Tuesday, May 8, 2012


date: Wed, 11 Jun 1997 08:59:09 -0400
from: drdendroatXYZxyzo.columbia.edu (Edward R. Cook)
subject: Re: Your paper
to: Keith Briffa <k.briffaatXYZxyz.ac.uk>

Hi Keith,

I just got the paper from Nature to review concerning Bayesian methods of
calibrating tree rings. Its title is "Implications of regression errors in
proxy palaeoclimatic reconstructions" by Robertson et al. I think you know
them because the tree-ring data they use for example is the dC13 data from
oak. It is hard to see exactly what they are doing because the paper is a
comment to Nature about 2 pages long. So, there is little description of
the method. Their Bayesian method appears to do somewhat better then
classical regression (just bi-variate between temperature and dC13), but
not startling so (unlike their use of words like "substantially"). Also,
they don't show any verification results. So, I am going to recommend that
it be rejected and also recommend that a more lengthy paper on it be
submitted as a forum article to The Holocene. What they claim is
provocative enough to warrant it and they will have more room to describe
the method.



No comments:

Post a Comment