You are currently viewing a new version of our website. To view the old version click .
MDPIMDPI
Search all articles
by
  • Chao Liu1,
  • Xiaoyan Chen2 and
  • Zekun Li1
  • et al.

Reviewer 1: Anonymous Reviewer 2: Adedibu Sunny Akingboye Reviewer 3: Anonymous Reviewer 4: Padam Jee Omar

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript presents a comprehensive and methodologically intensive approach to assessing roof water inrush risk in shallow-buried coal seams, using a combination of FAHP and CV weighting methods. The integration of GIS-based vulnerability indexing and multi-model comparison is commendable. However, in my opinion, some issues need to be improved.

（1）It is recommended to explain the advantages of choosing FAHP over AHP, as well as the reason for selecting the Coefficient of Variation (CV) method instead of other objective weighting methods.

（2）Does the 79 borehole data in the paper have regional representativeness? It is recommended to illustrate the spatial distribution of the samples.

（3）For Equation (3) “𝐻𝑙𝑖= 33.6 × ∑𝑀”, the citation source or derivation basis has not been clearly specified. It is recommended to supplement the literature citation for this empirical formula.

（4）Terminology consistency should be maintained throughout the paper, such as for “key stratum thickness” and “lithology association index”, to avoid inconsistent expressions.

（5）The quality of Figure 2 is poor; it is recommended to redraw it to improve its aesthetics. In addition, all figures should be replaced with high-resolution versions that are clear and distinguishable to ensure normal reading for readers.

（6）In Section 3.2, the statement that “the vulnerability of Panel 2 is higher than that of Panel 1” requires specific quantitative data for support, which is recommended to be supplemented.

（7）It is recommended to supplement the description of the Spearman correlation coefficient results in Table 9.

（8）A large number of equations are presented in the paper, and some of them have formatting inconsistencies, such as those in Lines 195-209 and 411-426.

（9）In Point (5) of the Conclusions section, it is mentioned that “some influencing factors may have been omitted”. It is recommended to clearly specify 1-2 specific potential factors.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript assesses roof-water inrush hazards using ten geological–hydrogeological factors combined through FAHP, CV, and multiple weighting schemes. Although the topic is important, major methodological gaps, unclear data processing steps, inconsistent equations, and insufficient reproducibility prevent verification of the results, warranting a major revision. Below are comments and suggestions to be addressed:

  1. The Abstract opening sentence is wordy and contains punctuation/spacing problems ("roof water inrush from coal seam roof is a core water hazard" — repetitive). Suggest rewriting to a single clear opening sentence (e.g., "Roof water inrush is a major hazard threatening coal-mine safety.") and removing duplicate phrases. The sentence listing the four combination methods is compressed and contains a run-on structure. Rephrase for clarity and ensure consistent capitalization of method names (e.g., "Linear Weighted Method; Multiplicative Synthesis Normalization Method; Minimum Information Entropy Method; and Game Theory Method").
  2. Lines 41–46: Several sentences in the Introduction are long and contain multiple clauses that obscure the message. Tighten sentences, avoid repeated phrases such as "height of the caved zone" and "height of the water-conducting fracture zone" in the same clause without a clear distinction. Also provide explicit references for the empirical formulas mentioned (currently cited generically as [5]). In Lines 53–59: The statement "three maps-dual prediction method" needs a reference and a brief one-sentence explanation or citation of how it differs from the FAHP-CV approach. Currently, it reads as jargon without context. Most importantly, Figure 1 has poor resolution; please improve.
  3. Lines 99–104: Reported mine water inflow range ("40.46 to 178.52 m³/h") needs a brief statement about measurement period and method (e.g., measured at mine outlets between YYYY–YYYY, monthly averages, instantaneous peaks). State whether these values are per panel, daily, hourly, or from pumping records.
  4. Lines 124–131: Selecting the ten controlling factors is fine, but the manuscript does not document the criteria for excluding other potentially relevant factors (e.g., hydraulic conductivity, permeability tests, groundwater pressure head). Add a short justification and a list of excluded factors with reasons.
  5. Lines 152–161 and Equations (1)–(5) (lines 195–212): Equations need clearer definitions and consistent notation. For Eq. (1), the layout is confusing: define each symbol inline and ensure the equation formatting is correct (e.g., F = (∑LiHi)/S). For Eq. (2), the fractal formula "lg N(l) = −Ds lg(l) + C" should identify base-10 log or natural log consistently and explain N(l) precisely (number of fault segments longer than l). For Eq. (3) the coefficient "33.6 × ∑M" requires a citation and derivation/justification — why 33.6? Provide the source or derivation. For Eq. (4) and Eq. (5), supply units for coefficients and confirm the normalization procedure; the LCI formula should use parentheses to avoid ambiguity.
  6. Lines 166–174: The physical meaning and units of many derived indices (e.g., "sandstone effective thickness" and "lithological association index") need to be explicit. State whether these are in meters, dimensionless normalized indices, or percentages after normalization. Clarify the coefficient values (a, b, c, d) origin, and whether sensitivity analysis was performed on them.
  7. Lines 215–237: The normalization formulas (Eqs. 6–7) are standard, but the manuscript does not describe treatment for outliers, missing values, or repeated measurements. Specify how NaNs and outliers were handled (censoring, winsorizing, or removal). Also state whether normalization was applied per raster cell or per-borehole dataset prior to interpolation.
  8. Lines 248–273: For the FAHP subjective weighting, the manuscript should report (a) number of experts consulted, (b) backgrounds of those experts, (c) method of aggregating expert responses, and (d) uncertainty (e.g., standard deviation of expert weights). Currently the text describes FAHP generically but gives no procedural transparency. Add these details and an example of the pairwise matrix used.
  9. Lines 333–365 and Table 3: In the CV method results table the "sum of CV" and "weight value" columns appear misformatted. Confirm calculation steps and include a short reproducible snippet or summary of the Python code (or include code in supporting information). Also state the sample size n used for each CV computation (79 boreholes are mentioned but confirm per-factor effective sample size after filtering).
  10. Lines 376–404 and Tables 4–6: For each combined-weight method, include a short justification for parameter choices (e.g., linear weights α=β=0.5). Provide sensitivity analysis showing how result zoning changes with α varying (e.g., α=0.3–0.7). For Minimum Information Entropy (Eq. 23), you present a closed-form √(wA·wB) result; include derivation steps or a reference.
  11. Lines 430–458 and Eq. (27): The Game Theory combination method uses a consistency metric S and a threshold "S ≤ 0.4". Provide a citation or justification for choosing 0.4 as the threshold; show the distribution of S when bootstrapping weights or across alternative expert groups. Also include the exact computed S (you quote S = 0.1425 later; make sure that number is presented in this paragraph).
  12. Lines 512–541 and Eqs. (29)–(32): The vulnerability index expressions are correct in principle, but the manuscript must explicitly state the spatial resolution of rasters, interpolation method, and grid cell size used in ArcGIS (e.g., 10 m, 35 m, etc.). Also state coordinate reference system used when overlaying maps. Without these details, others cannot reproduce spatial maps.
  13. Lines 550–561 and Table 8: The Jenks Natural Breaks thresholds are given for each VI model. Explain the sample size used to compute Jenks breaks and whether the same break count (5 classes) was forced across all rasters. Also give rationale for choosing Jenks over other classing methods and test robustness (e.g., compare with quantiles).
  14. Lines 571–587 and Figure 7/8 discussion: The spatial differences between methods are described qualitatively. Provide quantitative area statistics (e.g., percentage area in each class per panel for each method) and a confusion matrix or kappa statistic comparing maps to support the qualitative claims about similarity/difference among methods.
  15. Lines 588–595 and Eqs. (33)–(34): For Spearman correlation the formula shown is fine; however, report the sample size n used to compute ρ, include p-values, and describe how tied ranks were handled (you mention a correction but do not show the correction terms). Also report the distribution of residuals or rank differences to support the claim that VI2 achieved ρ=0.9961.
  16. Lines 633–641 and Figure 8 caption: Figure captions must be self-contained. Currently "Figure 8. Zoning map..." does not specify panels shown, legend classes, projection, or data sources. Update caption with (a) map scale, (b) coordinate system, (c) raster resolution, (d) classification breaks, and (e) color scale description.
  17. Lines 645–656: The geological interpretation names faults (F19, DF2, F17, F20) but the manuscript lacks a clear fault map labeling these features or a table describing their attributes (strike, dip, throw, mapped length). Include a figure or table listing the faults referenced in the text with their parameters and data sources.
  18. Lines 690–695 and 693 (Spearman result): The claim that the multiplicative method has Spearman 0.9961 is very high; include the sampling methodology (number and distribution of random sample points), bootstrap confidence intervals for ρ, and alternative ranking metrics to ensure result robustness. Without uncertainty measures, 0.9961 may appear overstated.
  19. Lines 706–713 (Conclusions and Limitations): Good to mention generalizability limitations. Strengthen this by adding explicit suggestions on how to validate the workflow at other mines (e.g., cross-validation with measured inflow events, use of time-series pumping-test data, or blind testing at other panels). Also add an explicit statement on ethical/data permissions if company data are proprietary.
  20. Lines 726/727: The Data Availability Statement says "original contributions presented ... included in the article" but the spatial datasets, interpolation grids, and Python scripts are not included. For transparency and reproducibility, (a) include the version of Python and major libraries (numpy, pandas, arcpy or GDAL/OGR, scipy, sklearn), (b) include or attach the scripts used for CV/FAHP/combination and GIS raster overlay (or provide a DOI/URL to a repository), and (c) provide the borehole dataset (or a sanitized version) and raster grids as supplementary materials. Hence, update the Data Availability Statement accordingly. Clarify whether any datasets are confidential and provide guidance for accessing them.
  21. Numerous minor English grammar and typographical errors. I recommend careful language editing by a native/experienced technical editor.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This article presents an assessment of the risk of water irruption in the roof of the shallow coal seams at the Guojiawan mine. The authors combine a subjective (FAHP) and objective (coefficient of variation) weighting method with four integration models to optimize the assessment. Their approach, implemented via a GIS platform, identifies vulnerable areas and demonstrates that the fault intensity index and the sand/clay ratio are the most significant influencing factors. Despite the relevance of this work, several questions remain:

  1. Is the selection of ten control factors sufficiently exhaustive? Could the absence of dynamic factors, such as hydraulic pressure on aquifers or the influence of seasonal rainfall, limit the accuracy of the short-term risk assessment? Add a sentence or two in the section "2.2. Analysis of the Main Controlling Factors" to discuss the factors considered but ultimately discarded.
  2. Although the approach combines subjective and objective methods, the initial FAHP weighting phase relies on expert judgment. How were the potential biases of these experts quantified or minimized to ensure the neutrality of the subjective weights?
  3. Have the areas identified as "vulnerable" or "at risk" been correlated with historical water eruption data or in situ observations? Empirical validation would significantly strengthen the credibility of the results. 
  4. Are there any historical data on water eruptions or high flow rates in the galleries that could be used to spatially validate the results of your model?
  5. To what extent is this methodology, developed for specific geological conditions in Guojiawan, transferable to other coal basins with different hydrogeological configurations?
  6. The overall linguistic quality is very good for an academic publication. The English is perfectly understandable and professional. The identified issues are minor and do not affect the comprehension of the scientific content. A final review by a native speaker could eliminate any remaining imperfections, but the article is linguistically very publishable.
  7. The "Conclusions" section briefly addresses the limitations, but a dedicated "Limitations and Future Work" section would be more standard and rigorous. This would allow for an open discussion of the model's generalizability and the factors not taken into account.

The article is scientifically sound, innovative, and has clear practical utility. Acceptance after major revisions.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

The paper entitled “Risk Assessment of Roof Water Inrush in Shallow Buried Thick Coal Seam Using FAHP-CV Comprehensive Weighting Method: A Case Study of Guojiawan Coal Mine” addresses an important safety topic in coal mining—roof water inrush risk assessment—and proposes a combined FAHP–CV weighting framework along with four composite weighting strategies integrated in a GIS environment. The focus on shallow-buried Jurassic coalfields and the Guojiawan case study enhances the applied relevance.

However, this paper needs major modifications and clarifications at some places to make it more informative to the readers.

Reviewer comments for improvement are as follows:

  1. The abstract provides a concise summary, but adding a sentence about the practical implications of your findings could strengthen its impact.
  2. In the methodology section, the selection of 10 main controlling factors is logical, but no statistical test (PCA, multicollinearity, sensitivity analysis) is used to justify their independence and factor interactions (faults + thin aquitard + high WCFZ height) are not considered. The authors are requested to please make it correct.
  3. Correct the caption of figure 2, make it little bit more informative.
  4. Figure 3. the text of the figure is not readable. Please make the text of the figure readable.
  5. Same issue with the figure 4.
  6. Section 2.4.2. Coefficient of Variation (CV) gives high weight to variables with high dispersion, but dispersion does not always equal importance.
  7. The authors acknowledge CV biases but still use it without addressing limitations through sensitivity or robustness analysis.
  8. The paper provides no theoretical justification for why these four methods were selected over others (CRITIC, entropy–AHP hybrid, DEMATEL).
  9. The discussion section restates results rather than critically analyzing them.
  10. The paper claims superior performance of the Multiplicative Synthesis Normalization Method (VI2), but the justification based solely on Spearman correlation is weak.
  11. All references must be consistently formatted and correctly cited. The authors are advised to carefully check the following references and may use them in the study.

https://doi.org/10.1007/978-981-15-6564-9_16

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have substantially addressed my previous comments and suggestions.

However, they should further clarify the abstract by reducing its length and emphasizing the main research problem, novelty, objectives, key results, overall contribution to knowledge, and the transferability of the research.

In addition, several sections of the manuscript contain excessively long single-block paragraphs that need to be divided into two or three shorter paragraphs. The authors are advised to seek assistance from a native English speaker to improve the scientific sentence flow and overall clarity.

Author Response

Comments 1: The authors have substantially addressed my previous comments and suggestions.

However, they should further clarify the abstract by reducing its length and emphasizing the main research problem, novelty, objectives, key results, overall contribution to knowledge, and the transferability of the research.

In addition, several sections of the manuscript contain excessively long single-block paragraphs that need to be divided into two or three shorter paragraphs. The authors are advised to seek assistance from a native English speaker to improve the scientific sentence flow and overall clarity.

Response 1： Thank you for pointing this out. We agree with this comment and we further refined the abstract section and optimized and polished the content of the article.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have thoroughly addressed all of our comments and suggestions in this revised version. Consequently, I am pleased to recommend the article for publication in its current form.

Author Response

Comments 1: The authors have thoroughly addressed all of our comments and suggestions in this revised version. Consequently, I am pleased to recommend the article for publication in its current form.

Response 1: Thank you very much. We will continue to strive.

Reviewer 4 Report

Comments and Suggestions for Authors

I read the revised manuscript, entitled: “Risk Assessment of Roof Water Inrush in Shallow Buried Thick Coal Seam Using FAHP-CV Comprehensive Weighting Method: A Case Study of Guojiawan Coal Mine”. The manuscript shows improvement compared to the previous version, and some of the earlier comments have been addressed. However, several important issues remain that should be revised to enhance the quality and clarity of the paper. Please consider the following points:

  1. Provide the references for the equations used in the study.
  2. Legend of the figure 03 is not readable to the readers. The authors are requested to make it large so that it can be readbale.
  3. Same issue with the figure 04 also.
  4. A more text and discussion is needed for the table 02, 03, 04 and 05.
  5. In the previous review, the authors are suggested to follow the journal guidelines while formating the references and suggested a work ‘https://doi.org/10.1007/978-981-15-6564-9_16’ but the authors did not use them in the study. Please check the work and use them in the study.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf