Yesterday a comment was made of the old rule that if the dewpoint+air temperature is zero or below, you're likely to see snow falling. This works if there are no milder layers aloft and is an approximation of several much more complicated formulae.
There's actually been a *lot* of research over the years on snowfall versus rain or sleet and several laws of physics are involved, as well as an exhaustive amount of fieldwork. This leads to the formulae as discussed here:
http://www.sciencebits.com/SnowAboveFreezing
I've plotted these values below. It might be worth saving a copy of this as a "ready-reckoner" when snow is heading your way and conditions are looking marginal! NB, these aren't be-all and end-all figures as every snowfall is different. However, they're a notch above the old "air and dewpoint" method.