of long time scale changes in the underlying variability of climate that affected the growth of the trees; in effect providing a high-pass filtered version of past climate. However, this is generally not the case for chronologies used in the reconstructions illustrated in Figure 6.10. Virtually all of these used chronologies or tree ring climate reconstructions produced using methods that preserve multi-decadal and centennial time scale variability. As with all biological proxies, the calibration of tree ring records using linear regression against some specific climate variable represents a simplification of what is inevitably a more complex and possibly time-varying relationship between climate and tree growth. That this is a defensible simplification, however, is shown by the general strength of many such calibrated relationships, and their significant verification using independent instrumental data. There is always a possibility that non-climate factors, such as changing atmospheric CO2 or soil chemistry, might compromise the assumption of uniformity implicit in the interpretation of regression-based climate reconstructions, but there remains no evidence that this is true for any of the reconstructions referred to in this assessment. A group of high-elevation ring width chronologies from the western USA that show a marked growth increase during the last 100 years, attributed by LaMarche et al. (1984) to the fertilizing effect of increasing atmospheric CO2, were included among the proxy data used by Mann et al. (1998, 1999). However, their tree ring data from the western USA were adjusted specifically in an attempt to mitigate this effect. Several analyses of ring width and ring density chronologies, with otherwise wellestablished sensitivity to temperature, have shown that they do not emulate the general warming trend evident in instrumental temperature records over recent decades, although they do track the warming that occurred during the early part of the 20th century and they continue to maintain a good correlation with observed temperatures over the full instrumental period at the interannual time scale (Briffa et al., 2004; D’Arrigo, 2006). This “divergence” is apparently restricted to some northern, highlatitude regions, but it is certainly not ubiquitous even there. In their large-scale reconstructions based on tree ring density data, Briffa et al. (2001) specifically excluded the post-1960 data in their calibration against instrumental records, to avoid biasing the estimation of the earlier reconstructions (hence they are not shown in Figure 6.10), implicitly assuming that the “divergence” was a uniquely recent phenomenon, as has also been argued by Cook et al. (2004a). Others, however, argue for a breakdown in the assumed linear tree growth response to continued warming, invoking a possible threshold exceedance beyond which moisture stress now limits further growth (D’Arrigo et al., 2004). If true, this would imply a similar limit on the potential to reconstruct possible warm periods in earlier times at such sites. At this time there is no consensus on these issues (for further references see NRC, 2006) and the possibility of investigating them further is restricted by the lack of recent tree ring data at most of the sites from which tree ring data discussed in this chapter were acquired.


63 Comments

  1. Dave Dardinger
    Posted May 4, 2007 at 8:25 AM | Permalink

    Well, it looks to me that the last part of that paragraph, beginning with “Others,” is essentially your position. Too bad you’re “He who cannot be named,” but it seems to be talking about the inverted U growth pattern and the “bring the proxies up to date” without actually getting into the details.

    More importantly, note the phrase “there is no concensus.” Considering that in one sense these issues are at the core of the entire paleoclimate reconstruction picture, it looks to me like you’re having a major effect on how scientists look at the possibility of that effort, though they’re not yet willing to face it head-on. Congratulations!

  2. bender
    Posted May 4, 2007 at 8:44 AM | Permalink

    There is no consensus on these issues. Divergence is a real, big problem.

  3. Michael Jankowski
    Posted May 4, 2007 at 8:45 AM | Permalink

    and the possibility of investigating them further is restricted by the lack of recent tree ring data at most of the sites from which tree ring data discussed in this chapter were acquired.

    The “possibility of investigating them further” is wide-open. “Bring the proxies up to date.”

  4. John A
    Posted May 4, 2007 at 8:48 AM | Permalink

    Of course, if there’s a divergence between solar cycle length and temperature, then the scales fall away from the eyes. Its amazing.

  5. bender
    Posted May 4, 2007 at 8:48 AM | Permalink

    DD: the “others” are likely D’Arrigo & Wilson, referred to in the opening paragraph.

  6. Steve McIntyre
    Posted May 4, 2007 at 8:49 AM | Permalink

    bender, it’s amazing how, without achieving consensus on such a fundamental issue, they can out of the other side of their mouth claim 2-sigma confidence intervals based on calibration period residuals. These guys are a joke.

  7. bender
    Posted May 4, 2007 at 9:05 AM | Permalink

    “There has been expressed concern that the divergence problem challenges the uniformitarianism assumption in tree rings (e.g., National Research Council 2006).

    From D’Arrigo & Wilson (2007).

    However, if the divergence is in fact anthropogenic in origin then it will only directly impact reconstructions within the past few decades. Some evidence suggests that this is the case, and that the divergence is limited, and unique to this recent period (Briffa et al. 1998a, Cook et al. 2004a).”

    The Briffa line.

    Nevertheless, there are still significant implications for the development of dendroclimatic reconstructions, as we have noted in this paper. For example, reconstructions based on northern tree-ring data impacted by divergence cannot be used to directly compare past natural warm periods (notably, the MWP) with recent 20th century warming, making it more difficult to state unequivocally that the recent warming is unprecedented.”

    The CA line.

    Now … what about those AD1000 megadroughts? Shhh. Be very quiet.

  8. Mark T.
    Posted May 4, 2007 at 9:17 AM | Permalink

    There has been expressed concern that the divergence problem challenges the uniformitarianism assumption in tree rings (e.g., National Research Council 2006).

    There is not concern that divergence challenges the uniformitarianism assumption at all. It challenges the _stationarity_ claim of the sources as well as the mixtures, particularly given that the latter may hinge on correlations between the sources. Why don’t dendros understand the difference? Uniformitarianism != stationarity.

    Mark

  9. jae
    Posted May 4, 2007 at 9:23 AM | Permalink

    At least they are kinda sorta coming to grips with the issue. Progress, I’d say.

  10. Steve Sadlov
    Posted May 4, 2007 at 9:23 AM | Permalink

    RE” #3 – Yes … bring the proxies up to date!

  11. bender
    Posted May 4, 2007 at 9:25 AM | Permalink

    … the other side of their mouth …

    The right hand side of the mouth doesn’t know what the left hand side is saying. ;)

    That’s your next sign that the team is splitting. (The last was VS&Z trying to scoop as much credit as possible in breaking the HS. Before that you had Hughes in his interview distancing himself from the statistics of MBH.)

  12. Douglas Hoyt
    Posted May 4, 2007 at 9:45 AM | Permalink

    Perhaps if they bothered to remove the UHI effects from the temperature measurements, then the divergence problem would disappear. The solar cycle/temperature divergence would disappear as well.

    Then, too, the claimed GHG warming signal would also disappear, so we can’t have that.

  13. Mark T.
    Posted May 4, 2007 at 9:49 AM | Permalink

    Or maybe they should simply re-write their hypothesis that tree-rings are proxies for temperature. If there’s a divergence now, how can we know if, or when, any similar divergences occurred in the past? This alone refutes any reconstruction using just about any tree-ring, foxtails or otherwise.

    Mark

  14. bender
    Posted May 4, 2007 at 9:50 AM | Permalink

    Except there are positive and negative trenders. i.e. The divergence is not between trees and temperature, it’s between trees and trees.

  15. Craig Loehle
    Posted May 4, 2007 at 9:51 AM | Permalink

    It is stated:

    “There has been expressed concern that the divergence problem challenges the uniformitarianism assumption in tree rings (e.g., National Research Council 2006).”

    It is not “uniformitarianism” and is only partially “stationarity”. The biggest problems are 1) that if the tree growth response is parabolic and you have a limited temperature range in the calibration period then you can’t estimate the full model, so you are extrapolating beyond the range of the data and 2) if precipitation and temperature interact, you need to estimate both models and have data from the past on precip to factor it out, which no one has.

  16. bender
    Posted May 4, 2007 at 9:55 AM | Permalink

    Craig’s got it. Many dendros are in denial over the strength of this possibility. But you can see it in the guarded language about linear univariate models as useful approximations.

  17. Terry
    Posted May 4, 2007 at 10:14 AM | Permalink

    This is even worse than the dendros are letting on.

    They should really be saying that the reconstructions fail out-of-sample verification tests. This brings up the issue of whether the original results were cherry-picked.

  18. Mark T.
    Posted May 4, 2007 at 10:22 AM | Permalink

    It is not “uniformitarianism” and is only partially “stationarity”. The biggest problems are 1) that if the tree growth response is parabolic and you have a limited temperature range in the calibration period then you can’t estimate the full model, so you are extrapolating beyond the range of the data and 2) if precipitation and temperature interact, you need to estimate both models and have data from the past on precip to factor it out, which no one has.

    While I agree with this, I think the point I was making was a little different than what may have been inferred. Point 1, as you’ve noted, is that temperature is not linear, we know that. Point 2 is that the “sources” are actually dependent upon one another. Temperature and CO2 are both sources for plant growth (in general), but they are correlated not only by hypothesis, but historical record (cause and effect relationship is immaterial for my quickie discussion). I.e., the upside-down quadratic moves around, and may even warp, based on other growth factors, call them limiters or enhancers if you will. That said, the “stationarity” shows up when one factor limits or enhances another, so what you get, while not necessarily “non-stationary behavior” in a purely statistical sense, but a change in the limits on one factor or another. Which, I think, is in complete agreement with what you’ve said.

    Mark

  19. Mark T.
    Posted May 4, 2007 at 10:26 AM | Permalink

    They should really be saying that the reconstructions fail out-of-sample verification tests.

    Which could easily be the result of the issues both I and Craig raised.

    This brings up the issue of whether the original results were cherry-picked.

    They were. Search the site for D’Arrigo’s “you can’t make cherry pie without picking cherries” comments. The humor in that statement alone had me crying for days… nay, I’m still crying.

    Mark

  20. Demesure
    Posted May 4, 2007 at 10:27 AM | Permalink

    From D’Arrigo & Wilson (2007).
    However, if the divergence is in fact anthropogenic in origin then it will only directly impact reconstructions within the past few decades. Some evidence suggests that this is the case, and that the divergence is limited, and unique to this recent period (Briffa et al. 1998a, Cook et al. 2004a).”

    What they are saying is the calibration period may be correct but the proxied period is fine with a high degree of confidence.
    What a joke! A joke so obfuscatingly phrased you can’t even register it in your “bestof climate science”.

  21. Bill F
    Posted May 4, 2007 at 10:36 AM | Permalink

    What struck me in the quote Steve posted is that the phrase “temperature reconstructions” was only used once. Everywhere else, they talk about “climate reconstructions”. I don’t know how you can represent “climate” with a two dimensional line of time versus temperature. It appears to me that they are dancing around the temperature proxy vs precipitation proxy issue by calling tree rings a “climate proxy”. But that is pretty disingenuous when they then draw line of temperature vs time to represent the “climate” signal in the tree rings. They describe all the various variables that could cause trees to grow at different rates than temperature would suggest except precipitation, which is lumped in with temperature as part of “climate”. Pretty careful dancing if you ask me…

  22. Tom C2
    Posted May 4, 2007 at 10:58 AM | Permalink

    There has been expressed concern that the divergence problem challenges the uniformitarianism assumption in tree rings (e.g., National Research Council 2006

    How can one argue fine points of mathematical modelling with persons who are so confused about basic concepts like this?

  23. Mark T.
    Posted May 4, 2007 at 11:07 AM | Permalink

    You can’t. Go and try to post at realclimate.

    Mark

  24. Steve McIntyre
    Posted May 4, 2007 at 11:27 AM | Permalink

    #22. Keith Briffa, the author of this section of AR4, also said the following (which D’Arrigo held out as an explanation at the NAS Panel hearings):

    In the absence of a substantiated explanation for the decline, we make the assumption that it is likely to be a response to some kind of recent anthropogenic forcing. On the basis of this assumption, the pre-twentieth century part of the reconstructions can be considered to be free from similar events and thus accurately represent past temperature variability.

    I’ve called this Cargo Cult in the past – Briffa’s “uniformitarianism” is the same as Cargo Cult Uniformitarianism. And to think that IPCC says that 2500 scientists supposedly agreed,

  25. Roger Dueck
    Posted May 4, 2007 at 11:43 AM | Permalink

    the divergence is limited, and unique to this recent period

    Theoretically, the best temperature data IS the instrumental record of past few decades, which the HSTeam goes to great length to adjust earlier records to. The divergence problem also exists between land surface and SST as well as hemisheric. The above quote is laughable, if the consequences of it’s acceptance were not so dire.

  26. Steve McIntyre
    Posted May 4, 2007 at 12:18 PM | Permalink

    Several people have sent me copies of this article saving me a trip downtown to the university. Thanks. I’m making one quick point now which is not intended as a “review”. They have a typical climate science bit of check-kiting in which they use a new divergence-light reconstruction from Wilson et al (in prep). I hate this sort of check-kiting.

    The new reconstruction (Wilson et al. in prep) consists mainly of unpublished results and, because there’s a lot of Team involvement here, there is negligible archiving other than what happened to be archived by non-Team players. Here’s a list:

    COL = Colorado (Salzer and Kipfmueller (2005), – unarchived
    IDA = Idaho (Biondi et al. (1999) – archived last year (Biondi was on the NAS Panel and the non-archiving was a little embarrassing)
    MBC = British Columbia (Wilson and Luckman (2002) – unarchived and Rob Wilson tells us it will be a frosty day before it’s ever archived
    YUS = Yukon (south – Youngblut and Luckman (in press) – unarchived. No Luckman measurement data has ever been archived.
    YUN = Yukon (north – Szeicz and MacDonald (1994), – archived
    WRA = Wrangells (Davi et al. (2003) – I need to check. I think that it’s archived.
    NQU = Northern Quebec (Payette (in press); Wilson et al. in prep) – unarchived
    ALP = Alpine region (Büntgen et al. 2006, in press-a), – unarchived. No Esper data has ever been archived,
    TAT = Tatra Mountains (Büntgen et al. in press-b), – unarchived. NO Esper data has ever been archived.
    SCA = Northern Scandinavia (Kirchhefer (2001, Wilson et al. in prep), – unarchived
    WSI = Western Siberia (Wilson et al. in prep), – unarchived
    MON = Mongolia (D’Arrigo et al. 2000), – archived in part
    KYR = Kirgistan (Wilson et al. in prep), – unarchived
    TSH = Tien Shan (Esper et al. 2003b), – unarchived
    NEP = Nepal (Cook et al. 2003); – archived promptly. Cook’s the best actor in this crowd.

    Rob provides a location map and list of composite chronologies utilized by D’Arrigo et al. (2006). He mentions the use of:
    POL = Polar Urals,
    But we know that he really used Yamal in his long reconstruction and that there was a “good” reason for using this rather than the Polar Urals series which had an elevate MWP.

    Rob says that “There are no common tree-ring data between both these studies.” The Mongolia series looks like its common to the two series.

    The long reconstruction in D’Arrigo et al 2006, the one which covered the MWP, had only 6 series:
    CSTA = Coastal Alaska,
    ICE = Icefields,
    TORN = Tornetraesk,
    POL = Polar Urals [actually Yamal]
    TAY = Taymir,
    MON = Mongolia.

    This roster was virtually identical to Briffa 2000 and the medieval-modern relation is not robust to the change from Yamal to Polar Urals. It will be interesting to see what the divergence-light Wilson et al 2007 does. My recollection from AGU is that this is a short reconstruction. So it won’t resolve divergence problems in reconstructions that include the MWP. But we shall see.

  27. Jon
    Posted May 4, 2007 at 12:32 PM | Permalink

    Steve,

    In particular Briffa’s position can be criticized as a form of Begging the Question. I think we can all agree that this research is being performed to test the hypothesis of unprecedented warming. Consequently, arguing that the results are consistent when we assume the hypothesis is true means that this study is not fit for purpose.

    Assuming the hypothesis was true is often the precursor to designing any study. i.e., if the hypothesis is true, I expect to observe xyx. Do I observe xyz? It is at that point that the hypothesis can no longer be taken as given, but that’s precisely what Briffa seems to be doing.

    This is certainly an epistemological problem for the field.

  28. Don Keiller
    Posted May 4, 2007 at 3:07 PM | Permalink

    Oh, how they wriggle! “Global Dimming”- show me the evidence that C3 plants at high latitudes are light limited! “Drought” yet it is a proven fact that elevated CO2 increases water use efficiency. And I bet my last penny that this effect is more pronounced at high altitude, where CO2 is a limiting factor. Finally haven’t D’arrigo et al ever read any of the literature that shows that Global productivity has increased over the last 20-30 years? It is as plain as a pikestaff- the only anthroprogenic infuence on the “divergence problem” is the urban heat island effect and the Team’s constant rewriting of temperature history- ably named and shamed by Steve.

  29. Jon
    Posted May 4, 2007 at 9:05 PM | Permalink

    “Global Dimming”- show me the evidence that C3 plants at high latitudes are light limited!

    Don,

    This remark isn’t fair. “Global Dimming” as alleged has nothing to do with whether or not plants are light limited. Rather, it is the idea that sulfur in the air–arising from diesel combustion emissions–has increased the albedo of the earth. Consequently, these emissions have caused some amount of cooling in addition to their GHG warming.

    The US clean-air act and other measures in other countries have stemmed sulfur emissions. Consequently, this countervailing effect is diminishing. The argument is then: historical measurements of CO2 and temperature from which one might estimate the post-amplification constant of proportionality linking the two are distorted low.

    This is, though, a bit of rubbish. No one would think of estimating the climate sensitivity in such a naive fashion, and indeed GCM supposedly have modeling for this too.

  30. Rob Wilson
    Posted May 6, 2007 at 2:36 AM | Permalink

    Hi
    Wilson et al. (2007) – A matter of divergence: tracking recent warming at hemispheric scales using tree-ring data. JGR – Atmospheres – is now in press. The reconstruction only covers the period from 1750-2000. All chronologies were submitted along with the paper when it was submitted last year and will be available through the AGU website when the paper is published as auxiliary material. The data will also be archived at the NCDC.

    Don’t flatter yourselves that this paper is a response to discussions on this Blog. I have bee wanting to write this paper since 2003.

    regards
    Rob

  31. Steve McIntyre
    Posted May 6, 2007 at 7:54 AM | Permalink

    #30. Rob, no need to be petulant. We speak kindly of you here while reserving the right to inquire into multivariate methodology. I’ve reported on discussions of divergence by Wilmking, Pisaric and even D’Arrigo without the remotest hint that they were responding to issues raised at climateaudit. Quite the opposite. In respect to divergence, I’ve simply been reporting on issues latent in the specialist literature and connecting that to the handling of the problem in the multiproxy studies. Since you’re a thoughtful specialist, it is unsurprising that you should also be concerned about the matter and develop your own view on the matter. I for one welcome the fact that you’re thinking about the issues, although I may well not agree with either the methods or conclusions of the article, which I have in hand.

    I’m also glad that you will be archiving the data concurrent with publication. Obviously I think that this is good practice and I welcome your doing so, regardless of how you arrived at this decision. In your shoes, if I were writing to CA to advise that the data would be archived concurrent with publication, I’d have added a short and fairly cheerful sentence that acknowledged CA’s interest in the practice, that you’ve come to the conclusion that it’s a good idea and are happy to agree with us on this matter. I like emphasizing agreements whenever possible. Before writing anything critical about the NAS Report, I wrote a polite thank you letter to them. But I might not have done so when I was your age.

  32. bender
    Posted May 6, 2007 at 8:16 AM | Permalink

    Your objective approach to dendroclimatology is appreciated, Dr. Wilson. If others archived their data and methods the way you did, policy-makers would be far better off than they are.

  33. Don Keiller
    Posted May 6, 2007 at 8:36 AM | Permalink

    Jon, re#30. I’m fully aware that “Global Dimming” is supposed to be responsible for the decline in world temperatures over the 40’s to 70’s. What I was specifically referring to in my comment was that over the last 50 + years, tree rings have apparently shown a reluctance to agree with the instrumentally measured global temperature rise. Thus the trees are growing more slowly than expected. Some researchers have speculated that the growth responses to such increased temperatures have been confounded by reduced sunlight (global dimming) or drought. Bizarrely enougth many of the same researchers alos believe that the “preanthroprogenic” era is not confounded by other environmental influences. The point I am trying to make is that some people want to have their cake and eat it.

  34. bender
    Posted May 6, 2007 at 8:42 AM | Permalink

    Re #33
    No. The tree growth patterns are divergent. Some grwing faster, some growing slower.

  35. bender
    Posted May 6, 2007 at 9:43 AM | Permalink

    the trees are growing more slowly than expected

    This is the bit which #34 was responding to. SOME are growing more slowly than expected, others MORE quickly – and even the experts don’t know why.

    some people want to have their cake and eat it

    The database for AAGW double-standards now stands at 22 entries. Seems LOTS of people want cake and full bellies.

  36. Don Keiller
    Posted May 6, 2007 at 10:44 AM | Permalink

    Jon re#29, I refer you to Steve’s latest post “RMS and SO2 emmisions”.
    In any case, I am fully aware what “Global Dimming” means. My suggestion, in relation to the “divergence problem” is that “global dimming” has been suggested (by Rob Wilson, I believe- correct me if I am wrong) as a possible cause. My experience as a plant physiologist tells me that this is extremely unlikely, for a number of reasons. Not least that at low temperatures- supposedly the main limiting factor for tree ring growth as for as Dendos are concerned, light saturation of photosynthesis occurs at very low levels. To suggest that the divergence problem is something that results from the confounding of other environmental factors, exclusive to the “post-anthroprogenic” era, whilst excluding the role of elevated CO2 and arguing that temperature correlations were not confounded by environmental factors in the pre anthroprogenic era, is disingenuous to say the least.

  37. Willis Eschenbach
    Posted May 6, 2007 at 2:42 PM | Permalink

    bender, you say to Rob Wilson:

    Your objective approach to dendroclimatology is appreciated, Dr. Wilson. If others archived their data and methods the way you did, policy-makers would be far better off than they are.

    I agree with you about Dr. Wilson’s objective approach to dendroclimatology, it is a good thing. I was puzzled, though, by your comment on his archiving practices.

    Does your comment mean that Dr. Wilson has actually archived the data from “Cycles and shifts: 1,300 years of multi-decadal temperature variability in the Gulf of Alaska”, and if so, where? My thanks and appreciation to him if he has done so.

    However, if that data is still not archived, I would not suggest that others archive their data and methods the way Dr. Wilson does …

    w.

  38. fFreddy
    Posted May 6, 2007 at 4:43 PM | Permalink

    Re #30, Rob Wilson

    I have bee wanting to write this paper since 2003.

    What was preventing you from doing so ? And why have you been able to do it now ?

  39. Willis Eschenbach
    Posted May 6, 2007 at 7:32 PM | Permalink

    There is an interesting discussion of Rob’s paper at Climate Science.

    One of the more insightful comments is that of Doug Hoyt, who wrote:

    The surface temperatures diverge from temperatures derived from tree rings.

    The surface temperatures diverge from the lower tropospshere balloon observations.

    The surface temperatures diverge from the UAH MSU observations.

    The surface temperatures diverge from the solar cycle length estimation of solar forcing.

    In all cases, the surface temperatures diverge in giving a greater warming trend.

    Perhaps it would be logical to suspect the surface temperature record. Of course, if the surface record is in error on the high side, then the argument for large GHG warming goes away, so I don’t expect any critical examination of the surface temperature record.

    Makes sense to me …

    w.

  40. David Smith
    Posted May 6, 2007 at 8:49 PM | Permalink

    Is the divergence problem true for both ring width and for ring density?

    Perhaps rings change (expand) their width in the first few decades after formation, as the microstructure undergoes slow changes. After some period the changes end. If that is true, then recent rings could understate the actual growth conditions.

    Another dendro question: when a site is sampled, do the scientists report how many samples were taken and discarded? I can imagine someone taking 50 tree samples, examine them to see which fit a preconceived notion and discard 30 or so which don’t fit the notion. This could all occur subconsciously.

  41. Steve McIntyre
    Posted May 6, 2007 at 9:37 PM | Permalink

    Both – see the Graphic in Briffa 1998 that I put up about 15 months ago.

  42. Geoff Sherrington
    Posted May 7, 2007 at 5:34 AM | Permalink

    I’m a graduate chemist now retired. I have made several trips to west China collecting some extremely rare plants to grow for safe haven in Australia. The idea is to legally import one scion of a tree and then over the years, to multiply the numbers. This multiplication is sometimes done from seed of the first plant, grown under conditions that are reasonably uniform (a few hundred plants adjacent to each other in pots in a greenhouse – pardon that word). The point is that even with this environmental uniformity, the growth of individuals is far from uniform. Some start off slow then shoot along, others vice versa. Some become super trees. Their genes are not apparently uniform one to the next, as later flowering shows consistent differences in what is supposed to be a single species.

    Although I have done no tree ring analysis, I would suspect it would be hard to examine tree rings and find that all plants were derived from the same population and grown under similar conditions. This is on a microscale with young plants. Some are camellia spp which are known to reach 400 years age or more.

    I doubt the assumption that variation referred to above in bristlecones or whatever can be correlated over large distances to make a global pattern. To me, the genetics of the individual tree enter the equation as a confounding variable. Has anyone ever done a test on many trees from a solid stand of old timber to see how dendrology correlated over a scale of kms or less?

    The more I read about this debate the more shocked I am by the refusal of authors to release raw data and by the absence of short-term, low cost side experiments which start to put envelopes around total variabliity. The quote about cherry-picking is most appropriate.

  43. MarkW
    Posted May 7, 2007 at 8:05 AM | Permalink

    Geoff,

    Idle question. How does one study the chemistry of a graduate? Do they distill easily? ;*)

  44. Rob Wilson
    Posted May 7, 2007 at 9:19 AM | Permalink

    w.r.t.

    Is the divergence problem true for both ring width and for ring density?

    There is no simple black and white answer.

    To generalise, the large scale RW based reconstructions (Briffa00, Esper01 and D’Arrigo06) start diverging in the mid 1980s while the large scale MXD based reconstructions (e.g. Briffa01) start diverging around the 1960s.

    However, if one looks at individual chronologies/local reconstructions, then the time of divergence is quite different between sites.

    This issue can really be only addressed at the local site scale.

    Also, divergence does not happen at all locations. Hopefully the review paper highlights this issue.

  45. Ken Fritsch
    Posted May 7, 2007 at 9:44 AM | Permalink

    Re: #30

    Don’t flatter yourselves that this paper is a response to discussions on this Blog. I have bee wanting to write this paper since 2003.

    You da man, Rob when it comes to dendro thinking on the divergence problem. While I am not holding my breath on finding any definitive answers, I look forward to your publication(s) as much for the insights they give into the current thinking processes for dendros as for any revelations from new data or new manipulations of new and/or old data.

  46. Dave Dardinger
    Posted May 7, 2007 at 11:04 AM | Permalink

    re #43:

    How does one study the chemistry of a graduate?

    Obviously you use graduate cylinders.

  47. Earle Williams
    Posted May 7, 2007 at 11:48 AM | Permalink

    Re #30

    Dr. Rob Wilson,

    My comments not only apply to your posting in #30 of this thread but to all comments by researchers and other professionals. This and other blogs are read by a huge multitude that never makes comments. In reading the banter it may be easy to feel you are engaged in a dialog with just a few individuals that are criticizing your work, but the fact remains that the entire exchange is available for all the world to see. Why should this matter to you? This matters because the snide and unprofessional comments do nothing to support or refute a scientific argument. However they do reveal to me and everyone else that you are unable to gracefully accept criticisms of your work. Other inferences will be drawn as well. For posters using a pseudonym they can enjoy the freedom to cast barbs without damaging their professional and personal reputations. Those posting under their true names put these reputations at risk when they choose to engage in unprofessional web dialog.

    You may not care what I think. You don’t know me and I don’t know you but I’ve already drawn unfavorable conclusions about you as a person. Is this the effect you want to have on everyone who reads these posts? Good heavens man, give some thought to your professional demeanor. If you can’t resist the cheap shots then don’t post.

    This comments is a bit harsh but it needs saying. If your work speaks for iteslf you needn’t even post here. If you do wish to defend your work know full well that the work and your reputation will be judged in part by your professionalism in its defense.

    Kind regards,
    Earle Williams

  48. Rob Wilson
    Posted May 8, 2007 at 1:38 AM | Permalink

    #47
    I am sure Steve will agree that the odd sarcastic comment is good for the soul.
    It comes from frustration I guess.

    I am sorry you have such a bad opinion of me.

    I am not here to make friends, but rather to ensure a fair and balanced critique of dendro methods.

  49. Sara Chan
    Posted May 8, 2007 at 1:49 AM | Permalink

    Re #47, “I’ve already drawn unfavorable conclusions about you as a person”. Wow! What a man.

    I agree with #48: comments should be focused on the research; your opinions about the person are as irrelevant as they are valueless.

  50. Steve McIntyre
    Posted May 8, 2007 at 4:28 AM | Permalink

    #48. Rob said:

    I am not here to make friends, but rather to ensure a fair and balanced critique of dendro methods.

    Excellent perspective. You know something, Rob – so am I. (although I’m interested in more generally in multiproxy methods.)

  51. bernie
    Posted May 8, 2007 at 5:45 AM | Permalink

    There is a long tradition in British fiction of having academics as victims or as perpetrators of ghastly deeds: Seems that murder mystery writers have a keen sense of what lurks in souls of some academics.

    I for one greatly appreciate Dr. Wilson’s willingness to engage – with or without the sarcasm. I do think it would be interesting to hear from dendro specialists on Geoff’s (#42 above) comments. At a minimum it seems to me that his observations would point to a much larger sampling of trees per site and, even more importantly, research to ensure that averaging across trees actually serves to reduces measurement errors rather than compounding measurement errors by masking very different responses to temperature changes.

  52. bender
    Posted May 8, 2007 at 8:55 AM | Permalink

    Dendros are well aware of the issue of between-tree variability. Whether that arises mostly from genotype (G) or environment (E), or the GxE interaction is anybody’s guess. Not alot of thought is given to studying genotype/phenotype issues, as the focus is usually on the E (=site). The approach taken by dendros is replicate to reduce the individual tree as a source of variability. If there’s one thing they do right, it’s replicate. Must be in their genes.

  53. Earle Williams
    Posted May 8, 2007 at 12:28 PM | Permalink

    Re #48

    Dr. Rob Wilson,

    Sarcasm, done well, can be a pleasure to read. Snide comments on the other hand only detract from the message. Neither ensures a fair and balanced critique of dendro methods. These are just some thoughts of mine which you’ve no doubt already considered. Thanks for the response.

    Re #49

    Sara Chan,

    Thank you for the italics and the healthy dose of sarcasm. However it seems the message I was trying to convey is not the one you interpreted. Perhaps this is a case of not seeing the forest for the trees?

    Kind Regards,
    Earle Williams

  54. bernie
    Posted May 8, 2007 at 12:49 PM | Permalink

    #52
    bender:
    I am not sure I get how replication per se solves the problem versuses hides the problem if the source of variability is not random?

  55. MarkR
    Posted May 8, 2007 at 12:58 PM | Permalink

    From D’Arrigo, Wilson via Pielke, apparently…

    “A recent analysis by Cook et al. (2004a) suggests that the divergence is restricted to the recent period and is unique over the past thousand years. It is thus likely to be anthropogenic in origin.”

    Don’t any of these Dendros realize that their chosen proxies only correlate over the recent period because they’ve been cherry picked for the calibration period. Even then, the cherry picked proxies don’t correlate prior to the calibration periods they are chosen for.

  56. Willis Eschenbach
    Posted May 8, 2007 at 1:10 PM | Permalink

    I had a different question about the quote:

    “A recent analysis by Cook et al. (2004a) suggests that the divergence is restricted to the recent period and is unique over the past thousand years. It is thus likely to be anthropogenic in origin.”

    My question is, how would one detect divergence in, say, the year 1156? …

    w.

  57. Mark T.
    Posted May 8, 2007 at 1:26 PM | Permalink

    Using the reconstructions, of course. Bristlecones are welcome to apply for the job.

    Mark

  58. MarkR
    Posted May 8, 2007 at 1:26 PM | Permalink

    #56 Willis. Perhaps you could put up one of your excellent graphs, or tables of inter correlation showing the divergence between the proxies, while Mr Wilson is about?

  59. Steve McIntyre
    Posted May 8, 2007 at 1:50 PM | Permalink

    #55, 56. I posted last year on Cook et al 2004, describing it as more cargo cult science in Briffa style See post 586, also 570 and 529 are relevant.

  60. Steve Sadlov
    Posted May 8, 2007 at 3:20 PM | Permalink

    Some more observations vis a vis my coast live (evergreen) oaks and douglas fir. The oaks have already hit their growth maximum for the year and are rapidly slowing toward their (expected, annual) summer dormancy. The dormancy corresponds with our annual low-no precipitation period. They will be fully asleep by the end of the month and will not awaken again until sometime late in the fall or early in the winter, assuming of course we actually get rains then. The doug fir are still growing at a moderate rate. They are a bit more drought hardy than the oaks (although, I would add, it’s a wet-dry climate seasonal hardiness – I believe that during a protracted drought – e.g. two or more years of statistically significant below normal annualized precip – the oaks are actually better able to survive). They will go dormant prior to the 4th of July. How does all this stack up with the belief system of the Dendro Truth Squad?

  61. Steve Sadlov
    Posted May 8, 2007 at 3:27 PM | Permalink

    RE: #60 – A bit of clarication regarding coast live oak dormancy. There is not an overall leaf drop. The leaves are waxy and thick with short spines protroding at the ends of the lobes (like holly). There will be some slight drop of selected leaves during very hot periods. There will continue to be a very low, maintenance level of photosynthesis, not complete stoppage. But there will be no growth – no addition of wood mass, if anything, slight shinkage due to moisture loss. When the seasonal dry ends, the response is not immediate – it take a few weeks for the trees to “wake up.”

  62. Steve McIntyre
    Posted May 8, 2007 at 10:03 PM | Permalink

    #59. I re-read my post on Cook et al 2004. There’s nothing in D’Arrigo et al that addresses these issues.

  63. BobFJ
    Posted Jul 29, 2007 at 7:21 PM | Permalink

    The following quotes from Increased snow is shortening tree-growing season in subarctic Siberia – Science Daily, are exquisite!

    …”The recent weaker correlation between tree growth and temperature clearly affects the reliability of our reconstructions of the past. Actually, it means past climate reconstructions (before the 1960s) are better than we thought they were. And, as a result of this, it means that we underestimated the differences between the present century and past centuries,” Hughes said…
    …The contrast between this century and previous centuries may be greater than thought, Hughes now suggests, because “our calibration is contaminated partly by this recent weaker correlation.”…

    Did Hughes actually say that?
    Did they really calibrate their proxies in the period of divergence?
    Where is there data for snowfall at individual sites, say at 1200 AD?
    etc!

Follow

Get every new post delivered to your Inbox.

Join 3,249 other followers

%d bloggers like this: