General Comments:

The paper presents an investigation into the Quasi Stationary Wave Strength and its dependence on surface forcing and internal dynamical factors. The investigation is performed using eight 100 year simulations from the CAM6 model. 

I believe the subject is of interest to the community and the set of simulations has been systematically/logically designed 

to probe the relationship between different surface forcings and Quasi Stationary Waves.  

I found the paper quite difficult to read and spent a lot of time flicking back and forth between various sections to work out 

what was plotted in the various figures and why, so feel it would benefit from a careful revision to the text to make it more explicit in terms of analysis and interpretation and some restructuring to make the main scientific questions/hypotheses and methods clear. It is possible the paper is attempting to cover too much material so there are many multi-panel figures, which require a lot of explanation. Reducing the scope of the paper would in my view make it much clearer and stronger.    

The main body of the paper is devoted to testing two hypothesis: 

1 The spatial distribution of quasi-stationary waves is governed by that of stationary waves.

2 The spatial distribution of quasi-stationary waves is governed by that of dynamic factors such as Eady growth rate and local stationary wavenumber and transient wave strength. These hypotheses appear in the results section, but I think they should really appear in the introduction motivated by/linked to the discussion of literature. The methods section can then more clearly explain how the methods are designed to address the hypotheses  

The partition between stationary, quasi-stationary and transient waves defined by the authors is as follows:

- Stationary waves: A day of year climatological mean.

- Quasi-stationary waves: 15 day low pass filtered anomalies the climatology 

- Transient waves: 15 day high pass filtered anomalies to the climatology 

All three of these are converted to wave-envelope amplitude (as in Zimin et al 2003) including only wavenumbers 4-15.

A major methodology employed in the paper is comparison of climatological mean spatial patterns of quasi-stationary wave amplitude to that of stationary waves and transient waves as well as to climatological mean spatial patterns of Eady growth rate and stationary wavenumber. So a number of the conclusions hang on the assumption that a similar climatological mean pattern implies a significant relationship between quantities/lack of a similar pattern implies no relationship. I think this method is employed in all figures except fig 1and fig 5 which makes use of time variation in the data. The robustness of this basic assumption requires some discussion in the text - it isn't clear from what is written why this is a useful method and what its possible limitations are. The authors could also consider including more results which more carefully examine the relationship between stationary wavenumber and Quasi-stationary wave strength by utilising information about time covariation of these two quantities - this to me would add weight to the argument that the two a closely linked. 

The main conclusion of the paper is that stationary wavenumber has a clear association with QSW strength suggesting a simple barotropic interpretation. This is an interesting conclusion, but the more clarity is needed in the explanations of results to justify this.

Specific Comments:

- There is a long introduction which covers well the literature in the area. 

This ends with a short paragraph about the scientific question/hypothesis the authors' work seeks to address. 

The hypothesis in this final paragraph is too vague and doesn't really explain what the authors aims are and how these fit

 into the literature described prior to this. 

- Two hypotheses are given in section 3.3. The bulk of the paper seems to be aimed at testing these, so perhaps it would make sense to 

include these in the introduction and motivate them from the literature - why are these important and novel hypotheses to test. Then to include some discussion of the methods used to test these hypotheses in the methods section. 

- section 3.1 line 170 What does "frequency of all events is fixed" mean - explain this more clearly. Do you mean you are defining an amplitude threshold for an "event" such that the total number of events is the same between the different simulations and reanalysis?

- Main conclusions of Figure 1 seem to be that the model is biased towards shorter duration events, including stronger forcing increases the duration of events. You could consider whether these points could be demonstrated with a simpler figure with fewer panels. It would also be good to comment on whether these conclusions would be expected from what we know about climate models/other modelling studies or are specific to this particular model. 

- section 3.2 you state that amplitude of meridional wind is used as proxy for stationary wave amplitude. This implies

 you are calculating the zonal mean of magnitude of the time mean meridional wind, not the time-mean zonal-mean of the magnitude of the

meridional wind which would include transient waves. However you don't state clearly in the text what is calculated. 

- line 257: "As a first test, we calculate Spearman rank correlation coefficients". This appears to be the main test that is used in

 the paper to link QSWs to the other quantities. 

- Figure 3. Panels a and b are very clear, but it would be good to state the contour interval where they are described in the caption rather than at the end.

 

- The effects of different types of surface on QSWs are investigated by examining differences in climatological means between different pairs of simulations, rather than just looking at anomalies from the control simulation. This makes the results quite difficult to follow, my question is whether this complication is necessary or anomalies from the control could be used instead or whether the anomalies are necessary at all - i.e. could you make the same points just by looking at the fields themselves and comparing? This might simplify the analysis. If it is necessary to look at these anomalies as in the current paper to make the points, then it would be good to give more explanation to help the reader understand what they are learning from the different pairings. 

 

- A significant portion of the paper addresses the link between stationary wavenumber and quasi-stationary wave amplitude. This to me was the more interesting part (particularly fig 5), however the method for calculating stationary wave number is missing some key details: the authors state it is calculated from zonal wind, but don't specify whether this is a single level a vertical integral. In figures 3c and 3d, I can not determine what the contours are showing. At first I assumed from the caption/title that these were the stationary wavenumber with only values 4 to 8 shown, however the contours range from 2 to 10 so is this just stationary wavenumber?, if so why have 4-8 in the title - please provide a clear explanation.  

- Figure 5 is very interesting and appears to show a clear link between the Ks and the QSW amplitude. However it is not clear from the caption of text what is actually plotted here. The count refers to the number of days in DJF which is clear and one assumes this sums up to 90xlength of simulation in years. But how do the authors decide what the QSW strength and stationary wavenumber are for a given day - is this a spatial mean value?