Stochastic Generation of Accelerograms for Subduction Earthquakes

Bulletin of the Seismological Society of America, 2001

Acceleration time histories, recorded during the destructive 15 April 1979 (M 7.1) Montenegro earthquake, have been simulated using a stochastic modeling technique for finite faults proposed by Atkinson (1997, 1998a). In this approach, the ground-motion amplitudes are simulated as a summation of stochastic point sources. The length of the fault was taken as 70 km and its width as 25 km, and the fault plane was divided into 13 ‫ן‬ 5 elements. The applied methodology is tested against its ability to predict site-specific strong-motion records by the incorporation of mean frequency-dependent site-amplification factors, based on a gross characterization of the site class. The results show that the overall agreement between stochastic and recorded waveforms and spectra is quite satisfying. Nevertheless, significant discrepancies exist at certain stations, implying that site-amplification functions play an important role in the simulation process. A repetition of the simulations combined with the use of site-specific amplification function estimated by the horizontal-to-vertical ratios technique improved the fit to the observed time histories and spectra.

Bulletin of The Seismological Society of America, 2008

We simulated strong motion records from the Umbria-Marche, Central Italy earthquake (Mw 6) of September 1997 using a frequency-dependent S-wave radiation function. We compared the observed acceleration spectra, from strong-motion instruments located in the near field and at regional distances, with those simulated using the stochastic modeling technique of Beresnev and Atkinson (1997, 1998), and modified to account for a frequency dependent radiation pattern correction. By using the frequency-dependent radiation function previously obtained by Castro et al. (2006) we reduced the overall fitting error of the acceleration spectra by about 9%. In general, we observed that the frequency-dependent radiation pattern correction has a small effect on the spectral amplitudes compared with site effects, which is an important factor controlling the strong-motion records generated by the 1997 Umbria-Marche earthquake. In addition, we modeled the observed ground-motion records using the dynamic corner frequency model of to reproduce the directivity effects, reducing the average error of the spectral amplitudes by 24%. We concluded that although the frequencydependent radiation pattern correction affects the frequency content of the spectral amplitudes simulated, site and directivity effects are more relevant.

Probabilistic Engineering Mechanics, 2011

The sustained dissemination of databases of recorded accelerograms along with the increasing number of strong-motion networks installed worldwide revealed that the current methodologies for simulating artificial earthquakes possess the drawback that the simulated time-histories do not manifest the large variability of the seismological parameters as well as of the joint-time frequency distribution observed for natural accelerograms. As a consequence, the dispersion of the output of structural response analysis can be underestimated. In order to take into account the natural variability of earthquakes a methodology for simulating artificial earthquake accelerograms matching mean and mean ± standard deviation response spectra is proposed in this paper. This dispersion can be determined from attenuation relationships or evaluated from selected accelerograms of a strong-motion database. The procedure requires the definition of an evolutionary response-spectrum-compatible power spectral density function with random parameters. It is shown in the paper that the simulated ground motion time-histories will manifest variability similar to that one observed in natural records.

Doboku Gakkai Ronbunshu, 1988

A method for synthesizing strong-motion accelerogam from small earthquake records is presented with aid of the faulting source model. It is derived by extending the spectra relation satisfied in a point source model to ones in the faulting source model. The important terms involved in the method, namely, the scaling law of source spectra and the number of sub-faults are obtained statistically from the past strong-motion accelerograms. The method is applied to several representative earthquakes occurred in Japan, and it is shown that the synthesized accelerograms by the method agree relatively well with the observed ones in the points of amplitude, duration and spectra characteristics.

The great subduction earthquakes that occurred recently in Peru, Chile, and Japan have provided unprecedented information about the ground motions generated by such earthquakes. The 23 June 2001 M 8.4 Peru earthquake was recorded at eight strong-motion stations; the 27 February 2010 M 8.8 Maule, Chile, earthquake was recorded at over 10 strong-motion stations; and the 11 March 2011 M 9.0 Tohoku, Japan, earthquake was recorded at more than a thousand stations and produced the most extensive dataset of recordings for any earthquake. For the first time, data are available to guide the generation of ground-motion simulations from great subduction earthquakes. Broadband ground-motion simulations can enhance the usefulness of the recordings of these earthquakes by providing a means of interpolating and extrapolating the recorded data. Once they have been validated, broadband ground-motion simulations can be used for forward predictions of the ground motions of great subduction events in regions such as Cascadia, in which there are no strong-motion recordings of large subduction earthquakes. In this study, we test our ability to use a hybrid method to simulate broadband strong-motion recordings of megathrust earthquakes by demonstrating that our simulations reproduce the amplitudes of the recorded ground motions without systematic bias. We use simulations to study the distribution of various intensity measures of ground motion caused by these earthquakes and to validate our ground-motion simulation method by comparing the simulated ground motions with recorded ground motions as well as with empirical ground-motion prediction models.

Journal of physics. Conference series, 2024

Time-history analysis requires the use of suites of accelerograms that represent the input ground motion. A simple and computationally efficient stochastic methodology for the generation of suites of fully non-stationary artificial accelerograms is presented. The proposed methodology ensures that the record suites have a given target spectral mean and variability for the whole period range of interest. The model first produces an ensemble of target spectra with a given mean and variability and then a methodology based on spectral representation techniques is used to obtain the corresponding acceleration time-histories. The proposed approach also ensures that the produced ground motions are fully non-stationary and have different duration. The outcome is a suite of ground motion time-histories whose spectral mean and variability match those obtained from a ground motion model (GMM).

Soil Dynamics and Earthquake Engineering, 2009

An energy-based envelope function is developed for use in the stochastic simulation of earthquake ground motion. The envelope function is directly related to the Arias intensity of the ground motion as well to the manner in which this Arias intensity is built-up over time. It is shown that this build-up, represented by a Husid plot, can be very well modelled using a simple lognormal distribution. The proposed envelope makes use of parameters that are commonly available in seismic design situations, either following a deterministic scenario-type analysis or following a more comprehensive probabilistic seismic hazard analysis, either in terms of Arias intensity or the more common spectral acceleration. The shape parameters of the envelope function are estimated following the calculation of the analytic envelopes for a large number of records from PEER Next Generation of Attenuation (NGA) database. The envelope may also be used to predict the distribution of peak ground acceleration values corresponding to an earthquake scenario. The distribution thus obtained is remarkably consistent with those of the recent NGA models.

The strong motion accelerograms of the Matsushiro earthquake were analyzed by various methods. When the velocity and the displacement in the time domain and in the frequency domain are calculated from the accelerogram, various errors may be introduced. Therefore, it is necessary to check and select the methods of data processing according to the purpose of analyses. In this paper, it was shown that the comparison of observed values with theoretical in the form of velocity amplitude spectral density might be most preferable. The main factors affecting the ground vibration during disastrous earthquakes are generally conceived to be: source, path and local geology. The relations between the observed seismogram and the theoretical one were compared when a moving dislocation model was assumed. It was found that the direction of particle motion and the spectral density of observation could be reasonably understood from the theoretical point of view.

Bulletin of the Seismological Society of America, 2018

The 2011 Tohoku-Oki megathrust earthquake and its aftershocks were well recorded by the KiK-net network in accelerographs placed inside boreholes and on the surface. These data allow comparing strong-motion records with synthetic acceleration time histories for this large magnitude earthquake that caused extensive damage in Japan. Generating synthetic accelerograms at high frequencies can be approached using different techniques. We use the stochastic method to simulate horizontal and vertical strong-motion accelerograms in hard-rock boreholes; additionally, we incorporate P, SV, and SH soil amplification transfer functions to generate surface accelerograms. We reproduce the three components of the strong motion for 18 stations of the M w 9.0 mainshock event; additionally, we simulated 8 stations for an M w 6.9 aftershock. Our simulated acceleration time histories show similarity in time and frequency with the acceleration records for the period band between 0.05 and 1 s. Electronic Supplement: Table of the velocity model used in the modeling of our synthetic records, and figures showing comparison of time series and 5% response spectra of synthetic and real data of 2011 Tohoku-Oki megathrust earthquake and an M w 6.9 aftershock. BSSA Early Edition / 1

Bulletin of the Seismological Society of America, 2008

A waveform inversion with empirical Green's functions was conducted to estimate strong-motion generation areas (SMGAs) of the 2003 Tokachi-oki, Japan, earthquake (M w 8.3). Although the rupture process of this great subduction-zone earthquake has been revealed with waveform inversions based on low-frequency (lower than 0.2 Hz) ground motions, it is much more important from an engineering point of view to investigate how strong ground motions with higher frequencies were generated from the earthquake. Waveform data with higher frequencies have been excluded from the conventional inversions because of the difficulty in computing realistic theoretical Green's functions at higher frequencies. In this study, with the aid of empirical Green's functions, we extended the applicability of the waveform inversion to higher-frequency ground motions up to 1 Hz. We selected three aftershocks for use in the analysis, referring to the similarity of the group delay time between the mainshock and the aftershock records. The resultant slip model has three SMGAs, each of which is close to the hypocenter of one of the aftershocks used. Generally speaking, locations of the SMGAs identified in this study agree well with asperities identified from the inverted results using low-frequency (lower than 0.2 Hz) ground motions plus geodetic data and tsunamis. It implies that strong ground motions up to 1 Hz were generated from almost the same asperities producing lower-frequency ground motions. In terms of the complexity of slip models, although our analysis is focused on high frequencies, our slip model is at least as simple as conventional lowfrequency slip models. Such results would be useful in constructing source models of future great subduction-zone earthquakes for strong-motion prediction.