Induced Polarization Imaging: A Geophysical Tool for the Identification of Unmarked Graves

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is a interesting application of ERT/IP applied to the location of a historical grave. Such targets are always challenging for geophysical techniques and it is useful to see the potential success offered through IP. 

An initial area survey with closely sampled earth resistance using the twin electrode array (or similar) together with Ground Penetrating Radar (if conditions were suitable for this technique) would have been useful to provide a broader context to the ERT/IP results. Without this it is difficult to fully support some of the interpretation suggested for the ERT/IP data and, in particular, indicate why only a single grave anomaly has been encountered with the survey area?

It would be useful to provide details of the local geology and soils to provide context for the results and any additional historical notes, should they exist, on funerary practices at the site (were other wooden coffins buried here too?).

The quality of the figures is extremely good although I would suggest swapping the columns of Fig 7 so that the ERT and IP results are consistent with the placement in the earlier figures. Ideally, all sub-plots should share a common orientation with North at the top of the page but it might be worth mentioning in the caption that the orientation of the sub-plots varies and is indicated by separate North arrows (eg Fig 7e). In addition, the "plume" anomaly is shown to extend beyond the area covered by the survey profiles and it would be useful to explain how the data has been interpolated onto the resulting raster grids.

line 303 - this mentions "grain surface" and it is unclear whether this is the grain surface of the casket/coffin or the soil matrix?

Author Response

We would like to thank the reviewer for their thoughtful and constructive feedback. The comments provided valuable suggestions that helped us clarify key aspects of our methodology, interpretation, and presentation. We have carefully addressed each point below and revised the manuscript accordingly. Specific changes made in response to the reviewer’s input are detailed in the following responses.

 

Comment 1: An initial area survey with closely sampled earth resistance using the twin electrode array (or similar) together with Ground Penetrating Radar (if conditions were suitable for this technique) would have been useful to provide a broader context to the ERT/IP results. Without this it is difficult to fully support some of the interpretation suggested for the ERT/IP data and, in particular, indicate why only a single grave anomaly has been encountered with the survey area?

Response 1: We appreciate the suggestion of the reviewer.
We acknowledge that a high-resolution earth resistance survey or GPR could have provided additional context. However, the use of a twin-electrode array is limited by susceptibility to contact resistance effects and lacks the vertical resolution provided by four-electrode configurations and tomographic inversion. Mapping lateral resistance variations alone would not adequately resolve subsurface structures. Instead, ERT and IP surveys rely on hundreds to thousands of measurements with well-established inversion techniques to reliably characterize both lateral and vertical resistivity contrasts.
The integration of GPR with IP could indeed offer complementary information in future studies. However, at the time of this study, GPR acquisition was not feasible due to high soil moisture levels. Performing GPR now, under significantly different moisture conditions, would require repeating the IP survey for comparability, which falls beyond the scope of the present investigation.

Regarding the detection of only a single grave anomaly: this observation is consistent with the documented use of the cemetery section, which remains largely unused aside from the one grave that was established to reinter remains from a separate massacre site. This grave was clearly documented in historical photographs, which guided the selection of the survey location. Broader surveys at coarser resolution (1.0 m electrode spacing) did not reveal additional anomalies, underscoring the necessity of high-resolution measurements (0.25 m spacing) for detecting such features. Expanding this high-resolution approach to a wider area would be valuable but was beyond the current study's scope.

 

Comment 2: It would be useful to provide details of the local geology and soils to provide context for the results and any additional historical notes, should they exist, on funerary practices at the site (were other wooden coffins buried here too?).

Response 2: Thank you for the comment. Geological maps for the region encompassing our study area are available at scales of 1:200,000 and 1:50,000. However, no direct sedimentological or pedological investigations were conducted at the cemetery site itself, which limits the integration of site-specific subsurface information into our interpretation. Moreover, the conductivity values observed during our survey fall within a relatively narrow range, and therefore, larger-scale regional geological information does not significantly contribute to the interpretation of the geophysical data.

 

Comment 3: The quality of the figures is extremely good although I would suggest swapping the columns of Fig 7 so that the ERT and IP results are consistent with the placement in the earlier figures.

Response 3: We agree with the comment regarding the order of the columns and swapped the ERT and IP depth slices in th figure (Figure 8 of the revised manuscript).

 

Comment 4: Ideally, all sub-plots should share a common orientation with North at the top of the page but it might be worth mentioning in the caption that the orientation of the sub-plots varies and is indicated by separate North arrows (eg Fig 7e).

Response 4: We use variable orientations within and between figures to optimize layout and clarity. To ensure this is transparent to the reader, we have added a note to the relevant figure captions stating that subplot orientations may vary and that north is indicated in each panel (Figures 3, 5, 6, 7, and 8 of the revised manuscript).

 

Comment 5: In addition, the "plume" anomaly is shown to extend beyond the area covered by the survey profiles and it would be useful to explain how the data has been interpolated onto the resulting raster grids.

Response 5: The data collected along the survey profiles were jointly inverted in 3D, meaning the electrical potential was solved within a volumetric domain constrained by the spatial extent of the profile layout. This 3D inversion directly yields volumetric models of electrical conductivity, polarization effects, and coverage. The coverage serves as a quantitative measure of sensitivity, and we restrict our interpretation to regions where the model is well-resolved by the field data.

As a result, the observed anomaly extends not only along the profiles but also between them—not due to any interpolation of 2D results, but because the 3D inversion integrates all data simultaneously. Therefore, the anomaly outside the profile lines is a direct outcome of the 3D inversion, not of interpolation or extrapolation, respectively.

For visualization purposes, we extract horizontal slices from the 3D inversion model (stored in VTK format) and remap them onto regular grids using the griddata function from the scipy Python library. This remapping facilitates plotting with matplotlib, but it is important to emphasize that this step involves linear interpolation of model values onto a regular grid, not interpolation of 2D inversion results. It is purely a visualization aid and does not affect the physical extent or interpretation of the resolved anomaly.

 

Comment 6: line 303 - this mentions "grain surface" and it is unclear whether this is the grain surface of the casket/coffin or the soil matrix?

Response 6: We refer to the grain surface of the soil matrix, characterized by high surface area and surface charge, which can contribute to polarization effects detectable in IP measurements. Wood itself typically exhibits low surface area and low surface charge, and thus does not induce a strong polarization signal. However, during decomposition, organic compounds (e.g., dissolved organic carbon) may be released and subsequently adsorbed onto the surfaces of mineral grains in the surrounding soil. These interactions are known to influence charge accumulation and polarization responses.

We clarified the sentence in the manuscript accordingly on page 9, paragraph 3, line 375f: “… organic carbon adsorbed to mineral grain surfaces within the soil matrix.”

Reviewer 2 Report

Comments and Suggestions for Authors

This article was straight to the point in terms of methodological steps and approaches. It is very clear and informative, and I was able to understand its conclusions pretty well. In my opinion, it is fit for publication. It presents an important contribution both to Remote Sensing and to Forensic Sciences.

There are a few minor edits which I have considered relevant, and which must be addressed prior to publication. Here they are:

1) The comparison between DEMs and geophysical investigations should be more thoroughly discussed in terms of methods and calculations employed, especially to calculate the terrain ruggedness index.

2) Also, many image captions should be incorporated into the main text, which at times is too limited in detail.

3) The discussion is very limited and must be enriched with more details regarding:

3.1) The applicability of the method when detailed scale investigations are impractical.

3.2) What to expect from the large scale investigation in terms of clandestine cemeteries or mass graves (i.e. will the method be able to catch them without having to perform a more detailed investigation?)

3.3) To which extent the methodology changes with different types of soils and levels of moisture.

4) Also, please add more details to the conclusions, including - and most importantly - future directions and issues to be addressed in other branches of this research.

Overall, the research achieved its goals, but the authors have not pondered enough about its limitations and future possibilities.

Author Response

We thank the reviewer for their thoughtful and detailed comments, which have helped us to significantly improve the clarity and depth of the manuscript. We appreciate the recognition of the work and the constructive suggestions. In response, we have clarified the methodological workflow for calculating the terrain impact factor, enriched the discussion on the applicability and limitations of the induced polarization method under different conditions, and added a more comprehensive outlook on future research directions. Detailed responses to each point are provided below.

 

Comment 1: The comparison between DEMs and geophysical investigations should be more thoroughly discussed in terms of methods and calculations employed, especially to calculate the terrain ruggedness index.

Response 1: Agree. In the revised manuscript, we expanded the introduction to the section on subsurface discretization for geophysical inversion to more clearly explain the motivation and benefit of comparing the DEM with the geophysical survey design (page 4, paragraph 5, line 176ff). Specifically, we now emphasize that this analysis enables a quantitative evaluation of whether small-scale topographic variations—despite the overall flatness of the area—could affect the geophysical data or inversion. This strengthens the justification for assuming flat topography in the inversion model. We also clarify that this approach goes beyond visual inspection by providing a data-driven basis for model parameterization. The Terrain Ruggedness Index (TRI) calculation and the subsequent derivation of the Topographic Impact Factor (TIF) are already described in detail in the manuscript, and we now highlight more clearly how these metrics contribute to the overall assessment.
In addition, we have added a flowchart (Figure 4 in the revised manuscript) to further clarify the methodological workflow used to derive the TIF.

 

Comment 2: Also, many image captions should be incorporated into the main text, which at times is too limited in detail.

Response 2: We agree with the reviewer’s suggestion and have streamlined the figure captions accordingly. To avoid redundancy and improve narrative flow, we have incorporated key descriptive elements from the original captions into the main text where appropriate: page 7, paragraph 2, line 264ff

 

Comment 3: The discussion is very limited and must be enriched with more details regarding: What to expect from the large scale investigation in terms of clandestine cemeteries or mass graves (i.e. will the method be able to catch them without having to perform a more detailed investigation?)

Response 3:

The large-scale induced polarization (IP) survey was designed to characterize the unused section of the cemetery in the absence of historical cadastre data, with the specific aim of identifying grave-like anomalies exhibiting spatial characteristics comparable to those documented in the occupied sections. However, the large-scale IP investigation did not resolve any anomalies indicative of graves, including the known grave, which was only detected in the subsequent high-resolution survey targeted at the expected burial dimensions and depth.

This outcome underscores that the detection capability of IP surveys is strongly contingent on the survey design parameters relative to the spatial scale of the target features. The results demonstrate that IP is an effective method for detecting and characterizing clandestine or mass graves, but only when the survey geometry and acquisition parameters are appropriately matched to the target’s size and burial depth.

Therefore, large-scale surveys should be considered as preliminary reconnaissance tools that can exclude or highlight areas of interest but cannot replace detailed, higher-resolution investigations tailored to anticipated target dimensions. In situations lacking prior information about burial characteristics, a multi-scale investigative approach, as implemented here, is necessary to reliably detect and delineate clandestine burial features.

We have, accordingly, added a paragraph to the Discussion section: page 8, paragraph 4, line 330ff

 

Comment 4: The discussion is very limited and must be enriched with more details regarding: To which extent the methodology changes with different types of soils and levels of moisture.

Response 4: The IP response is primarily governed by surface charge and surface area. In our study, we attribute the observed polarization effects to the enrichment of organic carbon—characterized by extremely high surface area and surface charge—adsorbed onto mineral surfaces. In contrast, variations in soil moisture predominantly influence electrical conductivity, i.e., the real component of the complex conductivity.

We have expanded the Discussion accordingly: page 9, paragraph 4, line 382ff

 

Comment 5: Also, please add more details to the conclusions, including - and most importantly - future directions and issues to be addressed in other branches of this research.

Response 5: We agree with the reviewer and have revised the Conclusions section to include future directions and unresolved issues relevant to the broader application of this methodology (page 10, paragraph 1, line 429ff; page 10, paragraph 2, line 435ff; page 10, paragraph 3, line 446ff; page 11, paragraph 2, line 456ff).

Reviewer 3 Report

Comments and Suggestions for Authors

Dear authors,

This is a strong and well-executed manuscript. With some clarifications and enhancements in interpretation and visualization, it will make a valuable contribution.

Please find below the review report according the provided guidelines.

 

Summary

This manuscript describes a new use of time-domain induced polarization (IP) imaging to identify and characterize unmarked graves using non-invasive textile electrodes.

The study was conducted at a documented burial site in an inactive cemetery, combining large-scale and high-resolution measurements to evaluate the effectiveness of IP in forensic geoscience.

The research shows IP imaging provides a clearer delineation of burial features than electrical resistivity alone and can attribute anomalies to organic carbon resulting from decomposition.

The methodology, results and interpretations are well structured and supported by comprehensive modelling and inversion procedures.

 

General comments:

1. Methodological and Conceptual Considerations

Improvement 1: Regarding survey strategy (Section 2.2): The manuscript would benefit from a more explicit reasoning for choosing 0.25 m vs. 1.0 m electrode spacing beyond spatial resolution considerations. For example, authors should address:

• Justification of spacing relative to target size: Relate the chosen spacings to the expected dimensions of the grave. For example, if the burial box is known to be around 2 m long and 0.75 m wide, then the 0.25 m spacing provides at least 3–4 measurement nodes across the grave's width, supporting detailed delineation.
• Trade-off between resolution and depth of investigation: You could elaborate on how 1.0 m spacing balances sensitivity at greater depths (~2 m) but may miss small targets, as shown by the unresolved grave in the large-scale dataset.
• Survey optimization: Was any forward modeling done before fieldwork to evaluate resolution versus spacing trade-offs? Including even a brief note on synthetic model tests or experience guiding these choices would be helpful.

 

Improvement 2: Regarding topographic impact factor (Section 2.3): It is new way to assess the influence of microtopography on inversion results. However, for broader readership, these improvements could be made:

• Streamlining terminology: The manuscript introduces several derived metrics (TRI, ∇z, TIF, P(TIF)) that could overwhelm a reader unfamiliar with topographic correction practices. A diagram summarizing the workflow (e.g., flowchart of how DEM → TRI → TIF → S-curve) would be fine.
• Practical implications: The authors could provide concrete examples of how TIF results guided inversion strategy.
• Comparative context: How does the TIF compare to other methods for incorporating topography into inversion frameworks (e.g., built-in elevation handling in pyGIMLi or BERT)?

 

Improvement 3: The transformation of time-domain chargeability to phase assumes a constant phase approximation, which may oversimplify subsurface properties sometimes. This practice is accepted, but the manuscript should:

• Clarify the assumption's validity: Specifically, reference studies (e.g., Binley & Slater 2020; Flores Orozco et al. 2018) that show the range of applicability for constant-phase models in shallow, organic-rich soils.
• Address potential frequency dependence: Acknowledge that polarization phenomena can be dispersive, and under what soil/sediment conditions this assumption might break down.
• Comment on time window limitations: The manuscript mentions that the short decay window justifies the assumption. This point should be reinforced by showing that the decay curve is effectively modeled within that window (e.g., by including a representative decay curve and fit).

 

2. Results Interpretation

Improvement 4: Figure 7 (Interpretation): The visualization is fine. Consider supplementing the interpretation with a 3D perspective or cross-section to reinforce spatial relationships between anomalies.

 

Improvement 5: Conductivity vs. Polarization Anomaly (Section 4): The explanation of differences between conductivity and phase responses is insightful. Yet, the authors should also consider and mention Clay content and type (e.g., montmorillonite vs. kaolinite), Ionic strength of infiltrating fluids and soil moisture and porosity

 

3. Literature and Context

Citations: The paper is well-referenced with a strong foundation in both forensic and geophysical literature.

The references are recent, relevant, and not overly reliant on self-citation.

 

Improvement 6: Comparison to GPR and ERT: While GPR and ERT are mentioned, the discussion could benefit from a more thorough comparative analysis. For example, would the observed polarization anomaly have been missed or misinterpreted using ERT or GPR alone?

 

4. Broader Impact and Future Work

The conclusions are reasonable and are supported by the findings. The suggestions for future work are practical and relevant.

Improvement 7:  Applying IP in archaeological and forensic contexts is significant and should be emphasized more strongly in the abstract and conclusion to appeal to the broader Remote Sensing readership.

 

Minor Comments

Sentence 314: Clarify that the “plume-shaped anomaly” may be influenced by preferential flow paths or topography-induced drainage.

Sentence 335: Consider a brief mention of how textile electrodes compare in performance to traditional electrodes in terms of contact impedance or signal strength.

 

Data Availability and Ethics

The data availability statement is complete.

Ethical considerations are adequately addressed, particularly the use of non-invasive textile electrodes in sensitive burial contexts.

Author Response

We sincerely thank the reviewer for their extensive and detailed comments and suggestions. We truly appreciate the time and effort invested in carefully evaluating our manuscript. Your insightful feedback has been invaluable in improving the clarity, depth, and overall quality of our work. Below, we address each point in turn and outline the corresponding changes made to the revised manuscript.

Comment 1: Regarding survey strategy (Section 2.2): The manuscript would benefit from a more explicit reasoning for choosing 0.25 m vs. 1.0 m electrode spacing beyond spatial resolution considerations. For example, authors should address:

• Justification of spacing relative to target size: Relate the chosen spacings to the expected dimensions of the grave. For example, if the burial box is known to be around 2 m long and 0.75 m wide, then the 0.25 m spacing provides at least 3–4 measurement nodes across the grave's width, supporting detailed delineation.

• Trade-off between resolution and depth of investigation: You could elaborate on how 1.0 m spacing balances sensitivity at greater depths (~2 m) but may miss small targets, as shown by the unresolved grave in the large-scale dataset.

• Survey optimization: Was any forward modeling done before fieldwork to evaluate resolution versus spacing trade-offs? Including even a brief note on synthetic model tests or experience guiding these choices would be helpful.

Response 1: Thank you for the constructive feedback. We have revised the manuscript to clarify the rationale for the selected electrode spacings. The 1.0 m spacing was chosen to achieve a balance between depth of investigation (down to ~2.0 m) and lateral coverage, aiming to identify larger-scale subsurface structures. The 0.25 m spacing was employed to enhance near-surface resolution, particularly in the context of investigating the expected burial features, where fine-scale detail was essential. The corresponding added or revised parts in the revised manuscript:

• page 4, paragraph 3, line 153ff
• page 8, paragraph 4, line 330ff

Regarding survey optimization, no formal forward modeling or synthetic resolution analysis was conducted prior to fieldwork. Instead, the choice of electrode spacing was guided by extensive experience with induced polarization and resistivity surveys in comparable contexts. While numerous studies address survey design and resolution capabilities of resistivity methods (e.g., Goes et al., 2004; Martorana et al., 2017; Urruela et al., 2021), there is no universally applicable solution due to the complexity introduced by varying subsurface conductivity contrasts and heterogeneous target geometries. As demonstrated in studies such as Funk et al. (2024), the detectability of targets depends not only on electrode geometry but significantly on the electrical contrast between the target and background.

Our study underlines this point: the 1.0 m spacing failed to resolve the known burial, while the 0.25 m spacing provided sufficient resolution. This is not merely a function of depth sensitivity, but also of the ability to resolve lateral and vertical resistivity gradients in the uppermost layers. Although refining the mesh resolution for the coarser spacing would be possible, it would increase computational cost and model non-uniqueness, without necessarily improving the imaging results. We have chosen not to include a detailed numerical analysis of survey design, as this would exceed the scope and focus of our manuscript.

References:

Goes, B.J.M. and Meekes, J.A.C., 2004. An effective electrode configuration for the detection of DNAPLs with electrical resistivity tomography. Journal of Environmental & Engineering Geophysics, 9(3), pp.127-141.

Martorana, R., Capizzi, P., D’Alessandro, A. and Luzio, D., 2017. Comparison of different sets of array configurations for multichannel 2D ERT acquisition. Journal of Applied Geophysics, 137, pp.34-48.

Urruela, A., Rivero, L., Casas, A., Garcia-Artigas, R., Sendrós, A., Lovera, R. and Himi, M., 2021. Improving the resolution of investigation using ERT instruments with a reduced number of electrodes. Journal of Applied Geophysics, 186, p.104239.

Funk, B., Flores-Orozco, A. and Steiner, M., 2024. Possibilities and limitations of cave detection with ERT. Geomorphology, 462, p.109332.

 

Comment 2: Regarding topographic impact factor (Section 2.3): It is new way to assess the influence of microtopography on inversion results. However, for broader readership, these improvements could be made:

• Streamlining terminology: The manuscript introduces several derived metrics (TRI, ∇z, TIF, P(TIF)) that could overwhelm a reader unfamiliar with topographic correction practices. A diagram summarizing the workflow (e.g., flowchart of how DEM → TRI → TIF → S-curve) would be fine.

• Practical implications: The authors could provide concrete examples of how TIF results guided inversion strategy.

• Comparative context: How does the TIF compare to other methods for incorporating topography into inversion frameworks (e.g., built-in elevation handling in pyGIMLi or BERT)?

Response 2: We thank the reviewer for these valuable suggestions. To improve accessibility, we have added a schematic flowchart to the revised manuscript (new Figure 4) that illustrates the derivation and conceptual flow from the digital elevation model (DEM) to the topographic ruggedness index (TRI), gradient magnitude (|∇z|), topographic impact factor (TIF), and ultimately the probability function P(TIF). This visual aid is intended to help guide readers through the metric derivation and clarify the role of each parameter.
We agree that it is important to clarify the practical implications of the TIF. The TIF was developed not as a method for incorporating topography into the inversion per se, but rather as a pre-inversion diagnostic to quantitatively assess whether microtopographic variations in a given area are significant relative to the electrode spacing. Incorporating high resolution topography directly into the inversion mesh requires a higher number or mesh cells, leading to increased model complexity and ill-posedness. The TIF offers a probabilistic criterion to inform whether such complexity is justified. In this study, TIF results indicated that microtopographic variation was negligible in relation to electrode spacing, and thus elevation was omitted from the inversion mesh.
We added this background and intentions in the revised manuscript: page 4, paragraph 5, line 176ff.

 

Comment 3: The transformation of time-domain chargeability to phase assumes a constant phase approximation, which may oversimplify subsurface properties sometimes. This practice is accepted, but the manuscript should:

• Clarify the assumption's validity: Specifically, reference studies (e.g., Binley & Slater 2020; Flores Orozco et al. 2018) that show the range of applicability for constant-phase models in shallow, organic-rich soils.

• Address potential frequency dependence: Acknowledge that polarization phenomena can be dispersive, and under what soil/sediment conditions this assumption might break down.

• Comment on time window limitations: The manuscript mentions that the short decay window justifies the assumption. This point should be reinforced by showing that the decay curve is effectively modeled within that window (e.g., by including a representative decay curve and fit).

Response 3: The IP data were acquired using a Syscal Switch Pro instrument, which includes a low-pass digital filter at 10 Hz. This defines the upper bound of the frequency content of the measurements. Given the 0.5 s current pulse length used during acquisition, the fundamental frequency corresponds to approximately 2 Hz. As such, the effective bandwidth of the data lies between 2 Hz and 10 Hz.

This bandwidth is too narrow to meaningfully assess frequency dependence in the IP response. Consequently, the conversion of the measured integral chargeability values to apparent phase values using a linear relationship is valid within this limited frequency range. Moreover, the interpretation of the IP imaging results in this study is qualitative in nature and focuses on the spatial mapping of anomalous features rather than a quantitative characterization of subsurface material properties, such as organic carbon content.

Even if the linear conversion introduces minor uncertainties in the calculated phase values, these do not affect the overarching interpretation or conclusions of the study.
We have clarified this in the revised manuscript: page 6, paragraph 2, line 218ff.

 

Comment 4: Figure 7 (Interpretation): The visualization is fine. Consider supplementing the interpretation with a 3D perspective or cross-section to reinforce spatial relationships between anomalies.

Response 4: We appreciate the reviewer’s suggestion regarding the use of a 3D perspective. However, we have opted for depth slices to present our 3D imaging results, as this approach allows for a more detailed and interpretable visualization of subsurface anomalies at specific depths.
While 3D renderings can be visually compelling, they often suffer from limitations such as shading effects, occlusion of internal structures, and restricted visibility of deeper anomalies. These challenges can obscure important details, particularly when subtle variations in geophysical parameters are critical for interpretation. In contrast, depth slices provide a clear, layer-by-layer representation of the subsurface, which is especially useful for identifying the lateral extent and precise depth of anomalies.
Depth slicing is an established and widely accepted method in archaeogeophysics, particularly in ground-penetrating radar (GPR) studies, and we apply the same rationale to the visualization of our induced polarization results. This approach enables a direct comparison across depth intervals and facilitates the correlation of geophysical anomalies with potential archaeological features. For these reasons, we believe that depth slices offer the most informative and reliable means of presenting and interpreting the 3D data in this context.

 

Comment 5: Conductivity vs. Polarization Anomaly (Section 4): The explanation of differences between conductivity and phase responses is insightful. Yet, the authors should also consider and mention Clay content and type (e.g., montmorillonite vs. kaolinite), Ionic strength of infiltrating fluids and soil moisture and porosity.

Response 5: We appreciate the reviewer’s suggestion. However, no site-specific information is available regarding mineralogical variations in the soils or changes in pore-water composition. Incorporating such parameters into the interpretation would be speculative and potentially misleading, as they cannot be substantiated for the study area.
While we acknowledge that these factors can be important in hydrogeological studies or in quantitative interpretations of induced polarization (IP) data, we believe they are not critical to the interpretation of our results, which is primarily qualitative and based on observed contrasts in the IP signal. Furthermore, to the best of our knowledge, the influence of mineralogy and pore-water composition on field-scale IP signatures remains an open question and is still under active discussion in the scientific community. We have chosen not to elaborate on these parameters in the manuscript to avoid interpretation beyond the available data.

 

Comment 6: Comparison to GPR and ERT: While GPR and ERT are mentioned, the discussion could benefit from a more thorough comparative analysis. For example, would the observed polarization anomaly have been missed or misinterpreted using ERT or GPR alone?

Response 6: We acknowledge the suggestion regarding the use of GPR; however, GPR data were not collected in this study and are therefore not addressed in our interpretation. While GPR can be a valuable tool in archaeological prospection, its sensitivity is primarily to dielectric permittivity contrasts. In contrast, our study focused on the IP method, which is more directly sensitive to changes in bulk electrical conductivity and polarization effects associated with organic enrichment.
As demonstrated in our results, the conductivity models alone do not reveal the grave anomaly, highlighting the limited contrast in conductivity. Given that GPR is generally not sensitive to conductivity changes related to organic material enrichment, we would not expect it to detect anomalies that are not apparent in the ERT data. Furthermore, due to the current differences in soil moisture and the need to co-locate repeated surveys for meaningful comparison, a combined GPR/IP investigation is outside the scope of this study but may be considered in future work.

 

Comment 7: Applying IP in archaeological and forensic contexts is significant and should be emphasized more strongly in the abstract and conclusion to appeal to the broader Remote Sensing readership.

Response 7: We thank the reviewer for this suggestion. In response, we have revised both the abstract (page 1, paragraph 1, line 3ff; page 1, paragraph 1, line 14ff) and the conclusion (page 10, paragraph 2, line 408ff; page 11, paragraph 2, line 456ff) to more clearly emphasize the significance of applying induced polarization (IP) imaging in archaeological and forensic contexts.

 

Comment 8: Sentence 314: Clarify that the “plume-shaped anomaly” may be influenced by preferential flow paths or topography-induced drainage.

Response 8: We modified the sentence accordingly: page 9, paragraph 2, line 356ff

 

Comment 9: Sentence 335: Consider a brief mention of how textile electrodes compare in performance to traditional electrodes in terms of contact impedance or signal strength.

Response 9: Agree. In response, we have added a sentence (page 3, paragraph 3, line 110ff) briefly summarizing the performance comparison between textile and conventional steel electrodes. This addition highlights the advantages of textile electrodes in terms of contact resistance, reciprocal error, and field logistics, as demonstrated by Bast et al. (2025).

Reviewer 4 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review this paper. Overall I have little to add as all sections are presented effectively and succinctly, and the one thing that would fundamentally improve the study would be a wider range of case studies which may be already in mind for a future publication. With that in mind can I suggest you add a more detailed characterisation of the geomorphology of your study area and the ground conditions when you undertook the fieldwork.

Line 286. You refer to this as a large-scale investigation. I don't think you can accurately describe this study as large-scale as you one investigate a small portion of a single site.

In the conclusion can you also add to future research the need to test the system in a much wider range of geomorphologies and archaeological cemeteries predating the 20th century. I would also suggest that this method needs to be field-tested alongside alternative techniques such as thermal imaging.

 

Author Response

We thank the reviewer for their positive and constructive feedback on our manuscript. We appreciate their recognition of the clarity and succinctness of our work. We also value the helpful suggestions regarding the inclusion of additional case studies, more detailed characterization of the study site’s geomorphology and ground conditions, and expanded future research directions. We have carefully considered these comments and addressed them in the revised manuscript. Detailed responses to each comment are provided below.

 

Comment 1: Thank you for the opportunity to review this paper. Overall I have little to add as all sections are presented effectively and succinctly, and the one thing that would fundamentally improve the study would be a wider range of case studies which may be already in mind for a future publication.

Response 1: We thank the reviewer for the positive feedback and fully agree that a wider range of case studies would substantially strengthen the applicability and robustness of the approach. While the current study is focused on a single site, future research will aim to extend the application of the induced polarization (IP) method to other settings and burial environments. This will allow us to further evaluate the potential and limitations of the method across a broader range of conditions.

 

 

Comment 2: With that in mind can I suggest you add a more detailed characterisation of the geomorphology of your study area and the ground conditions when you undertook the fieldwork.< !-- Similar to Comment 2 of Reviewer 1. -->

Response 2: Geological maps for the region encompassing our study area are available at scales of 1:200,000 and 1:50,000. However, no direct sedimentological or pedological investigations were conducted at the cemetery site itself, which limits the integration of site-specific subsurface information into our interpretation. Additionally, the conductivity values observed during our survey fall within a relatively narrow range, so larger-scale regional geological data provide limited benefit for the interpretation of our geophysical results. It is important to note that field activities such as excavations, sampling, or intrusive testing are subject to strict ethical guidelines due to the sensitive nature of the site. These ethical considerations necessarily limit the extent to which ground-truthing or invasive subsurface characterization can be performed.

 

Comment 3: Line 286. You refer to this as a large-scale investigation. I don't think you can accurately describe this study as large-scale as you one investigate a small portion of a single site.

Response 3: Thank you for pointing this out. The terms “large-scale” and “detailed” were used here in a strictly local context to distinguish between the broader characterization of the entire unused section of the cemetery and the more focused investigation of the small area immediately surrounding the documented grave.

 

Comment 4: In the conclusion can you also add to future research the need to test the system in a much wider range of geomorphologies and archaeological cemeteries predating the 20th century. I would also suggest that this method needs to be field-tested alongside alternative techniques such as thermal imaging.

Response 4: We thank the reviewer for this valuable suggestion. We have incorporated the recommendation into the revised Conclusions section: page 10, paragraph 6, line 446ff.

Reviewer 5 Report

Comments and Suggestions for Authors

This article addresses the challenges associated with identifying unlabeled tombs, which often arise due to the limitations of traditional invasive methods imposed by ethical and cultural considerations. It investigates the potential of induced polarization imaging as a non-invasive alternative for identifying and characterizing unlabeled tombs. The following sections present a series of observations and insights.

 

(1) To enhance the author's comprehension and facilitate a quick understanding of the paper's key contributions, it is recommended that these contributions be explicitly outlined in the "1. Introduction" section.

(2) The Section 2 currently lacks essential visual aids such as flowcharts and schematic diagrams, which would aid in the understanding of the proposed methodology. It is advised to include these supplementary materials.

(3) Based on the description provided in Section 2 , the innovative aspects of the proposed method appear limited. Although the paper discusses the impact of varying geological conditions on detection depth and resolution, as well as methods for distinguishing electrical changes caused by organic matter decomposition from those influenced by other geological factors (e.g., soil moisture and clay content), and includes induced polarization data for subsurface discretization processing, the approach largely represents a combination of established techniques. As a result, the overall level of innovation remains relatively modest.

Author Response

We thank the reviewer for their careful reading of the manuscript and for the constructive feedback aimed at improving both the clarity and the scientific contribution of our study. We appreciate the helpful suggestions regarding the need for a clearer presentation of the study’s key contributions, enhanced methodological visuals, and a more explicit articulation of the innovative aspects of our approach. In response, we have revised the Introduction to explicitly outline the main contributions, incorporated a new flowchart to clarify the methodology, and elaborated on the novel elements of the study in both the methods and conclusions sections. Detailed responses to each of the reviewer’s comments are provided below.

 

Comment 1: To enhance the author's comprehension and facilitate a quick understanding of the paper's key contributions, it is recommended that these contributions be explicitly outlined in the "1. Introduction" section.

Response 1: We appreciate the reviewer’s helpful suggestion. To improve clarity and highlight the key contributions, we have now explicitly outlined the main novel aspects of our study in the Introduction section (page 3, paragraph 2, line 89ff). This addition aims to facilitate readers’ quick understanding of the paper’s objectives and significance.

 

Comment 2: The Section 2 currently lacks essential visual aids such as flowcharts and schematic diagrams, which would aid in the understanding of the proposed methodology. It is advised to include these supplementary materials.< !-- Should we also add a flowchart illustrating the inversion approach? -->

Response 2: We appreciate the reviewer’s suggestion. In response, we have included a detailed flowchart illustrating the computation workflow of the Topographic Impact Factor (TIF) based on DEM data to enhance clarity in Section 2 (Figure 4 in the revised manuscript).
While we provide the necessary mathematical foundations of the geophysical inversion used in this study, we consider a schematic overview of the inversion workflow to be beyond the intended scope of this manuscript. Readers seeking a more comprehensive introduction to the inversion methodology and implementation details are referred to the publication on the pyGIMLi framework \citep{ruecker2017}, which we employ for all inversion-related computations and cite in this study.

 

Comment 3: Based on the description provided in Section 2 , the innovative aspects of the proposed method appear limited. Although the paper discusses the impact of varying geological conditions on detection depth and resolution, as well as methods for distinguishing electrical changes caused by organic matter decomposition from those influenced by other geological factors (e.g., soil moisture and clay content), and includes induced polarization data for subsurface discretization processing, the approach largely represents a combination of established techniques. As a result, the overall level of innovation remains relatively modest.

Response 3: We thank the reviewer for their thoughtful feedback. While our study builds upon established methods like induced polarization and textile electrodes, we believe it also offers several novel and valuable contributions. These include (1) a new quantitative approach to evaluate the potential influence of microtopography on geophysical data and inversion results, (2) the first application of the IP method specifically targeting the detection and characterization of clandestine graves, and (3) a tailored survey design optimized for conducting IP investigations in sensitive areas. We see these aspects as important advancements that add meaningful value beyond a typical case study. Additionally, the revised Conclusions section (page 10, paragraph 4, line 429ff) highlights future research directions that we believe will further enhance the innovative potential of the approaches we present. We hope these points help to clarify the originality and relevance of our work.

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

Dear authors,

You have addressed almost all reviewer comments with clear, well-supported revisions.

Please find below the comments to the revision:

Comment 1:

Satisfactory. The manuscript now connects electrode spacing to target dimensions and operational goals. While forward modeling would have added value, the authors’ justification grounded in literature and field experience is acceptable for this applied context.

 

Comment 2:

Satisfactory. The addition of a visual schematic improves clarity for non-specialist readers.

The practical purpose is well-positioned within the workflow, and the comparison with existing topography-inversion strategies provides sufficient context.

 

Comment 3:

Satisfactory. The technical justification is well-reasoned.

 

Comment 4:

Satisfactory. The rationale is convincing and aligns with best practices in forensic geophysics. While 3D visuals can enhance communication, their omission does not detract from the scientific merit of the data presentation.

 

Comment 5: Regarding confounding Soil Factors

Acceptable, but with Caution. While the decision not to speculate is scientifically responsible, a brief acknowledgement in the discussion that these variables can influence IP responses (even if not constrained here) would help generalize the findings for broader applications. Consider adding a single cautionary sentence in future revisions.

 

Comment 6:

Satisfactory. Given the absence of GPR data, the authors appropriately clarify methodological limitations and suggest potential for future studies.

 

Comment 7:

Satisfactory. The revised abstract and conclusion improve the manuscript’s appeal to Remote Sensing readers.

 

Comment 8:

Satisfactory. The revision appropriately clarifies the origin of the anomaly.

 

Comment 9:

This addition strengthens the justification and is appropriate for the intended audience.

Author Response

We thank the reviewer for their thoughtful and constructive feedback throughout the review process. We are pleased that most of our revisions have addressed the reviewer’s concerns.

Regarding the confounding soil factors (Comment 5) we agree with the reviewer’s suggestion to acknowledge additional factors influencing induced polarization (IP) responses. As recommended, we have revised the discussion to include a cautionary statement noting that parameters such as clay content and type, pore-fluid ionic strength, and mineralogical composition may affect both the real and imaginary components of complex conductivity. We emphasize that these parameters were not explicitly considered in our study due to a lack of site-specific data and that their influence on field-scale IP signatures remains an open question under active investigation in the scientific community. The revised paragraph (page 9, paragraph 4, page 385ff) provides this broader context while maintaining a focus on the qualitative nature of our interpretation.

Reviewer 5 Report

Comments and Suggestions for Authors

The author has made revisions in response to my suggestions, especially in the aspect of innovation that I am concerned about. The author has provided explanations and corresponding modifications. I think the article has improved to a certain extent compared to the previous version. However, I think the writing mode of the article can still be further optimized. The current way of writing text and pictures separately always feels not very friendly to read. Of course, this is just a suggestion.

Author Response

We sincerely thank the reviewer for their constructive and thoughtful comments throughout the review process. We appreciate the time and effort invested in the review.

Regarding the figure placement, the current manuscript format follows the journal’s submission guidelines, which require figures to be placed separately from the main text.