The authors use shipboard acoustic and microstructure turbulence observations to study a dense overflow in the Baltic Sea. Results show how topographic lee waves propagate upwards into the halocline where they break and cause high levels of turbulent mixing.

This is a very nice paper and I don't have any major comments. There are a few minor points that would be good to be addressed before the paper is published. I trust that the authors will include my comments where they feel that changes are necessary and do not need to review a revised version before publication.

Minor comments:

14-15: This sentence reads maybe a bit generic?

15: overflow -> overflows

19: Possibly split sentence where you start talking about when data were collected.

20: Year incomplete.

20: I had the impression that the fast timescale of the flow reversals was not previously known. If so it would be worthwhile to highlight the timescales here.

23: You only speculate about submesoscale processes in this paper, this shouldn't be stated here as a result.

23: leading -> led

24: Mixing -> Turbulent mixing? What specific quantity are you referring to here?

29: The last sentence of the abstract reads a it generic, consider removing.

33: You don't talk about your results (breaking lee waves cause turbulent mixing in the vicinity of the sill here). It remains a bit unclear why you refer to numerical models if you don't mention parameterizations. As stated above, I would move this into the discussion section and focus on the results of your study in this summary.

43: Do you have a reference for the Bothnian Sea being well oxygenated?

54: lead -> led

55: lead -> led

62: "despite the fact that" or "despite anthropogenic, land-based sources having declined significantly"

64: over the last decades

68-70: Consider moving this sentence to the end of the next paragraph or removing it as turbulent mixing and associated water mass transformation have not been discussed at this point.

77: remove comma after Bothnian Sea

77: parentheses missing around 2007

86: Consider removing "Here," or replacing it with "In this paper,".

87: velocities -> velocity

87: You could be more precise and replace "mixing parameters" with "microstructure shear and temperature".

87: Remove one "and"

95: Panel d is too small.

97: I suggest adding a reference for the red stars and removing the reference for selecting the area of your overview map.

196: effect of mixing -> effect of turbulent mixing

197: mixing data -> turbulence observations

198 mixing -> turbulent mixing

211: Change to published version.

226: Possibly mention here that the "internal shear sensor" (I am still a bit puzzled by the name) is used to determine noise due to instrument vibration?

247: You mention a quality control algorithm but the data in Figs. 2a & 2b appear to show some noise. Maybe the quality control was only applied to the moored ADCP data?

266: towards the north?

270: Maybe it would be better to phrase this as "data collected along the transect are expected to show spatial and temporal variations" since you don't resolve this with one transect?

271: This sentence seems to terminate early? I can't quite make sense of it.

277: How do you determine that some of the water is blocked by the sill? Looking at your (impressive!) acoustic observations in Fig. 4, I get the impression that most of the deep water is being drawn up to the sill height?

284: What is your definition of deep water here? Denser than a specific isopycnal? Some depth level?

308: Black line -> The black line

324: Do you mean tilting density interfaces indicate that deep water is lifted up to the sill height? "due to" implies a physical process. Also, do eroding density interfaces imply that there is near-bottom turbulent mixing before the water reaches the sill crest? Could this explain the loss of oxygen and temperature minima you show in Fig. 2 without any blocking necessary? 

356: 3degC or 3K

358: 6degC or 6K

380: The map inset is too small.

365: Reiterating my points above, how do you determine that this is due to blocking and not mixing and entrainment?

391: I could not figure out how to retrieve the videos. In general I would like to suggest to treat the videos as what they are - supplementary material - and to include a figure that summarizes what you show in the video. Lines 399 to 407 could refer to one figure with several panels showing the acoustic observations at the times discussed in the text?

423: shown below -> discussed below

449: Individual profiles in this figure are shown with such faint lines that are impossible to read. Consider increasing their thickness or not showing them at all.

464: You are still in section 4.2 and the jump into the discussion with this paragraph seems rather abrupt here. Why not reserve section 4 for your results and move the discussion into the next section?

481: It would be good to state here why higher oxygen near the sill is a consequence of mixing - I am guessing it's because oxygen-rich waters from the north are entrained?

495: This sentence reads very generic, consider removing or being a bit more specific. This and the following paragraph may benefit from being moved into a discussion section.

505: change etc to something more specific, or at least "and other tracers".

510: make -> makes

510: "established in-situ methods for observing dissipation rates" - you probably mean microstructure turbulence observations? If so good to spell this out.

511: "prone to fail" seems a bit harsh - you show with this paper that there is a lot to be learned even when undersampling in space and time.

515: Maybe it's just me but 10^{-0} just looks a bit odd...

516: in a reversing

522: Your reference profiles south of the sill in Fig. 8d show that there are regions with relatively small turbulent mixing! Doesn't this contradict your statement of continuous modification?

525: Wake eddies have not been discussed prior to this. It may be helpful to summarize your GRL paper in one or two sentences when you refer to it for the first time.

533: Consider starting a new paragraph at "The observed mixing".

In the paper by Muchowski et al. new observations around a sill in the Southern Quark region (Baltic Sea), i.e. the area connecting the Northern Baltic Proper with the Bothnian Sea, are presented. The new dataset is massive and comprehensive as it includes velocity and hydrographic data but also microstructure measurements as well as high-res acoustic observations of turbulent mixing. Results show that turbulent diffusivities, dissipation and vertical flux rates are very large and about 3-4 orders of magnitude bigger near the sill with respect to reference unperturbed stations. Such a strong mixing is thought to result from hydraulic jumps and stationary lee waves and shown to affect also oxygen values, impacting the ventilation and residence times of the deep layers in the region.

The paper is well written and organized and fits well the scope of the journal. I have only a major concern related to the large diffusivity values shown in Figure 8 which are reported to reach 10^{⁻1}-1 m2/sec in the deeper layers and even be larger than 1 at about 160-m of depth. I urge the authors to discuss these large values and compare with those observed in other areas. Can this be related to the choice of a constant mixing efficiency?

A process of revisions is suggested to address also the following minor concerns:

1. L24-25: the large values of mixing should be reported in the abstract
2. Fig1: please add a rectangle in panel a to show the close-up of panel b like it is done in b with the yellow rectangle for panel c. Many dots in panel c are barely visible also for the color choices (green and turquoise over a blue or light-blue bathymetry). I suggest to have panel c and its inset (panel d) way larger and below panels a and b
3. L135-141: Not sure these lines are relevant. Why is the bedrock geology important for this study?
4. L151-152: very true but also consider that ocean models need to have enough horizontal resolution to fully resolve the bathymetric features present in the DBM dataset and that it is usual practice in some ocean models (e.g. sigma models) to smooth bathymetry
5. L161: what kind of mixing values (i.e. diffusivity values) are expected to sustain such an estuarine-type circulation? How do they compare with those observed in this study?
6. L264-270: really confused mainly by the text here. It looks to me that at the beginning the ship ADCP data show northward velocities and not southward as written in the text while after 1.5 hours velocities are back towards the north and not the south. Please check, clarify and rephrase
7. L271-275: confused again as ADCP data are showing northward velocities also below the halocline but above 140m. Is this consistent with an estuarine-type circulation? If yes please explain better
8. Figure 2, panels c-f: please explain how is interpolation performed
9. L325: why not saying simply white arrow instead of bright grey? Initially I got lost try to find a different grey arrow
10. L334: how can you be sure to say that below 100 m the echoes are due to eddies or large overturns?
11. L339-340: what does “where there is little temperature and salinity microstructure” mean? Do you mean gradient?
12. L374: the dark blue dots in Fig.5f are barely visible
13. Figure 5: what about the two relative maximum temperature values for the black line across the 50-m depth? Intrusion of intermediate waters?
14. L409-425: the first and last lines of this paragraph seem to me to contradict each other as currents are indeed stronger and more persistent during EL19 and thus not comparable during both cruises. Or am I not grasping something here?
15. Figure 6: It looks there is an argument here that the wind sets up a ssh difference responsible for a barotropic signal. Wouldn't be possible to filter out the pressure-based signal to show the residual (estuarine-type?) circulation? Will a EOF-based approach work?
16. Figure 7: just to point out that the velocities are indeed weaker but also less barotropic
17. Figure 8: impressed by the large numbers here. Why do kz values increase below 200 m for the unperturbed profiles? Why are the yellow lines useful or, in other words, what is the point of reporting values for the north-western part? For comparison?
18. L515: please report here (and also in the abstract) also the diffusivity values