Process-Based Standardization in Dendroclimatology
CRU : Tom Melvin

Process-Based Standardization in Dendroclimatology

Thomas M Melvin, Keith R Briffa

The critical "standardisation" process in dendroclimatology, involves converting series of tree-growth measurements into a chronology of dimensionless indices, calculated as deviations from some underlying statistical model of expected ring growth. A major control of tree growth is via the effect of climate, and its annual variability, on photosynthesis. Tree-growth models use meteorological data to calculate the annual mass of carbon (growth material) produced in foliage by photosynthesis. Tree-growth models use empirically defined processes and parameters to allocate this carbon to the growth of new tree structure, including annual stem increments, and can be considered to use detailed knowledge of tree growth process and meteorological data over the life of a tree to insert an age related (geometric) trend into series of ring-width measurements. Reversing these processes has the potential to remove the age-related trend from series of measurements and in doing so will perform dendroclimatic standardisation.

The name "process-based standardization" (PBS) is coined to describe this novel approach to removing non-climatic variance apparent in series of radial tree-growth measurements. A prototype PBS model has been developed and tested using chronologies of actual tree-ring measurements taken from Pinus sylvestris sites in northern Scandinavia. Mechanical strength constraints are used to estimate tree sizes and foliage efficiency is used to model the rise of crown base. Pith estimates are used to generate initial saplings and then, for each ring in turn, the mass of carbon consumed in the growth of that ring is calculated and divided by the estimated foliage mass at the start of that year to generate a tree index based on foliage production rates. A count of sapwood rings beneath the bark gives an estimate of sapwood area in the final year of growth and an exact measure against which to test that the PBS model is correctly estimating sapwood area for each individual tree. Indices based on foliage production rate may not suffer from the low-frequency limitations implicit in traditional standardization techniques and this will be tested.

This work is funded by the Leverhulme Foundation.

Last updated: February 2008