Eddy covariance measurements of the dual-isotope composition of evapotranspiration
Citable Link (URL):http://resolver.sub.uni-goettingen.de/purl?gs-1/16182
Measurements of the isotopic composition of water vapor (H2O) provide valuable insights into the hydrological cycle, whereas eddy covariance (EC) measurements are widely used to quantify biosphere-atmosphere exchange processes. Yet, the direct combination of water isotope approaches and the EC method remains technically challenging. Here, we present the first EC measurements of δ18O and δD of evapotranspiration (ET) over the full growing season of a beech forest in central Germany. This EC implementation is based on 2 Hz measurements of the mole fraction and isotopic composition of water vapor (CH2O,v, δD and δ18O) with a customized version of a commercially available water vapor isotope analyzer. The isotopic composition of ET showed a seasonal variability from −19 to 0‰ for δ18OET and from −140 to −25‰ for δDET. The setup-specific limitations of our measurements yield a mean bias of 0.03 mmol m−2 s−1 of the measured net water flux. Spectral and cospectral analysis showed that high-frequency dampening was less pronounced for our EC setup (which was equipped with heated tubing) than for the standard EC setup at this site. Thus, we conclude that direct EC measurements of the isotopic composition of ET are feasible for both, δ18OET and δDET. We propose that EC-based measurements of the isotopic composition of ET are feasible to improve our understanding of the hydrological cycle, especially where flux gradient methods show limited applicability. Our simultaneous measurements of δDET and δ18OET reveal the difference between transpiration-dominated and evaporation-dominated periods. This study highlights the potential of simultaneous measurements of δ18OET and δDET.
Sponsored by EU
Please note: The access to the full text of the author's version will be granted after the end of the embargo period (24 months) on 17 February 2021.
These documents are avalilable under the license: