Tree ring analysis can provide annual reconstructions of climate variation for thousands of years (Bradley et al 2003). These records form a basis for our understanding of climate change (ACIA 2004).
My master’s degree has focused on ageing Arctic shrubs in Churchill, MB (Fig. 1), however I am also interested in testing whether or not Arctic shrubs can provide chronologies similar to those developed for trees. In order to understand the basic procedures for chronology development, I worked through a tree ring chronology example using common statistical techniques. I began with tree ring series and climate data from Churchill. (See Data References)
A number of software packages are available for studying tree ring climate relationships, however, it was beneficial from an instructional point of view to do the entire analysis manually. I used Principle Component Analysis (PCA) followed by Multiple Regression (MR) in SAS. It was also instructive to view basic plots of correlations and normality and to compare univariate and multivariate temperature chronologies.