Probabilistic quantification of tsunami current hazard using statistical emulation

Devaraj Gopinathan; Mohammad Heidarzadeh; Serge Guillas

doi:10.1098/rspa.2021.0180

Probabilistic quantification of tsunami current hazard using statistical emulation

Devaraj Gopinathan, Mohammad Heidarzadeh, Serge Guillas

Research output: Contribution to journal › Article › peer-review

13 Citations (SciVal)

Abstract

In this paper, statistical emulation is shown to be an essential tool for the end-to-end physical and numerical modelling of local tsunami impact, i.e. from the earthquake source to tsunami velocities and heights. In order to surmount the prohibitive computational cost of running a large number of simulations, the emulator, constructed using 300 training simulations from a validated tsunami code, yields 1 million predictions. This constitutes a record for any realistic tsunami code to date, and is a leap in tsunami science since high risk but low probability hazard thresholds can be quantified. For illustrating the efficacy of emulation, we map probabilistic representations of maximum tsunami velocities and heights at around 200 locations about Karachi port. The 1 million predictions comprehensively sweep through a range of possible future tsunamis originating from the Makran Subduction Zone (MSZ). We rigorously model each step in the tsunami life cycle: first use of the three-dimensional subduction geometry Slab2 in MSZ, most refined fault segmentation in MSZ, first sediment enhancements of seabed deformation (up to 60% locally) and bespoke unstructured meshing algorithm. Owing to the synthesis of emulation and meticulous numerical modelling, we also discover substantial local variations of currents and heights.

Original language	English
Article number	20210180
Number of pages	28
Journal	Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences
Volume	477
Issue number	2250
Early online date	9 Jun 2021
DOIs	https://doi.org/10.1098/rspa.2021.0180
Publication status	Published - 30 Jun 2021

Bibliographical note

Funding: D.G. and S.G. were supported by the Alan Turing Institute under the EPSRC grant no. (EP/N510129/1). M.H. was supported by the Royal Society grant no. (CHL/R1/180173). D.G. was partially funded by the Royal Society-SERB Newton International Fellowship (NF151483). D.G., M.H. and S.G. acknowledge support from the NERC grant no. (NE/P016367/1). Acknowledgements. This work has been performed using resources provided by the Cambridge Tier-2 system (CSD3 Wilkes2) operated by the University of Cambridge Research Computing Service funded by EPSRC Tier-2 capital grant no. EP/P020259/1. The authors would like to acknowledge the use of the University of Oxford Advanced Research Computing (ARC) facility (JADE) in carrying out this work (doi:10.5281/zenodo.22558). Preparative simulations were performed on the EMERALD High Performance Computing facility provided via the EPSRC funded Centre for Innovation (EP/K000144/1 and EP/K000136/1), owned and operated by the e-Infrastructure South Consortium formed by the universities of Bristol, Oxford, Southampton and UCL in partnership with STFC Rutherford Appleton Laboratory. We thank Eric Daub and Oliver Strickson for active development of MOGP UQ suite, Daniel Giles for improvements to second-order FV scheme and boundary conditions in VOLNA-OP2, István Reguly for its installation and running on CSD3, Deyu Ming and Mariya Mamajiwala on truncated G–R distribution and prediction intervals for L-O-O diagnostics, Frédéric Dias on meshing strategies and sediment amplification curve, Theodoros Mathikolonis on emulation, and Simon Day, Kusala Rajendran and C.P. Rajendran on seismicity of M.S.Z. We thank the referees whose detailed comments were instrumental in enhancing the content and presentation of the article.

Data accessibility: The data and codes used have been cited and/or linked in footnotes at first mention. Authors’ contributions. M.H. and S.G. conceptualized the problem. S.G. and D.G. conceptualized the employment of large-scale statistical emulation and the inclusion of the effect of sediments. M.H. digitized the bathymetry for Karachi port. D.G. designed the problem with inputs and supervision from S.G. and M.H., developed codes, curated data, carried out the simulations with associated validation, analysis and data processing, and created visualizations for the main article and electronic supplementary material. All authors drafted and critically reviewed the manuscript. All the authors give final approval for publication and agree to be held accountable for the work performed herein. Competing interests. We declare we have no competing interests.

Publisher Copyright:
© 2021 The Author(s).

Keywords

coastal engineering
hazard assessment
Karachi port
Makran subduction zone
sediment amplification
unstructured mesh

ASJC Scopus subject areas

General Mathematics
General Engineering
General Physics and Astronomy

Access to Document

10.1098/rspa.2021.0180Licence: CC BY

Cite this

@article{f80bac01385e4c1d9016a325680324cc,

title = "Probabilistic quantification of tsunami current hazard using statistical emulation",

abstract = "In this paper, statistical emulation is shown to be an essential tool for the end-to-end physical and numerical modelling of local tsunami impact, i.e. from the earthquake source to tsunami velocities and heights. In order to surmount the prohibitive computational cost of running a large number of simulations, the emulator, constructed using 300 training simulations from a validated tsunami code, yields 1 million predictions. This constitutes a record for any realistic tsunami code to date, and is a leap in tsunami science since high risk but low probability hazard thresholds can be quantified. For illustrating the efficacy of emulation, we map probabilistic representations of maximum tsunami velocities and heights at around 200 locations about Karachi port. The 1 million predictions comprehensively sweep through a range of possible future tsunamis originating from the Makran Subduction Zone (MSZ). We rigorously model each step in the tsunami life cycle: first use of the three-dimensional subduction geometry Slab2 in MSZ, most refined fault segmentation in MSZ, first sediment enhancements of seabed deformation (up to 60% locally) and bespoke unstructured meshing algorithm. Owing to the synthesis of emulation and meticulous numerical modelling, we also discover substantial local variations of currents and heights.",

keywords = "coastal engineering, hazard assessment, Karachi port, Makran subduction zone, sediment amplification, unstructured mesh",

author = "Devaraj Gopinathan and Mohammad Heidarzadeh and Serge Guillas",

note = "Funding: D.G. and S.G. were supported by the Alan Turing Institute under the EPSRC grant no. (EP/N510129/1). M.H. was supported by the Royal Society grant no. (CHL/R1/180173). D.G. was partially funded by the Royal Society-SERB Newton International Fellowship (NF151483). D.G., M.H. and S.G. acknowledge support from the NERC grant no. (NE/P016367/1). Acknowledgements. This work has been performed using resources provided by the Cambridge Tier-2 system (CSD3 Wilkes2) operated by the University of Cambridge Research Computing Service funded by EPSRC Tier-2 capital grant no. EP/P020259/1. The authors would like to acknowledge the use of the University of Oxford Advanced Research Computing (ARC) facility (JADE) in carrying out this work (doi:10.5281/zenodo.22558). Preparative simulations were performed on the EMERALD High Performance Computing facility provided via the EPSRC funded Centre for Innovation (EP/K000144/1 and EP/K000136/1), owned and operated by the e-Infrastructure South Consortium formed by the universities of Bristol, Oxford, Southampton and UCL in partnership with STFC Rutherford Appleton Laboratory. We thank Eric Daub and Oliver Strickson for active development of MOGP UQ suite, Daniel Giles for improvements to second-order FV scheme and boundary conditions in VOLNA-OP2, Istv{\'a}n Reguly for its installation and running on CSD3, Deyu Ming and Mariya Mamajiwala on truncated G–R distribution and prediction intervals for L-O-O diagnostics, Fr{\'e}d{\'e}ric Dias on meshing strategies and sediment amplification curve, Theodoros Mathikolonis on emulation, and Simon Day, Kusala Rajendran and C.P. Rajendran on seismicity of M.S.Z. We thank the referees whose detailed comments were instrumental in enhancing the content and presentation of the article. Data accessibility: The data and codes used have been cited and/or linked in footnotes at first mention. Authors{\textquoteright} contributions. M.H. and S.G. conceptualized the problem. S.G. and D.G. conceptualized the employment of large-scale statistical emulation and the inclusion of the effect of sediments. M.H. digitized the bathymetry for Karachi port. D.G. designed the problem with inputs and supervision from S.G. and M.H., developed codes, curated data, carried out the simulations with associated validation, analysis and data processing, and created visualizations for the main article and electronic supplementary material. All authors drafted and critically reviewed the manuscript. All the authors give final approval for publication and agree to be held accountable for the work performed herein. Competing interests. We declare we have no competing interests. Publisher Copyright: {\textcopyright} 2021 The Author(s).",

year = "2021",

month = jun,

day = "30",

doi = "10.1098/rspa.2021.0180",

language = "English",

volume = "477",

journal = "Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences",

issn = "1364-5021",

publisher = "Royal Society of London",

number = "2250",

}

TY - JOUR

T1 - Probabilistic quantification of tsunami current hazard using statistical emulation

AU - Gopinathan, Devaraj

AU - Heidarzadeh, Mohammad

AU - Guillas, Serge

N1 - Funding: D.G. and S.G. were supported by the Alan Turing Institute under the EPSRC grant no. (EP/N510129/1). M.H. was supported by the Royal Society grant no. (CHL/R1/180173). D.G. was partially funded by the Royal Society-SERB Newton International Fellowship (NF151483). D.G., M.H. and S.G. acknowledge support from the NERC grant no. (NE/P016367/1). Acknowledgements. This work has been performed using resources provided by the Cambridge Tier-2 system (CSD3 Wilkes2) operated by the University of Cambridge Research Computing Service funded by EPSRC Tier-2 capital grant no. EP/P020259/1. The authors would like to acknowledge the use of the University of Oxford Advanced Research Computing (ARC) facility (JADE) in carrying out this work (doi:10.5281/zenodo.22558). Preparative simulations were performed on the EMERALD High Performance Computing facility provided via the EPSRC funded Centre for Innovation (EP/K000144/1 and EP/K000136/1), owned and operated by the e-Infrastructure South Consortium formed by the universities of Bristol, Oxford, Southampton and UCL in partnership with STFC Rutherford Appleton Laboratory. We thank Eric Daub and Oliver Strickson for active development of MOGP UQ suite, Daniel Giles for improvements to second-order FV scheme and boundary conditions in VOLNA-OP2, István Reguly for its installation and running on CSD3, Deyu Ming and Mariya Mamajiwala on truncated G–R distribution and prediction intervals for L-O-O diagnostics, Frédéric Dias on meshing strategies and sediment amplification curve, Theodoros Mathikolonis on emulation, and Simon Day, Kusala Rajendran and C.P. Rajendran on seismicity of M.S.Z. We thank the referees whose detailed comments were instrumental in enhancing the content and presentation of the article. Data accessibility: The data and codes used have been cited and/or linked in footnotes at first mention. Authors’ contributions. M.H. and S.G. conceptualized the problem. S.G. and D.G. conceptualized the employment of large-scale statistical emulation and the inclusion of the effect of sediments. M.H. digitized the bathymetry for Karachi port. D.G. designed the problem with inputs and supervision from S.G. and M.H., developed codes, curated data, carried out the simulations with associated validation, analysis and data processing, and created visualizations for the main article and electronic supplementary material. All authors drafted and critically reviewed the manuscript. All the authors give final approval for publication and agree to be held accountable for the work performed herein. Competing interests. We declare we have no competing interests. Publisher Copyright: © 2021 The Author(s).

PY - 2021/6/30

Y1 - 2021/6/30

N2 - In this paper, statistical emulation is shown to be an essential tool for the end-to-end physical and numerical modelling of local tsunami impact, i.e. from the earthquake source to tsunami velocities and heights. In order to surmount the prohibitive computational cost of running a large number of simulations, the emulator, constructed using 300 training simulations from a validated tsunami code, yields 1 million predictions. This constitutes a record for any realistic tsunami code to date, and is a leap in tsunami science since high risk but low probability hazard thresholds can be quantified. For illustrating the efficacy of emulation, we map probabilistic representations of maximum tsunami velocities and heights at around 200 locations about Karachi port. The 1 million predictions comprehensively sweep through a range of possible future tsunamis originating from the Makran Subduction Zone (MSZ). We rigorously model each step in the tsunami life cycle: first use of the three-dimensional subduction geometry Slab2 in MSZ, most refined fault segmentation in MSZ, first sediment enhancements of seabed deformation (up to 60% locally) and bespoke unstructured meshing algorithm. Owing to the synthesis of emulation and meticulous numerical modelling, we also discover substantial local variations of currents and heights.

AB - In this paper, statistical emulation is shown to be an essential tool for the end-to-end physical and numerical modelling of local tsunami impact, i.e. from the earthquake source to tsunami velocities and heights. In order to surmount the prohibitive computational cost of running a large number of simulations, the emulator, constructed using 300 training simulations from a validated tsunami code, yields 1 million predictions. This constitutes a record for any realistic tsunami code to date, and is a leap in tsunami science since high risk but low probability hazard thresholds can be quantified. For illustrating the efficacy of emulation, we map probabilistic representations of maximum tsunami velocities and heights at around 200 locations about Karachi port. The 1 million predictions comprehensively sweep through a range of possible future tsunamis originating from the Makran Subduction Zone (MSZ). We rigorously model each step in the tsunami life cycle: first use of the three-dimensional subduction geometry Slab2 in MSZ, most refined fault segmentation in MSZ, first sediment enhancements of seabed deformation (up to 60% locally) and bespoke unstructured meshing algorithm. Owing to the synthesis of emulation and meticulous numerical modelling, we also discover substantial local variations of currents and heights.

KW - coastal engineering

KW - hazard assessment

KW - Karachi port

KW - Makran subduction zone

KW - sediment amplification

KW - unstructured mesh

UR - http://www.scopus.com/inward/record.url?scp=85111099822&partnerID=8YFLogxK

U2 - 10.1098/rspa.2021.0180

DO - 10.1098/rspa.2021.0180

M3 - Article

AN - SCOPUS:85111099822

SN - 1364-5021

VL - 477

JO - Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences

JF - Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences

IS - 2250

M1 - 20210180

ER -

Probabilistic quantification of tsunami current hazard using statistical emulation

Abstract

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this