Risk Evaluation of the Sanalona Earthfill Dam Located in Mexico Using Satellite Geodesy Monitoring and Numerical Modeling

Round 1

Reviewer 1 Report

• As part of this investigation, two geodetic satellite techniques and mathematical modeling are used to analyze the risk of the dam-reservoir system at Sanalona Dam in Mexico. Using InSAR technology, the PSI technique has been applied to measure millimeter displacements of the dam crest. An analytical model based on the finite element method (FEM) and GNSS systems is compared for monitoring radial displacements in dams using the InSAR technique. Considering the reliability analysis and the small probability of failure, it is concluded that the Sanalona dam does not pose a considerable risk to society.

  As a reviewer of this study, everything was going well, and there were some suggestions to improve the work so that it would be excellent. It can be said that everything is good from the beginning to the end until you consider the accuracy analysis part (4.2). There is a sharp drop, and I am in a dilemma. Before I address my suggestions and comments below, I raise some questions and reasons here:

  When you have only one GPS station in the area, you should have done the InSAR processing in the surrounding areas to compare the time series of the corresponding points obtained from it with that GPS station and say that the accuracy and evaluation of the measurements are valid. But the problem is that your GPS point is located in a green area, which could tell you this point; maybe you should have used another processing method, such as SBAS, which will also give you the DS points in this area, and your evaluation will be done thoroughly. Also, the number of PS points that you considered on the crust and crown of the dam is small, while you know that when benchmarks are used to evaluate the displacements of the dam, more than a few tens are fixed in the dam. I still believe this is a valuable and content-rich study, but they should be compared in the evaluation section by matching the PS points near the GPS station so that the work can be confirmed, or you should use another InSAR method, such as SBAS. Below you can find some comments to improve this study:

  • Comment #1: In the abstract section, the authors did not explain the 'probability of failure,' while one of the main parts of this study related to it. You can expand it and make some numerical results more meaningful (as you know, you put these words in your keywords.)  
  • Comment #2: The English language is sufficient for the peer-review process. However, for accepting and publishing, it must be changed which some of the listed below:

  Sub Comment: Before resubmitting the paper, please pay minute attention to the noun you must mention,"' the,' 'an', 'a' or 'S.'" use the common word for better understanding and prevent to use of repetitive words.

  Sub Comment: Two figures (8 and 11) are unreadable without quality and not clear enough to read directly. (Change the figures to larger with high quality.) you prepare figures for better understanding; now, your figures are so small and unreadable with such a bad quality.

  • Comment #3: Introduction section, the references are not completed and could be more. You can use old and primary references for some conventional facts. Also, when you put an important fact, such as on page 2, lines 68-70, you must add references. Moreover, sentence lines 82-83 lack sufficient references.

  Similar to the abstract, 'quantify the probability of failure' is not discussed and lacks references. It would help if you prepared some description about it.

  •  Comment #4: In the Methodology section, it is better to remove the sentence on page 3, lines 52-57 ('this issue will be discussed in Section 3?'). Also, in the subsection of 2.1, you can open this issue and explain more because it is an important part of your paper.
  • Comment #5: In the Study Area section, you could use some geological setting as a text or figure with a description, and also, I believe that it is necessary to prepare time series of rainfall data as a figure and description in this part (you can use reanalysis of ERA5 data.) in addition, the figure of this part can be modified: what is the black circle? 
  • Comment #5: In the Methodology section, lines 174-177, add valuable references to illustrate the vertical displacement of the dam is insignificant.

  You do not describe the 'probability of failure' and say in the literature, and we describe it?

  In figure 4, K_R σ_s or K_R σ_R?

  • Comment #6: In the result section,  Accuracy analysis of radial displacements is a problem that I mention before comments.

Author Response

We appreciate your comments. We tried our best to follow your suggestions.

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript entitled “Risk Evaluation of the Sanalona Earthfill Dam Located in Mexico Using Satellite Geodesy Monitoring and Numerical Modeling” by J. René Vázquez-Ontiveros is an important work to evaluate the risk of failure of the dam. I think the manuscript is worth publishing in Remote Sensing, but it has some important problems, especially in treating the InSAR analysis results. Therefore, I recommend that the authors should revise the following points.

1.     The authors should show the surface displacement or velocity maps estimated by their MT-InSAR analysis. This is because the areal extent of displacement or velocity is also important in indicating the plausibility of the results. If the authors estimate the displacement or velocity map, a comparison between GNSS and InSAR displacements at the same point becomes possible. I believe it is useful to ensure the plausibility of both results.

2.     When combining the LOS displacement, the authors considered only horizontal displacements. However, the plausibility of this idea is unclear. Firstly, the authors should show the quasi-upward and quasi-eastward components as the result of the typical combined result.

3.     The authors estimated the radial components of the displacements of the dam. However, the radial direction of the dam is almost the N-S direction. Even if considering only horizontal displacements are allowed, the sensitivity of the N-S direction of the InSAR displacements/velocities is not good. If the authors want to discuss the radial components, they should show the error of the component.

Author Response

We appreciate your comments. We tried our best to follow your suggestions.

Please see attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

The manuscript titled “Risk Evaluation of the Sanalona Earthfill Dam Located in Mexico Using Satellite Geodesy Monitoring and Numerical Modeling” used different geodetic and numerical modeling techniques to analyze the Sanalona Dam Risk. Overall, this study is significant for the stability and running of the dam, and the methodology can be used for risk evaluation of other long-term operational dams. However, in my opinion, there are still some issues that should be clearer and provide more information as follows:

Major Comments:

1) In the PS-InSAR or so-called MT-InSAR processing, the authors identified corresponding PSs in ascending and descending orbital results, please provide detailed approach information on how to determine these points.

2) As the authors mentioned in the GNSS displacements section, the site station only recorded data transmitted by the GPS constellation, therefore, using the ‘GNSS’ in many sections is not suitable.

3) The authors processed the GPS data under the IGb14 reference frame, whereas the InSAR data is often processed under the WGS-84 reference frame. How did the authors think about the difference between the reference frames?

4) The authors only analyzed the precision of external coincidence, how are the internal ones of GPS and PS-InSAR?

5) As we all know that the PS-InSAR displacement is relative, did the authors calibrate the PS-InSAR data?

6) For the displacement transformation, the authors assumed that there is not any vertical subsidence for the dam, which is a good and reliable assumption for the extraction of displacement components. However, one point that the InSAR LOS displacement is the projection of vertical, E-W and N-S components should be noted. In the manuscript, the authors mentioned the InSAR results as E-W, this statement is not rigorous, especially in this case  the N-S displacement is predominant.

As the dam is curved, did the authors use unique azimuth or multi azimuths for different sections of the dam when transform the InSAR horizontal displacement into the radial direction?

7) The FEM technique is another technique used in this study, why the authors didn’t use the geodetic data as the boundary constraint when determining the model parameters? I suggest the authors combine the geodetic data and FEM, but not mandatory.

Minor Comments:

1) Line 36: I think the “the” is repeated.

2) Figure 6: Suggest the displacement vectors of PSs should be plotted on the figure.

3) Figures: The resolution of all the figures is too low, and the information is not clear.

Author Response

We appreciate your comments. We tried our best to follow your suggestions.

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Thank you for addressing my previous comments. I think the manuscript was improved from the earlier version.

But, in Figure 6, I cannot understand why the PS points in decomposed results were increased from the original ascending and descending results. Please explain your decomposing method in the manuscript.

Author Response

We appreciate your comments and apologize if our explanations were not clear or if our Figure 6 in the revised version led to confusion.

The decomposition procedure is done by the software SARPROZ. If ascending and descending datasets are available, the software allows us to automatically decompose the movement into the vertical and horizontal E-W directions. First, the software computes possible shifts between the ascending/descending datasets estimating the offsets in longitude, latitude, and height. Precise offsets are then estimated using points with distances closer than the range specified as a maximum planar distance and with height differences closer than a maximum height distance. For processing the ascending/descending pairs, a maximum planar distance and maximum height distance are used as well to identify the corresponding points in the two datasets. Once offsets have been estimated and ascending/descending pairs identified (in the case of ascending and descending dataset pairs), results can be combined, plotted, and exported. Several options are available for that. For example, one can plot all points, ascending/descending pairs, or a grid created on overlapping areas.

In the revised version we provided the decomposition plots (vertical and E-W) in Figure 6 (6c and 6d) using a grid of 15 m. Certainly, the grid size may lead to confusion about the existence of more points as they are somewhat scattered. We apologize if this has led to any confusion. In order to improve the visualization, we have selected a 5x5 m mesh superimposed on the common points, improving the visualization of Figure 6 (6c and 6d).

Following the requirement of the reviewer, we have added the following text in lines 188-198 of the manuscript explaining the decomposition procedure: “First, the software computes possible shifts between the ascending/descending datasets estimating the offsets in longitude, latitude, and height. Precise offsets are then estimated using points with distances closer than the range specified as a maximum planar distance and with height differences closer than a maximum height distance. For processing the ascending/descending pairs, a maximum planar distance and maximum height distance are used as well to identify the corresponding points in the two datasets. Once offsets have been estimated and ascending/descending pairs identified (in the case of ascending and descending dataset pairs), the decomposed movement can be computed, plotted, and exported. Several options are available for that. For example, all points, ascending/descending pairs, or a grid created on overlapping areas can be generated.”

Author Response File: Author Response.pdf