Application of GALDIT in Assessing the Seawater Intrusion Vulnerability of Jeju Island, South Korea
Round 1
Reviewer 1 Report
The paper presents a well-established vulnerability assessment approach to seawater intrusion phenomenon in Jeru Island in South Korea. The paper is well structured and organized, presenting in great details the GALDIT approach and the application to Jeru coastal aquifer.
However, the results obtained from the vulnerability assessment are lacking scientific interest in the research community worldwide. What are the novelty items of this research?
It is important for the authors to show the contribution of their analysis and the impact of their results besides the boundary of Jeru Island.
GALDIT methodology is well known in the literature, a comparative analysis of the proposed approach to assess seawater intrusion phenomenon impacts to other coastal areas will enrich the robustness of GALDIT approach and improve the scientific impact of the paper.
Also, a sensitivity analysis should be performed to better justify the classification of all parameters used in the analysis.
Author Response
1. The paper presents a well-established vulnerability assessment approach to seawater intrusion phenomenon in Jeru Island in South Korea. The paper is well structured and organized, presenting in great details the GALDIT approach and the application to Jeru coastal aquifer.
However, the results obtained from the vulnerability assessment are lacking scientific interest in the research community worldwide. What are the novelty items of this research?
Answer)Jeju is a relatively large volcanic island that has hydrogeological heterogeneity. A vulnerability assessment of the whole area, which compares different areas of the island is needed by the government and society to support coastal water resource management. We believe that this study serves as a good example for the application of GALDIT to volcanic islands. We considered groundwater levels falling below sea level, which is strong evidence of increasing vulnerability to anthropogenic activities. To improve the GALDIT assessment method to fit the Jeju model, we explored the possibility of electrical conductivity instead of standard parameters representing the existing impact of saltwater intrusion (SWI). The conversion between three different parameters representing electrical conductivity is the first to use monitored (or collected) data in Korea and could be widely used for other vulnerability studies, including numerical model simulations, as well as index-based approaches. Finally, this study used daily and monthly continuous monitoring data in the GALDIT model. The resulting temporal changes in GALDIT assessment results (Figure 7) enable stakeholders to easily understand variations in the vulnerability.
2. It is important for the authors to show the contribution of their analysis and the impact of their results besides the boundary of Jeru Island.
Answer) Thank you for this helpful suggestion. We have included a comparison with El-Kadi’s (2014) results to highlight the contribution of our study:“The previous study by El-Kadi (2014) was the first numerical study for the whole of Jeju Island to assess the sustainability of the water resources that compared the overall groundwater levels based on simulated and observed data. These models showed a good match when the groundwater level was higher than 2 m, which was far from the coastal regions. However, discrepancy occurred between observed and simulated results when the groundwater level was less than 2 m. In Jeju Island, most saltwater intrusion occurs where groundwater levels are less than 2 m, which comprises a large proportion of the area. Moreover, in the GALDIT model, 2 m is an important level for determining the risk of saltwater intrusion. Unfortunately, El-Kadi (2014) only modeled the saline distribution for the eastern Jeju watershed and Jung et al. (2010) only considered low levels of ion concentration for springs in coastal areas.”
3. GALDIT methodology is well known in the literature, a comparative analysis of the proposed approach to assess seawater intrusion phenomenon impacts to other coastal areas will enrich the robustness of GALDIT approach and improve the scientific impact of the paper.
Anwer) We agree that additional comparative analyses would enrich the robustness of the GALDIT approach and improve the scientific impact of the paper. Previous GALDIT studies employed more than two index-based models such as DRASTIC, AHP, etc. to enrich the scientific insight of their studies. Therefore we have included the following text in the revised manuscript at line 70-74:
“Some of the common methods used for classification and mapping of groundwater vulnerability zones include DRASTIC [16], GALDIT [17], SINTACS [18], AVI [19],etc. With the exception of GALDIT, the methods were developed from the concept that pollutants originate from the land surface and are transported through the vadose zone with infiltrated water. Thus, they were mostly used for noncoastal aquifers.”
4. Also, a sensitivity analysis should be performed to better justify the classification of all parameters used in the analysis.
In some previous studies, a sensitivity analysis was implemented to evaluate the impact of subjective ratings and weights (mostly weights) of GALDIT parameters on the output vulnerability map, and modified weights were computed based on the specific field conditions in the area under investigation (Kura et al., 2015; Luoma et al., 2016). For example, Kura et al. (2015) conducted sensitivity analyses for DRASTIC and GALDIT weighting factors to modify models for contamination from anthropogenic activities and seawater intrusion in Kapas Island. Based on these effects and variation indices of the said parameters, new effective weights were determined and were used to create modified DRASTIC and GALDIT models. However, in this study, the weighting factors were not considered to be modified and only the range of the parameter values were modified to reflect the severity of the massive pumping due to anthropogenic activities.
In our study, the weightings were not considered to be modified. We believe that weight modification would be possible to conduct a sensitivity analysis for Jeju Island is future studies.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The objectives of the manuscript are the assessment of the vulnerability to seawater intrusion of coastal aquifer(s?) at Jeju Island, and the improvement of the techniques of the used model.
The manuscript is not clear about the aquifers relationships and settings. Initially, the authors roughly refer to an unconfined aquifer and a confined ones [111-112 lines]. After, they reveal " the study was on the unconfined aquifers of Jeju" [195 line]. Please, right away declare the aquifer (or the aquifers) you focused in the study. However the hydrogeological features of the Jeju Island must be described in 2.1. Site Description. About the "unconfined", how many aquifers are placed at your study site? Again, are the unconfined acquifers all in coastal setting?
Please, explain how you have obtained the correlations in Figure 3. Where came from the molar values of chloride and bicarbonates ions? If they are literature data ([30]?), write the Reference(s) in the caption.
The relevance of the hydrogeological structure of the island seems to be neglected. Thus data coming from "newly installed monitoring wells" [305 line] determine changes in the assessment. Moreover, it seems that the new wells have intercepted impermeable bottoms rather than seawater, because "Areas that were forcibly assigned a value of 2.5 because their base was located above sea level were excluded" [305-306 lines]. Please, explain this sentence. Have you found bounded aquifer(s) in these areas?
A further basic question follows: to run the model, the authors use a number of wells; are both the depth and the nature of the bottom of each of these known?
The parameter D rates the weight of the distance from the shore. In Figure 4b the authors draw the aquifer depth and in the text write "Figure 4b shows a map of the aquifer depth distribution up to 330 m, which corresponded to importance ratings of 2.5–10 for parameter D". Please explain.
What can the differences in hydraulic conductivity highlighted in Figure 4a be attributed to? Again, what can the differences in electric conductivity highlighted in Figure 4d be attributed to?
What new knowledge does the study provide about the problem you concern? Please, in Discussion critically use the literature on Jeju Island (especially Jung et al 2010; and El-Kadi et al 2014).
What do the authors consider to have improved of GALDIT?
Minor comments:
Redraw Figure 4 in two rows and two columns.
In the text, locality names are reported (e.g. Mureung, Gosan, Mt. Halla). Please, indicate the location of these in the figures.
Author Response
The objectives of the manuscript are the assessment of the vulnerability to seawater intrusion of coastal aquifer(s?) at Jeju Island, and the improvement of the techniques of the used model.
The manuscript is not clear about the aquifers relationships and settings. Initially, the authors roughly refer to an unconfined aquifer and a confined ones [111-112 lines]. After, they reveal " the study was on the unconfined aquifers of Jeju" [195 line]. Please, right away declare the aquifer (or the aquifers) you focused in the study. However the hydrogeological features of the Jeju Island must be described in 2.1. Site Description. About the "unconfined", how many aquifers are placed at your study site? Again, are the unconfined aquifers all in coastal setting?
Answer) Line 136-145: The hydrogeologic formation of Jeju island is assumed to consist of four different aquifer formations. The top formation consists of relatively permeable basalt; Seogwipo Formation comprises relatively low permeability volcanic sediments (average thickness of 100 m); the U-Formation occurred with average thickness of 150 m); the oldest formations is the relatively impermeable basement.
We considered the top formation as the unconfined layer that is the object site of this study. The unconfined aquifers are all in the coastal setting. You can see the outline of the formations from the cross-section of Jeju island shown in Figure 1(b).
Please, explain how you have obtained the correlations in Figure 3. Where came from the molar values of chloride and bicarbonates ions? If they are literature data ([30]?), write the Reference(s) in the caption.
Answer) We collected data from the literature [30] and created the figure from these data. The references are now written in the caption. As far as we know, this is the first attempt to obtain the correlations between molar values of chloride and bicarbonates ions vs EC data in a GALDIT assessment. (inserted at line 277-279)
The relevance of the hydrogeological structure of the island seems to be neglected. Thus data coming from "newly installed monitoring wells" [305 line] determine changes in the assessment. Moreover, it seems that the new wells have intercepted impermeable bottoms rather than seawater, because "Areas that were forcibly assigned a value of 2.5 because their base was located above sea level were excluded" [305-306 lines]. Please, explain this sentence. Have you found bounded aquifer(s) in these areas?
Answer) The relevance of the hydrogeological structure of the island is represented by the height of the groundwater level. The new monitoring wells were mostly installed in the southwestern area. Before the new monitoring wells were installed, high vulnerability to SWI was estimated for a relatively wide area due to the long distance between monitoring wells. The newly installed wells made the interpolation distance shorter and the high vulnerability zone became more accurate, as it was not based on interpolated values between wells.
The assigned value (2.5) is the default value, which is not a calculated value nor set for bounded aquifers. Technically, the areas where the bottom layer was above the sea level should not be included in the study site. However we did not want to show the mountainous area - cropped map. Instead we choose a value of 2.5 that never appeared from the calculation results.
A further basic question follows: to run the model, the authors use a number of wells; are both the depth and the nature of the bottom of each of these known?
Answer) We have data from subsurface drilling core samples from boreholes for the wells, as well as near core samples. In this way, we were able to determine the top of the Seogwipo formation, which is assumed to be the bottom of the unconfined aquifer.
The parameter D rates the weight of the distance from the shore. In Figure 4b the authors draw the aquifer depth and, in the text, write "Figure 4b shows a map of the aquifer depth distribution up to 330 m, which corresponded to importance ratings of 2.5–10 for parameter D". Please explain.
Answer) Thank you. We have corrected the typo from to “~2.5–10 for parameter T”
What can the differences in hydraulic conductivity highlighted in Figure 4a be attributed to?
Answer) The difference in hydraulic conductivity is due to geological heterogeneity. Won et al. (2005) studied the groundwater distribution for Jeju Island and showed the spatial distribution of the Seogwipo Formation, which is a relatively low permeability layer, and the hydraulic gradient of the groundwater. They concluded that these two factors resulted in extensive basal groundwater emergence along the coast of the eastern sector, which was less common along the coast of the western sector.
Again, what can the differences in electric conductivity highlighted in Figure 4d be attributed to?
Answer) The difference in electrical conductivity is still being investigated by Korean researches, currently assumed due to hydrologeologic heterogeneities.
What new knowledge does the study provide about the problem you concern? Please, in Discussion critically use the literature on Jeju Island (especially Jung et al 2010; and El-Kadi et al 2014).
Answer) A paragraph has been inserted in line with your suggestion:
“Finally, as Gogu and Dassargues (2000) suggested, new research on vulnerability assessments should be conducted that develops dynamic links between numerical models and overlay and index methods. Although almost 20 years have passed since Gogu and Dassargues’ (2000) study, there has been no clear outcomes in terms of linking numerical and overlay and index approaches. The previous study by El-Kadi (2014) was the first numerical study for the whole of Jeju Island to assess the sustainability of the water resources that compared the overall groundwater levels based on simulated and observed data. These models showed a good match when the groundwater level was higher than 2 m, which was far from the coastal regions. However, discrepancies occurred between observed and simulated results when the groundwater level was less than 2 m. In Jeju Island, most saltwater intrusion occurs where groundwater levels are less than 2 m, which comprises a large proportion of the area. Moreover, in the GALDIT model, 2 m is an important level for determining the risk of saltwater intrusion. Unfortunately, El-Kadi (2014) only modeled the saline distribution for the eastern Jeju watershed and Jung et al. (2010) only considered low levels of ion concentration for springs in coastal areas. Therefore, this study is the only one to consider the saline distribution for the entire island”.
What do the authors consider to have improved of GALDIT?
Answer) The ranges of the groundwater parameters were modified according to the local characteristics of Jeju Island to clearly distinguish vulnerable regions and show falling groundwater levels, which is strong evidence of increasing vulnerability due to anthropogenic activities. To improve the GALDIT assessment method to fit the Jeju model, we explored the possibility of using electric conductivity instead of the standard parameters representing the existing impact of SWI. The results of the GALDIT assessment using modified ranges and weights were mapped to show the SWI vulnerability of Jeju Island. We modified the objectives of the study in Lines 113-119 to clearly show these improvements.
Minor comments:
Redraw Figure 4 in two rows and two columns.
Answer) Thank you. We have redrawn this figure in two rows and two columns.
In the text, locality names are reported (e.g. Mureung, Gosan, Mt. Halla). Please, indicate the location of these in the figures.
Answer) Thank you. We have added the locality names in the figures as you suggested.
Author Response File: Author Response.pdf
Reviewer 3 Report
In this paper, the authors use GALDIT index to assess the intrinsic vulnerability to seawater intrusion of Jeju Island in Korea. The topic is suitable for the journal. The paper is well structured with well-defined objectives and results well related to the objectives. The paper is in general well written, except some unclear sentences. I would suggest major revisions. I have some technical concerns and some minor comments related to the form. Both of them are listed below. Technical comments: 1- The paper applies an existing method to a real site. From a technical point of view the paper, the paper is not new. But I think the application itself is interesting. It is very good to see the results (figures 5 and 6). As suggested by the authors somewhere in the paper (in the conclusion I think), this work can serve as a good example for the application of GALDIT to volcanic islands. I think this should appear somewhere in the introduction. 2- As mentioned in the section material and method, there are some numerical simulations studies for the Jeju Island site. It would be interesting to see (if it is possible) some qualitative comparisons between the results of this study and previous ones based on numerical simulations. If it is not possible please explain why. Minor comments: 1- I think there is no need for lines 40 tà 46. 2- Line 47 : “to assessing” should be “to asses” 3- Lines 47-48: “numerical, analytical, and overlay and index method” I would suggest “numerical simulations with the variable density flow-transport model, analytical solutions of the sharp interface approach, and overlay and index method” 4- Lines 48-49: “Numerical methods simulate the dynamics of groundwater flow in the coastal aquifer and are accurate at applying site characteristics [2,7-9]”. I think this sentence is not clear. In numerical simulation of SWI, the simulators reproduce the dynamics of groundwater flow and salt transport (not only groundwater flow). I would suggest: “Numerical simulations reproduce the dynamics of groundwater flow and salt transport in the coastal aquifer and are accurate at applying site characteristics [2,7-9]”. 5- To highlight the accuracy of the numerical models, please cite these works: - Q. Shao, M. Fahs, H. Hoteit, J. Carrera, P. Ackerer, A. Younes: A 3D semi-analytical solution for density-driven flow in porous media, accepted in Water Resources Research, 2018. https://doi.org/10.1029/2018WR023583 - M. Fahs, B. Ataie-Ashtiani, A. Younes, C.T. Simmons, P. Ackerer: The Henry problem: New semi-analytical solution for velocity-dependent dispersion, water resources research, 2016. https://doi.org/10.1002/2016WR019288 - A. Younes, M. Fahs: A semi-analytical solution for saltwater intrusion with very narrow transition zone, Journal of Hydrogeology, 2014. http://dx.doi.org/10.1007/s10040-014-1102-8 6- Lines 49-50: “However, despite the accuracy and usefulness of the results, numerical methods are limited because precise modeling requires both a considerable amount of hydrogeologic data as input and a difficult calibration process, so substantial computing power is needed for data processing and modeling.” I would suggest: “However, despite their accuracy and usefulness, numerical simulations require both a considerable amount of hydrogeologic data as input and a difficult calibration process, so substantial computing power is needed for data processing and modeling.” 7- Line 55: “and instead a sharp interface is assumed to be between the two media [10]” I would suggest: “and instead a sharp interface is assumed [10]” 8- Lines 55-57; “The sharp interface model has been employed to assess the vulnerability of an aquifer to the effects of SWI because of the relatively easy and rapid computations on the mathematical basis of SWI principles.” I didn’t understand what you mean by “on the mathematical basis of SWI principles”. Please remove this expression. I think the message here is to say that the main advantage of the sharp interface approach is its simplicity and efficiency (from a computational point of view) 9- Line 58: “this sharp interface” should be “this sharp interface approach” 10- Lines 59-60: please delete : “in which the sea level, recharge, and extraction are mathematically expressed” 11- Line 62: I think, before discussing the site, you should develop more consistent review (some lines) about the overlay and index techniques as they are used in this work. At this stage is good to cite a good reference for these techniques. 12- Line 89: What do you mean by “analytical methods”, please explain 13- Lines 101-102: “In order for these results to support water resource management in Jeju Island, the GALDIT model results were visualized as a tool for a SWI vulnerability test.” I cannot follow it, please reformulate? 14- Line 106: “island and south” please revise 15- Line 112: it is good to have a figure for a cross section. 16- Lines 124-126: “Fish farms have been installed along the shoreline, and groundwater wells have been developed near the shoreline that take in a huge amount of brackish groundwater.” This sentence should be before the previous one. 17- Lines 132-133: “A SWI vulnerability test using overlay and index techniques has never been performed for Jeju Island.” “SWI vulnerability tests using overlay and index techniques have never been performed for Jeju Island”. 18- Lines 391-392: please revise this sentence “The application of the SWI vulnerability method based on numerical ranking will be extended because the analytical method is familiar” I am not sure that the use of “will be extended” is good. I think this section should open future questions about the topic. It is not dedicated to describe future work. Please shorten this section and reformulate it in order to be as general as possible in opening further research questions related to this topic. 19- Please check the caption of figure 6 (I am not sure that during is the good word)
Author Response
In this paper, the authors use GALDIT index to assess the intrinsic vulnerability to seawater intrusion of Jeju Island in Korea. The topic is suitable for the journal. The paper is well structured with well-defined objectives and results well related to the objectives. The paper is in general well written, except some unclear sentences. I would suggest major revisions. I have some technical concerns and some minor comments related to the form. Both of them are listed below.
Technical comments:
1- The paper applies an existing method to a real site. From a technical point of view the paper, the paper is not new. But I think the application itself is interesting. It is very good to see the results (figures 5 and 6). As suggested by the authors somewhere in the paper (in the conclusion I think), this work can serve as a good example for the application of GALDIT to volcanic islands. I think this should appear somewhere in the introduction.
Answer)Thank you. We have inserted a sentence at line 109 “This study is the first to use the GALDIT method to assess its intrinsic vulnerability to SWI for the coastal aquifer throughout Jeju Island, which can serve as a good example for the application of GALDIT to volcanic island”.
2- As mentioned in the section material and method, there are some numerical simulations studies for the Jeju Island site. It would be interesting to see (if it is possible) some qualitative comparisons between the results of this study and previous ones based on numerical simulations. If it is not possible please explain why.
Answer)Thank you. We agree with your opinion. Your suggestion is similar to that proposed by Gogu and Dassargues (2000) for a topic of future work.
We are currently developing a finite difference numerical model by SEAWAT for the western Jeju area and are in the process of calibration for saline distribution. We will finish the project next year and are planning to publish the link between the SEAWAT numerical approach and the index-based GALDIT approach. A paragraph has been inserted as “Finally, as Gogu and Dassargues (2000) suggested, new research on vulnerability assessments should be conducted that develops dynamic links between numerical models and overlay and index methods. Although almost 20 years have passed since Gogu and Dassargues’ (2000) study, there has been no clear outcomes in terms of linking numerical and overlay and index approaches. The previous study by El-Kadi (2014) was the first numerical study for the whole of Jeju Island to assess the sustainability of the water resources that compared the overall groundwater levels based on simulated and observed data. These models showed a good match when the groundwater level was higher than 2 m, which was far from the coastal regions. However, discrepancies occurred between observed and simulated results when the groundwater level was less than 2 m. In Jeju Island, most saltwater intrusion occurs where groundwater levels are less than 2 m, which comprises a large proportion of the area. Moreover, in the GALDIT model, 2 m is an important level for determining the risk of saltwater intrusion. Unfortunately, El-Kadi (2014) only modeled the saline distribution for the eastern Jeju watershed and Jung et al. (2010) only considered low levels of ion concentration for springs in coastal areas. Therefore, this study is the only one to consider the saline distribution for the entire island.”
Reference :
Gogu, R. & Dassargues, A. Environmental Geology (2000) 39: 549. https://doi.org/10.1007/s002540050466
Minor comments:
1- I think there is no need for lines 40 tà 46.
Answer)Thank you. Lines 40-46 have been deleted.
2- Line 47: “to assessing” should be “to asses”
Answer)Thank you. The sentence has been modified.
3- Lines 47-48: “numerical, analytical, and overlay and index method” I would suggest “numerical simulations with the variable density flow-transport model, analytical solutions of the sharp interface approach, and overlay and index method”
Answer)Thank you. The sentence has been modified.
4- Lines 48-49: “Numerical methods simulate the dynamics of groundwater flow in the coastal aquifer and are accurate at applying site characteristics [2,7-9]”. I think this sentence is not clear. In numerical simulation of SWI, the simulators reproduce the dynamics of groundwater flow and salt transport (not only groundwater flow). I would suggest: “Numerical simulations reproduce the dynamics of groundwater flow and salt transport in the coastal aquifer and are accurate at applying site characteristics [2,7-9]”.
Answer)Thank you. The sentence has been modified.
5- To highlight the accuracy of the numerical models, please cite these works: - Q. Shao, M. Fahs, H. Hoteit, J. Carrera, P. Ackerer, A. Younes: A 3D semi-analytical solution for density-driven flow in porous media, accepted in Water Resources Research, 2018. https://doi.org/10.1029/2018WR023583 - M. Fahs, B. Ataie-Ashtiani, A. Younes, C.T. Simmons, P. Ackerer: The Henry problem: New semi-analytical solution for velocity-dependent dispersion, water resources research, 2016. https://doi.org/10.1002/2016WR019288 - A. Younes, M. Fahs: A semi-analytical solution for saltwater intrusion with very narrow transition zone, Journal of Hydrogeology, 2014. http://dx.doi.org/10.1007/s10040-014-1102-8
Answer)Thank you. The three papers have been referenced as you suggested.
6- Lines 49-50: “However, despite the accuracy and usefulness of the results, numerical methods are limited because precise modeling requires both a considerable amount of hydrogeologic data as input and a difficult calibration process, so substantial computing power is needed for data processing and modeling.” I would suggest: “However, despite their accuracy and usefulness, numerical simulations require both a considerable amount of hydrogeologic data as input and a difficult calibration process, so substantial computing power is needed for data processing and modeling.”
Answer)Thank you. The sentence has been modified as you suggested.
7- Line 55: “and instead a sharp interface is assumed to be between the two media [10]” I would suggest: “and instead a sharp interface is assumed [10]”
Answer)Thank you. The sentence has been modified as you suggested.
8- Lines 55-57; “The sharp interface model has been employed to assess the vulnerability of an aquifer to the effects of SWI because of the relatively easy and rapid computations on the mathematical basis of SWI principles.” I didn’t understand what you mean by “on the mathematical basis of SWI principles”. Please remove this expression. I think the message here is to say that the main advantage of the sharp interface approach is its simplicity and efficiency (from a computational point of view)
Answer)Thank you. The sentence has been modified.
9- Line 58: “this sharp interface” should be “this sharp interface approach”
Answer)Thank you. The sentence has been modified as you suggested.
10- Lines 59-60: please delete : “in which the sea level, recharge, and extraction are mathematically expressed”
Answer)Thank you. This part has been deleted.
11- Line 62: I think, before discussing the site, you should develop more consistent review (some lines) about the overlay and index techniques as they are used in this work. At this stage is good to cite a good reference for these techniques.
Answer) Thank you. We cited good references at 70-74
12- Line 89: What do you mean by “analytical methods”, please explain
Answer)Thank you. “analytical method” has been changed to “ overlay and index techniques”.
13- Lines 101-102: “In order for these results to support water resource management in Jeju Island, the GALDIT model results were visualized as a tool for a SWI vulnerability test.” I cannot follow it, please reformulate?
Answer)We agree that the sentence is not clear and it has been modified.
14- Line 106: “island and south” please revise
Answer)Thank you. This has been revised.
15- Line 112: it is good to have a figure for a cross section.
Answer)Thank you. The figure of a cross section has been inserted as Figure 1(b) and the layer system has been explained based on Figure 1(b).
16- Lines 124-126: “Fish farms have been installed along the shoreline, and groundwater wells have been developed near the shoreline that take in a huge amount of brackish groundwater.” This sentence should be before the previous one.
Answer)Thank you. The sentence has been moved as you suggested.
17- Lines 132-133: “A SWI vulnerability test using overlay and index techniques has never been performed for Jeju Island.” “SWI vulnerability tests using overlay and index techniques have never been performed for Jeju Island”.
Answer)Thank you. The sentence has been modified as you suggested.
18- Lines 391-392: please revise this sentence “The application of the SWI vulnerability method based on numerical ranking will be extended because the analytical method is familiar” I am not sure that the use of “will be extended” is good. I think this section should open future questions about the topic. It is not dedicated to describe future work. Please shorten this section and reformulate it in order to be as general as possible in opening further research questions related to this topic.
Answer)Thank you. The sentence has been revised.
19- Please check the caption of figure 6 (I am not sure that during is the good word)
Answer)Thank you. The caption has been modified from “during “ to “for the year”
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
English needs a revision by a native speaker.
The first author of the reference [29] is Goh (see caption of Fig.1b)? Or is he/she Koh (see 516 line in References)?
Author Response
Point 1: English needs a revision by a native speaker.
Response 1: Thank you. This manuscript was revised by a native speaker according to your comment. We attach “Certificate of English Editing”.
Point 2: The first author of the reference [29] is Goh (see caption of Fig.1b)? Or is he/she Koh (see 516 line in References)?
Response 2: Thank you. We corrected “Goh” to “Koh” in the caption of Fig.1b.
Author Response File: 
Author Response.docx
Reviewer 3 Report
The authors revised the paper according to my comments, Thank you.
Author Response
Point 1: The authors revised the paper according to my comments, Thank you.
Response 1: We like to take this opportunity to thank you for your kind review.
Author Response File: Author Response.docx
