Search

Probabilistic seismic hazard analysis (PSHA) is moving from ergodic ground-motion models (GMMs) to nonergodic GMMs that account for site-source-specific source, path, and site effects and which require a much larger number of GMM branches on the logic tree to capture the full epistemic uncertainty. An efficient method for computing PSHA with a large number of GMM branches was developed by Lacour and Abrahamson (2019) using polynomial chaos (PC) expansion with the key assumption that the epistemic uncertainty in the median ground motion is fully correlated. In the current study, we remove the assumption of full correlation using a multivariate PC expansion. The correlation structure of the available median GMMs across scenarios is computed empirically. The median ground motion is modeled as a Gaussian random process with the correlation structure of the GMMs across the range of relevant earthquake scenarios. This Gaussian random process is discretized using the Karhunen–Loeve expansion, which leads to multivariate PC expansions of uncertain hazard curves. The hazard fractiles can be reconstructed during an efficient postprocessing phase that includes the effects of partial correlation between the GMMs. Multivariate PC expansions require significantly more terms than for the fully correlated case, which increases the calculation time by about a factor of 5, but it is still much more efficient than direct sampling of the branches of the GMM logic for a large number of branches. An example hazard calculation shows that the effect of using partial correlation in place of full correlation of the GMMs is small for the Next Generation Attenuation-West2 (NGA-West2) set of GMMs, indicating that the fully correlated assumption may be adequate for many applications. The multivariate PC method can be used to evaluate the effects of the partial correlation of the GMMs for sets of GMMs that are different from the NGA-West2 GMMs.

As a companion article to Goulet et al., we describe implementation of the NGA-East ground motion characterization (GMC) model in probabilistic seismic hazard analysis (PSHA) for sites in the Central and Eastern United States (CEUS). We present extensions to the EPRI/DOE/NRC seismic source characterization (SSC) model for the CEUS needed for full implementation of NGA-East. Comparisons are presented to the EPRI GMC, the currently accepted model by the U.S. Nuclear Regulatory Commission for hazard assessment at nuclear facilities. Comparisons are presented both in terms of GMC model components and in the resulting seismic hazard assessments for a range of site locations in the CEUS. Illustrations of the effect of various components of the NGA-East GMC on seismic hazard results are also presented. Finally, we present recommendations for application of the NGA-East GMC in PSHA.

Two methods were developed to estimate updated mean seismic hazard for existing probabilistic seismic hazard analyses (PSHAs) due to a change in the ground motion model (GMM). Both methods were used to estimate updated hazard at nuclear power plant (NPP) sites in the Central and Eastern United States (CEUS) for a change from the Electric Power Research Institute (EPRI) 2013 GMM to the Next Generation Attenuation (NGA)-East GMM. These methods present efficient tools to inform decisions on whether to perform a full PSHA revision or other detailed evaluations, especially when a large number of sites must be analyzed. A Simplified Hazard (SiHaz) method was developed to estimate mean hazard explicitly using a reduced PSHA logic tree that incorporates the updated GMM and potential changes in the site response model. An alternative scaling method was independently developed to be applied directly to current CEUS NPP hazard. Both methods were validated using updated PSHA results at several sites. Estimates at 46 NPP sites using both methods showed good agreement for mean annual frequencies of exceedance between 1E-4 and 1E-5/yr.

The interest to build a nuclear power plant in West Nusa Tenggara Province has caused a concern related to the seismic hazard. One of the challenges was that the existing low-resolution national seismic hazard map did not cover the entire islands of West Nusa Tenggara Province. In this paper, preliminary probabilistic seismic hazard analysis of West Nusa Tenggara was performed to complement the existing national seismic hazard map for West Nusa Tenggara. The analysis was carried out using OQ software based on the seismotectonic model and GMPE developed by PUSGEN. Similar seismotectonic model and seismic event trees were implemented using a finer grid spacing of 500–1000 m. The seismotectonic model included active shallow crustal, subduction interface, and background sources with a total number of 50 seismogenic sources and was calculated using 108 GMPE paths. Two hazard levels were considered, each representing 1% and 2% POE in a 100-year return period. A comparison to the existing national seismic hazard map on the PGA level has been presented in this study and yielded a 13–25% difference in the ground acceleration value. Spectral hazard maps representing the shorter period (0.2 s) showed a maximum ground acceleration value of more than 2 g, especially in the northern part of Sumbawa Island. Beyond the longer spectral period (1 s), ground acceleration gradually decreases to the level below the PGA level.

In the past decade, Oklahoma has experienced unprecedented seismicity rates, following an increase in the volumes of wastewater that are being disposed underground. In this article, we perform a probabilistic assessment of the time-dependent seismic hazard in Oklahoma and incorporate these results into an integrated seismic risk model to assess the evolution of the statewide economic losses, including a conservative forecast through 2030. Our risk model employs an injection-driven earthquake rate model, a region-specific ground motion model, a recent Vs30 map, HAZUS exposure data and updated vulnerability curves for both structural and nonstructural elements, and contents. The calculations are performed using a stochastic Monte Carlo–based approach implemented in the OpenQuake engine. The resulting seismic hazard maps illustrate the incompatibility of the regional seismic provisions with the current seismicity. In 2015, in particular, the induced seismic hazard in several places in Oklahoma was higher than along the San Andreas fault. During the peak of seismicity in 2015, the seismic risk was 275 times higher than the background level, with the vast majority of losses originating from damages to nonstructural elements and contents. Our direct economic loss estimates are in reasonable agreement with the paid insurance claims, but show significant sensitivity to the ground motion model selection. The proposed risk model, with possible regular updates on the seismicity rate forecast, can help stakeholders define acceptable production levels.

The incorporation of local amplification factors determined through site response analyses has become standard practice in site-specific probabilistic seismic hazard analysis (PSHA). Another indispensable feature of the current state-of-practice in site-specific PSHA is the identification and quantification of all epistemic uncertainties that influence the final hazard estimates. Consequently, logic trees are constructed not only for seismic source characteristics and ground-motion models (GMMs) but also for the site amplification factors, the latter generally characterized by branches for alternative shear-wave velocity (VS) profiles. However, in the same way that branch weights on alternative GMMs can give rise to unintentionally narrow distributions of predicted ground-motion amplitudes, the distribution of amplification factors obtained from a small number of weighted VS profiles will often be quite narrow at some oscillator frequencies. We propose an alternative approach to capturing epistemic uncertainty in site response in order to avoid such unintentionally constricted distributions of amplification factors using more complete logic-trees for site response analyses. Nodes are included for all the factors that influence the calculated amplification factors, which may include shallow VS profiles, deeper VS profiles, depth of impedance contrasts, low-strain soil damping, and choice of modulus reduction and damping curves. Site response analyses are then executed for all branch combinations to generate a large number 2 of frequency-dependent amplification factors. Finally, these are re-sampled as a discrete distribution with enough branches to capture the underlying distribution of amplification factors (AFs). While this approach improves the representation of epistemic uncertainty in the dynamic site response characteristics, modeling uncertainty in the AFs is not automatically captured in this way, for which reason it is also proposed that a minimum level of epistemic uncertainty should be imposed on the final distribution.

We apply three data-driven selection methods, log-likelihood (LLH), Euclidean distance-based ranking (EDR), and deviance information criterion (DIC), to objectively evaluate the predictive capability of 10 ground-motion models (GMMs) developed from Iranian and worldwide data sets against a new and independent Iranian strong-motion data set. The data set includes, for example, the 12 November 2017 Mw 7.3 Ezgaleh earthquake and the 25 November 2018 Mw 6.3 Sarpol-e Zahab earthquake and includes a total of 201 records from 29 recent events with moment magnitudes 4.5≤Mw≤7.3 with distances up to 275 km. The results of this study show that the prior sigma of the GMMs acts as the key measure used by the LLH and EDR methods in the ranking against the data set. In some cases, this leads to the resulting model bias being ignored. In contrast, the DIC method is free from such ambiguity as it uses the posterior sigma as the basis for the ranking. Thus, the DIC method offers a clear advantage of partially removing the ergodic assumption from the GMM selection process and allows a more objective representation of the expected ground motion at a specific site when the ground-motion recordings are homogeneously distributed in terms of magnitudes and distances. The ranking results thus show that the local models that were exclusively developed from Iranian strong motions perform better than GMMs from other regions for use in probabilistic seismic hazard analysis in Iran. Among the Next Generation Attenuation-West2 models, the GMMs by Boore et al. (2014) and Abrahamson et al. (2014) perform better. The GMMs proposed by Darzi et al. (2019) and Farajpour et al. (2019) fit the recorded data well at short periods (peak ground acceleration and pseudoacceleration spectra at T=0.2 s). However, at long periods, the models developed by Zafarani et al. (2018), Sedaghati and Pezeshk (2017), and Kale et al. (2015) are preferable.

Recent research has highlighted the usefulness of cumulative absolute velocity [Formula: see text] in several contexts, including using the [Formula: see text] at the ground surface for earthquake early warning and using the [Formula: see text] at rock reference conditions for evaluation of the liquefaction risk facing structures. However, there are relatively few ground motion prediction equations for CAV, they are based on relatively small data sets, and they give relatively similar results. This study develops nine ground motion prediction equations for [Formula: see text] based on a global database of ground motion records from shallow crustal earthquakes. Its provision of nine models enables characterization of epistemic uncertainty for ranges of earthquake characteristics that are sparsely populated in the regression database. The functional forms provide different perspectives on extrapolation to important ranges of earthquake characteristics, particularly large magnitude events and short distances. The variability and epistemic uncertainty in the models are characterized. Spatial autocorrelation of the models’ errors is investigated. The models’ predictions agree with existing broadly applicable models at small to moderate magnitudes and moderate to long distances. These models can be used to improve hazard analysis of [Formula: see text] that incorporates the influence of epistemic uncertainty.

This study presents hazard estimation of Kashmir Basin, NW Himalaya using regional ground motion relations, representing one of the most seismically active region in the Himalayan belt. Fault-level seismic recurrence parameters are determined from an updated earthquake catalogue spanning from 25 to 2018 AD for all possible seismic sources. The estimated hazard maps are presented for three ground motion parameters (PGA, short and long period spectral acceleration) for 50, 100, 500 and 2500 years return periods. Moreover, uniform hazard response spectrums and hazard curves are presented for all ten districts of the basin. The southern section of the basin consisting districts of Budgam, Shopian, Pulwama and Kulgam show higher hazard levels due to presence of numerous seismogenic structures in close vicinity. Our results highlight that the imposed seismic hazard in Kashmir basin is highly underestimated which need to be redressed by modifying the current provisional design standards.

The Taiwan Earthquake Model (TEM) published the first version of the Taiwan probabilistic seismic hazard assessment (named TEM PSHA2015) 5 years ago. For updating to the TEM PSHA2020, we considered an updated seismogenic structure database, including the structures newly identified with 3D geometry, an earthquake catalog made current to 2016, state-of-the-art seismic models, a new set of ground motion prediction equations, and site amplification factors. In addition to earthquakes taking place on each individual seismogenic structure, the updated seismic model included the possibility of an earthquake occurring on multiple structures. To include fault memory for illustrating activity on seismogenic structure sources, we incorporated the Brownian passage time model. For the crustal seismicity that cannot be attributed to any specific structure, we implemented both area source and smoothing kernel models. A new set of ground motion prediction equations is incorporated. In addition to the calculation of hazard at engineering bedrock, our assessment included site amplification factors that competent authorities of governments and private companies could use to implement hazard prevention and reduction strategies.

A probabilistic seismic hazard assessment is applied, for the first time in the case of Sakhalin Island, using a regional ground motion prediction equation in terms of the peak ground acceleration. The seismic hazard maps for 475-year-long and 975-year-long recurrence intervals are made for southern Sakhalin Island.

We present an evaluation of the 2018 Northern Southeast Asia Seismic Hazard Model (NSAHM18) based on a combination of smoothed seismicity, subduction zone, and fault models. The smoothed seismicity is used to model observed distributed seismicity from largely unknown sources in the current study area. In addition, due to a short instrumental earthquake catalog, slip rate and characteristic earthquake magnitudes are incorporated through the fault model. To achieve this objective, the compiled earthquake catalogs and updated active fault databases in this region were reexamined with consistent use of these input parameters. To take into account epistemic uncertainty, logic tree analysis has been implemented incorporating basic quantities such as ground-motion models (GMMs) for three different tectonic regions (shallow active, subduction interface, and subduction intraslab), maximum magnitude, and earthquake magnitude frequency relationships. The seismic hazard results are presented in peak ground acceleration maps at 475- and 2475-year return periods.

We construct a probabilistic seismic hazard model for mainland China by integrating historical earthquakes, active faults, and geodetic strain rates. We delineate large seismic source zones based on geologic and seismotectonic characteristics. For each source zone, a tapered Gutenberg–Richter (TGR) distribution is used to model the total seismic activity rates. The TGR a- and b-values are calculated using a new earthquake catalog, while corner magnitudes are constrained using the seismic moment rate inferred from a geodetic strain rate model. For hazard calculations, the total TGR distribution is split into two parts, with smaller ( MW

The Caucasus has a documented history of cataloging earthquakes stretching back to the beginning of the Christian era. Instrumental seismic observation in the Caucasus began in 1899, when the first seismograph was installed in Tbilisi, Georgia. During the Soviet era (1921–1991 in Georgia), the number of seismic stations increased in the region, providing better network coverage and a valuable dataset for seismic research. Data from many thousands of earthquakes recorded by this regional network was stored on paper in seismic bulletins. As part of the project outlined in this article, we pulled together and digitized all available paper bulletins from Georgia and neighboring countries. This allowed significant Limprovements in location accuracy and recalculation of more robust moment magnitudes for earthquakes in this region. It also paved the way for future collaboration and data exchange among the countries in the Caucasus.The resulting earthquake catalog with the new locations and magnitudes was used to conduct a probabilistic seismic hazard assessment to support a major update to the building code in Georgia to align it with the European codes. This article outlines the improvements made to the earthquake catalog in Georgia using legacy data and the new hazard assessment based on this improved dataset.

We use an epidemic-type aftershock sequence (ETAS) based approach to develop a regionally optimized background earthquake rates from ETAS (ROBERE) method for probabilistic seismic hazard assessment. ROBERE fits parameters to the full seismicity catalog for a region with maximum-likelihood estimation, including uncertainty. It then averages the earthquake rates over a suite of catalogs from which foreshocks and aftershocks have been removed using stochastic declustering while maintaining the same Gaussian smoothing currently used for the U.S. Geological Survey National Seismic Hazard Model (NSHM). The NSHM currently determines these rates by smoothing a single catalog from which foreshocks and aftershocks have been removed using the method of Gardner and Knopoff (1974; hereafter, GK74). The parameters used in GK74 were determined from subjectively identified aftershock sequences, unlike ROBERE, in which both background rate and aftershock triggering parameters are objectively fitted. A major difference between the impacts of the two methods is GK74 significantly reduces the b-value, a critical value for seismic hazard analysis, whereas ROBERE maintains the original b-value from the full catalog. We apply these methods to the induced seismicity in Oklahoma and Kansas and tectonic activity in the San Francisco Bay Region. Using GK74 gives lower overall earthquake rates but estimates higher hazard due to the reduction in the b-value. ROBERE provides higher earthquake rates, at the magnitude of completeness, but lower hazard because it does not alter the b-value. We test two other declustering methods that produce results closer to ROBERE but do not use objectively fit parameters, include uncertainty, and may not work as well in other areas. We suggest adopting ROBERE for the NSHM so that our hazard estimates are based on an objective analysis, including uncertainty, and do not depend strongly on potentially biased b-values, which was never the goal of the existing methodology.

The Philippine archipelago is tectonically complex and seismically hazardous, yet few seismic hazard assessments have provided national coverage. This article presents an updated probabilistic seismic hazard analysis for the nation. Active shallow crustal seismicity is modeled by faults and gridded point sources accounting for spatially variable occurrence rates. Subduction interfaces are modeled with faults of complex geometry. Intraslab seismicity is modeled by ruptures confined to the slab volume. Source geometries and earthquake rates are derived from seismicity catalogs, geophysical data sets, and historic-to-paleoseismic constraints on fault slip rates. The ground motion characterization includes models designed for global use, with partial constraint by residual analysis. Shallow crustal faulting near metropolitan Manila, Davao, and Cebu dominates shaking hazard. In a few places, peak ground acceleration with 10% probability of exceedance in 50 years on rock reaches 1.0 g. The results of this study may have utility for defining the design base shear in the National Structural Code of the Philippines.

Aftershock probabilistic seismic hazard analysis (APSHA) has a key role in risk management after a major earthquake. The main goal of the current study is to assess aftershock hazard in a strategic and earthquake-prone region of Iran (Bushehr province). Bushehr province is a strategic region in the Middle East due to its major petroleum export facilities, industrial corridors and the Bushehr nuclear power plant. To prepare APSHA for Bushehr province, a seismic source is selected which surrounds the active faults in the study area. A uniform earthquake catalog is collected which contains information on a total of 1143 earthquakes (Mw > 4) occurred in the study area from 1900 to 2018. Aftershock parameters are calculated using the epidemic-type aftershock sequence model. Aftershock sequences follow a non-homogenous Poisson’s process, and their magnitude and location depend on the size and location of the mainshock. In this study, APSHA is performed for the intervals of 1, 7 and 30 days after the mainshock, by assuming occurrence of mainshocks with return periods of 475 and 2475 years. The results show that the aftershock hazard curve is greater than that of the mainshock hazard curve.

We present the first probabilistic seismic hazard assessment (PSHA) specifically for the Shanxi rift system, north China, which has been defined as one of the areas of highest seismic hazard and risk in China in recent decades. We applied a Monte Carlo-based approach to PSHA, based on so far the most complete earthquake catalog available, a detailed zonation considering both seismicity distribution and local tectonic features, a logic tree of carefully selected ground-motion prediction equations, as well as a cautious consideration of actual local site effects for this region. Both areal sources (for Ms

We would like to make some comments on the paper by Hamidatou et al. (2019). Initially, these comments are motivated to reveal that, previous results on probabilistic seismic hazard analyses, some of them computed and published by our research group, are wrongly quoted in the paper by these authors. In our opinion, some other points are worthy of debate, mainly, but not only, the used seismic source zone model, the used logic-tree, and also the comparison of estimated values of peak ground horizontal acceleration (PGA) with previous results.

This study aims to build on the existing knowledge and improve the overall PSHA results by modifying source, path and site characteristics for Bangladesh. Firstly, six potential seismotectonic zones have been re-defined based on the recent study of Wang et al. (J Geophys Res Solid Earth 119:3576–3822, 2014) and Nath and Thingbaijam (J Seismol 15(2):295–315, 2011), and the updated earthquake catalogue has been declustered using two methods. Important source parameters, such as recurrence b-values and maximum magnitudes, have been determined using the Maximum Likelihood and cumulative moment methods, respectively, and their uncertainties have been addressed using a logic-tree approach. Secondly, based on literature review and studies in neighboring countries, suitable GMPEs have been selected for the seismic zones and the uncertainties have been addressed using a logic tree approach. A significant novelty of the study lies in the consideration of the site effects by integrating Vs30 values throughout the country. The ground motions—PGA and SA (at 0.2, 1.0 and 2.0 s) are computed using GEM’s OpenQuake and presented in form of hazard maps for 2% and 10% probabilities of exceedance in 50 years as well as mean hazard curves and uniform hazard spectra. Disaggregation for capital city Dhaka has also been carried out to show the hazard contributions of magnitude–distance pairs. The spatial distribution of PGA and SA are found remarkably higher than previous findings, likely due to differences in parameters and uncertainties. The results show a marked increase (by almost 20%) in the observed ground motions with respect to those carried out previously by uniformly characterizing the whole country as a firm rock.

Peak ground acceleration (PGA) and study area (SA) distribution for the Patna district are presented considering both the classical and zoneless approaches through a logic tree framework to capture the epistemic uncertainty. Seismicity parameters are calculated by considering completed and mixed earthquake data. Maximum magnitude is calculated using three methods, namely the incremental method, Kijko method, and regional rupture characteristics approach. The best suitable ground motion prediction equations (GMPEs) are selected by carrying out an “efficacy test” using log likelihood. Uniform hazard response spectra have been compared with Indian standard BIS 1893. PGA varies from 0.38 to 0.30 g from the southern to northern periphery considering 2 % probability of exceedance in 50 years.

Two approaches are used to develop the vertical design spectra in probabilistic seismic hazard assessment (PSHA): (i) performing PSHA for the vertical component using vertical ground motion models (GMMs), and (ii) utilizing the V/H ratio GMMs to scale the horizontal spectrum during or after the hazard calculations for the horizontal component. This study intends to develop a new framework for building a vertical ground motion logic tree that combines both vertical and V/H ratio GMMs in the PSHA analysis. For this purpose, the Turkish strong motion database (TSMD) (Akkar et al. in Journal of Seismology 14(3): 457–479, 2010) is updated with recordings from earthquakes that occurred between 2008 and 2015, including the Mw = 6.1 Elazig and Mw = 7.2 Van earthquakes. The updated TSMD contains 2698 recordings, with the earthquake metadata, source-to-site distance metrics for the recordings, measured shear wave velocity profiles for the recording stations, and spectral accelerations for horizontal and vertical components. Four candidate V/H ratio GMMs (proposed by Gulerce and Abrahamson in Earthquake Spectra 27(4): 1023–1047, 2011; Gulerce and Akyuz in Seismological Research Letters 84(4): 678–687, 2013; Akkar et al. in Bulletin of Earthquake Engineering 12(1): 517–547, 2014a; Bozorgnia and Campbell in Earthquake Spectra, 32(2): 951–978, 2016) are selected, and the model predictions are compared with the actual data in the updated TSMD using the analysis of the residuals and data-driven techniques. Analysis results showed that the median predictions of the V/H ratio GMMs proposed by Gulerce and Akyuz (2013), Akkar et al. (2014a), and Bozorgnia and Campbell (2016) are compatible with the V/H ratios in the Turkish strong motion dataset. The findings of this study and the compatibility analysis of the vertical GMMs with the updated TSMD are combined in the accompanying manuscript (Gulerce et al. in Pure and Applied Geopyhsics, 2019, submitted) to provide the complete framework of ground motion characterization for vertical ground motion component.

In this paper, an evaluation of the variation of earthquake ground motions with a focus on site amplification factors based on spectral analysis is presented. By using strong motion record obtained at six sites in Japan, probability distributions of site amplification factors were shown. The relations between standard deviations of site amplification factors and distances between the sites were studied. The variations of representative values of earthquake ground motions based on the variations of site amplification factors were discussed by using probabilistic seismic hazard analysis with focus on Fourier amplitude and group delay time. The distributions of peak ground accelerations and peak ground velocities were shown. It is suggested that design earthquake ground motions considering the average site amplification factors may lead the engineering design on the dangerous side.

We assessed the applicability of several ground‐motion models (GMMs) against Iran’s local data. Candidate GMMs are selected from those developed for shallow crustal regions such as Iran, Turkey, Japan, Europe and the Middle East, and the western United States. We made the evaluation database prospective to all candidate GMMs to assess their predictive capability. The evaluation database is composed of 643 records from 240 earthquakes with magnitudes ranging from 3.9 to 7.3 and Joyner–Boore distances up to 300 km. We implemented the log‐likelihood method of Scherbaum et al. (2009), the Euclidean distance‐based ranking proposed by Kale and Akkar (2013), and the multivariate logarithmic score of Mak et al. (2017) to evaluate the candidate models. We ranked GMMs by paying attention to the issue of score variability. To assess the score variability, we generated resampled datasets from the whole database using the cluster bootstrap technique and ranked models based on their relative performance among bootstrap samples. Overall, Sedaghati and Pezeshk (2017), Zafarani et al. (2018), and Farajpour et al. (2019) local models outperform remaining models considering the whole database over the entire frequency range. For high‐seismicity regions, the Zhao et al. (2006) model can be used in line with the first two local models to better quantify epistemic uncertainties associated with the process of model selection. In addition to aforementioned local models, Bindi et al. (2014) show acceptable performance against small‐to‐moderate magnitude data and may be considered for estimating seismic hazard in low‐seismicity regions of Iran.

The article describes the stages and results of research for developing a prototype of general seismic zoning maps of the Russian Federation (GSZ-2016). The standard GSZ-2015 and GSZ-2016 maps are compared. The need to transition to seismic hazard mapping of the Russian Federation in parameters of physical characteristics is shown. The global experience of using peak ground accelerations and amplitudes of the response spectra as mapped values is considered. Existing software packages for probabilistic seismic hazard analysis are briefly reviewed and compared. The problem of using empirical seismic wave attenuation models to calculate GSZ maps for the Russian Federation is examined. The first results of such calculations are presented. Further research trends are defined for preparing the initial database required for creating new-generation GSZ maps for the Russian Federation.

The study focuses on improving the methodology and technology of detailed seismic zoning of the Sakhalin region and adjacent areas. A new regional GMPE (ground motion prediction equation) model derived from instrumental data of accelerometer and seismometer networks are considered, as well as imported new-generation advanced (NGA 2) models. The results of probabilistic seismic hazard analysis (PSHA) for Yuzhno-Sakhalinsk has made it possible to compare the hazard curves of physical parameters produced by different attenuation models. The advantages and disadvantages of the regional GMPE model are discussed based on this comparison. Practical recommendations are given.

We used three state-of-the-art machine learning techniques (boosted regression tree, random forest, and support vector machine) to produce a multi-hazard (MHR) map illustrating areas susceptible to flooding, gully erosion, forest fires, and earthquakes in Kohgiluyeh and Boyer-Ahmad Province, Iran. The earthquake hazard map was derived from a probabilistic seismic hazard analysis. The mean decrease Gini (MDG) method was implemented to determine the relative importance of effective factors on the spatial occurrence of each of the four hazards. Area under the curve (AUC) plots, based on a validation dataset, were created for the maps generated using the three algorithms to compare the results. The random forest model had the highest predictive accuracy, with AUC values of 0.994, 0.982, and 0.885 for gully erosion, flooding, and forest fires, respectively. Approximately 41%, 40%, 28%, and 3% of the study area are at risk of forest fires, earthquakes, floods, and gully erosion, respectively.

SUMMARY The increasing density of geodetic measurements makes it possible to map surface strain rate in many zones of active tectonics with unprecedented spatial resolution. Here we show that the strain tensor rate calculated from GPS in the India–Asia collision zone represents well the strain released in earthquakes. This means that geodetic data in the India–Asia collision zone region can be extrapolated back in time to estimate strain buildup on active faults, or the kinematics of continental deformation. We infer that the geodetic strain rates can be assumed stationary through time on the timescale needed to build up the elastic strain released by larger earthquakes, and that they can be used to estimate the probability of triggering earthquakes. We show that the background seismicity rate correlates with the geodetic strain rate. A good fit is obtained assuming a linear relationship ($\dot{N} = \lambda \ \cdot \dot{\epsilon }$, where $\dot{N}$ is the density of the rate of Mw ≥ 4 earthquakes, $\dot{\epsilon }$ is strain rate and λ = 2.5 ± 0.1 × 10−3 m−2), as would be expected from a standard Coulomb failure model. However, the fit is significantly better for a non-linear relationship ($\dot{N} = \gamma _1 \cdot \dot{\epsilon }^{\gamma _2}$ with γ1 = 2.5 ± 0.6 m−2 and γ2 = 1.42 ± 0.15). The b-value of the Gutenberg–Richter law, which characterize the magnitude–frequency distribution, is found to be insensitive to the strain rate. In the case of a linear correlation between seismicity and strain rate, the maximum magnitude earthquake, derived from the moment conservation principle, is expected to be independent of the strain rate. By contrast, the non-linear case implies that the maximum magnitude earthquake would be larger in zones of low strain rate. We show that within areas of constant strain rate, earthquakes above Mw 4 follow a Poisson distribution in time and and are uniformly distributed in space. These findings provide a framework to estimate the probability of occurrence and magnitude of earthquakes as a function of the geodetic strain rate. We describe how the seismicity models derived from this approach can be used as an input for probabilistic seismic hazard analysis. This method is easy to automatically update, and can be applied in a consistent manner to any continental zone of active tectonics with sufficient geodetic coverage.

As part of a feasibility study for a potential electronuclear power plant in Tunisia, a new seismic catalogue and a new seismic source model have been proposed, for Tunisia and adjacent areas. These findings are used in this study to evaluate the seismic hazard for two selected nuclear power plant sites: Marsa-Douiba in the north and Skhira in the southeast. The investigations and the assessment are conducted at a regional scale. The location of these sites and the choice of a regional review extent allow us to consider the entire Tunisian territory as the study area and also to conduct the hazard evaluation at a ‘national level’. In this study, a probabilistic seismic hazard analysis is performed using the R-CRISIS software, and a logic tree is developed to capture uncertainties related to the characterisation of both seismic sources and ground-motion. First, a hazard map of Tunisia for the return period of 475 years is generated to compare it with the ones obtained within previous studies. Then, the 10,000-year return period is considered, and the recommendations of the International Atomic Energy Agency are applied to determine the uniform hazard spectrums and perform disaggregation for the selected sites.

The growth of global ground-motion databases has allowed generation of non-ergodic ground-motion prediction equations (GMPEs) based on specific on-site recordings. Several studies have investigated the differences between the hazard estimates from ergodic versus non-ergodic GMPEs. Here instead we focus on the impact of non-ergodic PSHA estimates on the seismic risk of nonlinear single-degree-of-freedom systems representing ductile structures and compare it with the traditional risk estimates obtained using ergodic GMPEs. The structure-and-site-specific risk estimates depend not only on the difference in the hazard estimates but also on the different hazard-consistent ground-motion record selection that informs the response calculation. The more accurate structure-and-site-specific non-ergodic risk estimates show that traditional ones may be biased in a way impossible to predict a priori. Hence, the use of the non-ergodic approach is recommended, whenever possible. However, further advancements of non-ergodic GMPEs are necessary before being routinely utilized in real-life risk assessment applications.

Seismic source characterization (SSC) for probabilistic seismic hazard assessment (PSHA) in regions characterized by subduction megathrust involves a considerable ambiguity. Lack of detailed geologic, seismic, and geodetic data increases the uncertainties. The enigma is enhanced in regions where thin-skinned accretionary prism faults are part of active deformation. In this study, a planar SSC model for seismically active eastern Makran subduction zone, its associated accretionary prism faults and Chaman transform fault zone is proposed based on a kinematic block model accounting strain partitioning. Sensitivity tests for various parameters of the SSC model are performed by computing peak ground acceleration (PGA) maps for 475-year return period. Among alternative magnitude distribution models, the truncated exponential model gives ~10% higher PGA values than the composite recurrence model, which is favored by the observed subduction seismicity. Especially in gently dipping subduction zones such as Makran, estimated PGA values and their spatial distribution are highly sensitive (changing up to 0.3 g) to the megathrust interface geometry near the surface, dip amount, and defined depth limits. Gentler and deeper extending subduction interface maximizes the rupture width and results in higher PGA values towards inland, while the inclusion of shallow aseismic portion produces larger PGA values near the trench. Through eastward transition from subduction to transform motion, thrust faults within the Makran accretionary prism bend northward, forming oblique fault systems with higher slip rates and accommodate part of the compression in accordance with the lateral slip rate variations identified on the Chaman transform fault zone. Using the selected SSC model, the PGA values for short return period (475-year) are estimated to be between 0.5 g and 0.7 g within the accretionary prism down to the trench across the Makran subduction and along southern section of Chaman transform fault zone where slip rates are higher relative to the north.

Although Brazilian seismic activity is defined as low to moderate, it is known that intraplate earthquakes can also be associated to high intensities. In Brazil, the state of Rio Grande do Norte (RN) is one of the most seismically active areas, but there is no specific study to evaluate the seismic hazard in this region. This paper presents analyses towards improving the seismic hazard map, the peak ground acceleration value and the response spectrum of RN. The methodology is based on Probabilistic Seismic Hazard Analysis, comparing the results to the design criteria defined in the Brazilian code NBR 15421:2006 (Design of seismic resistant structures – Procedure). The analyses show that, in general, the code sets conservative values for the peak ground acceleration and for the design response spectrum; however, related to this last one, the shape is quite different.

A method is proposed to identify within seismic catalogs those earthquakes that are most relevant to the seismic hazard. The approach contrasts with the classical approach to decluster the seismic catalog with the expectation that the remaining main shocks will be the relevant events for the seismic hazard analysis. We apply a time window like in the window declustering approach of Gardner and Knopoff, but the time window is motivated by relevance to engineering. A ground motion criterion replaces the spatial window. An event in the time window is included in the “Maximum Shaking Earthquake Catalog (MSEQ catalog)” if the median ground motion at its epicenter exceeds the predicted median ground motion there from the main shock, using a locally appropriate ground motion prediction equation. Ground motion can be measured by any parameter that is estimated by a ground motion prediction equation. We consider peak acceleration and spectral amplitude (SA) at periods of 0.2, 1.0, and 3.0 s. The longer period parameters systematically remove more small events. The purpose is not to produce a declustered catalog, in which each group of physically related earthquakes is represented by its largest event. Statistical properties of the MSEQ catalog somewhat resemble the corresponding declustered catalog in three tested regions, but the MSEQ catalogs all retain more large-magnitude earthquakes. The MSEQ catalog may better represent the potential hazard in a region, and thus might be considered as an alternative to a declustered catalog in developing the seismicity model for probabilistic seismic hazard analysis.

In this study focused on France, we explore the uncertainties related to choices made while building a source model for hazard assessment and we quantify the impact on probabilistic hazard estimates. Earthquake recurrence models are initially built from the French Seismic CATalog (FCAT, Manchuel et al. in Bull Earthq Eng, 2018.10.1007/s10518-017-0236-1). We set up a logic tree that includes two alternative seismogenic source models (ESHM13 and Baize et al. in Bull Soc Géol Fr 184(3):225–259, 2013), two versions of FCAT catalog, two alternative declustering algorithms, and three alternative minimum magnitudes for earthquake recurrence modeling. We calculate the hazard for six cities (i.e. Nantes, Lourdes, Clermont-Ferrand, Briançon, Nice and Strasbourg) that are located in source zones with a minimum amount of data to work with. Results are displayed for the PGA and spectral period 0.2 s, at return periods 475 and 5000 years. Exploration of the logic tree shows that the parameters with the most impact on hazard results are the minimum magnitude used in the recurrence modeling (up to 31%) and the selection of the seismogenic source model (up to 30%). We also use the SHARE European Earthquake Catalog (SHEEC, Woessner et al. in Bull Earthquake Eng, 2015.10.1007/s10518-015-9795-1) to build earthquake recurrence models and compare hazard values obtained with the FCAT logic tree. Comparisons are limited because of the low number of events available in some sources in SHEEC; however, results show that, depending on the site considered, the earthquake catalog selection can also strongly impact the hazard estimates (up to 50%). The FCAT logic tree is combined with four ground-motion models (Bindi et al. in Bull Earthq Eng 12(1):391–430, 2014; Boore et al. in Earthq Spectra 30(3):1057–1085, 2014; Cauzzi et al. in Bull Earthq Eng 13(6):1587–1612, 2015.10.1007/s10518-014-9685-y; Drouet and Cotton in Bull Seismol Soc Am 105(4):1883–1902, 2015) to account for the epistemic uncertainty on the prediction of ground-motion. Exploration of the logic tree shows that the contribution of ground-motion model uncertainties can be larger than, equivalent to, or lower than the contribution of the source-model uncertainties to the overall hazard variability. Which component controls overall uncertainty depends on the site, spectral period and return period. Finally, exploring the logic tree provides a distribution for the ratios between hazard levels at 5000 and 475 years return periods, revealing that the ratios only slightly depend on source-model uncertainties, vary strongly from site to site, and can take values between 3 and 5, which is significantly higher than what is commonly assumed in the engineering community.

Earthquake Monitoring Center (EMC) at Sultan Qaboos University (SQU) initiated evaluating the seismic hazard in the Sultanate of Oman in 2009. EMC has produced the first probabilistic and deterministic seismic hazard maps for Oman in 2012 and 2013, respectively. In the current study, the probabilistic seismic hazard assessment (PSHA) is revisited to provide an updated assessment of the seismic actions on the Sultanate. The present study has several advantages over its predecessor: using an updated homogeneous earthquake catalogue, recently developed seismic source model; inclusion of epistemic uncertainties for the source models, recurrence parameters, maximum magnitude, and more recent and applicable ground-motion prediction equations (GMPEs). Epistemic uncertainties were treated using a combination of the best available databases within a properly weighted logic tree framework. Seismic hazard maps in terms of horizontal peak ground acceleration (PGA) and 5% damped spectral accelerations (SA) at the bedrock conditions (VS = 760 m/s) for 475- and 2475-year return periods were generated using the classical Cornell-McGuire approach. Additionally, uniform hazard spectra (UHS) for the important population centers are provided. The results show higher values at the northern parts of the country compared to the hazard values obtained in the previous study.

The Southwestern part of India investigated in the present study mainly comprises of states such as Goa, north Kerala and a major portion of Karnataka. A comprehensive regional seismic catalog has been compiled spanning over 190 years apart from a few prehistoric events from the early 16th century. The classical Cornel–McGuire approach has been incorporated in the estimation of seismic hazard. The seismic sources are modeled as area sources and the entire study region is divided into four seismogenic source zones. The uncertainties involved in the formulation of the seismic source model and ground motion prediction model has been discussed in detail. Further, the procedure for selecting appropriate GMPEs involves the evaluation of multidimensional (M, R, T) ground motion trends and performance against observed macroseismic data. The epistemic uncertainty in the estimation of seismicity parameters and ground motion prediction equations (GMPEs) has been addressed using logic tree computation. The results of the hazard analysis demonstrate that the existing seismic code underestimates the seismic potential of seismic zone II (BIS 1893) areas. The de-aggregation of the predicted seismic hazard revealed earthquakes of magnitude range (Mw) 4–6 occurring within a distance of 35kms to be most influential for any given site of interest. Sensitivity analysis has been performed for crucial input parameters in the formulation of seismic source and ground motion models. Site amplification study has been carried out using topographic slope as a proxy to shear velocity in the top 30 m (Vs30). A maximum of 60% to 80% amplification has been observed in the study area. The seismic hazard maps in terms of PGA have been plotted for the seismic hazard estimated at the bedrock level as well as the surface level for 2% and 10% probability of exceedance in 50 years. The hazard estimation specifically for the southern part of the west coast is the first of its kind. The investigation suspects mining-induced seismicity in Bellary and Raichur districts though there is no mention of this in the prior literature.

We describe a first principles methodology to evaluate statistically the hazard related to non-stationary seismic sources like induced seismicity. We use time-dependent Gutenberg–Richter parameters, leading to a time-varying rate of earthquakes. We derive analytic expressions for occurrence rates which are verified using Monte Carlo simulations. We show two examples: (1) a synthetic case with two seismic sources (background and induced seismicity); and (2) a recent case of induced seismicity, the Horn River Basin, Northeast British Columbia, Canada. In both cases, the statistics from the Monte Carlo simulations agree with the analytical quantities. The results show that induced seismicity affects seismic hazard rates but that the exact change greatly depends on both the duration and intensity of the non-stationary sequence as well as the chosen evaluation period. The developed methodology is easily extended to handle spatial source distributions as well as ground motion analysis in order to generate a complete methodology for non-stationary probabilistic seismic hazard analysis.

Use of faults in seismic hazard models allows us to capture the recurrence of large-magnitude events and therefore improve the reliability of probabilistic seismic hazard assessment (PSHA). In the past decades, fault segmentation provided an important framework for quantifying fault-based PSHA. Recent complex coseismic ruptures (e.g., 2010 Mw 7.1 Canterbury, 2012 Mw 8.6 Sumatra, 2016 Mw 7.8 Kaikōura, 2016 Mw 6.5 central Italy) have shown the need to consider different possible combinations of rupture scenarios in PSHA. Here we present two new methodologies that model rates of ruptures along complex fault systems, one based on a floating rupture approach (FRESH) and another one based on assumed rupture scenarios (SUNFISH). They represent alternatives to a recently proposed approach (SHERIFS), and further step to overcome the segmented and un-segmented approaches commonly used in PSHA in Europe. Differences among SHERIFS, SUNFiSH and FRESH are related to the way slip rate, rupture geometries and magnitude–frequency distributions are modelled. To quantify the differences between these three methodologies, we compared PSHA results based on geometries and slip rates of a fault system located in northeastern Italy, assuming a given maximum magnitude and the same seismic moment rate target. Differences up to 20–30% in the peak ground acceleration at 2% and 10% in 50 years are observed. Finally, we show that the three methodologies are able to solve for the long-term rate of ruptures with resulting PSHA that reflect the fault system geometry and slip rates, without any assumption on segment boundaries. Using fault-based approaches in PSHA requires collecting as much local geological information as possible. Now that multi-fault rupture approaches are available, simplifying assumptions often made to model complex fault systems (uniform slip rate, segmentation hypothesis) are no longer necessary. On the other hand, local data collection should be strongly encouraged to better characterize the actual fault slip rate variability and the complex 3D geometries.

Probabilistic seismic hazard analysis (PSHA) can accommodate various sources of uncertainties and it provides a rational framework for the precise portrayal of the hazard of a given region. Often, the information used in the PSHA is incomplete and uncertain; therefore, the question arises how the uncertainty of the input data affects the estimated hazard characteristics. In this study, sensitivity analysis (SA) was conducted to identify the most dominant inputs affecting the assessment of the key seismicity parameters (SPs), including the mean seismic activity rate λ, b value of Gutenberg–Richter, and the maximum possible earthquake magnitude $$m_{ \hbox{max} }$$ . The study was applied in five areas of Iran, for which such analyses have not been conducted in previous studies. Subsequently, Monte Carlo simulation was employed to determine the effects of the uncertain input parameters on PSHA relevant to spectral accelerations corresponding to 10% and 2% probability of exceedance at least once in 50 years. For this purpose, a unified and declustered earthquake catalogue was used for the five major seismotectonic provinces of Iran (Alborz-Azarbayejan, Zagros, Central-East Iran, Koppeh Dagh, and Makran). The results showed that the last (complete) part of the catalogue has a significant effect on the estimated value of seismic activity and the b value. In contrast, its influence is insignificant on the area-characteristic maximum possible earthquake magnitude, for which the most influential inputs are the maximum observed earthquake and its uncertainty. Furthermore, the uncertainties of the input SPs affected the seismic hazard estimates substantially and led to significant variability in the estimated ground motion characteristics.

Background: Seismic hazard assessments produced in a probabilistic form have become the most common in the world in recent decades. Thus, specifically designed for the purpose of detailed engineering and seismic works, methodology for the probabilistic assessment of the potential seismic effects of earthquakes is proposed. Objective: The focus of this paper is probabilistic hazard assessments for the territory of the Republic of North Ossetia-Alania. Methods: The technique involves a set of models to calculate the potential seismic impacts both from certain earthquakes with different magnitude and distance and from the total number of all potentially hazardous earthquakes with different magnitude and distance, taking into account their recurrence. Conclusion:Specifically designed for the purpose of detailed engineering and seismic works, methodology for the probabilistic assessment of the potential seismic effects of earthquakes is proposed. Results:The results of calculations are presented in the form of the probability distribution functions optimally suitable for subsequent seismic risk assessments, as well as various soil reactions such as seismogenic liquefaction, seismogenic landslides, subsidence, etc

It has been suggested that a better knowledge of fault locations and slip rates improves seismic hazard assessments. However, the importance of detailed along‐fault‐slip‐rate profiles and variable fault geometry has not yet been explored. We quantify the importance for modeled seismicity rates of using multiple throw‐rate measurements to construct along‐fault throw‐rate profiles rather than basing throw‐rate profiles on a single measurement across a fault. We use data from 14 normal faults within the central Italian Apennines where we have multiple measurements along the faults. For each fault, we compared strain rates across the faults using our detailed throw‐rate profiles and degraded data and simplified profiles. We show the implied variation in average recurrence intervals for a variety of magnitudes that result. Furthermore, we demonstrate how fault geometry (variable strike and dip) can alter calculated ground‐shaking intensities at specific sites by changing the source‐to‐site distance for ground‐motion prediction equations (GMPEs). Our findings show that improved fault‐based seismic hazard calculations require detailed along‐fault throw‐rate profiles based on well‐constrained local 3D fault geometry for calculating recurrence rates and shaking intensities.

A site-specific probabilistic seismic hazard analysis (PSHA) has been performed for the only nuclear power plant site in Brazil, located 130 km southwest of Rio de Janeiro at Angra dos Reis. Logic trees were developed for both the seismic source characterisation and ground-motion characterisation models, in both cases seeking to capture the appreciable ranges of epistemic uncertainty with relatively few branches. This logic-tree structure allowed the hazard calculations to be performed efficiently while obtaining results that reflect the inevitable uncertainty in long-term seismic hazard assessment in this tectonically stable region. An innovative feature of the study is an additional seismic source zone added to capture the potential contributions of characteristics earthquake associated with geological faults in the region surrounding the coastal site.

The probabilistic seismic hazards in the larger cities of western Nevada are dominated by the Mount Rose and Carson Range faults. These are normal faults; Reno and Carson City are on the hanging wall. This paper finds the sensitivity of the hazard posed by these faults to epistemic uncertainty of dip, slip rate, magnitude, and choice of ground motion prediction equation (GMPE) for an earthquake that ruptures the length of each fault. SA(0.01 s) with exceedance probability of 2% in 50 years was determined for each branch of a logic tree for those properties. This paper uses tools of probability theory to characterize this logic tree distribution with a lognormal distribution function, and to suggest an estimate of uncertainty in the mean hazard. In this case, which should not be considered general, on the hanging wall, the uncertainties in fault dip and the choice of the GMPE are the dominant contribution to the standard deviation of the lognormal distribution function. Elsewhere nearby, uncertainties in fault slip rate dominate, while at larger distances uncertainties in the magnitude dominate.

The evaluation of evidence provided by geological studies and historical catalogs indicates that in some seismic regions and faults, multiple large earthquakes occur in cluster. Then, the occurrences of large earthquakes confront with quiescence and only the small-to-moderate earthquakes take place. Clustering of large earthquakes is the most distinguishable departure from the assumption of constant hazard of random occurrence of earthquakes in conventional seismic hazard analysis. In the present study, a time-dependent recurrence model is proposed to consider a series of large earthquakes that occurs in clusters. The model is flexible enough to better reflect the quasi-periodic behavior of large earthquakes with long-term clustering, which can be used in time-dependent probabilistic seismic hazard analysis with engineering purposes. In this model, the time-dependent hazard results are estimated by a hazard function which comprises three parts. A decreasing hazard of last large earthquake cluster and an increasing hazard of the next large earthquake cluster, along with a constant hazard of random occurrence of small-to-moderate earthquakes. In the final part of the paper, the time-dependent seismic hazard of the New Madrid Seismic Zone at different time intervals has been calculated for illustrative purpose.

A novel approach to construct site-dependent ground-motion prediction equations (GMPEs) was implemented to build a uniform hazard response spectra (UHRS) using a reservoir site in southern Taiwan as an example. This innovation minimizes the GMPE uncertainties to avoid over-estimations in a probabilistic seismic hazard analysis (PSHA). We selected ground-motion records with similar site conditions from the stations within 70 km of the studied site to build a site-dependent ground-motion prediction model for crustal earthquakes in the region. This site-dependent GMPE set obtains a significantly smaller scattering of residuals compared with the existing regional GMPEs from the Taiwan data set. This indicates that a set of site-dependent GMPEs could provide a more accurate and reasonable prediction of ground-motion levels for the design or safety evaluation of important infrastructures.

This paper reviews the evolution of a sequence of seismological models developed and implemented as part of a workflow for Probabilistic Seismic Hazard and Risk Assessment of the seismicity induced by gas production from the Groningen gas field. These are semi-empirical statistical geomechanical models derived from observations of production-induced seismicity, reservoir compaction and structure of the field itself. Initial versions of the seismological model were based on a characterisation of the seismicity in terms of its moment budget. Subsequent versions of the model were formulated in terms of seismic event rates, this change being driven in part by the reduction in variability of the model forecasts in this domain. Our approach makes use of the Epidemic Type After Shock model (ETAS) to characterise spatial and temporal clustering of earthquakes and has been extended to also incorporate the concentration of moment release on pre-existing faults and other reservoir topographic structures.