Trends and signals in tree-ring parameters
Date issued
Authors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
License
Abstract
Tree-rings are frequently used to develop annually resolved paleoclimate reconstructions of periods in time when meteorological instrumental measurements are not available. Commonly used proxy records derived from tree-rings are ring width, maximum latewood density, and stable carbon and oxygen isotopes. Calibrating the different tree-ring proxies against available meteorological data is a prerequisite for reconstruction approaches, as results provide the essential statistical relationships to convert proxy data into the targeted climate data. Additionally, the evaluation of potential non-climatic biases is crucial for accurate calibration results. In this dissertation, methods for identifying proxy-specific biases are developed and evaluated for associated impact on various calibration setups. Focus is placed on age-related climate sensitivity trends in growth and density data, insect-induced disturbances in interannual growth patterns, and effects of elevated atmospheric CO2 concentrations on stable isotopic compositions.
Age-related alternation in the climate sensitivity of tree-ring width data has been previously reported and is here revisited by analyzing a large dataset of 692 Pinus sylvestris L. series from northern Fennoscandia. Additionally and for the first time, maximum latewood density measurements of the same trees are included. Results indicate significant decreasing climate sensitivity with increasing age in both tree-ring parameters, while density data are less affected, thus, more suitable for the development of climate reconstructions. Temperature reconstructions for this region can dismiss age-related biases, by using density data from evenly distributed cambial ages, i.e. including young and old trees, as a function of time.
The larch budmoth (Zeiraphera diniana Gn.) is characterized by regular population oscillations and cyclic mass outbreaks (8-10 years), causing interannual disturbances in tree-ring width chronologies from host trees and reduced climate signal strength. In contrast to reported historical evidence over 1200 years from the European Alps, cyclic mass outbreaks in the Slovakian High Tatra Mountains are, in this dissertation, found to be absent during the last 300 years. Low stand densities in the spatially limited larch forests in this habitat prevent larch budmoth populations from peaking at mass outbreaks levels. The significantly synchronized Larix decidua Mill. (host) and Pinus cembra L. (non-host) chronologies, both free of insect-induced pulsed disturbances, exhibit distinct early-summer temperature signals ideal for East-European temperature reconstructions.
Stable isotope data from tree-rings provide extensive information on interactions between tree-physiological processes and external climatic forcing. Annually and decadally resolved stable carbon (δ13C) and oxygen (δ18O) isotope time series from Pinus uncinata Ramond ex DC tree-rings from the Spanish Pyrenees exhibit climate sensitivity - δ18O values are controlled by spring precipitation and other more complex factors, and δ13C data embeds a significant summer temperature signal. However, different correction methods addressing effects of elevated atmospheric CO2 on the carbon isotopic composition add low-frequency information to the data, and alter the magnitude of the climate signal strength and its temporal robustness. Selecting the most appropriate correction method remains challenging, as the habitat-specific natural adaptation of tree-intrinsic physiological processes to elevated atmospheric CO2 could not be clarified, and objective criteria for best-fit corrections are still missing. Using the correction method resembling summer temperature characteristics, a newly developed 750-year decadal summer temperature reconstruction comprises greater amplitude compared to previously reported reconstructions derived from ring width and density data. Additionally, the novel approach in tree-ring stable isotope chronology development of detrending-techniques are applied to the δ13C isotopic data, providing the background of this temperature reconstruction.
The detection and evaluation methods for potential biases in calibration setups presented in this dissertation can be utilized as conceptual basis to evaluate existing and to advance future reconstruction, thereby, improving our understanding of long-term climate variations.