It quickly turns into one of those almost but not quite fractal problems. By the time you get down to the scale of individual grains of sand, you are having to

It quickly turns into one of those almost but not quite fractal problems. By the time you get down to the scale of individual grains of sand, you are having to make some tricky calls. If this grain of sand is sitting on top of this other one, do we count the full surface area of the top grain minus the point of contact, or do we declare that the grain is not actually part of Colorado but merely a grain sitting on top of Colorado? Whoops, the grain moved while you were thinking about it. Then once you've classified every single grain as in or out, you get down to the subatomic scale. Perhaps you could define the surface in terms of a single electronic density isosurface, but you'd be choosing your isosurface somewhat arbitrarily.