Obviously, water can't freeze until it meets a temperature of 0 degrees C.
Can this occur in the engine intakes, or plenums, for example? If so, you might encounter an engine icing situation although the airframe remains ice free.
Water can exist in a liquid state below that temperature, if there is nothing for droplets to freeze onto or around (supercooling). If the airframe, or part thereof, is below zero and it meets supercooled water droplets - you then almost certainly have icing conditions.
One other consideration - how accurate is the OAT guage on a particular aircraft?
In my outfit, if any build up is noticed (the wiper blades are an excellent indicator) THAT is icing conditions. However, as always, the Flight Manual is the paramount authority.