How Well Understood Are the Processes that Create Dendroclimatic Records? A Mechanistic Model of the Climatic Control on Conifer Tree-Ring Growth Dynamics - Dendroclimatology - page 68

Additional uncertainties derive from model parameter specification. The model
has 31 adjustable parameters (Vaganov et al. 2006 ) , many of which were originally
developed from empirical research on cold (Siberia) or dry (Arizona) environments,
but which are potentially poorly constrained in other regions. With our collaborators,
we are currently developing and testing two complementary approaches to evaluat-
ing and reducing uncertainties associated with parameter choices. Ongoing studies
in Siberia are focused on improving the empirical basis for model parameteriza-
tion across a range of environments and site conditions. Another approach is to use
Monte Carlo bootstrapping techniques to develop probabilistic estimates of the sen-
sitivity of simulation results to parameter choices, and to identify those that are most
important. An additional complication is the potential for species-specific parame-
ters. Although we have successfully applied the model with the default ('Siberian')
parameters to a range of species (Evans et al. 2006 ) , there are indications that the
model performs better for some species than for others. We are currently investigat-
ing the extent to which this differential skill is a consequence of site environment or
species characteristics.
As was discussed in the previous section, our specific research questions as
well as data requirements have led us to use a variety of approaches in simulat-
ing tree-ring width chronologies. Evaluations of the model skill as compared with
observations have therefore employed a range of statistical techniques, including
correlation, principal components (Anchukaitis et al. 2006 ) , and spatiotemporal
analysis of large overlapping networks (Evans et al. 2006 ) . Each of these approaches
necessarily has its own underlying assumptions and potential biases, which must be
kept in mind in testing the model against existing tree-ring data.
3.9 Conclusion
The biological, chemical, and physical processes that link the external environment
to the characteristics of annual growth rings in trees form the necessary basis for
dendroclimatological research. Although ultimately the incremental basal diame-
ter growth of trees cannot proceed without biochemical pathways, structures, and
substrates associated with both the crown and the roots, the fundamental processes
that give rise to basal growth and the characteristics and anatomical features of the
annual ring are those associated with cellular growth and division in the cambium.
Here, we have demonstrated that understanding and modeling the environmental
controls on tree-ring formation as arising in part from direct climatic influences
on the dynamic cambium allow us to reproduce the observed phenomenonology of
intra-annual patterns of tree growth as well as the annual and interannual patterns
of ring width variability as observed in a variety of species and sites, and across
different spatial scales. However, much future work is necessary at the intersection
of dendrochronology, tree physiology, and biochemistry in order to better describe
and understand the relationships between photosynthesis, carbon assimilation and
storage, and tree-ring width, as well as the cellular processes or structures that might