Comparison between MODFLOW Groundwater Modeling with Traditional and Distributed Recharge

Round 1

Reviewer 1 Report

This paper presented a solution to the recharge distribution using the PGR map, and verified the results with numerical models. The starting point of the study is clear, but the proof is not believable enough. Some comments are as follows.

(1) The PGR map is obtained from weighting spatially distribution information, and the explanation is lacking. First of all, the factors may be not independent, especially, hydrogeology. Second, because each factor is time-dependent, the consideration of processing time-series data is not discussed. Last, the point-scale field observation or experiment should be used to verify the results.

(3) The numerical model is not well demonstrated. From Fig.10, the results are not improved for traditional and PGR method. No observation well data is compared. Only the basic step is introduced, and it looks like an exercise. The results may be also influenced by parameter uncertainty. However, no further discussion about parameter is carried out.

(4) The writing should be improved greatly. Many paragraphs are consisted of a sentence.

This paper presented a solution to the recharge distribution using the PGR map, and verified the results with numerical models. The starting point of the study is clear, but the proof is not believable enough. Some comments are as follows.

(1) The PGR map is obtained from weighting spatially distribution information, and the explanation is lacking. First of all, the factors may be not independent, especially, hydrogeology. Second, because each factor is time-dependent, the consideration of processing time-series data is not discussed. Last, the point-scale field observation or experiment should be used to verify the results.

(3) The numerical model is not well demonstrated. From Fig.10, the results are not improved for traditional and PGR method. No observation well data is compared. Only the basic step is introduced, and it looks like an exercise. The results may be also influenced by parameter uncertainty. However, no further discussion about parameter is carried out.

(4) The writing should be improved greatly. Many paragraphs are consisted of a sentence.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

hydrology-2724525 “Comparison between MODFLOW subsurface modeling with 2 traditional and distributed recharge” María del Mar Navarro-Farfán, Liliana García-Romero, Marco Antonio Martínez-Cinco, Mario Alberto Hernández-Hernández and Sonia Tatiana Sánchez-Quispe

This is very basic material that reads like a chapter from a Masters thesis. Upon reading the “Data Availability” statement it was confirmed that this was part of a university level project. As such this should not be published in an international journal without something more substantial.

Here are some tips and suggestions for the student on their PhD work.

1 – If you run a groundwater model, please supply a water balance table, either annually, or as a whole model average. We will then know how much recharge is being applied, well extractions, boundary fluxes, any lake interactions, etc. It would also be good to show maps of the optimised values of conductivity and storage coefficient for both the constant and PGR recharge.

2 – I think it’s implied that the total recharge as a constant or distributed by PGR is the same but this needs to be explicit. Also show a map of the distribution of relative (or absolute) recharge, rather than the PGR value in Figure 6. Did you use PGR as a linear scalar on recharge, or experiment with other functions, such as logarithmic, polynomial or exponential, as the range in recharge is likely to be an order of magnitude or more, certainly more than a factor of three across your area.

3 – In Figure 6 you should make Lake Cuitzeo a gray colour as it mixes with the upper limit of the PGR, unless the presence of the lake is a recharge factor and then needs to be described.

4- Figures 7c, 8 and 9a,b all show piezometric level yet all have unique colours and water level scales. It is much clearer to have them all use the same set of scales at least, and similar colours if it is convenient. This allows the reader to get a visual idea of where differences occur, seasonality, and the groundwater recovery mentioned in the text.

5 – On the subject of difference, water level difference maps are a very useful tool. You can use these at a point in time, say 2017 in the middle of the simulation, to have the constant recharge results of piezometric levels and then a level difference map with the PGR distributed recharge results. This would illustrate how the distribution affects water levels that are generally similar. Further, for each of the 10 wells indicated in Table 2, on the constant recharge water level map each well can be shown with a circle that is coloured with the water level difference from the observed value. This would also show if PGR distributed recharge altered levels locally that are more in line with observation.

6 – You have a conclusion that says the difference in water levels using either uniform recharge or PGR distributed recharge is very small. This does not validate, or invalidate, either approach, but may indicate that the system is so tightly controlled by the high conductivity areas and the large lake, nearly 11% of all model cells, that only total recharge is important to reproduce water levels.

English language is generally good, but with some small grammar issues that do not significantly reduce the readability.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Groundwater resources are one of the most important elements of water management. One method of quantitative assessment of groundwater resources is based on the determination of groundwater recharge. Therefore, the research problem formulated in this article is of fundamental importance for water management in the country. The authors used PRG analysis and a numerical groundwater flow model to determine groundwater recharge. The topics of the article are in line with the scope of the journal Hydrology.

1) I propose to discuss the title of the article again. The phrase "subsurface modelling" does not accurately describe the groundwater flow model done in MODFLOW.

2) p. 2 v. 55-56 - I agree with the authors that assuming a constant recharge value is used in groundwater flow modelling, but this is a major simplification and not a traditional approach. There are numerous publications on reliable determination of groundwater recharge e.g. doi.org/10.3390/w15112112, doi.org/10.1007/978-981-13-7067-0_10

3) The description of the model, the discetization etc.  should be included in the Materials and Methods chapter.  In this chapter also, I suggest to include a detailed description of the PGR method, for example, as in doi.org/10.3390/land9100364.

4) The complete lack of information on the units in which the data are expressed (e.g. Table 1), the lack of description of the structure of the modelled system, the lack of description of the boundary conditions adopted on the model, the lack of a calibration graph of the model makes it impossible to assess the reliability of the results obtained. Please complete this information.

5) The Authors highlight that the study area is poorly recognised in terms of hydrogeology.  Chapter 4 lacks a detailed discussion on the impact of uncertainty in the determination of the hydraulic conductivity on the obtained groundwater recharge results. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

Dear Authors,

I have thoroughly reviewed your paper comparing traditional and PGR maps' recharge options while calibrating the MODFLOW, and I appreciate your research efforts. However, I would like to offer some suggestions for improving the overall quality of your work.

Introduction

The problem within the research area lacks clarity. It would be beneficial to explicitly outline the knowledge gap and highlight the contribution your research aims to make.

Literature Review

The introduction and overall discussion sections lack a thorough discussion of previous research within this modeling domain. Consider revising the introduction to better engage readers with the existing body of knowledge.

Methodology

The methodology section does not sufficiently explain the steps taken to assign weighted scales for developing PGR Maps. Clarify the methodology adopted and provide reasoning behind the assignment of weighted scales.

Results

The description of Table 2 is unclear, particularly regarding the upper column values (1, 2, 3, ...). Additionally, it would be beneficial to elaborate on the piezometric value and error value in the table.

Spatial Consideration

Consider incorporating piezometric locations on the map and assessing whether the Piezometer well location density influences model calibration errors in a spatially geographic manner. Refer to existing work in this area for guidance on how piezometer well density may impact the calibration process. Here is the link for your reference; https://www.mdpi.com/2076-3417/10/15/5200 

I believe these enhancements will further strengthen the clarity and impact of your research. Thank you for considering these suggestions, and I look forward to seeing the continued development of your work

Dear Authors,

Your paragraph sentence length is too short and it does not look impressive. Please improve and revise your writing style and consider merging the relevant sub-topics theme into one paragraph. Thanks   

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 5 Report

the paper is described about the research. the contents of research process is not enough. if possible,  please add some detailed research process. 

no 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The writing of introduction can be improved because it is hard to understand. Several sentences can be merged into a paragraph.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

hydrology-2724525-v2 “Comparison between MODFLOW subsurface modeling with 2 traditional and distributed recharge” María del Mar Navarro-Farfán, Liliana García-Romero, Marco Antonio Martínez-Cinco, Mario Alberto Hernández-Hernández and Sonia Tatiana Sánchez-Quispe

The authors have taken the suggestions very seriously and added significant data in tables and figures as suggested. It has made understanding their method and results much easier.

Revisiting the last round of suggestions:

1 – Thank you for including a water balance table (Table 1) and figures of optimised parameters. I suggest you change Table 1 entries to “ML/year” which means dividing each entry by 10, and then using only 1 decimal place for the values. It is important to recognise the accuracy of a model such as MODFLOW, and implying that it has the power to discriminate 1 part in 10,000 (e.g., 169.67) is incorrect. Note also that columns rounded off this way do not have to add up perfectly. I’m not sure of the value associated with the label “Subway Entrance”: is it inputs from the constant head boundary (lake) or loss from storage?

2 – Thank you for clarifying that PGR is a linear scalar for recharge.

3 – Thank you for changing Figure 6 and showing the lake explicitly.

4, 5, 6 - Thank you for showing piezometric head maps with the same colour scales. This map, along with the figure showing the position of fitting wells illustrates the issues with your study, and many in general. The wells used are all in the lowest head category and all close to the lake (probably flatter land and where population centres and roads are concentrated), which has a fixed head boundary condition. This means that their level is quite insensitive to recharge changes or aquifer state parameters except if they are very radically different locally. Most of what happen upslope to the west and south is almost immaterial as the local heads are controlled by the fixed head in the lake. This is most obvious in Table 4 where it is difficult to find any significant differences.

I would go a little further with Table 4 using only 2 decimal places for entries, and then showing the PGR results as differences from the values using “traditional” recharge (with 3 decimals as required). This is to emphasise where changes are observable.

An additional conclusion is that while your method may have merit, the combination of a strong fixed head boundary near all the available fitting wells makes the modelling insensitive to most other input parameters.

Please be careful with units and English expression in the Table captions. For example, Table 1 has units of hm3/ano. A better unit would be ML, which means dividing each of the entries by 10, and using the English word "year".

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Dear Authors,

Thank you for improving the article according to my comments. I have a few more comments on the changes made.

 1) Tab. 1 - Please write the units in English. 

2) Fig. 8 - The units in which hydraulic conductivity and storage coefficient are given are missing.

3) No calibration graph of Modflow to verify the model calculations done.

4) Tab. 2 - Studies by Sharm et al. (1983) and Fox et al. (1997) found that for slopes greater than 10o the slope has no effect on effective infiltration. In my opinion, the introduction of class 5 is not necessary and class 4 should be in the range > 10.

Sharma K.D., Singh H.P., Pareek O.P. 1983 Rainwater infiltration into a bare loamy sand. Hydrological Sciences Journal 28 (3): 417-424.

Fox D.M., Bryan R.B., Price A.G. 1997 The influence of slope angle on final infiltration rate for interrill conditions. Geoderma 80 (1-2): 181-194.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 4 Report

Dear Authors,

Thanks for responding to my comments. The incorporation of the comments does not improve the quality of the paper too much. 

The English language persists the same issue. Please revise it carefully.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 5 Report

Thank you for incorporating the changes based on my previous feedback. While reviewing the paper, I noticed that the language segmentation throughout the manuscript appears somewhat fragmented, and there may be room for improvement in terms of logical flow. Here are some suggestions to enhance the overall coherence:

1. Introduction: Clearly emphasize the background and objectives in the introduction to provide readers with a robust understanding of the study's motivation and significance. Explicitly state the unique contribution of this paper in terms of innovation, highlighting what sets it apart from existing work.

2. Methods Section: Reorganize the methodology section systematically, detailing the steps and processes of the research. Clearly explain the models, parameters, and techniques used, and emphasize any novel or innovative aspects introduced in the approach.

3. Results Presentation: Organize the results section thematically or by experimental stages to facilitate a gradual understanding of the study's findings. Utilize clear graphs and tables to succinctly present the results, emphasizing any novel insights derived from the data.

4. Discussion and Conclusion: In the discussion, provide in-depth explanations of the research results and compare them with existing literature, underscoring the distinctive contributions of this paper. In the conclusion, reiterate the key innovations and summarily highlight the paper's unique insights.

These suggestions aim to assist in improving the overall structure and logical coherence of the paper. It is crucial to underscore the unique contributions and innovations introduced in this work to distinguish it from existing research.

Author Response

Please see the attachment.

Author Response File: Author Response.docx