A TIME SERIES ANALYSIS APPROACH TO TREE RING STANDARDIZATION (DENDROCHRONOLOGY, FORESTRY, DENDROCLIMATOLOGY, AUTOREGRESSIVE PROCESS).

Abstract

The problem of standardizing closed-canopy forest ringwidth series is investigated. A biological model for the tree-ring standardization problem indicated that one class of non-climatic variance frequently responsible for standardization problems could be objectively minimized in theory. This is the variance caused by endogenous stand disturbances which create fluctuations in ringwidth series that are non-synchronous or out-of-phase when viewed across trees in a stand. A time series method based on the autoregressive process is developed which minimizes the timewise influence of endogenous disturbances in detrended ringwidth series. Signal-to-noise ratio (SNR) properties of this method are derived which indicate that autoregressive modelling and prewhitening of detrended ringwidth indices will result in a higher SNR when endogenous disturbances are present in the series. This enables the verification of the SNR theory and the error variance reduction property of the standardization method.