*REVIEW OF Rhode: Identifying orographic gravity waves... (ACP 4946)*

This paper demonstrates a method to identify orographic gravity waves from three-dimensional data with backward ray tracing. The dataset is IFS simulation prepared as if a satellite would observe it, called by the author "satellite simulation". Gravity waves are identified with he S3D method and rays are started at 35 km altitude. This method is new and could also be used in general to identify orographic gravity waves. However, the data cover the 2018/2029 "New Year SSW" which deserves some more physical interpretation, as specified in the major comments. That is why I suggest a minor revision.

** Major comments

1) Propagation: Once a mountain wave is excited at the surface, the further propagation is determined by the wind field. This issue deserves some discussion and presentation with typical examples. The key process is likely the refraction into the stratospheric jet, and this should be identified in the data. Further, there are comparable field campaigns which are documenting such oblique propagation in detail - and these should be discussed.

2) Stratwarms: The wind field is changing drastically during sudden stratospheric warmings, which concerns the phases before, at, and after the central date. The specific spatial structure of the wind should be documented and used to argue for the local gravity wave appearance above Mongolia respectively Atlantic/Canadian.

** Technical comments

L10: You write that orographic gravity waves may "dominate episodically, including prior to the onset of SSW" - but you do not show it. In line 132 you write of "40 % on individual days" but this is far from dominance. Please, reformulate this passage to the features you are documenting with the analysis.

L69: You write the "spectra are smoothed" - so, you did not execute this filtering in space? Further, the cutoff at zonal wavenumber 7 (~2900 km at 60 °N) and 10° in latitude (~1100 km) is not consistent. At least, this is not orientation preserving. Some arguments for this procedure are given in Mathew et al. (2025), which I after a while found at

  https://egusphere.copernicus.org/preprints/2025/egusphere-2025-4602/

(a DOI in the reference list would have been helpful). These are more technical reasons like strong jumps and applicability of smoothers. Please, adjust the text accordingly.

L119: I suggest to write "before and past" and leave "way" out.

L126: Do you mean "time series" with "timeline"?

L131: I do not see that the strongest GW events are "mostly driven by orographic GWMF" - please, reformulate.

Fig. 2: Please, specify what the contours are. In view of the major comments, I suggest to indicate the horizonal wind speed contours in order to show the refraction effect.

Fig. 3: Do you mean "time series" with "timeline"? Perhaps, indication of the central date with an arrow would help identification of the special situation. So, what you show here is a polar cap averaged of absolute momentum flux, right? Is it possible to say something on the sign which could change when the zonal wind turns easterly?

L175: "it show" --> "it shows"

Fig. 4: Also here, overplots of wind speed could help interpretation. From which time are these plots, or are they averages?

L181: This sentence is a bit confusing, I guess you mean "simulated observations of a space-based infrared imager", or?

Fig. 5: May be, another horizontal line in the left plot for 0.1 fraction would visualize the 90th percentile

L183: With reference to L108ff, I see three instead of two criteria.

Fig. A1: Please, specify "orographic GWMF" as you did for the other figures.

L254: May be, "limit case" is better to read than "edge case".

L264: The mentioning of the "horizontal propagation layer" is interesting - is it the tropospheric jet or the lower edge of he stratospheric jet? A further documentation of wind profiles could make this point clearer and worth to be placed in the main text.

This paper presents a method to isolate the contribution of mountain waves from gravity-wave observations that often contain a mix of a broad spectrum of gravity waves. The method involves the use of infrared limb measurements with backward ray tracing and is evaluated using simulated infrared limb observations during the 2018/2019 sudden stratospheric warming (SSW). The results yield gravity-wave momentum flux variations suggesting that the isolated waves do originate near well-known mountain wave source regions, which the author use to argue for the robustness of the method. Overall, I agree that the concept of using limb measurements together with ray tracing to isolate mountain-wave contributions is worthy of investigation. The results are encouraging, particularly in their apparent correspondence with established mountain wave source regions. My primary concern is the test scenario. The 2018/2019 SSW represents a dynamically complex period. While this makes it an interesting case study, it also raises the question of whether it is the most suitable initial scenario for validating the method. Demonstrating the method first under more quiescent, more dynamically simple, non-SSW conditions could help establish a clearer baseline for the method’s performance. I specifically suggest repeating the analysis during periods outside of an SSW event when it is well known that these same mountain wave sources are known to clearly generate significant mountain waves. Together with the results during an SSW event that is already included in the paper, this would add more credibility to the method. On the other hand, if the author really believes that an SSW provides the best test for the method, this rationale should better explained.