Evaluation of InfraRed Thermography Supported by UAV and Field Surveys for Rock Mass Characterization in Complex Settings

Round 1

Reviewer 1 Report

See attached PDF file

Comments for author File: Comments.pdf

Author Response

Dear reviewer,

thank you very much for providing us a detailed review report, which helped us to improve the quality of our manuscript.

Here's the list of the changes made throughout the text:

Lines 11, 14-15, 55: the incorrect information was removed and the bibliographic research was improved.

Line 34 (now 42): the period was added.

Figure 1: unfortunately, this is the resolution of the original paper from which the figure was taken. The legend for the white circle was added.

Line 45 (now 50): the reported references were accurately reported in the introduction and discussed in the discussion section.

Line 47-49 (now 99-107): the influence of urbanization and vegetation was clarified. Some examples on the role of “disturbance elements” on other remote sensing techniques (Terrestrial Laser Scanning and photogrammetry) for the geostructural characterization in the same study site can be retrieved from the article published by the same authors, to which reference is made.

Figure 2: the figure was improved by adding the field of view and location of the IRT sensor.

Line 84-103 (now 161-181): the description of the rock materials was reduced to the information necessary to understand the aims and scope of this research.

Line 109-119 (now 187-198): the table resuming the geotechnical parameters was added below the text.

Line 120-125 (now 201-223): the stereoplots were added.

Figure 4: the graphical scale was improved.

Line 192-194 (now 305-310): the position of the thermocouples was clarified.

Line 216 (now 334): the word “software” was replaced by “algorithm”.

Lines 223 and 247 (now 343): the definition of “fracture frequency” was added.

Figure 5: Due to logistical constraints, it was not possible to take a photo of the field of view of the camera (it was placed at the edge of a balcony). We did our best to make the figure clearer.

Line 247 (now 369): probably, the thermal delay is due to the fact that the thermocouples were placed in small voids of the rock, to avoid direct insulation on the sensors.

Line 275 (now 417) and Figure 7: we proceeded to clarify in the text that the correlations are referred to the reference profile. We did some tests to add the aspect along the vertical in Figure 7, but it is not feasible. Indeed, since the point cloud (used to support the correlations) has very few points along a linear profile, the presence of wrong normals leads to a bad calculation of the aspect. In addition, the aspect is rather constant along the reference line and, in general, along the surveyed rock wall. We discussed the effects of the aspect in section 3.3 (2-D thermal images, where some variations of the aspect were observed), lines 473-478 and 498-503.

Figure 8: we added the color scale, thanks for noticing that!

Section 3.3.2: we do not have information on the thermal conductivity of rocks, but this could be an interesting research topic for future studies.

Figure 9: the position of the reference profile was added.

Line 365-368 (now 545-550): in this case, it was not possible to obtain stereoplots from the thermal images. We could not measure the strike of the traces due to the perspective of the thermograms with respect to the sub-vertical rock wall.

Line 369-371 (now 550-554): we specified that the vegetation does not disturb the interpretation of the thermograms acquired during the night aimed at detecting the discontinuities. As a matter of fact, during the night, the discontinuities are detected in the form of warm thermal anomalies, whilst the vegetation gets colder.

General remarks:

• we did not consider the marine spray as disturbance element, but it would be interesting to further investigate its role in IRT acquisitions. This would be a nice research topic for future studies, but probably the surveys should be carried out in a more homogeneous zone (same lithology, no vegetation or urbanization) to monitor only the effects of marine spray and interpret the results.
• We added the reference to the article published in Remote Sensing.

Best regards,

the authors

Reviewer 2 Report

The manuscript entitled “Applicability of InfraRed Thermography for Rock Mass Characterization in Carbonate Environments” deals with the survey of a carbonate rock mass by UAV and IRT surveys to shed light on the type of information that can be achieved and on the limitation of IRT applied to rock masses.

The topic of this research is interesting from a scientific point of view, but with a low novelty. In fact, numerous studies on the same topic and with the same aims (i.e. the applicability of IRT, even combined with UAV, TLS and other remote surveying methodologies, for the rock mass characterization) already exist in international literature.

There are several points that should be addressed and clarified before being able to consider this manuscript suitable for publication:

1. The title does not reflect the content of the manuscript. In fact, based on it the manuscript should deal only with IRT, while there are UAV and TLS point cloud presented, with no apparent utility for this specific study except for a correlation that should be better explained. Therefore, I suggest to either adjust the title or adjust the manuscript content (see next points).
2. Line 12: Authors state that “only a few studies dealt with its applicability for rock mass characterization” talking about IRT. This is not true. There is a consistent literature, with a lot of recent studies published in several international journals. Authors are invited to improve their bibliographic research and to report this aspect, by considering also recent scientific contributions.
3. Introduction: a brief description of the methodological approach would be welcomed herein, along with an emphasization of novelty and scientific relevance of the study.
4. Figure 1 has a very poor quality.
5. Line 47: better define what Authors mean with “complex conditions”.
6. Lines 51-56: all these points have already been addressed by other studies. Authors should cite them, with reference to each single point, and then they should explain what novelty this study is bringing in the frame of the international state of the art. In fact, based on these points, the presented research fails in presenting new information.
7. Line 59: Authors should explain the utility of UAV surveys and laboratory tests in this work. See my previous comment at point 1.
8. Line 59: Authors state that they have performed conventional field surveys. Talking about a rock mass, a conventional field survey should be aimed at the rock mass characterization from the geostructural and geomechanical points of view, according to ISRM recommendations. No data on such aspects is reported in the manuscript and should be added.
9. Lines 60-64: this part is a conclusion and should be moved at the end of the manuscript.
10. Based on the manuscript aim, section 2.1 is too long and rich of useless information. Authors should reduce it.
11. Figure 3: in my opinion this figure is not necessary.
12. Lines 121-126: this part should be greatly improved and enriched with stereograms, showing the main discontinuity systems, and figures of some details of the fracturing condition. A table with the main geomechanical parameters of the rock mass would be useful to understand the condition of the surveyed rock cliff.
13. Lines 143-145: the GSI scores attribution, “according to the local conditions” should be better discussed and supported by data on the rock mass geomechanical conditions.
14. Line 152: the term “hazard” is not properly used. If elements at risk are involved, it would be more correct to talk about “risk”.
15. Line 160: although it is a known technique, in a scientific article a brief description of SfM should be reported.
16. Figure 4: probably this figure is out of the aims and scope of the research.
17. Lines 203-205: the application of the “reflector method” should be better described, even with the help of a figure. I wonder how big the aluminum foil-covered cardboard was to ensure a reliable survey from a spot located 40m away from the rock face. It is not clear from Figure 5. Some additional data supporting the reliability of such parameter should be reported.
18. Section 2.3: so, UAV model was used only for a remote geomechanical rock mass characterization. Any field validation of the UAV model? Moreover, what do authors mean with “discontinuity conditions” observed on a high-resolution point cloud? (line 226).
19. Section 2.3: line 216, please be more specific with respect to the employed software.
20. Line 257: authors state that “rock temperatures are related to the air conditions”. This is not a complete information. First of all they are talking about the “surface temperature” (please check throughout the manuscript), also depending upon the solar radiation, with specific reference to time of exposure to sunrays, and conditioned by the slope morphology and orientation. Moreover, the presence of external factors such as wet sector, weathering films (and so on…) condition the surface temperature of rocks.
21. Figure 6: not clear if these surface temperatures are average values, representative of the whole rock face, or related to fixed measurement spots. Since the rock surface temperature varies along the same rock mass, even under the same environmental condition, I wonder what kind of information this graph can bring. In fact, there is a fundamental difference between the heating phase and the cooling phase of a rock mass, i.e. the surface temperatures are strongly influenced by direct solar radiation for a certain phase of the day. Therefore, I believe this graph should not include the heating phase in order to avoid a misleading interpretation of the thermal response of rock masses.
22. Figure 7: literature experiences demonstrated that the best correlation between IRT outcomes and rock mass fracturing are related to the cooling phase, especially in dark environmental conditions. This is why, in my opinion, authors find the correlation “complex” (line 271). I think that they should separate heating phase information from those arising from the cooling phase.
23. Line 275-277: it is hard to believe that this is a constant outcomes resulting in the all-day IRT monitoring. Authors should report what time this consideration refers to. I guess this could be likely related to a morning thermogram, in direct irradiation condition, when the protruding rocks are directly hit by sunrays and hollow portions are partly shadowed. There are several studies on the application of IRT on rock mass surveys under natural light condition. Authors should improve the literature reference on this aspect.
24. Lines 278-284: same as previous point 23. Authors should report their consideration with respect to the different environmental conditions. This aspect, should be addressed in detail, providing also further thermograms of the same slope portions taken in different times, where differences should be highlighted and supported. This could be done even by improving Figure 9 and providing more information therein.
25. Lines 287-289: this consideration has no relevance if it is not referred to a specific survey time.
26. Line 291: authors should better explain what kind of “disturbance” was brought by vegetation in the thermograms and it would be interesting to show if this behavior is constant over the whole monitoring time.
27. Figure 8: not clear the surveying time. RGB and IRT images in inset “b” have a very poor quality. Probably the great surveying distance is not suitable for detailed consideration. Even with reference to this point, there are already published literature cases on potentials and limitations of IRT applied to rock mass survey.
28. Lines 311-312: this consideration is not new. Please, report the literature reference.
29. Lines 314-315: how a void can be heated by solar radiation? Authors should be careful in involving physical concepts with no supporting data. I would rather talk about rock+voids/fractures. Also this aspect has already been discussed in literature and the reference(s) should be reported.
30. Lines 324-326: not clear if open fractures are labeled with high or low surface temperature. After having explained this outcome, authors should report if this is in accordance with literature data or not.
31. Lines 330-332: consideration taken from published literature that has to be cited also herein.
32. Figure 9 is confusing and the caption is too long. I suggest to improve the readability of this figure by adding some key information within the image and to reduce the caption length.
33. Lines 372-374: already stated by previous articles that must be cited herein.
34. Lines 375-376: literature experience proved that the IRT survey of rock mass under direct solar radiation can give useful information on the rock mass morphology and so on. What stated here is certainly an opinion of the Authors and it is reasonable that they express this, but they should also report that there are other studies which took advantage from the direct irradiation of rock masses, for IRT surveys, with a positive feedback.
35. Line 381-385: this limitation is due to the great distance occurring between the rock face and the surveying spot. On the other hand, the resolution of the thermogram is related to the resolution of the employed thermal camera!

I hope that my comments will be useful to improve the content of this manuscript and I look forward to reading it after revisions!

Best regards

Author Response

Dear Reviewer,

we are grateful for your efforts in helping us improve the manuscript. We have made some modifications to better contextualize our research and explain its aims and scope, as well as to illustrate what is the novel contribution.

Here's the list of our modifications and comments to your observations:

• We adjusted the title of the manuscript according to the aims and scope of the research.
  • We improved the bibliographic research and revised the introduction, to contextualize our research and explain which were the goals and what novel contribution was introduced. In addition, we compared our results with the cited papers throughout the text.

4)    Unfortunately, this is the resolution of the original paper from which the figure was taken.

5)    Line 47 (now 99-107): we clarified what type of disturbance was caused by the vegetation and the urbanization. We added the reference to some examples on the effects of these elements on the interpretation of data acquired by means of other remote sensing techniques on the same study site.

6)   see comment 2-3.

7)   Line 59 (now 110-118): we explained the role of laboratory tests, UAV and field surveys in this research.

8)  We reported the results of the conventional geostructural and geomechanical surveys in section 2.1.

9)  We removed the sentence instead of moving it in the conclusion to avoid repetitions.

10) We reduced section 2.1. to report only the data necessary to contextualize the research.

11) We changed Figure 3 and gave more relevance to the information on the geostructural and geomechanical setting.

12) We added the resuming table of the geomechanical parameters (line 214).

13) We specified how the GSI was calculated (lines 234-238).

14) We changed the term “hazard” with the term “risk”, thanks for noticing this mistake.

15) We briefly described the SfM technique and added the references of papers describing this technique into detail (lines 255-264).

16) We believe that Figure 4 is relevant to understand the methodology, with particular reference to the role of UAV and TLS surveys.

17) We better explained the reflector method. Further information can be found in the references (lines 319-323).

18) We clarified why we performed the UAV surveys (lines 249-251). The data of the geostructural and geomechanical setting were mostly obtained from the field surveys, whilst the UAV point cloud and high-resolution photos were used to identify the elements that caused the thermal anomalies. We explained what “discontinuity conditions” was referred to (lines 346-347).

19) We added information on the algorithm used to extract the temperatures from the thermograms (lines 334-336).

20) We removed the incomplete information because unnecessary in this section.

21) Figure 6: we specified that the rock temperatures are those measured by the thermocouples. The aim of this plot is to show how the rock materials behave during the 24-hours monitoring. The results were compared with literature data.

22-25) Figure 7: we adjusted the plots and produced two figures, respectively for the cooling and the heating phase. We appreciate your hint because, by eliminating the “noise” caused by the Sun radiation, we were able to better interpret the results. However, the interpretation was still complicated because of the several factors contributing to the final thermal output. In general, we observed similar results on the plots of the heating phase (Figure S1 of Supplementary Materials), although less clear. In addition, we compared our results with those of other researches carried out by analyzing 2-D data. The thermograms of the same area taken in different moments of the day are illustrated in Figure 9 and discussed in section 3.3.

26) The disturbance caused by the vegetation was described in lines 440-442 and in Figure 8c.

27) Figure 8: the surveying time was added. The low resolution of inset b is due to the distance sensor-target, but, in our opinion, it is sufficient to observe that darker levels (other factors being equals) provided higher temperatures in the thermograms.

28-31) We corrected the unprecise sentences, compared our results with the literature data and provided a discussion.

32) The caption length was reduced to the information necessary to understand Figure 9, to which the legend was added.

33-34) The references were reported and discussed according to the results of this research.

35) The observation that low-resolution issues can be solved using high-resolution sensors was added in lines 569-572.

Thank you again for your support.

Best regards,

the authors

Round 2

Reviewer 2 Report

The manuscript “Evaluation of InfraRed Thermography supported by UAV and field surveys for Rock Mass Characterization in Complex Settings” has been revised by Authors. Some misleading aspects have been corrected and other unclear parts have been restyled.

Nevertheless, I found some further aspects that should be addressed before accepting it for publication. Please, find below my comments:

• Figure 1: the reason provided by the authors cannot justify the publication of a low-resolution image. Authors are invited to provide a better quality figure.
• Line 109: correct typo.
• Section 2.1: international recommendations followed for the field rock mass survey should be reported.
• Table 2: the persistence values have a precision to the second decimal place. How was it measured?
• Table 2: aperture values are reported to be up to 0.5 mm for all the 3 discontinuity sets. Nevertheless, by looking at figure 8, some evident open fractures are visible…please verify and correct this.
• Line 172-173: check written English
• Lines 257-259: authors did not addressed my initial comment on the “reflector method” procedure. I have serious doubts on the applicability of such methodology from such a distant IRT surveying point. How did the authors manage to calculate the surface temperature of an aluminum foil placed on a rock mass 35 m away from the IRT surveying points, even considering the low resolution of the used thermal camera? Probably it would be useful to add in Figure 5 a thermal image showing the aluminum foil and the related area considered for the surface temperature calculation.
• In the text it would be better turning “surficial temperature” into “surface temperature”.
• Lines 351-353: based on Figure 8, this interpretation seems wrong. It seems to me that the higher temperature are not related to the jutting block, but rather to the lower hollowed portion between the jutting block and the bedding surface. The linear positive anomaly, retracing the bedding trace below the jutting block is clear. Since this morphological setting determines the presence of a hollow rock mass sector below the jutting block, the higher surface temperatures are related to this preserved part, while the jutting volumes are labeled by lower temperatures.
• Lines 398-400: see previous comment.
• Lines 425-433: this comparison is hardly sustainable, because the cited literature is about laboratory setting, with oven heated specimens, while we have an irradiated rock mass herein. Please, modify or delete this part.
• Line 440: the term “darker” is referred to the specific chromatic scale used and its meaning can be misunderstood in case of the use of different chromatic scales. Please refer to the temperature range (“lower temperatures”) instead of the chromatic tone. Moreover, this concept has already been said in literature, please report the citation.

I hope that my further comments can be useful for the improvement of this research.

Best regards

Author Response

Dear Reviewer,

thank you for dedicating us some time to improve the quality of our manuscript.

We addressed all of your observations, as reported in the list below.

• Figure 1: the figure was changed to a higher resolution one.
• Line 109: correction done.
• Section 2.1: the reference was added at line 164.
• Table 2: the method used to calculate the persistence was clarified at line 167. The measurements of the discontinuity apertures were checked and corrected.
• Lines 258-262: a brief explanation of the reflector method was added in the 1^st review, as well as specific references. Following your suggestions, we added in Figure 5 an example of how the reflected temperature was measured in the thermograms, using the reflective paper. Indeed, the resolution of the thermal camera (2.4 cm/pixel) was enough to identify the reflector even if located at 35m away. As concern the resolution, in the text it is specified that the issue is related to the low resolution of the RGB camera of the thermal imager, which does not allow a correct interpretation of the thermograms. For this reason, conventional field surveys and UAV systems were used as support.
• The term “surficial” was changed to “surface”.
• Lines 351-353: we warmly thank you for your help with the interpretation of the plots. The inaccuracies were corrected in the paragraph 351-364 and in the captions of Figures 7, S1.
• Lines 425-433 (now 433-438): the sentence was deleted.
• Line 440 (now lines 450-451): the term was corrected and the references were added.

We  are fully available to  respond  to any further  questions  and  comments  you  may have. 

Best regards,

the authors