Ground Motion Model Research Articles

Design standardisation and seismic protection of SMRs through modular metafoundations

This paper investigates the seismic protection of the Nuward™ small modular reactor (SMR) building, focusing on design loading and beyond design basis earthquake (bDBE) conditions. The study aims to achieve two primary objectives: i) to enhance seismic mitigation of a SMR building under bDBE conditions, through the use of the innovative modular single-layer (SLM) and multi-layer (MLM) metafoundations (MFs); ii) to effectively standardise and harmonise SMR building designs in locations prone to beyond design basis conditions. To accomplish these goals and demonstrate the protective capabilities of the MFs, the study employs non-linear time-history analyses (NLTHAs) for both DBE and bDBE conditions. Along these lines, a reduced-order model was developed from a refined finite element (FE) model of the SMR building using the Craig-Bampton mode synthesis technique. Then, finite locally resonant modular MFs were designed and analysed using NLTHAs. Specifically, physics-based ground motion models (GMMs) were used to generate and select seismic triplets that mimicked DBE and bDBE scenarios for NLTHAs. Successively to achieve improved seismic performance, the optimization of the MFs was pursued by targeting the optimal number of columns, resonator parameters, and unit cell dimensions. Additionally, the deployment of inerters was considered, to significantly reduce the size of the MFs and enable their application in multiple layers for ultra-low frequency attenuation. The overall findings suggest that modular MFs meet seismic protection requirements, and positively contribute to the standardization process of SMR buildings, even in areas characterized by beyond-design seismic conditions.

Nonparametric ground motion models of arias intensity and significant duration for the Italian dataset

Abstract In the field of ground motion simulation, the stochastic site-based methodology relies on the existing database of ground shaking. Based on these methodologies, several properties of seismic signals are used to simulate seismic waves. These parameters could be evaluated either parametrically via linear or nonlinear regression techniques or non-parametrically via sophisticated machine-learning algorithms. Nonetheless, parametric models, which consist of a particular mathematical formulation, can be a source of large bias. In this study, machine learning techniques are employed to develop predictive models for two main input parameters of a stochastic site-based ground motion model: Arias intensity and significant duration, which control the time variation of the simulated ground shakings. The Arias intensity, defined by the integral of the square of the acceleration time series, and the significant duration, which is related to the strong shaking phase of an earthquake, are also of particular interest in structural and geotechnical engineering fields. For this purpose, the random forest approach is employed to develop prediction models for the Italian database. To guarantee the prediction accuracy of the models also for unseen future data, only 80 percent of the data is used for training, and the rest is reserved for testing the trained model. The model hyperparameters are tuned to control bias and variance trade-offs by k-fold cross-validation. For each model, a set of hyperparameters is selected, and a possible range is given. Then, a Bayesian optimization technique is implemented to find the best set of these hyperparameters among the given range. All these models provided promising results compared to the prior models in the literature.

Event-specific ground motion anomalies highlight the preparatory phase of earthquakes during the 2016–2017 Italian seismicity

Although physical models are improving our understanding of the crustal processes that lead to large earthquakes, observing their preparatory phases is still challenging. We show that the spatio-temporal evolution of the ground motion of small magnitude earthquakes can shed light on the preparatory phase of three main earthquakes that occurred in central Italy between 2016 and 2017. We analyze systematic deviations of peak ground accelerations generated by each earthquake from the values predicted by a reference ground motion model calibrated for background seismicity and refer to such deviations as event-specific ground motion anomalies (eGMAs). The eGMA temporal behavior indicates that during the activation phase of the main earthquakes, the ground shaking level deviates, positively or negatively, from the values expected for the background seismicity. eGMA can be exploited as beacons of stress change and help to monitor the mechanical state of the crust and the nucleation of large earthquakes.

Evaluation of the applicability of ground motion models (GMMs) for South Korea

Evaluation of the applicability of ground motion models (GMMs) for South Korea

The importance of corner frequency in site‐based stochastic ground motion models

AbstractSynthetic ground motions (GMs) play a fundamental role in both deterministic and probabilistic seismic engineering assessments. This paper shows that the family of filtered and modulated white noise stochastic GM models overlooks a key parameter—the high‐pass filter's corner frequency, . In the simulated motions, this causes significant distortions in the long‐period range of the linear‐response spectra and in the linear‐response spectral correlations. To address this, we incorporate as an explicitly fitted parameter in a site‐based stochastic model. We optimize by individually matching the long‐period linear‐response spectrum (i.e., for ) of synthetic GMs with that of each recorded GM. We show that by fitting the resulting stochastically simulated GMs can precisely capture the spectral amplitudes, variability (i.e., variances of ), and the correlation structure (i.e., correlation of between distinct periods and ) of recorded GMs. To quantify the impact of , a sensitivity analysis is conducted through linear regression. This regression relates the logarithmic linear‐response spectrum () to 7 GM parameters, including the optimized . The results indicate that the variance of observed in natural GMs, along with its correlation with the other GM parameters, accounts for 26% of the spectral variability in long periods. Neglecting either the variance or correlation typically results in an important overestimation of the linear‐response spectral correlation.

Modelling seismic ground motion and its uncertainty in different tectonic contexts: challenges and application to the 2020 European Seismic Hazard Model (ESHM20)

Abstract. Current practice in strong ground motion modelling for probabilistic seismic hazard analysis (PSHA) requires the identification and calibration of empirical models appropriate to the tectonic regimes within the region of application, along with quantification of both their aleatory and epistemic uncertainties. For the development of the 2020 European Seismic Hazard Model (ESHM20) a novel approach for ground motion characterisation was adopted based on the concept of a regionalised scaled-backbone model, wherein a single appropriate ground motion model (GMM) is identified for use in PSHA, to which adjustments or scaling factors are then applied to account for epistemic uncertainty in the underlying seismological properties of the region of interest. While the theory and development of the regionalised scaled-backbone GMM concept have been discussed in earlier publications, implementation in the final ESHM20 required further refinements to the shallow-seismicity GMM in three regions, which were undertaken considering new data and insights gained from the feedback provided by experts in several regions of Europe: France, Portugal and Iceland. Exploration of the geophysical characteristics of these regions and analysis of additional ground motion records prompted recalibrations of the GMM logic tree and/or modifications to the proposed regionalisation. These modifications illustrate how the ESHM20 GMM logic tree can still be refined and adapted to different regions based on new ground motion data and/or expert judgement, without diverging from the proposed regionalised scaled-backbone GMM framework. In addition to the regions of crustal seismicity, the scaled-backbone approach needed to be adapted to earthquakes occurring in Europe's subduction zones and to the Vrancea deep seismogenic source region. Using a novel fuzzy methodology to classify earthquakes according to different seismic regimes within the subduction system, we compare ground motion records from non-crustal earthquakes to existing subduction GMMs and identify a suitable-backbone GMM for application to subduction and deep seismic sources in Europe. The observed ground motion records from moderate- and small-magnitude earthquakes allow us to calibrate the anelastic attenuation of the backbone GMM specifically for the eastern Mediterranean region. Epistemic uncertainty is then calibrated based on the global variability in source and attenuation characteristics of subduction GMMs. With the ESHM20 now completed, we reflect on the lessons learned from implementing this new approach in regional-scale PSHA and highlight where we hope to see new developments and improvements to the characterisation of ground motion in future generations of the European Seismic Hazard Model.

Ground-motion models for inelastic response spectra using NGA-West2 database

This article presents ground-motion models (GMMs) for inelastic response spectra using the NGA-West2 database. The inelastic response spectra are defined in terms of constant ductility. The GMMs are used to observe the effect of event scenarios, site conditions, and oscillator properties on maximum displacement. Using a large database enabled us to quantify the magnitude scaling of inelastic response as compared to that for the traditional elastic GMMs. We observe that using the magnitude scaling of elastic GMM for hazard analysis of inelastic structural response can be unconservative for large magnitudes and large ductility ratios. To compare our models with models in the literature, we use the developed GMMs to compute the Reduction Factor due to inelasticity. Comparison of this model with models in the literature shows the importance of considering the effect of magnitude on the Reduction Factor. We also observe that the ratio of RotD100/RotD50 is larger for inelastic single-degree-of-freedom system than that for elastic system, and the ratio increases with increasing ductility.

Rapid Finite-Fault Models for the 2023 Mw 7.8 Kahramanmaraş, Türkiye, Earthquake Sequence

Abstract In the immediate aftermath of devastating earthquakes such as in the 6 February 2023 Kahramanmaraş sequence in southcentral Türkiye, key stakeholders and the public demand timely and accurate earthquake information. Especially for large events, finite-fault models provide important insights into the rupture process and enable interpretation of the observed ground shaking, which can improve situational awareness and facilitate rapid assessment of future hazards. Using strong-motion waveforms recorded during the Kahramanmaraş sequence, we simulate a real-time playback and calculate how a finite-source model computed with the Finite-fault rupture Detector (FinDer) algorithm would evolve for the Mw 7.8 Pazarcık, Mw 7.6 Elbistan, and Mw 6.4 Yayladağı earthquakes. Using template matching FinDer compares observed and predicted ground-motion acceleration amplitudes to determine the orientation and spatial extent of fault rupture. We test both generic crustal and fault-specific templates from ground-motion models and rupture geometries of the east Anatolian and Çardak–Sürgü faults. In the second step, we estimate the seismic slip along the source models from the backprojection of the seismic displacement amplitudes. The algorithms achieve excellent performance for all three earthquakes, and the final source models and slip profiles available within tens of seconds of the rupture nucleation match well with models computed days to weeks after the events occurred. The temporal evolution of the source models for the Pazarcık and Elbistan earthquakes suggests that FinDer can provide insight into the rupture kinematics of large earthquakes. Cascading instrument failures as well as power and data telemetry interruptions during the Pazarcık earthquake led to an early termination of signals at a significant number of near-source stations. We show that FinDer is robust enough to cope with this type of degradation in network performance that can occur in large earthquakes, in general.

A machine learning estimator trained on synthetic data for real-time earthquake ground-shaking predictions in Southern California

After large-magnitude earthquakes, a crucial task for impact assessment is to rapidly and accurately estimate the ground shaking in the affected region. To satisfy real-time constraints, intensity measures are traditionally evaluated with empirical Ground Motion Models that can drastically limit the accuracy of the estimated values. As an alternative, here we present Machine Learning strategies trained on physics-based simulations that require similar evaluation times. We trained and validated the proposed Machine Learning-based Estimator for ground shaking maps with one of the largest existing datasets (<100M simulated seismograms) from CyberShake developed by the Southern California Earthquake Center covering the Los Angeles basin. For a well-tailored synthetic database, our predictions outperform empirical Ground Motion Models provided that the events considered are compatible with the training data. Using the proposed strategy we show significant error reductions not only for synthetic, but also for five real historical earthquakes, relative to empirical Ground Motion Models.

Directionality characteristics of horizontal response spectra from the 2022 Mw 6.9 Chihshang, Taiwan earthquake

Horizontal earthquake ground motion intensity, and specifically response spectral ordinates, vary with orientation. This phenomenon is usually referred to as ground motion directionality and can be separated into two aspects: the orientation where the maximum spectral response occurs and the variation of response spectral ordinates as the orientation moves away from the orientation of maximum spectral response. This work studies both aspects using the recent 2022 Mw 6.9 Chihshang, Taiwan earthquake, which was recorded by a dense network of strong motion stations with various geological and topographical settings. The mean variation of response spectral ordinates with orientation is found to be slightly more significant than that of previous shallow crustal earthquakes in active tectonic regimes. Moreover, the orientation of maximum spectral response is found to be close to the transverse orientation, which is perpendicular to the orientation at a given site that points to the earthquake epicenter, confirming prior observations made for strike-slip earthquakes. These results suggest that the location of a site relative to the seismic source could be used to modify the outputs of ground motion models to estimate spectral responses at specific horizontal orientations.

Estimation of Geometrical Spreading Factor and Coda-Wave Attenuation Characteristics for the Saurashtra Horst in Western Deccan Volcanic Province, Gujarat, India

ABSTRACT In the present study, we have estimated the coda Q (QC) and the geometrical spreading factor (gamma) for the Saurashtra region that is located in the western part of the Deccan Volcanic Province, Gujarat, India. The waveform data from the seismic network operated by the Institute of Seismological Research, Gandhinagar, are used for the analysis. We have used records of a total of 221 local earthquakes with epicentral distances less than 70 km from the seismic stations. The moment magnitude range of the earthquakes is 2.5–5.3, and the depth range is 1–22 km. The QC has been estimated at five central frequencies (1.5, 3, 6, 12, and 20 Hz) using 20, 30, and 40 s window lengths for vertical, north–south, and east–west components, respectively. The obtained QC values show that the attenuation in the region has a strong dependence on frequency and points toward a heterogeneous medium. The geometrical spreading factors are also estimated for two cases, that is, with and without Q at the same central frequencies and window lengths. The average geometrical spreading factors with Q are 1.37, 1.48, 1.40, 1.08, and 1.04, at central frequencies 1.5, 3, 6, 12, and 20 Hz, respectively, whereas it is 1.59, 1.53, 1.47, 1.19, and 1.14, respectively, in the absence of Q. The estimated values also indicate that the medium is heterogeneous in the Saurashtra region. The estimated region-specific coda Q and the geometrical spreading factor may help in the modeling of strong ground motions and assessment of the seismic hazard for the Saurashtra region.

A prolegomenon to Design Input Energy Spectra for the Himalayan region

Energy-based design is becoming progressively significant as a new age of earthquake-resistant design unfolds. Thitherto, very little is known about energy-based design in India despite the fact that the Himalayan region is one of the most seismically active regions in the world. Therefore, the current work provides a preliminary analysis of the elastic input energy and equivalent velocity (Vea) spectra for the active regions in the Himalayas. In this regard, the ground motions of Himalayan databases are blended with NGA WEST2 to estimate the Vea spectra, an inertia-independent representation of the input energy spectra. Incipiently, an Artificial Neural Network-based ground motion model (GMM) is developed for Vea spectra based on the predictor variables as magnitude, distance, hypocentral depth and site conditions. Consequently, a Probabilistic Seismic Hazard analysis is performed to obtain the uniform hazard energy spectra for a 2475-year return period. In addition, a Design Input Energy Spectra is proposed for the active regions in the Indian Himalayas. In addition, the relationship between the Vea and pseudo-spectral velocity (PSV) is analyzed to develop a Vea-informed GMM for PSV. Furthermore, the PSA estimated from the Vea-PSV model is compared with the design spectra recommended by the Indian Standard Code of Practice (IS1893-1:2016).

The 2023 US National Seismic Hazard Model: Subduction ground-motion models

The US Geological Survey National Seismic Hazard Models (NSHMs) are used to calculate earthquake ground-shaking intensities for design and rehabilitation of structures in the United States. The most recent 2014 and 2018 versions of the NSHM for the conterminous United States included major updates to ground-motion models (GMMs) for active and stable crustal tectonic settings; however, the subduction zone GMMs were largely unchanged. With the recent development of the next generation attenuation-subduction (NGA-Sub) GMMs, and recent progress in the utilization of “M9” Cascadia earthquake simulations, we now have access to improved models of ground shaking in the US subduction zones and the Seattle basin. The new NGA-Sub GMMs support multi-period response spectra calculations. They provide global models and regional terms specific to Cascadia and terms that account for deep-basin effects. This article focuses on the updates to subduction GMMs for implementation in the 2023 NSHM and compares them to the GMMs of previous NSHMs. Individual subduction GMMs, their weighted averages, and their impact on the estimated mean hazard relative to the 2018 NSHM are discussed. The updated logic trees include three of the new NGA-Sub GMMs and retain two older models to represent epistemic uncertainty in both the median and standard deviation of ground-shaking intensities at all periods of interest. Epistemic uncertainty is further represented by a three-point logic tree for the NGA-Sub median models. Finally, in the Seattle region, basin amplification factors are adjusted at long periods based on the state-of-the-art M9 Cascadia earthquake simulations. The new models increase the estimated mean hazard values at short periods and short source-to-site distances for interface earthquakes, but decrease them otherwise, relative to the 2018 NSHM. On softer soils, the new models cause decreases to the estimated mean hazard for long periods in the Puget Lowlands basin but increases within the deep Seattle portion of this basin for short periods relative to the 2018 NSHM.

Empirical models for Fourier amplitude spectrum of ground-motion calibrated on data from the Iranian plateau

Ground-motion models (GMMs) are frequently used in engineering seismology to estimate ground motion intensities. The majority of GMMs predict the response spectral ordinates (such as spectral acceleration) of a single-degree-of-freedom oscillator because of their common application in engineering design practices. Response spectra show how an idealized structure reacts to applied ground motion; however, they do not necessarily represent the physics of ground motion. The functional forms of the response spectra GMMs are built around ideas taken from the Fourier spectral concept. Assuming the validity of Fourier spectral concepts in the response spectral domain could cause physically inexplainable effects. In this study, using a mixed-effects regression technique, we introduce four models capable of predicting the Fourier amplitude spectrum that investigates the impact of incorporating random-effect event and station terms and variations in using a mixed-effects regression technique in one or two steps using truncated dataset or all data (nontruncated dataset). All data consists of 2581 three-component strong ground motion data resulting from 424 events with magnitude ranging from 4.0 up to 7.4, from 1976 to 2020, and 706 stations. The truncated dataset’s records, events, and stations are reduced to 2071, 408, and 636, respectively. As part of this study, we develop GMMs to predict the Fourier amplitude spectrum for the Iranian plateau within the frequency range of 0.3–30 Hz. We adopted simple, functional forms for four models, and we included a limited number of predictors, namely Mw (moment magnitude), Rjb (Joyner–Boore distance), and VS30 (time-averaged shear-wave velocity in the top 30 m). Due to statistical analyses, the style-of-faulting term was excluded from the final functional forms. The robustness of the derived models is indicated by unbiased residual variation with predictor variables.

Machine learning ground motion models for a critical site with long-term earthquake records

Ground motion models derived using data from widespread stations and earthquakes induce significant aleatory variability when applied to predict ground motions for a single site. This issue can be even worse when the sites of interest show complex site response characteristics and do not have enough similar sites in the ground motion model development dataset. This work provides a machine-learning-based ground motion modeling framework for sites with numerous earthquake recordings. The Onahama Port Array, a liquefiable site in Japan that has recorded more than 1200 earthquakes in the last 25 years, was selected as a case study. The gradient boosting method was employed to predict response spectra with periods from 0.01 to 10 s for earthquakes with Japan Meteorological Agency (JMA) magnitude ranging from 2.4 to 9.0. Different predictor combinations were tested, and two gradient boosting models were recommended: the basic model that requires earthquake magnitude and epicentral distance as inputs, and the optimal model that requires three additional inputs, focal depth, azimuth, and rainfall. The basic and optimal gradient boosting models have the root mean square error of 0.004 and 0.001 g, average r2 of 5-fold cross validation of 0.915 and 0.983, respectively. The results shed light on seismic hazard assessment for critical sites with long-term earthquake recordings.

A geological, petrophysical and ground motion model of the Pavoncelli bis hydraulic tunnel (Southern Apennines, Italy): Implications for reservoir analyses

The strong 1980 M6.81 Irpinia earthquake in Southern Italy critically damaged the Pavoncelli hydraulic tunnel. Based on geological, petrophysical and geomechanical investigations and seismological data, a multi-scale geological model was inferred. The detrital mode and key petrophysical properties (porosity, permeability, and nanopore volumes) data have been used for defining the water-reservoir potential of the arenitic successions, with a focus on defining the hydrogeological model. The model was then used to perform a novel back-analysis of the spatial distribution of the peak ground acceleration (PGA) caused by the 1980 earthquake accounting for site-effects in a robust manner. This analysis, which accounts for site-effects in a robust manner, shows high PGA values along the length of the Pavoncelli tunnel, which is located between the causative fault of the 1980 earthquake in the vicinity of the epicenter. The main outcomes of the study are that: (1) the multi-scale and interdisciplinary approach developed in this study can be used in future studies in this and other areas, (2) the spatial 2D-3D geological model must be accounted for when analyzing distributed systems, (3) the hydrogeological model coupled with porosity data allows for reservoir analyses and (4) the complex tectonic environment in the area hosting the Pavoncelli tunnel plays a strong role in the analysis of the spatial distribution of PGA, creating the need for site-specific seismic microzonation studies and the evaluation of near-fault effects.

Assessment of Site Effects and Numerical Modeling of Seismic Ground Motion to Support Seismic Microzonation of Dushanbe City, Tajikistan

In the territory of Dushanbe city, the capital of Tajikistan, detailed geological and geophysical data were collected during geophysical surveys in 2019–2020. The data comprise 5 microtremor array measurements, 9 seismic refraction tomography profiles, seismological data from 5 temporary seismic stations for standard spectral ratio calculations, 60 borehole datasets, and 175 ambient noise measurements. The complete dataset for Dushanbe was used to build a consistent 3D geologic model of the city with a size of 12 × 12 km2. The results of the seismological and geophysical surveys were compared and calibrated with borehole data to define the boundaries of each layer in the study area. The Leapfrog Works software was utilized to create a 3D geomodel. From the 3D geomodel, we extracted six 12 km long 2D geological cross-sections. These 2D geological cross-sections were used for 2D dynamic numerical modeling with the Universal Distinct Element Code software to calculate the local seismic response. Finally, the dynamic numerical modeling results were compared with the amplification functions obtained from the seismological and ambient noise data analysis. The 2D dynamic numerical modeling results allowed a better assessment of the site effects in the study area to support seismic microzonation and the determination of local peak ground acceleration changes in combination with regional seismic hazard maps. In addition, our results confirm the strong seismic amplification effects noted in some previous studies, which are attributed to the influence of local topographic and subsurface characteristics on seismic ground motions.

Exploring the Use of Orientation-Independent Inelastic Spectral Displacements in the Seismic Assessment of Bridges

ABSTRACT Seismic intensity measures (IMs) provide a link between the seismic hazard and the dynamic response of structures subjected to earthquake shaking. The spectral acceleration at the first and usually dominant vibration mode, Sa(T 1), is a popular choice for building structures. Meanwhile, the IM selection for bridges is non-trivial since they do not typically possess a single dominant mode. Even for ordinary bridges with a dominant mode, the behavior can change significantly in each principal direction through the activation, or yielding, of its different components. This study examines the performance of a novel IM that incorporates ground motion directionality and structure non-linearity in this context: the nn th percentile of all rotation angles of the inelastic spectral displacement, Sd i,RotDnn. This evaluation is carried out within the context of an ordinary bridge structure and is compared with other conventional IMs used in regional risk assessment of bridges. The case study bridge utilized is a highway overcrossing located in California with two spans and a continuous prestressed reinforced concrete box girder deck section. A large ground motion set was selected from the NGA-West2 database, and incremental dynamic analysis was carried out on the structure to assess the IM performance to characterize collapse. The results indicate that Sd i,RotDnn performs very well compared to other IMs for the bridge structure and could be a prudent choice to characterize inelastic response of bridges with several possible mechanisms in different principal directions. Also, using the RotD50 definition, typically used in ground motion models, showed a 17.3% increase in efficiency compared to RotD100 definition typically used in engineering practice.

Partially non-ergodic ground motion model for the Bucaramanga seismic nest in Northern South America (NoSAm Nest GMM)

The Bucaramanga seismic nest has the highest concentration in volume of intermediate depth seismic events with light and moderate magnitudes (Mb > 4.8) worldwide. Despite the nest depth (Zhypo ≥ 100 km), these events play a crucial role in the seismic hazard assessment for several cities in northern South America (NoSAm), particularly those situated close to their epicenters. Moreover, these earthquakes are widely felt across the country, even hundreds of kilometers away from the nest. This paper assesses the performance of commonly used subduction intraslab Ground Motion Models (GMMs) in predicting the updated nest database gathered and processed by the Colombian Geological Survey. This analysis confirms the necessity of a partially non-ergodic ground motion model for estimating the spectral accelerations produced by Bucaramanga nest earthquakes in northern South America. The GMM presented herein is the first model developed specifically for seismic nests in the region and is conceived as a regionalization of the recent NGA-Sub Abrahamson and Gülerce (Earthq Spectra 38(4):2638−2681, 2022) GMM. Additionally, we provide estimates of the variance components at both the earthquake and station levels and compare the developed GMM with the spectral accelerations recorded for typical earthquakes, for the highest magnitude earthquake in the database, and for a recent Bucaramanga nest earthquake having a moment magnitude over 5.

Seismological Ground-Motion Models for Generic Hard-Rock Site Condition in Western Canada

ABSTRACT We develop ground-motion models (GMMs) to characterize the Fourier amplitude spectrum (FAS) for earthquakes in British Columbia (B.C.), Canada. GMMs developed for FAS are useful for understanding the underlying seismological parameters and can be transformed into GMMs for response spectral values for other applications (e.g., probabilistic seismic hazard analysis [PSHA]). The GMMs are calibrated using a compiled FAS database, referenced to a B.C. generic hard-rock site condition (shear-wave velocity ∼2285 m/s). The GMMs are developed separately for crustal, in-slab, transition, offshore, and Haida Gwaii events. The GMMs are calibrated using data mostly from M 2.5 to 5.5 earthquakes recorded at rupture distances of ∼50 to 500 km. Outside this range, the models are constrained by a seismological model supplemented with sparse observational data. The amplitude decay rate for crustal earthquakes in B.C. is very similar to that given by the empirical model of Bayless and Abrahamson (2018; hereafter, BA18), developed from California data on soil sites. However, we observe magnitude- and frequency-dependent differences between the models for ground-motion amplitude levels. We attribute these to (1) the different reference-site conditions of the models, with the B.C. GMMs being referenced to hard rock, and (2) steeper magnitude scaling at small-to-moderate magnitudes for events in B.C. in comparison to the BA18 model.