[@LHCollision]

@LHCollision asked the Naked Scientists:
   
Could you explain to me exactly how supercooled water can stay liquid even temperatures well below freezing?

What do you think?

[freecw]

Low pressure.
Also, for water to freeze you need to form exact hydrogen bond lengths between the molecules to create the crystal structure of ice. If you added some solutes in water it can disrupt the formation of these bonds so the water remains liquid.

[Geezer]

I thought it was more the other way around.

Shouldn't the question really be:

"Why does water crystallize at temperatures below 0°C at STP?"

I think it requires some sort of agitation or impurity to trigger the crystallization process. Without those conditions, it will remain liquid.

[Bored chemist]

A very tiny ice crystal in water has to push the liquid water away in order to grow. That means it has to overcome the surface tension of the water.
Sometimes a few water molecules' random motion will get them out of the way, but until that happens the ice can't start to form.

[hamdani yusuf]

It's said that a picture is worth a thousand words. I think a video is worth even more.

At least there are 2 interesting facts shown here:

Nucleation can start from a place seemingly lack of contamination.

Density of supercooled water doesn't seem to change when it freezes.

[hamdani yusuf]

Here are some explanations.

by Scishow

by Veritasium

In their explanations they don't mention the necessities for degassing the purified water, either by vacuuming nor by boiling it beforehand. So, dissolved air doesn't seem to be a problem.

[hamdani yusuf]

This video shows that temperature increases during freezing of supercooled water.

It also shows that metal from the thermometer doesn't trigger the freezing.

[hamdani yusuf]

Here's another video to analyze the behavior of supercooled water.
 

It doesn't start to freeze until the poured supercooled water reaches some volume. A hypothesis can be raised, like it waits until the glass cool enough to let it freeze. 

Another fact we can see here is that air bubbles are not good enough to start freezing.

[hamdani yusuf]

Here's another interesting experiment.

A single droplet of Pseudomonas syringae culture freezes a volume of 600 ml supercooled water of -7°C instantly.

[hamdani yusuf]

The freezing can also happen seemingly unprovoked.

See unprovoked supercooled water turn into ice.

12 000 Ws heat is released when the supercooled water unprovoced turn into ice. The temperature goes up 6 degC in less than 10sec.

It seems like you need cool air outside of the container to see this happen.

[hamdani yusuf]

It's said that a picture is worth a thousand words. I think a video is worth even more.

At least there are 2 interesting facts shown here:

Nucleation can start from a place seemingly lack of contamination.

Density of supercooled water doesn't seem to change when it freezes.

In their explanations they don't mention the necessities for degassing the purified water, either by vacuuming nor by boiling it beforehand. So, dissolved air doesn't seem to be a problem.

Since the water used was purified, then the cause of cloudiness in the frozen ice is likely come from dissolved gas in the supercooled water before it freezes, which become separated from the solid ice. If it's the case, then making superheated water from degassed/boiled pure water should produce clear ice. I think it's worth to try in an experiment.