I don't think the atmosphere is as "local" as you imply (diffusion of the water molecules should spread them out of state far faster than they can condense in-state).