Abstract
Different nonergodic Ground-Motion Models based on spatially varying coefficient models are compared for ground-motion data in Italy. The models are based different methodologies: Multi-source geographically weighted regression (Caramenti et al. 2022), and Bayesian hierarchical models estimated with the integrated nested Laplace approximation (Rue et al. 2009). The different models are compared in terms of their predictive performance, their spatial coefficients, and their predictions. Models that include spatial terms perform slightly better than a simple base model that includes only event and station terms, in terms of out-of sample error based on cross-validation. The Bayesian spatial models have slightly lower generalization error, which can be attributed to the fact that they can include random effects for events and stations. The different methodologies give rise to different dependencies of the spatially varying terms on event and station locations, leading to between-model uncertainty in their predictions, which should be accommodated in a nonergodic seismic hazard assessment.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Data availability
Availability of data is declared under Data and Resources section.
Code availability
Availability of code is declared under Data and Resources section.
Data and resources
The data used in this study is described in Lanzano et al. (2019) and Caramenti et al. (2020). Code and data can be found at https://github.com/nikuehn/NonErgGMM_Italy.
Notes
See https://haakonbakkagit.github.io/btopic114.html for some informal advice on mesh generation.
Also called Watanabe-Akaike information criterion.
Using control.inla = list(int.strategy = "eb", strategy = "gaussian") in the INLA call.
References
Abrahamson N, Kuehn N, Walling M, Landwehr N (2019) probabilistic seismic hazard analysis in california using nonergodic ground motion models. Bull Seismol Soci Am 109(4):1235–1249. https://doi.org/10.1785/0120190030
Akaike H (1973) Information theory and an extension of the maximum likelihood principle. In: Second international symposium on information theory, pp 267–281
Al-Atik L, Abrahamson N, Bommer JJ, Scherbaum F, Cotton F, Kuehn N (2010) The variability of ground-motion prediction models and its components. Seismol Res Lett 81(5):794–801. https://doi.org/10.1785/gssrl.81.5.794
Ancheta TD, Darragh RB, Stewart JP, Seyhan E, Silva WJ, Chiou BSJ, Wooddell KE, Graves RW, Kottke AR, Boore DM, Kishida T, Donahue JL (2014) NGA-west2 database. Earthq Spectra 30(3):989–1005. https://doi.org/10.1193/070913EQS197M
Anderson JG, Brune JN (1999) Probabilistic seismic hazard analysis without the ergodic assumption. Seismol Res Lett 70(1):19–28. https://doi.org/10.1785/gssrl.70.1.19
Bachl FE, Lindgren F, Borchers DL, Illian JB (2019) Inlabru: an R package for bayesian spatial modelling from ecological survey data. Method Ecol Evol 10(6):760–766. https://doi.org/10.1111/2041-210X.13168
Bakka H, Rue H, Fuglstad GA, Riebler A, Bolin D, Illian J, Krainski E, Simpson D, Lindgren F (2018) Spatial modeling with R-INLA: a review. Wiley Interdiscip Rev Comput Stat 10(6):e1443. https://doi.org/10.1002/wics.1443
Bayliss K, Naylor M, Illian J, Main IG (2020) Data-driven optimization of seismicity models using diverse data dets generation evaluation and ranking using inlabru. J Geophys Res Solid Earth. https://doi.org/10.1029/2020JB020226
Bivand R, Keitt T, Rowlingson B (2022) rgdal: Bindings for the geospatial data abstraction library. https://CRAN.R-project.org/package=rgdal, r package version 1.5-30
Bivand RS, Pebesma E, Gomez-Rubio V (2013) Applied spatial data analysis with R, 2nd edn. Springer, NY
Bivand RS, Gómez-Rubio V, Rue H (2015) Spatial data analysis with R - INLA with some extensions. J Stat Softw 63(20):1–31. https://doi.org/10.18637/jss.v063.i20
Bommer JJ, Douglas J, Strasser FO (2003) Style-of-faulting in ground-motion prediction equations. Bull Earthq Eng 1(2):171–203. https://doi.org/10.1023/A:1026323123154
Boore DM (2010) Orientation-independent, nongeometric-mean measures of seismic intensity from two horizontal components of motion. Bull Seismol Soc Am 100(4):1830–1835. https://doi.org/10.1785/0120090400
Bozorgnia Y, Abrahamson NA, Atik LA, Ancheta TD, Atkinson GM, Baker JW, Baltay A, Boore DM, Campbell KW, Chiou BSJ, Darragh R, Day S, Donahue J, Graves RW, Gregor N, Hanks T, Idriss IM, Kamai R, Kishida T, Kottke A, Mahin SA, Rezaeian S, Rowshandel B, Seyhan E, Shahi S, Shantz T, Silva W, Spudich P, Stewart JP, Watson-Lamprey J, Wooddell K, Youngs R (2014) NGA-west2 research project. Earthq Spectra 30(3):973–987. https://doi.org/10.1193/072113EQS209M
Bussas M, Sawade C, Scheffer T, Landwehr N (2015) Varying-coefficient models with isotropic gaussian process priors. 0:1–17, arXiv: 1508.07192
Bussas M, Sawade C, Kühn N, Scheffer T, Landwehr N (2017) Varying-coefficient models for geospatial transfer learning. Mach Learn 106(9–10):1419–1440. https://doi.org/10.1007/s10994-017-5639-3
Caramenti L, Menafoglio A, Sgobba S, Lanzano G (2020) Multi-Source Geographically Weighted Regression for Regionalized Ground-Motion Models. Tech. rep., https://mox.polimi.it/publication-results/?id=917%5C &tipo=add_qmox
Caramenti L, Menafoglio A, Sgobba S, Lanzano G (2022) Multi-source geographically weighted regression for regionalized ground-motion models. Spat Stat 47:100610. https://doi.org/10.1016/j.spasta.2022.100610
Chao SH, Lin CM, Kuo CH, Huang JY, Wen KL, Chen YH (2020) Implementing horizontal-to-vertical Fourier spectral ratios and spatial correlation in a ground-motion prediction equation to predict site effects. Earthq Spectra. https://doi.org/10.1177/8755293020952449
Dambon JA, Sigrist F, Furrer R (2021) Maximum likelihood estimation of spatially varying coefficient models for large data with an application to real estate price prediction. Spat Stat 41:100470. https://doi.org/10.1016/j.spasta.2020
D’Angelo N, Abbruzzo A, Adelfio G (2020) Spatial bayesian hierarchical modelling with integrated nested laplace approximation. pp 1–22, arXiv: http://arxiv.org/abs/2009.03712
Datta A, Banerjee S, Finley AO, Gelfand AE (2016) On nearest-neighbor gaussian process models for massive spatial data. Wiley Interdiscip Rev Comput Stat 8(5):162–171. https://doi.org/10.1002/wics.1383
Dawood HM, Rodriguez-Marek A (2013) A method for including path effects in ground-motion prediction equations: an example using the Mw 9.0 tohoku earthquake aftershocks. Bull Seismol Soc Am 103(2B):1360–1372. https://doi.org/10.1785/0120120125
Esposito S, Iervolino I (2011) PGA and PGV spatial correlation models based on European multievent datasets. Bull Seismolo Soc Am 101(5):2532–2541. https://doi.org/10.1785/0120110117
Esposito S, Iervolino I (2012) Spatial correlation of spectral acceleration in european data. Bull Seismol Soc Am 102(6):2781–2788. https://doi.org/10.1785/0120120068
Finley AO (2011) Comparing spatially-varying coefficients models for analysis of ecological data with non-stationary and anisotropic residual dependence. Method Ecol Evol 2(2):143–154. https://doi.org/10.1111/j.2041-210X.2010.00060.x
Finley AO, Datta A, Cook BD, Morton DC, Andersen HE, Banerjee S (2018) Efficient algorithms for bayesian nearest neighbor gaussian processes. J Comput Gr Stat 111(514):1–14. https://doi.org/10.1080/10618600.2018.1537924
Fotheringham AS, Brunsdon C, Charlton M (2002) Geographically weighted regression: the analysis of spatially varying relationships. Wiley, Florida
Fotheringham AS, Yang W, Kang W (2017) Multiscale geographically weighted regression (MGWR). Ann Am Assoc Geogr 107(6):1247–1265. https://doi.org/10.1080/24694452.2017.1352480
Franco-Villoria M, Ventrucci M, Rue H (2019) A unified view on bayesian varying coefficient models. Electron J Stat 13(2):5334–5359. https://doi.org/10.1214/19-EJS1653
Fuglstad GA, Simpson D, Lindgren F, Rue H (2019) Constructing priors that penalize the complexity of gaussian random fields. J Am Stat Assoc 114(525):445–452. https://doi.org/10.1080/01621459.2017.1415907
Gelfand AE, Kim HJ, Sirmans CF, Banerjee S (2003) Spatial modeling with spatially varying coefficient processes. J Am Stat Assoc 98(462):387–396. https://doi.org/10.1198/016214503000170
Gelfand AE, Schmidt AM, Banerjee S, Sirmans CF (2004) Nonstationary multivariate process modeling through spatially varying coregionalization. Test 13(2):263–312. https://doi.org/10.1007/BF02595775
Gelman A, Hwang J, Vehtari A (2014) Understanding predictive information criteria for bayesian models. Stat Comput 24(6):997–1016. https://doi.org/10.1007/s11222-013-9416-2
Gollini I, Lu B, Charlton M, Brunsdon C, Harris P (2015) GWmodel : an R package for exploring spatial heterogeneity. J Stat Softw 63(17):1–50
Gómez-Rubio V (2020) Bayesian inference with INLA. Chapman and Hall/CRC, Boca Raton, FL
Huang C, Galasso C (2019) Ground-motion intensity measure correlations observed in Italian strong-motion records. Earthq Eng Struct Dyn 48(15):1634–1660. https://doi.org/10.1002/eqe.3216
Jayaram N, Baker JW (2009) Correlation model for spatially distributed ground-motion intensities. Earthq Eng Struct Dyn 38(15):1687–1708. https://doi.org/10.1002/eqe.922
Kowsari M, Sonnemann T, Halldorsson B, Hrafnkelsson B, Snaebjörnsson JP, Jonsson S (2020) Bayesian inference of empirical ground motion models to pseudo-spectral accelerations of south Iceland seismic zone earthquakes based on informative priors. Soil Dyn Earthq Eng 132:106075. https://doi.org/10.1016/j.soildyn.2020.106075
Krainski E, Gómez-Rubio V, Bakka H, Lenzi A, Castro-Camilo D, Simpson D, Lindgren FK, Rue H (2019) Advanced spatial modeling with stochastic partial differential equations using R and INLA. Chapman and Hall/CRC, Boca-Raton, FL
Kuehn N (2021a) A Primer for using INLA to Estimate Ground-Motion Models. Engrxiv pp 1–30. doi: https://doi.org/10.31224/osf.io/6ut3p
Kuehn N (2021b) Comparison of Bayesian Varying Coefficient Models for the Development of Nonergodic Ground-Motion Models. Engrxiv pp 1–25. doi: https://doi.org/10.31224/osf.io/tjxa3
Kuehn NM, Abrahamson NA (2020) Spatial correlations of ground motion for non-ergodic seismic hazard analysis. Earthq Eng Struct Dyn 49(1):4–23. https://doi.org/10.1002/eqe.3221
Kuehn NM, Abrahamson NA, Walling MA (2019) Incorporating nonergodic path effects into the NGA-west2 ground-motion prediction equations. Bull Seismol Soc Am 109(2):575–585. https://doi.org/10.1785/0120180260
Landwehr N, Kuehn NM, Scheffer T, Abrahamson N (2016) A nonergodic ground-motion model for california with spatially varying coefficients. Bull Seismol Soc Am 106(6):2574–2583. https://doi.org/10.1785/0120160118
Lanzano G, D’Amico M, Felicetta C, Puglia R, Luzi L, Pacor F, Bindi D (2016) Ground-motion prediction equations for region-specific probabilistic seismic-hazard analysis. Bull Seismol Soc Am 106(1):73–92. https://doi.org/10.1785/0120150096
Lanzano G, Luzi L, Pacor F, Felicetta C, Puglia R, Sgobba S, D’Amico M (2019) A revised ground-motion prediction model for shallow crustal earthquakes in italy. Bull Seismol Soc Am 109(2):525–540. https://doi.org/10.1785/0120180210
Lanzano G, Sgobba S, Caramenti L, Menafoglio A (2021) Ground-Motion Model for Crustal Events in Italy by Applying the Multisource Geographically Weighted Regression (MS-GWR) Method. Bull Seismol Soc Am. https://doi.org/10.1785/0120210044
Lavrentiadis G, Abrahamson NA, Kuehn NM (2021) A non-ergodic effective amplitude ground-motion model for california. Bull Earthq Eng. https://doi.org/10.1007/s10518-021-01206-w
Lezama-Ochoa N, Pennino MG, Hall MA, Lopez J, Murua H (2020) Using a bayesian modelling approach (INLA-SPDE) to predict the occurrence of the spinetail devil ray (Mobular mobular). Sci Rep 10(1):1–11. https://doi.org/10.1038/s41598-020-73879-3
Lin PS, Chiou B, Abrahamson N, Walling M, Lee CT, Cheng CT (2011) Repeatable source, site, and path effects on the standard deviation for empirical ground-motion prediction models. Bull Seismol Soc Am 101(5):2281–2295. https://doi.org/10.1785/0120090312
Lindgren F, Rue H (2015) Bayesian spatial modelling with R - INLA. J Stat Softw 63(19):1–25. https://doi.org/10.18637/jss.v063.i19
Lindgren F, Rue H, Lindström J (2011) An explicit link between gaussian fields and gaussian markov random fields: the stochastic partial differential equation approach. J Royal Stat Soc Ser B Stat Methodol 73(4):423–498. https://doi.org/10.1111/j.1467-9868.2011.00777.x
Liu C, Macedo J, Kuehn NM (2022) Spatial correlation of systematic effects of non-ergodic ground motion models in the ridgecrest area. Bull Earthq Eng. https://doi.org/10.1007/s10518-022-01403-1
Martino S, Riebler A (2020) Integrated nested laplace approximations (INLA). Wiley, New Jersey, pp 1–19
Meng X, Goulet C (2022) Lessons learned from applying varying coefficient model to controlled simulation datasets. Bull Earthq Eng. https://doi.org/10.1007/s10518-022-01403-1
Moraga P (2019) Geospatial health data: modeling and visualization with R-INLA and shiny. Florida, CRC
Murakami D, Lu B, Harris P, Brunsdon C, Charlton M, Nakaya T, Griffith DA (2019) The importance of scale in spatially varying coefficient modeling. Ann Am Assoc Geogr 109(1):50–70. https://doi.org/10.1080/24694452.2018.1462691
Paciorek CJ, Schervish MJ (2006) Spatial modelling using a new class of nonstationary covariance functions. Environmetrics 17(5):483–506. https://doi.org/10.1002/env.785
Pebesma E (2018) Simple features for r: standardized support for spatial vector data. R J 10(1):439–446. https://doi.org/10.32614/RJ-2018-009
R Core Team (2021) R: a language and environment for statistical computing. https://www.r-project.org/
Rasmussen CE, Williams CKI (2006) Gaussian processes for machine learning. MIT Press, Cambridge
Rennolet SB, Moschetti MP, Thompson EM, Yeck WL (2018) A flatfile of ground motion intensity measurements from induced earthquakes in oklahoma and kansas. Earthq Spectra 34(1):1–20. https://doi.org/10.1193/101916EQS175DP
Rue H, Martino S, Chopin N (2009) Approximate bayesian inference for latent gaussian models by using integrated nested laplace approximations. J Royal Stat Soc Ser B Stat Methodol 71(2):319–392. https://doi.org/10.1111/j.1467-9868.2008.00700.x
Rue H, Riebler A, Sørbye SH, Illian JB, Simpson DP, Lindgren FK (2017) Bayesian computing with INLA: a review. Ann Rev Stat Appl 4(1):395–421. https://doi.org/10.1146/annurev-statistics-060116-054045
Schrödle B, Held L (2011) A primer on disease mapping and ecological regression using INLA. Comput Stat 26(2):241–258. https://doi.org/10.1007/s00180-010-0208-2
Simpson D, Lindgren F, Rue H (2012) In order to make spatial statistics computationally feasible, we need to forget about the covariance function. Environmetrics 23(1):65–74. https://doi.org/10.1002/env.1137
Simpson D, Rue H, Riebler A, Martins TG, Sørbye SH (2017) Penalising model component complexity: a principled, practical approach to constructing priors. Stat Sci 32(1):1–28. https://doi.org/10.1214/16-STS576
Sung CH, Abrahamson NA, Kuehn NM, Traversa P, Zentner I (2021) A non-ergodic ground-motion model of fourier amplitude spectra for france. Bull Earthq Eng. https://doi.org/10.1007/s10518-022-01403-1
Trugman DT, Shearer PM (2018) Strong correlation between stress drop and peak ground acceleration for recent M 1–4 Earthquakes in the san francisco bay area. Bull Seismol Soc Am 108(2):929–945. https://doi.org/10.1785/0120170245
Vehtari A, Gelman A, Gabry J (2017) Practical bayesian model evaluation using leave-one-out cross-validation and WAIC. Stat Comput 27(5):1413–1432. https://doi.org/10.1007/s11222-016-9696-4
Vilela R, Burger C, Diederichs A, Bachl FE, Szostek L, Freund A, Braasch A, Bellebaum J, Beckers B, Piper W, Nehls G (2021) Use of an INLA latent gaussian modeling approach to assess bird population changes due to the development of offshore wind farms. Front Marin Sci 8(July):1–11. https://doi.org/10.3389/fmars.2021.701332
Villani M, Abrahamson NA (2015) Repeatable site and path effects on the ground-motion sigma based on empirical data from southern california and simulated waveforms from the cybershake platform. Bull Seismol Soc Am 105(5):2681–2695. https://doi.org/10.1785/0120140359
Walling M, Abrahamson NA (2012) Non-ergodic probabilistic seismic hazard analyses. In: 15th World conference on earthquake engineering (15WCEE)
Walling M, Kuehn NM, Abrahamson NA (2021) An induced seismicity non-ergodic ground motion prediction equation (GMPE) in the oklahoma region. Tech. Rep. NEHRP Grant G18AP00076, USGS
Watanabe S (2010) Asymptotic equivalence of bayes cross validation and widely applicable information criterion in singular learning theory. J Machi Learn Res 11:3571–3594 arXiv: 1004.2316
Watanabe S (2013) A widely applicable bayesian information criterion. J Mach Learn Res 14(1):867–897
Wheeler DC (2014) Geographically weighted regression. Springer, Berlin Heidelberg, Berlin, Heidelberg, pp 1435–1459
Wheeler DC, Calder CA (2007) An assessment of coefficient accuracy in linear regression models with spatially varying coefficients. J Geogr Syst 9(2):145–166. https://doi.org/10.1007/s10109-006-0040-y
Wickham H (2016) ggplot2: Elegant graphics for data analysis. Springer, New York
Wolf LJ, Oshan TM, Fotheringham AS (2018) Single and multiscale models of process spatial heterogeneity. Geogr Anal 50(3):223–246. https://doi.org/10.1111/gean.12147
Zhang R, Czado C, Sigloch K (2016) Bayesian spatial modelling for high dimensional seismic inverse problems. J Royal Stat Soc Ser C Appl Stat 65(2):187–213. https://doi.org/10.1111/rssc.12118
Acknowledgements
CV/MSLL and log-likelihood calculation follows ideas from the INLA Google group ( https://groups.google.com/g/r-inla-discussion-group/c/K49zsVxOYiw/m/GwEQGk1dAwAJ). Plots are made using the R package ggplot2 (Wickham 2016). The mesh plot was made via the package inlabru (Bachl et al. 2019). Other packages used are sp (Bivand et al. 2013), sf (Pebesma 2018), rgdal (Bivand et al. 2022). The MS-GWR code uses internally the package GWmodel (Gollini et al. 2015). Spatial plots (Figs. 5, 7, and 6) follow the style of Caramenti et al. (2020), from https://github.com/lucaramenti/ms-gwr/. I would like to thank Melanie Walling, Greg Lavrentiadis, Yousef Bozorgnia, Xiaofeng Ming, and Elnaz Seylabi, and the nonergodic working group at UC Berkeley/UCLA for fruitful discussions about INLA and nonergodic GMMs. I would like to thank Giovanni Lanzano and Luca Caramenti for clearing up some misunderstandings of MS-GWR on my part. Comments from Sreeram Kotha and an anonymous reviewer were very helpful to improve the manuscript. Partial support of Pacific Gas & Electric Company and California Department of Transportation are gratefully acknowledged. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect those of the sponsors.
Funding
This work was partially sponsored by the PG &E Geosciences Department and the California Department of Transportation.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
There is no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Kuehn, N. A comparison of nonergodic ground-motion models based on geographically weighted regression and the integrated nested laplace approximation. Bull Earthquake Eng 21, 27–52 (2023). https://doi.org/10.1007/s10518-022-01443-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10518-022-01443-7