From what I can observe, it doesn’t. It first snowed this winter a couple weeks ago, and has snowed several times since without a complete melting. Much of the snow has melted in between snowfalls, but some of it has stayed on the ground throughout.
The first snow to melt has been under trees and along fences. The snow that never melted is in the center of wide-open fields. No doubt that is mostly due to the fact that the fences and trees caught some of the snow on the way down, so the wide open fields ended up with a higher accumulation.
Despite the lack of an obvious causal relationship, I just wanted to point out that your hypothesis is observably false.
I read this post yesterday, and spent the entire commute last night and this morning (about 1:45:00 total) carefully observing snow remaining from last weeks snowfall. Yes, you’re correct that the snow directly under trees, and directly next to fences had less original accumulation. Therefore, at these “boundary conditions”, if you will, there was no snow.
However, the point I was making was that radiation due to the sun has a much greater effect than sublimation in removing snow. This can be easily observed (and I wish I had my camera, perhaps I’ll borrow one), by noticing that there are two main places where the snow hasn’t yet melted. The first is on the north slope of hills, because the radiation from the sun doesn’t shine there as often. I figured this one was a given. The second place, are small sections of snow, about as big as a cross section of the tree viewed from the side, and about as far away from the tree trunk as its circumfrence, as viewed from above. This snow is still there because the radiation from the sun doesn’t shine there as often.
So, my hypothesis was not “observably false”.