Experiments on fracture propagation and avalanche release

Avalanche release requires four elements: (1) snow layering: a weak layer beneath a slab, (2) a trigger, (3) a spatially rather uniform snowpack structure and (4) a steep enough slope (larger than 30°). If the stability of the snow cover is in near critical state, due to external loading (e.g. snowfall or a skier) an initial failure can occur in a weak layer underlying more cohesive layers, called the slab. If the initial failure in the weak layer reaches a critical size it will propagate outwards, provided the snowpack layers are continuous over a large enough area. Finally, if the slope is steep enough, a slab avalanche will release (Figure 1).

Field experiments

By observing cracks propagating through weak snowpack layers in field experiments with a high-speed camera, it is possible to observe the processes described above with high detail (Figure 2). With various image analysis techniques (Figure 3), we investigate the behavior of snow layers in field experiments mimicking the different stages leading to avalanche release, allowing us to derive relevant material properties of snow as well as calibrating and validating numerical simulations.

To obtain detailed snow cover information, we also measure snow cover properties with a high resolution snow micro-penetrometer (SMP). Mechanical properties of the snow layers can be derived from these high resolution penetration resistance measurements. The information derived from these measurements can then be used as input for numerical simulations of the mechanics of our field experiments.

Numerical simulations

One of the major drawbacks of field experiments is the lack of reproducibility, in part due to the spatial variability of snow cover properties as well as the complex interrelation of the relevant parameters, such as density, strength and elastic properties. It is therefore nearly impossible to measure all parameters for each single experiment, increasing the uncertainty in the results. Numerical simulations of fracture processes in snow (LINK) can help alleviate the problem and allow us to assess the influence of the relevant parameters in a more systematic manner. Of course, such simulations can only be used if they are able to accurately reproduce fracture processes in snow. A detailed comparison with results obtained from field measurements is therefore always necessary.