On the spatiotemporal characteristics of Fennoscandian tree-ring based summer temperature reconstructions

Abstract

We analyse the spatial representation of five previously published multi-century to millennial length dendroclimatological reconstructions of Fennoscandian summer temperatures. The reconstructions, ranging from local to regional scale, were based on either tree-ring width (TRW) or maximum latewood density (MXD) data or on a combination of the two. TRW chronologies are shown to provide reasonably good spatial information mainly for July temperatures, but a combination of TRW and MXD yields a better spatial representation for the whole summer season (June–August). A multiple-site reconstruction does not necessarily provide better spatial representation than a single site reconstruction, depending on the criterion for selecting data and also on the strength of the climate signal in the tree-ring data. In a new approach to analyse the potential for further developing Fennoscandian temperature reconstructions, we selected from a network of TRW and MXD chronologies those having the strongest temperature information a priori, to obtain a strong common climate signal suitable for a regional-scale reconstruction. Seven separate, but not independent, reconstructions based on progressively decreasing numbers of chronologies were created. We show that it is possible to improve the spatial representation of available reconstructions back to around AD 1700, giving high correlations (>0.7) with observed summer temperatures for nearly the whole of Fennoscandia, and even higher correlations (>0.85) over much of central-northern Fennoscandia. Further sampling of older trees (e.g. dry-dead and subfossil wood) would be needed to achieve the same high correlations prior to AD 1700. Our analysis suggests that it should be possible to select a few key sites for improving the reconstructions before AD 1700. Since tree-ring data from northern Fennoscandia are used in all available hemispheric-scale temperature reconstructions for the last millennium, there is also a potential for slightly improving the quality of the hemispheric-scale reconstructions, by including an improved reconstruction for Fennoscandia. However, adding new chronologies from previously unsampled regions would potentially improve hemispheric-scale temperature reconstructions more substantially.