An important reassessment of the physics of non-equilibrium fractionation has been made in Cappa et al. (2003). Under non-equilibrium conditions molecular diffusion of the different water isotopomers produces greater fractionation of the lighter water molecules and explains the global deuterium excess value of about d = +10%o (if only equilibrium processes controlled evaporation, global excess would equal d = 0%). Mathematically this process was first described in Merlivat & Jouzel (1979): isotopic fractionation during evaporation is controlled by an effective fractionation comprising an equilibrium and a non-equilibrium (kinetic) part. The latter is affected by the relation of the molecular diffusivities between the different isotopomers. In kinetic gas theory molecular diffusivity is a function of the mass and the collision parameter of the respective molecules. Mainly focusing on the global deuterium excess value, Merlivat & Jouzel (1979) assumed a significantly different collision parameter for the HD16O molecule than for the H218O molecule, basing their argument on the broken symmetry of the deuterized water molecule. With this ad hoc assumption on molecular collision parameters the deuterium excess could be described successfully in the model of Merlivat & Jouzel (1979).
Cappa et al. (2003) carefully conducted laboratory experiments to measure the influence of non-equilibrium evaporation on the effective fractionation. They controlled the relative humidity of an air stream over an evaporation chamber and measured finally the isotopic composition of the vapour and the liquid phase. In their interpretation no different collision parameters for the different isotopomers have to be assumed. In order to explain the relation between bulk water temperature and observed evaporation fractionation, however, they give a major role to an evapo rative cooling affecting a skin layer on top of the water surface. Depending on relative humidity this evaporative cooling varies and leads to typical cooling of 1-3°C in this skin layer. Recomputing with these modified temperatures the corresponding effective fractionations, the authors could not only successfully explain their own measurements but also reinterpret former experiments. Their laboratory experiment waits now for a global in situ evaluation by measuring the isotopic composition of the first vapour formed over the ocean surface, which astonishingly is still unknown even after decades of using water isotopes. If the model of Cappa et al. (2003) is confirmed this has major consequences both for the description of the water isotopes in global atmospheric models (Jouzel et al., 1987; Hoffmann et al., 1998; Kavanaugh & Cuffey, 2003) and for the interpretation of the deuterium excess in palaeorecords. For today's conditions they predict, for example, a deuterium excess of the evaporative flux a couple of parts per mil larger than former estimates.
Was this article helpful?