Science Interviews


Sat, 19th May 2012

Indoor Avalanches - Planet Earth Online

Nathalie Vriend at the University of Cambridge

Listen Now    Download as mp3 from the show From PC to Plane - Making New Metals

With summer on its way (apparently) scientists who study the physics of avalanches have had to develop ingenious ways of studying what happens when snow cascades down slopes at up to 100 miles per hour.  And they don’t have to go on the piste to do it.

Nathalie Vriend at the University of Cambridge’s Centre for Mathematical Sciences simulates avalanches by using what looks like a big metal phone boxAvalanche on Mt. Timpanogos Utah leaning on its side.  There are no glass walls – just a rectangular frame and a narrow conveyor belt on a slope. 

Planet Earth podcast presenter Sue Nelson went to see Nathalie’s experiment for herself and discovered something familiar sliding down that slope – and it wasn’t snow… it was sand...

Nathalie - Both snow and sand consist of small particles and these particles, when they start to move, they interact and collide with each other.  Because of that the physical processes that occur on this small scale are actually very similar.  So, we're doing our experiments with sand instead of snow.

Sue  - Can you demonstrate to me what you actually do?  Can we turn this huge contraption, this conveyor belt of sand, on?

Nathalie  - Yes of course, I am happy to...  I'm opening a valve that opens the centre flow.  The sand is being let in at the top of the incline.  The distance where the sand can avalanche is about three metres long, so we're bringing the sand to the top and it starts to avalanche down.

Sue - I can see now some of the sand coming out of a little tube at the top of the slope which is just a bit higher than me, so it's a couple of metres high.

Nathalie - At this moment the sand is not flowing fast.  What we can do is actually open the valve a little bit higher so that the sand starts to come out in bigger quantities.

[Sound of falling sand]

Sue - Oh gosh!  That's really interesting; it looks like your slope of sand - the central bit - now looks like a flowing molten river of sand.

Nathalie - Indeed, indeed.  And if you look very carefully you can see that the middle part is flowing very fast, but there are bits on the side, like dykes-

Sue - Oh yes, they're turning round and not moving so much.

Nathalie - Exactly, and there are actually some grains that are static or not moving at all.

Sue  - Is that what happens when snow goes down a slope?

Nathalie  - Yes indeed.  When a mass comes down a mountain it goes the fastest in the middle and the velocity goes further to zero towards the sides. Therefore if you look very carefully at deposits on the mountain you can always see that the snow carved its way through a deposit and left debris on the side.

Sue - How did you choose the angle of your slope?

Nathalie  - The angle that avalanches usually occur is between 30 and 45 degrees.  That's actually pretty steep, it's much steeper than you would ever drive on.  Let me change the flow rate right now because then you can see some other features as well.

Sue - Oh gosh, that's interesting. Instead of a free flowing river it was almost like a drop of treacle.

Nathalie - So in this case, because we reduced the flow rate there is not enough sand coming through to form this continuous river.  Because of that the sand accumulates to the top of the incline and just starts to avalanche, the slope fills when there is a certain amount of sand available.  So you get these intermittent avalanches and they almost look like tongues going down the incline.

Sue - What do you hope to learn by studying this flow, this avalanche of sand?  Because a certain amount is known already?

Nathalie - It is really difficult to understand exactly what's going on because as you can imagine, if you have a flow of water down a incline, researchers know pretty well how to investigate water down an incline.  But if you have a solid material the physical laws aren’t very well understood.  In this case you have this odd mixture of different phases; as you look in the middle of river of sand it looks like a flowing stream and it goes very fast and there is a lot of motion going on. But if you look at the side of the river it is static and that's really hard to model.  So the bigger picture is that we want to understand where the avalanches are going, how far they get, how forceful and how big they are and what kind of pressures occur in them.  The reason is that we want to understand where we can build buildings and roads and where it is safe for people to live.

In the past people relied on historical records but they may not be very accurate anymore because the climate is changing so, snow fall and temperatures are changing as well.  Also, not every point in the world has historical records.  We want to be able to model snow avalanches from a physical point of view to actually be able to apply it to every valley and every mountainside in the world.



Subscribe Free

Related Content


Make a comment

But, I wonder if the characteristics of sand (wet or dry?) is the same as snow which can also be rather icy, or not.  However, it is interesting that the rate of accumulation affects the scale of the avalanche.  I would think one would also gain information by looking in the natural environment, such as the Alps where the avalanches must be common.

As far as new avalanche technology, one of the newest things on the market is the Avalanche Airbag, which is supposed to significantly increase a person's chance of survival.  With some skiers already surfing an avalanche down the mountain. Video in Action

CliffordK, Tue, 22nd May 2012

Being a backcountry skier, I listened to the avalanche story with interest. Though I only listened to the podcasted portion, I found it curious that the focus only seemed to be on the flow dynamics. There is a whole lot more to snow avalanches than what I expect can be modeled with sand. There is the stratigraphy of the deposited layers, buried hoar, depth hoar, temperature faceting, windloading, slab formation, and weather (temperature/wind) changes. All these play a factor on whether an avalanche will initiate spontaneously, or with an external (human) trigger, or whether the slope will be inherently stable. The type of avalanche, dependent on the above conditions, is also important to the flow characteristics. A slab avalanche, where a large mass of snow begins to move at once is much more inherently dangerous to a skier / climber / structures as it can entrain a massive amount of snow in an instant. Loose snow slides can be relatively benign, as the snow may never build up sufficient mass to propagate (though a skier could potentially still be buried in a terrain trap). A wet slide can be so slow as to allow people to clear out of the way, but the weight can be destructive, and can also carry unwary people over terrain hazards. Obviously, sand as a model can only cover certain aspects of avalanche physics. It would be great to hear a more in-depth (no pun intended) discussion of avalanches, it is a fascinating subject. jschuh, Thu, 24th May 2012

interesting indeed,

My initial reaction was similar to Jschuh - there are many factors in an avalanche that cannot be easily modeled with sand. Most of these have been studied (at least emprirically) in the field.  However, as the purpose of the reserach appears to relate to macroscale hazards to infrastucture (rather than small scale hazards to individual skiers/ mountaineers) it clearly has a point.

That said, the aspiration to " able to model snow avalanches from a physical point of view to actually be able to apply it to every valley and every mountainside in the world." whilst noble would have to consider all of the other factors that sand cannot model - such as stratigraphy, initiation & substrate. Mazurka, Thu, 24th May 2012

I only see that the same way computer software for making models of crowd control in built up areas and results prediction of stampedes operates. With water or sand you could create some type of viscosity and stickiness model from the polar covalent attractions result but i don't think either will be truly successful at predicting the final result. nicephotog, Wed, 6th Jun 2012

See the whole discussion | Make a comment

Not working please enable javascript
Powered by UKfast
Genetics Society