Addendum prior to final acceptance in ACP:

Some improvements and clarifications have been made in the revised version, in response to comments of reviewers, e.g., concerning methodology, phase relationships, and power spectra.

Although the paper is acceptable in its present form, my opinion is that the Abstract and opening paragraphs of the Conclusion merit some modification in order to reflect the significance of the main findings of the paper more fully and accurately. Therefore I am providing the authors with a final opportunity to improve the presentation of major findings in these two places of the manuscript.

It is important to take advantage of this opportunity because the field campaign is a significant investment of research dollars, and for this campaign in particular, gravity waves provide a focal point for study as part of the broader objective regarding Ex-UTLS transport processes.

Here is an outline of my concerns and how the summary of findings can be modified:

1. The existing summary is quick to highlight measurement issues, and projects an overly negative tone. More emphasis is needed on positive findings.

2. Among these positive findings is evidence of 50-500 km mesoscale fluctuations with the expected signature of vertically propagating gravity waves.

3. Not all such fluctuations are consistent with a monochromatic wave, but admit other interpretations. For example, quadrature of u' and w' may indicate rolling turbulent motions in the vertical plane, such as inertial instability, or simply a superposition of upward and downward waves owing to reflection.

4. Perhaps it is my misunderstanding, but the paper seems to confuse measurement issues with alternative (possibly true) explanations of deviations from gravity-wave behavior. By all means, take care to disentangle the measurement issues from these alternative explanations.

5. From prior conversion with the lead author several years ago, I understand his concern about false signals owing to vertical fluctuations of aircraft motion. Unless I missed it, the paper doesn't make completely clear whether the pressure measurements alone are vulnerable in this respect, or if the problem spills over to the velocity component measurements. If necessary, please clarify.

6. Regarding power spectra, the revision has attempted to highlight the -3 power law in addition to the -5/3 power law. A general summary of power spectra is as follows: (i) horizontal velocity components at the larger mesoscales display the approximate -5/3 power law in agreement with GASP & MOZAIC; (ii) vertical velocity is flat in this range; (iii) approaching smaller mesoscales the fluctuations roll over to a -3 power law, except (iv) when this part of the spectrum is activated, as recorded beautifully during M2. The following clip is noteworthy:

"Despite the overall resemblance among the flight segments of RF02, there are some unique characteristics in the power spectral distributions for individual segments. For segments M1 and M2, for example, (i.e., the fourth column versus the fifth column in Fig. 4), the slopes of u and v during segment M1 are approximately consistent with a -3 power law for the scale of ~0.5-~8 km, while those during segment M2 follows a -5/3 power law instead. This is probably associated with the fact that segment M2 successfully captures a rapid decrease in u (from ~65 m/s to ~40 m/s) while segment M1 has no such a dramatic reduction in u (the fourth column in Fig. 3a versus the fifth column in Fig. 3a). Note that the aircraft during segment M1 flew away from the jet core region, as the jet was still moving eastward to the downhill side of the topography. In contrast, the aircraft during segment M2 flew directly toward the approaching jet core at a lower flight level than segment M1 (the fourth column in Fig. 3d versus the fifth column in Fig. 3d), and the observed decline of u (i.e., a potential jet exit region) is located roughly on the downhill side of the topography (the fifth column in Fig. 3d). This suggests that the spectral slopes for the aircraft measurements can, in fact, be extremely sensitive to changes in the background flow, even though sampling takes place in the same area only a few hours apart."

This example of activated spectrum is lost when the average is taken over all flight segments. Interestingly, the M1 segment immediately prior to M2 did not record the event. By the way, it should be noted (and stated explicitly, if not already) that the wavelet analysis doesn't reach into this spectral range of elevated power. Either the orographic wave in question did not penetrate to the flight level of M1, or it was excited suddenly as the jet approached from the west. By any chance, however slight, could the model simulation resolve this ambiguity?

7. Regarding the -5/3 slope at larger mesoscales, clearly it cannot be attributed to isotropic 3D turbulence. On the other hand, the activated spectrum at smaller mesoscales during might fit this interpretation.

In summary, I encourage the authors to address these concerns, mostly by way of additional summary inserted in the Abstract and Conclusions. A final thought is:

8. Since we anticipate several follow-on studies of these unique measurements, a brief list at the end of the Conclusion might highlight a few key issues and how they might be addressed.

--Tim Dunkerton
(tim@nwra.com)