ADAtools: Automatic Detection and Classification of Active Deformation Areas from PSI Displacement Maps

Round 1

Reviewer 1 Report

First of all, congratulations to the authors for presenting an important work for the readers in a very organized and systematic manner. Sharing of such type of outcomes from a project is definitely fruitful for the rest of the world.

Few minor suggestions for the authors:

Comment 1, Page 3, L 96: Use the preposition ‘on’ instead on ‘in’.

Comment 2, Page 3, L 100: Write 2019 instead of 272019.

Comment 3, Page 4: Figure 1 needs to be referenced also.

Congratulations to the authors for presenting such research in an organized manner. Good luck.

Author Response

Dear reviewer,

please find below our answer to your comments (however, the changes made should be clearly visible in the modified version of the paper, since the change tracking feature is activated).

Please note that when we review your requests, our document (word) has NO line numbers. Therefore, we cannot state where the changes we made take place using this mechanism. On the contrary, and besides activating the change tracking mechanism in word, we added some side-comments stating the number of the reviewer the change is targeted at. Therefore, look for comments labeled "Reviewer 1" to find our answers in the document. Please, excuse the inconveniences this might cause to you.

--- "Comment 1, Page 3, L 96: Use the preposition ‘on’ instead on ‘in’."

We corrected the typo; we also checked whether there were other typos and corrected these when found (very few of them after so many revisions).

--- "Comment 2, Page 3, L 100: Write 2019 instead of 272019."

We're very sorry, but at first we couldn't find this typo. We guess that the editors already corrected it since line 100 in page 3 said:

"ADAs was presented in [26] and [27] back in 2017 and 2019, respectively. In these works, the ..."

We noticed that someone else had changed this when we activated the change tracking feature. So it has been corrected.

--- "Comment 3, Page 4: Figure 1 needs to be referenced also."

Figure 1 was already referenced at the beginning of the paragraph preceding the figure itself. We guess that you didn't find the text, which says:

"Figure 1 depicts the Graphical User Interface (GUI) version of the ADAfinder tool. Note that some of the values..."

To finish: thank you very much for your very positive comments and for your effort reviewing our work! Thanks a lot!

Reviewer 2 Report

Dear Authors,

Here are my comments on your manuscript.

1. This paper presents a nice topic, which is developing software for interpreting the InSAR results. In terms of development, the InSAR method is already in the mature stage. However, to interpret their results is very important. InSAR results in interpretation is still an open issue to discuss. And this paper offers a solution to that issue. So, it is important to publish this paper, and collect more feedback to the author to enhance their software.
   In this software, there is a tool called los2hv which is used to decompose the Horizontal (East-West) component and Vertical displacement component. And its result is used as one parameter to classify the displacement type (Sinkhole, subsidence, etc). How if the real displacement is in North-South direction. Does this tool will not detect such type of movement and then classification result become interfered?
   Is this the limitation of this tool? If yes, it is better to state the limitation of this tool or software in the paper. If not, please explain your argument and add such a condition in the paper. It is better to present the advantage and the limitation of this software in a simple way.

2. Please check again for the typos.

3. Overall this manuscript can be published with a minor revision.

Author Response

First of all, thank you very much for devoting your time to review our paper!

Please note that when we review your requests, our document (word) has NO line numbers. Therefore, we cannot state where the changes we made take place using this mechanism. On the contrary, and besides activating the change tracking mechanism in word, we added some side-comments stating the number of the reviewer the change is targeted at. Therefore, look for comments labeled "Reviewer 2" to find our answers in the document. Please, excuse the inconveniences this might cause to you.

Please, find below our answer to your comment.

How if the real displacement is in North-South direction. Does this tool will not detect such type of movement and then classification result become interfered?

Is this the limitation of this tool? If yes, it is better to state the limitation of this tool or software in the paper. If not, please explain your argument and add such a condition in the paper. It is better to present the advantage and the limitation of this software in a simple way.

We understand that your comment is questioning if a displacement movement along the North-South direction would be detected by the tool or not, because in the latter case it should be presented as a limitation of the tool.

The reasoning explaining why we didn't talk about decomposing the movement in the North-South direction is described below:

• The InSAR technique can detect and measure movements along one direction only, its line of sight.
• Combining the results of processing two different datasets for the same scene, where these datasets have different lines of sight, increases the precision of the results.
• Combining the processing of two datasets with different geometries makes possible to detect and measure displacements within the plane identified by the two lines of sight.
• KEYPOINT: the most common orbital configuration for SAR satellites (which have polar orbits and side-looking sensors) results in line of sights that are almost orthogonal to the North-South direction
• Therefore, InSAR processing with the most common SAR satellites orbital configuration have little sensibility to North-South displacements.
• CONCLUSION: This is a well-known limitation of the InSAR technique with the most common orbital configuration, not of the los2hv tool in itself.

Since most of the authors of the paper are SAR experts, it wasn't even considered to include an explanation about this issue.

However, and since the orientation of the paper towards software engineering is quite strong, we gladly accept your suggestion and some lines explaining this problem have been added to the paper. Note that explaining this limitation is so much detail as above would be excessive for the SAR-aware readers, since they already know about it. The more software-engineering oriented ones have, we believe, enough information in the new lines as to clarify why the East-West direction is (let's say), chosen.

These new lines are clearly identified by the change-tracking system plus a comment stating that the change was motivated by your suggestion.

Thank you very much for noticing this issue!

Reviewer 3 Report

Systems for the detection and risk assessment are an important element in the interpretation of the deformations measurements.
I appreciate the effort put into the development and implementation of system functions, however I think that the article should be completely rebuilt. It is currently in the form of a technical report or instruction manual without presenting a methodology.
The introduction need to be revised in order to clearly delineate the state of the art, the problem and the aim of the work. There should be done a deep literature study in the field of existing ground surface deformation analysis systems and used methodology.
Please also highlight the elements that distinguish the developed system from others.

Due to the wide range of topics, the methodology should be presented at least to the extent that allows understanding of the parameters presented in Figures 1,4,5,6. Replacing these Figures with description of the methods will contribute to a better understanding of the overall process of identifying deformation areas.
First of all, the methodology should be presented in terms of:
- deformation mapping,
- identification of ADA and application of this method in areas other than landslides (line 107 - the author indicates the use of the method mainly in landslide areas),
- quality assesment of the PSs displacements,
- determination of vertical and horizontal displacements on the basis of PSI data,
- PSs filtering and classification,
It is not possible to evaluate the operation of the system and the presented results due to the lack of a substantive description.
In present form the paper is description of the "black box" operation.

Other comments:
Please provide the method and criteria for classification (line 166- ADAfinder).
Explain the meaning of the abbreviations Th1-Th11 in Figure 3. How are the threshold values determined?
Discuss in detail how you take into account geological conditions in the form of geological maps in the classification of displacements and identification of the type of deformation.
Is the 2D geology model sufficient to determine the thickness of Quaternary sediments and for subsidence and landslides identification?
How does the system work when deformation factors overlap, e.g. subsidence, landslide or sinkhole occur simultaneously?

Author Response

First of all, thank you very much for devoting your time and effort to review our paper.

Please note that when we review your requests, our document (word) has NO line numbers. Therefore, we cannot state where the changes we made take place using this mechanism. On the contrary, and besides activating the change tracking mechanism in word, we added some side-comments stating the number of the reviewer the change is targeted at. Therefore, look for comments labeled "Reviewer 3" to find our answers in the document. Please, excuse the inconveniences this might cause to you.

Concerning your comments, we are afraid that that the goal of our paper has not been understood. Obviously, this misunderstanding leads to the cascade of rightful complains you state about our work.

You say that our paper presents no methodology, and this is absolutely true. On the contrary, the goal of our work is to describe a set of software tools that automate a couple of methodologies that were developed previously by the same group of authors (among others). This, furthermore, complies with the targets of the special issue for which this paper has been written: extended papers from a conference on geographical/hazard tools

This situation (no new methodologies but a new toolset relying on existing ones) is clearly stated in section 2, page 3. Literally:

"The four tools presented on this paper rely in preexisting methodologies, i.e., these applications have automated a set of procedures that already existed and were executed manually, step by step, using GIS (Geographic Information System) tools."

Moreover, in this same section & page, the references describing the methodology on which ADAfinder relies are clearly given:

"In the case of ADAfinder, the methodology to perform the identification and assessment of ADAs was presented in [26] and [27] back in 2017 and 2019 , respectively. In these works, the authors explain in detail the procedure to identify the active deformation areas and to assess their quality.
(...)
The concept of ADA has been exploited to estimate geohazard-related risk (mainly landslides) in different European environments, see [28, 29, 30]."

And these are there references:

26. Barra, A.; Solari, L.; Béjar-Pizarro, M.; Montserrat, O.; Bianchini, S.; Herrera, G.; Crosetto, M.; Sarro, R.; González-Alonso, E.; Mateos, R. M.; Ligüérzana, S.; López, C.; Moretti, S. A methodology to detect and update Active Deformation Areas based on Sentinel-1 SAR images. Remote Sensing, 2017, 9(10):1002.
27. Tomás, R.; Pagán, J.I.; Navarro, J.A.; Cano, M.; Pastor, J.L.; Riquelme, A.; Cuevas-González, M.; Crosetto, M.; Barra, A.; Monserrat, O.; Lopez-Sanchez, J.M.; Ramón, A.; Ivorra, S.; Del Soldato, M.; Solari, L.; Bianchini, S.; Raspini, F.; Novali, F.; Ferretti, A.; Costantini, M.; Trillo, F.; Herrera, G. & Casagli, N. Semi-Automatic Identification and Pre-Screening of Geological–Geotechnical Deformational Processes Using Persistent Scatterer Interferometry Datasets. Remote Sensing, 2019, 11, 1675.
28. Solari, L., Barra, A., Herrera, G., Bianchini, S., Monserrat, O., Béjar-Pizarro, M., Crosetto, M.; Sarro, R.; Moretti, S. Fast detection of ground motions on vulnerable elements using Sentinel-1 InSAR data. Geomatics, Natural Hazards and Risk, 2018, 9(1), 152-174.
29. Solari, L., Bianchini, S., Franceschini, R., Barra, A., Monserrat, O., Thuegaz, P., Bertolo, D.; Crosetto, M.; Catani, F. Satellite interferometric data for landslide intensity evaluation in mountainous regions. International Journal of Applied Earth Observation and Geoinformation, 2020, 87, 102028.
30. Aslan, G.; Foumelis, M.; Raucoules, D.; De Michele, M.; Bernardie, S.; Cakir, Z. Landslide Mapping and Monitoring Using Persistent Scatterer Interferometry (PSI) Technique in the French Alps. Remote Sensing, 2020, 12(8), 1305.

Concerning ADAclassifier and THEXfinder, the fact that these tools rely on a existing methodology is stated too in section 2, page 3. Again, literally,

"ADAclassifier and THEXfinder rely on a methodology developed by the UA that is under constant research –although mature enough as to be automated, which eases the experimentation process and helps to improve the methodology itself."

The reference to the underlying methodology is given at the end of the same paragraph:

"For more details about the methodology on which these two tools rely, see [32]*."

* The reference number was 31 in the version of the paper you read. This number has changed to 32 after adding extra references that served to back additional statements inserted in the text to solve the requests made by other reviewers.

And this is the reference:

32. Tomás, R.; Cano, M.; Pastor, J. L.; Riquelme, A. (2018). Automatic classification of Active Deformation Areas. MOMIT project technical note. Available online: https://www.momit-project.eu/sites/default/files/2018-12/MOMIT_technical_note_%20Automatic%20classification%20of%20Active%20Deformation%20Areas.pdf (accessed 3 July 2020).

In the case of los2hv, the method implemented by the tool is described in section 2.3, pages 10-11. This is how the description of the algorithm starts:

"los2hv performs a tessellation of the whole area covered by the PSs included in the two input files. The size of the tesserae (grid spacing) is decided by the user. As a consequence, each PS belongs just to one of the resulting tesserae...."

The remaining of the paragraph completes the description.

In page 11, the reference to the mathematical apparatus used to compute the East-West component is also given:

"The resulting amount, measured along the LOS, is then converted to East-West horizontal and vertical components according to the formulae described in [33]*."

* Previously, 32. Changed due to the extra references.

and this is the reference:

33. Notti, D.; Herrera, G.; Bianchini, S.; Meisina, S.; García-Davalillo, J. C.; Zucca, F. A methodology for improving landslide PSI data analysis. International Journal of Remote Sensing, 2014, (35):2186–2214.

The whole set of previous literal paragraphs and references should clarify why our paper presents no methodology at all but just the tools that automate these.

It is true that the tools are described only stating what are their target, inputs and outputs. Although this may be considered as a "black box" approach, it is also true that references to the whole set of user guides, where full details are provided are given.

For ADAFinder, section 2.1, page 5 (just before the heading for section 2.2):

"For a detailed description of the ADAfinder tool, please refer to its user guide ([34]*)."

* Previously, 33. Changed due to the extra references.

and here's the reference:

34. Navarro, J. A.; Cuevas, M. ADAfinder - A tool to automate the detection of Active Deformation Areas. MOMIT project technical note. Available online: https://www.momit-project.eu/sites/default/files/2018-12/MOMIT_technical_note_ADAfinder%20documentation.pdf (accessed 3 July 2020).

For ADAclassifier, in page 7, section 2.2, this can be read:

"The ADAclassifier tool is fully described in its user guide ([35]*)."

* Previously, 34. Changed due to the extra references.

And the reference is this one:

35. Navarro, J. A.; Cuevas, M.; Crosetto, M. ADAclassifier - Classify ADAs automatically. MOMIT project technical note. Available online: https://www.momit-project.eu/sites/default/files/2018-12/MOMIT_technical_note_ADAClassifier%20documentation.pdf (accessed 3 July 2020).

The user guide for los2hv is referenced in page 11, section 2.3, just before the heading of section 3:

"Please refer to the complete user guide of the los2hv tool [37]* for more information."

* Previously, 35. Changed due to the extra references.

And this is the reference:

37. Navarro, J. A.; Cuevas, M. los2hv - Compute horizontal and vertical components of the movement. MOMIT project technical note. Available online: https://www.momit-project.eu/sites/default/files/2018-12/MOMIT_technical_note_los2hv%20documentation.pdf (accessed 3 July 2020).

Since (a) the methodologies backing the tools have been fully referenced - and described there - and (b) the software tools are explained in detail in their corresponding referenced user guides (including input, outputs, meanings of the parameters, etc.) no further details beyond those necessary to describe the toolset in a coherent manner are given in the paper.

We thus believe that the paper is fully compliant with the target of the special issue it is going to be included in: covering a conference on geographical/hazard tools.

Nonetheless, we have taken into account two of your comments that are not affected by the misunderstanding we talk about.

The first one refers to the use of the labels Th1 to Th11. Now, the caption of Figure 3 explain these (page 7).

The second one, and although details on how the quality assessment and outlier filter work were already given in reference [26], we have added an extra statement in page 4 to state clearly once more where information about these methods may be found.

Finally, we regret not being able of answering the remaining comments you made, because these, we believe, are caused by the erroneous interpretation of the target of our paper. We hope that the explanations here given will serve to reconsider your opinion about our work.

Thank you very much.

 

Reviewer 4 Report

The manuscript presents the set of tools integrated by ADAfinder, losh2v, ADAclassifier and, to a esser degree, THEXfinder.These applications are targeted at automating the identification and classification of ADAs. All of them rely on methodologies that had been thoroughly used in real use cases prior to their materialization as software modules. Their implementation is based on well proven techniques. Suggesting to replace Some simple figures with text descriptions，such as figures 10，11 and 12.

Author Response

First of all, THANK YOU VERY MUCH for your effort reviewing our paper and for your suggestions as well!

Please note that when we review your requests, our document (word) has NO line numbers. Therefore, we cannot state where the changes we made take place using this mechanism. On the contrary, and besides activating the change tracking mechanism in word, we added some side-comments stating the number of the reviewer the change is targeted at. Therefore, look for comments labeled "Reviewer 4" to find our answers in the document. Please, excuse the inconveniences this might cause to you.

Concerning your comment:

"Suggesting to replace Some simple figures with text descriptions，such as figures 10，11 and 12."

we have decided to remove only figure 10, since its contribution (a graphical depiction of the interface of QGIS where the buttons to start ADAFinder & ADAClassifier are shown) is very small and it can be considered as negligible.

It has been eliminated as well as a very short sentence that referenced the aforementioned figure.

On the contrary, we have decided to keep both figures 11 & 12.

In the case of figure 11, it shows very clearly what the syntax of options files is. This kind (format) of files are used both as option files themselves as well as read-map ones, which make them very important for the ADATools. Figure 11 also provides with an example of the kind of options (contents) included there.

Figure 12 is the only example of a read-map file; all the "syntactic sugar" (such as comments explaining the contents of the file) present in Figure 11 has been removed so it has been reduced to its minimalistic expression. We thus think that eliminating the figure would complicate the explanation about the contents of the points read-map file given in a previous paragraph, which we reproduce here literally:

"The meaning of the pairs or labels and values in Figure 12 is the following:
• the x-coordinate of the PSs must be read from column 5 in the .dbf file,
• the column to read to obtain the y-coordinate is the sixth one,
• the velocity may be found in column number 9 and, finally,
• the set of values making the time series start at column number 11 and there is a total of 50 of these values."

So these are the reasons why we believe that keeping figures 11 & 12 make sense for improving the understandability of the paper.

We hope that you share our opinion and decision now that we (believe that we) have motivated it.

Again, thank you very much!