On the errors involved in ice-thickness estimates III: error in volume