Published in
European Geosciences Union, Climate of the Past Discussions, p. 1-25, 2019
DOI: 10.5194/cp-2019-8
Links
Tools
Annually resolved &delta;<sup>2</sup>H tree-ring chronology of the lignin methoxyl groups from Germany reflects averaged Western European surface air temperature changes
,
Birgit Sehls,
Werner Thomas,
Claudia Hartl ,
Markus Greule ,
Denis Scholz,
Jan Esper ,
Frank Keppler
Full text: Download
Abstract
<p><strong>Abstract.</strong> Stable hydrogen isotopes ratios of lignin methoxyl groups (expressed as &delta;<sup>2</sup>H<sub>LM</sub>) of wood have been shown to reflect the climate-sensitive &delta;<sup>2</sup>H values of precipitation (expressed as &delta;<sup>2</sup>H<sub>precip</sub>) modulated by a large uniform negative isotope fractionation. However, a detailed calibration study among temporal variabilities of &delta;<sup>2</sup>H<sub>LM</sub> in tree-ring series, site-specific &delta;<sup>2</sup>H<sub>precip</sub> and climate parameters has not been performed yet. Here, we present annually resolved &delta;<sup>2</sup>H<sub>LM</sub> values from nine tree-ring series (derived from four <i>Fagus sylvatica</i> L. trees) collected near stations of the Global Isotope Network of Isotopes in Precipitation (GNIP) and the Deutsche Wetterdienst (DWD) meteorological observatory at Hohenpeißenberg (southern Germany; ~&thinsp;48&deg;&thinsp;N, 11&deg;&thinsp;E). The measured nine &delta;<sup>2</sup>H<sub>LM</sub> tree-ring series (common period of overlap 1916-2015) show a strong coherency as indicated by highly significant (<i>p</i>&thinsp;<&thinsp;0.001) inter-series correlations (mean value Rbar&thinsp;=&thinsp;0.52) and no indication of any known (stable isotope-specific) juvenile trend affecting the first decades of growth. When compared to local instrumental data, the combined &delta;<sup>2</sup>H<sub>LM</sub> chronology shows highest correlations with annually averaged data of &delta;<sup>2</sup>H<sub>precip</sub> as well as temperature particularly when using the year defined from previous September to current August (<i>r</i>&thinsp;=&thinsp;0.73 and 0.56, respectively, <i>p</i>&thinsp;<&thinsp;0.001). However, the Hohenpeißenberg &delta;<sup>2</sup>H<sub>LM</sub> chronology shows enhanced correlations with land and sea surface air temperature for multiple (broadly combined) areas across Western Europe (<i>r</i>&thinsp;>&thinsp;0.6, <i>p</i>&thinsp;<&thinsp;0.1). We subsequently established a linear regression model between averaged Western European surface air temperatures (range: 30&deg;&thinsp;W&ndash;20&deg;&thinsp;E, 35&ndash;60&deg;&thinsp;N) and the &delta;<sup>2</sup>H<sub>LM</sub> chronology (<i>r</i>&thinsp;=&thinsp;0.71, <i>p</i>&thinsp;<&thinsp;0.001). When comparing instrumental and reconstructed large-scale temperature anomalies from the year 1916 to 2015, an average absolute deviation in annual reconstructions of as low as 0.3&thinsp;&deg;C was found (<i>n</i>&thinsp;=&thinsp;100). Therefore, &delta;<sup>2</sup>H<sub>LM</sub> values of mid-latitudinal tree-ring archives are considered suitable for large-scale mean annual temperature reconstructions and are therefore able to improve the paleoclimatic potential of Late Holocene tree-ring archives.</p>