Climate signals along subcontinental transects
Assessing non-stationary climate signals using novel wood density
methods along sub-continental transects
Project founded by DFG
Principal Investigator: Univ. Prof. Dr. Jan Esper
Duration: 2024-2027
Our understanding of naturally forced climate variability prior to the onset of instrumental temperature observations relies on the interpretation of climate proxies and their quantitative reconstruction. Among these, tree-rings have been identified as a key proxy for reconstructing temperature variability. Notably, maximum latewood density (MXD) stands out for its strong climate signal among dendrochronological parameters. However, results from a hemispheric density network established in the late 20th century using classical density measurement methods have revealed a weakening association between tree growth and climate, known as the “divergence problem”. This problem not only affects the climate signal in tree-ring density chronologies, but also increases the uncertainty of climate reconstructions, especially during pre-instrumental warm periods.
This project therefore seeks to evaluate the performance and resilience of novel density parameters derived from both blue intensity measurements (BI) and quantitative wood anatomy measurements (QWA) across documented divergence hot spot regions in the Northern Hemisphere. To achieve this, we will develop well-replicated BI and QWA chronologies from multiple tree species along sub-continental transects and assess them for a potential non-stationary climate signal. We will compare the signals among tree species and along the transects as well as between novel and classical density parameters, while exploring approaches to mitigate divergence in modeled temperatures. This project will enable, for the first time, a comparative analysis of the reconstruction skills of density parameters, including both novel (BI and QWA), and classical (MXD) density parameters, over extended calibration and reconstruction periods, from multiple sites, across divergence hotspot regions. Consequently, we will be able to identify divergence of each tree-ring parameter and site, specify the severity of the problem (or lack thereof), in the high- and low-frequency domains, and utilize the more robust estimates as a reference and surrogate for the more biased estimates.
Deputatsky Site