SiSeRHMap v1.0: a simulator for mapped seismic response using a hybrid model

Grelle, Gerardo; Bonito, Laura; Lampasi, Alessandro; Revellino, Paola; Guerriero, Luigi; Sappa, Giuseppe; Guadagno, Francesco Maria

doi:https://doi.org/10.5194/gmd-9-1567-2016

Articles | Volume 9, issue 4

https://doi.org/10.5194/gmd-9-1567-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/gmd-9-1567-2016

© Author(s) 2016. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 9, issue 4

Model description paper

|

26 Apr 2016

Model description paper |

| 26 Apr 2016

SiSeRHMap v1.0: a simulator for mapped seismic response using a hybrid model

Gerardo Grelle, Laura Bonito, Alessandro Lampasi, Paola Revellino, Luigi Guerriero, Giuseppe Sappa, and Francesco Maria Guadagno

Download

Final revised paper (published on 26 Apr 2016)
Supplement to the final revised paper
Preprint (discussion started on 17 Jun 2015)
Supplement to the preprint

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC C1732: 'Review of the article: SiSeRHMap v1.0: Simulator of mapped Seismic Response by Hybrid Model by G. Grelle, L. Bonito, A. Lampasi, P. Revellino, L. Guerriero, G. Sappa, and F.M. Guadagno', Anonymous Referee #1, 17 Aug 2015
- AC C2000: 'Reply to the Referee #1', Gerardo Grelle, 15 Sep 2015
RC C1735: 'Referee's Comments', Anonymous Referee #2, 18 Aug 2015
- AC C2014: 'Reply to the Referee #2', Gerardo Grelle, 15 Sep 2015

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Gerardo Grelle on behalf of the Authors (16 Sep 2015)

ED: Reconsider after major revisions (23 Sep 2015) by Lutz Gross

AR by Gerardo Grelle on behalf of the Authors (28 Oct 2015) Author's response

ED: Referee Nomination & Report Request started (04 Nov 2015) by Lutz Gross

RR by Anonymous Referee #3 (06 Mar 2016)

RR by Anonymous Referee #2 (09 Mar 2016)

Suggestions for revision or reasons for rejection

Review of Resubmission (version 4)

The authors have addressed most of the comments of this reviewer. My recommendation is to accept for publication, but I also include a list of suggested modifications to increase the readability and technical content of the paper. Overall, I think this is an interesting model and I hope that the authors continue its development. To me, the biggest limitation of this model is that it is a model of models, and its validation relies uniquely on matching other model predictions. For this reason, the model has to be used with caution.
One overarching comment is that the model aims to train site response in terms of the spectral shape of the response at the surface (or at depth zo). I am not sure this is the best option, because the output spectral shape will change significantly with each input motion. An alternative would be to train the model to capture the amplification function (e.g., spectral acceleration at the surface divided by the spectral acceleration of bedrock) for each spectral period. This is more stable across different input motions than the response spectra at the surface.

1. Throughout the paper the authors use “seismic response” to denote the computed spectral acceleration at the surface. I am not sure this is common use. The seismic response can relate to any response to seismic input (e.g., the stress-strain curve of a soil element can be “seismic response”). I recommend that the authors use more direct terms. (e.g., “computed spectral acceleration”)
2. The reply of the authors to comment 6b is factual (e.g., that is what their model implements), but does not address the larger issue of whether geometric parameterization at a constant scale can capture topographic effects across a wide range of frequencies. Analytical work has shown that longer period waves are affected differently by topographic features than shorter period waves.
3. In the abstract, line 19, the authors state: “… the one-dimensional linear equivalent analysis produces acceleration response spectra of shear wave velocity-thickness profiles…”. This should be modified because equivalent linear analyses need more input than simply profiles of shear wave velocity and thickness. The resulting response spectra is also a function of: a) input motion, b) modulus reduction and damping versus strain curves, c) soil density. I would recommend modifying the sentence to simply state that ‘the one-dimensional linear equivalent analyses produces acceleration response spectra for site profiles for a given input motion.’
4. Page 1, Line 37. What is meant by “local grassroots hazard”? the term grassroots, at least in the United States, tends to be used to define political movements that arise from the people (rather than led by party leaders). It is not clear how it applies to hazard.
5. Page 6, line 23. 1D site response assume vertical propagation of a plane wave, not a line wave.
6. Page 1, Line 58, please clarify the sentence “contextualized to the applied seismic response”. Do you mean to say “Contextualized to practical application in seismic site response studies …”?
7. The statement in page 2, lines 4-6: “Therefore, the map-sets of 4 seismic response provided by SiSeRHMap are the result of an advantageous compromise between intrinsic and 5 epistemic uncertainties and the accuracy and robustness required.” Are valid insofar as they refer to spatially distributed analysis, which is obviously the focus of this study. For site-specific studies, such as those for nuclear power plants or other critical facilities, the interplay of intrinsic (aleatoric) and epistemic uncertainty is much more complex. The authors may want to highlight that the intended application of their code is not site-specific studies but geographically distributed studies.
8. Some of the terminology used is not familiar to me (I am an expert in earthquake engineering, not in GIS). For example, page 3, line 14 uses “rigid reliefs”. I am not sure if the term rigid means that the reliefs are on exposed rock or if “rigid relief” refers to some type of relief. Given that the readership of the paper is likely going to reflect my own professional profile, I advise that the language be made clearer.
9. Page 3, line 25. The use of the word “bedrock” usually implies the rock layer that is underlying soil layers and is the base of the profile in a site response analyses. Since the authors are introducing two bedrock layers, I recommend to use the term “rock layer”, and reserve bedrock to the layer termed “rigid bedrock”
10. Sentence in page 3, lines 26 to 31 is very convoluted and difficult to understand.
11. Page 3, lines 53-54. The use of “formal physic dynamic behavior” is wrong. All behavior is physical, so the qualifier is not needed. I would suggest replacing the whole sentence by simply saying “the term qualifier “rigid” for the bedrock implies only relative rigidity and does not imply a rigid (e.g., infinite stiffness) layer”
12. Equation 2 should be properly justified. The text states (line 21). “The linear law used for bedrock (Eq. 2) meets the linear nature trend of the stiff soil in depth.” This statement is confusing. If it is used for bedrock, why is it relevant that it would meet trends in stiff soil? Moreover, I am not aware of any publication that justifies a linear increase with depth for shear wave velocity in stiff soil. The authors should give a reference or show data to prove their claim.
13. Is Equation 3 correct? How do we infer a constant velocity for bedrock from this equation?
14. A standard deviation is mentioned in page 4, line 23. This is the first reference to standard deviation and it is not clear what it refers to.
15. Page 4, line 33. Why is the output requested at a depth z(out) and not at the surface? This is not consistent with traditional output of microzonation or with building code recommendations (they all provide surface motions). If the output is at depth zo, are the authors providing “outcrop” or “within” output motions?
16. Page 4, lines 38-44 are not clear to me (again, I do not have background in GIS). For example, the Layer_n.txt files refer to the extension of the covered layers, which to me would mean a length unit, however the authors say that the input is in terms of zeros and ones. Also, the extension of Zones.txt is given as an integer. Why? Why not a real number? My lack of understanding is likely because of my limitations in GIS, but I would recommend that the authors try to make the paper accessible to earthquake engineers.
17. Page 5, line 55. Clarified what the term “dispersed” is used for. Page 6, line 1, clarify what “circumstantiated” is used for.
18. Page 8, line 38. Is the choice to envelope the computed response spectra (the authors use seismic response spectra, which I find confusing, since the input motion is also a seismic response spectra) dictated by code decisions? Or is it arbitrary? The authors should document this. Note that modern codes generally have probabilistic targets (e.g., 2% probability of exceedance in 50 years), so an envelope is not generally the most adequate solution. I am not recommending that the authors perform a probabilistic analysis, but they should document the reasons for their choice of enveloping the computed response spectra.
19. Page 8, line 49. What does it mean “lithodynamic is not present in the layer?” please clarify. In my understanding a layer is lithodynamic if it is characterized by Vs depth curve (page 2, line 32).
20. Page 10, line 47. Maufroy is misspelled. (this occurs elsewhere in the text as well)
21. Figure 13 is visually fascinating, but understanding it is akin to building a puzzle; a wonderful and fun enterprise when the pieces fit nicely, and a horribly frustrating one when they don’t. In either case, it baffles this reviewer why the authors would want to put this burden on the reader. For example, it is not clear at all in Page 11, line 5 what the labels (seven) and (three) imply. Reference in the caption to “top right of the panel” makes it confusing because usually the panel refers to the whole figure, not just part a or b. In part a, insert captions directly into the figure (top right of the panel) so that it is easier to read. What are the units on the figure in part b, at the middle of the three figures? What is “empty cycles” (in caption)? Does “senddle” mean “saddle”?
22. Page 11, line 33. What does “over imposing” mean?
23. Page 13, line 8. There is an orphan “fig.”
24. Explain what Figure 16b is. The contours plot shown are, I assume, computed from the proposed model. Where is the comparison with Quake-W?
25. The computation of the response spectra is generally computed using an algorithm by Nigam and Jennings (1969, in BSSA 59, 2, pp 902-922). This algorithm is exact for piecewise linear input, which is what is generally assumed for digital acceleration time histories. The deviations in the computed response spectra from other algorithms are not significant, so this is not a big issue.

Hide

ED: Publish subject to minor revisions (Editor review) (15 Mar 2016) by Lutz Gross

AR by Gerardo Grelle on behalf of the Authors (25 Mar 2016) Author's response Manuscript

ED: Publish subject to technical corrections (08 Apr 2016) by Lutz Gross

AR by Gerardo Grelle on behalf of the Authors (10 Apr 2016) Manuscript

Short summary

SiSeRHMap is a computer methodology for the mapped simulation of site seismic response (SSR) by means of a hybrid model. It combines physically based methods and adaptive and predictive numerical models that are solved in a layering GIS-geometric model. SiSeRHMap develops sets of different multispectral maps of site seismic response and other relating them, taking into account topography effects. Easy to use, SiSeRHMap seems to assure accuracy and robustness in the prediction performances.