Dry Snow Temperature Gradient Metamorphism: Is Our Picture Correct?
Abstract
Temperature gradient metamorphism (TGM) was the subject of many studies in the past 100 years. However, a direct observation of the transport phenomena was impossible, since an observation automatically destroyed the sample. Using 4D micro-tomography, we could observe the evolution of structures in situ under temperature gradients. Combining the time-lapse tomography with microscopic temperature and concentration field modeling, the temporal evolution of heat and mass transport can be calculated and visualized. We found that the prevailing concept of larger grains growing on the expense of smaller ones is wrong under temperature gradient conditions. Instead, the whole ice matrix is continuously replaced by sublimation and deposition. The slow average growth of the structures is caused by population dynamics: larger structures have a longer residence time than smaller ones (Fig. 1). This observation solves one of the larger mysteries of snow metamorphism, namely why the diffusion coefficient (Deff) of water vapor should be much higher than in air (Yosida et al, Cont. Inst. Low. Temp. Sci, 7:19-74 1955). Our measurements and simulations show that Deff is not enhanced, but is the same as in air. This was already suspected by Giddings and LaChapelle (JGR, 67:2377-2383,1962), as they interpreted the results of Yosida as an experimental artifact. Later models seem to have overlooked this interpretation, and are based on an incorrect quasi one-dimensional arrangement of ice spheres. The fundamental and dominating process during TGM is therefore vapor diffusion, and the intensity dictated by the temperature gradient. Our measurements of mass flux show that the vapor flux on a macroscopic scale is independent of the microstructure. The development of future snow simulation models incorporating changes in microstructure not based on empirical observations, but on effective vapor mass flux, will now be feasible. The picture of snow metamorphism is much more dynamic than anticipated, and our results may also improve the understanding of chemical processes within snow. Figure 1: Residence time of ice matrix in a depth hoar structure after 27 days at a temperature gradient of 50 K m -1. The oldest ice is younger than 200 h.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2009
- Bibcode:
- 2009AGUFM.C31B0443S
- Keywords:
-
- 0736 CRYOSPHERE / Snow;
- 0798 CRYOSPHERE / Modeling