On the Computations of Specific Surface Area and Specific Grain Contact Area from Snow 3 D Images

Estimating the Specific Surface Area (SSA) of snow and firn using three-dimensional (3D) images is now widely used in the snow and ice community. However, little information is available about parameters that may impact the final quantitative measurements. In particular, important questions such as the accuracy of the numerical methods used, the possible limitations due to image resolution or the minimal Representative Elementary Volume (REV) for this particular quantity remain. To try to fill in the gap, we investigated different features of the numerical SSA measurement from 3D images. We obtained the principal following results: 1) the main standard SSA numerical approaches provide globally similar results. However, each method has its own inherent drawbacks: the stereological approach does not handle anisotropic structures properly, the marching cubes tends to overestimate the surface and the voxel projection method is more highly sensitive to image resolution. 2) The resolution limit strongly depends on the numerical method and on the snow type. 3) The REV seems to be attained with a cube of 2.5 mm edge for all of the studied samples. Based on a recently developed grain segmentation algorithm, a new SSA-related numerical estimator was defined and applied to snow samples: it provides an estimation of the Specific Grain Contact Area (SGCA), which seems a promising parameter to describe snow microstructure, e.g. to forecast the SSA that can be released by mechanical processing (neck breaking). The present study showed that the proportion of SGCA is particularly high for aged snow types. The results obtained provide a more comprehensive insight on the microtomographic approach and can be useful as well for snow modeling or other SSA measurement methods.