The hockey stick: a retrospective - Climate Change: The Facts - page 152

directories at Mann's FTP site (the 'CENSORED' directories), which, through detective
work,werefoundtocontainassessmentsoftheimpactfromdroppingthebristleconesfrom
the underlying data. In light of the claim in Mann et al., 27 this should not have made any
difference, but it did. In our NAS presentation we showed graphs of the data in Mann's
'CENSORED' results, in which the hockey stick shape completely disappears. That is,
even applying Mann's biased methods, after dropping the few bristlecone pine series there
is no remaining hockey stick shape. The claim in Mann et al. about robustness to the
exclusion of the tree ring data was obviously misleading.
In the letter from the Congressional Oversight committee to Mann he was asked:
Did you run calculations without the bristlecone pine series referenced in the article and, if so, what was the result? 28
Mann's answer was lengthy, but included the following:
For a complete scientific response, you should consult the article my coauthors and I published back in 1999 addressing
precisely these issues … As my co-authors and I explained … given the proxy data available at that time, certain
key tree-ring data … were essential, if the reconstructed temperature record during early centuries were to have any
climatologic “skill” (that is, any validity or meaningfulness). These conclusions were of course reached through analyses
in which these key datasets were excluded, and the results tested for statistical validity. Our conclusions have been
confirmed by Wahl and Ammann. 29
Translation: Yes. When we removed them the graph collapsed and the statistical scores
went to zero. Oh dear, didn't we mention that? Anyway, to avoid the problem, we kept
them in.
Mann's claim that the 1999 paper addressed 'precisely these issues' was misleading. In
thatpapertheydidmentionthattheirtop-weightedPCwas'essential'buttheydidn'treport
the results of excluding the bristlecones. 30 Instead they applied a 'correction' that they
claimed (without proof) fixed the contamination pattern in the bristlecones, even though it
only applied to the nineteenth century portion. And their 2000 paper claimed robustness
both to contamination of bristlecones and removal of tree ring data. 31 Wahl and Ammann
later offered the argument that since the hockey stick fails all statistical tests without the
bristlecones they ought to be retained (the logic really was that bad). 32
In our letter to North we pointed out that we agreed with Wahl and Ammann (and
Mann) that the reconstruction without the bristlecones is no good.
But, we added:
Our contention is that the reconstruction with bristlecones is also no good, as evidenced by the failure of verification r 2
and CE statistics. 33