Assessment of liquefaction-induced hazards using Bayesian networks based on standard penetration test data

Tang, Xiao-Wei; Bai, Xu; Hu, Ji-Lei; Qiu, Jiang-Nan

doi:https://doi.org/10.5194/nhess-18-1451-2018

Articles | Volume 18, issue 5

https://doi.org/10.5194/nhess-18-1451-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/nhess-18-1451-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 18, issue 5

Research article

|

29 May 2018

Research article |

| 29 May 2018

Assessment of liquefaction-induced hazards using Bayesian networks based on standard penetration test data

Xiao-Wei Tang, Xu Bai, Ji-Lei Hu, and Jiang-Nan Qiu

Download

Final revised paper (published on 29 May 2018)
Preprint (discussion started on 17 Mar 2017)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC1: 'grammar and English usage', R. E. S. Moss, 10 Oct 2017
- AC1: 'Reply to Prof. Moss's comments', Jilei Hu, 11 Oct 2017
- AC2: 'A new version of the manuscript', Jilei Hu, 21 Oct 2017
RC2: 'review of NHESS-2017-80', Anonymous Referee #2, 20 Nov 2017
- AC3: 'reply to review #2', Jilei Hu, 24 Nov 2017

Peer-review completion

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

ED: Reconsider after major revisions (further review by editor and referees) (22 Dec 2017) by Andreas Günther

AR by Anna Wenzel on behalf of the Authors (02 Jan 2018) Author's response

ED: Referee Nomination & Report Request started (18 Jan 2018) by Andreas Günther

RR by Anonymous Referee #3 (31 Jan 2018)

Suggestions for revision or reasons for rejection

This paper proposed a BN model to predict liquefaction and also deduce the chain reaction process of liquefaction-induced hazards and perform backward reasoning. It integrates multiple factors related to seismic liquefaction, sand boils, ground cracks, settlement, and lateral spreading into a model based on standard penetration test data. This method outperforms ANN and simplified/empirical methods in terms of Accuracy, Brier score, Recall, Precision, and AUC of ROC. The manuscript is well organized, figures and tables are informative, language use is good. I am sure that the current version has been substantially improved based on comments from previous reviewers. Some minor revision is suggested before acceptance of this manuscript.

1. Page 2, line 12, Liquefaction empirical models, please try to include research by (Goh 1994; Zhang and Goh 2013, Pal 2006; 13 Toprak et al. 1999, Zhang et al. 2015, Zhang and Goh 2016) based on historical data:
Zhang WG, Goh ATC. 2013. Multivariate adaptive regression splines for analysis of geotechnical engineering systems. Computers and Geotechnics 48: 82-95.
Zhang WG, Goh ATC, Zhang YM, Chen YM, Xiao Y. 2015. Assessment of soil liquefaction based on capacity energy concept and multivariate adaptive regression splines. Engineering Geology 188: 29-37.
Zhang WG, Goh ATC. 2016. Multivariate adaptive regression splines and neural network models for prediction of pile drivability. Geoscience Frontiers. 7: 45-52.
2. Page 11, Furthermore, there are no effective simplified methods for estimating ground cracks and sand boils, and simplified methods for calculating lateral spreading (Bartlett and Youd 1995, Wang and Rahman 1999, Goh et al. 2014) require the free face ratio or ground slope, which were not included in the data collected for this study. Please check Goh et al. 2014, the reviewer think it is Goh and Zhang 2014.
3. About the use of liquefaction potential index LPI, other researchers have used the similar term-seismic liquefaction potential, the authors are encouraged to include MARS_LR approach:
Zhang WG, Goh ATC*. 2016. Evaluating seismic liquefaction potential using multivariate adaptive regression splines and logistic regression. Geomechanics and Engineering, 10(3): 269-284.
4. Line 21, Page 15, discussion part, it should be noted that the Goh and Zhang (2014) method for estimating lateral spreading is the multivariate adaptive regression splines (MARS) method instead of the multiple linear regression (MLR).
5. As is well known, the ANN is a black-box method, that is, quite difficult to interpret the built model (maybe mainly through the weights, bias values, as well as the transfer functions). The authors should also explain the interpretability of the developed BN model and compare it with the MARS method and the empirical equations.
6. As a data-driven method, what is the main limitation of the proposed BN model?

Hide

RR by Anonymous Referee #2 (04 Apr 2018)

ED: Publish subject to minor revisions (review by editor) (19 Apr 2018) by Andreas Günther

AR by Jilei Hu on behalf of the Authors (24 Apr 2018) Author's response Manuscript

ED: Publish subject to technical corrections (09 May 2018) by Andreas Günther

AR by Jilei Hu on behalf of the Authors (11 May 2018) Author's response Manuscript