The reliability of rating curve estimates of suspended sediment yield: some further comments