Using InSAR Time Series to Monitor Surface Fractures and Fissures in the Al-Yutamah Valley, Western Arabia

Round 1

Reviewer 1 Report

The article reports an analysis of InSAR displacement time series and velocities for a specific region of the Western Arabia, extracting and discussing peculiar local deformation occurring within the Al-Yutamah Valley, where surficial fractures and fissures can generate a high natural hazard for infrastructures and anthropogenic settlements. The paper presents the issue in an exhaustive way allowing the reader to understand the problem. The methodology and data used are also well described. However I think that the article should be modified in the last part where the specific analysis and discussions are piecemeal and sometimes inaccurate. Moreover a Conclusion section is missing and I suggest to add it in order to highlight the main results and a conclusive point of view of the proposed work.   

In the following there are reported some issues I ask you to address, while at the bottom there are some specific comments referring to line numbers, paragraphs and figures

- Along with the several corrections the authors perform within the InSAR time series analysis, there is one correction I think you should better describe and justify, i.e. the long wavelength correction. I understand you are interested to local deformation signals but the long wavelength ones are not only due to InSAR technological artifacts, but are also representative of the ongoing tectonic processes, that are not negligible at all in this region and actually are not an error. In order to remove correctly only the technological errors due to ionosphere and orbital effects, the most common procedure is to compare the resultant InSAR velocity map with a GNSS one, allowing to tie them to the same reference frame and avoiding to refer to an arbitrary local point that can be unstable. However, I guess this kind of comparison could not be possible for you for several reasons (because otherwise you would have done it) and I do not pretend that you do such kind of correction, but what I would like you do is to add a more in-depth discussion about this choice.

- Since you have referred several times along the manuscript to the various signals composing the resultant displacement time series (linear, annual, semi-annual, step components), I suggests to add in the supplementary some example of most meaningful time series to show the signals you model, not only the seasonal ones but also the “jumps” occurred in 2018 (see the specific note below for the term “jump”). I see that you display the offset for the 2018 event in Figure 5 but I suggest to detrend all the time series and not only the first one, because it is more helpful to highlight the offsets, and you can refer to the different velocities using the map in Figure 3

- As I said at the beginning, I believe that the last part of the paper (paragraphs 3 and 4, i.e. Results and Discussion) is not well organized and it should be revised. In the Results paragraph I have found not only inaccurate and misleading description of your results but also possible approximate explanations of what you are showing (such as the correlation with the rainfalls that is not so strong for the whole time series), even in the wrong position like the captions of the Figures. Moreover you talk about of some long wavelength signals both in the seasonal displacements (L 210- ongoing) and in the displacement time series themselves (Figure 5D) that are not so evident for the reader and they do not provide any important contribution to the analysis on a local scale, that is the focus of the work. Please, try to reorganize the last part of the manuscript in order to well separate the Results (where you present the obtained velocities and displacements and ancillary data highlighting both common features and discrepancies, NOT explanations) with the Discussions (where for the presented Results are provided possible explanations) improving in the meanwhile the clearness of your arguments (see the specific notes below).

- Please, do not use the term “anomaly” because it is not a scientific term. It does not mean anything in the specific, while you should define in a more accurate way what do you consider as anomaly (Which quantity? How much? Where? How?)   

- Last but not least: I propose all these corrections because I think the paper deserves to be improved, the idea is good and also the methodology and the dataset shown. For instance I appreciated a lot the gradient estimate of the velocity field because it is a good quantification of the specific features of the local deformation pattern. Just be more careful in defining this quantity in an appropriate way: the gradient is a mathematical operator providing the horizontal variation of a physical quantity, that in this case is the velocity field, do not define it as the horizontal displacement or the cumulative displacement, because it is wrong.

 

Specific notes:

Par. 2.3 from L 117: for the regional scale solution, specify along which orbit have been acquired the images you use

Figure 1c: I guess you mean Railway and not Trailway (same for Figure 7)

L 132-133: “The RMS residuals generally increase during the summers, which can be seen 132 clearly in the years 2017 and 2019” This is not really clear from Figure S1. Maybe you can try moving (o reducing) the legend covering the diagram and add a cumulative diagram for winter and summer to stress the difference of RMS between the two periods.

L 144-147: please provide reference for this correction

L 148-149: why do you do that? Which kind of errors you want to remove by this step? You describe this step in a second moment but you should add a reference to the specific section.

L 151: add “written” in PYthon, or in Python “language”

L 171: “The RMS residuals are generally highest during the summers” as said before, if you add a cumulative graph, this claim can be more clear to the reader

L 179: see the main comment above and for lines 148-149. At least provide an explanation for this correction.

L 211-214: which anomalies do you refer to? The observed phase residual magnitude is not shown anywhere so for the reader it is not possible to follow your reasoning. When you say “These observed anomalies are not validated nor well understood and are thus simply removed from our velocity” do you mean the annual and the semi-annual signals? If it is so, please refer to them with these terms. Or are you referring to removed pixels? However the seasonal signals cannot be considered an anomaly, these are quite common signals recorded worldwide by most geodetic techniques. I do not understand what is the anomaly, clarify this paragraph.

L 216-219: instead of using the term “anomalies”, that is not correct, I suggest “displacement rate variations” or “velocity variations”

Figures 2 and 3: the LOS arrow is wrong, the ascending satellite look direction is the same you show but in the opposite way; add in the caption what is the black square (it is missing also for the figures of the supplementary).

Only Figure 3 left panel: add a square to indicate the zoom made in figure 4; right panel: what is the difference between the gray and blue points? (same for Figure 5)

L 259-261: the term “jump” is not correct, please substitute with offset or displacement (check in the whole text)

Figure 4: rewrite the caption as follows: ‘(Left) Displacement map for Nov 18th, 2018  event after smoothing using a low pass gaussian filter. The distinct pattern is in agreement with the continued subsidence that is shown in Figure (4). A,B,C, and D denote the locations for time series shown in figure 5. (Right) Standard deviation of the displacement map. The estimated error is insignificant as compared with the observed displacement magnitude.’ Moreover, if I have understood well, in the two maps you are not showing velocities in cm/yr but the displacement due to the 2018 event, thus I suppose what it is shown in the maps should be displacements in cm (correct the palettes).  

L 272-onwards: there are some passages that you made for all the time series that are described in a specific section of one time series. Extrapolate all the common passages (e.g. detrending, rainfall estimate, etc) at the beginning before the specific sections for each local region.

L 283: which first region? Be more specific.

Par 3.2.2: this paragraph is a bit confusing and there are some considerations I think you should add or at least clarify:

• Why do you detrend only the first time series? This is misleading if you want to stress the correlation of the offsets with the rainfalls
• Looking at the picture I do not observe strong correlation for all the rainfall events. This can be due to the fact that rainfall data are extracted from a gridded dataset and the local accuracy could be missing. Add some consideration about this point.
• When you talk about subsidence and negative displacements due to rainfalls I understand what you mean, but sometimes you measure positive displacements (uplifts) in correspondence of the same process, how do you explain that?
• The comment of Figure 5D has no meaning (L 298-301), what kind of subsurface processes are occurring? Again, what do you mean with displacement anomaly?

Figure 5 caption: remove all the sentences related to possible correlations of displacements and rainfalls and put them into the main text; again, explain the difference between blue and gray dots

Par 3.2.3: nice idea to use the gradient amplitude of the velocity field to show the main features of the deformation. Just few corrections:

L 317: not transactions, maybe cross-sections?

L 317-321: the three sections do no show the horizontal displacement but the gradient amplitude of the LOS velocity field, or the horizontal variation of the LOS velocity field. Talking about horizontal displacement or apparent horizontal displacement is misleading. Please correct all the sentences.

L 322-324: “the cumulative apparent horizontal displacement threshold reaches 15 cm for the period between 2014-2021” I do not understand, what are you referring to? To which specific figure? Why threshold? Maybe do you mean a maximum value?

Figure 6: correct all the labels with “gradient of the LOS displacement rate (cm/(yr*pixel))” or “gradient of the LOS velocity (cm/(yr*pixel))” and correct the same mistake also in the caption.

L 355: correct the citation format

Author Response

Dear Reviewer,

Thanks you so much for your review, please find your comments reflected on our manuscript. Your insightful comments have significantly improved our work.

Regards,

Author Response File: Author Response.pdf

Reviewer 2 Report

This well-written manuscript provides another pertinent example of InSAR monitoring capability for geological and subsidence mapping. The authors seem to have proceeded with proven methodologies, applied required processing refinements with well know InSAR processing tools. Consequently, I recommend this manuscript for publication after minor revision. Below are suggestions and questions that might improve the quality of the document. Will let the authors and the editor judge their pertinence.

 

Comments/Questions:

-P5: For 16m high resolution, InSAR pairs were limited to 3 neighbors (so between -24 days and 24 days). For the Al-Yutamah Valley fracture in 20I8, I wonder if having extend the number of neighbors (ex. from mid 2018 to mid-2019), would have better defined the fracture and ensure there are phase unwrapping issue. 

-P6, L219: "The uncertainty of our results increases proportionally with distance away from the chosen spatial reference point..." Why not having taken multiple reference points?

-P6 section 2.4. Ground Truth Data: No differential GPS were performed to quantify the fracture displacement? Only visual interpretation on filed? It would have been nice to compare GPS measurements with InSAR results.

- Figure 3 and 5: What is the difference between gray and blue dots in the displacement figures?

- Figure 3: could you draw a box highlighting the region corresponding to figure 4?

-Figure 5:  Rain events are histogram gray bars and accumulation is shadow blue?. Please clarify

- No formal conclusion section?

Author Response

Dear Reviewer,

Thanks you so much for your insightful comments and feedback. Please find them reflected on our revised manuscript. 

Regards,

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper applied an InSAR time series analysis of Sentinel-1 data to cover wide areas of western Arabia, to investigate domains of earth fissuring and fracturing and to monitor their activities over time. This paper correlated the observed deformation fields with surface fractures and fissures, discuss the possible underlying mechanisms, and forecast future hazards. This paper does some work on InSAR data correction, but still have some problems. Major revisions are recommended, the main problems are as follows:

1. In lines 35-37, it is mentioned that a variety of technologies and methods have been studied in the experimental area, including InSAR technology. So, what is the difference of methods and conclusions between InSAR result in this paper and other studies?
2. Please supplement the specific information and situation of sentinel-1 data in the experimental with pictures.
3. It is suggested to elaborate on the proposed method, which is too brief at present.
4. According to the results in Figure 2, the deformation rate diagram is obtained by processing sentinel-1 data of multiple frames. How to mosaic InSAR velocity of different frames?
5. The image of the local area in Figure 2 looks smooth, is it interpolated?
6. Please introduce the four main correction methods of InSAR data in detail.

Author Response

Dear Reviewer,

Thank you for your comments. Please find our clarifications attached.

Regards,

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Thanks for having addressed all the comments and suggestions.

Reviewer 3 Report

The authors answered the questions and made corresponding changes to the paper. It is recommended to accept this paper.