Description: Physically based ground-motion prediction equations for soil and rock sites in the Zagros region have been developed based on the specific barrier model (SBM) used within the context of the stochastic model. Instead of direct time-domain simulation, random vibration theory was used to estimate measures of peak motion in terms of the pseudospectral velocity of anelastic harmonic oscillator with 5% viscous damping. To avoid the uncertainties, calibration of the source model uses a database of carefully selected strong motion data without ambiguity about the site condition. Therefore, only rock sites are selected for determining source parameters. Also, to avoid any inconsistencies caused by magnitude conversion formulas, we restricted the dataset only to events with available moment magnitudes. Regression analysis is performed using the random effects model that considers both interevent and intraevent variabilities to effectively deal with the problem of an unequal number of records from different earthquakes. No sign of self-similarity breakdown is observed between the source radius and its seismic moment. The local and global stress drops derived for the Zagros region (39 and 116 bars, respectively) are more consistent with the values obtained by other authors for an interplate regime than the values for an intraplate region. However, from the viewpoint of source heterogeneity (as the ratio of the stress drops is an indicator of the complexity of the source and heterogeneity of slip on the fault) the Zagros events, which have a stress-drop ratio of about three are more homogeneous than other interplate events. Stochastic simulations are then implemented to predict peak ground motion and response spectra parameters for rock and soil site conditions. Online Material: Catalog tables of earthquakes and attenuation tables.

Description: NW Iran is a region of active deformation in the Eurasia–Arabia collision zone. This high strain field has caused intensive faulting accompanied by several major ( M 〉 6.5) earthquakes as it is evident from historical records. Whereas seismic data (i.e. instrumental and historical catalogues) are either short, or inaccurate and inhomogeneous, physics-based long-term simulations are beneficial to better assess seismic hazard. In this study, a deterministic seismicity model, which consists of major active faults, is first constructed, and used to generate a synthetic catalogue of large-magnitude ( M 〉 5.5) earthquakes. The frequency–magnitude distribution of the synthetic earthquake catalogue, which is based on the physical characteristic and slip rate of the mapped faults, is consistent with the empirical distribution evaluated using record of instrumental and historical events. The obtained results are also in accordance with palaeoseismic studies and other independent kinematic deformation models of the Iranian Plateau. Using the synthetic catalogue, characteristic magnitude for all 16 active faults in the study area is determined. Magnitude and epicentre of these earthquakes are comparable with the historical records. Large earthquake recurrence times and their variations are evaluated, either for an individual fault or for the region as a whole. Goodness-of-fitness tests revealed that recurrence times can be well described by the Weibull distribution. Time-dependent conditional probabilities for large earthquakes in the study area are also estimated for different time intervals. The resulting synthetic catalogue can be utilized as a useful data set for hazard and risk assessment instead of short, incomplete and inhomogeneous available catalogues.

Description: In the present study, seismicity parameters (seismic activity, β -value, and maximum regional magnitude) of the Iranian plateau are computed for sites equally distributed all over the country on a grid of 1 x 1 decimal degree. The most complete available catalog including historical and instrumental earthquakes of the plateau from 734 B.C. to A.D. 2011 is first prepared from numerous resources. Earthquakes within a 200-km buffer surrounding each grid point are selected, and historical and instrumental parts are more accurately specified, based on the completeness test of Stepp (1972) . The instrumental part is also analyzed to identify the various completeness intervals, and the minimum completeness magnitude for each interval is determined by the maximum curvature method. Next, the seismicity parameters are calculated from uncertain and incomplete data by the maximum likelihood procedure of Kijko and Sellevoll (1989 , 1992) and Kijko (2004) , after removing aftershocks and foreshocks. The spatial variation of seismicity parameters is illustrated in the form of contour maps over the whole Iranian plateau for the first time. The -value corresponding to the magnitudes M w  4 varies over the plateau from 0.7 to 23.4, and the highest -values are located in the Zagros and Alborz mountains, confirming that a large portion of the seismic deformation in the plateau is accommodated in these regions. The estimated β -values for all points lie between 1.3 and 2.9. This spatial distribution of β -values demonstrates the crustal heterogeneity over the plateau. Low β -values in northeast and east Iran also indicate high probability of large-earthquake occurrence. It is observed that the estimated M max is between M w  6.5 and 8, and hence most of that part of the country may experience large earthquakes. The obtained results are important for two aspects: (1) new seismotectonic models could be proposed by combining the presented maps with geological, geodetic, and tectonic data; and (2) these results, together with a detailed geological data describing potential seismogenic sources, can be used directly to evaluate ground-motion hazard for engineering design and generate seismic-hazard maps. An approximate and gross estimate of peak ground acceleration (PGA) level at bedrock, for 2% and 10% probabilities of exceedance in a 50-year period, is also computed using an area-source model characterized by the presented seismicity parameters. The regional distribution of the estimated PGA sheds a new light on facilitating risk management and allocating national resources for mitigating potential losses at a country-based level. Online Material: Table of completeness magnitude and seismicity parameters.

Description: The original aim of the present study was to test the efficiency of some selected ground-motion prediction equations (GMPEs) against small-to-moderate data recorded in the Iranian plateau. For this purpose, we applied three statistical tests including the Nash–Sutcliffe model efficiency coefficient ( Nash and Sutcliffe, 1970 ) and likelihood and log-likelihood methods ( Scherbaum et al. , 2004 , 2009 ) to assess performance of eight GMPEs against a comprehensive databank from small-to-moderate magnitude data in Iran. Two of the candidate models were selected from Japanese models, and the rest were chosen from regions with shallow crustal earthquakes including local models corresponding to the Iranian earthquakes, regional models established for the Middle East and Euro-Mediterranean regions, and global models (developed for shallow crustal environments). We analyzed a high-quality dataset composed of 937 ground-motion records from 296 Iranian earthquakes with moment magnitude ranging between 3.9 and 5.0 and epicentral distances of up to 100 km. In conclusion, Zafarani et al. (unpublished manuscript, 2015; see Data and Resources ) as a ground-motion model valid for a magnitude range as small as M  4.0 is the only model that shows good consistency with the recorded data over all frequencies. This emphasizes the importance of region-dependent anelastic attenuation and the necessity to develop more elaborate local GMPEs for small-to-moderate earthquakes, which are also very important from the viewpoint of epistemic uncertainty.

Description: In the first part of this study, a set of 87 ground-motion records, with closest distance to the rupture plane ( R rup ) less than 200 km and averaged shear-wave velocity over the top 30 m of the subsurface ( V S 30 ) between 175 and 1400 m/s, recorded during the 2012 Ahar–Varzaghan dual earthquakes ( M w 1  6.4, M w 2  6.3) were taken into account to examine the predictive capabilities of the Next Generation Attenuation (NGA) ground-motion prediction equations (GMPEs) via a set of comparative analyses and tests. The first-applied method to assess the performance of the NGA GMPEs is based on the intraevent residual analysis. The primary database (i.e., 87 records) was also used to develop an event-specific GMPE in the case of the Ahar–Varzaghan dual earthquakes by means of regression analyses. The derived event-specific GMPE has been compared with the NGA GMPEs for two different site conditions, that is, V S 30 〉375 m/s (rock site) and V S 30 〉375 m/s (soil site). The residual analysis results indicate that the NGA GMPEs perform better in predicting data recorded at rock sites compared to soil sites. For soil sites and at large periods ( T =2.0 s), the observed spectral accelerations are overpredicted by the NGA GMPEs. Furthermore, as the second part of this study, to select the most adequate GMPEs, 14 strong-motion records from the 1997 Ardebil earthquake ( M w  6.1) were added to the primary database. The implementation of the likelihood (LH) and log-likelihood methods, as modern LH-based ranking assessment techniques, as well as the Nash–Sutcliffe index reveal that the NGA GMPEs show good compatibility at short-medium periods ( T 〈1.0 s) on the basis of the data recorded during the 2012 Ahar–Varzaghan dual and 1997 Ardebil earthquakes (i.e., 101 records). However, in the long-period range, the dispersion in the data does not allow the authors to draw a comprehensible conclusion.

Description: Here, the generalized inversion of S -wave amplitude spectra is employed for deriving the site response and S -wave attenuation (quality factor) in the northwest region of Iran. A total of 279 strong-motion accelerograms recorded at distances below 200 km from 41 earthquakes with moment magnitudes ranging from M w  3.8 to 6.5 are used to determine the region-specific attenuation model. The site responses have been determined for all 54 stations individually. The bulk of strong-motion data (i.e., 184 records) comes from the 2012 Ahar–Varzaghan dual earthquakes in northwest Iran and their aftershocks. Also, the modified finite-fault method, which is able to model nonuniform stress distribution on the fault plane, is employed to investigate the coseismic stress parameter for the first and second Ahar–Varzaghan dual earthquakes by minimizing the difference between the synthesized and observed pseudospectral accelerations for 5% damping. Results denote a concentration of stress drops on the northwestern part of the first event’s fault and on the central and eastern part of the second fault. Stochastic simulations, using the calibrated stress distribution have been performed on a regular grid covering the study area for both rock- and soil-site conditions (in total, 651 points). The results show that the maximum of peak ground accelerations (PGAs) during the Ahar–Varzaghan dual earthquakes for the rock-site condition reached 557 cm/s/s, and the largest peak ground velocities (PGVs) were estimated around 76 cm/s. In general, comparisons between simulated data and empirical ground-motion prediction equations (GMPEs) show that the stochastic predictions (PGAs and PGVs) are higher than those predicted by the empirical GMPEs for distances less than 30 km and are smaller for distances more than 30 km. The simulated data are also consistent with the modified Mercalli intensity observations and show reasonable agreement. This can be considered as another validation of the method.

Description: The time-averaged shear-wave velocity to a depth of 30 m ( V S 30 ) is widely used in various geotechnical and seismic issues such as ground-motion prediction equations, determination of site class in building codes, and evaluation of site effects. However, in many cases the velocity is not measured as deep as 30 m. Therefore, obtaining a correlation between V S 30 and velocities at depths lower than 30 m ( V Sz ) seems to be necessary. The present article proposes a rigorous model selection approach to find the best model for estimation of V S 30 by V Sz based on 257 Iranian profiles. The candidate models include the Boore et al. (2011) and Chandler et al. (2005) relations, as well as two types of power-form correlation presented by the authors. The presented approach includes hypothesis testing based on goodness-of-fit measures of normalized residuals, the likelihood method of Scherbaum et al. (2004) , and finally the log-likelihood (LLH) criterion based on the information theory. The results show that the best model obtained by each of these methods is not necessarily the same. Furthermore, it is observed that while hypothesis tests are somehow inconsistent and noninformative, the LLH criterion presents a coherent method for ranking various candidate models and selecting the best one.

Description: The eastern part of the Iranian plateau is characterized by large and infrequent earthquakes with recurrence intervals of more than several hundred years. Given that previous observations and paleoseismological studies are insufficient for forecasting large earthquakes, we have developed a physics-based synthetic seismicity model for the fault system in eastern Iran using the Virtual Quake simulator. The model is independent of the seismic catalogs, however, it is tuned to match available earthquakes records. We show that the modeled seismicity rates and empirical frequency–magnitude distributions are consistent within the uncertainty of the empirical relations. Furthermore, our synthetic catalog agrees with previous paleoseismological investigations. From the resulting catalog, we can obtain the statistical distributions of recurrence times and waiting times for large earthquakes in the region as a whole and for the individual faults. In agreement with previous earthquake simulator studies, we find that the Poisson (time-independent) distribution best describes the recurrence times of large earthquakes in the region as a whole. Large earthquakes on the individual faults show quasi-periodic behavior, and for most faults can be well represented by the Weibull distribution. We present the corresponding time-dependent conditional probabilities for large earthquakes throughout the region and for a few selected individual faults. Long-term simulations indicate that waiting times for large earthquakes on specific faults are strongly dependent on the fault system configuration. Faults confined by other active faults, or consisting of several segments, host large earthquakes with more distributed recurrence times. Moreover, the obtained long-term synthetic seismic catalog shows that fault interactions clearly have a major effect on occurrence of large earthquakes, and hence should be taken into account for seismic-hazard assessment.

Description: The experience of past earthquakes has shown repeatedly that the characteristics of ground motions and resulting damages are highly dependent on the local condition of the site. It is clearly known that the underlying layers may amplify the earthquake excitation to some extent. This phenomenon is called the site effects. The site amplification depends on several factors such as stiffness, thickness, and density of the subsurface layer. In the present article, the 1D yet acceptable square-root-impedance method is used to estimate the frequency-dependent amplification function for various site classes in Iran. To achieve this goal, a stochastic method is applied to generate several thousand velocity profiles down to 8000 m. These profiles are randomly generated and constrained based on the features of 458 available shallow velocity profiles and crustal structure models across Iran. Then, the resulted amplification functions are multiplied by near-surface attenuation function to obtain the site response in the frequency domain. The comparison of the mean spectra of different site classes demonstrates that although the softer classes (the ones with lower V S 30 ) have higher amplifications at lower frequencies, the trend is completely reversed at higher ones. Finally, the generated profiles are employed to derive models for Z 1.0 (the depth at which the shear-wave velocity is equal to 1 km/s). This depth is very important, because it is included in the new generation of ground-motion prediction equations.

Description: On 12 November 2017, an earthquake with a moment magnitude of 7.3 struck the west of Iran near the Iraq border. This event was followed about 9 and 12 months later by two large aftershocks of magnitude 5.9 and 6.3, which together triggered intensive seismic activity known as the 2017–2019 Kermanshah sequence. In this study, we analyse this sequence regarding the potential to forecast the spatial aftershock distribution based on information about the main shock and its largest aftershocks. Recent studies showed that classical Coulomb failure stress (CFS) maps are outperformed by alternative scalar stress quantities, as well as a distance-slip probabilistic model (R) and deep neural networks (DNN). In particular, the R-model performed best. However, these test results were based on the receiver operating characteristic (ROC) metric, which is not well suited for imbalanced data sets such as aftershock distributions. Furthermore, the previous analyses also ignored the potential impact of large secondary earthquakes. For the complex Kermanshah sequence, we applied the same forecast models but used the more appropriate MCC-F1 metric for testing. Similar to previous studies, we also observe that the receiver independent stress scalars yield better forecasts than the classical CFS values relying on the specification of receiver mechanisms. However, detailed analysis based on the MCC-F1 metric revealed that the performance depends on the grid size, magnitude cut-off and test period. Increasing the magnitude cut-off and decreasing the grid size and period reduce the performance of all methods. Finally, we found that the performance of the best methods improves when the source information of large aftershocks is additionally considered, with stress-based models outperforming the R model. Our results highlight the importance of accounting for secondary stress changes in improving earthquake forecasts.