Stable isotope fractionations in the evaporation of water: The wind effect

We performed pan evaporation experiments with the objective of exploring the behaviour of the long-standing Craig–Gordon (C–G) stable isotope model for evaporation under different conditions of air turbulence. The water lost through evaporation was automatically replenished so that a steady isotopic composition was reached, the value of which depended on the isotopic composition of the replenishment water and environmental parameters like temperature, relative humidity and isotopic composition of the atmospheric vapour, and the air turbulence index. The pans were exposed to artificial winds ranging from 0 to 2.5 m/s to change the air turbulence index, which governs the repartition between vapour transported by molecular diffusion and turbulent diffusion. Our data revealed that for wind speeds >0.5 m/s the isotopic composition of the evaporating water deviated from that predicted by the C–G model. This deviation was hypothetically attributed to microdroplets of liquid water removed by the wind without any isotopic fractionation. Isotope mass balance equations allowed us to quantify this water loss, which at wind speeds of ~2 m/s reached 10% of the total evaporation losses. An alternative kinetic evaporation model was proposed whereby the equilibrium layer and the atmospheric laminar layer above the evaporating water of the C–G model were destroyed by the wind and evaporated water molecules were directly injected into the atmosphere. In this model, the isotopic fractionations were due to the slower kinetics of hydrogen bond breakage between molecules in liquid water when heavy isotopes are involved. Accordingly, our data suggested that for isotope water balance studies where winds are frequently above 2 m/s, the C–G model may be inadequate without appropriate corrections for spray vaporization, or the introduction of appropriate kinetic isotope fractionation factors.