Isotopic enrichment of liquid water during evaporation from water surfaces

The predictions of isotopic composition of evaporation (δE) from a water body are often made with the Craig–Gordon model on the assumption that the isotopic composition of water surface undergoing evaporation (δL,e) is the same as that of the bulk water (δL.b). The validity of this well-mixed assumption is not known due to the lack of the δE measurements. In this study, a tunable diode laser (TDL) analyzer was deployed to make fine time-resolution measurements of H218O and HDO in vapor in a home-made flux chamber. Our results show that the surface enrichment was substantial: on average, the surface water was 7.5–8.9‰ more enriched in 18O and 12.6–16.5‰ in D than the bulk water, the exact value depending on the choice of the kinetic fractionation factor for evaporation. Also we reported greater uncertainties of D than that of 18O in the estimation of surface enrichment; The measured 18O surface enrichment was statistically different from zero but the D enrichment was not. The kinetic effect appeared stronger under conditions of higher evaporation. In addition to leaf water, we suggest that the surface enrichment should also exist in lakes and evaporation pans although it is likely lower than our enrichment values. The well-mixed assumption may have also affected previous efforts of determining the 18O/16O and D/H diffusivity ratios in air.