Some Thoughts on Declustering in Probabilistic Seismic-Hazard Analysis

Marzocchi, Warner; Taroni, Matteo

doi:10.1785/0120130300

Cited by 81 publications

(64 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The importance of aftershocks in seismic hazard has been highlighted previously. For instance, Marzocchi and Taroni (2014) concluded that aftershocks should be included in probabilistic seismic hazard analysis (PSHA), because only using the PSHA result from the declustered catalog underestimated hazard rates. Approaches for including aftershocks can be distinguished by the treatment of mainshocks: (1) long-term (years to centuries), timeindependent approaches that integrate aftershock hazard on unconditional mainshock occurrence; (2) short-term (days to years) approaches that estimate aftershock hazard conditional on a specified mainshock source.…”

Section: Introductionmentioning

confidence: 99%

Spatiotemporal Seismic Hazard and Risk Assessment of Aftershocks ofM 9 Megathrust Earthquakes

Zhang

Werner

Goda

2018

Bulletin of the Seismological Society of America

View full text Add to dashboard Cite

Current national seismic hazard models neglect time-dependent hazard due to triggered earthquakes, although these can certainly generate damaging ground motions. To understand the relative importance of aftershock hazard and risk in the context of a mega-thrust subduction-zone earthquake, we develop a new simulation framework for spatiotemporal seismic hazard and risk assessment of a mega-thrust earthquake and its aftershocks along the plate boundary and in the onshore continental crust. Tohoku region in the northeast Japan is considered as an example to show how the new simulation framework can be implemented to assess the spatiotemporal hazard and risk of aftershocks triggered by a M9 Tohoku-like earthquake. We generate quasi-3D synthetic aftershock catalogs using the Epidemic Type Aftershock Sequences (ETAS) model, modified to characterize aftershocks of large and anisotropic finite mainshock sources. By including the mainshock source model in the new simulation framework, the uncertainty of generating synthetic aftershock catalog is small in comparison with the observation. Therefore, should the mainshock source model is available right after the mainshock, the new simulation framework can be used for the quasi-real time hazard and risk assessments of aftershocks in different regions. For Tohoku region, we assess the relative importance of subduction-zone versus onshore-crustal aftershocks. The results show that the subduction-zone aftershocks tend to dominate hazard with peak ground velocity (PGV) < 60 cm/s (the boundary between VIII (severe) and IX (violent) of Modified Mercalli Intensity). On the other hand, onshore-crustal aftershocks control extreme hazards exceeding PGV of 60 cm/s. Moreover, on the day of the mainshock, aftershocks contribute about 23% of the onshore hazard with PGV > 60 cm/s, and the aftershock hazards remain relatively high for 4-5 days depending on different sites. From a seismic risk viewpoint, the subduction-zone and onshore-crustal aftershocks in the mega-thrust sequence affect buildings differently; both have similar potential to cause minor damage, whilst the latter tend to cause more severe damage.

show abstract

Section: Introductionmentioning

confidence: 99%

Spatiotemporal Seismic Hazard and Risk Assessment of Aftershocks ofM 9 Megathrust Earthquakes

Zhang

Werner

Goda

2018

Bulletin of the Seismological Society of America

View full text Add to dashboard Cite

show abstract

“…that an immigrant event is equally likely to occur at each point in time, and that the time between each pair of immigrant events (known as the 'inter-arrival times') follows a time-independent Exponential(μ) distribution. However, this conflicts with findings elsewhere in the seismology literature, where there is substantial doubt over whether the occurrence times of mainshock earthquakes is really Poissonian (Tahernia et al 2014;Ordaz and Arroyo 2016;Marzocchi and Taroni 2014). Although ETAS immigrant events are not strictly equivalent to mainshocks as defined elsewhere in the seismology literature (since there is no requirement that an ETAS immigrant should have larger magnitude than its offspring), this still seems to cast some doubt on the Poissonian assumption.…”

Section: Introductionmentioning

confidence: 76%

Inference for ETAS models with non-Poissonian mainshock arrival times

Kolev

Ross

2018

Stat Comput

View full text Add to dashboard Cite

The Hawkes process is a widely used statistical model for point processes which produce clustered event times. A specific version known as the ETAS model is used in seismology to forecast earthquake arrival times under the assumption that mainshocks follow a Poisson process, with aftershocks triggered via a parametric kernel function. However, this Poissonian assumption contradicts several aspects of seismological theory which suggest that the arrival time of mainshocks instead follows alternative renewal distributions such as the Gamma or Brownian Passage Time. We hence show how the standard ETAS/Hawkes process can be extended to allow for non-Poissonian distributions by introducing a dependence based on the underlying process' behaviour. Direct maximum likelihood estimation of the resulting models is not computationally feasible in the general case, so we also present a novel Bayesian MCMC algorithm for efficient estimation using a latent variable representation. Keywords ETAS • Stress release • Renewal process • Hawkes process • Brownian passage times • RHawkes

show abstract

“…In this appraisal, the fact that earthquakes follow a Poisson distribution in time has been considered to minimize any spatial bias of the earthquake occurrence process, as it usually is in most seismic-hazard assessments, both in zoning and nonzoning methods. Some recent papers (e.g., Boyd, 2012;Marzocchi and Taroni, 2014) advise the use of aftershocks in the seismic-hazard computation, noticing the underestimation of the seismic hazard when the declustering process is carried out, and proposing different initial approaches to correct it. Taking into account the characteristics of the catalog used in this work, especially its spatial completeness, as well as the different behavior of observed seismic sequences, among other issues, we were ultimately led to follow a typical zoning method.…”

Section: Methodsmentioning

confidence: 99%

Updated Probabilistic Seismic‐Hazard Values for Egypt

Sawires

Peláez

Fat-Helbary³

et al. 2016

Bulletin of the Seismological Society of America

View full text Add to dashboard Cite

Seismic hazard in terms of mean peak ground acceleration (PGA) and spectral acceleration (SA) values has been computed for Egypt using both historical and instrumental earthquake data. For this purpose, an updated earthquake catalog, spanning the time period from 2200 B.C. to 2013, has been compiled for Egypt as well as its surrounding regions and is prepared to be used in a new probabilistic seismichazard assessment of Egypt. The earthquakes sizes were unified in terms of the moment magnitude scale. A new seismic source model for the seismic activity in and around Egypt, consisting of a total of 88 seismic zones (for shallow-and intermediate-depth seismicity), was considered in this new assessment. The seismicity parameters have been specifically computed for 35 seismic sources covering the Egyptian territory and the Eastern Mediterranean region. A logic-tree design was set up to consider the epistemic uncertainty in the Gutenberg-Richter b-value, maximum possible magnitude (M max ), and the selected ground-motion prediction equations. Seismichazard computations for rock-site conditions with 10% and 5% probability of exceedance in 50 years were carried out. In addition, uniform hazard spectra for twelve, among the most important and populated cities in Egypt, are computed and compared with the most recent Egyptian building code values. It is interesting to highlight that the maximum hazard values are observed at the Gulf of Aqaba region, specifically around the epicentral location of the biggest Egyptian recorded earthquake of 22 November 1995 (M w 7.2) Aqaba earthquake. The obtained seismic-hazard values for Nuweiba city (located in this region) for mean PGA and SA (0.1 s) are 0:29g and 0:74g, respectively, for a return period of 475 years.

show abstract

Some Thoughts on Declustering in Probabilistic Seismic-Hazard Analysis

Cited by 81 publications

References 18 publications

Spatiotemporal Seismic Hazard and Risk Assessment of Aftershocks ofM 9 Megathrust Earthquakes

Spatiotemporal Seismic Hazard and Risk Assessment of Aftershocks ofM 9 Megathrust Earthquakes

Inference for ETAS models with non-Poissonian mainshock arrival times

Updated Probabilistic Seismic‐Hazard Values for Egypt

Contact Info

Product

Resources

About