Comment on “River loads underestimated by rating curves” by R. I. Ferguson