Early formation of preferential flow in a homogeneous snowpack observed by micro‐CT

Abstract

We performed X‐ray microtomographic observations of wet‐snow metamorphism during controlled continuous melting and melt‐freeze events in the laboratory. Three blocks of snow were sieved into boxes and subjected to cyclic, superficial heating or heating‐cooling to reproduce vertical water infiltration patterns in snow similarly to natural conditions. Periodically, samples were taken at different heights and scanned. Results suggest that wet‐snow metamorphism dynamics are highly heterogeneous even in an initially homogeneous snowpack. Consistent with previous work, we observed an increase with time in the thickness of the ice structure, which is a measure of grain size. However, this was coupled with large temporal scatter between consecutive measurements of the specific surface area and of the statistical moments of grain thickness distributions. Because of marked differences in the right tail, grain thickness distributions did not show shape invariance with time, contrary to previous analyses. In our experiments, wet‐snow metamorphism showed two strikingly different patterns: homogeneous coarsening superimposed by faster heterogeneous coarsening in areas that were affected by preferential percolation of water. Liquid water movement in snow and fast structural evolution may be thus intrinsically coupled by early formation of preferential flow at local scale. These observations suggest that further experiments are highly needed to fully understand wet‐snow metamorphism and infiltration patterns in a natural snowpack.

Liquid water infiltration into a layered snowpack: evaluation of a 3-D water transport model with laboratory experiments

“…Also, the diameter of samples was consistent with the thickness of similar experiments in soils (see e.g. Hill and Parlange, 1972), whereas Avanzi et al (2017a) showed that preferential flow may be intrinsically coupled with wet snow metamorphism at grain scale. This suggests that small-scale experiments are appropriate for understanding the physics of this process in snow.…”

“…This simplified condition for infiltration into dry snow may lead to an underestimation of the expansion of preferential flow. Neglecting quick metamorphism in preferential flow paths (Avanzi et al, 2017a) may represent another cause of underestimation of preferential flow path size as grain growth promotes lateral spreading of water and expansion of paths. Although this model includes grain growth following Brun et al (1989) and Tusima (1978), modelling some specific conditions such as wet snow metamorphism at the boundaries between preferential flow paths and drier snow is still an open issue.…”

Wet‐Snow Metamorphism Drives the Transition From Preferential to Matrix Flow in Snow

“…The cold‐laboratory experiment considered in this paper includes micro‐CT observations of wet‐snow metamorphism in a nature‐identical snow sample (Schleef et al, ) that was subjected to periodical melt cycles with no refreezing (see Figure S1 in the supporting information for the schedule of melt cycles and the estimated height of the snow sample throughout the experiment). More details about this experiment are reported in Avanzi et al () as Experiment M2. We replicated this experiment using a multidimensional water transport model (Hirashima et al, , ).…”

“…Even though models for wet snow metamorphism have been developed based of experiments on real snow samples (Brun et al, ), these models have rarely been validated for natural snowpacks (Hirashima et al, ), and their accuracy and impact on liquid water flow processes through snow remain largely unknown. Here, we first validate wet‐snow metamorphism patterns derived using a widely used parametrization of wet‐snow metamorphism (Brun et al, ) against observations from a cold‐laboratory experiment that was representative of seasonal snow (Avanzi et al, ). We then use the comparison between model simulations and observations to investigate the role that grain‐growth rates play in ruling the evolution of water flow regimes with time and in particular the transition from preferential flow to matrix flow.…”

“…By coupling wetting front propagation with grain growth (Avanzi et al, ), wet‐snow metamorphism further complicates the understanding of water flow through snow, because snow structure is not invariant with time. Even though models for wet snow metamorphism have been developed based of experiments on real snow samples (Brun et al, ), these models have rarely been validated for natural snowpacks (Hirashima et al, ), and their accuracy and impact on liquid water flow processes through snow remain largely unknown.…”