The authors present results on the adjustment to unbalanced initial conditions

on a beta plane set at mid-latitudes. The approach is elegant and the presentation

is effective at communicating the main results. My main point I would like to see

addressed is the addition of a discussion early in the paper on the application of

a channel for mid-latitude flows. There are no solid boundaries at mid-latitudes

in the Atlantic or Pacific Oceans. For wide channels, the solutions trapped wave

solutions depend on having a wall to support the waves. How would these waves

be supported in the real ocean? Is choice of the zero crossing sufficient? Is there 

a zonal pressure gradient at y=-L/2?

Detailed comments:

line 20: understanding

39: A brief statement of how the waves are trapped would be helpful.

Table 1: define what runs A-E are, either in the table or the caption.

132: What is lost when you neglect -2by and in what limit is this valid?

You calculate solutions for by=O(1).

146: This is a confusing way to state that the Bi term must be zero in

order for the solutions to be bounded.

Eqn (24) Can you provide some physical interpretation of this condition and

the role of b and E?

179: Mention the continuity equation (27) in going from (28) to (29).

183: Potential vorticity must also be conserved.  What is the physical

meaning of conservation of vorticity gradient?

198: missing =0 on the du/dt term.

205: Please provide justification for the d(eta)/dy =0 boundary condition.

210: If the solution is symmetric about y=0, don't all solutions satisfy

mass conservation?

Figure 4: So the variance is not well described by these modes since it

takes so many.

323: The tilting of wave crests in many previous studies has been found in the

absence of channel walls so it is most likely, at least in those cases, to be

due to the variation in the long wave phase speed with latitude. Linear Rossby

wave theory has also been applied to forced variability in closed basins with

close comparison to PE numerical solutions. The current discussion is too 

dismissive of the role of linear long Rossby waves at mid-latitudes.

345: Refer to Table 1 for model resolution.

Summary and Discussion: This section is very helpful and answered several

questions that I had while reading the paper.

437: state again what Z* and zeta are.

Add a subsection 6.7 on the assumption of walls at +- Ly/2, which do

not exist in the ocean or atmosphere.

This paper concerns geostrophic adjustment in the presence of a gradient of potential vorticity, specifically a beta plane. The standard Rossby adjustment problem is a staple of geophysical fluid dynamics and appears in many text books, but the extension to a beta plane has been studied much less extensively.  On the one hand this is surprising, since dynamics on the beta plane are also a staple of geophysical fluid dynamics.  However, the analytic/algrebraic difficulties are not trivial and often that entails a numerical approach.  Oddly enough, the equatorial beta-plane and the sphere have been studied more extensively than the mid-latitude beta plane, so the present submission fills a gap in the literature.  I have only a couple of comments. 

The specific domain used here is the zonal channel on the beta plane. This is arguably slightly restrictive but nevertheless useful. The presentation is generally clear.  What I think would be useful would be a explicit  demonstration of how the numerical results approach the standard and well-known f-plane results, as \beta tends to zero. both for the time evolution and the final state. This would both provide confidence in the numerical scheme and provide insight into the effects of beta. Some f-plane results are provided, and but the physical patterns arenot shown except in figure 10 (which is showing something different). 

Relatedly, potential vorticity is presumably conserved, and conserved pointwise in the linear problem. This is the basis for how the final state is calculated in the original Rossby problem.  Some discussion of how PV enters into this problem and provides constraints on the evolution and final state would be welcome.  Or if potential vorticity is not conserved or is somehow not relevant, some discussion of that would be useful.