Tuesday, June 21, 2011

The Tiljander Data Series Appear Again, This Time in a Sea-Level Study

At RealClimate.org, Stefan Rahmstorf has written "2000 Years of Sea Level" about a study published on June 20, 2011 in PNAS. Andrew Kemp and co-authors BP Horton, JP Donnelly, ME Mann, M Vermeer, and S Rahmstorf reconstruct sea levels from 500 AD to the present, and relate these levels to the temperatures of the past, using a multi-proxy reconstruction that was first presented in Mann et al. (PNAS, 2008). (The Kemp11 PDF can be downloaded at the RC post.)

It turns out that the chosen temperature recon is heavily dependent on the four three uncalibratable Tiljander data series. This reliance grows stronger as one goes back in time, and shorter (younger) records "drop out."

I tried to leave a remark on this subject at RealClimate.org. Apparently, that site is set to automatically fail any comment tagged with my user name, email, or IP address. Here is the local copy of what I submitted (21 Jun 3:50 PM EDT) --
I was surprised at the provenance of the paleotemperature reconstruction that was used in Kemp et al's Fig. 2A and Fig. 4A. According to Fig. 2A's legend, it is "Composite EIV global land plus ocean global temperature reconstruction, smoothed with a 30-year LOESS low-pass filter". The reference is Mann et al. (2008). In that paper's S.I., the unsmoothed version is in panel F of Fig S6, as the black line labelled "Composite (with uncertainties)".

This is one of the multiproxy reconstructions that employed the four (actually three) uncalibratable Tiljander lakebed sediment data series.

According to Gavin Schmidt, "...it's worth pointing out that validation for the no-dendro/no-Tilj is quite sensitive to the required significance, for EIV NH Land+Ocean it goes back to 1500 for 95%, but 1300 for 94% and 1100 AD for 90%" (link). Further remarks on this issue as Responses to other RC comments here (see numbers 525, 529, and 531).

The incorrect inclusion of Tiljander could well make this EIV reconstruction progressively worse, as one goes from 1500 AD back to 500 AD. This might explain the increasing divergence between the temperature recon and the sea-level recon, as one travels back from 1100 AD to the beginning of the recons at 500 AD. This pattern is shown in Kemp11's S.I. Figs. S3, S4, and S5.

Did any of the peer reviewers comment on this issue, or request that you use a no-Tiljander temperature reconstruction?
I left a more detailed version at Collide-a-scape as #128 at Climate Critics That Won't Muzzle Themselves (June 21st, 2011 at 1:35 pm; lightly edited) --
...Yesterday, Kemp et al. 2011 was published in PNAS, relating sea-level variation to climate over the past 1,500 years (UPenn press release). Among the authors is Prof. Mann. Figs. 2A and 4A are "Composite EIV global land plus ocean global temperature reconstruction, smoothed with a 30-year LOESS low-pass filter". This is one of the multiproxy reconstructions in Mann et al. (2008, PNAS). The unsmoothed tracing appears as the black line labelled "Composite (with uncertainties)" in panel F of Fig. S6 of the "Supporting Information" supplement to Mann08 (downloadable from pnas.org).

This is one of the Mann08 reconstructions that made use of the four (actually three) uncalibratable Tiljander data series.

As scientist/blogger Gavin Schmidt has indicated, the early years of the EIV Global reconstruction rely heavily on Tiljander to pass its "validation" test: "...it's worth pointing out that validation for the no-dendro/no-Tilj is quite sensitive to the required significance, for EIV NH Land+Ocean it goes back to 1500 for 95%, but 1300 for 94% and 1100 AD for 90%" (link). Also see RealClimate here (Gavin's responses to comments 525, 529, and 531).

The dependence of the first two-thirds of the EIV recon on the inclusion of Tiljander's data series isn't mentioned in the text of Kemp11. Nor is it discussed in the SI (link), although it is an obvious and trivial explanation for the pre-1100 divergence noted in the SI's Figures S3, S4, and S5.

Peer review appears to have been missing in action on this glaring shortcoming in Kemp11's methodology.

More than anything, I am surprised by this zombie-like re-appearance of the Tiljander data series -- nearly three years after the eruption of the controversy over their misuse as temperature proxies!
Interestingly, M. Vermeer discussed the Tiljander proxies data series with me and others in the comments of Arthur Smith's posts Where's the fraud? and Michael Mann's errors. So at least one of Kemp11's authors was familiar with the (claimed) severe shortcomings of Tiljander-based reconstructions. And then... they went ahead and used just that type of recon.

24 comments:

  1. AMac, are you going to submit a comment to PNAS?

    Layman Lurker

    ReplyDelete
  2. Amac,

    For clarity sake, can we drill this down to very basic complaints, or whatever you want to call them.

    1. It is your understanding that part of the Tiljander proxy can not be calibrated to current temperature. You do not think that Mann or anyone else has answered this complaint.

    2. Gavin's answer, that Mann replaced a graph to show that the Tiljander proxy doesn't matter shows that they know there is an issue. If I am right about that, can you please elaborate on whether that answer is satisfactory, as far as the accuracy of the Tiljander-less proxy (I also understand there are other trees no included there).

    3. So if one and two are correct, are your complaints about scientific honesty, and not whether the conclusions from Mann 08 are correct? And this transfers to Kemp because the Tiljander series is still included?

    This is my understanding. Can't calibrate Tiljander. Why is it included?

    ReplyDelete
  3. IIRC not only is the Tiljander data inverted, but hopelessly polluted by some bridge construction that stirred up river sediment.

    The spikes in the last part of the time series might be related.

    ReplyDelete
  4. Grypo,

    What a pleasure to read your comment. Its clarity invites reflection, and a careful answer.

    1. Yes. The authors of Tiljander03 cautioned that local activities contaminated the data series beginning around 1720, and that these non-climate-related signals grew larger through the 18th, 19th, and 20th centuries.

    It's helpful to look at graphs of the raw data. I plotted them out in the post The Tiljander Data Series: Data and Graphs. You can see the uptick in the early part of the 18th century. Then things really start jumping around 1940. Also look at the mean and standard deviation of these series for the 20th century.

    I don't know why Tiljander03 chose "1720" for their caution. Seems to me from looking at the traces that local effects start to dominate in the late 19th/early 20th century. But I don't know the area's history, and Tiljander et al do.

    Recall that calibration to temperature refers to correlating proxies to the CRUTEM3v calculated temperature anomaly for the 5 degree by 5 degree gridbox that encompasses Lake Korttajarvi, for the years 1850 to 1995. Direct calibration is an essential step in the procedures of Mann08. Kaufman09 used some of the Tiljander data as atemperature proxy, employing a more complex indirect procedure. There is no mention of such a method in Mann08's text or S.I.

    2. As to whether Gavin's answer is satisfactory: My quick response is simple. "No." The longer response is that I'd first have to read "Gavin's answer" in plain, declarative English. No double negatives, subordinate clauses, or lawyerly hypotheticals. Can you link to a clear statement that forthrightly tackles the main issues surrounding the use of Tiljander by Mann08?

    Also, I don't believe that Gavin is on the record with "it doesn't matter" -- IIRC, that was my paraphrase of the pro-Mann08 defenses in the comments of Collide-a-scape and Arthur Smith's two posts (these and other posts are linked here). But "it doesn't matter" seems like a fair summary of these apologias. In my opinion, the phrase highlights the postmodern nature of pro-Mann08 advocacy. What, exactly, is "it"? And what, exactly, defines "matters"?

    3. No, my complaint is not about scientific honesty. I am not a mind-reader. I am not an expert in this field. But as a scientifically-literate layperson, I can read Mann08, follow the arguments about the use of the Lake Korttajarvi varve series, pull Tiljander03, locate data archives, and graph the series in Excel. Having done so, I can pose questions, and evaluate the quality of the answers.

    Any similarly-educated person can do the same, if they want to make the effort.

    As to Mann08's Tiljander problem transferring to Kemp11: Yes. As far as I can tell, the trace for "Composite EIV global land plus ocean global temperature reconstruction" in that paper's Figs. 2A and 4A is simply the smoothed version of the trace for "Composite (with uncertainties)" in Panel F of Fig. S6 of the Mann08 SI.

    If so, the Tiljander-related problems of the black trace of Mann08's Fig. S6 Panel F are also problems for Kemp11 Figs. 2A and 4A, as well as for Figs. S3, S4, and S5 of the Kemp11 SI.

    ReplyDelete
  5. "I was surprised at the provenance of the paleotemperature reconstruction that was used in Kemp et al's Fig. 2A and Fig. 4A. According to Fig. 2A's legend, it is "Composite EIV global land plus ocean global temperature reconstruction, smoothed with a 30-year LOESS low-pass filter". The reference is Mann et al. (2008). In that paper's S.I., the unsmoothed version is in panel F of Fig S6, as the black line labelled "Composite (with uncertainties)".

    This is one of the multiproxy reconstructions that employed the four (actually three) uncalibratable Tiljander lakebed sediment data series.

    According to Gavin Schmidt, "...it's worth pointing out that validation for the no-dendro/no-Tilj is quite sensitive to the required significance, for EIV NH Land+Ocean it goes back to 1500 for 95%, but 1300 for 94% and 1100 AD for 90%" (link). Further remarks on this issue as Responses to other RC comments here (see numbers 525, 529, and 531).

    The incorrect inclusion of Tiljander could well make this EIV reconstruction progressively worse, as one goes from 1500 AD back to 500 AD. This might explain the increasing divergence between the temperature recon and the sea-level recon, as one travels back from 1100 AD to the beginning of the recons at 500 AD. This pattern is shown in Kemp11's S.I. Figs. S3, S4, and S5.

    Did any of the peer reviewers comment on this issue, or request that you use a no-Tiljander temperature reconstruction?"


    Dude you are challenging McIntyre for prose density.

    Nice work though. You should write this up.

    ReplyDelete
  6. Thanks,

    1. So for this one, it could come down to a difference of opinion on how the data should be used between Tiljander and Mann. I noticed Mann 08 seemed unsure how well this proxy estimated temperature which is what led to the non-dendro charts. So, that being the case, would an explanation from Mann, or someone else on why a reconstruction with Tiljander included is useful suffice? Or an admittance that it is not useful? Or perhaps a mention of using Kaufman's methods, if this is the case?

    2. "What matter's"
    The easy answer to this question is of scientific importance, I believe. It doesn't come from the paper's authors (unless we count Vermeer for the new paper), it comes from someone trying to explain to laypeople, how the results tie into the big picture. I'd stay away from thinking of these people as 'apologias', because it is more likely that, like me, they just want to know how this effects our understanding of climate. We could look at that two ways. (i) Is it not necessary to know MWP exact temperature if trying to understand sensitivity. It would be different if we had data on forcings, but we don't. (ii) It IS necessary to know (within reason) what the MWP temp was if we come to conclusions about whether the last 3 decades is unprecedented in the NH over the last millennial. So Tiljander proxy data effects confidence levels in the conclusions. I'm sure it is one of the reasons that confidence levels are not at %100. I believe this is what people are talking about when they say, "it doesn't matter". Other "matters" are probably of small significance to them. You probably already understand that this issue is too narrow to engage people into it, with out proper cause to do so.

    3. I have it from two separate people, that don't know each other, but both know Mann, that he is very honest and only out to get to the closest proximity of the truth. Take this anecdotal evidence for what it is.

    On the other hand, this could be about 'best practices' which is just too ambiguous and a lot more complicated than people understand.

    In a sense, I think we can resolve this by simply getting an answer on number 1. It's not going to make everyone comfortable, but I believe this to be a reasonable request. perhaps the answer is out there, but we've yet to see it, I don't know. Without that answer, it's difficult to answer the "what matters" and "what does it mean for science" questions.

    ReplyDelete
  7. I forgot to add to the "what matters" discussion that, if we were discussing statistical methods, the sign used in the multivariate regression doesn't matter in Mann's method, according to his reply in PNAS. But I'm not sure that matters here.

    ReplyDelete
  8. Grypo,

    My thinking is much more concrete than yours.

    In the present instance, a party can collect varve data. That party can then propose an interpretation of this data. For lightsum, Tiljander said that more mineral particles (silt) wash into the lake after a colder, snowier winter than after a warmer, less-snowy winter. She also said that after 1720, silt influx was increasingly dominated by local factors such as farming and road-building.

    Now, Tiljander might be right, or partly-right, or wrong. E.g. "lightsum" might have an excellent pre-1720 correlation with winter cold (or snowiness) for centuries. Or the correlation might be poor. Or it might be intermittent (for instance, I'd expect higher lightsum in the years following a forest fire, which would correlate with hotter, dryer summers).

    Tiljander might also be wrong about post-1720 contamination (though, I suspect you haven't yet looked at the graphs in "The Tiljander Data Series: Data and Graphs", as I recommended at 7:47am).

    What's acceptable practice for a follow-on investigator's use of such data? I can see three:

    a. "We are relying on the prior authority, and using the data in line with their interpretations."

    b. "We have a new interpretation that's different from the prior authority's, and are using the data that way."

    c. "We view this data as a string of numbers, and are ignoring prior interpretations."

    The quotes here are meant to signify both what these new authors did, and what they state in their Methods that they did.

    So, which choice did Mann08's authors make?

    That's the problem.

    They picked the Lake Korttajarvi varve core information as potential proxies on the strength of Tiljander03. They considered then ignored Tiljander's post-1720 warning. They then used one series as Tiljander suggested (pre-1720, anyway), and used two in an upside-down orientation (pre-1720, again). They used a fourth series Tiljander hadn't commented on -- and in doing so, they double-counted (see the prior cited post).

    Mann08's Methods do not state or imply that they are doing anything other than following in Tiljander's footsteps (except extending the analysis post-1720).

    This is a mishmash of a, b, and c.

    I don't see this as a "difference of opinion". It's a series of mistakes.

    If Mann08's authors want to offer an explanation: well, they should. Likewise for the paper's defenders, e.g. Gavin Schmidt.

    In my opinion, there's no explanation that starts out other than, "We erroneously employed the Tiljander data series as temperature proxies." This is a road that the authors decided not to travel in their Response to McIntyre's and McKitrick's Comment in PNAS.

    And they just keep doubling down, as far as I can tell.

    ReplyDelete
  9. Looking at Grypo's 1. I think it is quite clear he does not understand the issues. The argument he is using is not really in play though I have seen a number of ill informed people try to use it elsewhere. He should look at the two posts at climateaudit where there is a reasonable description of what is actually wrong.

    http://climateaudit.org/2008/10/02/its-saturday-night-live/

    or probably better at

    http://climateaudit.org/2009/10/14/upside-side-down-mann-and-the-peerreviewedliterature/

    Personally I still regard these 2 posts as a bit opaque and it took me several minutes of squinting and rereading before the light bulb went on for me. I still believe there is a strong need for a "Tiljander for dummies" post which improves upon and condenses the above posts to better describe and illustrate the mechanisms at play. Maybe AMac would consider doing such a post.

    What Grypo needs to understand is two basic issues.

    1. As AMac points out the contaminated and very exaggerated modern portion makes the proxy totally invalid for the type of processing that is being applied in the papers it has appeared.

    2. The contaminated modern portion if it was a real proxy would actually represent a modern ice age. The valid MWP part of the proxy is on the opposite side of the axis to the invalid modern portion. The studies then invert the modern portion to perform calibration but the valid part is pinned so this also inverts the valid portion into the "upside down" state that Steve McIntyre and others are having so much entertainment with.

    For me the take home message is the that 4 large groups of scientists are reusing data where it is clear none of them understand the properties of at least one of the datasets they are using. It also illustrates a series of issues with the processing being applied for these proxy studies and which lead me to the conclusion that the hockeystick results are far more likely to be an artefact of dubious data being badly mishandled.

    Clivere

    ReplyDelete
  10. Mosh here.

    I would also suggest a Tiljander for dummies.

    A post that starts with a reprise of Mia. with charts, not references to charts.

    Then a clear exposition of what mann selected.

    Then addressing his variations/ dendro no dendro etc

    Then Kaufman.

    Then Kemp.

    A long piece. for dummies.

    ReplyDelete
  11. FYI I have posted the following at RealClimate:

    Your comment is awaiting moderation.
    23 Jun 2011 at 9:50 AM

    Dear Prof. Rahmstorf,

    This has two parts.

    1)

    I assume you have seen responses in the German media at Der Spiegel Online (translated from German by GWPF
    http://thegwpf.org/science-news/3267-new-sea-level-study-divides-climate-researchers-.html).

    The translation quotes you as saying, “The new study confirms our model of sea level rise – the data from the past sharpens our view in the future”.

    Then it continues, “But other experts doubt exactly this claim. They see a major problem of the new study in the fact that it is ultimately based only on the finding from the coast of North Carolina. That could be too limited for a statement regarding global developments.”

    I am surprised that no one has commented here on this objection or preempted it with discussion?

    2)

    The study has also been criticised on various blogs for using “one of the multiproxy reconstructions that employed the four (actually three) uncalibratable Tiljander lakebed sediment data series” e.g. http://amac1.blogspot.com/2011/06/tiljander-data-series-appear-again-this.html.

    I don’t claim to understand the objection but given that it has spread like wildfire on blogs I am surprised that no one has mentioned it here.

    Is this objection valid?

    Cheers,
    Alex Harvey

    ReplyDelete
  12. Amac,

    That last post you made helped me understand your issues a little better. They are reasonable. They are related to scientific behavior or honesty, in that it appears that there is an issue with the way the data used and the "it doesn't matter" explanation although likely true, doesn't suffice to someone looking into the details. This is compounded by multiple uses. I took the "it doesn't matter" argument from an overall science perspective, so my contention about Mann's explanation actually does matter to your argument, so I was wrong to exclude that. So in addition to answering to question 1 (why use post 1720 data), it would help if we had a reason as to why this data is still used (wrongly?), especially if it "doesn't matter".


    In my opinion, there's no explanation that starts out other than, "We erroneously employed the Tiljander data series as temperature proxies."


    And that would be what I am asking. Would that suffice? And do you know if it is still being used that way?


    Looking at Grypo's 1. I think it is quite clear he does not understand the issues.


    Right, and this is why it is phrased in a question. I've been quite clear that I came here to GET the information on Amac's issues. Plus, I don't have to worry about being told that I'm an idiot by 50 commenters, instead of just one or two.

    A long piece. for dummies.
    Small pieces of information by Twitter would do better.

    ReplyDelete
  13. Alex Harvey @ 7:51am --

    Thanks for submitting a polite and informative comment to RC. It is very unfortunate that all things AGW have become so polarized and personal. It ought to be possible to have a civil and informative discussion on the merits of the science, without participants always looking over their shoulders at other things (e.g. policy implications).

    The count on the RC post was 15 last night, and is 21 right now. So their moderators have allowed some fresh comments.

    At #18, Gilles asked, "if the model is so sensitive that the pre-1000 discrepancy could be removed by only a 0.2 °C change, what is the reliability of the agreement between 1000 and 1500? i don’t think that temperatures were known to this accuracy, so is the agreement just adjusted by a convenient choice of the “right” temperature reconstructions?"

    Corresponding author Stefan Rahmstorf responded, "There was little (or even no) choice here since we needed a global land + ocean reconstruction..."

    That may be. In my opinion, readers should have been made aware of the dependence of the temperature recon on the uncalibratable Tiljander data series. According to Gavin Schmidt and Mann09's SI, this dependence grows stronger as one goes back in time before 1500AD, so it is very relevant to Gilles' observation.

    At #19, CM referred obliquely to issues with the temperature reconstruction, saying "I think McIntyre... suggest[s] that a downweighting of the proxy data applies to the sea-level reconstruction, when it actually relates to the fit to that reconstruction of a model based on reconstructed temperatures..."

    Co-author Martin Vermeer replied in #21, "Yes, it only relates to the fit."

    Obviously, that doesn't address the issue I raised in this post.

    ReplyDelete
  14. Grypo @7:53am --

    > and the "it doesn't matter" explanation although likely true...

    Well, whether "it doesn't matter" is "likely true" or not will depend on the definition of "it" and the meaning of "matter" that are selected. As noted above, this strikes me as a postmodern grammatical construction.

    I would say this --

    [The incorrect use of the Tiljander data series as temperature proxies in multi-century temperature reconstructions] matters [as far as the validity of those reconstructions].

    Now, if [the validity of these reconstructions] "doesn't matter" -- then why do authors write manuscripts on this subject? Why do editors at high-impact journals send them out for peer review? Why do specialists and scientifically-literate laypeople read them, and comment on them?

    As you can probably tell, I remain baffled by this line of argumentation. I don't find that it leads to greater clarity of thought, or to improved understanding of the issues at hand.

    > Would [an explanation that starts out 'We erroneously employed the Tiljander data series as temperature proxies'] suffice?

    If the choice is between a wrong assertion ("Tiljander used correctly") and a right one ("Tiljander used erroneously"), I'll take the right one.

    The authors of Kaufman09 looked at this issue and decided to cut their losses by amending their manuscript. They could do this prior to formal publication in Science. The problem for Mann08's authors is that they have repeatedly doubled down on the claim that their use of Tiljander is A-OK. At this point, they would have a lot of revising to do. And a lot of explaining to the foot soldiers who have carried water on their behalf. That could be embarrassing.

    > And do you know if it [the Tiljander data series] is still being used that way [incorrectly]?

    Yes.

    > I've been quite clear that I came here to GET the information on Amac's issues.

    Right, and I'm glad of that. You certainly don't deserve dissing for having an inquisitive attitude. Or for being civil. But then, I say that as somebody who also accepts the validity of RTEs, and that global average temperature has risen ~1C in the past century or so, and that rising CO2 will significantly raise gloval average temperature. How much -- I don't know. That's why this is the beginning of the conversation, not its end. :-)

    ReplyDelete
  15. AMac
    Thanks for your comment on WUWT and the link back to here. My comment is based on general science and from memory. I have not studied in detail the modern era contamination of the varve record for Tiljander, I simply note that contamination leading to mis-calibration turns cold into warm & vice versa.

    ReplyDelete
  16. Alex Harvey @7:51am --

    Your submission to RealClimate passed moderation there and is up as #22.

    Your point #2) was presented as follows --

    "The study has also been criticised on various blogs for using [edit] Tiljander lakebed sediment data series [edit]
    I don’t claim to understand the objection but given that it has spread like wildfire on blogs I am surprised that no one has mentioned it here.

    Is this objection valid?
    "

    According to what you wrote upthread, the first sentence as submitted read --

    "The study has also been criticised on various blogs for using “one of the multiproxy reconstructions that employed the four (actually three) uncalibratable Tiljander lakebed sediment data series” e.g. http://amac1.blogspot.com/2011/06/tiljander-data-series-appear-again-this.html."

    Emphasis added.

    Mike [Mann] left this inline commentary --

    "[Response: No. Just more of the usual deception from dishonest mud-slingers. More on that in short order. -Mike]"

    Well. Ad homs aside. Despite the unappetizing prospect of a big serving of humble pie, I actually hope I'm wrong on this. From my perspective, it would mean that Climate Science is in much better shape than I have come to believe. If apologies and retractions turn out to be in order, apologies and retractions Prof. Mann will get.

    I was re-reading a contentious Collide-a-scape thread from a year ago. Boiling it down to a single, simple question, here's what I asked back then.

    Are the Tiljander proxies calibratable to the instrumental temperature record, 1850-1995?

    Emphasis added.

    ReplyDelete
  17. Prof. Mann has responded to this post, albeit without referring or linking to it.

    Update 2 (June 23) to "2000 Years of Sea Level".

    There is a new graph of a smoothed "Mann et al (2008) global mean (land+ocean) temperature reconstruction." It is shown with (blue) and without (red) 7 proxy records, which include the Tiljander data series.

    The graph is largely uninterpretable. It lacks uncertainty intervals for the two curves, as Mann et al (2008) would calculate them. Further, there is no discussion of what portions of the two graphs pass the "validation test" as described in Mann et al (2009), at the 95% or higher level.

    In the paragraph of substantive text, the following question remains unacknowledged and unanswered:

    Are the Tiljander proxies calibratable to the instrumental temperature record, 1850-1995?

    Perhaps the RealClimate bloggers are planning to post additional information in a while.

    ReplyDelete
  18. The new graph uses bristlecones.

    Newcomers to this controversy need to keep in mind that Mann 2008's original claim to fame was that it claimed to 'get" a stick without bristlecones. The no-dendro stick used the problematic Tiljander sediments. Mann purported to show that the problems with contamination of the sediments didn't "matter" by showing that he could get a stick without the sediments. This version used bristlecones.

    However, they didn't show the no-bristlecone no-dendro result. What they did eventually show needs to be carefully parsed and I plan to revisit this.

    The new graph looks to me as nothing more than a variation of the original no-Tiljander version i.e. it uses bristlecones.

    ReplyDelete
  19. Amac:
    I truly admire your tenacity and patience. I am just a humble layman, but when I first read the climateaudit post about the "upside down" tiljander series, I immediately understood the problem. I did not find the post even a bit confusing. When I read Mann's formal reply, I understood why it was unresponsive. Since then I have been utterly bewildered that this issue is still controversial is some circles.

    ReplyDelete
  20. Pauld/AMac - just as Climategate broke I was having quite an energetic discussion with Steve McIntyre and others on this thread at climateaudit. AMac even visited briefly.

    http://climateaudit.org/2009/11/05/core-count-in-phil-trans-b/

    2 years on I will modify by views slightly but am still mainly happy with what I wrote then.

    As far as I am aware the only places where a clear description of the problem resides is in the 2 posts I linked earlier for Grypos benefit. AMac may be aware if someone has done something more recently because I lost interest 12 months ago so have not been following it as closely. The Its Saturday Night live post is rather obscure and may be difficult for people to locate. The upside-side-down-mann-and-the-peerreviewedliterature may be a bit easier for people to identify as relevent but is still one of quite a large number of CA posts.

    One of my main arguments in 2009 was the need for a journal quality post which clearly illustrates the mechanisms and also shows why the "if we take them out it doesn't matter" argument being floated is incorrect. This post should avoid any snark to make it more acceptable for academics who may otherwise get diverted. I disagree with Steve Mosher that it needs to be a very lengthy post. Ideally it should be of a format that Anthony Watts would find useful for reposting.

    If you look at the wider world you will see that there are different levels of understanding. For people like Grypo who appears to be from the academic world as far as they are concerned Mike Mann et al refuted the concern raised by Steve McIntyre and Ross M. Therefore from an academic perspective they would regard the Mann reconstructions as good and are probably astonished by AMacs post about Tiljander and the derision for the Kemp paper from skeptics that resulted.

    Grypo appears to have visited one or two of the AGW blogs which have covered Tiljander. However I still find that most of them are not able to articulate the issue and this results in some poor arguments. We know that it took William Connelley 2 sets of posts in 2009 before we had his light bulb moment and he then went into a damage limitation exercise using the it does not matter argument for his third post.

    ReplyDelete
  21. AMac - for your amusement you should look at the following and pages 30 and 35 in particular. How to use a paleo reconstruction to influence government policy.

    http://downloads.globalchange.gov/ocp/ocp2010/ocp2010.pdf

    ReplyDelete
  22. AMac - some observations/comments based on the ongoing skirmishes that are breaking out at various venues.

    We now seem to have both sides implicating Mia Tiljander as being somehow at fault in different ways. I have copied the following from one of your earlier 2009 posts to save retyping.

    Mia Tiljander. "Holocene sedimentary history of annual laminations of Lake Korttajärvi, central Finland." PhD Dissertation, University of Helsinki, Dept. of Geology, Geology and Palaeontology. October 2005. Archived PDF. Includes the quote, "Since the early 18th century, the sedimentation has clearly been affected by increased human impact and therefore not useful for paleoclimate research." (pg. 24).

    http://ethesis.helsinki.fi/julkaisut/mat/geolo/vk/tiljander/

    In my opinion no fault can legitimately be placed on Mia Tiljander who has clearly provided the necessary qualifier in her paper. I also think the Steve McIntyre "silence of the lambs argument" is a stinker in this context. He appears to want Mia Tiljander and the whole of the academic community to rise up and condemn Mike Mann et al. At this rate he will next want everybody to go and picket their houses or something. There needs to be a sense of proportion here. Mike Mann et al have made a howler of a technical error that should be acknowledged but it is only a technical error.

    I still have the view that until the Kaufman correspondence (Climategate emails) in September 2009 Mann et al were probably in ignorance of their error and their actions including the PNAS reply to M&M reflects that ignorance. However post Kaufman they should certainly be aware. I now see clear signs of evasion including the recent edit deletion of uncalibratable from the query Alex Harvey placed at realclimate on your behalf. Grypo was implying that academics regards Mike Mann as honest. I do not regard the behaviour we are currently seeing with respect to Tiljander as academically honest.

    The eventual repercussions will depend on the overall assessment of the impact of the error.

    The impact on the non Dendro reconstruction looks terminal to me. Your redrawing of the graph is Figure S8 of the S.I. for Mann09 clearly shows that any claim of a skillfull non dendro reconstruction cannot really be justified. Steve McIntyre looks justified in pushing this.

    The impact on the overall reconstruction is a lot less dramatic. However the comment by Mike Mann here about the impact on the Kemp et al paper shows that there is some impact.

    http://www.meteo.psu.edu/~mann/supplements/MultiproxyMeans07/

    "interestingly, eliminating the proxies in question actually makes the reconstruction overall slightly cooler prior to AD 1000, which -- as noted in the article -- would actually bring the semi-empirical sea level estimate into closer agreement with the sea level reconstruction prior to AD 1000"

    I would make the observation here that it cannot be good academic practice to tout a reconstruction as suitable for further research where a known error makes the reconstruction somewhat unreliable. I would also make the observation that if by chance the valid Tiljander data had been fed in the right way up it would probably have shown as even cooler still prior to AD 1000. There must be a concern if one small set of proxies can have that level of impact.

    ReplyDelete
  23. Clivere (July 10, 1:43 AM) --

    For some reason, I had to fish your comment out of blogspot's spam filter. Do you sell C1al1s on the side? :-)

    I'd better learn how to adjust the settings.

    > I still have the view that until the Kaufman correspondence (Climategate emails) in September 2009 Mann et al were probably in ignorance of their error and their actions including the PNAS reply to M&M reflects that ignorance.

    Seems like the simplest explanation to me. It's consistent with the idea of "confirmation bias," which seems to be an important theme in the Mann08 saga.

    Recall that in the comments at Gavin Schmidt's notable quasi-guest-post at Collide-a-scape, Schmidt discusses Mann's revision of Mann08 SI Fig. S8 (the first of the two revisions placed on PSU's website, IIRC). Schmidt claimed that Mann did this as a response to blog-based criticisms. This implies that Mann et al were aware of the objections that McIntyre was raising at Climate Audit almost as soon as Mann08 hit the digital newsstand.

    On this issue, Schmidt seems willing to serve as Mann's spear-carrier, while Mann08's authors remain silent. Obviously, this provides an escape hatch, "No, that's not it at all -- Schmidt misinterpreted us."

    ReplyDelete
  24. I suspect you need to fix the settings!

    My recollection is the Tiljander issue was raised several weeks after Mann08 came out and after the initial blog activity. I suspect that Gavin Schmidt either missed it or failed to relay it on.

    ReplyDelete