Comparison of surface-wave techniques to estimate S-  and P-wave velocity models from active seismic data

Khosro Anjom, Farbod; Adler, Frank; Socco, Laura Valentina

doi:https://doi.org/10.5194/se-15-367-2024

Articles | Volume 15, issue 3

https://doi.org/10.5194/se-15-367-2024

© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/se-15-367-2024

© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 3

Research article

|

14 Mar 2024

Research article |

| 14 Mar 2024

Comparison of surface-wave techniques to estimate S- and P-wave velocity models from active seismic data

Farbod Khosro Anjom, Frank Adler, and Laura Valentina Socco

Download

Final revised paper (published on 14 Mar 2024)
Preprint (discussion started on 06 Feb 2023)

Interactive discussion

Status: closed

RC1:
'Comment on egusphere-2023-78', Anonymous Referee #1, 23 May 2023

I am glad to review this manuscript "S- and P-wave velocity model estimation from seismic surface-waves" by Khosro Anjom et al. This work presents the application of three different surface wave methods to active-source surface wave data collected in a mining site, and estimates Vs and Vp, respectively. It's an overall good and informative paper. I have only a few comments on the details.
1. About the velocity perturbation about 7%

To be frank, i feel like the model perturbation is a bit higher. For example, let's say fig14a, the velocity perturbation of the target structure might be smaller than 15% from my guess. Authors may calculate it and prove me wrong. If it's this case, then the uncertainty from different methods is half of local anomaly of the structure which is unacceptable. I suggest authors provide the average velocity variation at different depths and different methods, and add these information into fig 18 and table 2. It will help the reader to understand the relative scales between the uncertainty associated with methods and the real variation associated with targets.
2. About the data error and model error

I know it's challenging to collect high quality seismic data in area with stiff surface. figure 3 shows that the quality of the collected surface wave data is poor. Is it necessary to use some signal-enhance technical to denoise the data? I am afraid the data quality/error might be introduced into the final model error/uncertainty between different methods. because the SWT method uses only two-station pair which will definitely provide lower quality inputs than other two methods who employ the multi-channel inputs. I would expect some necessary discussions about this part.
3. About the imaging point

I have no idea how the authors define the imaging points for the different methods. i guess the middle point will be taken as the imaging point for SWT, then what's the imaging point of the other two? please clarify this point.
4. why SWT can't provide VS model at the southern zone?
5. page 8, line 170, "Fig. 6b and c, as well as Fig. 6b and c, we show the", a typo!

Citation: https://doi.org/10.5194/egusphere-2023-78-RC1
- AC1: 'Reply on RC1', Farbod Khosro Anjom, 25 Sep 2023
  
  Dear anonymous reviewer,
  Thank you very much for your time in reviewing our manuscript and giving useful remarks. We provide a point-to-point response to your questions and remarks in the PDF file.
  Authors,
  
  Citation: https://doi.org/10.5194/egusphere-2023-78-AC1
EC1:
'Comment on egusphere-2023-78', Caroline Beghein, 25 Jul 2023

The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2023/egusphere-2023-78/egusphere-2023-78-EC1-supplement.pdf

Citation: https://doi.org/10.5194/egusphere-2023-78-EC1
- EC2: 'Reply on EC1', Caroline Beghein, 25 Jul 2023
  
  I, the topical editor, decided to review the paper because it has been very difficult to find a second reviewer. Please, make the necessary changes to the manuscript. Thank you.
  
  Citation: https://doi.org/10.5194/egusphere-2023-78-EC2
- AC2: 'Reply on EC1', Farbod Khosro Anjom, 25 Sep 2023
  
  Dear topical Editor,
  
  Thank you for the time you spent reviewing our manuscript and for the useful remarks. In the following, we provide a point-to-point response to all your remarks and comments.
  
  Best Regards,
  Authors
  
  Citation: https://doi.org/10.5194/egusphere-2023-78-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Farbod Khosro Anjom on behalf of the Authors (23 Oct 2023) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (31 Oct 2023) by Caroline Beghein

RR by Anonymous Referee #2 (15 Dec 2023)

Suggestions for revision or reasons for rejection

This paper applies three different methods to real seismic data to estimate both Vs and Vp near-surface structures. The wavelength-depth method gives the Vp structure, while all three methods give Vs structures. As a reviewer joined from the revised version, I feel that the qualities of the sentences and figures are sufficient for the publication. Below are some comments that should be addressed before the publication.

1. The title is too broad or general. This title can be used for a variety of studies. There should be more specific words to identify this work.
2. The method for estimating Vp should be described more. I guess everyone needs to see Socco and Comina (2017). It is helpful if the data, model parameters (number of layers; yes, I realized that there is no “layers” for the W/D method later, but that was really confusing before realizing it), and outputs are clearly stated in each method.
3. The disadvantages of W/D method should be described in the introduction section. In my understanding, up to ~10% uncertainty exists for both Vs and Vp estimations. This uncertainty is significant compared to conventional inversion methods. The W/D method seems like an approximation (transformation) method.
4. The main purpose of this study was not clear and should be clarified in both the abstract and introduction. Why did you compare the three methods? Is it to prove that the W/D method gives Vs structure comparable to conventional methods despite the theoretical uncertainties (and also gives Vp structure in less computational cost)?
5. The spatial distribution of Poisson’s ratio should be shown in addition to the Vs and Vp structures. It is not clear whether the estimated ratio gives meaningful a priori information to the conventional surface-wave analyses.
6. Why are the differences in three Vp models larger than that of Vs even though a priori information is common?
7. Some words were not familiar to me: carpet, “interval” models, and “time-average” models. Maybe because I’m not a native speaker and/or the research field is slightly different.

Hide

RR by Anonymous Referee #3 (28 Dec 2023)

Suggestions for revision or reasons for rejection

This paper compares and contrasts three different surface-wave methodologies applied to the same study area and dataset. The description of each method is extensive and most of the required detail is included in the manuscript, however, the link to the geological implications for the study area itself appears to be lacking. What has each result revealed about the study area? Based on the analysis is there a preferred technique for tackling active-source surface-wave imaging in studies such as these? Overall, the manuscript is well-written and informative. Below I detail some mostly minor comments.

Comment R3-1:
In the introduction it is not clear if group or phase velocities are used in the surface-wave tomography, though this is made clear further down in the method sections. I would suggest clarifying this in the introduction too, perhaps on Page 2, Line 30.

Comment R3-2:
Is it possible to plot sensitivity kernels for surface-waves as a function of depth? E.g., phase velocity sensitivity as a function of period/frequency and depth? This would help indicate what depths the dispersion curves are sensitive to. Have a look at Figure 4c in Darbyshire et al., (2013; EPSL).

Comment R3-3:
My background is in passive seismology so I’m not quite clear on what a carpet recording method. A sentence or two to explain what this is in more depth would be useful. Perhaps on Page 2, Line 44.

Comment R3-4:
I was going to suggest adding latitude/longitude values to Figure 1 but noticed the authors response to reviewer 2’s similar comment. Nevertheless, is there a way to better link Figures b and d? Do they correspond to exactly the same area? Perhaps add a box outline in figure 1b to show what region figure 1d corresponds to.

Comment R3-5:
Page 4, Line 104: “918 receivers”. Might be useful to include some more instrumentation detail here. What kind of receivers?

Comment R3-6:
Page 5, Line 126: “The data exhibits a low signal-to-noise ratio as expected for hard rock sites.” I would have thought soft rock sits have a low signal-to-noise ratio, whereas hard rocks allow for high signal-to-noise-ratio because of less attenuation? Perhaps add a citation for this statement since you mention this is ‘expected’.

Comment R3-7:
Page 8, Lines 179-180 and Figure 6b:
It still looks like there is almost double the paths coming from 0-40 degrees and 140-180 degrees than all the other azimuths. Perhaps run a surface-wave tomographic inversion excluding the excess paths in those directions and see if the result is similar. I.e., remove a subset of paths from the most sampled directions and see how that changes the final surface-wave tomography models.

Comment R3-8:
Page 11, Lines 249-251: Is there a figure for this? Like a trade-off curve showing data misfit versus level of constraint, for example, and the ideal model chosen along the curve?

Comment R3-9:
Was regularisation used such as damping and smoothing in the inversions? It isn’t immediately clear in the method section if this was done.

Comment R3-10:
Page 17, Line 348: “dense model grid”. More description of the model grid is needed. E.g., what is the spacing of the nodes etc. Perhaps including Figure R2 from the response to reviewers in the main text would be beneficial.

Hide

ED: Publish subject to minor revisions (review by editor) (22 Jan 2024) by Caroline Beghein

AR by Farbod Khosro Anjom on behalf of the Authors (29 Jan 2024) Author's response Author's tracked changes Manuscript

ED: Publish as is (07 Feb 2024) by Caroline Beghein

ED: Publish as is (07 Feb 2024) by Susanne Buiter (Executive editor)

AR by Farbod Khosro Anjom on behalf of the Authors (12 Feb 2024)

Short summary

Most surface-wave techniques focus on estimating the S-wave velocity (VS) model and consider the P-wave velocity (VP) model as prior information in the inversion step. Here, we show the application of three surface-wave methods to estimate both VS and VP models. We apply the methods to the data from a hard-rock site that were acquired through the irregular source–receiver recording technique. We compare the outcomes and performances of the methods in detail.