date: Thu, 12 May 2005 15:50:36 +0200
from: Kurt Nicolussi <kurt.nicolussiatXYZxyzk.ac.at>
subject: for discussion
to: Tom Melvin <t.m.melvinatXYZxyz.ac.uk>, "firstname.lastname@example.org" <k.briffaatXYZxyz.ac.uk>, Andrea Johanna Thurner <Andrea.Thurner@uibk.ac.at>
Hi Tom and Keith,
with the completed data set - as I sent you - we established new RCS
standardized chronologies for the last 2000 years - up to now using
win-ARSTAN. Now we can them compare with the results obtained by using
TOM.exe (MRCS, ...).
I've added some figures established by Andrea:
Vergleich RCS-chronos mit_ohne pt 2.wmf --- compares the two chonos
(with / without pt)
Vergleich RCS-chronos mit_ohne pt 2 smooth.wmf --- smoothed version of
EPS_RBAR_mit_u_ohne_pt 1k.wmf --- comparison of the EPS and RBAR values
of these two chronos
Mittelk_mit power trans_Vergl5.wmf --- comparison of the chronologies of
the three groups of samples (living, subfossil, historical) obtained by
using different RCS curves for standardization, all data with pt.
Mittelk_ohne power trans_Vergl5.wmf --- the same as above, but without pt.
Vergleich RCS mit power trans.wmf --- the RCS curves of the three sample
groups, with pt.
Vergleich RCS ohne power trans.wmf --- the same, without pt.
1) Why did we divide the data - for data from living trees in comparison
to subfossil or historical data it ist necessary - see the different
RCS-curves: Vergleich RCS ohne power trans.wmf, Vergleich RCS mit power
the "historical" and the "subfossil" RCS differs not really, but up to
now we don't have tested what would be the effects.
2) Than we compared the chronologies of these different groups in the
overlapping period - see: Mittelk_ohne power trans_Vergl5.wmf;
Mittelk_mit power trans_Vergl5.wmf.
historical and subfossil differs only slightly - with the exception of
the period around AD 1700 - here one sites dominates the data set
(altitude 2200 to 2300 m a.s.l.), maybe that's the reason. The first
part of the living trees chronology is below the other series - the very
old trees start as slow growing trees - maybe we should divide the data
sets in two age (?) classes - what do you think?
3) we wanted to compare chronologies established by using power
transformation or not - altogether, there are not so big differences for
the last 1000 years, but especially for the last few decades is maybe
better - but e.g. around 500 AD - the pt version is very high, perhabs
too high. (see: Vergleich RCS-chronos mit_ohne pt 2.wmf and Vergleich
RCS-chronos mit_ohne pt 2 smooth.wmf). The RBAR values are slightly
higher with pt (see: EPS_RBAR_mit_u_ohne_pt 1k.wmf). What's your opinion?
Altogether, the temperature sensitive Pinus cembra chronologies show a
temperature variability with which I can explain most of the glacier
fluctuations within this period that I know. Also the different levels -
the LIA as the longest below-average period, the MWP as long, but not
extrem above-average period, the long Roman Warm Period on a level
similar the late 20th century.
Dr. Kurt Nicolussi, Assoc. Prof.
Tree-ring Group / Institute of Geography
University of Innsbruck
Tel +43 512 507 5673
Fax +43 512 507 2806
Attachment Converted: "c:\eudora\attach\Vergleich RCS-chronos mit_ohne pt 2.wmf"
Attachment Converted: "c:\eudora\attach\EPS_RBAR_mit_u_ohne_pt 1k.wmf"
Attachment Converted: "c:\eudora\attach\Mittelk_mit power trans_Vergl5.wmf"
Attachment Converted: "c:\eudora\attach\Mittelk_ohne power trans_Vergl5.wmf"
Attachment Converted: "c:\eudora\attach\Vergleich RCS mit power trans.wmf"
Attachment Converted: "c:\eudora\attach\Vergleich RCS ohne power trans.wmf"
Attachment Converted: "c:\eudora\attach\Vergleich RCS-chronos mit_ohne pt 2 smooth.wmf"