This work designs a series of numerical experiments using FVCOM model to investigate the drivers of Qingdao cold water mass. The study finds that both tidal forcing and winds are important in the formation of cold water. However, the manuscript is not well organized and include the information not directly related to this work. The objectives of this study are not very clear. Figure 12 is not even mentioned in the manuscript. In Conclusions, the authors confused “anti-clockwise” with “anti-cyclonic”. Therefore, I would recommend the publication of the manuscript after major revision. Below are my detailed comments and suggestions.

1. Lines 35 and 47: line 35 states that the cold water is featured by moderate salinity, while line 47 says it is a low salinity water mass. Please be consistent.
2. Lines 39-40 and Lines 47-49: information duplication.
3. Lines 59-70: I would recommend the authors to rewrite this portion. First, other cold water masses are not directly related to this work, so it is unnecessary to include so many details. Please be concise and remove those examples. Second, the question need to be answered in a study should be very specific. Why you still use an example to express the question?
4. Lines 71-77: This portion is poorly organized. In question one, why you say it is a special cold pool structure? Question two is too vague. Influence what features of the anti-cyclonic gyre? Morphology, Duration, Magnitude, Timing, or something else? There is no need to list only two questions using one paragraph.
5. Line 82: cite the wrong paper. It should be Chen’s 2003 or 2006 paper. Also there are a lot of citation format issues in this work. Please double check your citations.

Chen, C., Liu, H. and Beardsley, R.C., 2003. An unstructured grid, finite-volume, three-dimensional, primitive equations ocean model: application to coastal ocean and estuaries. Journal of atmospheric and oceanic technology, 20(1), pp.159-186.

Chen, C., Beardsley, R. and Cowles, G., 2006. An unstructured grid, finite-volume coastal ocean model (FVCOM) system. Oceanography, 19(1), pp.78-89.

1. Line 82: remove “while”.
2. Line 89: provide citation of CFSv2 dataset.
3. Lines 93-94: why only one sentence in this paragraph? Is it necessary to use one paragraph to state tidal forcing?
4. Line 101: Very confusing. Please rewrite this sentence.
5. Lines 103-104: why you choose these four time frames? Also there is no information regarding your 9-year climatological simulation.
6. Line 105: why you choose year 2019? If the cold water mass has strong interannual variability, can 2019 scenario represent the normal year condition?
7. Lines 115-116: why you say “these results are not different if we consider the mean T …”? In your example the benchmark’s mean T is around 5 while the sensitivity test’s mean T is ~7.5. The rationale behind this setup is very tricky.
8. Line 122: Results.
9. Lines 124-126: the structure of this part is weird. There is no need to have this portion as a paragraph, especially when it does not provide too much detailed information.
10. Lines 129-130: you already included the similar information in Introduction, no need to repeat it here.
11. Lines 131-133: the figure should be included in the supplementary material to help readers better understand your work.
12. Line 137: remove “moreover”. “the shape of … has a … direction”? Rephrase this sentence.
13. Line 138: it should be Shandong coastal current, not Bohai coastal current or Bohai coast current.
14. Line 139: In Fig. 2, why not show data in June?
15. Line 141: “Diao et al” citation issue.
16. Line 144: the figures directly jump from Fig. 2 to Fig. 15?
17. Line 147: Additionally?
18. 1. The color gradient of colorbar and the x axis labels are not well matched.
19. Line 161: to July?
20. Line 163: it is hard to tell the anticyclonic structure is closed from the figure.
21. Lines 199: In Figs. 5d and 6d, I don’t think the statement in the manuscript “vertical fricition term plays a role, especially east of 122E” match the figure.
22. Line 210: the subtopic is too long. Please be concise.
23. Line 216: why you choose 35.5 N transect? If you look at Fig. 1 this transect does not cross the center of the cold water mass.
24. Line 280: I don’t think the authors discussed Fig. 12 in the manuscript.
25. Lines 310-311: rephrase this sentence “for the eastern portion,….but for the eastern portion,…”
26. 15. The two transects show very similar results. I think one transect is enough.
27. Line 353: be concise. I feel this section does not relate to the topic of this research. I would recommend to remove it.
28. Line 369: Conclusions
29. Lines 372, 373, 378: anti-clockwise or anti-cyclonic?

This study presents the results of a local anticyclonic gyre around Qingdao cold water based on 3D numerical simulations. In short, the authors found that wind and tide are major factors influencing the anticyclonic structure, not the temperature and salinity gradients caused by the cold-water mass. The manuscript is generally well written but clarifications or more explanations and rearrangements of figures are suggested. There are also some typos. Please see below the detailed comments.

(1) A large bathymetry map identifying locations (e.g., Shandong Peninsula, Korean Peninsula), coastal seas (e.g., Yellow Sea, South Yellow Sea, Bohai Sea), and coastal currents (e.g., Bohai coast current, Yellow Sea warm current) mentioned in the paper are suggested to provide a big picture for readers who are not familiar with this region.

(2) Please supplement explanations on why ensemble experiments are needed. Why not directly using the realistic simulation but initializing the model for four times to get the ensemble?

(3) is the anticyclonic structure formed at depth shallower than 25 m?

(4) “temperature or salinity gradient” was used several times, are these horizontal or vertical temperature/salinity gradient?

(5) The short Section 3.1 Model validation could be merged into Section 2.1.

(6) ‘-fu’ in Equation (2) should be ‘+fu’. The expression of the barotropic pressure gradient force should be g*dζ/dx in Line 190.

(7) the identified different roles of wind and tide on upwelling that occurred east or west of the cold water mass is interesting. Yet, what leads to the different roles of wind and tide on different sides of the cold water mass seems lacking. The tide-induced front was mentioned. Could you add a plot showing where the tide-induced front is?

(8) All “anti-clockwise” in Conclusion should be “anti-cyclonic” or “clockwise”.

(9) Line 35: the cold pool is characterized by “moderate salinity” but later, it was described as “low salinity” (Lines 47, 53). Please check it.

(10) Line 41: “emergence” to “mergence”.

(11) Line 56: add reference(s) after “…the Qingdao cold water mass forms”.

(12) Line 57: “Yellow Sea cold water mass” should be “Qingdao cold water mass”?

(13) Line 70: “anticyclone” to “anticyclonic”

(14) Line 115: “However, these results are not different…” seems to be not correct. In the example, the mean temperature of this grid node’s ensemble is 4.875 ˚C for the control run and 7.375 ˚C for the no-tide run. Their difference is 2.5 ˚C, which is different from the “3 ˚C”.

(15) Line 141: (Diao, 2015) to (2015)

(16) Line 144: Figures should be mentioned in ordered sequence in the text.

(17) Line 161: “July” should be “June” (?)

(18) Lines 160, 168: “southeast monsoon” and “southward monsoon” are a bit confusing. Is it southeasterly monsoon or northward monsoon?

(19) Lines 178-179: please clarify “meaning that the Qingdao cold water mass is less affected by horizontal disturbances”.

(20) Lines 199-200: “east of 122 ˚E” should be “west of 122 ˚E”? change “southwest wind stress” to “southwesterly wind stress,” “northern current” to “northward current”.

(21) Line 299: please check “a decrease in the baroclinic pressure gradient force around the location of the Qingdao cold water mass in the no-wind experiment”. Is it “increase”?

(22) Line 335: “west side” should be “east side”?

(23) Add references in Section 4.4 for the Yellow Sea basin-scale cyclonic circulation.

(24) Question about the terminology: the anticyclonic circulation studied here disappears as the seasonal Qingdao cold water mass disappear. It’s not a permanent feature, maybe call it a ‘seasonal anti-cyclonic gyre’ is more accurate or just use ‘anti-cyclonic circulation’?

Specific comments about figures:

(1) Some figures can be combined for better visualization and comparison of results, for example, the following figures can be put together: Figures 1 and 2, Figures 5 and 6 (those two can be re-formatted into two rows and each row shows four plots), Figures 10 and 11.

(2) the unit of the momentum term in several figures is wrong: Figures 5, 6, 10, 11. It should be m/s².

(3) For those figures showing features at 25 m, some of them masked the nearshore region (I believe these are shallower than 25 m), like Figures 1-3, 5-6, but others have valid values in this shallow region, like Figure 10-11. Please check it.

(4) Figure 7: correct (e, f, g) to (d, e, f). Which direction does the positive value represent?

(5) Figure 8: there should be no unit for Ekman ratio.

(6) Figure 11: why is “time average over two M2 tide cycle” used here, rather than using “monthly mean for may” as Figure 10?

(7) Figure 12 is not cited in the main text. Figure 12a is the same plot as Figure 1b, keep one figure is fine.

(8) Figure 14 is repeated with one of the plots in Figure 15. You may consider put Figure 15 into supplementary. It’s also recommended to reduce the vector numbers in Figures 14-15 to make them easier to see.

(9) Figure A1: which period of the surface wind stress is shown in this figure? The caption indicates that this *constant* wind field was used in the control run (?)