Ground Motion Model Research Articles

Regional and Site-Specific Ground Motion Model for Probabilistic Seismic Hazard Analysis in Taiwan: A Case Study of I-Lan

In this study, a probabilistic seismic hazard analysis (PSHA) using a traditional approach with a simple logic tree that handles a single-site sigma is developed to present a site-specific ground motion hazard. The analysis relies on the accelerometer data collected from the I-Lan Plain, a deep sedimentary basin in Northeastern Taiwan where hundreds of earthquake sequences (with [Formula: see text] = 4.0–7.6) were recorded by the Taiwan Strong Motion Instrument Program (TSMIP) network between 1992 and 2016. The performance of PSHA on soil and hard sites is evaluated from two representative instrumented sites (ILA048 and ILA025), using residual measures available from a ground motion model (GMM) developed by Phung et al. [ 2023 ] (referred to as Ph23), from which the uncertainties in site terms ([Formula: see text] and single-site sigma ([Formula: see text] are estimated. We address key conceptual issues in the current PSHA approach and introduce a new region- and site-specific PSHA approach in which (1) site-to-site variability ([Formula: see text] is estimated as a random variance in a mixed effects GMM regression and (2) the GMM site-specific single-site sigma ([Formula: see text] is replaced with a generic site-corrected aleatory variability ([Formula: see text]). Comparison of the region- and site-specific hazard curves from our method against the traditional method estimates at two well-recorded sites in the I-Lan region shows an approximate 50% difference in prediction ground motion values considering for 2% and 10% probability of exceedance in 50 years.

REVEAL: A Global Full-Waveform Inversion Model

ABSTRACT We present REVEAL, a global-scale, transversely isotropic full-waveform inversion model. REVEAL builds upon the earlier construction of the long-wavelength Earth (LOWE) model by lowering the minimum period from 100 to 33 s and by more than doubling the number of included earthquakes to 2366. In the course of 305 quasi-Newton iterations, REVEAL assimilated a total of 6,005,727 unique three-component waveforms. The inversion method rests on the combination of a stochastic mini-batch optimization and wavefield-adapted spectral-element meshes. Although the former naturally exploits redundancies in the data set, the latter reduces the cost of wavefield simulations by reducing the effective dimension of the numerical mesh. As a consequence, the average cost of an iteration in this inversion is only around 0.62% of an iteration that uses the complete data set with a standard cubed-sphere-type mesh. We calculated 3D synthetic seismograms using a graphics processing unit-accelerated spectral-element wave propagation solver, accommodating the effects of anelasticity, topography, bathymetry, ocean loading, and ellipticity. For a diverse range of global wavepaths, REVEAL predicts complete three-component seismograms at 33 s period that have not been included in the inversion. This generalization to unseen data suggests applications of REVEAL in event location and characterization, as well as in ground-motion modeling.

Empirical ground-motion basin response in the California Great Valley, Reno, Nevada, and Portland, Oregon

We assess how well the Next-Generation Attenuation-West 2 (NGA-West2) ground-motion models (GMMs), which are used in the US Geological Survey’s (USGS) National Seismic Hazard Model (NSHM) for crustal faults in the western United States, predict the observed basin response in the Great Valley of California, the Reno basin in Nevada, and Portland and Tualatin basins in Oregon. These GMMs rely on site parameters such as the time-averaged shear-wave velocity ( VS) in the upper 30 m of Earth’s crust ( VS30) and depths to 1.0 and 2.5 km/s shear-wave isosurfaces ( Z1.0 and Z2.5) to capture basin effects and were developed using observations and simulations primarily from the Los Angeles region in southern California. Using ground-motion records from mostly small-to-moderate earthquakes and mixed-effects regression analysis, we find that the GMMs perform well with our local basin-depth models for the California Great Valley. With our local basin-depth models for Reno, the GMMs do not perform as well for this relatively shallow basin and exhibit little sensitivity to the basin parameters used in the NGA-West2 GMMs. We also find good performance for the local Z1.0 model across the Portland region, whereas the local Z2.5 model provides little predictive power except at sites in the deepest part of the Tualatin basin. Additional work could improve the performance of the site and basin terms in the NGA-West2 GMMs for regions with geologic structure different than the deep basins in southern California and the Great Valley. In addition, we find significant discrepancies among the GMMs in how the uncertainty in the ground motion varies with basin depth and pseudospectral period. Our results can help guide seismic hazard analyses on whether to include these local basin-depth models.

Adjusting an active shallow crustal ground motion model to regions with scarce data: application to France

The objective of this work is to test whether an empirical Ground Motion Model (GMM) developed for high-seismicity regions can be effectively adapted to a neighbouring region with lower seismic activity. We select the ITA18 suite of GMMs (Lanzano et al. in Bull Seismol Soc Am 109(2): 525-540, 2019a), developed for Italy, which is a region dominated by moderate-to-strong shallow crustal earthquakes, and assess their applicability to Continental France, where the seismic activity is less frequent and characterised by lower magnitudes. Based on a dataset of more than 2300 records of events with 3.0 ≤ MW ≤ 5.2, occurred in France in the time interval 1996–2019 (named FR20), we perform a residual analysis and calibrate an adjustment factor for both horizontal and vertical-to-horizontal (VH) components of Peak Ground Acceleration, Peak Ground Velocity, and 5% damped Spectral Acceleration (SA). Apart from the median correction, no modification of the scaling with magnitude, focal mechanism, and VS,30 is introduced, while the distance scaling is adjusted to capture the lower anelastic attenuation of the French data. In addition, to overcome the underestimation of the ITA18 model for the short period VH spectral amplitudes in the near-source region (Repi < 15 km), an additional empirical corrective factor is introduced. In spite of the good agreement of the adjusted model with respect to the median trends of the FR20 dataset, a regionalization of the source effects is introduced to reduce the relatively high between-event variability of the proposed model. The proposed model provides predictions similar to ITA18 in the most seismically active regions (Alps or Pyrenees), while, in the other zones, the predicted amplitudes are richer at high frequencies. Given the paucity of seismic records in these zones, this behavior should be confirmed on the basis of additional data (e.g. physics-based simulations, geologic and tectonic features). The use of the proposed model for hazard applications is recommended within the validity limits of the data (3.0 ≤ MW ≤ 5.2). However, the similarity of the ground motion in the Alps and Pyrenees with the predictions of ITA18 suggests that the adjusted model could be also employed for higher magnitudes, upon suitable checks.

Horizontal ground-motion model for subduction slab earthquakes using offshore ground motions in the Japan Trench area

The S-net, a large-scale network of permanent ocean-bottom seismographs in the Japan Trench, consisting of 150 stations, has recorded a large number of offshore ground motions that can be used to establish the empirical attenuation relationship of offshore ground motions in the subduction zone. Due to the different attenuation characteristics among different earthquake types in subduction zones, the earthquakes are classified into four tectonic types of subduction slab, interface, shallow crustal, and upper-mantle earthquakes according to the classification scheme of Zhao et al. (2015). Predictive models and uncertainties in offshore ground motions were investigated for different earthquake types. This study establishes a horizontal offshore subduction slab earthquake ground-motion model (GMM) and compares the offshore and onshore slab earthquake GMMs. As the site conditions at ocean-bottom stations are different from those at land stations, the effects of water depth and sediment thickness are taken into account in the offshore slab earthquake GMM. Due to differences in the burial methods of ocean-bottom stations, stations were divided into buried and unburied to investigate the effects of the burial methods. Therefore, regression analysis was used to propose an offshore slab earthquake GMM considering the magnitude, focal depth, distance, water depth, sediment thickness, and burial method. By separating the within-event residuals, the single-station standard deviation is presented. Compared to the onshore GMM, the predicted spectra of the offshore GMM are significantly larger at long periods. For the attenuation rate, the offshore attenuation rate is lower than that of the onshore attenuation rate for short periods, but is basically consistent with the onshore attenuation rate for long periods. The proposed GMM can be used to predict the offshore ground motions for slab earthquakes with rupture distances less than 300 km, focal depths less than 110 km, and moment magnitude between 4 and 7.4.

Ranking and developing ground-motion models for Southeastern Africa

The southern East African Rift System (EARS) is an early-stage continental rift with a deep seismogenic zone. It is associated with a low-to-moderate seismic hazard, but due to its short and sparse instrumental record, there is a lack of ground-motion studies in the region. Instead, seismic hazard assessments have commonly relied on a combination of active crustal and stable continental ground-motion models (GMMs) from other regions without accounting for the unusual geological setting of this region and evaluating their suitability. Here, we use a newly compiled southern EARS ground-motion database to compare six active crustal GMMs and four stable continental GMMs. We find that the active crustal GMMs tend to underestimate the ground-motion intensities observed, while the stable continental GMMs overestimate them. This is particularly pronounced in the high-frequency intensity measures (&gt;5 Hz). We also use the referenced empirical approach and develop a new region-specific GMM for southern EARS. Both the ranked GMMs and our new GMM result in large residual variabilities, highlighting the need for local geotechnical information to better constrain site conditions.

Basin effects from 3D simulated ground motions in the Greater Los Angeles region for use in seismic hazard analyses

We develop basin-depth-scaling models (i.e. “basin terms”) from the long-period ([Formula: see text]) simulated ground motions of the Southern California Earthquake Center (SCEC) CyberShake project for use in seismic hazard analyses at sites within the sedimentary basins of southern California. Basin terms use the Next Generation Attenuation (NGA)-West-2 ground-motion models (GMMs) as reference models and use their functional forms with slight modifications. We investigate the use of two approaches to incorporate the time-averaged shear-wave velocity in the upper 30 m ([Formula: see text]) in these calculations and find that the use of site-specific and uniform [Formula: see text] has minor effects on the resulting basin terms for this data set. By centering the simulated ground motions on the basin terms, we separate the information from the simulations about absolute ground-motion level from information relating to the relative amplifications, such as the differences between shallow- and deep-basin sites. Recent observations from sedimentary basins of southern California indicate that additional amplification effect may persist at relatively shallow basin depths (i.e. the GMM basin terms should have positive values when differential depths, [Formula: see text], are near zero), and we present models for “centered” and “adjusted” basin-depth scaling models that reflect this potential. The simulation-modified GMMs are appropriate for crustal sources and for deep-basin sites ([Formula: see text]) within basins of the Greater Los Angeles region, for the magnitudes and distances defined by each of the reference NGA-West-2 GMMs.

SSA Future Directions for Physics-Based Ground-Motion Modeling Meeting Report

SSA Future Directions for Physics-Based Ground-Motion Modeling Meeting Report

Ground-Motion Model for Small-to-Moderate Potentially Induced Earthquakes Using an Ensemble Machine Learning Approach for CENA

ABSTRACT Ground-motion models (GMMs) are vital in assessing probabilistic seismic hazards and uncertainties. This study develops new GMMs benefiting from nonparametric machine learning algorithms, including artificial neural networks, kernel ridge, random forest, and gradient boosting regression techniques for small-to-moderate potentially induced earthquakes in central and eastern North America (CENA). As part of this study, we evaluate the performance of different machine learning models in estimating peak ground acceleration (PGA) and 17 spectral accelerations based on the moment magnitude (Mw), hypocentral distance (Rhypo), and the timed-average shear-wave velocity of the upper 30 m of soil (VS30). To train the algorithms, we have utilized a database of nearly 31,000 ground motions with small and moderate moment magnitudes ranging from 3.0 to 5.8, recorded within a hypocentral distance of less than 200 km in CENA. Typically, for GMM development, analysts employ linear regression-based models with predefined functional forms. The requirement for predefined functional forms can restrict the use of complicated and nonlinear equations to improve performance. Although the conventional regression model is more interpretable, machine learning can achieve a better result given sufficient training data. The results of error metrics reveal that gradient-boosting regression provides a better performance. Furthermore, a machine learning ensemble method is used to combine the regression results of four machine learning algorithms. The ensemble method improves the GMM performance and provides smoother results.

Summary of the Discussions During the 2023 SSA Topical Meeting on “Future Directions for Physics-Based Ground Motion Modeling”

Summary of the Discussions During the 2023 SSA Topical Meeting on “Future Directions for Physics-Based Ground Motion Modeling”

Machine learning-based ground motion models for predicting PSAs of borehole motions in Japan

Machine learning-based ground motion models for predicting PSAs of borehole motions in Japan

A New Spatial Variation Model for Ground-Motion Intensities Combined with Correlation and Coherency

Abstract Regional seismic risk or loss assessments generally require simulation of spatially distributed ground motions using multiple intensity measures. Hence, in this study, ground-motion model estimation is performed with a spatial correlation. Previously, many researchers have analyzed spatial correlations and developed empirical models using ground-motion recordings. In this study, ground motions occurring in California between 2019 and 2023 were used to analyze spatial correlations using semivariograms for the peak ground acceleration and pseudospectral acceleration in various spectral periods. Based on the analysis results, two aspects need to be revised in the conventional correlation model: (1) the empirical exponential model cannot reasonably reflect the target spatial correlation at a separation distance &amp;lt;10 km, and (2) the variation in the spatial correlation ground-motion intensity cannot be described at a small separation distance &amp;lt;1 km. Owing to these limitations, we revised the fitting model of the semivariogram to better characterize the spatial correlation. In the model, another function called coherency, replaced the spatial correlation to characterize the variation in the Fourier phase rather than the intensity within a separation distance &amp;lt;1 km. This research shows that the spatial variation in any region can be analyzed by combining the coherence and correlation functions for practical seismic-risk or loss assessment problems.

Stochastic emulation with enhanced partial‐ and no‐replication strategies for seismic response distribution estimation

AbstractModern performance earthquake engineering practices frequently require a large number of time‐consuming non‐linear time‐history simulations to appropriately address excitation and structural uncertainties when estimating engineering demand parameter (EDP) distributions. Surrogate modeling techniques have emerged as an attractive tool for alleviating such high computational burden in similar engineering problems. A key challenge for the application of surrogate models in earthquake engineering context relates to the aleatoric variability associated with the seismic hazard. This variability is typically expressed as high‐dimensional or non‐parametric uncertainty, and so cannot be easily incorporated within standard surrogate modeling frameworks. Rather, a surrogate modeling approach that can directly approximate the full distribution of the response output is warranted for this application. This approach needs to additionally address the fact that the response variability may change as input parameter changes, yielding a heteroscedastic behavior. Stochastic emulation techniques have emerged as a viable solution to accurately capture aleatoric uncertainties in similar contexts, and recent work by the second author has established a framework to accommodate this for earthquake engineering applications, using Gaussian Process (GP) regression to predict the EDP response distribution. The established formulation requires for a portion of the training samples the replication of simulations for different descriptions of the aleatoric uncertainty. In particular, the replicated samples are used to build a secondary GP model to predict the heteroscedastic characteristics, and these predictions are then used to formulate the primary GP that produces the full EDP distribution. This practice, however, has two downsides: it always requires minimum replications when training the secondary GP, and the information from the non‐replicated samples is utilized only for the primary GP. This research adopts an alternative stochastic GP formulation that can address both limitations. To this end, the secondary GP is trained by measuring the square of sample deviations from the mean instead of the crude sample variances. To establish the primitive mean estimates, another auxiliary GP is introduced. This way, information from all replicated and non‐replicated samples is fully leveraged for estimating both the EDP distribution and the underlying heteroscedastic behavior, while formulation accommodates an implementation using no replications. The case study examples using three different stochastic ground motion models demonstrate that the proposed approach can address both aforementioned challenges.

Hybrid Empirical Ground-Motion Models for the Island of Hawaii Based on an Updated Strong Ground-Motion Database

ABSTRACT In this study, we develop ground-motion models (GMMs) for the Island of Hawaii. This area has been the site of several significant earthquake events with a growing database of strong ground-motion observations. Ground-motion modeling on the Island of Hawaii is challenging due to different anelastic attenuation characteristics, the volcanic origin of some of the events, and event depth distribution. Only a few GMMs have been developed for the Island of Hawaii. In this study, we apply a hybrid empirical method (HEM) to develop two separate GMMs for shallow (hypocentral depth ≤ 20 km) and deep (hypocentral depth &amp;gt; 20 km) earthquakes on the Island of Hawaii. We utilize the ratio of the stochastic point-source model in the target and host regions as an appropriate adjustment factor. We apply these adjustment factors to convert the GMMs from the host (western North America) to the target (Island of Hawaii) region. We considered five GMMs proposed in the Next Generation Attenuation Phase 2 project by the Pacific Earthquake Engineering Research Center to model ground motions in the host region. We developed GMMs to predict peak ground acceleration and 5%-damped pseudospectral acceleration at periods T = 0.01–10 s, for moment magnitudes (M) in the range of 3–7.5, and for Joyner–Boore distances in the RJB≤200 km range. The applicability of HEM to develop GMMs for the Island of Hawaii and the growing strong ground-motion data result in further improvements in the capability of GMMs.

Broadband Ground-Motion Synthesis via Generative Adversarial Neural Operators: Development and Validation

ABSTRACT We present a data-driven framework for ground-motion synthesis that generates three-component acceleration time histories conditioned on moment magnitude (M), rupture distance (Rrup), time-average shear-wave velocity at the top 30 m (VS30), and style of faulting. We use a Generative Adversarial Neural Operator (GANO)—a resolution invariant architecture that guarantees model training independent of the data sampling frequency. We first present the conditional ground-motion synthesis algorithm (cGM-GANO) and discuss its advantages compared to the previous work. We next train cGM-GANO on simulated ground motions generated by the Southern California Earthquake Center Broadband Platform (BBP) and on recorded the Kiban–Kyoshin network (KiK-net) data, and show that the model can learn the overall magnitude, distance, and VS30 scaling of effective amplitude spectra (EAS) ordinates and pseudospectral accelerations (PSA). Results specifically show that cGM-GANO produces consistent median scaling with the training data for the corresponding tectonic environments over a wide range of frequencies for scenarios with sufficient data coverage. For the BBP dataset, cGM-GANO cannot learn the ground-motion scaling of the stochastic frequency components (f &amp;gt; 1 Hz); for the KiK-net dataset, the largest misfit is observed at short distances (Rrup&amp;lt;50 km) and for soft-soil conditions (VS30&amp;lt;200 m/s) due to the scarcity of such data. Except for these conditions, the aleatory variability of EAS and PSA are captured reasonably well. Finally, cGM-GANO produces similar median scaling to traditional ground-motion models (GMMs) for frequencies greater than 1 Hz for both PSA and EAS but underestimates the aleatory variability of EAS. Discrepancies in the comparisons between the synthetic ground motions and GMMs are attributed to inconsistencies between the training dataset and the datasets used in GMM development. Our pilot study demonstrates GANO’s potential for efficient synthesis of broadband ground motions.

An approach for predicting surface strong motion using borehole seismometers

South Korea has implemented borehole-type seismometers for reliable earthquake observations and earthquake early-warning systems, with approximately 85% of seismometers being replaced by borehole-type seismometers after the Gyeongju earthquake. Although these seismometers are more effective at detecting earthquakes owing to the reduced artificial ambient noise, they do not record surface-level shaking. Therefore, it is necessary to estimate ground surface shaking directly associated with potential damage when using borehole-type seismometers without surface sensors. This study investigated and compared various methods, including the stochastic point-source ground-motion model, transfer function based on ambient noise, and one-dimensional site response, to estimate horizontal seismograms of the ground surface. We assessed the accuracy of these methods by comparing the waveforms generated in event cases (magnitude from 2.5 to 5.8, with epicentral distances spanning 22 km–209 km) in terms of Fourier spectra, intensity, and spectral acceleration. Among the methods assessed, the transfer function approach, which does not account for the geophysical characteristics such as VS30, proved to be the most appropriate for correcting ground-surface effects.

Surrogate model‐aided global sensitivity analysis framework for seismic consequences estimation in buildings

AbstractSeismic consequences estimation for individual buildings is valuable for various stakeholders, including government entities, building owners, and insurers. The robustness of estimation results in the presence of incomplete input information can typically be investigated through sensitivity analysis. However, the estimation process's complexity and the sensitivity analysis's computational burden hinder its practical application, which requires a more efficient procedure to facilitate broader use. This paper proposes a novel framework for sensitivity analysis of seismic consequences estimation to improve the efficiency and reliability of such analysis. The proposed approach encompasses three key components: (1) stochastic ground motion modeling (SGMM)‐based seismic consequences estimation to evaluate the economic, environmental, and social consequences given specific buildings by considering different hazard levels, (2) the training of surrogate model (Gaussian process model) for structural analysis to reduce the computational cost of the evaluation process, and (3) variance‐based global sensitivity analysis to investigate the importance of parameters of concern in the estimation process. The entire procedure is implemented in Python, adhering to object‐oriented programming, and does not rely on external software. Then, the proposed methodology is applied to two distinct three‐story steel moment‐resistant frames (SMRFs) subjected to four different hazard levels to demonstrate its effectiveness. The SGMM method can generate specific ground motions for each hazard level, mitigating the potential for result bias from using ground motions with unrealistic characteristics. Furthermore, the SGMM method is particularly suitable for automated analysis processes reducing the laborious task of screening ground motions from the database. Comparative analysis with surrogate‐free estimation reveals that the surrogate‐based analysis delivers reliable results with significantly reduced computational cost. The results of analyzing different structures under varying hazard levels reflect the variability in sensitivity analysis of consequences estimation, highlighting the necessity of the proposed flexible and efficient framework. Furthermore, the proposed framework's advantages, limitations, and future research needs are discussed.

Data-driven adjustments for combined use of NGA-East hard-rock ground motion and site amplification models

Model development in the Next Generation Attenuation-East (NGA-East) project included two components developed concurrently and independently: (1) earthquake ground-motion models (GMMs) that predict the median and aleatory variability of various intensity measures conditioned on magnitude and distance, derived for a reference hard-rock site condition with an average shear-wave velocity in the upper 30 m ( VS30) = 3000 m/s; and (2) a site amplification model that modifies intensity measures for softer site conditions. We investigate whether these models, when used in tandem, are compatible with ground-motion recordings in central and eastern North America (CENA) using an expanded version of the NGA-East database that includes new events from November 2011 (end date of NGA-East data curation) to April 2022. Following this expansion, the data set has 187 events, 2096 sites, and 16,272 three-component recordings, although the magnitude range remains limited (∼4 to 5.8). We compute residuals using 17 NGA-East GMMs and three data selection criteria that reflect within-CENA regional variations in ground-motion attributes. Mixed-effects regression of the residuals reveals a persistent pattern in which ground motions are overpredicted at short periods (0.01–0.6 s, including peak ground acceleration (PGA)) and underpredicted at longer periods. These misfits are regionally variable, with the Texas–Oklahoma–Kansas region having larger absolute misfits than other parts of CENA. Two factors potentially influencing these misfits are (1) differences in the site amplification models used to adjust the data to the reference condition during NGA-East GMM development relative to CENA amplification models applied since the 2018 National Seismic Hazard Model (NSHM), and (2) potential bias in simulation-based factors used to adjust ground motions from the hard-rock reference condition to a VS30 = 760 m/s condition. We provide adjustment factors and their epistemic uncertainties and discuss implications for applications.

Constraining Between-Event Variability of Kinematic Rupture Scenarios by Empirical Ground-Motion Model: A Case Study in Central Italy

ABSTRACT The region of central Italy is well known for its moderate-to-large earthquakes. Events such as 2016 Mw 6.2 Amatrice, generated in the shallow extensional tectonic regime, motivate numerical simulations to gain insights into source-related ground-motion complexities. We utilize a hybrid integral–composite kinematic rupture model by Gallovič and Brokešová (2007) to predict ground motions for other hypothetical Amatrice fault rupture scenarios (scenario events). The synthetic seismograms are computed in 1D crustal velocity models, including region-specific 1D profiles for selected stations up to 10 Hz. We create more than ten thousand rupture scenarios by varying source parameters. The resulting distributions of synthetic spectral accelerations at periods 0.2–2 s agree with the empirical nonergodic ground-motion model of Sgobba et al. (2021) for central Italy in terms of the mean and total variability. However, statistical mixed-effect analysis of the residuals indicates that the between-event variability of the scenarios exceeds the empirical one significantly. We quantify the role of source model parameters in the modeling and demonstrate the pivotal role of the so-called stress parameter that controls high-frequency radiation. We propose restricting the scenario variability to keep the between-event variability within the empirical value. The presented validation of the scenario variability can be generally utilized in scenario modeling for more realistic physics-based seismic hazard assessment.

Site-specific stochastic ground motion model utilizing deterministic physics-informed simulations: A Bayesian approach

Limited availability of recorded ground motions poses a challenge for reliable probabilistic seismic-hazard analysis (PSHA), even in highly seismic regions like the Western United States. Stochastic ground motions are commonly employed to address this challenge. However, the stochastic ground motion models (GMMs) may not consistently generate ground motions compatible with the site hazard due to their calibration using global data, failing to capture site-specific characteristics adequately. In the absence of recorded motions, physics-informed simulations provide a viable alternative but are deterministic with limitations of their own that makes them challenging to support PSHA. This article introduces a Bayesian framework that combines prior knowledge from a stochastic GMM, calibrated with global data, with site-specific data obtained from deterministic physics-informed simulations. The proposed framework utilizes the Rezaeian–Der Kiureghian (2010) model as the stochastic GMM and incorporates site-specific data from the CyberShake 15.12 study. By updating the mean and variance of the predictive relationships, along with the marginal distribution of the model parameters, through Bayesian inference, this framework allows for the simulation of site-specific ground motions consistent with the site characteristics. The statistics of peak ground acceleration distributions, as well as both the median and variability of the elastic response spectra, obtained from the calibrated stochastic GMM, demonstrate consistency with those derived using GMMs based on the Next Generation Attenuation (NGA) database.