Comments on: Splices in Crowley and Lowery

By: The Trick Timeline « Climate Audit

The Trick Timeline « Climate Audit — Wed, 28 Nov 2012 22:05:46 +0000

[...] Date: 21 Nov 2005, Steve [...]

By: Steve Sadlov

Steve Sadlov — Thu, 23 Nov 2006 02:14:23 +0000

That’s utterly outrageous. Chicanery at its worst.

By: Steve McIntyre

Steve McIntyre — Wed, 22 Nov 2006 23:50:37 +0000

This particular post didn’t attract much interest, but was useful to me in documenting a bit of dis-entangling. IT is relevant to Juckes. In Juckes’ collation of cited reconstructions, although he attributes the Crowley version in his spaghetti graph, he actually uses the CL2.JNs11 version, in which Crowley spliced the instrumental record from 1870 on due to poor performance of proxies (interestingly Crowley noted poor matching of bristlecone series and Mann’s “adjustment” of the MBH99 PC1 (bristlecones) to make them match better – more on this latter to come. The Crowly version used here was said by Mann in Eos to be the “wrong version”.

By: TCO

TCO — Tue, 22 Nov 2005 00:33:06 +0000

My experience in academic hard core science is that it is quite common to have math errors and such make it into publication. Without the source code (exact algorithm), how can you deconvolute errors in the original work versus errors in detective work of the methods.