This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).
This project seeks better ways to use tree ring data to infer changes in climate. Annual growth rings in trees have long been used as a primary means of reconstructing climatic conditions over the past two millennia, extending our knowledge of climate beyond the period of the instrumented record. However, tree ring growth depends on factors other than climate, and non-climatic influences must be removed when tree ring data is used for climate reconstruction. Traditional methods for filtering out non-climatic influences can be too extreme, removing some of the climate signal as well as the non-climate noise. The research in this project will improve upon traditional filtering methods using Bayesian hierarchical modeling, a technique in which statistical models will be constructed to specify how tree rings growth is expected to vary in response to both climatic and non-climatic (e.g. tree age) factors. The models will then be combined with tree ring data to produce both climate reconstructions and estimates of the level of uncertainty present in the reconstructions. Additional research will consider the extent to which tree ring-based reconstructions might benefit from changes in the way tree ring data is collected. In particular, the potential benefit from taking additional cores from each tree will be evaluated.
Tree ring-based climate reconstructions are an important source of information on the range of natural climate fluctuations, and thus they play an important role in assessing the extent to which recent climate changes are outside the envelope of pre-industrial climate variability. The improvements in tree ring analysis resulting from this research will thus have a broad impact on our understanding of both natural and human-induced climate variation and change. In addition, tree rings are an important source of information regarding the range of past variability in water resources, which is of interest to water managers in the semi-arid southwestern states. Beyond these benefits, the research will develop statistical techniques which will be applicable to a range of scientific problems in which the behavior of a complex system must be inferred from proxy data.
