Kinetic Fractionation of Gases by Deep Air Convection in Polar Firn
A previously unrecognized type of gas fractiona- tion occurs in firn air columns subjected to intense convec- tion. It is a form of kinetic fractionation that depends on the fact that different gases have different molecular diffusivi- ties. Convective mixing continually disturbs diffusive equi- lib...
| Published in: | Atmospheric Chemistry and Physics |
|---|---|
| Main Authors: | , , , |
| Format: | Text |
| Language: | unknown |
| Published: |
Dartmouth Digital Commons
2013
|
| Subjects: | |
| Online Access: | https://digitalcommons.dartmouth.edu/facoa/543 https://doi.org/10.5194/acp-13-11141-2013 https://digitalcommons.dartmouth.edu/context/facoa/article/1545/viewcontent/acp_13_11141_2013.pdf |
| Summary: | A previously unrecognized type of gas fractiona- tion occurs in firn air columns subjected to intense convec- tion. It is a form of kinetic fractionation that depends on the fact that different gases have different molecular diffusivi- ties. Convective mixing continually disturbs diffusive equi- librium, and gases diffuse back toward diffusive equilibrium under the influence of gravity and thermal gradients. In near- surface firn where convection and diffusion compete as gas transport mechanisms, slow-diffusing gases such as krypton (Kr) and xenon (Xe) are more heavily impacted by convec- tion than fast diffusing gases such as nitrogen (N2) and ar- gon (Ar), and the signals are preserved in deep firn and ice. We show a simple theory that predicts this kinetic effect, and the theory is confirmed by observations using a newly- developed Kr and Xe stable isotope system in air samples from the Megadunes field site on the East Antarctic plateau. Numerical simulations confirm the effect’s magnitude at this site. A main purpose of this work is to support the devel- opment of a proxy indicator of past convection in firn, for use in ice-core gas records. To this aim, we also show with the simulations that the magnitude of the kinetic effect is fairly insensitive to the exact profile of convective strength, if the overall thickness of the convective zone is kept constant. These results suggest that it may be feasible to test for the existence of an extremely deep (∼30–40 m) convective zone, which has been hypothesized for glacial maxima, by future ice-core measurements. |
|---|