Cargando…

Removing the no-analogue bias in modern accelerated tree growth leads to stronger medieval drought

In many parts of the world, especially in the temperate regions of Europe and North-America, accelerated tree growth rates have been observed over the last decades. This widespread phenomenon is presumably caused by a combination of factors like atmospheric fertilization or changes in forest structu...

Descripción completa

Detalles Bibliográficos
Autores principales: Scharnweber, Tobias, Heußner, Karl-Uwe, Smiljanic, Marko, Heinrich, Ingo, van der Maaten-Theunissen, Marieke, van der Maaten, Ernst, Struwe, Thomas, Buras, Allan, Wilmking, Martin
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Nature Publishing Group UK 2019
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6385214/
https://www.ncbi.nlm.nih.gov/pubmed/30792495
http://dx.doi.org/10.1038/s41598-019-39040-5
Descripción
Sumario:In many parts of the world, especially in the temperate regions of Europe and North-America, accelerated tree growth rates have been observed over the last decades. This widespread phenomenon is presumably caused by a combination of factors like atmospheric fertilization or changes in forest structure and/or management. If not properly acknowledged in the calibration of tree-ring based climate reconstructions, considerable bias concerning amplitudes and trends of reconstructed climatic parameters might emerge or low frequency information is lost. Here we present a simple but effective, data-driven approach to remove the recent non-climatic growth increase in tree-ring data. Accounting for the no-analogue calibration problem, a new hydroclimatic reconstruction for northern-central Europe revealed considerably drier conditions during the medieval climate anomaly (MCA) compared with standard reconstruction methods and other existing reconstructions. This demonstrates the necessity to account for fertilization effects in modern tree-ring data from affected regions before calibrating reconstruction models, to avoid biased results.