We’ve been discussing Briffa and Yamal at CA for a couple of years and many, if not most, regular readers understand the significance of the Yamal collapse. For example, as recently as Sept 19, we were observing the dependence of the Kaufman reconstruction on the Yamal series. So the long-sought fresh information on Yamal (and I’ve been aware of that the Yamal data had been finally placed online for only a few days) fell on fertile soil here.
Some commentators have been very quick to seize on one more example of perceived Team iniquity. That I had been publicly seeking this data for a long time and that Briffa had withheld the data not just from me (but also from D’Arrigo et al, for example) lent an unsavory aspect to CRU’s conduct, fresh after widespread unfavorable publicity for CRU’s withholding of temperature data (by Briffa’s long-time colleague and mentor, Phil Jones.)
Between these two camps, there are obviously readers who wish to understand the implications, if any, of the Yamal problems, without being particularly interested in the sordid backstory of past obstruction. I’ll try to provide some notes on this today. These are not comprehensive, but introductory.
IPCC AR4 said that the late 20th century was the warmest in 1300 years, relying not just on the Mann Hockey Stick, but on about 10 reconstructions by a relatively small group of authors (the “Hockey Team” or the “Team”). These reconstructions are typically presented in a smoothed version as what I’ve called here “spaghetti graph” – a term that has spread into wider usage. The IPCC AR4 spaghetti graph is shown below. Wikipedia has a similar spaghetti graph; the NAS Panel had a simplified spaghetti graph. In detail, these reconstructions seem to agree on very little other than that the modern warm period is slightly warmer than the Medieval Warm Period.
The IPCC spaghetti graph contains 10 “multiproxy” reconstructions, 9 of which go back to the MWP. These will be discussed below. It also contains a short reconstruction from boreholes and a short reconstruction from glacier advance and recession, neither of which pertains to the medieval-modern differential and will not be discussed here.
You will frequently see apologists state that these reconstructions provide “independent” evidence. However, this is not the case on two counts. The same proxies are used over and over again – a point reported at CA on many occasions and confirmed in Wegman et al 2006. Thus, if problems arise with (say) bristlecones or (say) Yamal, this will affect multiple studies and not be contained to one result. More on this below. In addition, there are not really 10 “independent” groups, as “independence” is understood in the real world. Mann and Jones 2003 is not independent of Mann, Bradley and Hughes (1998) or Jones (Briffa) et al 1998 or Rutherford et al 2005 (Mann, Osborn, Bradley, Briffa, Hughes and Jones).
The other point – which I will state here but not prove – is a point about reconstruction methodologies. It is well known in the statistical climate blogs (CA, Jeff Id, David Stockwell, Lubos) that typical paleoclimate operations applied to red noise will yield HS patterns – a line of argument that Ross and I discussed in detail in connection with MBH PC methods, but the same problem arises when you cherry pick from red noise or do ex post correlation weighting. To my knowledge, the climate “community” is in denial on this issue. There is a corollary to this “theorem” – a point that we discussed in our Reply to von Storch and Zorita, which is very important, but which hasn’t got as much attention. If you manually include a huge HS-shaped series in a bunch of red noise and apply standard paleoclimate methods to the network – CPS, as well as Mannian PCs – you get an enhanced HS back with a a minor amount of static.
Rather than paleoclimate methods being “robust” as they self-proclaim, they are profoundly non-robust as this term is used in statistics – denoting the lack of stability of results to individual series. There are two series that play a particular role in the current spaghetti graph population: strip-bark bristlecones/foxtails (especially as Mann’s PC1) and Briffa’s Yamal (and its predecessor). These two series are both shown in IPCC AR4 Box 6.4 Figure 1 as the two biggest HS series – going off the top right corner of the graphic.
Box 6.4, Figure 1. The heterogeneous nature of climate during the ‘Medieval Warm Period’ is illustrated by the wide spread of values exhibited by the individual records that have been used to reconstruct NH mean temperature. These consist of individual, or small regional averages of, proxy records collated from those used by Mann and Jones (2003), Esper et al. (2002) and Luckman and Wilson (2005), but exclude shorter series or those with no evidence of sensitivity to local temperature. These records have not been calibrated here, but each has been smoothed with a 20-year filter and scaled to have zero mean and unit standard deviation over the period 1001 to 1980.
One important “family” of spaghetti graph reconstructions are highly dependent on strip bark bristlecones/foxtails (a topic which has been much discussed here and elsewhere) but which do not use Yamal. These are “highly dependent” on strip bark bristlecones/foxtails in the sense that their methods do not yield a HS without them. Examples include MBH98-99, Crowley and Lowery 2000, Esper et al 2002 plus the re-statements of the NBH network in Rutherford et al 2005, Mann et al 2007 and Wahl and Ammann 2007.
A second “family” of spaghetti graph reconstructions are dependent on Briffa’s Yamal – again, in the sense, that the equivalent calculation using plausible alternatives (e.g. Esper’s Polar Urals version instead of Briffa’s Yamal) yield different MWP-modern relationships. Examples include Briffa 2000, the closely related D’Arrigo et al 2006 and very recently, Kaufman et al 2009 (despite its first impression of a very different network.)
A third “family” of reconstructions wears both belt and braces – i.e. using both strip bark and Yamal. Key examples are Mann and Jones 2003, Mann et al (EOS 2003), Osborn and Briffa 2006, Hegerl et al 2007. The recent UNEP graphic uses the Mann and Jones 2003 version. A common stratagem in these studies is a leave one out sensitivity – where they show that they can “get” a similar result by leaving out any individual proxy. They can do so safely because they have both Yamal and bristlecones.
There are a couple that are a bit sui generis, but these unfailingly have some serious problem.
Jones et al 1998 uses neither Yamal nor bristlecones, but still has a slight modern-medieval differential. In its early portion, it uses only three series, two of which are early Briffa series (Tornetrask and Polar Urals pre-update). Both these series have serious problems – Briffa’s original Tornetrask series contains a gross manual adjustment to increase the 20th century relative to the MWP. See early CA posts on this.
Moberg uses both bristlecones and Yamal, but I view it as sui generis as well. Moberg used some unorthodox wavelet methods, that I’ve sort of emulated, but gave up trying to do so precisely. However, I can confirm that the bristlecone versions used in Moberg are not Graybill versions and don’t affect the result; they are merely fill. I’m not sure what impact Moberg’s filtering method will have on Yamal – I’ve not analyzed that in detail, but may do so some day. I’ve discussed Moberg problems in the past and, for present purposes, merely note that it is not a safe haven, but that it does not appear to stand or fall with Yamal and thus is not discussed further today.
In summary, the apparent problems with Briffa’s Yamal series impact multiple other studies (IPCC AR4 spaghetti graph bolded):
Briffa 2000, Mann and Jones 2003 (used in the recent UNEP graphic), Mann et al (EOS 2003), Jones and Mann 2004, Moberg et al 2005, Osborn and Briffa 2006, D’Arrigo et al 2006, Hegerl et al 2007, Kaufman et al 2009 (and of course, Briffa et al 2008).
And because of the non-robust methods used in these studies, replacing the Briffa Yamal version with a more defensible alternative (such as Esper Polar Urals either individually or in combination with the subfossil Yamal data and Schweingruber russ035w in Yamal) is going to have a material impact on the medieval-modern differential.