This is a video about how natural or spontaneous dry slab avalanches release.
Jörg Schweitzer from the WSL Institute for Snow and Avalanche Research in Switzerland
helped with the content of this video and the research behind it.
For a slab avalanche to release, fractures or cracks must occur at the crown, both flanks,
the stock wall, and in the underlying weak layer shown in pink. Since the 1980s we have known that
the first failure occurs in the weak layer. The other fractures are a consequence of the
initial failure and the propagating crack in the weak layer. The early steps in the release
of a natural dry slab avalanche shown in red differ from human triggering shown in blue.
This video focuses on the steps of natural avalanche release. One, snowpack deformation due
to loading or warming. Two, formation of micro failures in the weak layer. Three,
subcritical growth of a deficit zone which, when it reaches a critical length,
results in a self-propagating crack. This is usually followed by crown,
flank, and stock wall fractures and slab release if the slope is steep enough.
The initial stage of human triggering shown in blue is simpler and covered in another video.
The latter stages, shown in black, are the same for natural dry slab avalanches and for human
triggering. This video starts with how the snowpack deforms on a slope and how that sometimes
leads to failure in a weak layer. If the failure zone becomes long enough in the down slope direction,
then a crack will propagate in the weak layer. As a consequence of a crack propagating in a weak
layer, the crown, flank, and stock wall fractures will occur, leading to slab avalanche release
if the slope is steep enough. Let's start with the deformation of the snowpack over time.
Consider a column of the snowpack. Over time, that column will creep. We can think of that creep
as having a component of shear parallel to the slope and a component of vertical settlement.
It is the slope parallel shear shown in red that plays a greater role in the failure of a weak
layer. This video by Thomas Exner shows the creep of a fresh dry snow layer over two and a half hours.
Toothpicks have been placed in the pit wall and I've circled the second one from the top.
Watch the toothpicks in the pit wall move. The green arrow shows the vertical settlement.
The red arrow shows the movement parallel to the slope. It is mostly slope parallel shear
that contributes to the failure of weak snowpack layers. This photo of a crown fracture of a
storm snow avalanche shows the slab, weak layer, and bed surface. It has long been believed that
the initial failure in a weak layer is caused by a sufficiently fast slope parallel shear.
Recent cold lab studies by Riveger and colleagues has shown that this is true for slopes over 30
degrees. This slide shows a weak layer in pink underneath the slab. Both are shearing, but the
weak layer is shearing faster than the slab above or the snowpack below because it is less stiff
than the adjacent layers. Snow is very sensitive to the rate of shearing. Shear it quickly and it
fails because bonds between grains are being broken faster than new bonds are forming.
Here is the puzzle. The shear deformation on a typical snow slope is about four orders of magnitude
too slow to cause a propagating crack. That is about 10,000 times too slow. Let's look at where
and when shear is faster. During snowfall or warming, shear speeds up, but not enough for a
weak layer to form a propagating crack. Shear deformation is likely faster on steeper sections
of a slope or near certain terrain features, but that is still not enough. As noted in previous
slides, shear is faster where adjacent stiff layers concentrate shear in a weak layer,
but that is not enough. Even when combined, these shear concentrations still do not shear
weak layers fast enough for a crack to propagate. We need an additional factor that is described
in the next few slides. During the shearing that is part of downslope creep, some bonds between
grains are being broken and other bonds are forming from new contacts. Where bonds are
breaking faster than they are forming, micro failures, probably less than a centimeter in
length, will form initially. These are most likely to form where shear is concentrated at the upper
and lower boundaries of the soft weak layer, that is next to the adjacent stiff layers.
If the shearing is fast enough, some of these micro failures may coalesce or grow until they
become a failure zone many centimeters long. When a failure zone does not fully support the shear
stress from the overlying slab, it is called a deficit zone or shear band. A deficit zone is a
temporary local failure in a weak layer. I say temporary because deficit zones don't last long.
Since a deficit zone does not fully support the overlying slab, there is some tension parallel
to the slope here and some compression parallel to the slope here. Because the overlying slab
is not fully supported, shear stress is concentrated at the boundaries of the deficit zone. So in
addition to the shear concentrations mentioned three slides ago, it is the shear stress concentrations
around deficit zones that represent the final condition necessary for the failure zone to grow
and potentially become a propagating crack in a weak layer. And we have a long list of researchers
to thank for the theory behind deficit zones. Now let's suppose we could look at the distribution
of shear strength of a weak layer on a slope. On most places on this hypothetical slope,
the shear strength is 700 Pascals. Some places are stronger than average, say 800 Pascals. And
there are some weak errors where the shear strength is only 600 Pascals. Suppose the overlying slab
is uniform 50 centimeters thick with the density of 210 kilograms per cubic meters. If the slope
angle is 38 degrees, then the uniform slab will cause a shear stress on the weak layer of 500
Pascals everywhere on the slope. So everywhere on the slope, the shear strength is greater than
the stress. This is a stable slope. Under these hypothetical conditions, natural avalanches will
not start. Now consider heavy snowfall on this slope. Although weak layers slowly gain strength
in response to gradual loading such as snowfall, let's suppose the snowfall is so fast that this
particular weak layer does not gain appreciable strength. To keep things simple, assume 25 centimeters
of new snow falls with the density of 100 kilograms per cubic meters. That adds a shear stress of 150
Pascals for a total shear stress of almost 650 Pascals on the weak layer. On most places on the
slope, the shear strength exceeds the shear stress. But the slope is no longer stable because in the
orange zones, the shear stress is now greater than the shear strength. These local zones of failure
are called deficit zones or shear bands. This slope is not stable. If either deficit zone grows to
a critical size, then a crack will propagate through the weak layer and a slab avalanche will
release if the slope is steep enough. Of course, these simple diagrams ignore variations in slope
angle, slab thickness, and slab density, all of which can contribute to the formation and growth
of deficit zones. Let's make our hypothetical slope slightly more realistic. The deficit zones in the
weak layer shown in orange are growing. Remember that the usual causes of growth are either loading
by snowfall or warming of the upper slab. The gray zones are fracture tough zones. These are
areas where the weak layer is stronger than average and, more importantly, with above average resistance
to a propagating crack in the weak layer. Now, let's suppose the large deficit zone in the lower
right becomes big enough to transition to a propagating crack in the weak layer. A crown
fracture might occur here. Alternatively, the snowfall or warming that was causing the deficit
zones to grow could stop. Then the deficit zone would shrink and heal. Little is known about
the life of a deficit zone, but researchers have estimated minutes to a few hours. Okay,
I know what you are thinking. There have not been nearly enough graphs in this video,
so here is one to get us back on track. The green curve shows the downslope length of
a deficit zone that is growing over time. It reaches its critical length at the end of the
curve. When this happens, a mostly brittle crack propagates so the length increases quickly as shown
by the vertical arrow. Moments later, the crack in the weak layer is long enough to cause a tensile
crown fracture through the slab and a slab avalanche will release if the slope is steep enough.
This graph is based on one by Bader and Saum, but of course other researchers contributed
to the development of the theory. Okay, it's about time for a photo of an avalanche. This one
released after the sun hit the slope, which would have increased creep in the upper snowpack,
which can increase shear in weak layers, including weak layers below the warmed upper
snowpack. We can't know where the deficit zone was located, but let's assume it was here.
When it reached a critical downslope length, the crack in the weak layer would have propagated
outward something like this. We can imagine that the crack that propagated along the weak layer
ran into fracture tough zones here, here, and here. Let's sum up. The initial failure in a
weak layer is caused by slope parallel shear, which can cause micro failures in the weak layer.
Under the right conditions, these sometimes grow into a deficit zone that locally and temporarily
does not support the overlying slab. Deficit zones are a plausible theory based on fracture
mechanics that explain how stress concentrations around failure zones can cause a deficit zone
to grow. Although no one has ever observed or measured a deficit zone in the field or in the
cold lab, it is our best theory for how a failure zone in a weak layer being sheared can grow.
The stress concentrations at the boundaries of a deficit zone enable it to grow. When the downslope
length of a deficit zone becomes long enough, that is, it reaches its critical length,
it will become a propagating crack in the weak layer. It is sometimes called self-propagating
because a crack growth of, say, one centimeter releases more than enough energy to drive the
crack one centimeter ahead. For natural avalanches, researchers believe the critical length is one to
a few times the slab thickness, so for a slab that is 50 centimeters thick, the critical length is
probably roughly between 50 and 200 centimeters. When the snowfall or warming or whatever is
driving the faster shear stops, we expect deficit zones to shrink and to heal. The life
of a deficit zone is likely less than a couple of hours, maybe much less. Within that time,
it will either grow and lead to a propagating crack or it will heal. Because a self-propagating
crack grows quickly, within seconds it is followed by a crown fracture. On sufficiently steep slopes,
the flank and stop-wall fractures occur next and a slab avalanche is released. The initial failures
and subcritical growth of a deficit zone are unique to natural avalanches. Localize dynamic loads,
such as skiers, snowmobiles, or explosions, skip this step, and directly cause cracks in
weak layers. The ideas in this video are based on research by many people over many decades.
The author of some of the key papers are listed here. We hope this video has clarified some
concepts behind the release of natural dry slab avalanches and we hope the research continues
into this important and mysterious process. Here are some questions that will help you think about
the release of natural dry slab avalanches. Press pause now. When you have tried to answer
each of these questions, press play to check your answers. If some of these answers are not
making sense, you can re-watch all or parts of the video. Thanks for watching.
