The NAS Panel used Esper et al 2002 as a comfort series in their spaghetti graph, but did not perform any due diligence on it. As some of you may recall, I’ve had prolonged correspondence with Science, after being stonewalled by Esper, and, as a result of that, have obtained versions of the chronologies used by Esper for all but one site (Mongolia). I’ve been unable to obtain any sort of coherent explanation of Esper’s methodology. However, any robust result should be valid under an average of the site chronologies. This is what I’ve done here, adding in the Mongolia version used in Osborn and Briffa 2006. A simple re-scaled average fits the Esper reconstruction well up until the late 20th century when they diverge.
I’va also showed the foxtail impact separately. Esper uses two foxtail sites despite them being extremely close to one another – this does not happen anywhere else. As noted elsewhere, foxtails interbreed with bristlecones; Graybill collected strip-bark foxtail sites. The Esper foxtail sites are from Graumlich rather than Graybill but no information is provided which permits one to conclude that strip-bark forms have been avoided. The two sites used here are reported in Bunn et al 2005 and are the two sites with the largest HS. The graph below shows my emulation of Esper’s results (red dashed) together with my estimate of results in which foxtails are excluded.
Figure 1. Esper et al 2002 reconstruction. Archived, emulation and without foxtails.
The Esper reconstruction is unusual in that it is the only one that uses a series from the Polar Urals area not calculated by Keith Briffa. Here Esper uses the Polar Urals Update rather than Briffa’s Yamal Substitution or the predecessor Briffa MXD series.
My emulation of Esper obviously has a substantial Divergence Problem – which becomes a twofold Divergence Problem: a divergence between ring widths and temperature and a divergence between results as archived by Esper and as emulated. Esper has been totally unresponsive to any attempts to examine his data and methodology beginning long before climateaudit existed. It is hard not to draw an unfavorable impression from this obstinacy, especially when there are major discrepancies between archived results and replicable results.