Partitioning Water Vapor Fluxes by the Use of Their Water Stable Isotopologues

Abstract

Water stable isotopes are powerful tracers for partitioning of the terrestrial ecosystem water vapor fluxes into process-based components, i.e. evapotranspiration (<em>ET</em>) into soil evaporation (<em>E</em>) and plant transpiration (<em>T</em>). The isotopic methodology for <em>ET</em> partitioning is based on the fact that <em>E</em> and <em>T</em> have distinct water stable isotopic compositions, which in turn are due to each flux being differently affected by isotopic kinetic effects. To use stable isotopologues of water in <em>ET</em> partitioning studies, a good knowledge of the isotopic (equilibrium and kinetic) fractionation effects is crucial. While the temperature-dependent equilibrium fractionation factor is well characterized (Majoube 1971), the kinetic fractionation factor (α_K), relevant, e.g., during soil evaporation, needs further investigation. <br /> In order to address this knowledge gap, we conducted a series of three different long-term bare soil evaporation experiments (differing in soil-water availability and aerodynamic conditions) to obtain α_K values from the collected isotopic data and the inversion of a well-known resistance-to-transfer model (i.e., the Craig and Gordon (1965) model). The isotopic composition of the soil water (δ_s) vapor was monitored non-destructively by using gas-permeable tubing (Rothfuss et al. 2013). The Craig and Gordon (1965) model was used in two different approaches. The first approach uses the Keeling (1958) plot to obtain values for the isotopic composition of the evaporation (δ_E). The second approach uses the slope of the linear regression between δ_s²H and δ_s¹⁸O. Results showed that the largest source uncertainty in the computation of α_K stemmed from the uncertainty associated with the δ_E values modeled with the Keeling (1958) plot method. In the second approach α_K values were within the theoretical range proposed by Dongmann et al. (1974) and Mathieu and Bariac (1996), which pointed to the prevalence of the turbulent transport of water vapor under saturated and unsaturated soil conditions.<br /> A variety of studies use different measurement techniques to estimate the isotopic composition of <em>ET</em> (δ_ET), <em>T</em> (δ_T) and δ_E for <em>ET</em> partitioning at the field scale. Here, especially the long-term monitoring of δ_E and δ_T is challenging. For this, non-destructive soil water stable isotopic monitoring using gas-permeable material is a promising tool. We tested the method of Rothfuss et al. (2013) to measure δ_s in an ET partitioning field campaign during one growing season of sugar beet (<em>Beta vulgaris</em>). To evaluate this method, the estimates of the transpiration fraction (<em>T/ET</em>) obtained from non-destructive soil profiles (P) were compared to the destructive soil sampling (S). In addition, the isotope-based approach was compared to <em>T/ET</em> estimates obtained from the combination of micro-lysimeter and eddy covariance (EC) measurements. Results showed discrepancies between the δ_E values obtained from S and P, which are in line with recent findings for different sampling methods (Orlowski et al. 2016a, Orlowski et al. 2018). However, the mean absolute deviations found between isotope-based and lysimeter-based <em>T/ET</em> estimates were more than three times higher than the differences between S and P. This underlines the great potential of gas-permeable tubing for long-term monitoring in the field and calls for further investigation of the isotopic offsets between direct measurement and extraction methods.<br /> The long-term monitoring of δ_E by using gas-permeable material is only one challenge for <em>ET</em> partitioning studies. To provide sub daily estimates of <em>T/ET</em>, the long-term monitoring of δ_ET and δ_T should be improved. Therefore, a review of the current and past literature was written about the progress and challenges of isotope-based ET partitioning. In total, we reviewed 31 studies and analyzed which methods provide the most promising approach for the long-term monitoring of δ_ET, δ_T and δ_E. Next to gas-permeable material for the determination of δ_E, we encourage the development of experimental setups allowing for the determination of <em>ET</em> isotopic fluxes by combining EC measurements and high-frequency laser spectroscopy. The use of the gas-permeable material has also a great potential to measure δ_T. Albeit up to now only one study (Volkmann et al. 2016) showed that in-situ monitoring of δ_T in tree xylem is possible, this approach should be further developed for medium-sized plants (e.g. maize) and, in the longer term, thin-stem (cereal) plants.