Statistical Relationships Between Tree Growth and Climate in Western North America
|Title||Statistical Relationships Between Tree Growth and Climate in Western North America|
|Year of Publication||1992|
|University||University of Arizona|
The objective of this study is to examine large-scale spatial patterns of tree growth and climatic variation and to investigate the possible role of climate in determining tree growth patterns over space. This study represents one of the first uses of geostatistical methods to extract information about the spatial variation of climate from tree rings in western North America. It is also one of the first uses of data in spatial series to study the relationships of spatial variations between climate and tree growth. Geostatistics analyzes the spatial structure of the variables by assuming that adjoining data are correlated with each other over space and that the particular relationship expressing the extent of spatial correlation can be analytically and statistically captured in a function. It is applied to both June Palmer Drought Severity Index (PDSI) and ring-width index data from western North America. One basic assumption of applying geostatistics in this study is that the spatially uncorrelated small-scale variations are insignificant and represent background noise in large-scale dendroclimatic studies. The statistical relationships between the spatial variations of June PDSI and ring-width index are studied by simple scatter diagrams and correlation analysis. This is done in terms of yearly variations and variations of spatial patterns. Both of them support the contention that the large-scale spatial variations in ring-width index data can be used to infer the spatial variations of climate variables. Based upon the results of this research it can be concluded that geostatistics is a viable method to characterize the spatially correlated variations in dendroclimatology. By applying geostatistics to data sets, information about the spatial variations of climate contained in tree-ring data are enhanced, and the large-scale variations of climate are emphasized. The analysis of yearly relationships over space is particularly useful for identifying statistical relationships between climate and tree growth in a geographic region. The main factors of climate controlling ring-width index are identified as well as the less frequent limiting events. Once the statistical relationships are validated, they can be used to infer the spatial variations of past climate from variations in tree-ring index.