Synchronous variability changes in Alpine temperature and tree-ring data over the past two centuries

D Frank, Rob Wilson, J Esper

Research output: Contribution to journalArticlepeer-review

20 Citations (Scopus)


The understanding of extremes and their temporal distribution is useful in characterizing the behaviour of the climate system, and necessary for understanding their social and economic costs and risks. This task is analogous to the study of pointer years in dendrochronological investigations. Commonly used dendroclimatological methods, however, tend to result in an equalization of variance throughout the record by normalizing variability within moving windows. Here, we analyse a larger network of high-elevation temperature-sensitive tree sites from the European Alps processed to preserve the relative frequency and magnitude of extreme events. In so doing, temporal changes in year-to-year tree-ring width variability were found. These decadal length periods of increased or decreased likelihood of extremes coincide with variability measures from a long-instrumental summer temperature record representative of high-elevation conditions in the Alps. Intervention analysis, using an F-test to identify shifts in variance, on both the tree-ring and instrumental series, resulted in the identification of common transitional years. Based on a well-replicated network of sites reflecting common climatic variation, our study demonstrates that the annual growth rings of trees can be utilized to quantify past frequency and amplitude changes in extreme variability. Furthermore, the approach outlined is suited to address questions about the role of external forcing, ocean-atmosphere interactions, or synoptic scale changes in determining patterns of observed extremes prior to the instrumental period.

Original languageEnglish
Pages (from-to)498-505
Number of pages8
Issue number4
Publication statusPublished - Nov 2005


  • ALPS


Dive into the research topics of 'Synchronous variability changes in Alpine temperature and tree-ring data over the past two centuries'. Together they form a unique fingerprint.

Cite this