Tree rings can provide continuous yearly palaeoclimatic records for regions of the Earth or periods of time with no instrumental climate data. However, trees of different species are expected to react differently to changes in climate, with some species more sensitive to rainfall and others to temperature or sunshine. In order to disentangle the effects of different climate variables on tree-ring growth rates of different species, and to explore the possibility that more than one species could improve the accuracy of the reconstructions of climate using tree rings, a methodical study of trees in close proximity to the Armagh Observatory climate station was initiated together with Ana Maria García Suárez, and continued as part of the latter's PhD thesis with John Butler and Professor Mike Baillie (QUB).
A selection of 10 to 20 specimens of each of five species that are common in Northern Ireland, and which had grown within 10km of the climate station at Armagh Observatory, was made. The trees were cored and the tree rings measured with a travelling microscope. Standard chronologies (tree-ring width series) from the late 18th or early 19th century were established for four of the species studied, namely ash, oak, beech and scots pine. Attempts to form a chronology for lime were abandoned after it was found that tree-ring series from this species did not replicate. This part of the work was undertaken with the kind assistance of the School of Archaeology and Palaeoecology at Queens University Belfast.
A detailed statistical analysis by Ana Maria García Suárez, using procedures developed and kindly made available to us by Dr Ed Cook (The Lamont-Doherty Tree-Ring Laboratory, Columbia University, New York), has shown that of the species we have studied beech and ash are the most sensitive to climate, with tree-ring growth more strongly influenced by precipitation and soil moisture in early summer than by temperature or sunshine. These meteorological parameters can also be relevant to growth in the autumn months, particularly in the previous year. Whereas previous studies of tree rings in Ireland (with oak) have been able to explain only about 30% of the variance in tree-ring widths by climate, we have found that we are able to explain 50% in beech, 30% for oak and ash and 25% for pine. Combinations of climate variables are able to explain a significantly higher fraction of the variance than single variables.
Though it has not proved possible to successfully reconstruct from tree rings annual mean meteorological data, some parameters such as rainfall and the Palmer Drought Index, and to a lesser extent maximum and mean air temperatures, can be reconstructed over limited portions of the year. As with the response of trees to climate, we find that combinations of species are better able to reconstruct meteorological parameters than single species.