On the proper motion of auroral arcs