Producing an enhanced millennium-length temperature reconstruction from tree-ring density for the European Alps
Final Report Abstract
This project aimed at producing an enhanced temperature reconstruction for Europe for the past millennium. We use tree-ring maximum latewood density as this is considered as ‘the current gold standard’ of high‐resolution paleoclimatology for temperature reconstructions. The requirements for the development of a robust temperature reconstruction for the past millennium are a strong temperature signal fingerprinted in the tree-rings, and sample availability covering this time scale. This is the case in the Swiss Alps where tree growth is temperature limited at elevational treeline sites and historical material is available from old buildings. However, the original growing site of the trees used for timber production is unknown so that the robustness of the temperature signal inherent to the tree-ring series is unsecure. With this project we extended the frontiers of dendroclimatology in many ways by integrating tree-ring density with tree-ring width data. We introduced machine learning algorithms to find the exact provenance and growing site of the trees by using several elevation specific tree-ring series features (e.g. mean, maximum, minimum, standard deviation of ring width and density, earlywood width and density, latewood width and density, earlywood latewood ratio etc.). This means that we actually reject available series but are able to only select high elevation temperature sensitive tree-ring series for our chronology. With this background information the necessary detrending procedure of tree-ring series is also fundamentally improved. Another progress to previous approaches is that we detect and reduce the signature of insect outbreak events which would feign cold temperatures, with a new stepwise multi-parameter approach. We produced a very reliable maximum latewood density chronology with very high replication and stable signal strength with which we can reconstruct summer temperatures for the period 802-2017 for large regions of Southwestern and Central Europe.
Publications
-
The effect of provenance of historical timber on tree-ring based temperature reconstructions in the Western Central Alps. iForest - Biogeosciences and Forestry, 13(1), 351-359.
Riechelmann, D. F.; Hartl, C. & Esper, J.
-
The temperature sensitivity along elevational gradients is more stable in maximum latewood density than tree-ring width. Dendrochronologia, 73, 125958.
Hartl, Claudia; Schneider, Lea; Riechelmann, Dana F.C.; Kuhl, Eileen; Kochbeck, Markus; Klippel, Lara; Büntgen, Ulf & Esper, Jan
-
Using machine learning on tree‐ring data to determine the geographical provenance of historical construction timbers. Ecosphere, 14(3).
Kuhl, Eileen; Zang, Christian; Esper, Jan; Riechelmann, Dana F. C.; Büntgen, Ulf; Briesch, Martin; Reinig, Frederick; Römer, Philipp; Konter, Oliver; Schmidhalter, Martin & Hartl, Claudia
