HR: 0830h
AN: GC31B-0192    [PDF]
TI: Stochastic Model of Complexity in Dendrochronology
AU: Beltrami, H
EM: hugo@stfx.ca
AF: Environmental Earth Sciences Laboratory, St. Francis xavier University, P.O. Box 5000, Antigonish, NS B2G2W5 Canada
AU: * Lavallee, D
EM: daniel@crustal.ucsb.edu
AF: Institute for Crustal Studies, University of California, Santa Barbara,, Santa Barbara, CA CA 93106 United States
AB: The basic assumption of this study is that a stochastic model can reproduce the variability in amplitude and the long-range correlation observed in dendrochronological data. The model is tested for four chronologies from the International Tree Ring Data Bank. For the four samples, we show that the power-spectra follows a power law behavior with scaling exponents close to 0.6. For the four samples, we have found that non-Gaussian distributions, i.e., the Levy distributions, are better suited to describe the temporal variability of the ring index. We also show that a stochastic characterization of the ring index based on a Gaussian distribution fails to capture the temporal variability observed in the original samples: especially, the 'extreme', large values. The results obtained for the samples suggest that some features of the complexity recorded in tree ring indices are universal and can be modeled accordingly. If this is proven correct, this will imply that the temporal variability and the long-range correlation of tree ring indices can be described with the help of five parameters: a scaling exponent controlling the spatial correlation and the four parameters of the Levy distribution constraining the temporal variability. We suggest that we can use these inferences as proxies of climatic variability for the assesing of extreme weather events in a changing climate.
DE: 1600 GLOBAL CHANGE (New category)
DE: 1699 General or miscellaneous
SC: Global Climate Change [GC]
MN: 2003 Fall Meeting