Rheology of the Northern Tibetan Plateau Lithosphere Inferred from the Post-Seismic Deformation Resulting from the 2001 Mw 7.8 Kokoxili Earthquake

Round 1

Reviewer 1 Report

Dear authors,

my questions were addressed properly. Therefore, I am recommending the paper to be published.

Best regards

Author Response

Dear reviewer,

Thank you for your agreement about the publishing! Thank you for your valuable suggestions. We highly appreciate them! Best regards, Xiaoran.

Author Response File: Author Response.docx

Reviewer 2 Report

Dear Authors,

I have thoroughly read your article titled: “Rheology of the Northern Tibetan Plateau litosphere inferred from the post-seismic deformation resulting from the 2001 Mw 7.8 Kokoxili earthquake” submitted again to Remote Sensing.

Every year, the world is hit by a slew of powerful earthquakes. The research on the impact of such earthquakes on the land surface is primarily focused on the short-term monitoring of land surface displacements up to a few weeks after the seismic event. The relaxation of the rock mass, on the other hand, can last for years after the earthquake. Nonetheless, there has been little research on long-term relaxation following an earthquake. As a result, the research presented has the potential to be innovative and appealing to a broad audience.

The authors conducted the viscoelastic relaxation simulations using the source parameters of the Kokoxili earthquake. Unfortunately, after reading the manuscript, I get the impression that it is intended to be more of a technical note than a research article. The study only looks at one case. There is no broader reference to the literature on the subject of long-term relaxation of the rock mass following earthquake occurrence. Furthermore, many chapters are written in a very generalized manner, with no detailed data that was used in the modeling of the phenomenon under study. The numerous uncertainties in the research methodology must be addressed significantly.

The manuscript, in its current form, should, in my opinion, be extensively revised by the authors before it can be processed further. At the moment, the manuscript presented is not suitable for publication in Remote Sensing.

Specifically:

• This chapter, in my opinion, is far too brief. The authors concentrate solely on one analyzed shock. However, in the article's introductory chapter, comprehensive information about the state of the art in the field of research on rock mass relaxation following large-scale earthquake occurrence should be provided. As a result, this chapter must be improved so that it covers the global scale rather than only describing the issue under study on a local scale.
• The geology description is inadequate. There is no information about the rock layers, their parameters, the hypocenter of the tremors, seismicity, or the tectonics of the study area. This chapter needs to be greatly improved. Furthermore, Figure 1 is unreadable; it should include information about the location of the research area in Asia, as well as more detailed geological, seismic, and tectonic information.
• I'm not sure what InSAR processing method you used in your study. It's also unclear whether you ever converted your LOS movements to vertical ground movements. This is an inadequate description. Please make it better.

 

Best regards,

Reviewer

Author Response

Dear reviewer,

Thank you for your valuable suggestions!

We answered the comments one by one. Please see the attachement.

And we also made the revised manuscript get polished by the professional polishing company.

Best regards

Author Response File: Author Response.pdf

Reviewer 3 Report

Recommendations for paper improvement:

The English text should be corrected by a native English editor, because many sentences are not correct written and the text is not clear.

The viscoelastic relaxation is very slow in time and it needs long-time monitoring. The determined parameters of viscosity by using 1700 satellite images, covering more than 6 years and processing by very accurate modern technique INSAR, based on satellite interferometry, yields valuable results about mechanic properties of the crust and upper mantle. These results are important for a large number of specialists in the field of Geology, Geophysics, Geodesy and Remote Sensing. The paper will be ready for print after English text edition.

Author Response

Dear reviewer,

Thank you for your valuable suggestions. We highly appreciate them!

We made our revised manuscript get polished by the professional polishing company.

Thank you again for your kind and helpful suggestions!

Best regards

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Dear Authors,

Thank you for your thorough replies to my remarks.
In my opinion the manuscript has been substanitally improved and now is suitable for publication in Remote Sensing.

Kind regards,

Reviewer 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Lv et al. used 6 years of Sentinel-1 InSAR data observed from October 2014 to July 2020 in northern Tibet to study postseismic deformation of the 1997 Mw7.6 Manyi and 2001 Mw7.8 Kokoxili earthquakes. Assuming the InSAR derived deformation field was caused by viscoelastic relaxation of the lower crust and upper mantle, they estimated the viscosities of the in situ lower crust and upper mantle. They claim that their results are consistent with that of previous studies and that viscoelastic deformation beneath the Manyi and Kokoxili regions have almost reached the stable state. I however find that the study is seriously flawed and the results are not credible. I therefore cannot recommend for publication of the paper.

The most serious problem this study has is that it ignores secular motion in crustal deformation modeling.  It is common knowledge that the Tibetan plateau deforms interseismically, particularly around major faults such as the Kunlun fault. One might ignore secular motion in postseismic deformation study only if the signal of secular motion is much smaller than that of postseismic deformation, which is true during the early stage of postseismic deformation of the Manyi and Kokoxili earthquakes. This study, however, used data collected 17 and 13 years after the two quakes, by then the postseismic deformation had decayed significantly and the interseismic deformation should be at least comparable in amplitude to postseismic deformation. Using these data as model constraints but ignoring the contribution of interseismic deformation would then seriously bias the modeling result. I don’t see a way to fix the problem using the datasets the authors have produced, since the interseismic and postseismic deformations are highly correlated in this dataset, both spatially (wrench style motion across a fault) and temporally (quasi-steady state motion).

The InSAR data they present also have some problem. I don’t understand why the processed data are shown as ‘cumulative postseismic deformation’ (Figures 3 and 4) but not as the deformation rate derived from fitting the 6 years of LOS data. The data also show strong spatial scattering, which, according to the authors, may be correlated with the regional topography.  This suggests that significant errors are involved in the LOS data, due to either inaccurate DEM model correction or atmospheric disturbance, or both.

Author Response

Dear Reviewer,

Thank you for your valuable comments. We response your comments in the word file. Please see the attachment.

Best regards,

Author Response File: Author Response.pdf

Reviewer 2 Report

Dear Authors,

Many strong earthquakes occur every year in the world. The research on the impact of such earthquakes on land surface is focused mainly on the short-term monitoring of land surface displacements up to a few weeks following the seismic event. However, the relaxation of the rock mass can last for years after the earthquake occurrence. Nevertheless, there is relatively little research on relaxation in the long-term after the earthquake occurred. Therefore, the research presented has the potential to be innovative and interesting to the general audience.

In the study, you used the source parameters of Hua and Wang for the Manyi and the Kokoxili earthquakes, respectively, to undertake the viscoelastic relaxation simulations. Unfortunately, after reading the manuscript, I get the impression that it is more of a technical note than a research article. Research only focuses on two cases. There is no broader reference to the literature on the subject, the issues of long-term relaxation of the rock mass after earthquake occurrence. Additionally, many chapters are written in a very general way, without providing detailed data that was used in the modelling of the phenomenon under study. The numerous uncertainties in the research methodology must be significantly corrected.

Please find below detailed comments.

Good luck,

Reviewer

 

Introduction:

In my opinion, this chapter is too short. The authors focus only on the two analyzed shocks. However, in the introductory chapter to the article, there should be comprehensive information about the state of the art in the field of research on rock mass relaxation following large-scale earthquake occurrence. For this reason, this chapter must be improved, so that it covers the global scale and does not describe the issue under study only on a local scale.

 

Geological settings:

The geology description is very poor. There is no information about rock layers, their parameters, location of the tremors hypocenter, seismicity and tectonics of the study area. This chapter requires significant improvement. Besides, Figure 1 is not readable - it should contain information about the location of the research area concerning Asia, as well as more extensive geological, seismic and tectonic information.

 

InSAR data and processing methodology:

Lines 93-94: In terms of long time-series data scarcity of about less than 20% is not a problem.

Lines 86-103: I do not know what InSAR processing technique you used in your study. It is also unknown whether you finally converted LOS movements into vertical ground movements. This description is too weak. Please, improve it.

 

Modelling approach:

Lines 118-124: What exactly parameters were used in modelling with the RELAX software? What is their source and what is the uncertainty of the data? Could you provide the reader with a Figure showing your model?

 

Results:

Lines 161-163: What is the reason for a significant long-wavelength noise in ascending data?

Figure 3 and Figure 4 must be enlarged as it is hard to read their content (especially grid mesh), lack of scale and projection, please improve the quality of these figures.

Lines 180-191: Where do those layers come from? What is the source of such data? This must be described in details in the data section.

Lines 201-210: I am not sure if this is correct to compare LOS movements with the vertical displacements induced by the earthquake occurrence. The reason for discrepancies for both descending and ascending orbits is simple – different look angle and different type of the observed movements. Therefore, you should have converted LOS into vertical movements and make a comparison.

Figure 7 and Figure 8: the same comment as for Figures 3 and 4.

 

Discussion:

No comments.

 

Conclusion:

In my opinion, this chapter is too short. There are only basic conclusions retrieved from the manuscript presented. There is a lack of the general overview on a studied issue nor plans for future research and implementation of modelling results as well.

Author Response

Dear reviewer,

Thank you for you valuable comments. We response to you comments in the word file. Please see the attachment.

Best regards,

Author Response File: Author Response.pdf

Reviewer 3 Report

The authors evaluate the rheology of the Northern Tibetan Plateau lithosphere by determining viscosity values several years after earthquake occurrence by comparison between modeled displacements and actual displacements determined through InSAR.

The manuscript is clearly written and well organized. However, I would like some points to be clarified:

1) More details about the InSAR processing should be provided:

1.1 - Which Sentinel-1 image acquisition mode was considered?

1.2 - Does the used InSAR processing method consider distributed or persistent scatterers?

1.3 - Which interferograms were built? How many?

2) In Sentinel-1 images, pixel spacing is larger along azimuth than along range. Why did the authors use a larger number of looks in azimuth than in range? 

3) The authors show displacements are influenced by atmospheric artifacts correlated to topography. Did they apply any correction to the displacements in order to mitigate this effect? If they did not, which is the impact of this effect on the determination of viscosities?

4) In Figure 2, it should be "residual" instead of "redisual".

Author Response

Dear reviewer,

Thank you for your valuable comments. We response your comments in the word file. Please see the attachment.

Best regards,

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors' response to my first comment shows their lack of basic understanding of the interseismic deformation field.  Interseismic deformation field is sourced by all the active faults in the region, not just the section of the fault ruptured during an earthquake as the authors assumed. For the postseismic study of the Kokoxili earthquake as an example, the east and west extensions of the Kunlun fault also contributed to the interseismic deformation, and the accumulated far field interseismic deformation (for a 2D model) is therefore (the interseismic slip rate across the fault) * (duration of the InSAR observation period), which is 10 mm/yr * 6 yr ~= 60 mm.  This is much greater than the 6 mm calculated by the authors.  The authors should check the GPS determined interseismic velocity field (e.g. Wang and Shen, JGR, 2020) to see the reality.

The authors' response to my second comment also shows their lack of some understanding about the right approach for modeling postseismic deformation. They cite some of the previous studies as a support for their using the accumulated displacements as data input for the modeling work. What they perhaps do not understand is that these previous studies were dealing with early postseismic deformation signals, which were nonlinear in time and could not be approximated by a linear velocity field. Instead, a viable approach would be to take interferograms of selected SAR image pairs and model these interferograms accordingly. The situation for this study is different. The SAR data that the authors acquired were observed more than a decade after the two earthquakes, and the deformation field was quasi-linear and could be well represented by a linear velocity field. According to the authors, hundreds, or even more than a thousand of interferograms were obtained for each track of the ascending and descending data, and the most precise and efficient way to quantify the deformation field is to solve for a linear velocity field, inverted by fitting all the interferogram time series of each track of the data. According to the authors, this seems what they did, as described in line 174-175 that: “the average line of sight (LOS) velocity is estimated on a pixel-by-pixel basis …” If so, why not use the velocity data as the data for model constraints? In the response to my previous comments the authors explained that one of the two reasons is that “the information contained in the velocity field is equal to the information contained in the cumulative deformation field.” This is probably true, but they ignored another consequence, that by multiplying an arbitrary time constant to the velocity field the data error becomes vague, which is not as simple as scaling the velocity error by the time constant.

Overall, although a lot of revisions have been done, the problem with data error associated with the interseismic deformation is still unsolved. As I stated in my last round of review comments, the interseismic deformation correlates with the postseismic deformation in spatial pattern and is comparable in amplitude, ignoring its effect would result in serious error in modeling the postseismic deformation process. Again I cannot recommend publication of this paper.

Reviewer 2 Report

Dear Authors,

Thank you for your in-depth answers to my inquires. In my opinion, the manuscript has been substantially improved, therefore, its current version is suitable for publication in Remote Sensing.

Congratulations!

Best regards,

Reviewer

Reviewer 3 Report

Dear authors,

thank you for your detailed answers. I think the paper is ready to be published.

Best regards