Simulation of Water Balance Components Using SWAT Model at Sub Catchment Level

Round 1

Reviewer 1 Report

The authors describe the modelling of the water balance of a catchment with the Soil Water and Assessment Tool (SWAT) in South India. This is generally not new. Many other authors have done it in other regions before. The paper could be seen as a case study for the region. But that requires dealing with a water-related problem by applying SWAT in the catchment area concerned. Most of the paper is a description of the site conditions in the catchment (model input), the calibration and validation of the model and there is nothing new. The authors only use the routines implemented in SWAT. The application of the model for a processing of water management problems within the catchment cannot been seen. The illustration of different model outputs is very simple and not really meaningful. Five figures show water balance components of 11 sub-catchments in the same way and allow only a visual comparison. Any statistical tests are missing. Because the model was only validated with the discharge of the whole catchment it is not sure that the other components are also valid. Their uncertainties are not discussed. In summary, this manuscript can only be the basis of a generally improved paper. The current scientific content is not sufficient for publication.

I propose a major revision.

Some detailed hints and suggestions for improvement:

The second paragraph of the introduction is very long. It contains a lot of general information about different hydrological catchment models, which are not relevant for this study. Here, more focus could be put on the benefits of SWAT for the study region and more applications of SWAT from other regions with similar problems could be integrated. What should be done with the model in the study area? A clear application goal is missing. The exclusive presentationof the model application in the case study of the Chittar catchments (line 99/100) is not sufficient for a publication.

Line 34: What do you understand under soil water (SW)? This can be a lot, soil water content, soil water storage, soil water storage change, … The abbreviation “SW” is used very often in the paper, but it is not clear how it is determined. In Table 1 you call it a process. What process is SW? I do not know what the authors really mean.

Line 118: Here, an equation should be given for SW. How was it calculated?

The order of the outline in the Material and Methods section (first SWAT, then Study Area) shows that it is less about solving a problem in catchment in this paper, but only about a further SWAT modelling. That makes no sense. There is already enough of that in literature. SWAT is described in the manual.

The structure of the paper should be modified. The introduction should more focus on the problems to be solved in the larger region (India) and especially in the study catchment. Emphasise why SWAT is suitable for this and the model of choice. In 2.1 only the catchment is described, nothing is said about the water related problems there. You should elaborate more on the water problems in the catchment.

Caption of Figure 3: last slope class in legend should be “>15”

line 165: Write out abbreviations the first time (VNIR?)

line 182/183: Check sentence. Perhaps the colours chosen are too difficult to distinguish, but Fig 4 shows a decrease for residential and agricultural land and an increase of waste land.

Section 2.7 is not sufficient. The measured discharge data should be described more in detail (some statistical values, gaps). The criteria are not all clear. What means r-factor, p-factor? Abbreviations are not described.

Line 240/241: Where are reaches, longest path flow monitoring outlets in Fig 6?

Section 3.1: First paragraph is more method than result. Should be shifted. Not clear which time step the model used for calculation and which time step the authors used for calibration and validation, daily or monthly time step?

Fig 7: What is “cumes” in y-axis unit? Are this average monthly values in the figure?

Line 280: Do not switch between 0 and 1 respectively 0 and 100% for R² and NS

Line 281/282: PBias and RMSE have a unit

Table 2: Criteria not addressed in the text are dispensable.

Line 302: I think the “monthly WBCs” are modelled values, not observed? Are values really monthly values or annual values?

Line 313/314: Monthly mean ET values between 25 and 43 mm/month seems to be very small. I would have expected values up to 100 mm/month in the catchment. How large is the potential ET? The plausibility of the values should be discussed.

Line 316: What does this mean? What do you mean with “trend”? There is no statistical trend analyses of any results. So, the authors should no speak about trends.

Fig 8: Unit y-axis? Inappropriate illustration, nothing can really be identified, the annual cycle is described, but this is not discernible, perhaps a figure with average annual cycle and range would be more helpful.

Many WBCs are presented in the results and discussion part but their validity is week. They should be compared with measured values or results from literature. Their uncertainty should be discussed.

Line 357: Repetition to line 343/344

Line 371: Do not use the word “trend” without statistical verification.

The units of values in the text should be precise. Sometimes you write “mm”, sometimes “mm/month”. The time reference should always be written (mm/d, mm/month, mm/a).

Line 397/398: “Statistical indexes support the perfection of model development and it is most efficient and more significance for the water balance modeling over the Chittar catchment.” What is used to prove this thesis?

Conclusion: too long. First part of the conclusion is a summary, not conclusion. Points 1 to 4 are summaries, points 5 to 8 could be conclusions. Last paragraph is an outlook.

 

Author Response

The authors thank the Reviewers for their constructive comments as well as suggestions, which are very helpful for improving the quality of the manuscript. Authors have put our efforts to respond to all the comments of reviewers.  Response to each comment is given as follows where the reviewers’ comments are shown in italics (red color). The modifications incorporated into the revised manuscript are highlighted (yellow color).

Author Response File: Author Response.pdf

Reviewer 2 Report

This manuscript deals with a relevant issue. I have some remarks:

1.      The topic is relevant to the journal's scope. The paper is well-organized and well written. The organization of the article is satisfactory. The paper's title is brief and reflects the main theme of the paper. 

2.      The abstract is sufficiently informative. It is completely self-explanatory, briefly presents the topic, states the scope of the experiments, it indicates significant data. 

3.      The keywords are suitable so the article can be found in the current registers or indexes.

4. Conclusion Section: The authors mention the major and specific conclusions of their research study. Conclusions are sound and justified by the data. 

5. Figures and tables: The tables and figures are numbered sequentially. Titles of tables and figures are brief and informative. All the tables and figures included are referred. 

6. References Section: The references are accurate and relevant for the subject of the paper. 

7. Although the paper is generally well written.

8. In my opinion, the authors need to emphasize the scientific novelty of their work.

9. In the introduction, it is worth emphasizing the advantages of the selected model and writing about its possible disadvantages in comparison with other models.

On my opinion, the work provides an advance towards the current knowledge.

Author Response

Authors are thankful to the reviewer for spending their valuable time and effort to review this manuscript. Authors thank the reviewer encouraging remarks.  Authors polished several sentences to make them concise and clearer for better clarity.

Author Response File: Author Response.pdf

Reviewer 3 Report

The manuscript is well revised. After correcting the grammatical errors, it is acceptable.

Author Response

The manuscript is well revised. After correcting the grammatical errors, it is acceptable.

Authors thank the reviewer encouraging remarks. Authors completely checked again the typo error, spelling and grammar in the revised manuscript. Authors also polished several sentences to make them concise and clearer. Since this has been done in many places, the line numbers to show the changes in the revised manuscript are not highlighted.

Reviewer 4 Report

General Comment:

This manuscript evaluated the Water Balance Components(WBCs) by building the SWAT model in the Chittar catchment, Tamilnadu, India. The topic is interesting, the methodology sounds reasonable, and the results could be used as a reference for local water research management.  Some issues that need to be addressed before this can be considered for publication. The following are some specific comments to this manuscript.

 

 

Specific Comments:

1.       Line 31: I think the word “specter” was not used properly, it should be replaced by a proper one.

2.       In part of “Method”, I think the description of SWAT was not enough, the authors should described more in detail.

3.       In part of “Method”, the innovation points of the method were not obvious enough, I think the authors should describe it more in detail and compared with others’ research.

4.       Fig 5: there are so many information that authors wanted to show, but it looks a little messy, readers can not catch the key points easily. I think authors could add a pie chart to show the several largest classes  of the soil map.

5.       I think authors should point out the similarities with existing researches and the advantages of their study in the part of “Results and Discussions”, which other researches didn’t perform.

Author Response

This manuscript evaluated the Water Balance Components(WBCs) by building the SWAT model in the Chittar catchment, Tamilnadu, India. The topic is interesting, the methodology sounds reasonable, and the results could be used as a reference for local water research management.  Some issues that need to be addressed before this can be considered for publication. The following are some specific comments to this manuscript.

The authors thank the reviewers for their support and constructive comments as well as suggestions, which are very helpful for improving the quality of the manuscript. Authors have put our efforts to respond to all the comments of reviewers.  Response to each comment is given as follows where the reviewers’ comments are shown in red color. The modifications incorporated into the revised manuscript are highlighted in yellow color.

Specific Comments:

1. Line 31: I think the word “specter” was not used properly, it should be replaced by a proper one.

The authors corrected the above mentioned sentences in the revised manuscript

2. In part of “Method”, I think the description of SWAT was not enough, the authors should described more in detail.

Thanks for your comment. As per reviewers suggestion we have reduced the description about SWAT during previous round of review.  They clam than SWAT is extensively used in literature and also have well described manual. Adequate description about SWAT is available in manuscript.

3. In part of “Method”, the innovation points of the method were not obvious enough, I think the authors should describe it more in detail and compared with others’ research.

Thanks for your suggestion. The innovations in methodology and study area selection were discussed in Introduction and Materials & Methods chapters. However it is further strengthened in the revised manuscript.

4. Fig 5: there are so many information that authors wanted to show, but it looks a little messy, readers can not catch the key points easily. I think authors could add a pie chart to show the several largest classes of the soil map.

Thanks for your input. Chittar catchment has 75 soil classes. Hence providing pie chart with major soil classes will mislead the readers.  Accordingly the details of soil classes cover more than 50 sq km are given in the revised manuscript.

5. I think authors should point out the similarities with existing researches and the advantages of their study in the part of “Results and Discussions”, which other researches didn’t perform.

Results and discussion part further improved in the revised manuscript.

Round 2

Reviewer 1 Report

The authors have improved the manuscript in many points compared to the first version. All specific points of criticism have been clarified. However, the main points of criticism have not been eliminated in any way. An appropriate revision is a prerequisite for publication of the manuscript. I repeat once again the text from the first review:

„The illustration of different model outputs is very simple and not really meaningful. Five figures show water balance components of 11 sub-catchments in the same way and allow only a visual comparison. Any statistical tests are missing. Because the model was only validated with the discharge of the whole catchment it is not sure that the other components are also valid. Their uncertainties are not discussed.“

A response was made to the comments, but no significant changes were made to the manuscript. The response of the authors referred to the poor data situation in developing countries. This is true, but it cannot be an excuse for the inadequate presentation of the results, the lack of discussion of the uncertainties of the results and the lack of comparisons with the literature. SWAT can deal with such initial situations and calculate useful results. Nevertheless, these must be critically evaluated. The points that were already critically raised in the first review were not addressed in a sufficient way. These were the main reasons for the proposed major revision, not the specific references to some text passages and figures.

There are, for example, freely available data on ET derived from satellite observations (https://www.ntsg.umt.edu/project/modis/mod16.php). These could be used for comparison. The average monthly ET values still seem questionable and contradictory. On the one hand, reference is made to the very high ET in the region (line 109), on the other hand, average monthly totals of 25 to 43 mm (line 358) are given. This is low and not high. In this context, the long dry season with monthly totals of less than 5 mm for the ET is argued (Authors response: About 4 to 6 month in a year the ET is less than 5 mm/month at all sub catchment.). No vegetation would be able to survive at these ET values.  SC8 is supposed to have the minimum ET in the whole simulation period (Z 364/365). SC8 has a high proportion of forest and agricultural land (Fig 4). A higher ET is to be expected there. For my understanding, this indicates that the model results are not really critically evaluated.

In the response to the uncertainty of ET, reference was made to the FAO 56 source as the standard method for ET calculation. This is true. The standard is the calculation method and the grass reference method derived from it. Crop ET is calculated using crop coefficients Kc, which may or may not be correctly fitted in the model. Comparing the reference ET also calculated in SWAT with the calculated crop ET would also be a way to evaluate the calculated crop ET.

The conclusions were not really changed. They still do not meet the requirements of a scientific paper for conclusions. They are much too long. Most of it is a summary of what was already said earlier in the text.

A few more notes on errors in the manuscript:

- Fig. 4: the values of % and km2 in the upper pie chart do not match.

- The unit of PBias is not mm. % or without unit.

In summary, it must be said that a more fundamental revision is expected in a request for a major revision. Since the authors have very inadequately complied with this request, I decline to publish the manuscript.

Author Response

Comments and Suggestions for Authors

The authors have improved the manuscript in many points compared to the first version. All specific points of criticism have been clarified. However, the main points of criticism have not been eliminated in any way. An appropriate revision is a prerequisite for publication of the manuscript. I repeat once again the text from the first review:

Authors are thankful to the reviewer for spending their valuable time and effort to review this manuscript. Response to each comment is given as follows where the responses are shown in red colour. The modifications incorporated into the revised manuscript are highlighted (yellow color).

„The illustration of different model outputs is very simple and not really meaningful. Five figures show water balance components of 11 sub-catchments in the same way and allow only a visual comparison. Any statistical tests are missing.

Authors partially agreeing with reviewer’s view. The preliminary statistical results such as standard deviation, mean, min and maximum values for individual sub catchment are discussed in manuscript. Please note that SWAT is a lumped model at hydrological response unit and at the same time able to estimate the spatial variability only at sub-catchment level. Accordingly five figures provided here to show the spatial variability of five WBCs at 11 sub-catchments. Authors agreeing that statistical test provide mechanism for readers to make quantitative decisions based on these results. Authors searched many multivate statistics techniques and literature survey to showcase the relative preference among five WBCs at catchment level. However authors not able to identify appropriate statistical technique to handle this.

Because the model was only validated with the discharge of the whole catchment it is not sure that the other components are also valid. Their uncertainties are not discussed.“

In the revised manuscript, the satellite observations based actual evapotranspiration derived from GLEAM (https://www.gleam.eu/) at outlet of the catchment is compared with SWAT predicted actual evapotranspiration. Authors included more statistical test results (P value & t value) about model parameter uncertainty in the revised manuscript.

A response was made to the comments, but no significant changes were made to the manuscript. The response of the authors referred to the poor data situation in developing countries. This is true, but it cannot be an excuse for the inadequate presentation of the results, the lack of discussion of the uncertainties of the results and the lack of comparisons with the literature. SWAT can deal with such initial situations and calculate useful results. Nevertheless, these must be critically evaluated. The points that were already critically raised in the first review were not addressed in a sufficient way. These were the main reasons for the proposed major revision, not the specific references to some text passages and figures.

There are, for example, freely available data on ET derived from satellite observations (https://www.ntsg.umt.edu/project/modis/mod16.php). These could be used for comparison. The average monthly ET values still seem questionable and contradictory. On the one hand, reference is made to the very high ET in the region (line 109), on the other hand, average monthly totals of 25 to 43 mm (line 358) are given. This is low and not high. In this context, the long dry season with monthly totals of less than 5 mm for the ET is argued (Authors response: About 4 to 6 month in a year the ET is less than 5 mm/month at all sub catchment.). No vegetation would be able to survive at these ET values.  SC8 is supposed to have the minimum ET in the whole simulation period (Z 364/365). SC8 has a high proportion of forest and agricultural land (Fig 4). A higher ET is to be expected there. For my understanding, this indicates that the model results are not really critically evaluated.

In the response to the uncertainty of ET, reference was made to the FAO 56 source as the standard method for ET calculation. This is true. The standard is the calculation method and the grass reference method derived from it. Crop ET is calculated using crop coefficients Kc, which may or may not be correctly fitted in the model. Comparing the reference ET also calculated in SWAT with the calculated crop ET would also be a way to evaluate the calculated crop ET.

Thanks for the reviewer input. The Evapotranspiration (ET) and Actual Evapotranspiration (AE) are interchangeably used in this manuscript by mistake. This is corrected in the revised manuscript. In the revised manuscript, the satellite observations based AE derived from GLEAM (https://www.gleam.eu/) at outlet of the catchment is compared with SWAT predicted AE. It is observed that two data set reasonable matching.

The conclusions were not really changed. They still do not meet the requirements of a scientific paper for conclusions. They are much too long. Most of it is a summary of what was already said earlier in the text.

Authors made the statements as per the finding from this the study, however, some sentences with the general meaning are removed from the conclusion.

A few more notes on errors in the manuscript:

- Fig. 4: the values of % and km2 in the upper pie chart do not match.

Thanks for pointing out the mistake. It is corrected in Figure 4 and in the discussion.

- The unit of PBias is not mm. % or without unit.

Thanks for the input. It is changed as “%”in the revised manuscript.

In summary, it must be said that a more fundamental revision is expected in a request for a major revision. Since the authors have very inadequately complied with this request, I decline to publish the manuscript.

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

The authors have further improved the manuscript compared to the 2nd version. They have included the suggestion to compare the calculated actual evapotranspiration (AE) with freely available data based on satellite measurements. However, they have insufficiently described how they obtained the comparison values from the GLEAM dataset and with which model results they compared them. Based on their description, the authors seem to use only the values from one GLEAM grid cell at the Puram gauge position. They are not looking at the entire basin. Which values do the authors compare with the values of the grid cell? Do they compare these values to the values of the HRU at the Puram gauge, to the values of the subbasin at the Puram gauge, or to the mean values of the entire basin? It is not clear from the descriptions in the methods section.

As in the first review, the fundamental criticisms from the previous reviews were not eliminated.

·        In chapter 3 "Results and Discussion" the results are presented and described. They are not discussed in the sense of a scientific discussion.

·        The form of presentation in Figures 9 to 13 is still not very suitable to identify anything.

·        Chapter 4 "Conclusions" has also not really been revised. It is almost exclusively a summary of what has already been written before. What are the new findings from the modeling? What conclusions can be drawn from it? I can't see it.

The authors have very poorly responded to the prompts from the previous reviews and revised their manuscript fundamentally. I reject the manuscript.

Author Response

Thanks for the reviewers input. Authors have provided the response to all the comments of reviewer. The necessary changes are also incorporated in the revised manuscript. Response to each comment is given as follows where the responses are shown in red colour. The modifications incorporated into the revised manuscript are highlighted in yellow color.

The authors have further improved the manuscript compared to the 2nd version. They have included the suggestion to compare the calculated actual evapotranspiration (AE) with freely available data based on satellite measurements. However, they have insufficiently described how they obtained the comparison values from the GLEAM dataset and with which model results they compared them. Based on their description, the authors seem to use only the values from one GLEAM grid cell at the Puram gauge position. They are not looking at the entire basin. Which values do the authors compare with the values of the grid cell? Do they compare these values to the values of the HRU at the Puram gauge, to the values of the subbasin at the Puram gauge, or to the mean values of the entire basin? It is not clear from the descriptions in the methods section.

As per reviewer suggestion detail description about GLEAM dataset and comparison process are included. The sub catchment level monthly AE is compared with GLEAM grid cell data and it is clearly described in the revised manuscript. Further two addition sub catchments AE are also compared and reported in the revised manuscript.

As in the first review, the fundamental criticisms from the previous reviews were not eliminated.

• In chapter 3 "Results and Discussion" the results are presented and described. They are not discussed in the sense of a scientific discussion.

Results and discussion part further improved in the revised manuscript.

• The form of presentation in Figures 9 to 13 is still not very suitable to identify anything.

Figures 9 to 13 provided to show the spatial variability of five WBCs at 11 sub-catchments. And it is extensively discussed in the manuscript. Authors agreeing the reviewer’s view, however authors not able to identify appropriate statistical technique to handle this.

• Chapter 4 "Conclusions" has also not really been revised. It is almost exclusively a summary of what has already been written before. What are the new findings from the modeling? What conclusions can be drawn from it? I can't see it.

The conclusions were improved in the revised manuscript.

The authors have very poorly responded to the prompts from the previous reviews and revised their manuscript fundamentally. I reject the manuscript.

The authors thank the reviewers for their constructive comments, which are very helpful for improving the quality of the manuscript. Most of the comments/criticisms (i.e. except Figures 9 to 13 issues) raised during round 1 & 2 were implemented in the manuscript.