date: Thu, 06 Apr 2006 12:45:16 +0100
from: Tom Melvin <t.m.melvin@uea.ac.uk>
subject: Re Hakan
to: Keith <K.Briffa@uea.ac.uk>

<x-flowed>
Keith,

I read Hakan's draft and find problems.

Would it be possible for you to ask Hakan to send his updated (and 
original) measures, TRW, MXD and Pith estimates for testing purposes?

The problem is that the update chronology appears to have a different 
growth rate to the original and this removes the growth increase of 
the modern century and hides the sensitivity problem. Both feature in 
Hakan's conclusions.

Mean MXD for old trees of 0.58 is considered similar to the mean for 
new trees at 0.57, but considering the slope of the RCS curve and the 
mean age of the constituent rings the new data should have a much 
larger mean density than the old data.

The presumption that the MXD chronology is "correct" and that the TRW 
chronology has excess slope over the modern period (Figures 4 and 9) 
may be wrong.

Signal-Free methods and BFM would offer a second opinion. They 
managed to reduce the excesses of the few < 50 old, very fast growing 
trees, of the 2002 paper Tornetrask chronology.

I have a few detailed comments I will let you have later.

The MXD data would be useful to the "Sensitivity" review paper if 
Hakan is happy for it to be used.

Tom



Tom Melvin

Climatic Research Unit
University of East Anglia, Norwich, NR4 7TJ, U.K.
Phone: +44-1603-593161   Fax: +44-1603-507784


</x-flowed>
