New Article on Igaliku

Shortly after the publication of PAGES2K, I pointed out that the Igaliku lake sediment proxy, had been contaminated by modern agricultural runoff. The post attracted many comments.

Nick Stokes vigorously opposed the surmise that the Igaliku series had been contaminated by modern agriculture and/or that such contamination should have been taken into account by Kaufman and associates. Stokes:

I see earlier demands that selection criteria be declared for proxies. Kaufman has done that, and appears to have stuck with them. But when a spike appears, suddenly the CA throng has a thousand a posteriori reasons why Kaufman is a reprobate for not throwing it out.

or

I see no reason to disagree with the original authors, Massa et al in saying that “pollen accumulation appears to document climatic changes of the last millennia nonetheless”. The Betula/Salix counts are not contaminated.

Subsequent to my CA post, the Igaliku specialists have published a new article entitled “Lake Sediments as an Archive of Land use and Environmental Change in the Eastern Settlement, Southwestern Greenland” (abstract here) which unambiguously connected soil erosion to agriculture, not just in the modern period, but in the medieval period, observing that modern mechanization in the 1980s had resulted in a “five times” the rate of erosion.

Palaeoenvironmental studies from continental and marine sedimentary archives have been conducted over the last four decades in the archaeologically rich Norse Eastern Settlement in Greenland. Those investigations, briefly reviewed in this paper, have improved our knowledge of the history of the Norse colonization and its associated environmental changes. Although deep lakes are numerous, their deposits have been little used in the Norse context. Lakes that meet specific lake-catchment criteria, as outlined in this paper, can sequester optimal palaeoenvironmental records, which can be highly sensitive to both climate and/or human forcing. Here we present a first synthesis of results from a well-dated 2000-year lake-sediment record from Lake Igaliku, located in the center of the Eastern Settlement and close to the Norse site Garðar. A continuous, high-resolution sedimentary record from the deepest part of the lake provides an assessment of farming-related anthropogenic change in the landscape, as well as a quantitative comparison of the environmental impact of medieval colonization (AD 985—ca. AD 1450) with that of recent sheep farming (AD 1920—present). Pollen and non-pollen palynomorphs (NPPs) indicate similar magnitudes of land clearance marked mainly by a loss of tree-birch pollen, a rise in weed taxa, as well as an increase in coprophilous fungi linked to the introduction of grazing livestock. During the two phases of agriculture, soil erosion estimated by geochemical proxies and sediment-accumulation rate exceeds the natural or background erosion rate. Between AD 1010 to AD 1180, grazing activities accelerated soil erosion up to ≈8 mm century-1, twice the natural background rate. A decrease in the rate of erosion is recorded from ca. AD 1230, indicating a progressive decline of agro-pastoral activities well before the end of the Norse occupation of the Eastern Settlement. This decline could be related to possible climate instabilities and may also be indirect evidence for the shift towards a more marine-based diet shown by archaeological studies. Mechanization of agriculture in the 1980s caused unprecedented soil erosion up to ≈21 mm century-1, five times the pre-anthropogenic levels. Over the same period, diatom assemblages show that the lake has become steadily more mesotrophic, contrary to the near-stable trophic conditions of the preceding millennia. These results reinforce the potential of lake-sediment studies paired with archaeological investigations to understand the relationship between climate, environment, and human societies.

I recently noticed that my criticism had been more or less conceded in McKay and Kaufman 2014, which purported to accommodate the contamination (or overprinting, as suggested by Mosher) by deleting the last two points. I was critical of their correction, arguing that their correction still leaves a heavily contaminated reading in 1970. (The next reading is dated circa 1910 – its’ very low resolution and actually below the resolution standard of the study).

It’s hard to tell whether this was intentional or not. I can see one way that they might have left in this value by accident. If they had deleted two points from the PAGES2K-2013 version, that would have also deleted the contaminated 1970 point. But the PAGES-2013 had already omitted or removed one point from the underlying NOAA version. The new McKay and Kaufman version deleted two points from the NOAA version, and thus only one point from the PAGES2K-2013 version, still leaving the contaminated 1970 reading.

Or, if pressed, perhaps they would argue that the most recent article only expressly referred to mechanization “in the 1980s”. However, this hardly precludes the likelihood that the elevated erosion observed in the sample dated circa 1970 could not similarly be attributed to mechanization occurring earlier than the 1980s (farm mechanization obviously occurring throughout the world long before the 1980s) or dating error.

The series should never have been used in a temperature reconstruction.

Note: Jean S and I have been doing some interesting analysis of paico and it is my present view that Igaliku does not have a large impact on the paico reconstruction, but does have a large impact on the “basic composite” reconstruction, one of the PAGES2K alternatives.


113 Comments

  1. Craig Loehle
    Posted Oct 11, 2014 at 9:45 AM | Permalink

    Contamination is always something that must be considered in any field work. It happens with carbon dating. In this case, the human effects are large because sediment is the proxy. It would also affect the early settlement period and there is no way to “factor it out”. You never keep a contaminated point because you “need” it. Toss the whole thing.

  2. Steven Mosher
    Posted Oct 11, 2014 at 10:40 AM | Permalink

    I think people will get a really good sense for the racehorse approach by reading
    https://climateaudit.org/2013/04/29/more-kaufman-contamination/#comment-416616

  3. tty
    Posted Oct 11, 2014 at 11:57 AM | Permalink

    The complete paper is available here:

    https://www.academia.edu/7230342/Bichet_et_al._2014._Lake_Sediments_as_an_Archive_of_Land_Use_and_Environmental_Change_in_the_Eastern_Settlement_Southwestern_Greenland

    In the body of the paper major change of vegetation caused by modern sheep-grazing is dated to c. 1950 with major soil erosion from 1960, so the 1970 data point is certainly contaminated.

  4. tty
    Posted Oct 11, 2014 at 12:15 PM | Permalink

    An interesting quote from the paper:

    “While lake sediments have the temporal resolu-tion necessary to capture the full range of climate variability, their suitability for palaeotemperature reconstructions is usually limited due to the lack of direct quantitative proxies. In suitable hard-water lakes (when the evaporative effect on lake-water δ18O is low), authigenic calcite may reliably record the lake-water δ18O, which is controlled by the δ18O of precipitation, itself strongly correlated to the mean annual temperature at high latitudes (Masson Delmotte et al. 2012). However, lakes in southern Greenland are poorly buffered, and calcite is largely absent there. So, our ongoing studies at Lake Igaliku are now focused on two ways of extracting palaeo-temperatures records. The first involves chironomid larvae for both the analysis of chitin δ18O and species assemblage-based temperature inference. The latter uses alkenone biomarkers, which are highly resistant organic compounds produced by phytoplankton, where molecular long-chain saturation depends on lake-water temperature (D’Andrea et al. 2011, Zink et al. 2001). Using these complementary ap- proaches, a quantified reconstruction of historical temperatures at decadal scale in the settlement may be feasible.”

  5. Nic Stokes
    Posted Oct 11, 2014 at 2:06 PM | Permalink

    “Nick Stokes vigorously opposed the surmise that the Igaliku series had been contaminated by modern agriculture”
    Actually, I didn’t. I was pointing out that CA had been inveighing against what it called ex post selection, and demanding that scientists should state rules in advance and stick to them. I pointed out that Kaufman et al had done just that, as in your first quote from me. Their rule was based upon a stated opinion by the original authors that pollen was a temperature proxy.

    Those authors, Massa et al, had said that there was a modern agricultural effect which affected most proxies, but, explicitly, that pollen was still OK. I said their view should be respected in the application of the rule.

    This simple but apparently incomprehensible argument earnt special derision at WUWT. I made my case here.

    CA was initially enthusiastic about Massa et al when they spoke of agricultural interference, but then not so much when it turned out that they believed pollen was not affected. I urged consistency. The vigorous condemnation of ex post decisions was CA’s, not mine. Kaufman seems to have made a limited ex post decision, which looks reasonable to me.

    Steve: Nick, show me a single place where I was “initially enthusiastic about Massa et al when they spoke of agricultural interference, but then not so much when it turned out that they believed pollen was not affected.” Please provide a quotation. Nor have I tried to set up a Napoleonic code for how specialists should handle data. The pathology of ex post selection of series from a population according to temperature correlation is something that ought to be understood even by a climate scientists. Observing this pathology does not entail the inclusion of data contaminated by agricultural activity. Supporting objective protocols does not imply that one supports stupid protocols. Further, the problem of this sort of contamination was well known from Mann’s upside-down Tiljander fiasco. Any protocol that did not consider this sort of problem in advance was either negligent or incompetent. That you think that such stuff is reasonable is just ClimateBaller inanity.

    • scarletmacaw
      Posted Oct 11, 2014 at 2:51 PM | Permalink

      So what you’re saying is that including contaminated measurements is OK because it wasn’t addressed BEFORE the analysis?

      • Nic Stokes
        Posted Oct 11, 2014 at 3:34 PM | Permalink

        Steve,
        “Please provide a quotation.”
        It was you who introduced Massa et al, with a long quote introduced thus:
        “The specialist publications make it overwhelmingly clear that modern agriculture has resulted in dramatic changes to the sediments, rendering the recent portion of the Igaliku series unusable as a climate proxy. Here are some quotes from the original article[Massa et al]:”

        Then in an update, with waning enthusiasm:
        “There is an obvious relationship between erosion (indicated by mineral matter accumulation) and pollen accumulation. Massa could just as easily “suggested” that erosion was a proxy for temperature.”

        Steve: you’re just fabricating again, Nick. My comment was further argument that both measures were nothing more than agricultural runoff, as in Tiljander. You’re in full racehorse mode.

        • Craig Loehle
          Posted Oct 11, 2014 at 3:53 PM | Permalink

          Nic, your first quote from steve fails to show him exhibiting ANY enthusiasm for Igaliku–try again.

        • Nic Stokes
          Posted Oct 11, 2014 at 4:00 PM | Permalink

          Craig,
          His enthusiasm was for Massa et al, and what they seemed to be saying about nitrogen run-off etc. I never said Steve was enthusiastic about Igaliku.

          Steve: NIck, what sort of fantasy world do you ClimateBallers live in? I consistently and unequivocally said that this proxy was contaminated by agricultural runoff, just like Mann’s upside-down Tiljander. D’oh. Your fabrications are getting annoying.

        • Craig Loehle
          Posted Oct 11, 2014 at 4:22 PM | Permalink

          The quote is about Igaliku and the article by Massa and does NOT show any enthusiasm.

        • Nic Stokes
          Posted Oct 11, 2014 at 5:26 PM | Permalink

          “I consistently and unequivocally said that this proxy was contaminated by agricultural runoff, just like Mann’s upside-down Tiljander. “

          No dispute there. The thing is, it turned out that Massa et al said that pollen was still a sound temperature proxy, and it was their proxy that Kaufman was using. That’s when Massa et al seemed to become not so well liked.

          Anyway, sorry about the annoyance. It will probably diminish, as my comments seem to be going into moderation again.

          Steve:no they didn;t say that it was a “sound temperature proxy”. Nor, ironically, did the proxy even meet Kaufman’s standards for resolution of 50 years.

        • Craig Loehle
          Posted Oct 11, 2014 at 8:55 PM | Permalink

          Nick, if a site gets a major introduction or cessation of agriculture, the pollen will radically change and not be reliable. This applies here to after 1950 and to the earlier agricultural period that came and went so overall this proxy is useless.

    • Bob
      Posted Oct 11, 2014 at 3:06 PM | Permalink

      Nick, I must say I have never encountered a scientist, such as yourself, that so consistently opposes any attempt to understand and not insist on quality. Imagine what the state medicine or surgery would be in if people with your approach to quality were in charge of data analysis.

      • Sven
        Posted Oct 11, 2014 at 3:50 PM | Permalink

        +1

      • Nic Stokes
        Posted Oct 11, 2014 at 4:03 PM | Permalink

        Bob,
        “consistently opposes any attempt to understand and not insist on quality”
        I’m all for quality, and getting stuff right.

        CA’s stern condemnation of ex post exclusion was directed at a series which indicated a massive temperature dive that obviously didn’t happen. T&H excluded it; the original authors didn’t think it was right, and gave reasons. Yet CA slammed T&H. I think their (T&H) exclusion was in the interests of quality.

        As said, I’m not a fan of rigid rules, and I think Kaufman’s relaxation here is sensible.

        Steve: More misrepresentation. You have not provided a shred of evidence that Mount Logan ice core was contaminated by agricultural runoff or other anthropogenic contamination. The authors said that the decline was due to “regional” weather. Not a reason for exclusion as “regional” weather can increase O18 elsewhere. This was no better than Briffa and Mann’s hide the decline. It’s amazing that purported scientists struggle with this.

        • Nic Stokes
          Posted Oct 11, 2014 at 11:08 PM | Permalink

          “The authors said that the decline was due to “regional” weather.”

          Fisher et al said more than that.
          They say, unequivocally (Intro)
          “At 5,400 m asl, d(18O) is not a temperature indicator, and a goal of this paper is to explain what it does indicate.”

          They contend that the variation depends which side of the divide the water is coming from. Tingley and Huybers said (your quote)

          “We exclude the Mount Logan series that is included in [35] because the original reference [36] indicates it is a proxy for precipitation source region and is out of phase with paleotemperature series.”

          Freedom from agricultural runoff does not ensure a temperature proxy.

      • Alan Poirier
        Posted Oct 11, 2014 at 7:06 PM | Permalink

        Is Nick Stokes a scientist?

    • Posted Oct 12, 2014 at 10:06 PM | Permalink

      > Those authors, Massa et al, had said that there was a modern agricultural effect which affected most proxies, but, explicitly, that pollen was still OK.

      Is this correct?

      Somehow, this sentence has been skipped in the parsing of Nick’s comments.

      • Steven Mosher
        Posted Oct 12, 2014 at 10:58 PM | Permalink

        nicks chorus is hiding in a black box.
        go find it climateballer

        • kim
          Posted Oct 13, 2014 at 4:25 AM | Permalink

          And, as anybody who knows anything about black boxes knows, he also ruined the meter.
          ================

      • RomanM
        Posted Oct 13, 2014 at 8:48 AM | Permalink

        These has been dealt with before in depth in a previous post.

        Nick still hasn’t given me a proper explanation as to why this proxy differs completely from eight similar proxies post-1920s. His explanation was that in the Arctic “pollen is hard to find, and where it is, people are likely to be not far away.” Strange that anyone would state that this one site exhibits a suggested “rapid warming” which must have been very local to have been missed by all of the other sites, rather than be a result of the people “likely to be not far away.”

        • Nic Stokes
          Posted Oct 13, 2014 at 10:55 PM | Permalink

          “Steve: the site was impacted by agricultural erosion. End of story.”
          In fact, the new 2014 paper is informative on the time course of erosion. It says:
          “From the beginning of the 20th century to ˜1960 (± 5 yr), corresponding to the re-establishment of farming at Igaliku, none of the sedimentary parameters reveals any significant increase of erosion around the lake of Igaliku (Fig. 6). Since ˜1960, the grazing pressure (shown by a decline in woody taxa and a rise in coprophilous fungi [Fig. 5]) caused a progressive increase in detrital parameters. Soil erosion accelerated beginning in 1969 (± 4 yr), with a sharp increase in Ti and C:N ratio, reaching ˜11 mm century-1 in 1988 (± 2.5 yr), slightly more than during the medieval period. Around 1988, major earthworks and digging of drainage ditches were carried out in both hayields (Fig. 3), which caused soil erosion to increase dramatically up to 21 mm century-1. After 1997 (± 2 yr), the erosion rate decreased continuously, whch may mark the stabilization of the material remobilized by the drainage work. “

          Soil erosion acceleration began in 1969 is hardly consistent with your claim of “a heavily contaminated reading in 1970.” In fact, they indicate that the farms we see in the pictures were established after 1980:
          “Since 1980, after the climate crisis of the 1960s/1970s (Egede and Thorsteinsson1982, Greenland Agriculture Advisory Board 2009), two sheep farms (more than 1000 sheep) were established in the catchment, and around 30 ha of hayield were created on the shore of the lake to produce winter fodder for stabling (Figs. 2a, 3). “

          In terms of windbaggery, I still prefer the judgment of the people who actually cored the lake and know about pollen to the various local accounts of why they were all wrong.

        • RomanM
          Posted Oct 14, 2014 at 5:02 AM | Permalink

          Nick, there does not appear to be any “sudden” change in the data in 1960(± 5 yr). The pattern starts much earlier. This series is substantially different in the 20th century from the others yet it is the ***only*** one of the nine shown which somehow picks up on the extreme warming. Such an outlier requires explanation which is not addressed in the quote given.

          Do you honestly believe the pattern was indeed the sole result of a “warming” which could only be found at that one location?

        • Nic Stokes
          Posted Oct 14, 2014 at 6:49 AM | Permalink

          Roman,
          First an apology – this comment wasn’t meant to follow yours, but to tag on the end of the next subthread. But since you’ve responded – yes, the pattern starts earlier. And the authors have measured erosion indicators. So the fact that it starts before 1960 says that the cause isn’t erosion. It could be temperature. I leave that to them.

          “Nick still hasn’t given me a proper explanation as to why this proxy differs completely from eight similar proxies post-1920s.”

          “This series is substantially different in the 20th century from the others yet it is the ***only*** one of the nine shown which somehow picks up on the extreme warming.”

          Mosh says that I would be lying if I said that people were making ex post judgments based on which way the curve goes. But it is the only pollen-based indicator.

        • TAG
          Posted Oct 14, 2014 at 6:59 AM | Permalink

          Instead of dwelling on particulars, could this issue be used to bring out some principles that could be used in assessing the validity of a proposed proxy.

          From what I see there are two principles being used in this discussion

          a) the opinion of the original specialist author
          – does this opinion pass scrutiny?

          b) the compatibility of the proxy measurement with similar proxies
          – is the proxy an outlier which will provide undue influence of the result
          – from what I have read of it here, this seems to be a basic principle behind the PAICO method in assessing the reasonableness of any proxy by comparing it pairwise to other proxies.

          c) I suppose there is another principle that proxies should not be used blindly but the suitability of any and all proxies should be directly assessed. For me, this is one of the issues that shows the problems of dealing with AGW by use of academic research and publishing. A report on a major reconstruction should not be confined to some 15 page paper.

        • Posted Oct 14, 2014 at 8:55 PM | Permalink

          > What is your point?

          So far, I made three requests. First, if Nick’s description of what Massa & al (2012) said is correct; this has been conceded, albeit indirectly.

          [Steve: Nick’s description of what authors say is seldom correct. The authors clearly stated that their data could not be used for paleoclimate of the last millennium and any paleoclimate information had been “overprinted” by erosion. Nor does Bichet contradict the early articles: it confirms them. FWIW Mosher had previously proposed the word “overprinted” as an alternative for “contaminated”. SO yes the authors said so directly but this was ignored by Kaufman and Stokes.]

          This request has a point: what the authors say matters if we’re to use them against Kaufman, CA’s scapegoat of the moment. The same point lies behind the second request, i.e. that we substantiate what was said in “the new Igalku paper”: do the authors themselves declared their data “contaminated,” or is this extrapolated from reading an abstract? My third request was to ask if the corresponding author has been contacted, or alternatively if someone like Richard Telford could be consulted. The point behind this last request is that the authors themselves are responsible for their nonsense, if nonsense there is, and perhaps by extension the community to which they belong.

          These two points follow from what amounts to be Nick’s argument, which may also be an argument made by the authors responsible for the selection of the Arctic proxy in Kaufman & al 2012. They also show that the accusation of “defending nonsense” burdens me with a commitment I do not need to hold, which was the point of my last comment. Note, and this is a new point, that Jean Goodwin already observed this tactic when she came here. It is a common ClimateBall ™ move on CA, for instance when Nick is being forced to racehorse into commitments he has no reason to make.

          My first request referred to the sentence preceding the one that led to an “enthusiastic” hurly burly. This hurly burly obfuscates that Massa & al (2012) has been published in an issue edited by Darrell Kaufman and that Vincent Bichet, the lead and corresponding author of “the New Article on Igaliku”, was also an author on Massa & al (2012). The most expedient way to verify if Bichet & al (2014) contradicts Massa & al (2012) would be to ask Bichet, instead of playing yet another ClimateBall ™ round with Nick.

          Hope this helps,

          W

          PS: If you could resurrect bender from Moshpit’s swimming pool, that would be nice.

      • Posted Oct 13, 2014 at 8:41 PM | Permalink

        > These has been dealt with before in depth in a previous post.

        Not “these”, only one question, about what Massa & al said. Here it is: “pollen accumulation appears to document climatic changes of the last millennia nonetheless”.

        https://climateaudit.org/2013/04/29/more-kaufman-contamination/#comment-416628

        Nick’s claim appears to be correct.

        Thank you, janitor.

        ***

        This question does not only matter for the article discussed “in depth” in that other thread, but also to this new article on Igaliku. Considering that Nick’s main argument regarding Massa & al amounted to defer to the authors themselves, what the authors said about the quality of their data to document climate changes should be disclosed and discussed. All we have is the abstract, where the authors declare that their results “reinforce the potential of lake-sediment studies paired with archaeological investigations to understand the relationship between climate, environment, and human societies”.

        If there is a disagreement with the authors about the “contamination” of their proxies, they should be contacted. Has anyone contacted the corresponding author, Vincent Bichet? We could also ask Richard Telford to comment. I am referring to him because he participated in the comment thread of that “previous post”, entitled **More Kaufman Contamination**. An interesting choice of title considering who collected and reviewed the proxy data.

        If this blog post is about Nick Stokes, none of this is relevant.

        Steve: That the site was impacted by modern agricultural runoff is clearly documented in the specialist literature. In the wake of Mann’s use of contaminated portion of Tiljander data, that ought to have been an ex ante criterion. Use of this data is a screwup and the point should just be conceded. Indeed, Kaufman has already conceded that the site was contaminated by agricultural runoff, though his correction is still screwed up. Trying to defend this sort of nonsense just makes you and Nick look like nothing more than windbags.

        • Posted Oct 13, 2014 at 9:29 PM | Permalink

          > Trying to defend this sort of nonsense […]

          By requiring that a quote be provided from the article which is used as evidence for a claim about “contamination,” or by telling to ask the authors themselves or Richard Telford?

          One can have a perfectly good sell recommendation without necessarily having a buy recommendation in mind [1].

          If all you can do with this stuff is to slime Nick, turn it over to those who collect and review the proxy data.

          [1] http://neverendingaudit.tumblr.com/post/34253462351

          Steve: the site was impacted by agricultural erosion. End of story. If Nick doesn’t want to be criticized for defending contamination, then he shouldnt defend contamination.

        • Steve McIntyre
          Posted Oct 13, 2014 at 9:53 PM | Permalink

          Willard, what is your point? That the data wasnt contaminated by agricultural erosion? that it should have been used anyway? that it was impossible for Kaufman to know that it was contaminated by erosion? that selection criteria that incorporated contaminated data were fine?

        • Steven Mosher
          Posted Oct 14, 2014 at 1:29 AM | Permalink

          More distractions and distortions from the liar in chief of climateball.

        • Steven Mosher
          Posted Oct 14, 2014 at 1:35 AM | Permalink

          “Considering that Nick’s main argument regarding Massa & al amounted to defer to the authors themselves, what the authors said about the quality of their data to document climate changes should be disclosed and discussed.”

          except that wasnt Nick’s main argument and it isnt his only argument.
          And further he continues to engage in material misrepresentations of fact.
          a chorus of black boxes so to speak.

          Climateballers at their best, all tutored by Willard.

        • Nic Stokes
          Posted Oct 14, 2014 at 4:20 AM | Permalink

          “Considering that Nick’s main argument regarding Massa & al amounted to defer to the authors themselves”
          It’s not just my argument. It’s items 5 and 6 in Kaufman’s ex ante rules:
          “The proxy records selected by the Arctic2k group for the Arctic continental-scale temperature reconstruction (Fig. S7) meet the following criteria:
          (1) situated north of 60°N,
          (2) extend back in time to at least 1500 CE,
          (3) have an average sample resolution of no coarser than 50 years,
          (4) include at least one chronological reference point every 500 years,
          (5) exhibit a documented temperature signal, and
          (6) are published in peer-reviewed literature as a proxy for temperature, although not necessarily calibrated to temperature (i.e., some records provide only a relative measure of temperature with unknown transformations between the proxy measurement and temperature).”

          If you believe in ex ante rules, they are following them.

        • Steve McIntyre
          Posted Oct 15, 2014 at 11:14 AM | Permalink

          Nick Stokes has cited the PAGES2K ex ante rules. I should have checked his quotation earlier as, once again, he’s opportunistically cut off a relevant sentence from the quotation. Here’s the next sentence, omitted by Stokes:

          In some cases the data were modified according to the original authors’ suggestions. For example, if the authors suggested that a record is compromised by human influence during some time period, these intervals were removed from the proxy record

          This is presumably the basis for PAGES2K’s truncation of the Korttajarvi, Lehmilampi and Nautajarvi records from the versions archived by the authors. In fact, the Igaliku authors directly state that their record “is compromised by human influence” during the last millennium in several different statements:

          Anthropogenic influences, which began with the Norse colonization around 1000 AD, preclude paleoclimatic interpretation during the last millennium.

          Consequently, the response to climate change over the last millennium was overprinted by land-use effects (Gauthier et al. 2010; Massa et al. 2012; Perren et al. 2012).

          and in the dataset readme:

          Since ~1 cal ka BP the climatic-driven changes were overprinted by the human influence of Norse and recent agriculture.

          Obviously, the inclusion of the last millennium of the Igaliku dataset is not in compliance with this policy. Had Stokes accurately quoted the SI, we would have saved some time.

          Interestingly, the authors’ comments in respect to Igaliku appear to me to be considerably more categorical than author remarks in connection with Nautajarvi citation (Ojala et al 2005, pdf). In looking through some of the FInnish literature, it also appears that pollen accumulation rates have been calculated for other sites. I did not notice any reports thus far indicating order-of-magnitude increases in modern pollen accumulation rates in uncontaminated sites.

        • Posted Oct 14, 2014 at 8:58 PM | Permalink

          I answered “what is your point?” above:

          https://climateaudit.org/2014/10/11/new-article-on-igaliku/#comment-736057

          Sorry about that.

          Steve: if you wish to contact the authors, it’s a free world. As with Mann et al 2008, the issue is that the multiproxy academics have used the data inappropriately. Kaufman has already admitted this for at least part of the Igaliku core. The procedural problem is that Kaufman did not issue a corrigendum, but acknowledged and (partly) corrected the error without a corrigendum and without withdrawing the erroneous data. Nick has obfuscated the issue by misrepresenting the original article, by failing to provide the quotes from the original authors stating that the series could not be used for paleoclimate of the last millennium.

        • Steven Mosher
          Posted Oct 14, 2014 at 10:39 PM | Permalink

          if its contaminated then its NOT a temperature signal.
          that is what the chorus sang, Nick.

          The chorus didnt make the argument you said they did. Liar.

        • Posted Oct 15, 2014 at 1:34 AM | Permalink

          “Nick has obfuscated the issue by misrepresenting the original article, by failing to provide the quotes from the original authors stating that the series could not be used for paleoclimate of the last millennium.”

          This gets bizarre. You provided a misleading and irrelevant quote about multiproxy. Kaufman used POLLENSUM, not a multiproxy. I provided a quote (from the same section) with them saying that pollen could be used, and spelling it out. I’ll say it again:
          “Juniperus/Alnus PAR,after ~4.5 cal kaBP, and Betula/Salix PAR, after ~3 cal ka BP, turned out to be valuable proxies of past summer temperatures. As in other locations in south Greenland, a more signicant cooling is recorded ~3 ka BP, with a decreasing pollen accumulation culminating with the end of Little Ice Age. This long-term cooling trend was reversed during the 1920s by recent warming.

          They are components of the series Kaufman used. The authors are using pollen as a temperature proxy. Right there.

        • Posted Oct 15, 2014 at 2:12 AM | Permalink

          My way of combining the apparently contradictory statements to remove the contradiction is:

          1) They note generally, and applying also to pollen, that human influence precludes the use of the proxy as climate indicator.

          2) They observe in spite of (1) that the pollen proxy has signals of right sign for MWP, LIA, and modern rise at around the right time telling that climate has probably contributed to the observed variability (my addition: either directly or indirectly by influencing first human activities).

          3) From the combination of (1) and (2) we learn more about the formation of the proxy than about climate.

          If the total input to multi-proxy analyses is so weak that accepting a proxy based on the type of support as given above, then they are really of little evidential value. Then they are still in the exploratory phase, where the properties of the approach are studied, but the results should not be given much weight.

        • Posted Oct 15, 2014 at 4:10 PM | Permalink

          Steve:
          “Had Stokes accurately quoted the SI, we would have saved some time.”
          You had only to brush up on the last thread, where you quoted the same text with the same argument. And the answer is the same. Kaufman did not use the proxies that the authors said were affected, or the multiproxy which you like quoting references to. He used pollen accumulation.

          I’ll quote my response from then:
          “It isn’t that they “did not concede”. It’s that they explicitly expressed their view that it wasn’t compromised.
          “Despite the possible influence of land use, pollen accumulation appears to document climatic changes of the last millennia nonetheless.”
          There’s no way you can construe that to be a suggestion that the data needed to be modified, which is what their “Tiljander clause” would require.”

        • Posted Oct 15, 2014 at 9:07 PM | Permalink

          > Nick has obfuscated the issue by misrepresenting the original article, by failing to provide the quotes from the original authors stating that the series could not be used for paleoclimate of the last millennium.

          In a thread where all we have about a “new paper” is an abstract, no less.

          Let’s see how this claim fares by looking at the relevant quote:

          The multiproxy approach used in this study highlights the interplay of ontogenetic processes (e.g. marine isolation, catchment maturation, soil and lake development, plant colonization) with overarching climatic factors (e.g. early Holocene aridity, Neoglacial cooling). Anthropogenic influences, which began with the Norse colonization around 1000 AD, preclude paleo-climatic interpretation during the last millennium.

          The authors do not state anything about “the series”. The sentence before “preclude” refers to their “multiproxy approach”. The paleo-climatic interpretation may very well refer to the multiproxy approach. According to this reading, what is precluded from a paleo-climatic interpretation during the last millenium is the multiproxy approach the authors use.

          This reading reconciles the authors’ statement that:

          Despite the possible influence of land use, pollen accumulation appears to document climatic changes of the last millennia nonetheless.

          How can the authors imply that each of their proxies are precluded from a paleo-climatic interpretation during the last millenium, and at the same time state that pollen accumulation appears to document climatic change of the last millenia?

          Something’s amiss. Let’s not ask the authors. It’s a free world, after all.

          ***

          There are two readings on the table. One misidentifies a statement with an implication, conflates “the multiproxy approach” with “their proxies”, and is contradicted by one of the authors’ statement. The other is Nick’s.

          But it is Nick who obfuscates the issue by misrepresenting the original article.

          A more appropriate title for what is going on may very well be “Sliming Nick Stokes”.

  6. Nic Stokes
    Posted Oct 11, 2014 at 3:13 PM | Permalink

    “So what you’re saying is that including contaminated measurements is OK because it wasn’t addressed BEFORE the analysis?”

    I can only quote the start of a CA post, just before the Igaliku post:

    “Perhaps the greatest single difference between being a “real climate scientist” and policies recommended here is that “real climate scientists” do not hesitate in excluding data ex post because it goes the “wrong” way, a practice that is unequivocally condemned at Climate Audit and other critical blogs which take the position that criteria have to be established ex ante: if you believe that treeline spruce ring widths or Arctic d18O ice core data is a climate proxy, then you can’t exclude (or downweight) data because it goes the “wrong” way.
    This seems trivially obvious to anyone approaching this field for the first time and has been frequently commented on at critical blogs.”

    BEFORE = ex ante. Then we had a chorus at CA demanding that pollen at Igaliku be excluded because it goes the “wrong” way.

    I’m not an enthusiastic for the rigidity that CA is demanding, and that Kaufman tried to follow, and is now apparently relaxing.

    In fact, the nitrogen runoff issue was considered before, by Massa et al. They thought pollen would not be compromised. Seems reasonable.

    Steve: Nope – nobody demanded that Igaliku be excluded because it went the “wrong way”, but because it was contaminated as a proxy by agricultural runoff, just as Mann’s upside down Tiljander had been. Kaufman ought to have known of the risks of such contamination and established protocols.

    • Steven Mosher
      Posted Oct 11, 2014 at 8:38 PM | Permalink

      “BEFORE = ex ante. Then we had a chorus at CA demanding that pollen at Igaliku be excluded because it goes the “wrong” way.”

      Lie.

      there was not a chorus demanding this. There was a chorus demand that contaminated proxies not be used.

      There is not a lot of comments there

      Why lie?

      more black box climateballing

      • Nic Stokes
        Posted Oct 11, 2014 at 10:32 PM | Permalink

        “There was a chorus demand that contaminated proxies not be used.”
        And how, of all the proxies in the study, do we know it should not be used? The authors said: “Despite the possible influence of land use, pollen accumulation appears to document climatic changes of the last millennia nonetheless.”

        Steve tells us how:
        “The contaminated series is readily identified as an outlier through a simple inspection of the data.”/i>

        “One of these series (top left – Igaliku) has the classic shape of the contaminated Finnish sediment series”

        We know it’s contaminated – just look at it!

        Steve: Nick, you’re once again misrepresenting. Here’s the full quote here

        The contaminated series is readily identified as an outlier through a simple inspection of the data. The evidence of contamination by recent agriculture in the specialist articles is completely unequivocal. This sort of mistake shouldn’t be that hard to spot even for real climate scientists.

        You left out the 2nd and 3rd sentences. This is deception on your part. I didn’t say “We know it’s contaminated – just look at it!” I said that one could tell that it was an outlier by looking at it.

        Proper data analysis – something that is obviously unfamiliar practice to ClimateBallers – requires examination of outliers. In this case, having identified the series as an outlier, consultation of the specialist literature showed that it was contaminated. The contamination was demonstrated by the specialist literature, not the shape itself. Notwithstanding the author’s weak association with climate, the actual evidence, as summarized in my post, showed beyond cavil that the series had been contaminated.

        Indeed, the contamination is so beyond dispute that even Kaufman has conceded on this point. But any competent data analyst should have been cognizant of the contamination in the first place, particularly given Kaufman’s previous experience with contaminated sediments.

        C’mon, this is getting annoying.

        • Nic Stokes
          Posted Oct 11, 2014 at 11:27 PM | Permalink

          “In this case, having identified the series as an outlier, consultation of the specialist literature showed that it was contaminated.”
          The specialist literature you cited consisted of papers by Massa et al. And they are the ones saying:
          “Despite the possible influence of land use, pollen accumulation appears to document climatic changes of the last millennia nonetheless.”

          Aside from that. your evidence consisted of plots showing a spike.

          BTW, I don’t disagree that an outlier should be checked, and if necessary excluded. But it is ex post analysis. I think T&H got it right, and probably Kaufman has too.

        • Steve McIntyre
          Posted Oct 12, 2014 at 9:11 AM | Permalink

          A simple ex ante policy is that varve sites should not have been impacted by agricultural activity. Kaufman should have excluded Igaliku on those grounds ex ante. Had he been more diligent, he would not have been put in a position of examining outliers. But the Igaliku series is clearly an outlier, and with similar properties to other series contaminated by agricultural activity. The primary literature documents that the series was impacted by agricultural activity. Thus, he ought then to have known that his original ex ante policy had not been implemented. Just because a grant-seeking author made a hopeful invocation of climate change doesn’t change the fact that Igaliku had been impacted by agricultural activity and that this made the series non-compliant with a policy against series impacted by agricultural activity. This seems like stuff that high school students can understand and should not be so difficult for climate scientists.

          As to Mount Logan, there are dozens of ice core O18 series used as temperature proxies. There are operational biases that specialists have to be careful with even with polar ice cores: is the core on a summit? has the elevation changed over time? has there been geological uplift?

          Mt Logan was a pristine site – there was no agricultural or similar impact. The only issue was that the data went the “wrong way” without the data itself being contaminated. It meets all ex ante criteria of being a valid ice core.

          Now whenever I’ve given examples of “ex ante criteria”, I’ve always expressed them in objective properties of the site e.g. tree line white spruce or polar ice core O18 or whatever,, or not impacted by agricultural activity, not in terms of opinions of academic authors about whether something is or isn’t a temperature proxy or inclusion in previous multiproxy studies.

          The behavior of the Mt Logan isotopes was unexpected for the authors and they ended up attributing this to changes in source precipitation. This is undoubtedly true. However, it raises that problem that such changes can go in both directions. once this problem has been raised, how do you know that, for example, Austfonna wasn’t impacted by a source change in the opposite direction? You don’t. If you ex post only keep O18 series that go up, then you introduce a very strong bias if there is a form of long-term autocorrelated “noise” arising from regional changes in source precipitation, as there appears to be. This has been repeatedly documented at “skeptic” blogs and, in my opinion, is established beyond any reasonable cavil. However, I realize that most climate academics don’t understand this point.

          The bottom line is this: if you want to use polar ice core O18 isotope series, then you have to use Mt Logan, as well as Austfonna.

          You cannot

          If you accept the authors’ conclusion that ice core O18 was not a temperature proxy in this inst

          An essential aspect of any model (e.g. O18 is a proxy for temperature) is out-of-sample verification. One interpretation of the Mt Logan data is that it showed that there was a flaw in the model. But that is an issue for all the series, not just Mt Logan. If you

          At Mount Logan, the O18 series went the “wrong way”. So what does this imply for the data as a proxy? One possibility is to not use any of the O18 data

          , which the authors attributed to “regional” fluctuations.

        • johnfpittman
          Posted Oct 12, 2014 at 5:42 AM | Permalink

          I don’t think this quote of Steve’s supports your claims Nic.

          Steve: to do proper analysis, I have criticized ex post screening of like proxies and argued for (sensible) ex ante rules. However, I have not blindly endorsed any old ex ante rules. I do not agree with Mannian garbage cans where you grab every series in the ITRDB data base and hope that you can sort things out with multivariate methods. The sort of ex ante rule proposed here has been high-altitude white spruce or Arctic O18 i.e. a class of proxy believed to have a relationship to temperature for physical reasons and with statistical support. I do not encourage the use of ad hoc methods or oddball proxies. To do engineering quality work designed for assessment reports, “pollen sum” would need to be qualified through numerous studies. That seemed obvious in the context of my remarks. Ex ante does not mandate stupidity. If this is too hard for the climate community to understand, no wonder people think so poorly.

        • Steven Mosher
          Posted Oct 12, 2014 at 10:57 PM | Permalink

          More lies nick.
          there was not a chorus.
          The data point was an outlier
          the literature notes contamination.

          Not even Willard can defend your lies.
          climateballer

        • TerryMN
          Posted Oct 15, 2014 at 3:33 PM | Permalink

          Can someone please take Nick’s shovel away?

    • John Bills
      Posted Oct 15, 2014 at 10:25 AM | Permalink

      Ny rapport: Sydvestgrønlands temperatur siden 1784

      http://www.dmi.dk/fileadmin/user_upload/Rapporter/TR/2014/tr14-06.pdf

  7. MikeN
    Posted Oct 12, 2014 at 11:56 AM | Permalink

    There is a conflict of saying you shouldn’t screen proxies for correlation, and at the same time wanting to remove proxies that are unrelated to temperature.

    Steve: if ice core O18 or whatever is a “proxy” for temperature, it is correlated to temperature. You want to prove that that correlation to temperature is a property of the class of data, not one by one on individual examples ex post.

  8. Posted Oct 13, 2014 at 12:22 AM | Permalink

    This particular post and series of comments has given me a bit of an insight as regards the behavior of Team Authors and their tireless water bearer, Nick Stokes.

    Years ago in business I learned a valuable lesson about decision making in a large company or organization. The decision maker will normally choose the MOST JUSTIFIABLE DECISION, even if they don’t believe it is the best decision. They do this to protect their job. If the s@#t hits the fan, they can point to all those justifications as the reason they went down the path, exonerating them self of responsibility for the failure.

    On this thread Steve McIntyre wondered:

    The NOAA archive contains many other measurements: it is unclear why Kaufman selected pollen accumulation rate out of all the available measurements.

    Now this has been answered by Nick Stokes on multiple threads, generally reading something like this.

    Those authors, Massa et al, had said that there was a modern agricultural effect which affected most proxies, but, explicitly, that pollen was still OK. I said their view should be respected in the application of the rule.

    So, while this behavior differs from the corporate decision-making I noted earlier, there are strong parallels.
    Whether caused by noble-cause corruption, confirmation bias, or even dishonesty, there is an appearance an effort to construct a narrative in support of The Cause.
    Decisions to include obvious contaminated outliers pass because the original researchers gave them a hand-waving get out of jail free pass. On this flimsiest of justification, BUT IT IS JUSTIFICATION, in goes the contaminated series. It doesn’t matter how egregious the choice, as long as there is a justification to placate a reviewer or a reporter, and it furthers THE CAUSE.

    • Posted Oct 13, 2014 at 2:49 AM | Permalink

      +1

      • EdeF
        Posted Oct 13, 2014 at 9:57 AM | Permalink

        +2

    • TAG
      Posted Oct 13, 2014 at 6:38 AM | Permalink

      The issue of the acceptability of proxies is not one that should be trivialized. I can remember the discussions of the Tijlander proxy in which much was brought up about the opinion of the original authors as having specialist knowledge. Also the Loehle reconstruction used the opinion of the original specialist authors as it criterion fro inclusion. If that is chosen as an objective rule then, as Nic Stokes points out, the Massa pollen proxy would be accepted.

      What are the criteria that should be used in the selection of proxies?

      Is it the opinion of the original specialist author?

      Is it a consensus of specialists in the field?

      Could it be concluded that a proxy is valid by comparison of its effect in use with other accepted proxies?

      Under what circumstances should a proxy be accepted or not?

      It would seem to me as a an outsider that a set of criteria in regard to proxy validity this would be an important contribution to this field

      • Steve McIntyre
        Posted Oct 13, 2014 at 10:12 AM | Permalink

        I’ve never tried to set down a set of rules. However, in my own thinking about the problem, I’ve emphasized objective properties of the site and the proxy, emphasizing the importance of having enough examples of the proxy to have a population the properties of which can be assessed and discouraging the introduction of “singleton” and “oddball” proxies or poorly understood proxies into multiproxy analyses. For example, in the present PAGES2K South American network, there’s an oddball proxy from Lago Aculeo that has never been used anywhere else, which the author puts forward as a “temperature proxy”. This cannot be placed in the same category as polar ice core O18 series, where there are now dozens of examples, and where biases and inhomogeneities can be appraised. Once a proxy category is established, all the examples have to be considered and explained. Information from original authors is relevant but has to be assessed in the context of the population.

        Any rule that supposedly mandates inclusion of contaminated data such as Igaliku is a stupid rule and should be causing its proponents to ask themselves how they came up with such a stupid rule and what sort of sensible ex ante criteria would have protected PAGES2K against this sort of egregious error. Rather than defending it. In this case, I can think of a couple of ex ante rules. Not using “singletons”. Not using sediment series impacted by modern agriculture – the latter should have been obligatory post-Mann et al 2008 and Kaufman et al 2009. Focusing more about the objective properties of the proxy and the site, than subjective author theories and explanations (as distinct from direct observations.)

        I find myself extremely reluctant to trying to average unlike proxies without plotting them in a multipanel graph and discussing them.

        I also think that there is utility in looking at populations of O18 series as a whole (for example) within a region prior to trying to average them with tree rings and varve thicknesses.

      • Salamano
        Posted Oct 13, 2014 at 10:52 AM | Permalink

        It seems to me that the criteria are:

        (a) Specialist author declares said proxy type(whether they were the one to first obtain the samples) to be a ‘valid temperature proxy’

        then

        (b) Obtained in a region that has been ex ante determined to be a key region potentially sensitive to larger climate-related temperature (like a Yamal, for example).

        then

        (c) Contains the key identifier of the post-1850 temperature-record uptick, or at the very least a huge chunk of it, whether then declining after 1960 or not.

        So far, those seem to me to be the criteria in use. Unfortunately, it also seems near-curious that some proxies are culled from a place like (b), mined for appearances like (c), then declared to be (a). Or perhaps strung through a methodology that makes these such samples 95% of the meat, whilst the others are down-weighted off the table.

  9. Posted Oct 14, 2014 at 4:28 AM | Permalink

    Having read the argumentation between Steve and Nic my impression is that both agree on the main principles:
    – Seriously faulty records must be excluded from the analysis.
    – Doing that without introducing bias is far from trivial.
    – Properly formulated set of rules and applying these rules effectively is needed to reach the goals in an acceptable way.

    To see, how these principles influence interpretation of the Massa et al paper I tried to pick the directly relevant parts keeping enough text to explain the context.

    Starting from Synthesis and conclusions the first paragraph is (emphasis mine)

    The multiproxy approach used in this study highlights the interplay of ontogenetic processes (e.g. marine isolation, catchment maturation, soil and lake development, plant colonization) with overarching climatic factors (e.g. early Holocene aridity, Neoglacial cooling). Anthropogenic influences, which began with the Norse colonization around 1000 AD, preclude paleoclimatic interpretation during the last millennium.

    That appears clear. The next paragraph includes at the very end of the paper, however, the sentences

    Juniperus/Alnus PAR, after ~4.5 cal ka BP, and Betula/Salix PAR, after ~3 cal ka BP, turned out to be valuable proxies of past summer temperatures. As in other locations in south Greenland, a more significant cooling is recorded ~3 ka BP, with a decreasing pollen accumulation culminating with the end of Little Ice Age. This long-term cooling trend was reversed during the 1920s by recent warming.

    Looking at the the earlier chapter Cooler neoglacial phase (after ~3 cal ka BP) we read after discussion of human influences

    Despite the possible influence of land use, pollen accumulation appears to document climatic changes of the last millennia nonetheless. PAR reached minimum values during the Little Ice Age from 1500 to 1920 AD, consistent with maximum glacial readvance at Qipisarqo (Kaplan et al. 2002) and elsewhere in south Greenland (Weidick et al. 2004; Larsen et al. 2011). It is also coeval with high rates of isostatically driven transgression, which caused the inundation of a Norse graveyard at Herjolfsnæs (Mikkelsen et al. 2008). The sharp increase of Salix/Betula pollen accumulation rate after 1920 AD (Fig. 6) suggests a rapid warming, which reversed the Neoglacial cooling trend similar to other locations in the Arctic (Kaufman et al. 2009).

    From the above we can see that observing the expected climate signal is used as the main argument for concluding that there’s a climate signal in the proxy. Making such an argument is correct, when evidence is searched for deciding, whether the proxy has a climate signal, but it’s absolutely unacceptable, when the proxy is used to study climate.

    Objective criteria for accepting or rejecting a proxy into a multi-proxy analysis must not be based on the data that goes into the analysis, except that outliers must be picked for further scrutiny and processing, which may or may not lead to their rejection. Other factors must be used in the selection. Typical factors concern the nature of the proxy (variable studied, location of trees etc.), technical observations from the empirical work and data processing, and also behavior of the proxy time series over periods that are not included in the final analysis either for calibration or in other way. When information is used in screening, using it in further analysis leads unavoidably to bias, how severely depends on the case.

    The optimal strictness in applying these principles depends on the availability of high quality data. The more such data is available, the stricter criteria are optimal. Here we are clearly near the other end. Therefore it’s justified to perform analyses relaxing the criteria significantly, but then it must also be understood that the outcome from the analysis is not reliable. The research has in many ways the nature of exploratory research rather than confirmatory research.

    Steve: a recommendation for multiproxy assessment that I’ve made frequently but is ignored by Stokes etc is not to use “singleton” proxies. If such proxies show consistent behaviour over multiple sites, then they can become the topic of analysis and assessment. I’ve also consistently urged that criteria be framed objectively in terms of the properties of the proxy and site, rather than theories or hypotheses by the authors.

    As a further reminder to Stokes and others, here is a direct image of Pekka’s quote (from here):

    massa excerpt 1

    • Nic Stokes
      Posted Oct 14, 2014 at 5:30 AM | Permalink

      Pekka,
      I generally agree. Remember, though, that Kaufman did state a set of objective rules. I’ve relayed them in a comment in moderation, but in my first comment upthread, I linked to where I said them at WUWT. Worth considering here.

      Steve: any ex ante rule which requires the use of contaminated data is obviously a stupid rule and points to a lack of competence. In the wake of Kaufman’s prior corrigendum, ex ante rules which did not screen for agricultural impact were incompetent. In this case, the authors reported a record over the Holocene and believed that the earlier part of the record contained climatic information, but that the recent record had been impacted by agricultural activity. The fault is entirely PAGES2K.

      • Craig Loehle
        Posted Oct 14, 2014 at 10:48 AM | Permalink

        Nic, your comment reminds me of the rule that led to a young boy being sent home because he chewed his pop-tart into a gun shape and “shot” someone with it. Any such rigid rule is “stupid” to quote Steve.
        In an experiment using mice, if you found out one of your techs was feeding them candy, might you not drop those mice? or would you “stick to your protocol”?

    • Craig Loehle
      Posted Oct 14, 2014 at 10:52 AM | Permalink

      If the authors feel that the proxy contains SOME climate signal (you can tell it got cooler over 1000 years) that does not mean a detailed reading of the signal is available or advised. One must also use one’s own brain. Whether sediment or pollen, the “signal” is contaminated in a big way at two historical periods. There is no method available to subtract out this contamination.

    • Nic Stokes
      Posted Oct 14, 2014 at 3:44 PM | Permalink

      “As a further reminder to Stokes and others,”
      The section you have highlighted refers to the multiproxy aggregation, and the issues they have identified with diatoms etc, which you noted in your original quote. It certainly does not preclude the use of pollen as temperature proxy. That very same section ends with (my bold):

      “Juniperus/Alnus PAR,after ~4.5calkaBP, and Betula/Salix PAR, after ~3 cal ka BP, turned out to be valuable proxies of past summer temperatures. As in other locations in south Greenland, a more signicant cooling is recorded ~3 ka BP, with a decreasing pollen accumulation culminating with the end of Little Ice Age. This long-term cooling trend was reversed during the 1920s by recent warming.

      You want agricultural influence to be included in the ex ante rules. But there are all sorts of things that could be wrong with a proxy. And a decision of some sort has to be made. For example, here a sheep farm was established after 1980. Does that invalidate the entire record? All proxies? Can the very reasonable argument by Massa et al that pollen counts (borne by wind from far away) are not affected by agricultural run-off not even be heard?

      The point of Kaufman’s ex ante rule is that he delegates these decisions to the original authors and their journal referees. They know about agricultural effects too. Maybe better than you.

      Steve: Nick, it is unfortunate that it is impossible to rely on anything that you say. Your statement that “here a sheep farm was established after 1980” is a total deception, since there are multiple statements that sheep farming was established at Igaliku from the 1920s on, though the severely cold weather of the 1970s (“climate crisis”) caused the death of thousands of sheep. (This is the period when Stokes says that there was a heat wave at Igaliku.)

      Massa et al, A multiproxy evaluation of Holocene environmental change from Lake Igaliku, South Greenland states:

      Agriculture in the region resumed in the 1920s and current farming consists of two farms that breed 400–450 sheep (Miki Egede, pers. commun.). One of these is established in the catchment and manages 30 ha of hay meadows on the fields surrounding the lake (Fig. 1).

      Igaliku was resettled during the 18th century (Arneborg 2007) and large-scale agriculture, based on sheep farming, was developed in the 1920s (Austrheim et al. 2008). Consequently, the response to climate change over the last millennium was overprinted by land-use effects (Gauthier et al. 2010; Massa et al. 2012; Perren et al. 2012).

      As sheep farming was reintroduced at a large scale in the area during the 1920s, we have also compared medieval and recent soil erosion to place the Norse impacts in a modern context

      Another related article, A 2500 year record of natural and anthropogenic soil erosion in South Greenland” states:

      The modern grazing management began in 1915 with 250 sheep ( Austrheim et al., 2008 ) and was intensified in 1924 under the impetus of the Danish government ( Hansen, 1991).
      Until 1976, sheep farming was based on extensive all year
      round grazing, with small supplements of winter fodder and
      occasional stabling. The number of sheep in South Greenland
      increased up to a maximum of 47800 in 1966 but heavy snow and strong frost during the winter 1966/67 starved to death nearly 60% of the herds. Similar disasters occurred in the winters of 1971/72 and 1976/77 ( Greenland Agriculture Advisory Board, 2009 ). Consequently, a new plan for sheep management was presented ( Egede, 1982 ): henceforth, sheep should graze 5 months on the outlying fields and be stall fed during the 7 months of the Greenland winters, requiring a considerable amount of fodder. Nowadays, there is one single farm in the catchment of Lake Igaliku and the lake surroundings are freely grazed by a few hundred sheep during the summer season.

      A thid article, “Lake Sediments as an Archive of Land Use and Environmental Change in the Eastern Settlement, Southwestern Greenland” states:

      A continuous, high-resolution sedimentary record from the deepest part of the lake provides an assessment of farming-related anthropogenic change in the landscape, as well as a quantitative comparison of the environmental impact of medieval colonization (AD 985–ca. AD 1450) with that of recent sheep farming (AD 1920–present). After centuries of abandonment, the same areas were re-occupied by farmers using conventional agricultural methods beginning in the early 20th century

      Modern pastoral agriculture began in 1915 (Austrheim et al. 2008). During the 20th century, the village of Igaliku and surroundings were central to the re-development of contemporary agriculture in southwestern Greenland. Since 1980, after the climate crisis of the 1960s/1970s (Egede and Thorsteinsson 1982, Greenland Agriculture Advisory Board 2009), two sheep farms (more than 1000 sheep) were established in the catchment, and
      around 30 ha of hayfield were created on the shore of
      the lake to produce winter fodder for stabling (Figs.
      2a, 3).

      After ca. 1950, trees and shrubs decrease, weeds, apophytes, and coprophilous fungi increase, all in response to the new phase of grazing pressure and the re-establishment of farming activities (IGA2e).

      • Posted Oct 14, 2014 at 10:24 PM | Permalink

        “Steve: Nick, it is unfortunate that it is impossible to rely on anything that you say. Your statement that “here a sheep farm was established after 1980? is a total deception, since there are multiple statements that sheep farming was established at Igaliku from the 1920s on,”

        My statement is very well established. Igaliku, which you are talking about, and Lake Igaliku are not the same thing. Massa et al said, quoted above:
        “From the beginning of the 20th century to ˜1960 (± 5 yr), corresponding to the re-establishment of farming at Igaliku, none of the sedimentary parameters reveals any significant increase of erosion around the lake of Igaliku (Fig. 6). Since ˜1960, the grazing pressure (shown by a decline in woody taxa and a rise in coprophilous fungi [Fig. 5]) caused a progressive increase in detrital parameters. Soil erosion accelerated beginning in 1969 (± 4 yr), with a sharp increase in Ti and C:N ratio, reaching ˜11 mm century-1 in 1988 (± 2.5 yr), slightly more than during the medieval period.” And they said, as I cited:
        “Since 1980, after the climate crisis of the 1960s/1970s (Egede and Thorsteinsson1982, Greenland Agriculture Advisory Board 2009), two sheep farms (more than 1000 sheep) were established in the catchment, and around 30 ha of hayield were created on the shore of the lake to produce winter fodder for stabling (Figs. 2a, 3). “

        Now there seems to have been some grazing some years before the establishment, but it was the farm itself (OK 2 farms post 1980) that made the big difference.

        But the real question I raised was – who decides? Does a modern sheep farm invalidate some part of the record? How much? Does prior grazing do so? Who is better placed to decide? The palynologists who were actually there and know stuff? Kaufman? Or bloggers at CA?

        Steve: Does modern sheep farming invalidate some part of the record? Yes, of course. There is no inconsistency between CA and the original authors on this point. Only between CA and Kaufman/PAGES2K. It is disquieting that specialists are so incompetent that they make such absurd errors.

        • Tony Hansen
          Posted Oct 15, 2014 at 4:10 AM | Permalink

          ‘Igaliku, which you are talking about, and Lake Igaliku are not the same thing’.
          Nick may be correct on this point. The sheep grazed aound Igaliku but not in Lake Igaliku 🙂

        • Carrick
          Posted Oct 15, 2014 at 6:36 AM | Permalink

          Tony, possibly you’re mistaken. 😉

        • Carrick
          Posted Oct 15, 2014 at 6:37 AM | Permalink

          (Note this body of water is just a proxy for Lake Igaliku.)

        • ehac
          Posted Oct 15, 2014 at 8:09 AM | Permalink

          I have missed something. Somewhere. The pollen producing sheep.

          Perhaps Steve can explain?

        • Craig Loehle
          Posted Oct 15, 2014 at 8:59 AM | Permalink

          Why sheep “matter”: any grazed area will produce different pollen than ungrazed areas because sheep or any livestock eat things they like (imagine that) and also trample vegetation. They can also muddy the water and change runoff. The pollen effects will travel a good distance.

        • Posted Oct 15, 2014 at 9:00 AM | Permalink

          Nick wrote, “Igaliku, which you are talking about, and Lake Igaliku are not the same thing.”

          Map here:
          https://www.google.com/maps/place/Igaliku,+Greenland/@60.9925236,-45.4505808,4724m/data=!3m1!1e3!4m2!3m1!1s0x4eabdc8096723ffd:0x40bb08a7d0b162

        • ehac
          Posted Oct 15, 2014 at 11:30 AM | Permalink

          A result of grazing sheep is more pollen. Because they eat pollen-producing plants.

          Nice.

          And pollen will increase in muddy water. Wonder where the pollen originated in the first place.

        • TAG
          Posted Oct 15, 2014 at 12:04 PM | Permalink

          ehac wrote

          Wonder where the pollen originated in the first place

          As quoted by SMc above from Massa et al, A multiproxy evaluation of Holocene environmental change from Lake Igaliku

          around 30 ha of hayield were created on the shore of the lake to produce winter fodder for stabling

          For the non-metric a hectare (ha) = 2.5 acres. So about 30 hectrares or 75 acres of hayfield are described as growing to maturity to provide hay for winter fodder. Perhaps the ecologists who read this blog would be willing to comment on that with respect to changes in the pollen in the area. My area has severe winters as well and large areas of land are repeatedly harvested for hay throughout the summer. two years ago, there was a severe drought and the hay shortage extended to night time theft of hay bales left in the fields after harvest.

          Steve: the issue with this (and other similarly contaminated/overprinted) sediments are modern increases in erosion, not increases in pollen generation. The huge increases in erosion result in vast increases in accumulation rate for everything, including pollen.

        • Carrick
          Posted Oct 15, 2014 at 5:15 PM | Permalink

          I agree with Craig and Steve about the effects of erosion and run off. But I have a feeling it’s more complicated that this.

          Start with this: sheep have hooves and they eat vegetation.

          When the sheep walk on the surface of the Earth (as opposed to when they are properly trained to fly), or when they munch on the plants, the sheep damage the plants.

          Damaged and distressed plants flower to attempt to reproduce before (potentially) they die, hence more pollen.

          If the plant is left undisturbed, it will expend more energy into growth.

          But if the plant is disturbed, either by breaking limbs or (as most gardeners are aware) by deliberate pruning, you will see more flowers. If you don’t prune the plant, some plants will produce few or no flowers on a given year.

          More flowers = more pollen.

          If there are sufficient sheep, you’ll also see replacement of woody species with herbaceous ones. Herbaceous plants have a shorter life cycle, and this also leads to more flowering and more pollen production.

          Feces from the sheep also change the amount of bioavailable nutrients.

          And their hooves increase the roughness of the surface and can actually allow more water to penetrate around the roots of plants.

          Unfortunately this is a field that seems to have few open publications, and few publications that are covered by my library. But what I can get ahold of doesn’t suggest a clear picture either way.

          The main thing though is grazing clearly can affect plant composition and growth patterns. This amounts to an anthropogenic contamination of the pollen signal.

          But if the anthropogenic contamination doesn’t lead to clear predictions that can be used to interpret the anthropogenic signal that is now present in the pollen, it seems blatantly obvious you can’t use this proxy, during the period of contamination, to study the climate signal alone.

        • Steve McIntyre
          Posted Oct 15, 2014 at 10:23 PM | Permalink

          Sheep grazing in the 19th century US had a very large impact on the landscape. Apparently they eat herbs right to the roots. That sheep grazing would impact fragile meadows in Greenland is certain.

        • Carrick
          Posted Oct 16, 2014 at 4:27 AM | Permalink

          Steve, I think there were issues with over-grazing of sheep and cattle in the US, which have had long-term consequences for the landscape. However, I think it is generally agreed that the impact of sheep farming is relatively limited compared either to cattle or to pig farming. Cattle ranching in particular leads to soil compaction, and pigs root and actively destroy tuberous plants. The prohibition against raising and eating pork in arid regions probably traces back to this fact.

          I also don’t doubt that Greenland has seen an impact from the introduction of sheep. But I believe it is actually agreed in the ecological community that the major impact of sheep farming in regions like Greenland is the loss of birch forests and their replacement by meadows. Similar issues have been noted in the US Southwest.

          If anything, I would think there are now more meadows, since the introduction of sheep. Soil erosion in forests is known to be significantly much smaller than in non-forested landscapes, so the increase in erosion in recent decades is certainly more complicated than simple denudement of the soil by e.g., over-grazing.

          However we also need to keep in mind that in addition to the introduction of sheep, there was an overhunting of muskox in Greenland to the point of near extinction as well as timbering and firewood gathering, also contributed to deforestation. Thus the impact of settlers is neither linear, and certainly not limited to the impact of sheep farming.

          I think it is important to point out that pollens of different plant species are easy for botanists to distinguish, at the difficulty level that amounts to a high-school science fair project (on par with leaf identification and being more challenging only in requiring a microscope).

          I believe it is the case that much of the interest in the original sample collections were not to study climate but to study the changing impact of human activity on the local ecology, which you can document by looking at the changing ratios of different plant species over time.

          This suggests that many core samples are consciously collected in areas that are known to be disturbed areas. The argument that the anthropogenic disturbance of the intrinsic climate signal is limited to some portion of the 20th century is not likely to be valid for these samples. Nor is it the case that the climate signal is expected to be limited to temperature.

          This discussion reminds me of why I don’t like these sorts of compendiums of large number of different samples. It takes quite a bit of work to take a single time series and turn it into a usable proxy for climate studies, especially if you’re interested in a temperature-limited response.

          But Igaliku Lake should be a particularly ignominious example. If we start with the published geographic location, 61º 00′ N, 45º 26′ W then plot it using Google Maps:

          It doesn’t take a high level of geographic proficiency to work out that this is not a good location choice for a temperature proxy that extends from the modern period to the to earliest documented time of the core sample.

          However, the use of this core sample as a temperature proxy was not the stated goal of the individuals who collected and studied this core. Rather it was to “describe the major environmental changes that affected Lake Igaliku and its catchment since the last glacial retreat”.

          Basically this is not a climate proxy. Rather it is a catchment proxy.


          Steve: yup. Anthony also put a similar figure online in discussion of this series last year. needless to say, Nick Stokes objected to any challenge to the immaculate-ness of the proxy. The original interest in this site was historical and archaeological – it was apparently where Erik the Red settled.

        • Steve McIntyre
          Posted Oct 16, 2014 at 10:59 AM | Permalink

          Carrick, I looked at a number of historical discussions of 19th century grazing and overgrazing in the US Southwest in connection with our 2005 discussion of bristlecones. My impression was strongly that specialist opinion was that sheep preceded cattle and pigs into these areas in the 19th century (though the 20th century pattern was different.) John Muir of the Sierra Club called sheep “hoofed locusts” and Carl Purpus, a prominent 19th century botanist, said that sheep had denuded all vegetation to the top of some ranges. Erosion increased in areas of sheep grazing/overgrazing leading to some dispute in the field over the relative roles of overgrazing and climate. See references in McIntyre and McKitrick 2005 (EE). Also work by Craig Allen e.g. http://landcover.usgs.gov/luhna/chap9.php

          When one sees an increase in erosion contemporary with the introduction of sheep grazing, only a real climate scientist would shut his eyes to the impact of sheep grazing.

        • Carrick
          Posted Oct 16, 2014 at 4:52 AM | Permalink

          Here’s a reference on the impact of plant coverage on erosion:

          http://www.fao.org/docrep/t1765e/t1765e0h.htm

          Looking at the history of musk ox in Greenland, the population loss from South and West Greenland long predates recent settlement from Europe. Whether there is even a role of humans in the population diminishment seems to be controversial.

        • Tony Hansen
          Posted Oct 16, 2014 at 5:48 AM | Permalink

          Ahh, Carrick. You’ve done me.
          Is it only the recently shorn sheep that graze in the water?

        • Carrick
          Posted Oct 16, 2014 at 7:02 AM | Permalink

          Tony, I can imagine the shorned sheep might prefer the water. 😉

          By the way, if you use that geographic coordinates and put it into Google Earth (and enable pictures), you can find an image of the farm (it’s to the west of the supplied coordinate by about a mile):

          Also, the nonlocal nature of the catchment (one issue brought up by Nick) does not rescue it from being influenced by human activity. In fact, it’s quite the opposite.

          What this means is, if you want to look at a catchment that is not affect by anthropogenic activity, you need to collect the data in a non-develoiped area. In other words, you can’t argue that your proxy was uninfluenced by surrounding activity due to some “sharpshooting ability” on the part of the people who originally collected the proxy.

        • Posted Oct 16, 2014 at 8:19 AM | Permalink

          Carrick,

          This photo seems to fit most convincingly with the satellite pictures of the lake

          http://www.panoramio.com/photo/12426959

          To me the place is not the same as in the photo you linked.

          Based on satellite pictures there’s only one farm near the lake

        • Steve McIntyre
          Posted Oct 16, 2014 at 10:39 AM | Permalink

          That photo was shown last year in the WUWT article http://wattsupwiththat.com/2013/05/01/the-stokes-kaufman-contamination-protocol-a-sticky-wicket/ on the incident as it stood in 2013.

          A year later, Kaufman has conceded that the modern portion of the series was contaminated, though his purported correction was incomplete, leaving Stokes and Willard as the lone defenders of the validity of the series. Of course, Kaufman has previous experience in issuing corrgendums for contaminated sediment series.

          In terms of Stokes’ and WIllard’s strained exegesis of the Massa et al text, it’s amusing to re-read the corresponding articles on the Finnish lakes (Nautajarvi, Lehmilampi, Korttajarvi) where PAGES2K truncated contaminated modern portions, as the caveat by Igaliku authors was just as clear, if not clearer, than anything in these other articles.

        • Posted Oct 16, 2014 at 11:29 AM | Permalink

          The photo on shown at WUWT is not the same that I linked but it shows the same area at a different time of the year. The aerial view shown at WUWT is not from the right location but from the Eastern side of the ridge. Comparing the maps shown in the papers and Google maps shows clearly that the lake is at the center of this page

          https://www.google.fi/maps/@61.0050798,-45.4449107,2477m/data=!3m1!1e3?hl=en

          The agricultural area is south of the lake.

          Steve: check the link at the end of the WUWT article – looks like the same photo to me/

        • Posted Oct 16, 2014 at 4:44 PM | Permalink

          Here is a map from a Massa et al paper

        • Carrick
          Posted Oct 16, 2014 at 8:29 PM | Permalink

          Pekka, yes there is only one farm (of indeterminate size), but there are many lakes.

          The farm is at the eastern end of Nunataaq Sheep Farm Road (the western edge ends up in Qassiarsuk).

          The picture I showed is the Nunataaq Sheep Farm main complex. This is the link:

          http://www.panoramio.com/photo/14643819?source=wapi&referrer=kh.google.com

          to the image I showed.

          If you get on GoogleEarth (the app not the web browser) it is easy to demonstrate that this picture is accurately located and correctly depicts the Nunataaq Sheep Farm.

          The location of the sample 61º 00′ N, 45º 26′ W given by Maasa is to the west by 2-km from this farm.

          The lake shown in the picture I provided is definitely not the same lake as bore-holed by Maasa, unless their geographical location is way off.

        • Carrick
          Posted Oct 16, 2014 at 8:43 PM | Permalink

          Pekka, I found a copy of Massa and looked at it, and I agree with you now. I’m at a hotel with very poor internet and I think Google Earth wasn’t firing on all cylinders.

          Nunataaq Sheep Farm is located well north of this location.

    • Kenneth Fritsch
      Posted Oct 14, 2014 at 9:52 PM | Permalink

      “The research has in many ways the nature of exploratory research rather than confirmatory research.”

      In my view this is very much the problem with most of climate science doing temperature reconstructions. In the hard sciences where one can explore and then confirm under controlled conditions the exploration is more or less a tool and confirmation is the end product, while with temperature reconstructions the exploration appears as the end product.

  10. barn E. rubble
    Posted Oct 14, 2014 at 9:54 PM | Permalink

    RE: Steve: “Nick, it is unfortunate that it is impossible to rely on anything that you say.”

    O.K. Obviously I’m not the only one tried of this. I can’t help but think Nick’s comments are nothing but chain-yanking posts for either personal pleasure &/or on assignment, and he sits back giggling with each reply he elicits. Whether ‘Nic-Picking’ (or better Nic-Stocking) the minutia of a post, or arguing over every possible meaning of every word in a post, he is looking for a response. When I see a ‘Recent Comment’ by those I like to read, I will click on the link. I’m now tired of finding those links are replies to the Race Horse’s latest misrepresentation &/or misquote &/or another take to schooling episode. Is it not time to snip those before posting? I think our Host and Moderators have been more than patient and courteous. I mean so those that have other things to do don’t have to deal with such an annoyance, and the rest of us won’t have to read about it either . . . just saying.

    Steve: when Nick was just racehorsing, it was one thing. But now he frequently and consistently misquotes and misrepresents and (what Mosher calls) “lies”. I’d prefer to be challenged than not challenged, but it is tiresome dealing with the inaccuracies. Obviously not “everything” is a misrepresentation, but the ratio is far too high to be useful.

    • Posted Oct 15, 2014 at 5:52 PM | Permalink

      I doubt if Nic is an official member of the Crusher Crew (a group united by the internet to introduce or repeat alarmist dogma in climate blog comment threads). He’s been doing it too long.

      However, he may have served as a role model.

      In any event, it becomes obvious that he’s Nic-picking after a few exchanges and can be ignored. It’s a pity that Mr. McIntyre is forced to respond to Nic as lack of response is considered admission of error by Nic.

      The comments here are not the issue. The baldfaced lies Nic writes at other venues are.

    • Steven Mosher
      Posted Oct 16, 2014 at 5:40 PM | Permalink

      Over the course of years, what nearly 7 now, Nick has always, in my opinion, gone up to the line, but never crossed it. That is, he’s pushed his lines of argument on points that I consider arguable. That’s good as it challenges people to make tight cases. I don’t mind the racehorse approach. But recently– I’d point to the black box incident and the present discussion– he has crossed the line from pressing a case to misrepresenting to lying or just plain making crap up. Folks don’t need to speculate as to why this is the case. That’s a distraction. But he has switched from forcing folks to tighten up their cases to a mode where he makes cases for his own position that are as twisted as anything Goddard has ever done. Imagine that.

      • Posted Oct 16, 2014 at 7:43 PM | Permalink

        One thing has changed – I am effectively semi-booted. Comments made under my WordPress ID go straight to spam. Comments with no ID (dropping a k for extra anonymity) are now going invariably into moderation, where they stay a long time. Under those circumstances, it is impractical to engage in dialogue.

        • TerryMN
          Posted Oct 17, 2014 at 11:45 AM | Permalink

          Under those circumstances, it is impractical to engage in dialogue.

          What you’re engaging in could be described as several things, but “dialogue” ain’t one of ’em, sorry.

      • Carrick
        Posted Oct 16, 2014 at 9:34 PM | Permalink

        S. Geiger, critical commentary from people like Nick improves the quality of the content of blogs like this. If he’s being paid, it should be from the skeptics group.

        • Layman Lurker
          Posted Oct 17, 2014 at 10:57 AM | Permalink

          Ditto that. As frustrating as Nick can be, if he motivates those engaged in debate to put meat on the bones of their arguments, then we are all the better for it.

        • Carrick
          Posted Oct 17, 2014 at 9:42 PM | Permalink

          +1 LL

      • thisisnotgoodtogo
        Posted Oct 16, 2014 at 11:58 PM | Permalink

        Comments on behaviour and being paid are turning discussion to a low level.

        • Carrick
          Posted Oct 17, 2014 at 12:28 AM | Permalink

          I agree wholeheartedly. And people can stop with the insults anytime now.

          I also think people need to stop complaining about the presence of comments that are critical to the consensus on a blog.

          That is effectively asking to dumb the conversation down.

      • Carrick
        Posted Oct 17, 2014 at 12:23 AM | Permalink

        I think slowing down the rate of commenting can improve the quality of the comments, so there is that.

        Merely uttering the banned word N**k can get one in moderation these days.

  11. Craig Loehle
    Posted Oct 16, 2014 at 12:25 PM | Permalink

    It is my impression that sheep and goats can denude land more than cattle because they eat more types of plants and because they will pull plants up to eat them.

  12. Nick Stokes
    Posted Oct 16, 2014 at 4:46 PM | Permalink

    Here is a map from a Massa et al paper

    • Carrick
      Posted Oct 17, 2014 at 12:44 AM | Permalink

      Thanks, I was able to find the Massa paper on line and confirm that Pekka is correct on this.

      I think my comment that this should be seen as a catchment proxy, rather than primarily a climate related one, remains a valid comment.

      The comments aren’t completely related to the commentator with initials NS whose name must not be uttered. They are just my reflections. Sometimes they might be seen relevant to his comments, and other times, possibly not. I think their validity has nothing to do with whether he would agree or disagree with them though.

      If you wanted to select proxies for climate change, you should select ones that are far away from agricultural activity. I also think one shouldn’t defend inclusion of very poor quality proxies based on the hypothetical that there may not be better ones to use.

      I’ve actually performed meta-analysis reviews of others work, including one that made it into Nature. This review included the analysis of 50 different experimental measurements. In order to fully understand the data, I had to model each class of experiment independently, and for most, there were unique elements that had to be considered for each experiment.

      As I did this analysis entirely on my own (I actually prefer that model to dividing the work across researchers as was done with PAGES2K), I can say it is entirely a tractable problem to look at each proxy, and to apply a pre-agreed to set of criterion as to whether that proxy is appropriate for inclusion into climate change studies.

      I can also say that simply because somebody has chosen to include it in a database of proxies is absolutely no basis for defending its inclusion. You should look at the data quality criteria for that. If the proxy is not a climate proxy, it’s still not a climate proxy simply because somebody later misclassified it as one.

      Also, the argument that it’s inclusion doesn’t vastly affect the outcome is not even an argument. If it’s not a valid signal, it shouldn’t be included period. It’s not even relevant what effect if any it has on the shape of the reconstruction curve.

      And I strenuously object to arguments of the sort “ooh it’s too hard to do it correctly”. This is absolute BS. Science done properly is hard work. People who don’t want to work hard, should stay in the front office selling toothpaste or whatever else gets them their giggles.

      Finally, people do make mistakes. By the time my work gets into final publication, I like to think there are few major issues remaining, but mistakes get made. Sometimes you get a good referee and the bigger errors get fixed and sometimes you get unlucky and a thumbs-up is inappropriately issued instead.

      The issue here isn’t that mistakes are made. It’s that when the mistakes are made, corrections aren’t being issued. It’s not the mistake, it’s the cover up that’s the real problem.

      Steve: I agree entirely with the above.

      BTW re “whose name must not be uttered” – Nick’s name and comments occur hundreds of times on this blog and is not restricted in any way. When I was being overwhelmed by spam a few months ago (before closing old threads), I added a few non-controversial words to the filter ( including “blog” itself) to try to stem the tide and have revised this list again. Nick has reported that he has been listed by wordpress, but that did not arise from CA.

      • Posted Oct 17, 2014 at 2:36 AM | Permalink

        Much of this thread has been about, how to fight biases in statistical analysis. The issue is so difficult, because fighting some specific type of bias tends to introduce a bias of another type. One example of such new bias is a biased conclusion that the data is insufficient of telling anything, when that’s not really the case.

        Everyone seems to agree that fighting biases requires that rules are set and adhered to. Everyone must agree that rules are of no help unless they are well chosen. Briefly stated generic rules (like the one that the original authors must indicate that there’s a temperature signal) are not good rules as they do not tell, how to interpret such conflicting information that we can see in the Massa et al paper. We should not assume that the choice of words in a free format scientific paper tells unambiguously and correctly, what the authors think about the suitability of the proxy for a specific use. That’s justified only, if they tell it very explicitly and in sufficient detail. Even then it should be accepted that the paper may tell that they have reached this conclusion for reasons that violate the rules we have set for accepting the proxy as part of the analysis.

        When the number of data sets is as low as in PAGES 2k Arctic (probably less than 100 including those that have been rejected without further notice) the scientists who build the database can and must check in detail the information provided by the authors of each study.

        When we are discussing specific cases on the net, it’s likely that we are also biased. We may pick particular studies for scrutiny in a biased way as it’s likely that most (or all) of us check carefully only a fraction of the studies. It’s unlikely that we start by defining in sufficient detail our own set of rules, and that we check in an unbiased way, how all the data sets perform against these rules. My understanding is that Nick is fighting against this bias in net discussion. This kind of bias is certainly to be expected in net discussion on a site like Climate Audit, how ever seriously the host tries to avoid the bias himself.

        Then we have also the bias in perception of the readers of a paper or the net discussion. Discussing in length certain details may give readers of the blog the impression of a much more widespread and essential problem that that particular issue really is.

        • N ιck Stoκes
          Posted Oct 17, 2014 at 7:32 AM | Permalink

          “My understanding is that Nicκ is fighting against this bias in net discussion.”
          That doesn’t sound right. My argument is this:
          1. CA originally voiced a strong demand for ex ante rules, and eschewing ex post. I am not so keen, but see some merits.
          2. Kaufman adopted ex ante (objective) rules. Some concerned scope, but basically it left judgment about whether a proxy was good for temperature to the authors.
          3. Ex post arguments were then used at CA to criticise this.

          To me, Kaufman’s rule made sense. He is simply saying that the judgment of the scientists on the spot (and their referees) will be on average more reliable than his. They may make errors, but the bias and errors created by overruling will be greater.

          Of course, there is then scope for argument about what the authors are actually saying. But it is easier to resolve these (well, maybe not here) than it is to redo the science.

          Now I’m not passionately claiming that this is the best way overall. But ex ante is what CA advocated, and I think it not be so easily thrown aside.


          Steve: as I’ve repeatedly said, ex ante does not require stupid. Ex ante rules that supposedly require inclusion of contaminated data were stupid in the first place and should not have been proposed by a competent specialist. This ought not to be difficult to understand.

        • Steve McIntyre
          Posted Oct 17, 2014 at 8:00 AM | Permalink

          Pekka, a thoughtful comment with a couple of different issues.

          On proxy selection, as with many topics, it’s easier to say where people have gone wrong than it is to establish alternative rules.

          In an assessment report, such as IPCC and perhaps PAGES2K, I think that there is an obligation to assess the proxies and to categorize based on objective properties of the site and proxy, rather than subjective assessments of the authors who have invested time and energy in their site.

          A common sort of rule is that “the original authors must indicate that there’s a temperature signal”. You’ve mentioned the problem that arises when there are conflicting comments by the original authors. But I think that there are other reasons in relying on such “indications”.

          I think that it would be much more helpful if assessments summarized the current position of the community on the validity of various proxies in common use, together with the biases of each one. I.e. Igaliku pollen accumulation rate wouldn’t be included in an assessment because the authors advertised the proxy, but because there had been enough pollen accumulation rate series developed and published so that the properties of the data could be studied as a group. It is evident that sediment accumulation rates are profoundly impacted by agricultural activity. This problem is well known outside of Igaliku (three other PAGES2K series were curtailed on this basis). I would say that one would need at least 10 such series before feeling confident in their properties. Since pollen accumulation rate is under discussion, I’ve located some other series and will try to comment on them.

          Another example in discussion is polar ice core O18 series, the workhorse proxies in deep time. If the specialist community believes that these are proxies, what happens if an individual series goes the “wrong way” in the 20th century e.g. Mt Logan and the site authors say that it is therefore not a temperature proxy at their site. I strongly object to the idea that specialists can just discard this adverse series and proceed as though nothing had happened. (Any more than that they can legitimately “hide the decline” in MXD series and substitute infilled data as in Mann et al 2008 and the earlier notorious WMO diagram.) If specialists continue to believe in the validity of the proxy as a type, then, at a minimum, they have to contemplate the presence of highly persistent “noise” (regional circulation patterns) that might persist long enough for an individual series to go the “wrong way” during the calibration period. But once you admit that possibility, then specialists have to chin up to the possibility of overshoot in the opposite direction. If you only reject the downward outliers and don’t also reject the upward outliers, you get a bias.

          Nick Stokes has vigorously argued the opposite: that you should include contaminated Igaliku and exclude uncontaminated Mt Logan, consistent with the practice of PAGeS2K and specialists, but, on this point, I strongly believe that both he and PAGES2K are wrong.

          You comment that “discussing in length certain details may give readers of the blog the impression of a much more widespread and essential problem that that particular issue really is.” You’ve noticed – more than most – that the true sample size of the paico method is much smaller than advertised. The true sample size of most multiproxy studies is even smaller. When you combine that with repetitive use of proxies with known (and questionable) properties (Graybill bristlecones, Briffa’s Yamal), then the actual knowledge base of the field is much smaller than advertised.

          For example, authors studying Lago Aculeo in Argentina recently studied its reflectance properties over time, claiming that their PC1 (or something like that) was a proxy for temperature. This is a novel and singleton proxy that was included in the SOAMER PAGES2K report and, because it was one of only a few long series, had a huge impact on its medieval period.

        • AndyL
          Posted Oct 17, 2014 at 9:28 AM | Permalink

          N ick
          Can you clarify, do you think Kaufmann was correct to include Igaliku because of his ex ante rules, correct to then (partly) remove Igaliku, or both?

          I have a horrible suspicion your answer will be “both”.

        • Posted Oct 17, 2014 at 9:40 AM | Permalink

          In the specific case of the lake Igaliku proxy there are more than one reason to leave it out based on the stated rules.

          We have discussed in many comments, what the Massa et al paper states about its applicability. My view is that their statements give strong reasons for leaving it out rather than including it in the analysis of past temperatures.

          In addition it contains 14 values for the first 1000 years and 20 values for the second 1000 years. The times of these samples are irregularly spaced. Thus some time gaps are rather long. It clearly fails on another stated requirement: have an average sample resolution of no coarser than 50 years as that would require 40 rather than 34 values over the full period.

          Steve: Pekka, the failure of this proxy to meet the resolution criteria has been pointed out on numerous occasions to Nick, but he has in the past ignored or disdained this simple observation and continued to say that the proxy meets the ex ante rules anyway.

        • taget
          Posted Oct 18, 2014 at 7:55 AM | Permalink

          Yes, the failure of the proxy to meet the ex ante resolution test seems clear cut. However, the existence of adverse facts has never served as an impediment to Nick in the past. I can’t wait to see what argument he and Willard contrives to racehorse out of this.

        • kim
          Posted Oct 18, 2014 at 10:55 AM | Permalink

          willard’s pounding along several lengths behind, and riderless.
          ==========

        • AndyL
          Posted Oct 18, 2014 at 11:35 AM | Permalink

          Slightly OT but possibly relevant to the use of proxies.

          In the context of a discussion at Climate Dialogue, Mike Lockwood – a mainstream scientist specialising in solar – begins his post with seven “general points” that he considers “fundamental”. The first five all seem relevant to selection and use of proxies to generate global trends.

        • AndyL
          Posted Oct 18, 2014 at 11:36 AM | Permalink

          I forgot to give a link:
          http://www.climatedialogue.org/wp-content/uploads/2014/10/Blog-Mike-Lockwood-def.pdf

        • Steve McIntyre
          Posted Oct 18, 2014 at 1:47 PM | Permalink

          I don’t want this comment to serve as a coatrack for solar discussion, which is better done at a site interested in the topic. To the extent that the article supports proper statistical testing, I agree. However, it does seem to me that the climate community is more supportive of such tests when applied against solar correlations than when applied against bristlecones or contaminated sediments as magic thermometers.

        • Posted Oct 18, 2014 at 1:29 PM | Permalink

          AndyL,

          All those 5 points are relevant. All of them have also been widely discussed in scientific literature on proxy-based climate reconstructions. People have developed numerous methods that to take those points better into account.

          When the set of available proxies is too small to allow taking all those points properly into account, the weight given on these considerations seems, however, to be limited. That’s justified up to a point, but going beyond that point happens easily. Determining objectively, when that has happened, may be impossible based on the available data. Stringent criteria set for safe statistical analysis are certainly often violated, but it can be argued that such criteria may be unnecessarily strict.

  13. Posted Oct 16, 2014 at 5:20 PM | Permalink

    I believe there are many who say the desertification of so much of the Middle East, Sahel and even some parts on the edges of the Sahara is down to the direct and indirect impact of the domestication and increased population of sheep and goats.

  14. Posted Oct 16, 2014 at 5:21 PM | Permalink

    Desertification is too strong a word. Environmental degradation probably more appropriate.

    • Tony Hansen
      Posted Oct 17, 2014 at 5:53 AM | Permalink

      I do not agree Tom. You were correct the first time. Desrtification is a process, it takes time (and it is degradation) but ’tis hard to see in the moment.
      Added to that is what at best is a poor basic dataset.
      Our understanding of what grows above the soil surface is really not that strong. Our understanding of what is happening below the soil surface is orders of magnitude less.

%d bloggers like this: