Snow is usually imaged in air, the single crystals laying flat on some substrate such as glass. The method is relatively simple, but one must work fast to image the crystal before it appreciably sublimates. Sublimation first rounds out the sharp edges and then causes the crystal to shrink. Generally, this sublimation happens because the photographer is radiating too much heat to the crystal. Conversely, particularly in very cold conditions, the photographer’s breath may deposit fog near the crystal, causing the crystal to grow.

Such issues vanish if one instead captures the snow in a cold fluid before taking the image. To work, this fluid should not dissolve the ice, be less dense than ice, be fluid enough to completely spread over the crystal surface, and be transparent. Other than preserving the crystal, the method has several other advantages. For example, in his laboratory experiments in Hokkaido, Japan, Tsuneya Takahashi lets the crystal fall into a cold suspension of two transparent, cold silicone oils. He sets it up so one fluid is denser than ice, one is lighter than ice, so the crystal falls through the light oil and rests on the (transparent) interface between the two fluids.

This method sounds complicated, so why use it? One, as the oils are immiscible with water, they block water molecules from arriving or leaving the crystal surfaces, so the ice crystal shape is preserved precisely for as long as the fluid is below 32 F (0 C). Two, after imaging the crystal, the fluid is warmed above melting such that the crystal melts into a spherical drop from which he can easily measure the volume and thus infer the mass of the original crystal. A third advantage, more difficult to exploit yet sometimes used, is that he can get top and side views of the same crystal. Other researchers in Japan have also used silicone oils to capture ice crystals in the lab, as well as naturally falling crystals, mainly for the first and third reason. They use just one oil type, a lighter oil. Charles Knight at NCAR in Boulder, Colorado had a fourth reason for using a cold fluid: better imaging. That is, one can image greater depth detail because light scattering off the surfaces is greatly reduced, particularly if the fluid is very clear and has an index of refraction close to that of ice. By reducing the scattering, one can see through surfaces to underlying surfaces. He would use gasoline or hexane fluid.

I don’t have a photomicrography setup to take detailed images of snow, and we rarely get snowfall with nice single crystals anyway, but I wondered how well the method might capture falling snowflakes. That is, could I at least see their rough shapes as they fell through the fluid?

Here, we typically get about one light snowfall per winter (2-3"). A relatively large snowfall happened this past weekend, depositing about seven inches. At first, the particles were small, probably highly rimed single crystals or small aggregates. Later, larger snow particles fell, and these particles were clearly snowflakes (i.e., aggregates). In preparation, the previous night I set out two covered wide-mount jars, one with water, the other with Coleman camping fuel (white gas), which has similar properties to gasoline. In the morning, the one with water had frozen, so I knew the other was also below 32 F.



Outdoors, I set the jar on top of my car, put a wooden chopstick in the jar both to focus on (my camera only has autofocus) and to provide a size reference, set a small LED light panel to the side for brighter illumination, and then opened the lid. The flakes fell into the fluid, and fell down to the bottom of the jar. They fell through the fluid slower than they fell through the air, but it was still too fast for me to see how well they were focused. Turned out that they were not very sharply focused, yet one can still see their general shape and fall orientation (below). In general, the flatter the flake, the more it tends to orient broadside to the fall direction.




Obvious improvements would be a better jar, such as one with a flat, smooth glass front, a better camera, and a thicker fluid to slow down the rate of fall.



Such improvements will have to wait at least until next winter.


--JN