More findings by Dr. Wegman:
Some of the issues have been addressed in the text of the description of the methodologies in the earlier part of our discussion. For completeness and clarity we will summarize again here.
1. In general we found the writing of MBH98 somewhat obscure and incomplete. The fact that MBH98 issued a further clarification in the form of a corrigendum published in Nature (Mann et al. 2004) suggests that these authors made errors and incomplete disclosures in the original version of the paper. This also suggests that the refereeing process was not as thorough as it could have been.
2. In general, we find the criticisms by MM03, MM05a and MM05b to be valid and their arguments to be compelling. We were able to reproduce their results and offer both theoretical explanations (Appendix A) and simulations to verify that their observations were correct. We comment that they were attempting to draw attention to the deficiencies of the BH98-type methodologies and were not trying to do paleoclimatic temperature reconstructions.
3. As we mentioned earlier, it is common for data analysis and pattern recognition experts to use a training sample for calibration. Normally one would seek to have the training data to be representative of what one might expect from the entire dataset. Because the temperature profile in the 1902-1995 is not similar, because of increasing trend, to the millennium temperature profile, it is not fully appropriate for the calibration and, in fact, leads to the misuse of the principal components analysis. However, the narrative in MBH98 on the surface sounds entirely reasonable on this calibration point, and could easily be missed by someone who is not extensively trained in statistical methodology. Dr. Mann has close ties to both Yale University and Pennsylvania State University. We note in passing that both Yale University and Pennsylvania State University have Departments of Statistics with excellent reputations. Even though their work has a very significant statistical component, based on their literature citations, there is no evidence that Dr. Mann or any of the other authors in paleoclimatology studies have significant interactions with mainstream statisticians.
4. In response to the letter from Chairman Barton and Chairman Whitfield, Dr. Mann did release several websites with extensive materials, including data and code. The material is not organized or documented in such a way that makes it practical for an outsider to replicate the MBH98/99 results. For example, the directory and file structure Dr. Mann used are embedded in the code. It would take extensive restructuring of the code to make it compatible with a local
machine. Moreover, the cryptic nature of some of the MBH98/99 narratives means that outsiders would have to make guesses at the precise nature of the procedures being used.
5. As mentioned in our introduction, much of the discussion on the ‘hockey stick’ issue has taken place on competing web blogs. Our committee believes that web blogs are not an appropriate way to conduct science and thus the blogs give credence to the fact that these global warming issues are have migrated from the realm of rational scientific discourse. Unfortunately, the factions involved have become highly and passionately polarized.
6. Generally speaking, the paleoclimatology community has not recognized the validity of the MM05 papers and has tended dismiss their results as being developed by biased amateurs. The paleoclimatology community seems to be tightly coupled as indicated by our social network analysis, has rallied around the MBH98/99 position, and has issued an extensive series of alternative assessments most of which appear to support the conclusions of MBH98/99
.
7. Our committee believes that the assessments that the decade of the 1990s was the hottest decade in a millennium and that 1998 was the hottest year in a millennium cannot be supported by the MBH98/99 analysis. As mentioned earlier in our background section, tree ring proxies are typically calibrated to remove low frequency variations. The cycle of Medieval Warm Period and Little Ice Age that was widely recognized in 1990 has disappeared from the MBH98/99 analyses, thus making possible the hottest decade/hottest year claim. However, the methodology of MBH98/99 suppresses this low frequency information. The paucity of data in the more remote past makes the hottest-in-a-millennium claims essentially unverifiable.
8. Although we have not addressed the Bristlecone Pines issue extensively in this report except as one element of the proxy data, there is one point worth mentioning. Graybill and Idso (1993) specifically sought to show that Bristlecone Pines were CO2 fertilized. Bondi et al. (1999) suggest [Bristlecones] “are not a reliable temperature proxy for the last 150 years as it shows an increasing trend in about 1850 that has been attributed to atmospheric CO2 fertilization.” It is not surprising therefore that this important proxy in MBH98/99 yields a temperature curve that is highly correlated with atmospheric CO2. We also note that IPCC 1996 stated that “the possible confounding effects of carbon dioxide fertilization need to be taken into account when calibrating tree ring data against climate variations.” In addition, as use of fossil fuels has risen, so does the release of oxides of nitrogen into the atmosphere, some of which are deposited as nitrates, that are fertilizer for biota. Thus tree ring growth would be correlated with the deposition of nitrates, which, in turn, would be correlated with carbon dioxide release. There are clearly confounding factors for using tree rings as temperature signals.
9. Based on discussion in Mann et al. (2005) and Dr. Mann’s response to the letters from the Chairman Barton and Chairman Whitfield, there seems to be at least some confusion on the meaning of R2. R2 is usually called the coefficient of determination and in standard analysis of variance; it is computed as 1 – (SSE/SST). SSE is the sum of squared errors due to lack of fit (of the regression or paleoclimate reconstruction) while SST is the total sum of squares about the mean. If the fit is perfect the SSE would be zero and R2 would be one. Conversely, if the fit of the reconstruction is no better than taking the mean value,
then SSE/SST is one and the R2 is 0. On the other hand, the Pearson product moment correlation, r, measures association rather than lack of fit. In the case of simple linear regression, R2 = r2. However, in the climate reconstruction scenario, they are not the same thing. In fact, what is called ß in MBH98 is very close what we have called R2.
10. We note here that we are statisticians/mathematicians who were asked to comment on the correctness of the methodology found in MBH98/99. In this report we have focused on answering this question and not on whether or not the global climate is changing. We have discussed paleoclimatology only to the extent that it was necessary to make our discussion of the statistical issues clear. The instrumented temperature record makes it clear that global temperatures have risen since 1850 CE. How this present era compares to previous epochs is not clear because the uncertainties in the proxies. However, it is clear that average global temperature increases are not the real focus. It is the temperature increases at the poles that matter and average global or Northern Hemisphere increases do not address the issue. We note that according to experts at NASA’s JPL, the average ocean height is increasing by approximately 1 millimeter per year, half of which is due to melting of polar ice and the other half due to thermal expansion. The latter fact implies that the oceans are absorbing tremendous amounts of heat, which is much more alarming because of the coupling of ocean circulation to the atmosphere. (See Wunsch 2002, 2006). |
