Error propagation in the Abel hill inverse problem