Machine Learning and Hyperparameters Algorithms for Identifying Groundwater Aflaj Potential Mapping in Semi-Arid Ecosystems Using LiDAR, Sentinel-2, GIS Data, and Analysis

Round 1

Reviewer 1 Report

Please read the attached file. There are some improvements and clarifications to be made.

Comments for author File: Comments.pdf

Author Response

Ref: Remote Sensing-1911924

Title: Machine Learning and Hyperparameters Algorithms for Identifying Groundwater Aflaj Potential Mapping in Semi-Arid Ecosystems Using LiDAR, Sentinel-2, and GIS Data and Analysis

Thank you for consideration of our manuscript. Below we lay out our rebuttal of comments made by reviewers in red. We also have made other changes to the manuscript as outlined as track changes in the edited document. A ‘clean’ version of the updated manuscript is also uploaded to reflect these changes.

Reviewer # 1

Different models were used to predict groundwater aflaj potential. However, what is this potential expressed in m3/s or millions of cm/year? How the classification into five classes (deficient, low, moderate, high and very high) is useful for water managers? Where are located the water users in relation to the groundwater aflaj potential? New aflaj are foreseen in the future? How the water will be delivered to users who are far from the water source? What should effective management look like? (Lines 52-53: “….due to excessive water extraction rates and ineffective management…”). How could the flow of water from aflaj be controlled or stored for future use?

Thank you. Based on the reviewer’s comment, several sentences have been modified in the revised manuscript. Please see lines 13–14, lines 52–66 and lines 67–79.  

Dear reviewer “Aflaj is made up of a tunnel, possibly several kilometres long, that taps water where it is concentrated in the ground and brings it to the surface. The gradient must be less than that of the ground's surface or the water table beneath the ground. As a result, the tunnels' design must be very precise, and they must be maintained on a regular basis; however, they operate without the use of any mechanical devices”.

Line 70: “…can provide data analysis of water stored in small dams”). The water is stored not in dams but in reservoirs. Please, modify the sentence accordingly.

The sentence has been revised in the manuscript. Please see line 77.  

Lines 85-86: “However, if surface inflows dominate groundwater, watershed characteristics….”. The groundwater resources are replenished by natural recharge where there are favorable conditions. There is no other source of natural recharge than the surface inflows. There is no question that the inflows dominate the underground water, but that they influence the quality of the groundwater depending on the different factors shown further by the authors. Please, modify the sentence accordingly.

We agreed with this comment. The sentence has been deleted in the revised manuscript. See line 114.  

Lines 91-92: “The literature suggests a variety of efficient models for mapping groundwater potential around the world.” Among the available techniques, the physically based models are not even mentioned. Of course, they cannot be used in the specific conditions of the current study, but they must be remembered.

Thank you so much. We contend that we discussed the physically based model techniques in the previous manuscript (see line 484). However, in response to the reviewer's comments, a paragraph with relevant citations about the physically based model has been included in the revised manuscript. Please see lines 505–511.

Dear reviewer’s “Number of physically based models have also been used to solve groundwater issues. However, due to a lack of data, physically based distributed models are rarely thoroughly calibrated and validated. In practice, validation is limited to comparing simulated and predicted catchment discharges or simulated and observed piezometric levels in some calibrated wells. Internal noncalibrated wells or discharge stations are only rarely included in model evaluation”.

Lines 187-188: “Drainage density influences permeability and surface runoff.” and lines 189-190: “Higher drainage density values result in lower permeability and higher surface values”. In fact, a low hydraulic conductivity (not permeability) of the soil leads to erosion and then to the creation of a drainage network. Thus, the cause-effect relationship should be seen the other way around.

The entire paragraph has been modified and updated in the revised manuscript in response to the reviewer's comment.  Please see lines 195–201.

Lines 207-208: “Positive curvature represents a flat convex area with zero curvature, while negative curvature represents a concave area”. Not very clear. If the negative curvature represents a concave area, then a positive curvature should represent a convex area, while a zero curvature indicates a flat area.

The sentence has been modified in the revised manuscript in response to the reviewer's comment. Look at lines 220–222.

Lines 211-212: “…and categorised into three groups, including < (-0.001), (-0.001) and (0.001) “. In fact, and in accordance with Fig. the third classes are: <(-0.001), (-0.001; 0.001) and (>0.001).

Thank you so much. The sentence has been updated and modified in the revised manuscript. Please see lines 224-225.   

Line 224: “According to the IDW results, the average rainfall ranged from 68 mm–190 mm…”. Probably it is about annual precipitation.

Based on the reviewer’s comment, the “average rainfall” has been changed to “annual precipitation”. Please see the lines. Please see line 237.

Fig. 2: Fig. 2d does not contain the geological map but the drainage density, as Fig. 2e refers to the geological map and not to the drainage density. Finally, (m) represents the distance to the faults.

Done.  All Figures have been updated and uploaded to the revised manuscript.  See lines 245. 

No mention or explanation in the text regarding Fig. 3.

Thank you. See line 126.  

Line 302 and line 311: the correct numbering of the paragraphs should be 2.5.2 and 2.5.3 respectively

The numbering of the sub-sections has been corrected in the revised manuscript. Please see lines 313 and line 321.   

Line 484: “Although there are currently few physical models available…”. I assume that it is about physically based models.

Done. See line 499.  

Lines 503-505: “Thus, this study provides valuable techniques for water resource management, such as aflaj systems, to overcome water stress, primarily during periods of low rainfall.” Please, see comment no. 1 and justify this statement.

Thank you so much. Lines 503–505 have been deleted. However, instead, sentences have been added in the revised manuscript. See lines 522-527 

All the best

Alkindi  

Reviewer 2 Report

In this study, the support vector machine (SVM) and extreme gradient boosting (XGB) machine learning models were used to predict the white spot potential of Nizwa groundwater, and three hyper-parametric algorithms, grid search (GS), random search (RS), and Bayesian optimization, were used to optimize the parameters of the XGB model. This is very meaningful work, but I hope the following suggestions can be adopted by the author:

1. Research background and research significance need to be added.

2. The middle part of the abstract needs a concise description of what methods were used, what problems were solved, and what the results were; I am not sure if the understanding is correct, "The prediction of groundwater white-spot potential was achieved using Support Vector Machines (SVM) and Extreme Gradient Boosting (XGB), as well as three models optimized for the XGB model. The results show that the following models exhibit high accuracy based on the accuracy evaluation in the training phase...Therefore, this study concludes that... "The authors should pay attention to the logical relationships, and for the description of the method it is recommended to add in the method section, such as the use of Sentinel-2 satellite data, etc. Here The grading of the groundwater macular pressure potential map is introduced, then the results section should have the percentage of groundwater potential under different grades and different methods.

3. Line26-27:“The study concluded that remote sensing data, GIS-based machines, and deep learning are effective techniques for mapping and investigating groundwater aflaj in Oman’s watersheds.” This seems to be a common sense, and such a common sense as a conclusion is not optimal, I hope to revise this part and add the corresponding conclusion.

4. It is suggested that the paragraphs Line33-40 and Line51-54 be combined into one paragraph to illustrate the preciousness of groundwater resources and the need for effective management methods.

5. It is suggested to combine Line41-Line50 and Line55-Line91 to illustrate the current status of groundwater research. It is suggested to add the advantages and disadvantages of Aflaj method and the Research Gap of this paper.

6. Line118-Line149: The study area is mainly about the characteristics of the study area (human and natural features) and why it was chosen as the study area. Line120-Line140 about the Aflaj system is not suitable here, and if it has to be placed here, it should be about the characteristics of the Aflaj system as expressed in the study area.

7. Line50:Figure1 latitude and longitude are not shown in the figure, suggest to modify and beautify Figure1.

8. The method section should explain how the information such as slope calculated in the geographic method is used and how this information is related to the prediction methods SVM and XGB.

9. How is the efficiency of the five algorithms quantified? The text should be clearly explained in the methods section.

Author Response

Reviewer # 2

Comments and Suggestions for Authors

In this study, the support vector machine (SVM) and extreme gradient boosting (XGB) machine learning models were used to predict the white spot potential of Nizwa groundwater, and three hyper-parametric algorithms, grid search (GS), random search (RS), and Bayesian optimization, were used to optimize the parameters of the XGB model. This is very meaningful work, but I hope the following suggestions can be adopted by the author:

Thank you so much.  We greatly appreciate it. We wholeheartedly concur with this assertion.

• Research background and research significance need to be added.

This comment is quite similar to one of reviewer #1’s comments about adding more information about background and research significant to the current manuscript, but we feel the inclusion of more variables in this paper will make it too long.

Although such information on the background and research significance were mentioned in the manuscript, several sentences about the background and research significance also have been added in the updated version.

 Please See lines 127–144. We highly recommend also reading lines 88–102.

Several spatial studies have evaluated groundwater aflaj potential in Oman [25-27]. For example, GIS and remote sensing were used to map groundwater potential in Wadi Al‑Jizi in northern Oman [27]. However, that study used slope, soil, geomorphology, LULC, and geology. At the same time, many critical factors, such as distance to drainage, elevation, topographic wetness index (TWI), drainage density, distance to faults, stream length, rainfall, and fault density, were not included. Despite extensive studies on the threats to aflaj irrigation systems [28,29], little has been done to identify the current distribution of aflaj potential groundwater, and there have been no concentrated efforts on mapping groundwater aflaj potential in the country. Periodic studies to monitor changes in groundwater levels have been conducted in a few locations in Oman. However, the data were not current, the studies were small in scale, and the results did not represent an accurate picture of the current conditions of groundwater aquifers.

Thus, in this study, five robust ML methods and hyperparameter algorithms, including SVM, extreme gradient boosting (XGB), extreme gradient boosting using random search (RS-XGB), extreme gradient boosting using grid search (GS-XGB), and extreme gradient boosting using Bayesian optimization (XGBO), were applied to model and map groundwater aflaj potential in the Nizwa watersheds. Although the Bayesian optimization methodology has been extensively applied to modeling flooding, this is the first study to use it to assess groundwater potential.

All the above sentences are mentioned in the last version of the manuscript. A few sentences have been added to the revised manuscript. 

The aflaj systems in the Sultanate of Oman have been under threat due to groundwater pumping and economic development since 1970. The current study would assess existing groundwater datasets, facilities, and future spatial datasets in order to design systems that could map groundwater aflaj using geospatial and ML techniques. In turn, this will assist groundwater protection service projects and integrated water source management (IWSM) programs in protecting the aflaj irrigation system from threats by adopting timely preventative measures. In addition, geospatial approaches will be used to extract novel information regarding spatial patterns, spatial correlations, and other groundwater aflaj-related potential parameters, including geology, hydrology, soil, land use, land cover, and water setting. These variables can assist authorities in the Sultanate of Oman in better understanding, managing, and controlling the groundwater in the Nizwa watershed.

Thus, advanced remote sensing and GIS-based machine, and deep learning are needed to be developed and implemented to map and control the groundwater on a large spatial scale. The mapping of groundwater aflaj potential can aid in the development of more effective management techniques for their control. In addition, mapping is necessary for developing predictive models that offer information on the likelihood of occurrence, spatial distribution, and density under various environmental variables. These updated maps will aid IWSM initiatives in educating and empowering authorities and organizations concerned with groundwater quality. Spatial modelling will also help reduce costs, like GIS and remote sensing-based methods developed to pursue this research promise more practical and cost-effective solutions. In addition, this research will save money on monitoring since the remote sensing-based technologies developed in this project will give a more efficient and cost-effective way to monitor water resources on a broad scale in Oman.

Please see lines 522-535 in the revised manuscript.

2. The middle part of the abstract needs a concise description of what methods were used, what problems were solved, and what the results were; I am not sure if the understanding is correct, "The prediction of groundwater white-spot potential was achieved using Support Vector Machines (SVM) and Extreme Gradient Boosting (XGB), as well as three models optimized for the XGB model. The results show that the following models exhibit high accuracy based on the accuracy evaluation in the training phase...Therefore, this study concludes that... "The authors should pay attention to the logical relationships, and for the description of the method it is recommended to add in the method section, such as the use of Sentinel-2 satellite data, etc. Here The grading of the groundwater macular pressure potential map is introduced, then the results section should have the percentage of groundwater potential under different grades and different methods.

Thank you so much, based on the reviewer’s comment, the whole sections have been updated. The mythology sections and sub-sections have been modified in the revised manuscript.  

3. Line26-27: “The study concluded that remote sensing data, GIS-based machines, and deep learning are effective techniques for mapping and investigating groundwater aflaj in Oman’s watersheds.” This seems to be common sense, and such a common-sense conclusion is not optimal, I hope to revise this part and add the corresponding conclusion.

Based on the reviewer’s comment, the sentence has been revised in the manuscript. See lines 27–30 in the abstract section and lines 522–532 in the conclusion section. 

Dear reviewer

 “The study concluded that evaluating existing groundwater datasets, facilities, and future spatial datasets is critical in order to design systems capable of mapping groundwater aflaj using geospatial and ML techniques. In turn, groundwater protection service projects and integrated water source management (IWSM) programs will be able to protect the aflaj irrigation system from threats by implementing timely preventative measures”.

4. It is suggested that the paragraphs Line33-40 and Line51-54 be combined into one paragraph to illustrate the preciousness of groundwater resources and the need for effective management methods.

As suggested by a reviewer, paragraphs lines 33-40 and 51-54 have been combined into one paragraph to demonstrate the value of groundwater resources and the need for effective management methods.

5. It is suggested to combine Line41-Line50 and Line55-Line91 to illustrate the status of groundwater research. It is suggested to add the advantages and disadvantages of Aflaj methods and the Research Gap of this paper.

Lines 41-50 and 55-line have been combined to show the current state of groundwater research. As a result, sentences describing the benefits and drawbacks of Aflaj methods, as well as the research gap, have been added to the revised manuscript.

6. Line118-Line149: The study area is mainly about the characteristics of the study area (human and natural features) and why it was chosen as the study area. Line120-Line140 about the Aflaj system is not suitable here, and if it has to be placed here, it should be about the characteristics of the Aflaj system as expressed in the study area.

Thank you some much. We totally agree that the study area is mainly about the characteristics of the study area. Based on the reviewer’s comment, Lines 120– line 140 have shifted to the interaction section.  Please see lines 55–87.

7. Line50:Figure1 latitude and longitude are not shown in the figure, suggest modifying and beautifying Figure1.

Thank you so much. Figure 1 has been updated and uploaded to the revised manuscript. 

8. The method section should explain how the information such as slope calculated in the geographic method is used and how this information is related to the prediction methods SVM and XGB.

The SVM, XGB, RS-XGB, GS-XGB, and XGBO models were applied to all aflaj factors using a raster grid. Finally, all leading aflaj factors were converted to a raster grid with 5 × 5 m cells. Please see line 284–292 and lines 294–299.

9. How is the efficiency of the five algorithms quantified? The text should be clearly explained in the methods section.

Thank you so much, section about the efficiency of the five algorithms has been added in the revised manuscript. See line 333–340.  

Thank you so much

Alkindi 

Author Response File: Author Response.pdf

Reviewer 3 Report

This manuscript, remotesensing-1911924-peer-review-v1- entitled "Machine Learning and Hyperparameters Algorithms for Identifying Groundwater Aflaj Potential Mapping in Semi-Arid ecosystems Using LiDAR, Sentinel-2, and GIS Data and Analysis," is well written and has potential, but it should be more organized.

In my opinion, a careful revision of the English language should be carried out as there currently are some unclear sentences. The study seems to be well designed. The methodology and results are technically sound. Discussions on the scientific and practical values of the study, the limitations of proposed models, and future work are meaningful. I recommend accepting this manuscript after revision. The main concerns are as follows:

1)     Aflaj irrigation systems are adopted as the case study. What are other feasible alternatives? What are the advantages of adopting this case study over others in this case? How will this affect the results? The authors should provide more details on this.

2)     It is better to clarify the difference between Qanat and Aflaj in the introduction section.

3)     More literature review about the other methods is needed. The manuscript could be substantially improved by relying and citing more on recent literature about contemporary real-life case studies of sustainability and/or uncertainty, such as the followings.

Vadiati, M., Rajabi Yami, Z., Eskandari, E., Nakhaei, M., & Kisi, O. (2022). Application of artificial intelligence models for prediction of groundwater level fluctuations: case study (Tehran-Karaj alluvial aquifer). Environmental Monitoring and Assessment, 194(9), 1-21.

4)     Please provide all software used in this study.

5)     Tab. 4 is the an important table in the manuscript, and, unfortunately, the authors did not try to discuss it in a specific way. A comprehensive discussion emphasizing would significantly improve the paper on the table.

6)     It is important to give a better description of the samples and the sampling protocol since we are trying to understand the data variability. What are the advantages of adopting these parameters over others in this case? How will this affect the results? More details should be furnished.

7)     It is better to add more error criteria to better understand the model's ability.

8)     It seems that conclusions are observations only, and the manuscript needs thorough checking for explanations given for results. The authors should interpret more precisely the results argument.

Author Response

Ref: Remote Sensing-1911924

Title: Machine Learning and Hyperparameters Algorithms for Identifying Groundwater Aflaj Potential Mapping in Semi-Arid Ecosystems Using LiDAR, Sentinel-2, and GIS Data and Analysis

Thank you for consideration of our manuscript. Below we lay out our rebuttal of comments made by reviewers in red. We also have made other changes to the manuscript as outlined as track changes in the edited document. A ‘clean’ version of the updated manuscript is also uploaded to reflect these changes.

Reviewer #3

This manuscript, remote-sensing-1911924-peer-review-v1- entitled "Machine Learning and Hyperparameters Algorithms for Identifying Groundwater Aflaj Potential Mapping in Semi-Arid Ecosystems Using LiDAR, Sentinel-2, and GIS Data and Analysis," is well written and has potential, but it should be more organized.

Thank you so much. The revised manuscript has been updated and now it is more organized. 

In my opinion, a careful revision of the English language should be carried out as there currently are some unclear sentences. The study seems to be well designed. The methodology and results are technically sound. Discussions on the scientific and practical values of the study, the limitations of proposed models, and future work are meaningful. I recommend accepting this manuscript after revision.

Thank you so much. We have performed a careful proofreading of the entire manuscript.  

The main concerns are as follows:

Aflaj irrigation systems are adopted as the case study. What are other feasible alternatives? What are the advantages of adopting this case study over others in this case? How will this affect the results? The authors should provide more details on this.

Thank you so much. Sentences have been added in the revised manuscript. 

It is better to clarify the difference between Qantas and Aflaj in the introduction section.

Done.

3)     More literature review about the other methods is needed. The manuscript could be substantially improved by relying and citing more on recent literature about contemporary real-life case studies of sustainability and/or uncertainty, such as the followings.

Vadiati, M., Rajabi Yami, Z., Eskandari, E., Nakhaei, M., & Kisi, O. (2022). Application of artificial intelligence models for prediction of groundwater level fluctuations: case study (Tehran-Karaj alluvial aquifer). Environmental Monitoring and Assessment, 194(9), 1-21.

Followed by the reviser’s comment, the above citation has been added in the revised manuscript. See line 465.    

4)     Please provide all software used in this study.

Thank you. Python, ArcGIS Pro 2.8, and ArcGIS Desktop were used in our research. All of them are already mentioned in the manuscript.   

5)    Tab. 4 is an important table in the manuscript, and, unfortunately, the authors did not try to discuss it in a specific way. A comprehensive discussion emphasizing would significantly improve the paper on the table.

We acknowledge this reviewer’s comment that the table 4 should be discuss it in a specific way; however, in this manuscript we tried discuss the most important variables to map the aflaj potential in the study areas we feel the inclusion of this analysis in the current manuscript will make it too long.

6)     It is important to give a better description of the samples and the sampling protocol since we are trying to understand the data variability. What are the advantages of adopting these parameters over others in this case? How will this affect the results? More details should be furnished.

About samples the method and how data was divided for modelling was described.

To model the aflaj groundwater potential, aflaj and non-aflaj points were generated in ArcGIS 10.8 using random point extensions. From 336 locations for modelled, 168 were aflaj locations, and 168 were non-aflaj locations. The datasets were divided into training (135 or 70%) and validation (101 or 30%) randomly.

About using hyperparameters algorithms for optimization and effect of these methods on results of model the optimize parameters of model and the results of optimization have been shown in the result section and we discussed ability of these algorithms in increasing accuracy.

7)     It is better to add more error criteria to better understand the model's ability.

Dear reviewer, the ROC AUC score are important tools to evaluate binary classification models. In summary they show us the separability of the classes by all possible thresholds, or in other words, how well the model is classifying each class. In this study the classification models were used for modelling groundwater afalj system, the most important criteria for validation of classification models include ROC parameters: Sensitivity, Specificity, NPV, PPV and AUC that other researchers have mentioned these criteria in their studies.

Chen, Y., Chen, W., Chandra Pal, S., Saha, A., Chowdhuri, I., Adeli, B., Janizadeh, S., Dineva, A.A., Wang, X. and Mosavi, A., 2022. Evaluation efficiency of hybrid deep learning algorithms with neural network decision tree and boosting methods for predicting groundwater potential. Geocarto International, 37(19), pp.5564-5584.

Naghibi, S.A., Ahmadi, K. and Daneshi, A., 2017. Application of support vector machine, random forest, and genetic algorithm optimized random forest models in groundwater potential mapping. Water Resources Management, 31(9), pp.2761-2775.

Rizeei, H.M., Pradhan, B., Saharkhiz, M.A. and Lee, S., 2019. Groundwater aquifer potential modeling using an ensemble multi-adoptive boosting logistic regression technique. Journal of Hydrology, 579, p.124172.

8)  It seems that conclusions are observations only, and the manuscript needs thorough checking for explanations given for results. The authors should interpret more precisely the results argument.

Dear reviewer, as you mentioned the arguments of results and also the application of results in the case study were added, please see lines 531-546.

Thank you so much 

Alkindi

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Please, read the attached file. There are some minor improvements to be made.

Comments for author File: Comments.pdf

Author Response

Ref: Remote Sensing-1911924

Title: Machine Learning and Hyperparameters Algorithms for Identifying Groundwater Aflaj Potential Mapping in Semi-Arid Ecosystems Using LiDAR, Sentinel-2, and GIS Data and Analysis

Thank you for consideration of our manuscript. Below we lay out our rebuttal of comments made by reviewers in red. We also have made other changes to the manuscript as outlined as track changes in the edited document. A ‘clean’ version of the updated manuscript is also uploaded to reflect these changes.

Additional comments on Aflaj potential

Lines 76-77

Therefore, assessing groundwater aflaj potential is critical for maintaining groundwater reservoirs, especially in

data-scarce areas

I suggest: Therefore, assessing groundwater aflaj potential is critical for maintaining groundwater resources,

especially in data-scarce areas

Thank you so much

Based on the reviewer’s comment the “reservoirs” has been changed to “resources” in the revised manuscript.

Lines 224-225

……including < (-0.001), (-0.001) and > (0.001) (Figure 2i).

It should be written:

…….including < (-0.001), (-0.001; 0.001) and > (0.001) (Figure 2i).

Thank you so much.

Based on the reviewer’s comment the sentence has been modified in the manuscript. Figure 2i has been updated and uploaded to the revised manuscript.  

Fig 2 (e) Geology map was not written in the box 2 (e)

 Thank you so much.

Figure 2 has been updated and uploaded to the revised manuscript.

Lines 498-504 I disagree with the following paragraph.

“Although there are currently few physically based models available, it is possible to enhance groundwater models by quantifying the variable influence on hydrologic parameters. However, due to a lack of data, physically based distributed models are rarely thoroughly calibrated and validated. In practice, validation is limited to comparing simulated and predicted catchment discharges or simulated and observed piezometric levels in some calibrated wells. Internal noncalibrated wells or discharge stations are only rarely included in model evaluation.”

- There are many physically based models, like Modflow or Mike She to cite only the most used of them.

- Generally, calibration and validation are based on registered piezometric levels. Discharges usually represent boundary conditions The best would be to delete completely this paragraph and to add a sentence related to the physically based models after the line 113 (in yellow is what I suggest introducing in your text): The literature suggests a variety of efficient models for mapping groundwater potential around the world. If good quality data about the aquifer are available the best option is to use physically based models, where the groundwater flow equation is obtained by combining the Darcy law and the balance equation. On the contrary, in the absence of the necessary information, the data-driven models are a valuable alternative for investigating groundwater resources. Available techniques include logistic regression……..

Thank you so much. Based on the reviewer's suggestion, sentences lines 498-504 (discussion section) have been completely removed from the revised manuscript, and the reviewer's suggested sentences related to physically based models have been added. Please see lines 120–124.

Reviewer 2 Report

All of my previous suggestions are reflected in the text, and we accept the additions. I have those suggestions before publication:

1. We still feel that the abstract should be a bit more abbreviated, and some descriptions of the methods are suggested to be placed in the methods section. For example, Line 15-26 is really lengthy.

2.the current introduction section only elaborates on the human significance of groundwater and ignores the natural ecological significance of groundwater, we suggest adding the natural significance of groundwater to the introduction section, consider adding and summarizing the contents of the following documents.

1.Yang, Kun, et al. "Spatial‐temporal variation of lake surface water temperature and its driving factors in Yunnan‐Guizhou Plateau." Water Resources Research 55.6 (2019): 4688-4703.

Author Response

Ref: Remote Sensing-1911924

Title: Machine Learning and Hyperparameters Algorithms for Identifying Groundwater Aflaj Potential Mapping in Semi-Arid Ecosystems Using LiDAR, Sentinel-2, and GIS Data and Analysis

Thank you for consideration of our manuscript. Below we lay out our rebuttal of comments made by reviewers in red. We also have made other changes to the manuscript as outlined as track changes in the edited document. A ‘clean’ version of the updated manuscript is also uploaded to reflect these changes.

All of my previous suggestions are reflected in the text, and we accept the additions. I have those suggestions before publication:

1. We still feel that the abstract should be a bit more abbreviated, and some descriptions of the methods are suggested to be placed in the methods section. For example, Line 15-26 is really lengthy.

Thank you incredibly much. The abstract section has been condensed in response to the reviewer's comments. We don't believe we need to replace some of the methods in the methodology section because we already mentioned all of this information.

2. The current introduction section only elaborates on the human significance of groundwater and ignores the natural ecological significance of groundwater; we suggest adding the natural significance of groundwater to the introduction section and consider adding and summarizing the contents of the following documents.

1.Yang, Kun, et al. "Spatial‐temporal variation of lake surface water temperature and its driving factors in Yunnan‐Guizhou Plateau." Water Resources Research 55.6 (2019): 4688-4703.

 Based on the reviewer’s comment, information about the natural ecological significance of groundwater have been added to the revised manuscript. See lines 75–82. Citation (the above citation) has also added to the revised manuscript. See line 82.

“Groundwater, which flows through the soil and protects the water level in rivers, lakes, and wetlands, is especially important during dry seasons when direct rainwater recharge is low. This helps to preserve wildlife and plants, and its role in keeping water levels stable during dry seasons helps to keep marine travel moving along inland waters and rivers. Water is stored in deeper layers beneath the earth's surface, which preserves its quality and protects it from pollution, making it suitable for direct consumption without high extraction or treatment costs, but it is critical to preserve this vital importance due to depletion or pollution”.

Reviewer 3 Report

Accept in present form

Author Response

Thank you very much