Modeling Chlorophyll a with Use of the SWAT Tool for the Nielba River (West-Central Poland) as an Example of an Unmonitored Watercourse

Round 1

Reviewer 1 Report

This work deals with the modelling chl-a with SWAT for the Nielba River (West-Central Poland) as an example of a small agricultural watershed. Some results are interesting and have potential signification for the management of the pollutant in watershed. However, some contents in this work may need to be further validated. My comments and suggestions are listed below for your consideration in improving the quality of the manuscript.

1. For the calculation equation of chlorophyll a in Section 2.3, there is no model improvement, and the literature of S.L. Neitsch et al. can be directly cited.
2. In the selection of influencing factors, flow velocity or hydraulic retention time are also important aspects that affect chlorophyll a in the channel. Please explain why the author did not choose it.
3. The discussion is too qualitative and lacks quantitative analysis. It is suggested that the author divide the flow and nutrients according to the upstream and the basin, focusing on the contribution or reduction of nutrients by the Łęknińskie Lake, Bracholińskie Lake and Rgielskie Lake.
4. The author chose Nielba river as the study area for SWAT modeling. Why the calibration and verification points are outside the watershed, which will cause inaccurate determination of model parameters, please explain.
5. Please simplify the conclusion, and the analysis of model uncertainty can be placed in the discussion。
6. There are some errors in the citation of the references in the manuscript. Please revise.

Author Response

Dear Reviewer,

I am very grateful for your review of my Manuscript and I appreciate your valuable comments. Below, I present a point-by-point answer to your remarks. They have all been implement into the text with a great benefit.

• For the calculation equation of chlorophyll a in Section 2.3, there is no model improvement, and the literature of S.L. Neitsch et al. can be directly cited.

As suggested by the reviewer the pertinent reference has been cited in the revised version of the manuscript

 

• In the selection of influencing factors, flow velocity or hydraulic retention time are also important aspects that affect chlorophyll a in the channel. Please explain why the author did not choose it

Indeed, as pointed out by the reviewer, there are many other parameters affecting chlorophyll a concentrations in the channel. However, to simplify the approach only parameters directly included in the SWAT model output were taken into consideration. In this way, any additional alterations of the received results were omitted. Since, the model indicates flow as a dominant feature of the riverine channel, the flow velocity has not been included in the final discussion. As for the hydraulic retention time the SWAT model considers “lateral flow travel time (LAT_TTIME)” as a proxy for this variable. LAT_TTIME is calculated based on hydraulic properties for each of the subbasins and relates variables involved in it. Therefore, was determined as not suitable for this discussion.

 

• The discussion is too qualitative and lacks quantitative analysis.

Taking into consideration all the reviewers’ comments the main goal of this study has been modified to better reflect the article content.  Moreover, the discussion and conclusion section have been modified accordingly. The study focuses mainly on the  SWAT model performance and its limitations in chlorophyll a estimations for unmonitored watercourse, for which no field data is available. Therefore, some quantitative aspects of are difficult to capture. The modified version of the manuscript offers two-stage approach to overcome this issue and enable reliable discussion on chlorophyll a behaviour in the studied watercourse.

 

• It is suggested that the author divide the flow and nutrients according to the upstream and the basin, focusing on the contribution or reduction of nutrients by the Łęknińskie Lake, Bracholińskie Lake and Rgielskie Lake.

The spatial pattern of chlorophyll a changes in the Nielba River has been presented already in a previous study (Orlińska-Woźniak et al. 2021, ES&T  55(16):11113-11124). Also, contribution of the lakes to the nutrient loads has been fully described there. In the current one, the main focus is on presentation of the SWAT model performance in relation to the chlorophyll a simulations in an unmonitored watercourse. To maintain manuscript brevity the issues not directly related to this subject were deliberately omitted. 

 

• The author chose Nielba river as the study area for SWAT modeling. Why the calibration and verification points are outside the watershed, which will cause inaccurate determination of model parameters, please explain.

The main goal of this study is to present the  SWAT model performance and its limitations in chlorophyll a estimations for unmonitored watercourse, for which no field data is available. Therefore, the Nielba River has been selected as a study venue. However, as mentioned in the section 2.2 Model calibration and validation, this watercourse constitutes a tributary of the Wełna River for which catchment the SWAT model has been successfully created. Therefore, calibration and verification points are indeed outside the Nielba River catchment, however they are located on the territory of the main river catchment. Since, the Wełna River SWAT model performance was assessed as more than satisfactory therefore, such study on the selected sub-catchment is fully justified.

 

• Please simplify the conclusion, and the analysis of model uncertainty can be placed in the discussion.

As suggested by the reviewer the conclusion section has been simplified and paragraphs dealing with the model limitations have been moved into the discussion section of the revised manuscript.

 

• There are some errors in the citation of the references in the manuscript. Please revise.

All the references cited in the manuscript have been checked and verified in the revised version of the manuscript.

Reviewer 2 Report

Review report

The paper presents a case-study of the Nielba River catchment in central-western Poland. The author applies the SWAT model to examine the relationships between Chla-loads and external (catchment) boundary variables. This is an interesting study, and the application of SWAT in this manner has good potential – in particular, the change in the relationships of the variables with Chla between sub-basins 1-4 and 5-7 - due to the lake influence - is interesting from a modelling perspective and illustrated nicely in Figure 4. This is also examined in the discussion section.

However, the author needs to present a stronger case for publication before the manuscript can be accepted in its current form. In addition, the text needs substantial revision.

 

Overall comments

The introduction states the aim of the study was to verify the SWAT model performance in algal biomass estimations for a pilot catchment. However, this does not seem to be discussed further within the manuscript. Instead, the following sections analyse the statistical relationships between Chla-load and various catchment variables (E.g., flow). After reading the manuscript it is still not clear how well the SWAT model performs in predicting algal biomass. There are no comparisons of model predictions for Chla versus observations for example. What is the conclusion? Can the SWAT model be used to predict Chla concentrations as boundary data for another model?

 

The author should ensure that citations are correctly used in the manuscript. I have not examined all the citations through the text but did note the citations in lines 50-51 of the manuscript (citations #20-#28). Here, some of these citations are used inappropriately. For example, the authors cited as reference #20 apply the CE-QUAL-W2 model in their work but have no connection to DELFT3D as cited in this current manuscript. Similarly, the authors cited as reference #28 apply the SWAT model in their work but have no connection to HSPF as cited in this current manuscript. Citations should be used when presenting results/concepts/conclusions etc. of other authors – citations should not be included indiscriminately to ‘back-up’ any piece of text. I suggest the author reviews each citation within the text and assesses if it is a) necessary, and b) used correctly.

 

There seems to be an issue with the editing software in that the references to figures, equations etc are in non-English characters. For example, lines 97, 129, 177, 187, 192, and so forth.

 

There is insufficient information on the calibration and set-up of the SWAT model. Lines 132-133 state the model was calibrated “using the SWAT-CUP program using the SUF1-2 algorithm”. The author should provide a short description in the text as to procedure behind this method for a reader unfamiliar with SWAT. After reading the section 2.2. I am left with questions such as how was the SWAT model parameterised – what were the important coefficients? How many sample observations of Chla, TP, & TN were there per sub-catchment? The supporting information S2 gives limited information as the full names of the parameters are not provided so it is unclear what any of the parameters are. In the supporting information S3 it would be good to see calibration plots of the model predictions for flow, TN, TP and Chla plotted against the sample observations.

 

In the results, I found I was confused in several sections as to what part of the study was being presented in each of the figures. The analyses seemed to be performed in several stages, but I found the results section hard to follow. Figures 2 and 4 seemed to be presenting the same data but just with Figure 2 plotting the results of the subbasins on one axis and Figure 4 plotting each subbasin separately. Figure 2 appears unnecessary if this is the case.

 

The text makes no attempt to demonstrate the importance of this work or how it can be applied to other aquatic systems. Can it be applied to other systems – this should be discussed? How has this work advanced current scientific knowledge? How can it be built upon – what is next?

 

Specific comments

Lines 49-57:

As written, the author appears to be arguing that models such as CE-QUAL-W2, WASP etc. are less appropriate than SWAT for modelling eutrophication in riverine conditions as SWAT simultaneously simulates hydrological, geochemical, and ecological processes, but the others (apparently) do not. I do not believe this is the intention of the author to infer this, and it is more an issue with how the paragraph is worded. I am not familiar with some of the models, but CE-QUAL-W2 and WASP can also track changes in nutrient transport, hydrology, and biological processes simultaneously in rivers. If SWAT is the model of choice, then that is completely fine, but it should not be implied that the other models listed are not adequate for modelling river eutrophication in general.

 

Lines 63-65:

“Earlier studies have shown….. “ Which studies in particular? The author cites only one (#33)

 

Lines 168-169 (also lines 303-304):

“Due to the low data number of Chla…” What is low? It is not specified how many samples there are in the manuscript

 

Lines 353-355:

“Although, the high…..  nutrient loads” This sentence is not very clear.

Author Response

Dear Reviewer,

I am very grateful for your review of my Manuscript and I appreciate your valuable comments. Below, I present a point-by-point answer to your remarks. They have all been implement into the text with a great benefit.

• The introduction states the aim of the study was to verify the SWAT model performance in algal biomass estimations for a pilot catchment. However, this does not seem to be discussed further within the manuscript. Instead, the following sections analyse the statistical relationships between Chla-load and various catchment variables (E.g., flow). After reading the manuscript it is still not clear how well the SWAT model performs in predicting algal biomass. There are no comparisons of model predictions for Chla versus observations for example. What is the conclusion? Can the SWAT model be used to predict Chla concentrations as boundary data for another model?

Taking into consideration all the reviewers’ comments the main goal of this study has been modified to better reflect the article content. The study focuses mainly on the SWAT model performance and its limitations in chlorophyll a estimations for unmonitored watercourse, for which no field data is available. The watercourse of choice (Nielba River) constitutes the part of the Wełna River model which has been calibrated and verified (additional information included in S2 and S3). The performance of this model has been assessed as more than satisfactory in respect to TN and TP. In case of chlorophyll a there is a limited information available on the calibration/validation coefficient values range. Therefore, this study offers a possibility to assess the SWAT model performance in respect to chlorophyll a changes in a two-step approach presented in this study. Indeed, the SWAT model chlorophyll a results could be used as a boundary data for another model if necessary. Pertinent changes have been introduced in the all sections of the revised manuscript.

 

• The author should ensure that citations are correctly used in the manuscript. I have not examined all the citations through the text but did note the citations in lines 50-51 of the manuscript (citations #20-#28). Here, some of these citations are used inappropriately. For example, the authors cited as reference #20 apply the CE-QUAL-W2 model in their work but have no connection to DELFT3D as cited in this current manuscript. Similarly, the authors cited as reference #28 apply the SWAT model in their work but have no connection to HSPF as cited in this current manuscript. Citations should be used when presenting results/concepts/conclusions etc. of other authors – citations should not be included indiscriminately to ‘back-up’ any piece of text. I suggest the author reviews each citation within the text and assesses if it is a) necessary, and b) used correctly.

All the citations have been verified in the revised version of the manuscript.

 

• There seems to be an issue with the editing software in that the references to figures, equations etc are in non-English characters. For example, lines 97, 129, 177, 187, 192, and so forth.

Unfortunately, the final appearance of the manuscript has been awfully altered during the submission process. All the editorial issues, including font changes, have been rectified in the in the revised version of the manuscript.

 

• There is insufficient information on the calibration and set-up of the SWAT model. Lines 132-133 state the model was calibrated “using the SWAT-CUP program using the SUF1-2 algorithm”. The author should provide a short description in the text as to procedure behind this method for a reader unfamiliar with SWAT. After reading the section 2.2. I am left with questions such as how was the SWAT model parameterised – what were the important coefficients? How many sample observations of Chla, TP, & TN were there per sub-catchment? The supporting information S2 gives limited information as the full names of the parameters are not provided so it is unclear what any of the parameters are. In the supporting information S3 it would be good to see calibration plots of the model predictions for flow, TN, TP and Chla plotted against the sample observations.

Both supporting information files (S2 and S3) have been amended, and pertinent information on model calibration provided in the revised version of the manuscript. Moreover, the requested information on parameters, coefficients, and also flow, TN ,and TP model performance (plots) have been provided.

 

• In the results, I found I was confused in several sections as to what part of the study was being presented in each of the figures. The analyses seemed to be performed in several stages, but I found the results section hard to follow.

The two-step approach applied in the current study has been clarified in the result and discussion sections.

 

• Figures 2 and 4 seemed to be presenting the same data but just with Figure 2 plotting the results of the subbasins on one axis and Figure 4 plotting each subbasin separately. Figure 2 appears unnecessary if this is the case.

Indeed, the database presented in both figures is basically the same. However, the presented in the article approach requires two steps. The first one is performed on the pooled information for the whole studied river. The second step gives opportunity to discuss results for the separate subbasins. This issue has been clarified in all the pertinent manuscript sections. Moreover, Figure 2 has been modified accordingly to avoid further distraction. 

 

• The text makes no attempt to demonstrate the importance of this work or how it can be applied to other aquatic systems. Can it be applied to other systems – this should be discussed? How has this work advanced current scientific knowledge? How can it be built upon – what is next?

  The importance of the current study has been presented in the modified version of the conclusion sections.

 

Specific comments

• Lines 49-57:

As written, the author appears to be arguing that models such as CE-QUAL-W2, WASP etc. are less appropriate than SWAT for modelling eutrophication in riverine conditions as SWAT simultaneously simulates hydrological, geochemical, and ecological processes, but the others (apparently) do not. I do not believe this is the intention of the author to infer this, and it is more an issue with how the paragraph is worded. I am not familiar with some of the models, but CE-QUAL-W2 and WASP can also track changes in nutrient transport, hydrology, and biological processes simultaneously in rivers. If SWAT is the model of choice, then that is completely fine, but it should not be implied that the other models listed are not adequate for modelling river eutrophication in general.

  As rightly pointed out by the reviewer the SWAT model was just a model of choice in the current study. The wording of the pertinent paragraph has been modified.

 

• Lines 63-65:

“Earlier studies have shown….. “ Which studies in particular? The author cites only one (#33)

The pertinent part of the text has been modified and the citations amended.

 

• Lines 168-169 (also lines 303-304):

“Due to the low data number of Chla…” What is low? It is not specified how many samples there are in the manuscript

The table with classification of value ranges for the statistical measures used in calibration, verification and validation of the model for flow, nitrogen, total phosphorus and suspended sediment has been added to the supporting information file (SI). It was also emphasized that no such classification is available for chlorophyll a. On this basis, the manuscript has been corrected in line with the reviewer's suggestions.

 

• Lines 353-355:

“Although, the high…..  nutrient loads” This sentence is not very clear.

The indicated sentence has been modified in the revised version of the manuscript.

Reviewer 3 Report

The article under review entitled "Modeling chlorophyll a with use the SWAT tool for the Nielba River (West-Central Poland) as an example of a small agricultural watershed", concerns the use of the application of the SWAT model to simulate the load on riverine systems with chlorophyll a. The results of the analysis gives interesting conclusions and in my opinion, the paper may be published in the “Water” journal with a few improvements. The comments are below.

Comments:

Please standardise the font throughout the article (for example lines 15, 474-80).

I do not understand what mean the references in Chinese/Japanese language in the manuscript? For example, lines: 97, 129, 177 etc. I suppose these are some editing errors. Please fix this.

Line 126: What is "sediment" parameter? Should be "suspension". This chapter should also include the abbreviations of the analysed parameters used in the further part of the manuscript (TP, TN, Chla etc.).

Line 187: Please use the abbreviation "Chl a" or "Chla" consistently.

Lines 183-4: Kmoles are the load unit?

Equation 1: Please perform the operation on the units in the equation 1. Actions on provided by the author units do not give the expected result. Please explain that.

Lines 194-6: Please use consistent unit notation throughout the manuscript (for chlorophyll
µg chla/l or µg/l etc.). This applies to the entire manuscript and all units.

Discussion: The discussion in relation to literature is very poor. Please enrich individual elements of the discussion with references.

Figure 3: Some p-values are unreadable.

Figure 4: “Nutrient load” or “nitrogen load”?

Supporting Information S2: The abbreviations of the analysed parameters are not explained and therefore incomprehensible to the reader. Parameters: nitrogen, phosphorus, suspension should be in English. The "flow" parameter is missing from the table.

Supporting Information S3: Replace "sediment" with "suspension" and "CHLA" with "Chl a" or "Chla".

Author Response

Dear Reviewer,

I am very grateful for your review of my Manuscript and I appreciate your valuable comments. Below, I present a point-by-point answer to your remarks. They have all been implement into the text with a great benefit.

• Please standardise the font throughout the article (for example lines 15, 474-80). I do not understand what mean the references in Chinese/Japanese language in the manuscript? For example, lines: 97, 129, 177 etc. I suppose these are some editing errors. Please fix this.

Unfortunately, the final appearance of the manuscript has been seriously altered during the submission process. All the editoral issues, including font changes, have been rectified in the in the revised version of the manuscript

 

• Line 126: What is "sediment" parameter? Should be "suspension". This chapter should also include the abbreviations of the analysed parameters used in the further part of the manuscript (TP, TN, Chla etc.).

The “sediment” parameter indeed means suspended sediment being transported in the river bed - the nomenclature has been corrected. Due to the fact that the article uses, for example, two different chlorophyll parameters, in order to maintain the readability of the entire text, only the full names of individual parameters were used (suspended sediment, total nitrogen, total phosphorus, chlorophyll a load and chlorophyll a land delivery).

 

• Line 187: Please use the abbreviation "Chl a" or "Chla" consistently.

The pertinent abbreviation has been unified in the entire revised manuscript.

 

• Lines 183-4: Kmoles are the load unit? ???

Equation 1: Please perform the operation on the units in the equation 1. Actions on provided by the author units do not give the expected result. Please explain that.

The pertinent equation follows the functional relationship presented by Cluis et al (1988) which has been considered a baseline for the SWAT nutrient modeling module. Indeed, the units for TN and TP loads were quoted wrongly, therefore this part has been corrected in the revised manuscript.

 

• Lines 194-6: Please use consistent unit notation throughout the manuscript (for chlorophyll µg chla/l or µg/l etc.). This applies to the entire manuscript and all units.

The pertinent unit notation has been unified in the entire revised manuscript.

 

• Discussion: The discussion in relation to literature is very poor. Please enrich individual elements of the discussion with references.

In the revised manuscript the discussion have been enriched with references.

 

• Figure 3: Some p-values are unreadable.

The readability of Figure 3 has been improved in the revised version of the manuscript.

 

• Figure 4: “Nutrient load” or “nitrogen load”?

The pertinent term has been unified in the entire revised manuscript.

 

• Supporting Information S2: The abbreviations of the analysed parameters are not explained and therefore incomprehensible to the reader. Parameters: nitrogen, phosphorus, suspension should be in English. The "flow" parameter is missing from the table.

The abbreviations of the analysed parameters have been explained in the revised version of the S2. Also, the flow parameter has been added into the table.

 

• Supporting Information S3: Replace "sediment" with "suspension" and "CHLA" with "Chl a" or "Chla".

The pertinent changes have been introduced into the revised version of the S2.

Round 2

Reviewer 1 Report

The manuscript has been revised according to the comments. It can be accepted.

Reviewer 2 Report

The author has revised the manuscript to make the sections more focused and clearer to read. Substantial information has been incorporated into the supplemental information that provides sufficient information with respect to the methodology. It is an interesting study and I recommended the manuscript for publication subject to (very) minor corrections of a small amount of typing and spelling errors through the text (a couple of examples are below, although there are more). 

• Section 2.3: Eq. 9, I believe there is a small error in the description (?) and it should say
  Where:  I_phosyn,hr   (it currently says I_physyn,hr)
• Section 3.2: line 6 - Should it say Figure 4 after "colour representation"? This section of the paragraph is still a little confusing (between "The analysis of chlorophyll a load in relationship..... and ..... as do the data with a large dispersion" lines 5-8)

Author Response

Thank you very much for pointing out the errors. They have been corrected and part of the paragraph edited. The text has been checked again for typing mistakes.

Reviewer 3 Report

Dear Author,
I accept the improvements made and I recommend the publication of an article entitled "Modelling chlorophyll a with use the SWAT tool for the Nielba River (West-Central Poland) as an example of unmonitored watercourse" in the Water journal. 

Author Response

I would like to thank you again for your valuable comments and for the opportunity to improve the manuscript.