Search

A contribution to the 2k Network. ABSTRACT: Despite the spatially homogenous summer temperature pattern in Fennoscandia there are large spreads among the many existing reconstructions resulting in an uncertainty in the timing and amplitude of past changes. Also there has been a general bias towards northernmost Fennoscandia. In an attempt to provide a more spatially coherent view of summer (June–August JJA) temperature variability within the last millennium we utilized seven density and three blue intensity Scots pine (Pinus sylvestris L.) chronologies collected from the altitudinal (Scandinavian Mountains) and latitudinal (northernmost part) treeline. To attain a JJA temperature signal as strong as possible as well as preserving multicentury scale variability we used a new tree ring parameter where the earlywood information is removed from the maximum density and blue intensity and a modified signal free standardization method. Two skilful reconstructions for the period 1100–2006 CE were made one regional reconstruction based on an average of the chronologies and one field (gridded) reconstruction. The new reconstructions were shown to have much improved spatial representations compared to those based on data from only northern sites thus making it more valid for the whole region. An examination of some of the forcings of JJA mean temperatures in the region shows an association with sea surface temperature over the eastern North Atlantic but also the subpolar and subtropical gyres. Moreover using Superposed Epoch Analysis a significant cooling in the year following a volcanic eruption was noted and for the largest explosive eruptions the effect could remain for up to 4 years. This new improved reconstruction provides a mean to reinforce our understanding of forcings on summer temperatures in the North European sector.

Dendroclimatological sampling of Scots pine (Pinus sylvestris L.) has been made in the province of Jamtland, in the west-central Scandinavian mountains, since the 1970s. The tree-ring width (TRW) chronology spans several thousand years and has been used to reconstruct June–August temperatures back to 1632 bc. A maximum latewood density (MXD) dataset, covering the period ad 1107–1827 (with gap 1292–1315) was presented in the 1980s by Fritz Schweingruber. Here we combine these historical MXD data with recently collected MXD data covering ad 1292–2006 into a single reconstruction of April–September temperatures for the period ad 1107–2006. Regional curve standardization (RCS) provides more low-frequency variability than “non-RCS” and stronger correlation with local seasonal temperatures (51% variance explained). The MXD chronology shows a stronger relationship with temperatures than the TRW data, but the two chronologies show similar multi-decadal variations back to ad 1500. According to the MXD chronology, the period since ad 1930 and around ad 1150–1200 were the warmest during the last 900 years. Due to large uncertainties in the early part of the combined MXD chronology, it is not possible to conclude which period was the warmest. More sampling of trees growing near the tree-line is needed to further improve the MXD chronology.