The statement in LaCasce (2004) that the amplitude, A, of the western boundary layer solution (Eq. 25 in the article) is determined by matching at the northern wall to the streamfunction from the northern boundary inner layer solution, ϕin, is incorrect. The correct procedure is to match to the full boundary layer solution at the northern wall, with contributions from both the inner and outer layers. Doing this ensures that the full transport in the northern layer is fed into the western layer...