ATP - Automated Intra-Annual Tree-Ring Profiling for Dendroecological Research

• Applicants: Professor Dr. Rolf Backofen; Professor Dr. Heinrich Spiecker
• Subject Area: Forestry
• Term: from 2014 to 2019
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 269191667

Tree-rings integrate and conserve precisely datable environmental information on the growing conditions of trees during the time of their development. Dendrochronology builds on the basic principle of cross-dating, i.e. the identification and synchronization of time specific patterns between multiple tree-ring sequences to develop absolutely dated tree-ring chronologies. The integration of statistical findings and consolidated knowledge about the relations between environmental factors and the tree growth response as manifested in the tree-ring structures serve as the basis for the reconstruction of past environmental conditions. Such tree-ring based studies provide significant contributions to the fields of climate reconstruction (dendroclimatology), reconstruction of human history (dendroarchaeology) and to a wide range of ecological aspects of the environment.Within this interdisciplinary project, we aim at implementing methods of pattern matching to the analysis of intra-annual tree-ring density profiles for dendrochronological and dendroecological research questions. The project joins the expert knowledge of the Dendroecology group (Prof. Spiecker et al.) in tree-ring based research with the extensive experience of the Bioinformatics group (Prof. Backofen et al.) concerning alignment and clustering method development for bioinformatics applications. Our activities are targeted towards the development of improved methods (1) to enhance environmental signal strength when aggregating multiple intra-annual tree-ring density profiles, (2) to generate high-resolution temporal annotations for intra-annual tree-ring positions, (3) to cross-date tree-ring sequences and build tree-ring chronologies.To this end, we will extend and improve our MICA approach (Multiple Interval-based Curve Alignment), which generates signal-enhanced consensus profiles building on established curve and sequence alignment techniques. Applied on the available extensive intra-annual tree-ring density data, we will generate spatial consensus profile chronologies for various tree species. Using appendant quasi-continuous point dendrometer data, temporal annotation of these intra-annual density data via dedicated MICA adaptions will result in time-annotated chronologies. Both types of chronologies are used to establish new alignment-based cross-dating techniques through the exploitation of intra-annual density profile information. Due to the extended information content of the latter compared to e.g. just ring-width, these new methods will require less sample material while enhancing dating precision. Finally, we will establish a pipeline to investigate the correlation of environmental factors with tree-ring density profiles. An initial study will focus on the application to probabilistically allocate a location or site to an un-located tree-ring sample. All tools and pipelines will be made publicly available and accompanied with extensive documentation.

• DFG Programme: Research Grants