If we can have a rainbow, why can't we have a snowbow?
Posted September 25, 2007
MIKE MOSS SAYS: Diane, Raindrops have a size and shape that interacts with light waves in a manner that is favorable for refracting incoming rays, internally reflecting them off the back of the droplet, and then refracting them again on the way out with a concentration of rays traveling along a path that is about a 138 degree angle from the direction in which they arrived (so they appear at about a 42 degree angle from the anti-solar point of any given observer) , with a slight variation in this direction that depends on wavelength and therefore separates the colors into the familiar ROY G BIV pattern. Snow crystals, especially the more complex dendrites or dendritic plates, and especially snow flakes, which are larger agglomerations of multiple snow crystals that are stuck together, are more chaotic and do not tend to refract or internally reflect light in so organized a fashion, so we wouldn't typically see a "snow bow" in the same sense as a rainbow.
That isn't to say that some types of snow crystals can't produce colorful optical phenomena, but these are typically small, less complex ice crystals in the form of fairly simple six-sided plates and/or columns, usually more associated with high altitude cirrus cloud formations than with significant falling snow at the surface. These tiny crystals are responsible for the "rainbow colors" sometimes seen in the form of parhelia (sun dogs), haloes and circumzenithal or circumhorizontal arcs. For lots more on many types of sometimes colorful atmospheric phenomena, including a good visualization of how rainbows form, see