Dear Authors

Many thanks for your submission and your comments during the discussion phase, and apologies there has been some delays in reaching a decision.

As someone who is familiar with the challenges of evaporation reduction and also the complexities of lake mixing I believe the study has significant merit and in particular is in an area where there is currently limited literature. Bringing a theoretical energy balance/aerodynamic approach to the problem I think will make the study a useful contribution.

The reviewers have a somewhat mixed view of the status of the manuscript in its present form, and so I have spent a fair bit of time myself reviewing and assessing the paper. In short, Reviewer 2 was supportive of publication with some suggestions for improvements, whilst Reviewer 1 indicated the results were too speculative and requested the paper has more evidence the model fits with observations, plus other theoretical justifications. Following my own assessment, I agree fully with the comments by both, and also share the concerns of Reviewer 1. Further, based on the reply to Reviewer 1 comments I was not convinced how the manuscript would be improved to adequately address the concerns. Therefore, it is recommended that the manuscript be re-considered after major revisions.

For the re-submission, I would suggest the structure of the paper is reorganised and believe this may address some of the concerns. I make the following points:

* The Lake Mead validation example appears out of context with the underlying motivation of the paper. The text refers many times to shallow reservoirs, and Figure1 sets this context clearly. Yet Lake Mead is a very deep reservoir where the surface mixed layer is impacted by complex processes (and I note the simple diffusivity model appears to have an error in the thermocline depth by about 10m during summer stratification). I find it hard to reconcile that validating the uncovered model on a reservoir of this size is also proof that the model is able to capture heat fluxes from a 3m deep reservoir (covered or uncovered). Further, this un-covered model validation is presented as a result (Figure 4; Section 3.1), but this does not clearly align to an aim presented in the introduction, and seems better suited to an appendix - i.e. supporting material to help bench-mark the model prior to its application here. A second bench-mark simulation on a "shallow" reservoir would greatly help strengthen this argument - that your (uncovered) model is a reasonable tool for both shallow and deep reservoirs. Accessing public thermistor chain datasets through something like the GLEON network may provide a case-study?

* Following on from the comment above about the structure, the aims of the paper should more clearly align to the results and key discussion points. Currently, the connection is not always obvious - for example, there are two sections in the "Results and Discussions" that relate to ecology /gas fluxes (Section 3.3) and economics of covers (Section 3.4), yet neither of these sections present actual results, and also neither of these sections match the objectives listed in the introduction - usually sub-sections in the results/discussion should logically map to the objectives. Therefore, a revised submission should not use a combined results/discussion section, and better organise the sub-headings to separate results and discussion and to more obviously align with the objectives of the paper as they are argued in the introduction - i.e. every figure or sub-section should clearly align with one of the aims/objectives.

* The Lab Experiment presented in Figure A1 is poorly integrated into the paper, currently just mentioned as a "motivation" for the model context and approach. In the reply to Reviewer 1 another Lab based figure is presented (this one has black discs whereas the first figure has white discs). This is currently not clear - in the reply you state that you "incorporate insights" from your lab, and later state that the lab scale experiments "were in good agreement with model predictions for evaporation suppression and effects of cover color (as mentioned in page 15, line 3)" - however on Page 15 line 3 there is no specific comparison of the model with the lab data. I note it is stated it will be published a separate paper, however, that is not adequate for this paper. In the revision, the authors should better integrate the lab aspect of the study as a component of the paper (with a corresponding aim and result), or not mention it, or wait until the other paper is published. If it is included or referred to then the insights that link the lab and model should be clear rather than general statements that the agree. Related to my point above, I note that the validation of the Lake Mead is presented as a key result of this paper whereas the link with the lab results are an appendix - I would consider this should be the other way around given the aim of the paper is focused on covered systems.

* To address the lack of validation data for covered reservoirs, it may be worth refocusing the aims on exploring the potential sensitivities of the model to guide further experiments. For example, I found Appendix C to be an interesting result, yet it is only an appendix. More systematically demonstrating how the results vary in response to poorly known variables associated with the cover or lake will help make the model results more compelling, even in the absence of a full validation.

* The limitations of the model and needs for further research need to more comprehensively be discussed (in a discussion section). Currently there are aspects of this in the combined results/discussion. One issue raised by the reviewers, and is a big difference between Lake Mead and a small turbid reservoir, is the potential impact of non-neutral atmospheric boundary layer conditions - how would this impact the results? Others are raised in the discussion, and addressing them will enrich the paper

* The summary needs to be re-written as a conclusion to more clearly align with the specific aims and highlight the most important findings and limitations/recommendations rather than repeat the study in general.

* The paper needs to remove the speculative statements and more carefully word findings to match the specific results presented - for example, where it is stated the model was validated against the lab (page 15), or COMSOL simulations (page 13), this needs to be fully supported/justified rather than just asking us to trust and have faith that things are good. Statements such as "Floating elements efficiently suppress evaporative loss from water reservoirs and alter energy storage within the reservoir and oxygen exchange at the water-air interface" .... are not entirely supported by the current results and need to be much more carefully worded (there are no results related to oxygen exchange so how is this statement proven by this study?). To address the Reviewer 1 concerns, the paper must more clearly highlight that this study is assessing dynamics with a model that is yet to be validated on covered conditions.

* The paper needs a thorough review for grammer and expression, as many proof-reading errors occur throughout.

Whilst I realise there are some substantial changes recommended here, I hope you will find these comments useful. Should you decide to re-submit a revision that can address these concerns and those of the reviewers, please provide a point by point response to the 2 reviewers comments and my comments above, along with the revised manuscript outlining specifically how you have addressed the issues (or where no change is made then please provide a justification). Given some of the revisions may take time please contact us should you require an extension to the revision submission date.

I look forward to seeing the revised manuscript and please feel free to contact me or the editorial team should you wish to seek further clarification.

Many thanks
Matt Hipsey

Dear Authors

Thank you for the revised paper and the replies to the discussion comments. I have sent the paper to the reviewers for their further comment on the suitability for publication. Again, the reviewers remain divided on whether it is fit for publication and so I have also had a detailed look over the new version to come to a decision.

Whilst the paper has had some modifications, they don’t appear to resolve the fundamental concerns. Usually at this point we would reject the paper, however, I do ultimately want to see this work published as I feel it is a worthy contribution. I acknowledge that this is a first step, but nonetheless it must also be robust enough for people to refer and cite the paper without ambiguity as to how accurate the approach is.

In order to progress the paper, the main issues that were not resolved in this revision must be addressed. I am paraphrasing here, but in short, there is currently weak or ambiguous connections between the Lake Mead, hypothetical reservoir, and lab scale components of the study. A more logical connection of the 3 work components is needed.

* The Lake Mead example is somewhat out of place in the paper as it currently stands. In your reply you assert that it provides a reference case for the model. If the Lake Mead case is the "reference", then the "test" should be a simulation comparing a covered Lake Mead with uncovered Lake Mead (which is not provided). This would not validate the floating cover module, but at least would logically fit the argument of using this particular validation data-set as a reference. I completely agree with the need to validate your model with the data you have available and provide confidence that it works, but the 1D mixing model chosen is well-documented and widely used and is not a novel contribution of this study. In fact there are many far more advanced lake mixing model approaches that have been published since H-S (1985). In this case, the thermal profile validation is simply a bench-marking exercise that was important for you to do, but is not core to the main aims of the paper. To make my point I note that this aspect of the paper is not referred to in the title, abstract or discussion and therefore is not adding to the overall story.

* The issue of model verification has not been addressed, and the connection between the 10m deep lake simulation and the lab-scale study has not been made. The Reviewer had requested empirical verification of lakes covered with floating elements but this was not provided. I had also previously suggested to move the lab study from the Appendix to the main paper on the assumption that the lab data would be used somehow to validate the model. However, the lab study now seems tacked on after the model is presented and the links with the model remain weak. Obviously, a model simulation of the lab-scale study is ideal, but even a basic comparison of model performance in reproducing the trend in Fig 11a could be possible?.

Overall, for the paper to be robust enough for publication in HESS, the methods and results in particular need a major overhaul to more logically integrate the lab-scale, small reservoir scale and deep reservoir scale aspects of the analysis. This must include attempts at model verification for covered predictions. Ideally simulations comparing cover and no-cover for each scale could also be provided to really demonstrate the power of the new model. And depending on how robust your verification is at any given scale, more careful use of wording about the suitability of the model as a design tool is also necessary. I realise these requests may seem frustrating given the time invested in your paper to date, but do hope you are able to make these modifications as I can see the potential of this paper. Feel free to contact me should you wish to clarify any aspects before committing to the revision. Or alternatively, if you are not able to make such changes then please let myself or the editorial team know so we can withdraw the paper and allow you to re-submit elsewhere.

All the best
Matt Hipsey

Dear Authors

Thank you again for your revised submission and adding the validation of the model at the experiment scale.

This addition addresses the issue raised by the reviewer and justifies the link to the empirical work. However - the quality of the text needs some further improvement. See attached for specific details and suggestions.

Kind Regards
Matt

Dear Dr Aminzadeh and co-authors
I think the final revisions have improved clarity and I am now pleased to accept your manuscript for publication.
I think this is a significant contribution and will be of wide interest. Thank you for the submission and best wishes for your ongoing research in this area.
Best wishes
Matt