Tree Rings and Climate: Sharpening the Focus - Dendroclimatology - page 331

the past few hundred years. Knowledge of the variability of circulation features and
its impact on regional climates depended almost entirely on the century or less of
direct meteorological observations available for most locations. Great progress has
been made on all three of these issues, and this progress will be briefly discussed
here with particular emphasis on the role of dendroclimatology. Then we will con-
sider the growth in recent decades of applications of dendroclimatology to other
fields, including ecology, hydrology, and the study of interactions between human
societies and climate. Finally, we will look to the future of dendroclimatology—
in particular, we will call for a reframing of the way we use tree rings as archives
of past climate. This recommendation arises from a maturing understanding of the
strengths and limitations of dendroclimatology as it is currently practiced.
11.2 Spectrum of Climate Variability
In the mid-twentieth century, a notional spectrum of climate variability would have
had no robust features between the annual cycle and those related to celestial
mechanics on millennial and multimillennial timescales. For example, Kutzbach and
Bryson ( 1974 ) pointed out the lack of information on Holocene climatic variabil-
ity at what they called intermediate frequencies or timescales—the interval of the
variance spectrum ranging in periods from about 100 to 1000 years. Furthermore,
because of the relative shortness of the instrumental record, estimates of periodici-
ties greater than about 30 years were deemed to have large uncertainties. The authors
expressed an expectation that tree rings and other high-resolution proxy climate
records would play an important role in defining details of the climatic spectrum
at these intermediate timescales (see also Mitchell 1976 ) . Indeed, that expectation
has proven justified, as multiple authors, using both instrumental and proxy records,
including tree rings, have documented the existence of robust features of variabil-
ity on multidecadal (e.g., Schlesinger and Ramankutty 1994 ; Mann and Park 1994 ;
Cayan et al. 1998 ; Allan 2000 ; Delworth and Mann 2000 ; Briffaetal. 2001 ; Gray
et al. 2003 ; Villalba et al. 2003 ) and longer timescales (e.g., Mann et al. 1995 , 1998 ,
2000 , 2008 ; Cobb et al. 2003 ; Cook et al. 2004b ; Graham et al. 2007 ) .
Following the very strong El Niño event of 1982-1983 (Kiladis and Diaz
1986 ) , the connection between the El Niño/Southern Oscillation (ENSO) and
climate anomalies worldwide (Kiladis and Diaz 1989 ) had been demonstrated.
Reconstructions of a number of ENSO indices have been published, using networks
of tree-ring chronologies and in some cases, other proxies such as coral annual band
properties (Lough and Fritts 1985 ; Stahle and Cleaveland 1993 ; Stahle et al. 1998 ;
Cook 2000 ; Mann et al. 2000 , 2009 ; D'Arrigo et al. 2005a ) . These reconstructions
were based on the strength of the large-scale associations between ENSO and local
climate patterns. They extended the available record of El Niño events with annual
resolution back several hundred years. Tree rings have also been used to reconstruct
other aspects of climate variability on these timescales (see, for example, Cayan
et al. 1998 ; Dettinger et al. 1998 ) .