Radiational Recrystalization is probably my favorite type of near surface faceting. I find it intriguing because of the delicate balance of incoming and outgoing radiation that's needed for this to form. First, the sun penetrates the snow surface without melting it and heats snow UNDER the surface producing a damp layer. This is hard to get a grasp on and seems like Voodoo to me.
The damp layer is now the heat source to produce the faceting. Warm air moving to cold air causes the faceting. (it's actually vapor transfer but for simplicity we'll concentrate on temperature) While the layer is damp there are very strong temperature differences (temperature gradients) between that layer and the adjacent snow. This causes the faceting above the melt layer and below it to a certain extent as well.
As the sun shifts off the slope, the voodoo stops occurring and the melt layer freezes forming a crust which can be very very thin with small grained facets above and below it. During field observations, this is generally where I notice that radiational recrystalization has occurred. To the layman, he notes this as just a thin melt freeze crust formed from the sun and doesn't pick up on the thin layer of snow above it. To the slightly more experienced, it can be mistaken as a thin melt freeze crust with a skiff of new snow or surface hoar on top of it. To a true "Radiational Recrystalization Conspiracy Theorist" who knows there wasn't any new snow since the crust formed, he recognizes very small grained facets above the crust.
This is an example of a snow profile diagram showing the end result.
Conditions need to be just right for this voodoo to happen. The sun needs to be positioned at just the right angle (A) against the slope and temperatures need to be cool enough so the surface doesn't just melt. Slope angle, aspect and the seasonal postition of the sun all play a role. At this time of year (early to mid season) with the low angle of the sun, we can see the steeper slopes get this effect where lower angled slopes during the same time will stay cold powder.
Later in the year when the sun is higher in the sky, lower angle slopes will get this effect and steeper slopes often just melt.
It appears that this layer is not often a major contributor to avalanching here in the central Wasatch because of its spotty distribution and it often gets destroyed by subsequent days of warmer temperatures following the day of its formation. The faceted grains simply melt then freeze into a crust. Wind blowing these grains can also destroy and redistribute them.
However, when conditions are right, this structure can get buried and produce avalanches. When looking at crown faces of avalanches that you suspect may involve radiational recrystalization as a weak layer, you often will have to feel for the crust with your bare hand as it can be very minimal and quite fragile.
For more on Radiational Recrystalization as well as other near surface faceting, check THIS DOCUMENT.