Field investigation into the diffusion of semi-volatile organic compounds into fresh and aged snow

Empirically derived field diffusivities were determined over a 24 h period for a selection of anthropogenic chlorinated chemicals in both fresh and aged snow at Tromsø, Norway. Diffusivities in fresh snow were 4.91×10−2, 4.79×10−1, 4.75×10−2 and 4.75×10−2 cm2 s−1 for PCB-6, HCB and α-/γ-HCH, respectively. These field diffusivities were compared to theoretical diffusivities (assuming no interaction between the chemical vapour and ice surfaces) and effective diffusivities (assuming sorption of the chemical vapour to the ice surface). Theoretical diffusivities were the highest values and were in good agreement with the empirical diffusivities derived from the field data. This suggests that the test chemicals used in this study have a low affinity for the snow/ice surfaces during diffusion into the snow. Differences in calculated diffusivities between these compounds can be largely accounted for by their physical–chemical properties, notably their snow interfacial-air partition coefficients (Kia snow). However, using calculated values of Kia snow to describe sorption to the snow surfaces greatly reduced the effective diffusivities of the chemicals relative to the empirical field diffusivities, and calls into question the accuracy of Kia snow and/or its suitability for describing snow/air interactions at the relatively mild temperatures encountered in this study. Comparison of diffusivities between fresh and aged snow revealed similar values (within the same order of magnitude) and may be due to similar porosities between the snow types, although this parameter was not measured. Mass transfer coefficients were determined using the empirical field diffusivities and depth of the snow-layer, allowing chemical fluxes from snow to air to be calculated through use of the Whitman two-film resistance model. Using derived fluxes the chemical half-lives (t1/2) in fresh snow were calculated assuming that snow metamorphosis did not occur. The half-lives were in good agreement to half-lives measured in a separate field study for POPs, and serve to validate the empirical diffusivities measured in this study.