By Craig Loehle, Ph.D.
Since I am mentioned in Mann’s book, The Hockey Stick and the Climate Wars: Dispatches from the Front Lines and not kindly, it seems necessary to set the record straight. I do not take credit for any angry posts that follow.
Mann begins discussion of my work on page 187 (see bottom of this post) with the title “The Hockey Fight Continues” where my work is categorized as another assault from the denialosphere, with me being part of the “Hydra” that is hatefully out to get Mann. Simply because I published a paper that does a reconstruction and expressed a view that tree rings might have issues (which the Climategate emails show was a hidden view of many in the field) I was engaging in a “fight” against Mann? Really? I think it is because they didn’t like my results, which is rather post-modernist, no? So, I am lumped in with politically motivated and evil “deniers” and “denialists”. I find these terms and the entire context for discussing my work offensive. I am not a “denialist” and my recent paper (Loehle, C. and N. Scafetta. 2011. Climate Change Attribution Using Empirical Decomposition of Historical Time Series. Open Atmospheric Science Journal 5:74-86) attributes about 40% of recent warming to human activity – estimating that this equates to a no-feedback atmosphere.
What I would deny is that tree rings are good thermometers, but this is a scientific view based on my knowledge of trees, not a political view.
I point this out in the paper Mann criticizes and also in (Loehle, C. 2009. A Mathematical Analysis of the Divergence Problem in Dendroclimatology. Climatic Change 94:233-245), which interestingly was published while Steve Schneider was ed in chief. As a final note on my work on this problem, I showed in (Loehle, C. 2005. Estimating Climatic Timeseries from Multi-Site Data Afflicted with Dating Error. Mathematical Geology 37:127-140) that combining time series with measurement error or especially dating error will flatten out peaks like the MWP and troughs like the LIA. This means that comparing the MWP peak temperature, which is likely smeared (damped) to recent annual temperatures will show recent temps warmer simply due to data resolution (the warmest years are not averaged out in the recent data). It is an apples and oranges comparison.
I am not part of a conspiracy and am not directed by anyone. I have never received money from fossil fuel interests, as Mann states is true of all sceptics. On the contrary, I work for the US wood products industry (which has no official position on climate change and does not tell me what to say or think). This industry is the largest single renewable fuels user because it uses wood waste to generate steam and power at paper mills. It is also the largest recycler (of paper), plants millions of trees every year, and manages millions of acres of forest land to ensure protection of wildlife. Only a fraction of my work concerns climate change.
My disagreements with the use of tree rings (by anyone, not just Mann) have nothing to do with a conspiracy, are not organized or directed by anyone, and are not personal. I just think tree rings (especially strip bark) are not valid more than about 100 years back in time no matter how fancy the statistics used.
In general also I would like to defend the implicit charge that I am unqualified to enter the arena with Mann and dispute him. I studied forestry and ecological modeling, so I know a few things about trees and models. More than many of the dendro people. I have published 138 peer-review papers in ecology, forestry, statistics, control theory, topology, landscape ecology, evolutionary theory, animal behavior, hydrology, and psychology. I have a bunch of papers where I model forest growth and other dynamic biological systems. 28 of my papers are on climate or ecosystem responses to climate change. I have studied and modeled photosynthesis and tree growth processes.
Mann notes that I got the dates wrong on a few series in my original paper (Loehle, C. 2007. A 2000 Year Global Temperature Reconstruction based on Non-Treering Proxy Data. Energy & Environment 18:1049-1058). Some of the first papers I found use the year 2000AD as 0, and count BP back from there. Others used calendar dates. Most used 1950 as 0 year but did not necessarily say so in the archives, they just start with 0 in the data file with no explanation. Mann’s assertion that everyone dates from 1950 as year 0 for BP is simply not true. So during the period when I gradually pulled the data together, I got some of this mixed up. Unfortunate. Wish I had Steve McIntyre’s memory. On the other hand, I got the corrected version in print in record time (thanks to E&E being very understanding). This benefitted from Hu’s input (Loehle, C. and Hu McCulloch. 2008. Correction to: A 2000 Year Global Temperature Reconstruction based on Non-Treering Proxy Data. Energy & Environment 19:93-100). Mann mentions the correction but does not discuss it. This is what he should have discussed since it is in the record. I would point out that Mann has never publicly admitted or corrected any mistake. As an ironic aside, it was Gavin who found my dating errors and notified me and was helpful in resolving them. I thanked Gavin for his help in print.
Mann complains that I could not make a comparison of recent with the MWP with “such a paucity of records” in footnote 44—referring to recent decades particularly. When I was pulling data together for my paper, many authors did not respond to requests for their data, and one refused. Many interesting papers had not archived their data. So I was able to find 18 data sets that seemed least problematic. This is more than the 11 low frequency series in Moberg et al. 2005 in Nature, which IPCC uses and the Team were coauthors on (9 of which I used). The oxygen isotope record that Mann objects to was used in Moberg. But Mann exaggerates how much data he himself uses. When you look at the weighting of series by Mann’s methods, the effective number of series in early centuries is often only a few or even 1. It doesn’t matter how much data is in the hopper if a couple of series have all the weight. SM has documented this exhaustively at CA. Furthermore, the point of the paper was not a claim that I had correctly reconstructed climate history, but rather that non-tree ring data does not agree with tree ring data. I think the paper showed this but Mann completely missed this point.
Mann focuses on the MWP vs recent temperature differential in my paper and says the important part is the post-1950 (post mid-century) trend. But he could only get his strong recent warming uptick in his papers by reflecting and smoothing past the end of the data. This has been parsed extensively at CA and I explicitly avoided smoothing past the end of the data for obvious reasons.
Mann repeats that his work is valid because it has been “independently replicated”. A cluster of papers with overlapping authors and heavily overlapping data is not any sort of independent test. Furthermore, it is widely known that Mann’s results depend strongly on a few key proxies like stripbark pines,which are damaged trees and should not be used. When a few proxies are heavily weighted, it does not matter much what the other data are.
So much material! Mann also says
By contrast with the hockey stick studies-and every other peer reviewed scientific article on the subject-Loehle claimed that medieval warm period temperatures were warmer than “20th century values.”
Now this is a curious comment because Mann seems unaware of a very large literature. Just counting published studies that enable quantitative estimates to be made, Craig Idso at http://www.co2science.org/data/mwp/quantitative.php published the following summary of about 100 studies:
So here we have a statement by Mann that NO literature exists suggesting a warmer MWP when in fact there are scores of them. I think Mann only views Team publications as being valid, but we are not required to share his view. So his claim that “every other…article” refutes my work is simply false. Whether you like how Idso did his analysis or not, there IS literature out there on this question. The famous Soon and Baliunas paper was also about this very question, and Mann surely remembers them? Oh, yes, and we remember the reaction to S&B.
Mann states that E&E is not peer reviewed. This is untrue and the editor of E&E has publicly refuted this charge repeatedly. My paper was peer reviewed in the same manner as every other journal I ever submitted to, except that the reviewers were able to keep their political biases under control.
Mann says I should have done “statistical validation emphasized in all serious paleoclimate reconstruction studies”. In the studies he mentions, an elaborate calibration is used that leads to a need for “validation” but I was using simple averaging of studies which independently estimated a temperature model for each data set. There was nothing to “validate” but Mann may not understand that point.
Mann objects to some records not being sufficiently time-resolved and that I “improperly” filled in, but Moberg did the identical infilling that I did (linear interpolation) for the same reasons, as have others, including Mann himself. I explicitly stated in my paper that the time resolution limitations meant that decadal details could not be resolved wth my results. There is no sense in which this is “invalid” it just means that fine temporal scales will be screened out. Some of the data used in any case represent a time window rather than a particular year. If the goal is to use non-tree ring proxies, there are none that are annual, so Mann’s requirement can’t be met.
Mann ends by praising me for a more rigorous approach than many contrarians, but this is an insulting comparison because he is essentially saying that for a shill I did a better than usual job. Furthermore, I don’t accept that my work “didn’t stand up to the scrutiny” because it was never, to my knowledge, addressed in subsequent literature. Of 32 citations in Scholar Google, 5 are in other languages and only one, that of Swanson, attempts to criticize my paper (a criticism to which I replied adequately, I believe) Just because the “Team” didn’t like it in private does not mean it did not stand up to scrutiny (remembering the Team admonition that only published peer-reviewed work counts). I repeat that I do not believe that my reconstruction is “right”, only that it shows that non-tree ring records give a different result than tree ring records. There are many problems with the various proxies, including dating error, calibration issues, data continuity, data time resolution, measurement error, local geology changes over time, etc, but the problems are in many cases more subject to resolution than for tree rings which face, I believe, insurmountable problems (e.g., see the new paper Brienen, RJW, E Gloor, & PA Zuidema. 2012. Detecting evidence for CO2 fertilization from tree ring studies: The potential for sampling biases. Global Biogeochemical Cycles 26 GB1025).
The last sentence is a dig at those who criticize from blogs and don’t make their own reconstructions (guess who he means!). While I am sure SM can defend himself, I would note that someone who points out that a perpetual motion machine is impossible is not obliged to build one, nor is any critic obliged to engage in an activity that they view as ill-conceived or incorrect. Mann exhibits a misunderstanding of the nature of science. If something is wrong it is wrong.
In summary, Mann’s setup for discussing my work is borderline libel, as it implies many things about me that are false and detrimental to my reputation. It is unacceptable to portray those who disagree with you scientifically as evil and politically motivated. Science is full to the brim with disagreements about everything, from which treatment is best for coronary blockage to whether frequentist or Bayesian methods are best. By Mann’s logic, we should all be using slanderous language to refer to anyone who disagrees with us. I don’t think so. There are also multiple objectively false statements in the section discussing my work.
Mann’s section on Loehle starting page 187:
The Hockey Fight Continues
The gaze of the Hydra remained largely focused, however, on the denialists’ bete noir, the hockey stick. Despite the fact that the NAS, the IPCC, and more than a dozen independent peer reviewed scientific studies had now not only reaffirmed the key conclusions of our work, but in fact extended them further back in time, the denialosphere was still fond of claiming that the hockey stick had been “totally discredited” or “broken.” Most of the attacks represented some version of the myth that the hockey stick was a statistical artifact, combined with a studied neglect of the numerous confirmatory independent studies. Some of the attacks were new, however.
In late 2007, the home journal of climate change denial, Energy and Environment, published a paper by Craig Loehle that purported to present a new two-thousand-year reconstruction of global temperature.39 By contrast with the hockey stick studies-and every other peer reviewed scientific article on the subject-Loehle claimed that medieval warm period temperatures were warmer than “20th century values.” Had the paper somehow identified key new sources of information or a more appropriate methodology to overrule the findings of all other recent studies?
The paper, in fact, suffered from serious problems that would presumably have been identified had it been submitted to a peer reviewed scientific journal and reviewed by individuals with the relevant paleoclimatological expertise.4o Loehle was evidently unaware of the dating convention in paleoclimatology that in “BP” (nominally, “before present”), “present” actually refers to the standard reference year ofA.D. 1950. By assuming that “BP” instead meant “relative to A.D. 2000,” Loehle erroneously shifted many of his records forward by fifty years, in essence portraying the warmth of the records in the mid-twentieth century as if it pertained to the end of the twentieth century. This error thus had the effect of erasing all of late-twentieth-century warming. Most paleoclimate reconstructions, including the original hockey stick, show peak peak medieval warmth to be comparable to that of the early and mid-twentieth century. It is only the late twentieth century that stands out as anomalous.
Among other problems, many of the sediment records Loehle used in his analysis had chronologies that were determined by just a few radiocarbon dates distributed over the past two thousand years. The dating in these records is consequently uncertain by as much as four hundred years or so, precluding their use in reconstructing centennial
timescale temperature variations.41 There were several records that Loehle wrongly assumed to reflect temperature but instead reflected some other quantity,42 and he inappropriately averaged records that had different temporal resolutions. Loehle did issue a correction that appears to have dealt with some of the most glaring problems, but the fundamental problem remained that his estimates were insufficient to allow for a meaningful comparison of past and present global temperatures.44 Yet even so, had he performed the critical step of statistical validation emphasized in all serious paleoclimate reconstruction studies, and had he published the work in an actual scientific peer reviewed journal, the paleoclimate community might not have so readily dismissed it.
Loehle’s approach was laudable by comparison with that of many of the contrarians. He did attempt to make a positive contribution, putting his own reconstruction out there to be scrutinized and criticized. While the reconstruction didn’t stand up to the scrutiny (and the venue for its publication was dubious), he made an attempt to contribute to the scientific discourse in a meaningful and constructive manner. This can be contrasted with many others who are more than happy to take potshots at peer reviewed studies from their blogs but are unwilling to produce a reconstruction themselves, or even to provide evidence that genuinely contradicts the current scientific consensus that recent warmth is anomalous.