In the boreal forests of the Northern Hemisphere, time series of tree-ring width (TRW) and maximum density in the latewood (MXD) are highly correlated to local instrumental summer-temperature data and are thus widely used as proxies in high-resolution climate reconstructions. Hence, much of our present knowledge about climatic variability in the last millennium is based on tree-rings. However, many tree-ring records have a lack of data in the most recent decades, which severely hampers our ability to place the recent temperature increase in a longer-timescale perspective of natural variability.

The main objective of this thesis is to update and extend the Torneträsk TRW and MXD records in northern Sweden. Local instrumental climate-data is used to calibrate the new tree-ring records. The results show that TRW is mainly forced by temperature in the early growing season (June/July) while MXD has a wider response window (June – August) and has a higher correlation to temperature. Two reconstructions of summer temperature are made for (i) the last 7,400 years based on TRW, and (ii) the last 1,500 years based on a combination of TRW and MXD. The reconstructions show natural variability on timescales from years to several centuries. The 20th century does not stand out as a notably warm period in the long timescale perspective. A medieval period from AD 900 – 1100 is markedly warmer than the 20th century.

The environmental impact from a large explosive volcanic eruption in 1628/1627 BC is analysed in the tree rings of 14C-dated bog pines in south-central Sweden and in absolutely-dated subfossil pines from Torneträsk. The results show evidence of an impact in the southern site at approximately this time but no detectable impact in the North.

Subfossil trees of Fitzroya cupressoides in southern Chile were 14C-dated to approx. 50,000 years BP and amalgamated into a 1,229-year TRW chronology. This tree-ring record is the oldest in the world. The variability in this Last-glacial chronology is similar to the variability in present-day living trees of the same species. These results suggest that the growth–forcing mechanisms 50,000 years ago were similar to those at present.