Finding the permeability-porosity-relationship of porous media is a classical problem in many scientific fields. As the permeability is often difficult to measure, percolation theory has been popular to interpret limited observations. In recent years Digital Rock Physics – to compute the relationship on X-ray microtomographic (XRT) images – has received growing interest. Here we present such an approach for the porous medium sea ice, using GeoDict.
Sea ice consists of brine, solid salts, and air in pore networks and inclusions within a matrix of almost pure ice crystals. This microstructure changes during cooling/ warming, when the brine partially freezes/melts in pores to stay in thermodynamic equilibrium. This affects many physical sea ice properties, in particular the permeability, being of high relevance for sea ice in the climate system: By influencing the salinity and pore space evolution of sea ice, the permeability is a key factor for many other properties. E.g., by controlling the drainage of snow melt through sea ice, it has a considerable impact on the ice albedo.
The sea ice permeability has been suggested to exhibit percolation behavior with a critical brine volume fraction of 5 %. So far, this threshold was based on sparse observations. We have analyzed an extensive dataset of XRT images of Arctic Sea ice. Linking characteristic pore scales (PoroDict) to fluid permeability (FlowDict), we propose a percolation-based permeability model with a revised threshold porosity of 2-3%.
The data sets are made available to all researchers in the field and hosted by Math2Market under https://doi.org/10.30423/scandata.scan-2021-01
