Do some of you remember Steig et al 2009, a pre-Climategate Nature cover story? Like so many Team efforts, it applied a little-known statistical method, the properties of which were poorly known, to supposedly derive an important empirical result. In the case of Steig et al 2009, the key empirical claim was that strong Antarctic warming was not localized to the Antarctic Peninsula (a prominent antecedent position), but was also very pronounced in West Antarctic. Their claims are set out firmly in the opening sentences of their abstract as follows:
Assessments of Antarctic temperature change have emphasized the contrast between strong warming of the Antarctic Peninsula and slight cooling of the Antarctic continental interior in recent decades. This pattern of temperature change has been attributed to the increased strength of the circumpolar westerlies, largely in response to changes in stratospheric ozone. This picture, however, is substantially incomplete owing to the sparseness and short duration of the observations. Here we show that significant warming extends well beyond the Antarctic Peninsula to cover most of West Antarctica, an area of warming much larger than previously reported. West Antarctic warming exceeds 0.1 deg C per decade over the past 50 years, and is strongest in winter and spring. Although this is partly offset by autumn cooling in East Antarctica, the continent-wide average near-surface temperature trend is positive.
Their claims were illustrated on the cover of Nature as shown below:
Again, to be very clear about this, the “novelty” of Steig et al 2009 were their results for West Antarctica – the location of the West Antarctic Ice Sheet. Otherwise, there was nothing in their article that warranted an article in Nature, let alone a cover.
There were a variety of problems with their analysis, many of which were documented at CA and tAV at the time. The remarkable episode of Gavin, International Man of Mystery, plagiarizing a small point on a CA thread arose in the context of Steig et al. At an early stage, it seemed that Steig’s rather complicated method was spreading results from the Antarctic Peninsula into other parts of the Antarctic, a surmise that proved correct.
After an abusive peer review process in which the Team were evidently involved, an article has been accepted by Journal of Climate (O’Donnell [Ryan O], Lewis [Nic L], McIntyre and Condon [Jeff Id]) refuting the West Antarctic claims of Steig et al 2009.
Their results proved to be sensitive to the number of retained principal components, with Steig et al selecting only 3 PCs, using arguments that contradicted Wahl and Ammann’s efforts to retain bristlecones in MBH. Steig et al themselves did not appear to realize that Chladni patterns emerge from spatially autocorrelated time series, a peril known for many years (“Castles in the Air”) but apparently forgotten by Steig et al and the Nature reviewers.
Using a more rational (though not magical) policy on retained PCs, a very different picture emerged as illustrated below. The “traditional” picture of very strong warming on the Antarctic peninsula re-emerged, together with large areas of cooling on the continent. Instead of the West Antarctic being a location of anomalous warming, as Steig et al had purported to show, parts of it were actually cooling.
The gauntlet that had to be run shows that practices in climate science journals remain unchanged despite Climategate. Horton’s essay for Muir Russell noted that conflicts of interest were not simply financial. In this case, the Journal of Climate appointed a reviewer – or shall we say a representative of a Team of reviewers – whose energy in attempting to suppress the article went far beyond an unconflicted reviewer. Ultimately, the reviews and responses totalled 88 pages! And Andy Revkin and others blame critical authors for not running such gauntlets. The reviewers for Ross’ and my comment at IJC on Santer et al 2008 were even worse. All too often, in the case of Santer et al, after Team reviewers sabotaged a straightforward and correct comment, Santer and other Team members criticized us for not publishing in the PRL after getting data and continued to put their results forward as unblemished despite knowing of unrebutted criticisms. Getting this article accepted is entirely to the credit of Ryan O’Donnell who did more than the lion’s share of the work.
Climategate documents show the asymmetry between the puffball “pal reviews” that Jones submitted for Mann or Schmidt or Wahl and Ammann, as compared to the Team “going to town” on criticism. Unfortunately, the Muir Russell “inquiry” did not investigate the peer review incidents evidenced in the emails (not even the “going to town” incident) nor did it comment on pal review nor did it make the slightest effort to see if there were other peer review incidents in other CRU documents (part of their terms of reference.)
The gauntlet that was created in this particular incident had nothing to do with additional due diligence occasioned by perhaps overturning a well established result. Steig’s results, showing West Antarctic as a particular locus of warming, were themselves novel and, if anything, contradicted prior views of Peninsula warming. Our results were straightforward – the 88 pages of review and response were nothing more than obstruction, “going to town” on the comment rather than the original article.
Substantively, what is actually left of the signature results about the West Antarctic, which were:
Assessments of Antarctic temperature change have emphasized the contrast between strong warming of the Antarctic Peninsula and slight cooling of the Antarctic continental interior in recent decades… Here we show that significant warming extends well beyond the Antarctic Peninsula to cover most of West Antarctica, an area of warming much larger than previously reported. West Antarctic warming exceeds 0.1 deg C per decade over the past 50 years, and is strongest in winter and spring.
Nothing. Steig’s West Antarctic warming results from a spreading of warming in the Peninsula to the West Antarctic through choices made in their principal components. Different choices – ones more plausible in the circumstances – lead to opposite results.
Update Dec 7: Online SI here http://www.climateaudit.info/data/odonnell/