`Valve-effect' at a critical layer
Bretherton and others showed, in the late 1960s, that internal
gravity waves propagating in a shear flow can be suddenly absorbed at a
so-called critical layer, with a consequent change in the mean
horizontal flow there.
My own study of this topic showed that in systems with less symmetry the critical layer will typically act like a `valve', absorbing waves which approach from one side, but letting through those which approach from the other.
Another result of that work, which caused some surprise at the time, was
that a shear flow is not the only way in which critical layer absorption
can occur. If the fluid is electrically-conducting, for instance, a
non-uniform magnetic field can give rise to essentially the same
phenomenon.
J. Fluid Mech. Vol 53, pp 401-415, 1972
J. Fluid Mech. Vol 58, pp 27-37, 1973