date: Tue, 3 Jul 2007 23:15:26 +0200
from: Jan Esper <esper@wsl.ch>
subject: Re: Mitrie
to: Martin Juckes <m.n.juckes@rl.ac.uk>, anders.moberg@natgeo.su.se, Eduardo.Zorita@gkss.de, hegerl@duke.edu, k.briffa@uea.ac.uk, m.allen1@physics.ox.ac.uk, weber@knmi.nl, t.osborn@uea.ac.uk

   Martin

   This is quite a task, as I do not really remember which version of a dataset was used in
   which paper.

   For ECS2002, I detrended all data via two RCS runs applied to the "linear" and "non-linear"
   sub-groups as identified in that paper. All data except for Boreal and Upper Wrigth (both
   from Lisa Graumlich) and Mongolia (from Gordon Jacoby) were measured at WSL.

   I wouldn't necessarily claim that the regional chronologies from the ECS approach are
   highly useful records, i.e. for a regional analysis I would use data that are detrended
   region-by-region.

   (that used by ECS2002 is based on the same tree-ring data as that used by MSH2005, but with
   a different standardisation method.)

   Not fully sure what MSH2005 did, but this is very likely correct, i.e. they likely used a
   "regional" version from Briffa and/or Grudd.

   (The Fennoscandia data used by JBB1998, MBH1999 also come from the Tornetraesk area, but
   from a different group of trees.)

   Hm..., I don't believe that these studies used different trees. Up to the recent update by
   Hakan Grudd, that is currently in review with Climate Dynamics, there was effectively only
   one dataset from Tornetrask. Keith or Tim might know this better.

   (The Polar Urals series used by ECS2005 is also a reanalysis of the data used to create the
   Northern Urals series used by JBB1998, MBH1999.)

   I wouldn't necessarily call this a reanalysis. Perhaps better say 'differently detrended'.
   Anyway, I doubt that there is a long dataset from the Northern Ural as there is little wood
   preserved in that area. This is likely the same data, i.e. both are Polar Ural.

   (The Taymir data used by HCA2007 is a smoothed version of that used in ECS2002, MSH2005.)

   This I really don't know but it would be better to use a regionally detrended version of
   the data...

   (The Greenland stack data used by MBH1999 is a composite of data analysed by
   \citet{fisher_etal1996}, but the precise nature of the composite is not described by
   \citet{fisher_etal1996}.")
   Agreed. Just read the paper again, and it is indeed difficult to say which data was
   combined.
   (I've kept the phrase about "serious flaws" in the conclusion, despite Tim's suggestion,
   supported by Nanne, of a weaker wording, because I think it is important to draw attention
   to the serious flaws which are there.)

   I also think that a less aggressive wording would be more effective.
   -- Jan

   At 16:41 Uhr +0100 3.7.2007, Martin Juckes wrote:

     Hello,
     another version of our paper is attached.
     I've added the following paragraph to the discussion of Table 1, and I'd be
     grateful if Jan and Keith could check that it is accurate:
     "Evaluation of past work is further compicated by confusion between closely
     related proxy series. In Tab.~1 there are two series referred to as
     Tornetraesk: that used by ECS2002 is based on the same tree-ring data as that
     used by MSH2005, but with a different standardisation method. The
     Fennoscandia data used by JBB1998, MBH1999 also come from the Tornetraesk
     area, but from a different group of trees. The Polar Urals series used by
     ECS2005 is also a reanalysis of the data used to create the Northern Urals
     series used by JBB1998, MBH1999. The Taymir data used by HCA2007 is a
     smoothed version of that used in ECS2002, MSH2005.
     The Greenland stack data used by MBH1999 is a composite of data analysed by
     \citet{fisher_etal1996}, but the precise nature of the composite is not
     described by \citet{fisher_etal1996}."
     I've also moved a few things around and tried to follow most of the
     suggestions from Anders and Nanne. I've kept the phrase about "serious flaws"
     in the conclusion, despite Tim's suggestion, supported by Nanne, of a weaker
     wording, because I think it is important to draw attention to the serious
     flaws which are there. One reviewer has implied that we should not discuss
     flawed work at length because in oding so we give it credibility it does not

     deserve. I believe that since this stuff is published and influential in some
     quarters we should discuss it and draw attention to the fact that it is
     seriously flawed.
     cheers,
     Martin
     Attachment converted: Hennes:cp-2006-0049-rv 3.pdf (PDF /IC) (001588D6)

--

   Jan Esper
   Head Dendro Sciences Unit
   Swiss Federal Research Institute WSL
   Zuercherstrasse 111, 8903 Birmensdorf, Switzerland
   Voice: +41-44-739 2510 or +41-44-739 2579
   Fax:   +41-44-739 2515
   http://www.wsl.ch/staff/jan.esper
