The authors investigate the dynamics of a narrow tidal ocean channel using velocity observations from a single mooring downstream of the channel and idealised and realistic 2D hydrodynamic model implementations. 

The observations show high frequency asymmetric flow peaks anchored around the timing of peak ebb/out flow. The analysis of idealised and realistic 2D model implementations indicate that the channel is characterised by the regular formation of tidal dipoles as explained by classical dynamics combining flow separation and side-wall friction. The authors dispute earlier suggestions of quasi-stationary momentum-balance in similar settings, and instead explain the asymmetric high frequency velocity signals as arising from the formation, growth, detachment and outflow advection of the tidal eddy-pair.  Both idealised and realistic model simulations reproduce high frequency along-channel velocity fluctuations similar to those recorded by an ADCP. The eddies detached when the low-pressure anomalies associated with the eddies can overcome the Bernoulli flow associated pressure gradient and large scale pressure gradient.

These asymmetric velocity signals appear to have been identified in other similar tidal channels but were adscribed to errors in the measurements rather. Here the authors provide a well thought out experiment to investigate the origin of those high frequency signals and explore the implications for the theoretical approach to studying eddy dipoles in tidal channels.

This is an excellent example of how observations and model studies complement each other when used intelligently.

 

The manuscript is well designed and written. The introduction gives a good summary of previous efforts and evidence for the motivation of the work. The methods are well described with sufficient details. The results are clearly stated and

easy to follow. The discussion is equally well developed with the possible caveats to their approach honestly discussed.

Overall I feel this is an interesting and well executed manuscript worthy of publication here.

My only suggestion refers to the figures. I think all the figures are helpful and required but there are some inconsistencies between figures in terms of style that could be addressed. Figures 3b, 6, 7, 11 and 13 could be drawn with the same style. Include variable name with the units on the y axis and remove title (as appropriate). Increase the font size of labels and axis ticks. Remove shaded background but keep some of the grid lines (enough to be able to compare between figures).

The authors present the results of an observational and numerical study of flow within a constricted tidal channel. Though their realistic and idealized simulations, the authors demonstrate that velocity extrema within the flow can be explained by the formation and propagation of dipolar eddies. I thoroughly enjoyed reading the manuscript: it is very well written with clear methods and results. The authors present the work from a fundamental fluid-dynamics perspective, but with an accessible style and practical purpose. Barring the very minor revisions below, I recommend publication.



1. ADCP is not defined until L57, but the acronym is used earlier (abstract and L18). Please define earlier.



2. L44: Should be `exCited’, not exited.



3. It is referenced to a small degree in the discussion, but I think it would be nice if the authors could add greater discussion surrounding i) the choice of running the model in 2D, and ii) how the results of a 3-D simulation might differ.



4. L87: ``the model performed better’’. Please elaborate on what `better’ means/how it was quantified.



5. Please add missing/appropriate labels (axes, colorbars, and/or titles) to all figures as necessary. Providing only the units on the axis and variable name in the caption is difficult for the reader to take in the meaning of the plot.



5. Fig. 3, left panel: Some of the symbols are a bit hard to make out on the map. Consider using colours that stand out more against the teal?



6. Fig. 3, right panel: Other than the M2 tide, the constituents appear to follow more flat lines than a 1:1 slope. This is, for example, most apparent for the K1 tide. Why are you not able to capture the regional variability in the smaller-amplitude tides like you are with the M2 tide? How significant is this discrepancy for your tidal channel? (It is briefly mentioned starting on L95, but I would appreciate more analysis).



7. Fig 4: It might be nice to add a middle panel here showing what the channel looks like at an intermediate scale. As it stand it is difficult to get a sense for the entire channel from the full domain and zoomed-in domain.

Consider also using mesh lines other than green. Because of the increased resolution near the channel, the domain looks shallower in the left panel.



8. Fig 5: Is the `m below sea level’ relative to a mean SSH, or is it dynamic? (i.e., does the figure take tides into account). Would it make more sense to refer to height above sea floor to have a constant reference point?



9. Fig. 7, 11, 13. Because of the grey background on the axes, the grey shading is not obvious. Please consider removing the grey background, or choosing a different colour for shading.



10. Fig 7 caption: `studies’ -> `studied’



11. L136: remove the hyphen between nine-time



12. Fig. 8: the white star is difficult to see in some of the panels (e.g., A-C) due to the pale yellow. Consider outlining the star, using a different colour, or only plotting it in one panel.



13. Fig. 8: The lower panel with the time series could benefit by being  made larger. It is difficult to see the different points and lines clearly otherwise.



14. L138: remove `above’ when referring to the snapshots (it is ambiguous if you mean above the text or above the lower panel, but is unnecessary)



15. L140: remove `(+15 mins)’ as this is already clear.



16. L153: Consider replacing `So a main take home…’ with ‘A key…’ (or similarly concise phrase)



17. The jet in Fig. 9 appears to be angled toward the right. Why? Is it simply because of the Coriolis force, or are there other factors involved?



18. Figure 10, with its clear message and simplicity is beautiful!



19. Fig 11 and elsewhere: m/s2 are units of acceleration, not a force. Please clarify terminology.



20. Fig 11 legend: along-SHORE pressure gradient



21. I think there is a word missing from L240: `on southward and inward what?’ Flow?



22. L247: `…fluctuates between positive and negative values in quite a noisier manner…’ -> ``…fluctuates noisily between positive and negative values…’



23. L248: `So, despite the more noisy situation here…’ -> `So, despite the increased noise…’ ->



24. Fig 13 (caption): perhaps referring back to the idealized equivalent in F11 would help the reader interpret the plot?



25. L253: remove hyphen in `in-viscid’



26. L320: `changes’ -> `changed’



27. L348: fix backwards apostrophe on quote of `propagating constrictions’



28. L348/349: Remove `pretend to’ from `The current study does not pretend to offer a complete description of this complex flow dynamics.’ While I appreciate you acknowledging the short comings, somehow that phrase seems to lessen the great work you have done!



28. L352 on: I appreciate this nice, practical finish to the manuscript.

29. There is discussion around the alternating rotation of the eddies as well as the role mixing plays within the channels. Are the eddies large enough to lead to local differences in vertical mixing on either side of the channel (because of their alternate rotation), or would that be small compared to background mixing?



30. L375: `sat’ -> ‘set’