Climate Audit

One of the central claims of The Trick, if not the most central claim, was that “hiding the decline” was nothing more than an inopportune phrase about a single diagram.

It wasn’t. 

The “trick to hide the decline” was an inopportune, if revealing, phrase, but rather than the issue being limited to a single diagram, the inconsistency between the Decline (in observed tree ring widths and densities) and the Hockey Stick temperature reconstructions (primarily based on tree ring widths) was, together with the looting of the Baghdad Museum, the issue that inspired my original examination of Michael Mann’s Hockey Stick and was the driving theme of Climate Audit from its origin up to Climategate.  There are dozens, even hundreds, of Climate Audit articles that, in one way or another, relate back to the conundrum arising from the inconsistency of the underlying proxies and the superficial consistency of the reconstructions.

In this and a couple of follow-on articles, I’ll illustrate the centrality of The Decline vs The Stick in the controversies in the years prior to Climategate.  For the benefit of people that may be new to these disputes, I re-iterate that I never interpreted the late 20th century decline in ring widths as evidence of a decline in temperatures, but as a seriously problematic inconsistency for “reconstructions” relying in large part on tree rings.

April 2003

When I say that the Decline inspired my original examination of Michael Mann’s Stick, it is literally true.

I had become mildly interested in climate issues in late 2002 when the Canadian government was promoting the Kyoto treaty, including in its promotion the assertion that 1998 was the “warmest year” in 1000 years.  This was based on the 2001 IPCC Assessment Report, which included multiple versions of the Hockey Stick graph, one of which was the following:

In early April 2003 (while the Iraq invasion was going on), I read Briffa et al (2001, JGR) and noticed its passing reference to an “anomalous decline in tree density measurements over recent decades”.  I wondered why this effect had not been observed and/or reported in Mann et al 1998. On April 3, 2003, I posted a comment on this conundrum at a Yahoogroups chatline. In this comment, I also observed that there wasn’t anything exceptional in the “non-tree-ring proxy series” used by Mann. (I was referring here to data from Mann et al 1999 then available online at UMass. This was the small subset of Mann et al 1998 proxies used in the AD1000 reconstruction step. The data for the much larger MBH98 dataset was then not publicly available.)

Here’s what I wrote (at the very dawn of my interest) in a comment entitled “Briffa on post-1960 tree rings”:

Keith R. Briffa, Timothy J. Osborn, Fritz H. Schweingruber, Ian C. Harris, Philip D. Jones, Stepan G. Shiyatov, Eugene A. Vaganov, 2001, Low-frequency temperature variations from a northern tree ring density network, JGR, 106 D3 (16-Feb-2001) 2929-2941, http://www.cru.uea.ac.uk/cru/people/briffa/jgr2001/, “We used simple linear regression, fitting the regression equations over the period 1881-1960, or over the total available period prior to 1960 when the instrumental record was shorter. The period after 1960 was not used to avoid bias in the regression coefficients that could be generated by an anomalous decline in tree density measurements over recent decades, that is not forced by temperature [Briffa et al 1998b: Nature 391, 678-682]”

At face value, this looks to me like quite peculiar statistical methodology … As to Mann’s methodology, I must confess that I am unable to understand it. But in any event, if Briffa observed the above effect, then (1) Mann should have observed the effect also and his failure to observe it would diminish the value of his study; (2) if Mann observed the effect, but failed to report it, then that would equally diminish the value of his study. Either way, I’d be inclined to rely on Briffa’s evidence as to observations and to rely on neither series as a proxy reconstruction for the obvious reasons above.

I have examined most of the non-tree-ring proxy series used by Mann e.g. Quelccaya O18 and accumulation, etc. and, through inspection only, have seen no evidence of 20C variables breaking new ground, although they are on the warm side.  I’ve posted up previously a note on O18 in a Greenland snow-pit. I collected this information specifically as an O18 measurement in the 1990s. 

Over the next few days, I wrote several further comments in the Briffa on post-1960 tree rings thread. In one comment, I compared the need to look at underlying data to the requirement that independent geologists look at drill core in mining projects (a key step in due diligence that was omitted in the Bre-X fraud.)

On April 8, 2003, I sent my initial request to Michael Mann for the FTP location of Mann et al 1998 data in order to look at the underlying data.

April 8, 2003 was also the date that looting began in Baghdad. On April 9-10, 2003, the Baghdad Museum, a totally unique repository of cuneiform tablets from early world civilization, was looted as a result of appalling negligence of cultural heritage by the US occupation, which devoted resources to protecting Toyota Land Cruisers, but not unique world history. I had a considerable prior interest in Bronze Age history and watched the spectacle with both dismay and anger.  I later wrote some (now lost) comments entitled “Day of infamy at the Baghdad Museum”. 

The connection between the looting of Baghdad and my decision to request MBH98 data wasn’t just coincidence. The looting of Baghdad was a radicalizing moment for me: whatever the rights or wrongs of the invasion itself, it was evident that there were no plans on what to do now. It was a mess. Making matters worse, I had begun with substantial skepticism of the WMD rationale and presumed that Iraqi emigres were telling too eager US intel agencies exactly what they wanted to hear.  Was it possible that something similar was going on with climate? In the sense that everyone in climate, like everyone in the days leading up to the Iraq invasion, was too eager to go to war. 

I discussed this issue on a few occasions in some early Climate Audit posts. The primary rationale for the Iraq war was supposed proxy evidence of WMD, such as the now infamous aluminum tubes.  The (then) lead argument from Canadian and other governments – the late 20th century climate uniqueness  of the Hockey Stick – was similarly based on proxy evidence.  I likened my parsing of the proxy reconstructions of temperature to what a pre-war (CIA) analyst ought to have asked in respect to proxy evidence of WMD: does the proxy evidence unequivocally prove the conclusion? Or are there alternative possibilities that would account for the proxy evidence?  Here are a couple of excerpts:

In an emotional debate, I think that there’s an important role for analyzing individual arguments being relied upon. I’ve focused on the multiproxy studies and have come to the conclusion that all the hockey-stick studies are flawed and biased. De-constructing each individual study is very time-consuming. I view this exercise as not dissimilar to that of a pre-war analyst studying proxy evidence for WMD such as aluminum tubes. At the end of the day, an analyst is sometimes obliged to say that maybe an aluminum tube is just an aluminum tube. That does not mean that some other piece of evidence may not be valid – only that the aluminum tube wasn’t.: https://climateaudit.org/2006/06/14/a-few-inconvenient-truths/

Back when views on Iraq were more evenly divided, I sometimes compared what I do to being a CIA analyst arguing that sometimes an aluminum tube is just an aluminum tube and not evidence of WMD. That wouldn’t mean that proponents of the war couldn’t argue the matter using different arguments or that the war was or wasn’t justified, or that the subsequent occupation of Iraq was or wasn’t botched. All it means is that policy-makers shouldn’t be basing their decisions on questionable information about aluminum tubes. This was a line of argument that used to rub right-wing people who liked part of my message the wrong way, but I hope that it says something about me.:  https://climateaudit.org/2007/11/08/why-is-this-left-and-right/

{Nor would] a belated discovery of some other type of WMD in Iraq vindicate Powell’s claim that aluminum tubes were conclusive proxy evidence of WMD.): https://climateaudit.org/2006/08/09/ipcc-and-glaciers/

I’ve been very consistent in viewing the parsing of proxy reconstructions as a very narrow point in the larger climate debate.  In the sense that I’ve never argued that climate policy should be disdained because Hockey Stick studies were flawed. Or that we know that it was warmer in the medieval period than at present. My point was very narrow and consistent with my opening question about the Decline vs the Stick: that there was a fundamental inconsistency between the observed decline in tree ring widths and densities in the late 20th century (Briffa et al, 1998) and temperature reconstructions from proxy networks dominated by tree rings (Mann et al 1998).  This was a scientific question. 

Ironically, I recently read the original 2005 WMD Commission report in connection with the re-consideration of the Russiagate hoax in our “corner” of Twitter.  It turns out that US intel agencies committed more or less exactly the same procedural errors in respect to Danchenko (Steele) as they had previously with Curveball and claims that the Iraq WMD programme had been reconstituted. With one important difference: in the case of Iraqi WMD, there was a concerted and prompt effort to understand how the intel agencies had got their analysis so wrong, whereas in the case of the Russiagate collusion hoax, there was a concerted effort by intel agencies to conceal the origins of the false information and perpetuate the hoax long after they knew otherwise. 

As a closing comment and teaser for subsequent posts, I am well aware that there are now at least several dozen proxy reconstructions that supposedly “vindicate” the MBH Hockey Stick and that a new avatar of the Hockey Stick (PAGES2K, 2019) has returned to its place of glory in the most recent IPCC Assessment Report by inclusion in the Summary of Policy-makers.  But none of these studies squarely confront, let alone resolve, the problems that occasioned my original interest in the field – a point that I hope to return to in this series.