Spatial Analysis and Fairness Evaluation of Seismic Emergency Shelter Distribution in High-Density Cities Based on GIS: A Case Study of Seoul

Round 1

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

The article is interesting and the topic is worthy of investigation.  However, I believe that the article suffers from some relevant shortcomings, which need to be corrected with improvements both from the point of view of its content. Next, I detail the issues that should be addressed by the authors.  If the authors manage to correct the detected deficiencies and incorporate the requested improvements in a satisfactory manner, the manuscript may be reconsidered for publication in the journal.

GENERAL ISSUE
The manuscript looks sometimes a technical report rather than a scientific paper. One of the main weaknesses of the manuscript is that it does not have a clear common thread, which clearly guides the reader to identify the objective of the investigation.  It is not clear whether the intention is to present a new innovative methodological framework or simply to carry out an the calculation of three GIS indicators for a local case study.  In addition, the manuscript should be given a more international focus to make it attractive to a wider audience, beyond local readers who are familiar with the case study. These issues should be addressed to enhance research scholarship and scientific relevance of the manuscript (see literature review, methodology and discussion comments for more detail).

INTRODUCTION AND LITERATURE REVIEW
The review of the state of the art is presented from an overly local perspective and is excessively focused on the case study. The references are not geographically diversified (Most of references are Chinese, Japanese or Korean) and address the topic from a perspective excessively limited to the analysis of the case study, neglecting the necessary theoretical reflection on the problem of seismic risk management from a broader perspective. The authors should analyze and cite, in the first two introductory sections, more strategic frameworks and methodological approaches to spatial analysis in seismic risk management and its derivatives (see, for example, the use of AI applied to the subject in https://doi.org/10.32604/rig.2024.051788, fuzzy methods for predisaster planning in https://doi.org/10.1016/j.habitatint.2025.103478 and geostatistics for seismic risk analysis in city neighborhoods in https://doi.org/10.3390/app9153182 ), and justify why their proposal represents a relevant contribution to this scientific field that deserves further study so that the manuscript does not appear merely justified as an simple analysis of a local case study.

METHODOLOGY
The methodological approach, as presented, is excessively simplistic and rudimentary. At times, it resembles a technical report more than a scientific article, as it is presented as the mere calculation of GIS indicators in a local case study, when it could be better framed as the proposal of an innovative methodological framework for addressing the challenges of seismic risk management in cities. In this regard, the authors are advised to integrate the various indicators into a comprehensive methodological strategic framework rather than simply a series of isolated calculations. This framework should be reflected in a more sophisticated, step-by-step analysis than the current one, one that genuinely leads to improved management of the problem from a policy perspective. This process should be summarized schematically in the methodology section for clarity and to clearly demonstrate its application in the results section.

SCIENTIFIC DISCUSSION
This part is a bit poor.  The analysis is too self-indulgent and lacks sufficient scientific, bibliographic, and academic references. In fact, a true section of scientific discussion is globally missing in the text, since what is mainly all an analysis of results for a local case study to enhance the scientific relevance of the manuscript. Consequently, this section should be improved to elevate scientific and academic erudition of the paper.  The authors must analyze to what extent the results obtained corroborate, contradict, or improve the results of previous studies on the subject, citing more international studies from other sites to enhance scientific and academic scholarship of the manuscript. 

The limitations of the proposed methodological framework should also be addressed here from a self-critical approach, explaining what issues could be improved in future lines of research.  Lastly, the policy implications of the results obtained should have the level of scientific and academic erudition of the text raised a little, pointing out relevant case studies in other parts of the world, where this methodological framework could be applied making a comparative analysis of the interest of this study for a wider audience beyond Korean readers.

Author Response

Comments 1: The manuscript looks sometimes a technical report rather than a scientific paper. One of the main weaknesses of the manuscript is that it does not have a clear common thread, which clearly guides the reader to identify the objective of the investigation.  It is not clear whether the intention is to present a new innovative methodological framework or simply to carry out an the calculation of three GIS indicators for a local case study.  In addition, the manuscript should be given a more international focus to make it attractive to a wider audience, beyond local readers who are familiar with the case study. These issues should be addressed to enhance research scholarship and scientific relevance of the manuscript (see literature review, methodology and discussion comments for more detail).

Response 1: We sincerely appreciate your valuable feedback. We have constructed an innovative methodological framework and visualized it through a new technical roadmap (Figure 4, Line 362). The research objectives are now clearly emphasized in the Introduction (Lines 114–117). Furthermore, we have expanded the international perspective throughout the Introduction, Methodology, and Discussion sections to enhance global relevance.

Comments 2: The review of the state of the art is presented from an overly local perspective and is excessively focused on the case study. The references are not geographically diversified (Most of references are Chinese, Japanese or Korean) and address the topic from a perspective excessively limited to the analysis of the case study, neglecting the necessary theoretical reflection on the problem of seismic risk management from a broader perspective. The authors should analyze and cite, in the first two introductory sections, more strategic frameworks and methodological approaches to spatial analysis in seismic risk management and its derivatives (see, for example, the use of AI applied to the subject in https://doi.org/10.32604/rig.2024.051788, fuzzy methods for predisaster planning in https://doi.org/10.1016/j.habitatint.2025.103478 and geostatistics for seismic risk analysis in city neighborhoods in https://doi.org/10.3390/app9153182 ), and justify why their proposal represents a relevant contribution to this scientific field that deserves further study so that the manuscript does not appear merely justified as an simple analysis of a local case study.

Response 2: We have refined the Literature Review to include a broader international scope. For example, new international references were incorporated (References 1 at Line 41; 6–7 at Lines 51–52; and 44 at Line 142), emphasizing strategic frameworks and global approaches to seismic risk management.

Comments 3: The methodological approach, as presented, is excessively simplistic and rudimentary. At times, it resembles a technical report more than a scientific article, as it is presented as the mere calculation of GIS indicators in a local case study, when it could be better framed as the proposal of an innovative methodological framework for addressing the challenges of seismic risk management in cities. In this regard, the authors are advised to integrate the various indicators into a comprehensive methodological strategic framework rather than simply a series of isolated calculations. This framework should be reflected in a more sophisticated, step-by-step analysis than the current one, one that genuinely leads to improved management of the problem from a policy perspective. This process should be summarized schematically in the methodology section for clarity and to clearly demonstrate its application in the results section.

Response 3: We have enhanced the methodological section to highlight the proposed innovative framework and integrated its analytical logic into a clear, step-by-step process, summarized in the Technical Roadmap (Figure 4, Line 362).

Comments 4: The methodological approach, as presented, is excessively simplistic and rudimentary. At times, it resembles a technical report more than a scientific article, as it is presented as the mere calculation of GIS indicators in a local case study, when it could be better framed as the proposal of an innovative methodological framework for addressing the challenges of seismic risk management in cities. In this regard, the authors are advised to integrate the various indicators into a comprehensive methodological strategic framework rather than simply a series of isolated calculations. This framework should be reflected in a more sophisticated, step-by-step analysis than the current one, one that genuinely leads to improved management of the problem from a policy perspective. This process should be summarized schematically in the methodology section for clarity and to clearly demonstrate its application in the results section.

Response 4: We have substantially rewritten the Discussion and Conclusion sections to improve their academic depth. Comparative insights from international case studies have been added (Lines 510–545), together with an expanded reflection on methodological limitations and future research directions (Lines 622–634). We also explicitly discuss how this framework can be applied to other global contexts (Lines 635–641).

 

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

This paper is an attempt to conduct an integrated analysis of the accessibility, redundancy, and spatial- and population-based fairness of seismic emergency shelters, using Seoul as a case study.  Overall, the research framework and analytical methods appear relatively appropriate; however, the completeness of the manuscript could be further enhanced by making several improvements to the detailed exposition, the description of data and indicators, and the structure of the discussion and conclusion sections.

 

1. Abstract duplication and structure (Lines 11–55)

The manuscript effectively contains two abstracts: the first from “Seismic disasters pose significant challenges …” (Lines 12–32) and a second beginning with “Seismic disasters pose major challenges …” (Lines 33–55). Please merge these into a single, concise abstract and remove redundant sentences or repeated results. Clarify which version you wish to keep and ensure consistency of key figures and conclusions. In addition, I would like to ask the authors to limit the number of keywords to no more than five.

 

2. wording issues and repeated phrases (e.g., Lines 17, 66, 97–101, 170–175, 242–250)

 

3. Please present the legend of the picture clearly (Line 198)

In Figure 1(b), the distinction between administrative-dong and legal-dong in Seoul is presented using color, but an explanation of the color scheme is missing. Please either add a legend to clarify the colors or revise the figure to use black-and-white lines without color differentiation. In addition, please change all content currently presented in Korean in Figure 1(c) into English.

 

4. Data description and terminology issues (Lines 203–222)

In Section 3.2, the study states that four main categories of spatial data are used, but then lists five (shelter locations, land use, roads, resident populations, administrative boundaries; Lines 203–205). Please correct this inconsistency. Lines 218–221: The English label “Number of resident registration population by gender and age by region (Buddhist Dong)” appears to mistranslate the Korean term (likely “legal-status neighborhood/administrative neighborhood ” rather than “Buddhist Dong”). Please verify and correct the terminology.

Please also clarify the reference year(s) for the shelter, land-use, road and population datasets, and state explicitly whether all datasets are aligned to the same year.

 

5. Conceptually, classifying all transportation zones as “built-up” while excluding some green/open spaces may have implications for fairness analysis, as many seismic shelters are located in parks and open areas. Please add one or two sentences discussing this classification choice and its potential impact on the results (Lines 247–256).

 

6. Justification of the 500 m accessibility threshold and text redundancy (Lines 270–278)

The justification for the 500 m walking threshold cites several factors (disruption of roads, elderly/children, density, other studies; Lines 273–276), but remains qualitative.

Consider: Providing one or two numerical references, Or please quote a previous study that was conducted in units of 500m were used.

 

7. Lines 358–366 and 468–469: The population fairness indicator FiF_iFi​ is truncated to the range [0.05, 3.0]. Values below 0.05 are classified as “severe undersupply”, 0.05–1.0 as “partial insufficiency”, and values above 1.0 as “adequate or surplus capacity”. The text notes that extremely high values can occur in commercial areas with very small registered populations (Line 363), but the choice of 0.05 and 3.0 as cut-off values is not fully explained. Because these thresholds directly affect the interpretation of fairness, please clarify:

Whether 0.05 and 3.0 are based on policy standards, prior studies, or exploratory analysis of your data; and approximately how many dongs are affected by this truncation (e.g., number or percentage).

A brief justification will help readers better understand the robustness of your classification.

 

8. Discussion and Conclusion

I recommend tightening these sections by:

Avoiding repeated definitions of fairness (keep one concise formulation);

Summarizing the methodological framework only once in the Conclusion; and

Using the freed space to emphasize more concrete theoretical and practical implications, such as how the proposed fairness metrics could be transferred to other high-density cities or integrated into real-world shelter planning guidelines.

This will make the Discussion and Conclusion more focused and impactful.

Author Response

Comments 1: The manuscript effectively contains two abstracts: the first from “Seismic disasters pose significant challenges …” (Lines 12–32) and a second beginning with “Seismic disasters pose major challenges …” (Lines 33–55). Please merge these into a single, concise abstract and remove redundant sentences or repeated results. Clarify which version you wish to keep and ensure consistency of key figures and conclusions. In addition, I would like to ask the authors to limit the number of keywords to no more than five.

Response 1: Thank you for your observation. The duplication resulted from previous tracked revisions. The issue has been corrected, and the abstract now appears as a single, concise version. Two keywords were removed, and we retained six, which complies with the journal’s 3–10 keyword guideline.

Comments 2: wording issues and repeated phrases (e.g., Lines 17, 66, 97–101, 170–175, 242–250)

Response 2: These repetitions were caused by revision marks in the previous tracked version. The final clean manuscript has been carefully reviewed to ensure there are no redundant phrases.

Comments 3: Please present the legend of the picture clearly (Line 198). In Figure 1(b), the distinction between administrative-dong and legal-dong in Seoul is presented using color, but an explanation of the color scheme is missing. Please either add a legend to clarify the colors or revise the figure to use black-and-white lines without color differentiation. In addition, please change all content currently presented in Korean in Figure 1(c) into English.

Response 3: As requested, we have revised Figure 1(b) and converted all Korean text in Figure 1(c) into English (Line 178).

Comments 4: Data description and terminology issues (Lines 203–222). In Section 3.2, the study states that four main categories of spatial data are used, but then lists five (shelter locations, land use, roads, resident populations, administrative boundaries; Lines 203–205). Please correct this inconsistency. Lines 218–221: The English label “Number of resident registration population by gender and age by region (Buddhist Dong)” appears to mistranslate the Korean term (likely “legal-status neighborhood/administrative neighborhood ” rather than “Buddhist Dong”). Please verify and correct the terminology. Please also clarify the reference year(s) for the shelter, land-use, road and population datasets, and state explicitly whether all datasets are aligned to the same year.

Response 4: We corrected the inconsistency by changing “four” to “five” main data categories (Line 183). The mistranslated term “Buddhist Dong” has been replaced with “Legal-status neighborhood” (Line 201). The reference years of all datasets are now explicitly aligned and stated (Line 209-210).

Comments 5: Conceptually, classifying all transportation zones as “built-up” while excluding some green/open spaces may have implications for fairness analysis, as many seismic shelters are located in parks and open areas. Please add one or two sentences discussing this classification choice and its potential impact on the results (Lines 247–256).

Response 5: We added a clear explanation of land-use classification and discussed its potential implications for fairness analysis (Lines 237–244).

Comments 6: Justification of the 500 m accessibility threshold and text redundancy (Lines 270–278). The justification for the 500 m walking threshold cites several factors (disruption of roads, elderly/children, density, other studies; Lines 273–276), but remains qualitative. Consider: Providing one or two numerical references, Or please quote a previous study that was conducted in units of 500m were used.

Response 6: We supplemented the justification with three previous case studies from high-density East Asian cities: Nanjing, Tianjin, and Beijing, all adopting 500 m as the seismic shelter threshold (Lines 264–271).

Comments 7: Lines 358–366 and 468–469: The population fairness indicator FiF_iFi​ is truncated to the range [0.05, 3.0]. Values below 0.05 are classified as “severe undersupply”, 0.05–1.0 as “partial insufficiency”, and values above 1.0 as “adequate or surplus capacity”. The text notes that extremely high values can occur in commercial areas with very small registered populations (Line 363), but the choice of 0.05 and 3.0 as cut-off values is not fully explained. Because these thresholds directly affect the interpretation of fairness, please clarify: Whether 0.05 and 3.0 are based on policy standards, prior studies, or exploratory analysis of your data; and approximately how many dongs are affected by this truncation (e.g., number or percentage). A brief justification will help readers better understand the robustness of your classification.

Response 7: We added a detailed justification for the truncation thresholds [0.05, 3.0], explaining that they were determined from the actual data distribution characteristics and the number of affected dongs (Lines 348–352).

Comments 8: Discussion and Conclusion. I recommend tightening these sections by: Avoiding repeated definitions of fairness (keep one concise formulation); Summarizing the methodological framework only once in the Conclusion; and using the freed space to emphasize more concrete theoretical and practical implications, such as how the proposed fairness metrics could be transferred to other high-density cities or integrated into real-world shelter planning guidelines. This will make the Discussion and Conclusion more focused and impactful.

Response 8: Both sections have been substantially revised. Redundant definitions of fairness were removed, and the discussion now emphasizes the framework’s transferability and theoretical implications for other high-density cities (Lines 510–641).

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

First, I would like to thank the journal editors for providing this opportunity to review the manuscript.

Addressing the insufficient assessment of the effectiveness and equity of the spatial distribution of earthquake emergency shelters in high-density cities (taking Seoul as an example), this article proposes a GIS-based "space-population" two-layer equity assessment framework. Using built-up neighborhoods (Dong) as the basic analysis unit, it conducts in-depth research from three dimensions: 500-meter walking accessibility, facility redundancy, and spatial and population equity. The study shows that Seoul's overall shelter accessibility is relatively high, but there are significant coverage gaps in the city center (such as Jung-gu) and mountainous areas. Redundancy analysis shows that the northeast and southwest regions perform better, while the core area performs worse. The equity assessment reveals a mismatch between physical spatial distribution and population demand; only about 45% of communities meet the per capita shelter capacity standard, and specific "low-low" (low accessibility, low redundancy) vulnerable areas have been identified. Overall, this study, through refined spatial analysis, reveals structural inequalities in the disaster prevention system of high-density cities, providing empirical evidence and policy recommendations for urban disaster prevention planning and resource optimization. However, the following shortcomings exist in content and format.

Suggested Revisions

1. The article currently primarily uses "Registered Resident Population" data (line 218). However, Seoul, as a high-density metropolis, experiences significant daytime and nighttime population movement in core areas such as Jung-gu and Gangnam-gu. Using only resident population data may severely underestimate the daytime shelter needs of commercial centers, leading to biases in the "population equity" analysis. It is recommended to try introducing "Floating Population" or "Daytime Population" data for supplementary comparative analysis.
2. The article lacks an overall technical roadmap; it is recommended to add one.
3. Figure 1(c) is merely a photograph of a shelter sign, contributing little to the academic paper and taking up valuable space. It is recommended to delete this sub-figure and allocate space for a more detailed study area map or to enlarge and display key areas.
4. Several figures in the article (Figures 1(b), 3(b), 4(b), 5(a), 5(b), 6(b), 7(a), and 7(b)) show densely packed lines in Seoul's central area. It is recommended to use magnified images of specific areas to illustrate the situation in Seoul's central region in detail.
5. The weighting of the population coverage model needs further justification. The current allocation of shelter capacity based on service area proportions may not reflect actual overcrowding, differences in accessibility, and localized unavailability of shelters. It is recommended to include explanations of limitations or sensitivity analysis.
6. The conclusions section should be further refined. The current conclusions overlap with the discussion; a greater focus could be placed on key findings, their value to planning, and future research directions.

It is recommended to revise and publish.

Author Response

Comments 1: The article currently primarily uses "Registered Resident Population" data (line 218). However, Seoul, as a high-density metropolis, experiences significant daytime and nighttime population movement in core areas such as Jung-gu and Gangnam-gu. Using only resident population data may severely underestimate the daytime shelter needs of commercial centers, leading to biases in the "population equity" analysis. It is recommended to try introducing "Floating Population" or "Daytime Population" data for supplementary comparative analysis.

Response 1: Thank you for your insightful comment. In Korea, floating population statistics are collected based on “administrative-dong” units, while our analysis uses “legal-status neighborhoods”. Hence, integrating the two datasets directly is technically infeasible within this study’s framework. As the core objective is to establish an innovative methodological model, the floating population will serve as a valuable direction for future research (Lines 622–634). Considering its importance, we have added a detailed discussion in the Discussion section analyzing population fluctuation across 25 districts on a typical weekday and its implications for population fairness (Lines 546–572).

Comments 2: The article lacks an overall technical roadmap; it is recommended to add one.

Response 2: We added Figure 4 (Technical Roadmap) to clearly present the analytical workflow (Line 362).

Comments 3: Figure 1(c) is merely a photograph of a shelter sign, contributing little to the academic paper and taking up valuable space. It is recommended to delete this sub-figure and allocate space for a more detailed study area map or to enlarge and display key areas.

Response 3: We have revised Figure 1(c) into an English version to illustrate actual shelter signage in Korea. Given its contextual relevance, we retained the figure. However, all key regional enlargements have been moved to the Results section for more meaningful visualization.

Comments 4: Several figures in the article (Figures 1(b), 3(b), 4(b), 5(a), 5(b), 6(b), 7(a), and 7(b)) show densely packed lines in Seoul's central area. It is recommended to use magnified images of specific areas to illustrate the situation in Seoul's central region in detail.

Response 4: Following your suggestion, key central regions have been magnified in the corresponding results figures (Lines 383, 394, 425, 451).

Comments 5: The weighting of the population coverage model needs further justification. The current allocation of shelter capacity based on service area proportions may not reflect actual overcrowding, differences in accessibility, and localized unavailability of shelters. It is recommended to include explanations of limitations or sensitivity analysis.

Response 5: We acknowledged this methodological limitation in the Discussion and Conclusion sections (Lines 622–634). The revised text explicitly notes that shelter capacity weighting assumes uniform accessibility and suggests sensitivity analysis or scenario-based modeling for future validation.

Comments 6: The conclusions section should be further refined. The current conclusions overlap with the discussion; a greater focus could be placed on key findings, their value to planning, and future research directions.

Response 6: The Conclusion has been refined to eliminate redundancy with the Discussion and to highlight key findings (Lines 593–602), the study’s planning relevance (Lines 635–641), and future research directions (Lines 622–634).

Round 2

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

All the concerns exposed in my previous report have been answered and the suggestions made to correct different shortcomings detected have been globally implemented in the new version of the manuscript. In my opinion the article can be considered for publication now.

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

The author has made significant modifications to the questions raised, and the revised paper content is relatively comprehensive. It is recommended to accept the manuscript.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The study demonstrates certain practical significance in its topic selection; however, it exhibits notable deficiencies in theoretical depth, methodological rigor, data currency, completeness of analytical dimensions, and feasibility of policy recommendations. Specific comments are as follows:
1. Lack of innovation: The GIS analytical methods employed in this study (service area analysis, accessibility measurement, and redundancy assessment) are already conventional in the field of emergency shelter planning. There is a lack of significant methodological or theoretical innovation, and the study fails to demonstrate a substantive contribution to the advancement of research in this area.
2. Absence of population dimension: The analysis focuses solely on spatial patterns of accessibility and redundancy, completely overlooking key social factors such as population distribution, density, and age structure. As a result, the "equity" analysis remains superficial and lacks explanatory power in terms of social fairness.
3. Overly simplistic method description: Key methodologies (such as service area generation, redundancy counting, and two-dimensional matrix classification) lack necessary technical details and parameter specifications, resulting in poor reproducibility and undermining the credibility of the analytical process.
4. Vague policy recommendations: The proposed suggestions (e.g., "strengthen cross-district resource sharing" and "raise public awareness") lack concrete implementation pathways or strong connections to the findings of this study. They are of limited practical use and fail to reflect the practical guidance value of the research.
Additional suggestions:
1. Line 13: “built-up neighborhoods (dongs)” should be revised to “built-up neighborhoods (called dongs in Korean)” to clarify the terminology.
2. Lines 39–40: “often amplifies secondary risks” is suggested to be changed to “often amplify secondary risks” to maintain subject-verb agreement.
3. Lines 85–86: “This gap provides the basis for the present study.” is overly subjective. It is recommended to rephrase it as “This gap motivates the present study.”
4. Lines 266–267: “the median accessibility was found to be 0.85, while the median redundancy was 0.08” should include an explanation of how these median values were calculated (e.g., whether they are based on normalized values).
5. Lines 311–312: “This overlap between low shelter access, high proportions of vulnerable populations, and tourism pressure…” should be revised to “This overlap between low shelter accessibility, high proportions of vulnerable populations, and tourism pressure…” to maintain terminological consistency.

Author Response

Please see attached document.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript focuses on the spatial distribution analysis of earthquake emergency shelters in high-density cities, using GIS tools to evaluate accessibility, redundancy, and fairness, and has certain academic value. The article topic meets the requirements of the journal, but overall, the manuscript still has shortcomings in terms of structural logic, methodological details, data processing, literature citation, and result presentation.

1. The abstract lacks specificity and in-depth description of quantitative results and policy recommendations. The abstract mentions that Seoul has high overall accessibility, but does not provide specific values (such as coverage percentage), which makes it difficult for readers to quickly grasp the core findings. At the same time, the author's viewpoint of prioritizing the provision of shelters in vulnerable areas has not been supported by actual data. In addition, it is necessary to ensure that the word count of the abstract is controlled within 250-300 words to ensure independent writing.
2. The keyword selection is not comprehensive and does not cover core methods and technologies, ignoring core concepts such as "accessibility", "redundancy", and "fairness".
3. The Introduction section lacks sufficient emphasis on earthquake risk in South Korea and cites outdated data. The article mentions that the earthquake risk in South Korea is relatively low and cites the 2016 Gyeongju and 2017 Pohang events, but recent data has not been updated.
4. The research objectives are expressed vaguely and the gap with existing research is not clearly defined. At the end of the introduction, it is mentioned to examine three aspects (walking accessibility, redundancy, and spatial fairness), but it is not explained why these dimensions were chosen or how to fill existing research gaps.
5. In the Literature Review section, there are very few references to local research in Korea (such as only mentioning earthquake related studies [13-20]), and most of them cite international literature, resulting in a disconnect from the Seoul case. Meanwhile, the fairness section did not discuss the impact of population mobility on Seoul. The summary of the research gap at the end of the review is too broad and does not specify the innovation of this study.
6. The data preprocessing details in the article are insufficient, and the definition of built-in areas is vague. The authors mentioned using Dissolve and Intersection tools to extract built up caves, but did not explain the land use classification criteria or quantify the exclusion area.
7. The method construction section lacks explanation for the setting of some parameters. If the authors use 500m as the accessibility threshold but do not explain why, it may overlook the actual walking ability of the elderly and children. The expression of the redundancy formula (Equation 2) is not precise and does not explain how to handle the counting of overlapping or boundary service areas, which may lead to counting bias. The selection of median threshold in fairness analysis is subjective, without explaining why mean or quantile are not used, and without considering data distribution. The district level fairness ratio (Equation 4) only focuses on Low Low and does not comprehensively compare all categories.
8. The Results section lacks quantitative support and Figure references for the description of the results. The result section is too simple and now looks more like a display of calculation results without in-depth analysis and discussion.
8.There is a lack of discussion of the research results, especially regarding the similarities and differences between this study and existing research results. At the same time, there is a lack of discussion on the shortcomings of this study and their impact on the results.
9.The conclusion part needs to be further improved. In addition to the need to summarize and analyze the results of this study, it is necessary to make a certain outlook on the future research direction.

Author Response

Please see attached document.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

While the study is commendable, there are several key areas where it could be substantially strengthened. My primary concerns relate to the omission of population data and shelter capacity, which are fundamental to a comprehensive fairness evaluation, and the need for stronger justification for certain methodological parameters. The following points are offered for your consideration to improve the manuscript for publication.

1. The study's definition of "fairness" is purely spatial, based on the accessibility and redundancy of shelter locations. This is a significant limitation. A truly equitable distribution must consider the population that needs to be served. A dense neighborhood with low accessibility is a much greater problem than a sparsely populated one. Similarly, all shelters are treated equally, but a small park has a vastly different capacity than a large school playground. The analysis in its current form might misrepresent the actual level of fairness.
2. The 500m walking distance needs a clear rationale. Why was this specific distance chosen? Is it based on a Korean government planning standard for emergency facilities, an average walking time (e.g., 5-10 minutes) cited in disaster literature, or another established precedent? Please provide a citation or a clear explanation to support this choice. This will enhance the methodological rigor of the study.
3. Please specify the classification method used to create the choropleth maps (e.g., Natural Breaks (Jenks), Quantiles, Equal Interval). This is standard practice for cartographic representation and aids in the interpretation of the maps.

Author Response

Please see attached document.

Author Response File: Author Response.pdf