General comments

This manuscript evaluates the role of the background circulation, tides, and rivers on the South Vietnam Upwelling (SVU) system during the summer of 2018. The methodology and analysis is well though and convincing evidence is presented supporting that:

1. Wind forcing is the leading mechanism explaining the chronology of the upwelling in the 4 areas studied
2. The upwelling during the summer of 2018 in the Mekong shelf (MKU) is due to the convergence and divergence of currents (and the associated downwelling and upwelling), which are strongly influenced by topography and to a lesser extent by tides. 
3. Tides increase the horizontal velocity gradient and hence the associated downwelling/upwelling
4. Tidal mixing enhances the upwelling by mixing of water column (a local effect) and by advecting water mixed upstream (a remote effect)
5. River discharge inhibits the development of MKU through enhanced stratification

The only general comment/doubt I have is that this study focuses on a single year of intense upwelling (summer of 2018). I wonder how representative the identified mechanisms are for other years. The narrative in the conclusions seem to imply this is the case. The role of the tides should be the same for other years, but the details regarding the convergence/divergence of currents could be more variable, both in intensity and perhaps in localization, despite the fact that topographic steering plays an important role. I also think that providing some model validation could be useful. Is the upwelling well represented by the model when confronted by observations?. Is the MKU upwelling observed by the satellite SST products? If so, describing this could help the reader to assess how representative the mechanisms described here are.

One additional note: Fig. 9 nicely summarizes the findings described in the manuscript. Thank you!. However it seems to me that such Figure is not referenced in the manuscript (or maybe I missed it).

Specific comments

P4,L17. It is mentioned that the simulation covers 2017-2018. However just the summer of 2018 seems to be analyzed. Why? Please clarify

P4: since Kz is reported later in the analysis of the sensitivity to river discharge experiment, it could be good to mention what vertical mixing parameterization is used in the model

P5: why MKU does not include the coast?

P7: in “SEJ and eddy dipole are well established in July-August, preventing NCU to develop, and much weaker at the beginning (June) and end (September) of summer, allowing NCU to develop.” A similar effect is observed for the OIV, could you comment on this?

In “Strong mid-July and mid-August wind peaks indeed induce peaks of upwelling intensity for MKU, SCU and OFU in the FULL ensemble analyzed in this previous study, while NCU does not develop during the core of the summer but in late June and late August (Fig. 2).” Could you comment on how different is the wind stress for other years?. That is, I wonder how representative these results are for other years.

P7: In “The effect of river discharge on the intrinsic variability of in MK is much weaker :”  This is a bit counter-intuitive. River discharge will introduce strong buoyancy gradients in the mixed layer, which promote the development of submesoscale variability. Is this because BoxMK excludes the coastal region?

P8: In “ We take M17, i.e. the simulation where both the small and large scale states are given by conditions of January 2017 of the PSY4QV3R1 analysis”, Is there any validation of the model solution over this period? At least for the SST, SLA and maybe SSS ... If it was reported previously, it would be good to reference that.

P9: In “the edge of the continental slope, this barotropic warm northeastward current meets a cold bottom northwestward current that flows from the open area towards the coast (Figs. 3d,4d)” Is this really at the edge of the continental slope?

P13: In “This stratification weakening makes the water column easier to mix vertically, facilitating the tidal vertical mixing, which is the main contributor to MKU in this area and during the transition period, as explained above” Is there any extra contribution due to wind forced coastal upwelling in the NoRiver case?

P13: In “Three sensitivity”, the FULL experiment is the reference state.Therefore two sensitivity experiments were performed (NoTide and NoRiver).

Technical corrections

Note: English is not my first language. Therefore non-typo corrections are just suggestions to be taken with a (huge) grain of salt.

P1: Maybe change “investigating” by “investigates”?

P2: In “∼ 14°E”,  is this 14 deg N?

P2, L56: change “and” by “are”?

P3,L67: Change “tidal” by  “the tidal”?

P4: Maybe change “ensemblist” by “probabilistic”?

P4:L108: Change “refsec:conclusion” by 5.

P7: “n :”  There are numerous spaces before punctuation colons. I wonder if this was just an effect of the pdf rendering …

P11: Change “°C” by “deg E”

P12: In “tidal mixing is a major factor of MKU maintaining.” Maybe “on the maintenance of MKU”?

 P12, L360: “in” by “on”?

P12: In “fresh hence light ..” maybe “fresher (and hence lighter) water ..”

This article describes the authors' analysis of an ensemble of simulations to analyze the SST variability in the Gulf of Thailand off the coast of Vietnam. They use a ten-member ensemble computed with a high-resolution ocean model to compare the impacts of eddies, tides, and river runoff during a two-year period. They find that the SST, which they characterize in terms of an "upwelling index", is mostly determined by winds, but in the Mekong Delta region they find that tides alter the currents and vertical turbulent heat transport enough to significantly influence the SST. This paper should be of interest to oceanographers and fisheries scientists interested in the detailed analysis of this region.

Overall, I found the paper well-written and logically laid out. My main questions and suggestions for improvement are as follows:

(1) I found the initial characterization of "upwelling" in terms of SST confusing. I think readers will be confused because upwelling normally refers to vertical velocity, but it is clear that it is really SST which is the focus of this article. I understand that this may be done because they want to use language and diagnostics consistent with previous analyses of this area, but it would be helpful to be explicit about this distinction.

(2) Because there is no discussion of degrees of freedom, I don't think that the t-test and F-test estimates of statistical significance are justified or useful. The discussion also mentions correlations between the FULL NoTides and NoRivers simulations, and I think this is a perfectly acceptable qualitative characterization of the results instead.

(3) Because the results find, essentially, that the tides influence Mekong upwelling, and the other regions are not much affected, I think the other regions could mostly be left out of the discussion. Perhaps the discussion of the other regions could be emphasized only in the introductory material. Similarly, in the abstract, consider re-ordering the presentation so that the positive or significant results are stated first, and the no-impact results are stated as a contrasting follow-up.

(4) I think the article would be more impactful if it emphasized the analysis of the term balance of the temperature evolution equation. You could map the Simpson-Hunter number (the ratio of surface heat flux to tidally-generated vertical turbulent heat flux) to delineate the regions where you would might expect tidal currents to be significant in the SST budget. Since you mention the significance of locally-created vs non-local (horizontally-advected) stratification, could you quantify this by mapping an appropriate non-dimensional number. Likewise, in the analysis of divergence you highlight the role of topographic features. This could be captured by comparing bottom u*grad(H) vs. H*div(u) at mid-depth.

(5) There is no discussion to justify the randomization technique used to create the ensemble. The approach taken seems perfectly reasonable, but it would be good to discuss why the 100km cutoff between large-scale and small-scale is appropriate, and also to mention why other sources of randomness (such as the winds or large-scale stratification) were not used.

Overall, I think this careful analysis will be of interest to researchers studying this region. I recommend publication after minor revision, if the authors choose; or they might wish to pursue the more extensive revision implied by item (4), above.

[Dear authors and editors:

I spent the last 2.5 hours reading this manuscript and enumerating my detailed comments in this text box. Unfortunately, when I clicked "Intermediate save" the website asked me to authenticate again, and all of my comments were lost when this web page again re-opened. I am not willing to re-create my detailed comments. I have re-created my general comments, above.]

This manuscript presents an investigation of the effect of tides and rivers on upwelling intensity over upwelling areas in Vietnamese waters in the summer of 2018 at different time scales. In addition, the physical mechanism of Mekong upwelling development was further investigated. The results of this paper is very interesting, comprehensive, and deepen the understanding of upwelling and the dynamics of upwelling areas. The manuscript is quite fully done and well-constructed. I suggest the manuscript to be accepted after minor revisions. Here are my comments:     

1. Some technical mistakes should be corrected:
   1. “:” should not have a space before it.
   2. Line 29: “14 E” should be “14 N”
   3. Line 56: “and” should be “are”
   4. Line 81-82: two “moreover”, line 91-92: two “however”: should use one word!
   5. Line 108: “refsec:conclusion”?
   6. Line 173: “Fig.1e,f” should be “Figs. 1e,f”; similar to others!
   7. In section 2.1, the limits of four numerical domains should be added with longitudes and latitudes in the caption of Fig. 1 or in the main text. It is not so good to read if we always need to check the paper of To Duy et al. (2022) or Herrmann et al. (2023).
   8. In Fig. 1: the lower limit of the color bar is 26^oC so the center of the upwelling area shows a white (blank) color, this color bar should be extended!
2. The authors already simulated, showed, and emphasized the spatial differences of SST of only a representative year of 2018 (Fig. 1). Why did the authors discuss that representative year? Do you get similar conclusions for 2017?
3. Line 261: “It belongs to the large scale cyclonic circulation…” => I could not see this cyclonic circulation, is it not shown in Fig. 3 or it is anticyclonic?