What makes snow stick?

It doesn't matter where you are or how old you might be, when it snows, there's always a great temptation to go outside and have a snowball fight. But some snow just doesn't cut it. Either it's too powdery to hold together, or it's so hard that it could be classified as a weapon. To clear up this conundrum, Tricia Smith spoke to Gareth Rees and Becky Dell, in search of the science behind that satisfying splat...

Tricia - I visited the Scott Polar Research Institute at the University of Cambridge to find out what makes snow stick.

Gareth - Snow is a mixture of ice in the form of pretty small crystals, and air, and nearly always some liquid water as well.

Tricia - Does snow always have a little bit of water in it or can it sometimes exist without water?

Gareth - This is not completely known at the moment, I think. But what we suspect is true is that there is always a tiny bit of liquid water in snow. If the temperature of the snow is not too far below zero Celsius, say down to minus five minus ten, then there can be a noticeable amount of water, maybe 1%.

Tricia - So why is it that some snow will pack down when I, when I squeeze it and some won't?

Gareth - The main difference between kinds of snow that will compress into a nice snowball, the kind that you can throw satisfactorily and the kind that doesn't, that just stays as powder and falls apart again, is really its temperature. The powdery stuff is too cold. What you want to happen when you make a snowball is that you want to weld the ice crystals together. And the way you do that is you apply some pressure to the snowball, you squeeze it with your hands. Ice is unusual as a material in that when you apply pressure to it, its melting point actually decreases. So it melts a little bit when you squeeze it and the water, then refreezes when you take the pressure off, so when you stop squeezing the snowball, but now what you've got is ice crystals that have been welded to each other by more ice. So the thing becomes mechanically stronger than it was. So it's a little bit denser and quite a lot stronger, so it won't fall apart immediately.

Tricia - When I was at university, I was shown an experiment where they hung a metal wire on a cube of ice and they hung weights off each end. And what happened was the wire passed through the block of ice, but it didn't leave a gap. There was ice left behind as well. So can you explain that experiment to me?

Gareth - That experiment is usually called regelation, that just means refreezing, and it's really a good model of what happens when you make a snowball. The weighted wire is applying pressure to the ice and that succeeds in melting the ice. But above the wire, the pressure is lower again, so the water refreezes to ice. So what happens is that there is a little area of melted ice below the wire, but reformed ice above the wire. So the wire just travels down through the ice refreezing behind it, as it goes. And then eventually the wire drops out at the bottom, but the ice hasn't been cut into two.

Tricia - One of the interesting properties of snow, or of water and ice, is that ice floats on water. So are those things related?

Gareth - Yes, these things are related. What happens in ice, is that the water molecules, the H2O molecules, they arrange themselves in a very specific way. The forces between the molecules are very strong. So they have a very fixed idea about how to arrange themselves, and it's actually quite open. If you melt it into water, the average distance between the molecules actually decreases. So water is more dense than ice, and that's the reason why ice floats in water. And it's also the reason why the melting point of ice decreases when you squeeze it, because you're trying to force it into a smaller volume and it can do that by turning into water.

Tricia - Where in the world would we find snow that is the best for making snowballs?

Gareth - It would need to be somewhere not too cold. So I wouldn't go to the middle of Antarctica to have a snowball fight. I think we'd just find powder snow there. I think you'd want to be somewhere on the edges of regions that tended to get a lot of snow, edges of mountain ranges, fringes of cold regions.

Tricia - Who'd have thought a snowball fight in Antarctica might not be the best idea. Well, just to be sure, I reached out to my friend, who's in Antarctica right now. She agreed to go outside, make a snowball and let us know how it went.

Becky - My name's Becky Dell, and I'm a glaciologist working at the University of Cambridge. Currently, I'm in Antarctica on Alexander Island, living in Fossil Bluff hut, working on George VI ice shelf. Luckily it was pretty snowy this morning, so the snow is now fairly wet and malleable. However, a couple of days ago, it was much drier here, and it was almost too powdery to form a snowball. I'm going to try and make a snowball now, and I'll talk you through what happens.

Snowball - <snowball noises>

Becky - I'm not going to lie, the snow's actually pretty tough. It's formed a ball pretty quickly, but it's a fairly angular jagged ball, and I'm having to apply quite a lot of pressure. So rather than a ball, I'd call this... a cube. I've made a snow cube. No one to throw it at though, because they're all inside having a cup of coffee. So I'm going to go and join them.

Tricia - I'm not sure I'd want to be hit by a snow cube... sounds painful. So there you have it. Leave the skiing for the mountain tops because the snow in the valley is much more likely to weld together into the perfect snowball. Before I left, Gareth had one more, very important piece of advice for me.

Gareth - The very best snow for a snowball is the snow that's right next to you, because that's when you want to make a snowball. So if it snows in your garden, use that snow.