The magnitude of isotopic fractionation during sublimation of ice remains poorly constrained. Field and laboratory studies that span decades have persistently shown conflicting results. A better understanding of fractionation during sublimation is needed to improve interpretation of alpine hydrology, glaciology, paleoclimate and planetary histories that rely on stable isotopic records stored in icy reservoirs. At the core of the problem is the question of whether sublimation occurs as a layer-by-layer process with no fractionation or whether diffusion within the ice and vapor-ice exchange generate fractionation. We present results from an experiment where we suspended ice spheres in temperatures ranging from −25 to -10 °C in an unsaturated atmosphere and used a Rayleigh distillation model to estimate fractionation of the spheres. A small, yet statistically significant and repeatable, isotope fractionation (103lnα18O of ~ −0.6 ‰ and 103lnα2H between −3 and − 4 ‰ or α = 0.999, α = 0.994 to 0.997, respectively) was found during the sublimation of ice. The results definitively indicate the presence of fractionation yet the values are an order of magnitude smaller than would be predicted for equilibrium fractionation at this temperature and humidity. By assuming a porosity for the ice of 0.0005 % - which is typical for ice frozen under these conditions - we estimate a sufficient increase in diffusivity relative to solid ice to explain the observed fractionation values. The results help to reconcile how fractionation estimates during sublimation could vary between experimental and observational studies where porosity is not controlled for and can vary substantially across a continuum from porous firn layers to low porosity ice deep in glaciers.
Read full abstract