25

I just don't get it. Isn't snow just another form of water? Also are all ices transparent or do they go white after a certain temperature?

Community
  • 1
Huzo
  • 1,013

3 Answers3

38

That's same for cloud, fog, wave splash and so on. Because of tiny size (but in large number) and irregular appearance, reflection and refraction occur in an irregular manner so the light is scattered randomly so that every images and colors mixed together. As a result, it looks white. By the way, under thick clouds, they look dark because all the light from the top has scattered away. Under microscope, snow flakes (ice crystal aggregate in fractal way) look crystal clear.

enter image description here

Photo credit goes to Henry David Thoreau:

  1. snowcrystals.com
  2. photos' link
  3. Method of illumination is mentioned at the end of this link.
Ng Chung Tak
  • 1,377
  • 2
    Given this, could a group of snowflakes cause enough create a rainbow? :D – Anoplexian Feb 09 '17 at 15:55
  • 19
    @Anoplexian No, rainbows rely on the relative uniformity of the liquid drops to scatter light consistently across the curcular arcs of relative angle to the sun and viewer. But small ice crystals can create a similar effect. – MooseBoys Feb 09 '17 at 16:04
  • 6
    It should be noted that, because they scatter light, they only look white when most of the light hitting them is white. If the sky were red the snow would look red as well. – Dave Feb 09 '17 at 17:47
  • 4
    @Dave This property is true of most things that we attribute a color to. A white wall will also only appear white if the light hitting the wall is mostly white. It is also not the color of the sky that determines the color of the clouds. The clear sky is blue (due to Rayleigh scattering), but the clouds are still white because the direct sun light is white. – jkej Feb 09 '17 at 19:00
  • @Anoplexian: Not a rainbow, but you can get conditions (large crystaline flakes on top of snow cover) that produces lots of diamond-like reflections/refractions in the sunlight. – jamesqf Feb 09 '17 at 19:08
  • @jkej Good point. When we speak about what color something is, we typically mean "what color does it appear when illuminated with white light?" In this case, the snow appears white. – Dave Feb 09 '17 at 20:10
  • 1
    @Anoplexian I suspect the slight rainbowish hues you can see in these photos is owing to the wavelength dependence of the Fresnel equations describing the angled facets of the crystals, and the fact that different facets present different angles. There is probably some Fabry-Perot etalon effects and maybe even stress birefringence effects (the lighting for these micrographs is probably polarized). – Selene Routley Feb 09 '17 at 23:35
  • Dear all: Photo credit is updated in my answer, please feel free to see the link for "method of illumination" at the bottom. Many thanks to all of your responses. – Ng Chung Tak Feb 10 '17 at 12:19
17

Snow is simply a random collection of snowflakes and bits of irregularly shapen bits of ice. Each of these is clear, but a small fraction of the light incident on each clear entity is reflected and scattered. The fraction is of the order of 2%.

Light passing through snow encounters huge numbers of these randomly oriented interfaces. So a large fraction of it is scattered. The wavefronts of the incident waves are completely randomized by this scattering, so the scattering can no longer form images. The same goes for any light that passes through the snow: its wavefront is highly randomized and cannot form images.

The scattering happens roughly equally to all visible wavelengths, which is why we see white, diffuse light from snow when it is lit by a broadband source like the Sun.

Selene Routley
  • 88,112
4

It isn't white. You perceive it to be white because it is clear and it is reflecting the full spectrum of light applied to it.

As for why snow appears white and, say, raindrops do not, the answer lies in the crystalline and faceted nature of snow. Consider a clear ice cube which, when crushed, appears white.

Jake Watrous
  • 473
  • 3
    Isn't anything that reflects the full spectrum considered as white? – Huzo Feb 09 '17 at 14:05
  • When it is reflecting the full spectrum, it can be. If you shine a red (or any other color) light at it, no. – Jake Watrous Feb 09 '17 at 14:06
  • 12
    Isn't that what "white" means? In other words, are there any white things whose whiteness is not for exactly this reason? – hmakholm left over Monica Feb 09 '17 at 16:28
  • 4
    @HenningMakholm you could have something that just emits (through reflection, or as a direct source) just red, green and blue, in order to stimulate the rods in your eye to make you believe you are seeing white. In fact, you are looking at a device that does that, right now – Baldrickk Feb 09 '17 at 17:02
  • 1
    @Baldrickk: Sure, but I'm not sure I would say that device "is white" itself. – hmakholm left over Monica Feb 09 '17 at 17:15
  • @HenningMakholm true. Though 'white' often isn't; the light spectra isn't necessarilly uniform. See https://cdn.comsol.com/wordpress/2016/01/Combined-plot-of-emission-spectra.png from https://www.comsol.com/blogs/calculating-the-emission-spectra-from-common-light-sources/ for an example. Our brains are quite good at doing "white balance" like cameras do. Try wearing rose tinted glasses for a bit, and then noticing how everything looke blue when you take them off. Usually, when we talk about colours, we are usually referring to how we observe them. (looking at the full spectra is also valid) – Baldrickk Feb 09 '17 at 17:20
  • 2
    @Baldrickk: I don't think that is relevant for when a thing is white. A white object is one that reflects the ambient illumination at whatever spectrum that ambient illumination has. – hmakholm left over Monica Feb 09 '17 at 17:24
  • @HenningMakholm Yes indeed. I was noting that something that something that just reflects Red, Green and Blue is indistinguishable (to us at least) from something that is "true" white, as it will behave the same under the same light source. In terms of "white" being uniform reflectiveness/emittance, then it doesn't meet the criteria, in terms of "white" being indistinguishable from the above, then no. Depending on the definition of "white" chosen, the answer can be either. If you just want the colour as we see it, then the answer to your question is "yes" (RGB sources) – Baldrickk Feb 09 '17 at 17:30
  • @Baldrickk: At best that would produce a grey object that appears white when viewed under a particular jagged-spectrum illumination. (And under a light with peaks elsewhere in the spectrum it might appear to be black). – hmakholm left over Monica Feb 09 '17 at 17:41
  • @Baldrikk Your point is sort of technically valid, but also kind of irrelevant. An object that reflects all wavelengths roughly equally (and relatively highly) is the type of thing that is typically called white. The type of object that you are describing (only reflecting specific bands of RGB) would be highly exotic and human made precisely for this purpose. Should this object be described as white? Maybe. It will appear white in ordinary white light, but I could construct other "white" light sources in which your object would not appear white. – jkej Feb 09 '17 at 19:23