This project investigated water movement in porous materials and snow using magnetic resonance imaging (MRI). The study was divided into two main sections: experiments on model porous media, including 3D-printed snow samples, and experiments on natural snow under melting and infiltration conditions.
MRI techniques were used to capture spatial and temporal dynamics of water movement in porous materials. While high-resolution 3D imaging was used to observe the before and after states of drainage, faster one-dimensional (1D) vertical profiling was employed to track rapid localized fluid flow fluctuations. These so-called Haines jumps occur in less than 0.1 seconds and were captured using imaging speeds down to 17 milliseconds per frame. Additionally, the phase-angle data and second-echo imaging provided insights into higher-order fluid flow properties such as turbulence and acceleration.
To better understand flow behavior in snow, 3D-printed porous structures based on CT scans of real snow samples were tested. These models allowed controlled studies of water movement in porous media that resemble natural snow at the microstructural level. Although differences in scale, wettability, and phase transitions meant they did not fully replicate natural snow conditions, they revealed a non-trivial increase in signal heterogeneity.
Experiments on natural snow posed significant challenges due to temperature control constraints and limited space for meltwater collection. The focus of these experiments was therefore on snow melting, meltwater accumulation, and interactions with soil-like porous media. The rapid profile images were interleaved with 2D slices to track meltwater distribution and calibration of liquid water content. An example of such an experiment is included in Figure 2, where we observed meltwater accumulating at the interface between snow and soil, where direct percolation was prevented by a capillary barrier. The local pressure head increased with continued melting, eventually overcoming the entry pressure, at which point percolation into the underlying material began. The snow gradually compacted, preserving layered liquid water distributions, suggesting slow restructuring of the ice matrix.
In rain-on-snow experiments, a key comparison was made between water infiltration in snow and a bead-pack sample. While water infiltrated gradually and evenly in the bead pack, snow exhibited discontinuous and heterogeneous liquid water distribution from the start. This suggests that melting ice has unique wetting properties, which facilitate rapid water transport into smaller pores at a much larger length scale than soil-like porous media. This distinction indicates that the hydraulic behavior of snow is fundamentally different from other porous materials.
These findings provide valuable new insights into snowmelt and water infiltration dynamics, which can help refine hydrological models and improve predictions of snowmelt and runoff behavior.
