Exploring Northern Climate History Through Tree Rings: Proxy Development, Climate Signals, and Reconstructions

Abstract

Tree rings provide annually resolved and precisely dated information about past climate variability, but their value for climate reconstruction depends on where suitable climate-sensitive records can be developed, how strongly and consistently different tree-ring properties record climate, and how faithfully the measured proxy preserves this information through time. This thesis aims to improve the use of tree-ring proxies for reconstructing climate variability in northern environments by addressing these interconnected aspects of dendroclimatic reconstruction. The first aim was to evaluate the strength, seasonality and temporal stability of temperature signals in different tree-ring proxies. Tree-ring width temperature signals in northern Fennoscandia remained relatively stable during twentieth-century warming, despite substantial changes in the thermal growing season and spring phenology (Paper I). This indicates that changes in the climatological growing season do not necessarily result in corresponding changes in the seasonal temperature signal recorded by radial growth. A comparison of tree-ring parameters from the Canadian Rockies further showed that signal strength and stability differ among proxies (Paper II). Wood anatomical properties, particularly radial cell-wall thickness and anatomically derived density, contained strong and temporally stable summer-temperature signals, comparable to X-ray-based maximum latewood density. Together, these results demonstrate the importance of considering both the seasonal timing and temporal stability of climate signals when selecting tree-ring proxies for reconstruction. The second aim was to develop a long hydroclimate-sensitive record in northern Fennoscandia, where such records remain sparse. At the High Coast of Sweden, local site conditions created strong moisture limitation in Scots pine despite the relatively high latitude (Paper III). This enabled May–June precipitation to be reconstructed over approximately seven centuries and shows that suitable local environments can provide hydroclimate information beyond the regions where moisture-sensitive tree-ring records are normally concentrated. The third aim was to improve the preservation of climatic information in the accessible image-based Blue Intensity proxy. A new parameter, Proportional Blue Reflectance (PBR), was developed to reduce colour-related biases affecting conventional Delta Blue Intensity in long chronologies (Paper IV). PBR showed improved agreement with X-ray-based density at lower frequencies while retaining the accessibility and low cost of image-based measurements, thereby improving its potential for preserving long-term climate variability. Together, the results show that improving northern tree-ring climate reconstructions requires not only longer records, but also suitable climate-sensitive environments, a clear understanding of the seasonality and stability of the signals they contain, and proxy parameters that preserve climatic variability across relevant timescales. By addressing these components, this thesis contributes towards more robust, climatically interpretable and spatially extensive tree-ring reconstructions from northern environments.