Tree Rings and Climate: Sharpening the Focus - Dendroclimatology - page 336

It is important to note that most of the reconstructions referred to above were
based either completely or substantially on tree-ring data and on the work of den-
droclimatologists. Even so, the major conclusions are robust to the exclusion of
tree-ring data. Mann et al. ( 2008 ) brought together the largest dataset of tree-ring and
other annual- and decadal-resolution proxy data so far used for a hemispheric recon-
struction, and they compared alternative algorithms for extracting a reconstruction
of Northern Hemisphere temperature since the mid-first millennium AD. They con-
clude that, 'Recent warmth [the past decade] appears anomalous for at least the
past 1300 years whether or not tree-ring data are used. If tree-ring data are used,
the conclusion can be extended to at least the past 1700 years, but with additional
strong caveats. The reconstructed amplitude of change over past centuries is greater
than hitherto reported, with somewhat greater medieval warmth in the Northern
Hemisphere, albeit still not reaching recent levels.' The strong caveats referred to
concern decreasing numbers of records for earlier times, the effects of the 'seg-
ment length curse' on the capacity of tree-ring chronologies to conserve century- to
multicentury-scale variability, and time-dependent biases introduced by standard-
ization procedures (see Hughes, Chapter 2 , this volume, and Briffa and Melvin,
Chapter 5 , this volume).
The Mann et al. ( 2008 ) results suggest that, at least for the past 1300 years, these
effects do not produce a different general pattern of hemispheric temperature vari-
ability than is found with the other proxies used (namely, marine and lacustrine
sediments, speleothems, ice cores, corals, and historical documentary series). This
fact, combined with the general agreement between climate model results and the
existing, largely tree-ring-based, literature of hemispheric temperature reconstruc-
tions, suggests not only that tree rings, appropriately handled, are no worse than
other proxies, but also that it will be worth the effort to extend and expand the global
dendroclimatic network back in time, out to the present, and to presently uncovered
regions.
11.6 Circulation Features and Regional Climates
Although hemispheric or global mean temperature reconstructions have special
value for studying the effects of changing radiative forcings—such as solar irradi-
ance, volcanic gases and ash, and greenhouse gases—and eventually for estimating
climate sensitivity, they conceal a rich history of regional variations. It is this that we
referred to at the end of Section 11.3 . At very large scales, it has recently emerged
that, far from conditions being uniformly warm in so-called 'medieval' times, some
large regions of the globe experienced relatively cool sea surface temperatures,
as recorded in coral skeletons, compared to modern values (Cobb et al. 2003 ) .
Stine ( 1994 ) , using geomorphic data, identified century-long extreme low stands
in Mono Lake, California, and other midlatitude locations in both North and South
America for the tenth through fourteenth centuries. He believed these low stands to
be so unusual as to merit the designation of this period as the 'Medieval Climate