ERT and GPR Prospecting Applied to Unsaturated and Subwater Analogue Archaeological Site in a Full Scale Laboratory

Round 1

Reviewer 1 Report

This paper presents some results of geophysical tests carried out in a full-scale laboratory test. Several different configurations were analyzed, in particular varying the water level in the physical model.

Although the methods used (GPR and ERT) are not innovative, the equipment, testing techniques and analysis of results, as well as the characteristics of the analyzed structures, are always challenges for researchers involved in this type of studies.

In this case, the presented results allow us to assess the potential and difficulties of each method in detecting archaeological structures, depending on the depth at which they are located and the water level in the model.

The use of different methods allowed the authors to conclude that it is possible to obtain better information when the results are analyzed together.

In general, the article is well written and well organized.

The materials and methods used in the study are well explained.

The conclusions are supported by the results presented.

The references are comprehensive.

In the opinion of the reviewer some aspects should be reviewed, namely:

The authors mention several times the inadequate compaction of some parts of the model, as interfering with the results obtained (e.g. "Further some reflections not associable to the structures due to a low soil compaction have caused a blurry image of the buried objects"). It is not clear to the reviewer, without having the results of compaction control tests, how this association of ideas can be made. This matter should be better explained.

Some figures are too big (eg Fig. 3) and others are too small (eg Fig. 16).

It is necessary to correct some sentences and misspelled words, for example:

Line 153 – “parameters” instead “paramters”

Line 165 – “area limited area”

Line 203 – “(WL2)and 1.50 m”

Line 245 – “In WP2,simulated”

Line 279 – “selceting only the most higher reflecions”

Line 292 – “selceting only the most higher reflecions”

Line 301 – “selceting only the most higher reflecions”

Line 395 – “the em reflections”

Author Response

Reviewer No 1 comments:

This paper presents some results of geophysical tests carried out in a full-scale laboratory test. Several different configurations were analyzed, in particular varying the water level in the physical model. Although the methods used (GPR and ERT) are not innovative, the equipment, testing techniques and analysis of results, as well as the characteristics of the analyzed structures, are always challenges for researchers involved in this type of studies.

In this case, the presented results allow us to assess the potential and difficulties of each method in detecting archaeological structures, depending on the depth at which they are located and the water level in the model.

The use of different methods allowed the authors to conclude that it is possible to obtain better information when the results are analyzed together.

In general, the article is well written and well organized.

The materials and methods used in the study are well explained.

The conclusions are supported by the results presented.

The references are comprehensive.

Authors: Dear Reviewer n.1, we thank you very much for your positive comments and we appreciated your work a lot. We have completed our revision, considering all your comments and suggestions, which improved the paper.

 

In the opinion of the reviewer some aspects should be reviewed, namely:

The authors mention several times the inadequate compaction of some parts of the model, as interfering with the results obtained (e.g. "Further some reflections not associable to the structures due to a low soil compaction have caused a blurry image of the buried objects"). It is not clear to the reviewer, without having the results of compaction control tests, how this association of ideas can be made. This matter should be better explained.

Authors: In a sand box or “pool box”, as our laboratory, one of the main problems during man made analogue experiments is the sand filled work. The refill was made with the help of an excavator and the installation of the archaeological structures by hand. Therefore, it is not possible to check the correct sand layering and compaction. Even if, we made several different waters loads to help the sand compaction, it is possible that somewhere was not well compacted defining em waves reflections.

 

Some figures are too big (eg Fig. 3) and others are too small (eg Fig. 16).

Authors: We changed them.

 

It is necessary to correct some sentences and misspelled words

Authors: Dear Reviewer n.1, we made all your suggested corrections

 

Reviewer 2 Report

Focusing on the performance of ERT and GPR method on prospecting unsaturated and subwater analogue archaeological site, authors designed a full scale laboratory test and analyzed the testing results. The experiments were interesting, however, the research results are not well analyzed and discussed. Some revisions are needed and the manuscript is not recommended to be accepted without major revisions.

General issues:

1. The Abstract section is suggested to be rewritten. Authors described too much on the research significance and objective. However, results and conclusions were not well introduced. It is suggested to state the performance differences of ERT and GPR technique in unsaturated and subwater conditions respectively.
2. Page 2, Line 50-60. Exploration using GPR or ERT under water is still a challenge. Previous underwater applications and characteristics of geophysical methods are not well documented and reviewed. It is one of the main topics concerned of the study. I suggest to add reviews of underwater applications.
3. Page 5, Line 187-189. The material of archaeological framework was described. However, the physical parameters (like density, modulus, resistivity) of archaeological framework were not mentioned. What is the physical property difference between each archaeological construction and background soil? It could provide reference for analyzing the prospecting performance of ERT and GPR method in different condition.
4. The description of performance of ERT and GPR in different working phases is qualitative, which seems like the general knowledge of geophysical methods. It is suggested to define the detection error form different aspects (like position, shape, or size) further quantitatively.

Minor issues:

1. Page 6, Line 192. The top of archaeological constructions is 0.3m below the soil surface. Why is the buried depth of archaeological constructions determined as 0.3m in experiment design? In addition, the electrode spacing was 0.6m in dipole-dipole array. With such electrode spacing, does the resolution of ERT match the small depth of archaeological constructions?
2. Page 11, Line 303. What inversion method is adopted for 3D ERT?
3. Figure 9. Some resistivity profiles are shown in Figure 9. What is the depth of cross section?
4. Page 17, Line 455. Should it be “in WL1 (b-e), WL2 (c-f) and WL3 (d-g)”?
5. Figure 17. How are the results of GPR and ERT integrated and co-rendered? It seems like that the results figures are superimposed after adjusting the transparency.
6. Page 18, Line 486. Should “threes” be “trees”?

Author Response

Reviewer No 2 comments:

Focusing on the performance of ERT and GPR method on prospecting unsaturated and subwater analogue archaeological site, authors designed a full scale laboratory test and analyzed the testing results. The experiments were interesting, however, the research results are not well analyzed and discussed. Some revisions are needed and the manuscript is not recommended to be accepted without major revisions.

Authors: Dear Reviewer n.2, we thank you very much for your review and we appreciated your work a lot. We have completed our revision, considering all your comments and suggestions, which  improved the paper.

General issues:

1.The Abstract section is suggested to be rewritten. Authors described too much on the research significance and objective.

Authors: We rewrote the Abstract

 

2.However, results and conclusions were not well introduced. It is suggested to state the performance differences of ERT and GPR technique in unsaturated and subwater conditions respectively. Page 2, Line 50-60. Exploration using GPR or ERT under water is still a challenge. Previous underwater applications and characteristics of geophysical methods are not well documented and reviewed. It is one of the main topics concerned of the study. I suggest to add reviews of underwater applications.

Authors: We improved the introduction part adding several reviews of underwater applications and some comments on them in order to well introduce our results and conclusions.   

 

3.Page 5, Line 187-189. The material of archaeological framework was described. However, the physical parameters (like density, modulus, resistivity) of archaeological framework were not mentioned. What is the physical property difference between each archaeological construction and background soil? It could provide reference for analyzing the prospecting performance of ERT and GPR method in different condition.

Authors: We add a table with the physical parameters (permittivity and electrical resistivity) of the archaeological structures and the background sand.

 

4.The description of performance of ERT and GPR in different working phases is qualitative, which seems like the general knowledge of geophysical methods. It is suggested to define the detection error form different aspects (like position, shape, or size) further quantitatively.

Authors: Following your suggestion, we defined a semi-quantitative approach taking in account the project plane of the buried archaeological structures and the slices coming from the GPR and ERT 3D results. Therefore, we added the figure 20 and 21 and they are discussed in the text in order to define a semi-quantitative table (Table n.3) where we defined the ability of each used geophysical methods to detect each buried archaeological structure.

 

 

Minor issues:

1. Page 6, Line 192. The top of archaeological constructions is 0.3m below the soil surface. Why is the buried depth of archaeological constructions determined as 0.3m in experiment design? In addition, the electrode spacing was 0.6m in dipole-dipole array. With such electrode spacing, does the resolution of ERT match the small depth of archaeological constructions?

Authors: The first question is not so clear to us. The archaeological constructions followed the experiment design and the shallower structure (top of the wall) was defined to be at 30cm below the surface. About the second question, we used an electrode spacing of 60cm for the 3D Dipole-Dipole acquisition experiment because we would like to have the best compromise between the resolution and the depth of exploration. Moreover, the top of the wall was 30cm deep, but the large part of the remains was between 60cm to 130cm deep.

 

1. Page 11, Line 303. What inversion method is adopted for 3D ERT?

Authors: We used ERTLab software (Geostudi Astier srl and Multi-Phase Technologies LLC). ERTLab is an electrical resistivity inversion software that offers full three-dimensional modelling and inversion. The inversion procedure of the software is based on a smoothness constrained least-squared algorithm (LaBrecque et al., 1999) with Tikhonov model regularization, where the condition of the minimum roughness of the model is used as a stabilizing function.

 

1. Figure 9. Some resistivity profiles are shown in Figure 9. What is the depth of cross section?

Authors: We forget to add this information in the figure text. We added it. The cross-section between the horizontal slices and the vertical sections is 0.45cm.

 

1. Page 17, Line 455. Should it be “in WL1 (b-e), WL2 (c-f) and WL3 (d-g)”?

Authors: We did a mistake; thanks for your suggestions.

 

1. Figure 17. How are the results of GPR and ERT integrated and co-rendered? It seems like that the results figures are superimposed after adjusting the transparency.

Authors: The co-rendered work is the second level of an integration framework, as discussed in the paper (Dell’Aversana, P; From rock physics to geophysical data integration: theory, applications and implications. First break, 32, 2014, 159-161). The integration process can be joined at the raw data level (data co-rendering), at the output level (model co-rendering), in a co-operative way (cooperative and constrained inversion) or taking into account the cross-relations at the petrophysical and/or structural level. The GPR and ERT data could be joint at different integration level, but at this stage we are not able to make a complex correlation such as joint inversion approach. We are starting to work on simple statistical approach with Machine Learning methods addressed on to time series forecasting, but we are at the early level (Giampaolo et al., Combining Multi-temporal Electric Resistivity Tomography and Predictive Algorithms for supporting aquifer monitoring and management. Near Surface Geoscience Conference & Exhibition 2020).  

 

 

 

1. Page 18, Line 486. Should “threes” be “trees”?

Authors: We did the word correction.

Reviewer 3 Report

This study presents an advanced ERT and GPR approach to detect ancient and buried remains. To demonstration usefulness of ERT and GPR, the authors performed the real scale laboratory experiment. Overall, the study provided a very interesting and helpful experiments for the study of ancient remains where it is inevitable to use GPR and ERT.

 

However, I have one question in the comparison of performances between ERT and GPR. I’m not sure we can say ERT has more benefit for underwater case because Fig.14 (c) and (d) is not underwater condition and it is not fair to compare GPR results for underwater case.

 

The clearer comparison about weakness and strength will make readers to understand better about the integration of two approaches will be pursued in the further study.

 

Overall, I would recommend for the publication after clarifying following minor comments.

 

Further comments:

 

p.3 line 100

The reference is required for Eq.(2).

 

P.13, lines 350-351

- This statement is not valid based on the Fig.13 where the readers cannot see the information of location and shape. For the underwater case a 2D plot similar to Fig. 8 will be helpful.

- Also, by overlapping fig. 1 to this 2D plot might provide useful insight for a shape comparison.

 

P.13, lines 355

Is there further analysis about why road (C) is undefinable? Is it due to the similar dielectric properties with soil or smaller size than others?

 

P.14, lines 356

I’m just curious about why the boundary between water and soil has a high signal?

 

p.17, lines 453-456

- I’m wondering how they are co-rendered. Are they simply overlapped?

- Can you provide a definition about the theoretical depth applied in the Fig.17.

- Also, the full notations of alphabets are missed for (B) and (E) in Fig.17.

 

1. 18, line 468

The number of Fig.17 should be corrected to Fig.18.

 

1. 19, lines 497-501

Authors emphasized that the integration with two approaches is the best.

ERT can be arranged in the underwater floor, but GPR cannot be. Therefore, is this integration recommended only for the underwater case?

Author Response

Reviewer No 3 comments:

This study presents an advanced ERT and GPR approach to detect ancient and buried remains. To demonstration usefulness of ERT and GPR, the authors performed the real scale laboratory experiment. Overall, the study provided a very interesting and helpful experiments for the study of ancient remains where it is inevitable to use GPR and ERT.

However, I have one question in the comparison of performances between ERT and GPR. I’m not sure we can say ERT has more benefit for underwater case because Fig.14 (c) and (d) is not underwater condition and it is not fair to compare GPR results for underwater case. The clearer comparison about weakness and strength will make readers to understand better about the integration of two approaches will be pursued in the further study. Overall, I would recommend for the publication after clarifying following minor comments.

Authors: The 2D ERT images in figure 13 (c) and (d) were obtained with electrodes located on the water floor, therefore we call them “underwater” acquisition. Moreover, the paper would like to emphasize the integrated use of different geophysical methods and the proposed experiments has been leaded just to achieve this goal. It is far from us the idea to make a competition between the two methods. From our point of view, it is important to highlight the weakness and strength of an integrated use of the two geophysical methods, trying to help the readers to understand better their application in special conditions. The revised paper was modified with some improvements, and we hope that it goes in the reviewer suggestion direction.        

 

Further comments:

p.3 line 100

The reference is required for Eq.(2).

 Authors: We added it

 

P.13, lines 350-351

- This statement is not valid based on the Fig.13 where the readers cannot see the information of location and shape. For the underwater case a 2D plot similar to Fig. 8 will be helpful.

- Also, by overlapping fig. 1 to this 2D plot might provide useful insight for a shape comparison.

Authors: The two profiles are indicated in figure 5. We add in the text the link. Moreover, in the new submitted version, we added new figures.

 

 P.13, lines 355

Is there further analysis about why road (C) is undefinable? Is it due to the similar dielectric properties with soil or smaller size than others?

Authors: The reflections associated to the road (C) are weakly identifiable, may be because the used material were not well consolidated and for its smaller size.

 

P.14, lines 356

I’m just curious about why the boundary between water and soil has a high signal?

Authors: We think about the dielectric contrast between tap water and sandy background.

 

p.17, lines 453-456

- I’m wondering how they are co-rendered. Are they simply overlapped?

Authors: The co-rendered images of different attribute or model parameters can be considered a sort of “level-zero” geophysical integration. In that case, we have just an a-posteriori comparison of models obtained separately through independent process. Despite the simplicity of this approach, significant benefits can be obtained from discovering correlations or inconsistencies between geophysical images of data, models and anomalies of different types (Dell’Aversana, P., Integrated geophysical method. Combining rock physics with Seismic, Electromagnetic and gravity data. 2014, EAGE publication). Obviously, the first “real” integration approach is constrained inversion or cooperative inversion or simultaneous joint inversion. The latest approaches are the Machine Learning methods. We are starting to work on using ML addressed on to time series forecasting, but we are at the early level (Giampaolo et al., Combining Multi-temporal Electric Resistivity Tomography and Predictive Algorithms for supporting aquifer monitoring and management. Near Surface Geoscience Conference & Exhibition 2020). Finally, we used a model co-rendered approach in this paper, as described in the cited paper.

 

- Can you provide a definition about the theoretical depth applied in the Fig.17.

Authors: We corrected the used terms. The depth was estimated from the GPR data and the ERT depth coming from the inversion.

 

- Also, the full notations of alphabets are missed for (B) and (E) in Fig.17.

Authors: We added a correct image.

 

    18, line 468

The number of Fig.17 should be corrected to Fig.18.

Authors: We did it.

 

    19, lines 497-501

Authors emphasized that the integration with two approaches is the best.

ERT can be arranged in the underwater floor, but GPR cannot be. Therefore, is this integration recommended only for the underwater case?

Authors: The paper would like to emphasize the integrated use of different geophysical methods everywhere. We know that the ERT could be used on underwater floor and the GPR has some obvious difficulties, but it is not a competition. On the contrary, the experiment highlights the good results with the GPR on the small boat, even if it is not widely used in the field. Therefore, the paper would be a great sponsor of these geophysical methods and the integration is the best solution when we have particular field survey, such as wetland scenarios.

Reviewer 4 Report

The article is devoted to the important problem of using geophysical methods in paleoarchaeological reconstructions. The integration of various methods allows taking into account the water saturation of soils and minimizing the degree of errors. The results of laboratory experiments and their effectiveness are clearly demonstrated. Further directions of research are indicated. The article is of scientific and practical importance and is recommended for publication.

Author Response

The article is devoted to the important problem of using geophysical methods in paleoarchaeological reconstructions. The integration of various methods allows taking into account the water saturation of soils and minimizing the degree of errors. The results of laboratory experiments and their effectiveness are clearly demonstrated. Further directions of research are indicated. The article is of scientific and practical importance and is recommended for publication.

Authors: Dear Reviewer n.3, we thank you very much for your review and we appreciated your positive comments a lot.