Feasibility of Applying Kriging for Earthquake Ground Motion Intensity Measures in South Korea

Strong ground motion intensity measures, for example Peak Ground Acceleration (PGA) or Peak Ground Velocity (PGV), are important dynamic features, or predictive variables, in most regional earthquake induced landslide susceptibility models. Despite global reliance on these ground motion intensity measures, little work has been done to evaluate how dynamic feature selection, and underlying ground motion models, influence the predictive performance of landslide susceptibility models. Here, we conduct a feature sensitivity analysis, training a suite of 131 comparative logistic regression models on the distribution of landslides from the 2016 Mw 7.8 Kaikōura earthquake on the South Island of New Zealand. This analysis uses a combination of common susceptibility features (e.g., slope, curvature), distance to a surface fault rupture (both a susceptibility and dynamic feature), and 9 ground motion intensity measures (PGA, PGV, Arias Intensity, PSA - Peak Spectral Acceleration at 0.3, 1.0, 3.0, and 10.0 seconds, MMI - Modified Mercalli Intensity, and Duration of Shaking) derived from 4 published ground motion models for the Kaikōura earthquake. Ground motion is highly correlated with distance to a surface fault rupture (a Pearson R2 as high as 0.86). Models trained using both distance to surface fault rupture and a ground motion intensity measure produce high model performance but are overfit to the Kaikōura landslide distribution with negative model coefficients for most ground motion intensity measures. Excluding distance to a surface fault rupture still produces high model performance (less than a 0.04 drop in Model AUC) when including the most predictive ground motion intensity estimates (typically MMI, PSA at a period of 0.3 seconds, PGA, or PGV from the USGS ShakeMap) and results in more explainable, and likely more applicable, model coefficients. Although MMI and PSA at a period of 0.3 seconds (3.3 Hz) appear to be good predictors of the landslide distribution from the Kaikōura earthquake, MMI can be influenced by the availability of felt reports and the frequency of shaking can vary in different earthquakes. PGA and PGV provide acceptable model performance for the Kaikōura landslide distribution and are likely more applicable to other events. Highly variable performance is observed when applying the same ground motion intensity measures from different published ground motion models. The choice of ground motion model may, therefore, introduce a high degree of uncertainty into the landslide susceptibility analysis that remains relatively underappreciated in most studies. Additional recorded strong motion data will likely be required to further improve ground motion models, and thereby landslide susceptibility models, for future events.

Ground motion models (GMMs) play a pivotal role in both deterministic and probabilistic seismic hazard assessments, which are essential for identifying the seismic safety of nuclear power plants. In regions with abundant seismic data, especially strong earthquake records, GMMs could be empirically derived. However, in areas like South Korea with scarce strong earthquake records, development of empirical GMMs is impractical, leading to the utilization of alternative methods such as stochastic simulations. There have been a few GMMs developed in South Korea, all of which relied on stochastically simulated motions. In this study, GMMs are developed for rock sites in South Korea using the hybrid empirical method (HEM) suggested by Campbell (2003). Western United States (WUS) is selected as a host region and five Next Generation Attenuation (NGA)-West2 GMMs are used as GMMs of the host region. The seismological parameters employed in the simulation, including effective point source distance, source and path duration, and path attenuation, duly encompass the findings of recent studies. The high-frequency spectral attenuation parameters, kappa, utilized as site attenuation parameters in ground motion simulations for the target region, are estimated in this study. It is primarily estimated using the classical method proposed by Anderson and Hough (1984). Additionally, the estimation process considers the standardized procedure and the recommended lower bound magnitude decisions put forth by Ktenidou et al. (2013) and Van Houtte et al. (2014), respectively. Since the shear wave velocity for bedrock is considered to be 760 m/s in South Korea, the site amplification functions have been applied with reference to this velocity for both the host and target regions. The adjustment factors obtained from simulated ground motions in both the host and target regions are applied to adjust NGA-West 2 Ground Motion Models (GMMs). Derived GMMs are for magnitudes from 5.0 to 7.5 and rupture distances from 10 to 500 km. Median GMMs are provided with aleatory standard deviations. Predictive GMMs are compared with observed ground motions from the available earthquake records for moment magnitudes 5.0 and 5.5. The notable advantages of the GMMs developed in this study are as follows: Distinct from previous researches utilizing stochastic methods, the implementation of HEM served to complement the limitations inherent in stochastic approaches such as lack of near-source ground motion characteristics. Defining the sites where GMMs are employed at Vs30 = 760m/s enables the derivation of seismic motions applicable to rock layers having Vs30 of 760m/s. Since aleatory standard deviations are quantitatively defined, they can serve as the sigma parameter within GMMs in Probabilistic Seismic Hazard Analysis (PSHA).

Iceland is seismically the most active region in northern Europe. Large single earthquakes (~ $$ M_{\text{w}} $$ 7) and seismic sequences of moderate-to-strong earthquakes (~ $$ M_{\text{w}} $$ 6–6.5) have repeatedly occurred during past centuries in the populated South Iceland Seismic Zone (SISZ). The seismic hazard in Iceland has mainly been evaluated using ground motion models (GMMs) developed from strong-motion observations in other countries and only to a very limited extent from Icelandic data, despite a particularly rapid attenuation of ground motions with distance in Iceland. In this study, we evaluate the performance of these GMMs against the Icelandic strong-motion dataset, consisting of peak ground accelerations of moderate-to-strong ( $$ M_{\text{w}} $$ 5–6.5) and local (0–80 km) earthquakes in the SISZ. We find that these GMMs exhibit both a strong bias against the dataset and a relatively large variability, which calls their applicability and earlier hazard analyses into question. To address this issue, we recalibrate each of the GMMs to the dataset using Bayesian regression and Markov Chain Monte Carlo simulations. This approach allows useful prior information of the GMM parameters to be combined with the likelihood of the observed data and provides posterior probability density functions of model residuals and regression parameters. The recalibrated GMMs are unbiased with respect to the data and have a low total standard deviation of around 0.17 (base-10 logarithmic units). The model-to-model variability in the median predictions vary primarily with distance, reaching 0.05 the lowest for $$ M_{\text{w}} $$ 6.3–6.5 at intermediate distances. While the lack of near-fault and far-field data, particularly at large magnitudes, and the different functional forms of the GMMs calibrated to the same dataset may affect the results, the recalibrated GMMs should represent well the ground motions of a typical sequence of moderate-to-strong SISZ earthquakes. We present the recalibrated GMMs of this study as promising candidates for future use in ground motion prediction in Iceland e.g., in the context of either a logic tree or the backbone approach in probabilistic seismic hazard assessment.

A ground motion model (GMM) for interface subduction zone earthquakes of Northeast India (NEI) and its adjacent countries is developed for the first time. Countries adjacent to NEI are Bangladesh, Bhutan, China, Myanmar and Nepal. High-magnitude earthquakes occur frequently in these regions due to buildup of high-stress parameters in the subduction zone of the Indian tectonic plate. Strong motion data are too few and sparse to develop a robust GMM for this region. We used both finite-fault simulations and a stochastic point-source model in developing our GMM. In our GMM, we used 50,000 ground motion samples which were stochastically simulated for different moment magnitudes (Mw) of 5.0–9.0 and hypocentral distances of 30–300 km using a point-source seismological stochastic model and finite fault model. In this study, we calculated stress drop (∆σ), quality factor Q(f) and all other region-specific seismic input parameters from the past strong motion records of interface subduction zone earthquakes of NEI and its adjacent countries. We used these seismic input parameters in ground motion simulation. Sensitivity analyses of the input parameters were also performed to check the bias of the present model. Our GMM was validated by comparing it with the existing NEI interface strong motion records. We compared our GMM with other GMMs developed for interface subduction zone earthquakes for different regions in the world. We also compared our GMM with point-source and finite-fault simulation models. Ground motion parameters estimated using the point-source model are comparatively higher than the finite-fault simulation model. Horizontal components of peak ground acceleration (PGA) and spectral acceleration (Sa) can be estimated for NEI using our GMM.

ABSTRACTThe selection and weighting of ground-motion models (GMMs) introduces a significant source of uncertainty in U.S. Geological Survey (USGS) National Seismic Hazard Modeling Project (NSHMP) forecasts. In this study, we evaluate 18 candidate GMMs using instrumental ground-motion observations of horizontal peak ground acceleration (PGA) and 5%-damped pseudospectral acceleration (0.02–10 s) for tectonic earthquakes and volcanic eruptions, to inform logic-tree weights for the update of the USGS seismic hazard model for Hawaii. GMMs are evaluated using two methods. The first is a total residual visualization approach that compares the probability density function (PDF), mean and standard deviations σ, of the observed and predicted ground motion. The second GMM evaluation method we use is the common total residual probabilistic scoring method (log likelihood [LLH]). The LLH method provides a single score that can be used to weight GMMs in the Hawaii seismic hazard model logic trees. The total residual PDF approach provides additional information by preserving GMM over- and underprediction across a broad spectrum of periods that is not available from a single value LLH score. We apply these GMM evaluation methods to two different data sets: (1) a database of instrumental ground motions from historic earthquakes in Hawaii from 1973 to 2007 (Mw 4–7.3) and (2) available ground motions from recent earthquakes (Mw 4–6.9) associated with 2018 Kilauea eruptions. The 2018 Kilauea sequence contains both volcanic eruptions and tectonic earthquakes allowing for statistically significant GMM comparisons of the two event classes. The Kilauea ground observations provide an independent data set allowing us to evaluate the predictive power of GMMs implemented in the new USGS nshmp-haz software system. We evaluate GMM performance as a function of earthquake depth and we demonstrate that short-period volcanic eruption ground motions are not well predicted by any candidate GMMs. Nine of the initial 18 candidate GMMs fit the observed ground motions and meet established criteria for inclusion in the update of the Hawaii seismic hazard model. A weighted mean of four top performing GMMs in this study (NGAsubslab, NGAsubinter, ASK14, A10) is 50% lower for PGA than for GMMS used in the previous USGS seismic hazard model for Hawaii.

A synthesis of ground motion and corresponding design response spectrum is presented here. A semi-empirical Green’s function approach based on envelope summation technique is used to model ground motion for evaluation of expected peak ground acceleration (PGA) from future large earthquake in Garhwal Himalaya, northeast (NE) and south India. The synthetic site-specific response spectra for the single degree of freedom (SDOF) structures corresponding to the synthetic ground motion records are obtained for both rock and soil conditions for engineering design of reservoir and nuclear power plant in different parts of India. The predicted PGA for NE India is found to be ~ 3.5 m/s2 at 60 km, decaying to ~ 0.1 m/s2 at 500 km distance, while it is ~ 3.0 m/s2 at 15 km, decaying to ~ 1.4 m/s2 at 80 km distance for Garhwal Himalaya. These PGA estimates are in good agreement with those obtained from the Global Seismic Hazard Assessment Program (GSHAP) exercise. The response spectrum attains its maximum relatively earlier for rock site than for soil site. The response spectra decay gradually at periods later than 0.5 s and approach asymptotically constant values at longer periods. The synthetic ground motions are used to perform seismic response analysis of engineering structures. For such analysis, a PGA of 0.027 g is estimated for the Tail pond reservoir in south India for a target earthquake of M 6.5 at 50 km distance, while a PGA of 0.015 g is found for the Gorakhpur nuclear power plant in north India for similar magnitude and distance. The difference in the PGA estimates between the two is mostly attributed to regional difference in attenuation and site response. The findings of this study are expected to facilitate the seismic design of engineered structures, safe from seismic hazard, in regions where strong motion data are sparse.

Offshore ground motion characteristics on the horizontal PGA, spectral acceleration, frequency content and significant duration from the 2021 Mw 7.1 and 2022 Mw 7.4 offshore earthquakes near the Japan Trench area

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 ( V S30 ) = 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 V S30 = 760 m/s condition. We provide adjustment factors and their epistemic uncertainties and discuss implications for applications.

Prediction of peak ground acceleration by genetic expression programming and regression: A comparison using likelihood-based measure

SUMMARY Turkey is characterized by a high level of seismic activity attributed to its complex tectonic structure. The country has a dense network to record earthquake ground motions; however, to study previous earthquakes and to account for potential future ones, ground motion simulations are required. Ground motion simulation techniques offer an alternative means of generating region-specific time-series data for locations with limited seismic networks or regions with seismic data gaps, facilitating the study of potential catastrophic earthquakes. In this research, a local ground motion model (GMM) for Turkey is developed using region-specific simulated records, thus constructing a homogeneous data set. The simulations employ the stochastic finite-fault approach and utilize validated input-model parameters in distinct regions, namely Afyon, Erzincan, Duzce, Istanbul and Van. To overcome the limitations of linear regression-based models, artificial neural network is used to establish the form of equations and coefficients. The predictive input parameters encompass fault mechanism (FM), focal depth (FD), moment magnitude (Mw), Joyner and Boore distance (RJB) and average shear wave velocity in the top 30 m (Vs30). The data set comprises 7359 records with Mw ranging between 5.0 and 7.5 and RJB ranging from 0 to 272 km. The results are presented in terms of spectral ordinates within the period range of 0.03–2.0 s, as well as peak ground acceleration and peak ground velocity. The quantification of the GMM uncertainty is achieved through the analysis of residuals, enabling insights into inter- and intra-event uncertainties. The simulation results and the effectiveness of the model are verified by comparing the predicted values of ground motion parameters with the observed values recorded during previous events in the region. The results demonstrate the efficacy of the proposed model in simulating physical phenomena.

The development of a design motion requires some input on ground motion characteristics. For a country such as Malaysia, where historical data is lacking and seismic activities are low, the information on characteristics of ground motion may be obtained by utilizing established attenuation relationships. This study aims to investigate the characteristics of distant ground motions in Peninsular Malaysia, which originated from the active tectonic plate of Sumatra, by comparing recorded peak ground acceleration (PGA) and peak ground velocity (PGV) values with those estimated using four attenuation models. Selected attenuation models are the Atkinson and Boore (1995), the Toro et al. (1997), the Dahle et al. (1990), and the Si and Midorikawa (1999) models. The analysis for comparison was made by employing a maximum of 46 horizontal component accelerograms, recorded at 14 seismic stations. These are data derived from 15 interplate earthquakes between May 2004 and July 2007. Recorded PGA and PGV values were plotted on respective attenuation curves to examine ground motion attenuation characteristics. Results indicated that attenuation models, established for stable tectonic regions, provide good estimation of PGA and PGV for Sumatran earthquakes for magnitude range of 5.9 to 9.0. The study shows that the Dahle et al. (1990) model best represents the characteristics of ground motions in terms of PGA, while the Atkinson and Boore (1995) model gives appropriately estimate ground motions in terms of PGV

Improving accuracy and reducing uncertainty in ground motion models (GMMs) are crucial for the safe design of infrastructure. Traditional GMMs often oversimplify source complexity, such as stress drop, due to high variability in estimation. This study aims to address this issue by extracting robust spatial variations in stress drop estimates and ground motion residuals. We introduce a non‐ergodic modeling framework using Bayesian Gaussian Process regression to analyze data from over 5,000 earthquakes (M2‐4.5) in the San Francisco Bay area. Our findings reveal consistent spatial patterns in non‐ergodic stress drop and peak ground acceleration (PGA), providing a reliable approach to understanding the spatial distribution of stress drop and its link to regional tectonics. Furthermore, integrating source models derived from the non‐ergodic stress drop into GMMs can effectively account for source effect in ground motions and reduce aleatory uncertainty. This study establishes a framework for utilizing stress drop data sets to enhance seismic hazard assessment.

Two powerful earthquakes (magnitudes 7.8 and 7.6) struck south-central Türkiye on February 6, 2023, causing significant damage across an extensive area of at least ten provinces in Türkiye as well as in multiple cities in northwestern Syria, making them one of the deadliest earthquakes in Türkiye for multiple centuries. The first mainshock started close to the well-known East Anatolian Fault (EAF) and then rupturing more than 300 km of that fault, whereas the second large earthquake occurred nine hours later around 90 km north of the first mainshock, on an east-west trending fault. In this study, we analysed recorded strong ground motions from the two events to better understand the factors contributing to the devastation caused by the earthquakes. For this, we collected 250 and 200 strong ground motion records for the first and the second event, respectively, from the Disaster and Emergency Management Authority (AFAD) in Türkiye. Maximum peak ground accelerations (PGA) of 2g were observed at a distance of 31 km northeast of the first mainshock epicenter and 0.6g for the second event 65km west to its epicenter. In addition, we find particularly high amplitude ground motions in the Hatay province for the first event, which is consistent with the extent of damage reported in that region. High shaking levels in Antakya and other parts of Hatay can be explained by a combination of strong directivity and local site effects. The results of our analysis imply that the PGA values derived from two local ground motion models (GMMs), adopted for the 2018 Turkish hazard map, are underestimated in comparison to observed strong motion recordings. In addition, we also compared observed peak and spectral ground motion characteristics with estimated seismic hazard values (10% probability to exceed in 50 years) in the East Anatolian Fault region (extracted from the 2018 Turkish seismic hazard map). Furthermore, we compare the recorded response spectra with the Turkish design code for several locations around the main faults.  The results show that the observations greatly exceed the hazard values and code guidelines in the Hatay province.

ABSTRACTIn this study, we develop a new nonergodic ground motion model (GMM) for Chile, which better captures the trade-off between the aleatory variability and epistemic uncertainty on ground motion estimates compared with existing GMMs. The GMM is developed for peak ground acceleration and pseudospectral acceleration at a period of 1 s. Most existing GMMs for subduction earthquake zones were developed based on an ergodic assumption, and this is not the exception for the subduction zone in Chile. Under the ergodic assumption, the ground motion variability at a given single site–source combination is considered the same as the variability observed in a global database. However, recent efforts have highlighted significant location-specific systematic and repeatable effects for ground motions recorded within a particular region. These systematic effects promote the relaxation of the ergodic assumption and the transition to the development of nonergodic GMMs. The nonergodic GMM developed in this study uses an ergodic GMM as a backbone, the systematic source and site effects are modeled using Gaussian processes, and the path effects are modeled using the cell-specific attenuation approach enhanced with a computer graphics-based algorithm. The coefficients of the nonergodic GMM are estimated using Bayesian inference via Markov chain Monte Carlo (MCMC) methods, in which we use an integrated nested Laplace approximation approach to address the computational burden involved in MCMC. The developed nonergodic GMM reveals spatially varying and correlated location-specific source, path, and site effects in Chile, which cannot be captured by existing Chilean ergodic GMMs. Moreover, the developed nonergodic GMM shows a reduced aleatory variability compared to existing ergodic GMMs that are commonly used in Chile. In addition, the developed nonergodic GMM shows small epistemic uncertainty for regions with large ground motion data and large epistemic uncertainty for regions with few ground motion data. Finally, we provide guidelines on how to use the developed nonergodic GMM in the context of probabilistic seismic hazard analysis, which is important for performance-based earthquake engineering assessments in Chile.

In the context of assessing seismic hazard, accurately predicting ground motion stands out as a crucial task. Achieving precision in ground motion modeling proves valuable in revealing the actual pattern, even when faced with insufficient data on soil structure, provided there is precise information about the seismic source. This study introduces a methodology for calculating local- and near-field ground motion, expressed in peak ground acceleration (PGA) and intensity values. The deterministic approach is employed, incorporating source characteristics and one-dimensional (1D) site effects. For the chosen test area, the Ismayilli-Shamakhi region on the southeastern slope of the Greater Caucasus in Azerbaijan, two seismic scenarios are investigated: the 1902 Shamakhi earthquake (magnitude M = 6.8) and the November 25, 2000 Baku-Caspian earthquake (two shocks with moment magnitude Mw = 6.08 and 6.18). Different soil types are considered to validate the proposed methodological procedures. The analysis involves the computation of peak ground acceleration motion for two scenario earthquakes: a local-field event with M = 6.8 and a near-field event with Mw = 6.5, representing the average magnitude of the 2000 Baku-Caspian earthquake. The computed peak ground acceleration values are then used to derive intensities. Notably, the 1902 Shamakhi earthquake and the 2000 Baku-Caspian earthquake exhibit similar trends on surface PGA values. The local-field scenario estimates PGA values ranging from 77 to 328 gal, corresponding to MSK-64 scale intensity levels of VII-IX. The near-field scenario, with PGA values ranging from 28 to 62 gal, aligns with MSK-64 intensity levels of VI–VII. In the final assessment, the amplification factor in the study area varies between 0.55 and 0.83. The seismic hazard level is identified as high in the southern and southeastern regions, particularly in areas with soft shallow and medium-depth soils, indicating a high potential for ground motion amplification.