Broadband Seismograms Research Articles

The fk-based direct exact stiffness matrix method for broadband seismogram synthesis of a multi-scale crustal structure due to finite fault kinematic sources

The fk-based direct exact stiffness matrix method for broadband seismogram synthesis of a multi-scale crustal structure due to finite fault kinematic sources

Upper Mantle Velocity Structure Beneath the Yarlung–Tsangpo Suture Revealed by Teleseismic P-Wave Tomography

We applied teleseismic tomography to investigate the 3D P-wave velocity (Vp) structure of the crust and upper mantle at depths of 50–400 km beneath the Yarlung–Tsangpo suture (YTS), by using 6164 P-wave relative travel-time residuals collected from 495 teleseismic events recorded at 20 three-component broadband seismograms. A modified multi-channel cross-correlation method was adopted to automatically calculate the relative arrival-time residuals of all teleseismic events, which significantly improved the efficiency and precision of the arrival-time data collection. Our results show that alternating low- and high-Vp anomalies are visible beneath the Himalayan and Lhasa blocks across the YTS, indicating that strong lateral heterogeneities exist beneath the study region. A significant high-Vp zone is visible beneath the southern edge of the Lhasa block at 50–100 km depths close to the YTS, which might indicate the rigid Tibetan lithosphere basement. There exists a prominent low-Vp zone beneath the Himalayan block to the south of the YTS extending to ~150 km depth, which might be associated with the fragmentation of the underthrusting Indian continental lithosphere (ICL) and induce localized upwelling of asthenospheric materials from the upper mantle. In addition, significant low-Vp anomalies were observed beneath the Yadong–Gulu rift and the Cona–Sangri rift extending to ~300 km depth, indicating that the tearing of the subducted ICL might provide pathways for the localized asthenospheric materials upwelling, which contributes to the widespread distribution of north–south trending rifts and geothermal activities in southern Tibet.

Investigation of Site Amplification and Attenuation Effects in the Changjiang Delta

Abstract Site response in the Changjiang Delta in eastern China was studied using Lg Fourier amplitude spectra. We used broadband seismograms recorded at 70 stations from 62 earthquakes with magnitude (Ms) varying from 3.5 to 5.0 during 2009–2021. The crustal quality factor Q and site response in the Changjiang Delta were obtained simultaneously from regression of Lg Fourier acceleration amplitude versus frequency. The κ0 of each individual station was subsequently calculated from a regression of the high-frequency site terms versus frequency. The site terms exhibit obvious dependence on sediment thickness in the Changjiang Delta. The site amplification factor reaches ∼7–10 for stations overlying sediments ∼8–9 km thick in the northern Jiangsu basin. The site terms were found to behave consistently as a function of sediment thickness over the frequencies of 0.56–24.86 Hz. Site amplification shows a positive correlation with sediment thickness at lower frequencies (<7.26 Hz) but transitions to a negative correlation as frequency increases to 12.86 Hz and higher. Linear functions versus sediment thickness were used to model the site response terms at individual frequencies. We also showed that site terms calculated using the κ0 model as a function of sediment thickness fit the site response terms well at frequencies higher than 9.66 Hz. Results of this study can be incorporated in ground-motion prediction models for the Changjiang Delta. In addition, the site response estimates determined here can be used to reduce bias due to site effects in studies of earthquake source parameters.

Near-source ground motion estimation for assessing the seismic hazard of critical facilities in central Italy

We apply the Probabilistic Seismic Hazard Analysis (PSHA) and compute Physics-Based Simulations (PBS) of ground motion for three dams in the Campotosto area (Central Italy). The dams, which confine an artificial water reservoir feeding hydroelectric power plants, are located in an active seismic zone between the areas that experienced the 2009 L’Aquila and 2016–2017 Central Italy seismic sequences. The probabilistic disaggregation estimated for a return period of 2475 years, corresponding to the collapse limit state for critical facilities, indicates that the most dangerous fault is associated with a maximum magnitude of 6.75 ± 0.25 within a distance of 10 km. This fault is used in PBS to emulate the Maximum Credible Earthquake scenario. To capture the ground motion variability, we input a pseudo-dynamic source model to encompass spatial and temporal variations in the slip, rise time and rupture propagation, heavily affecting the near-source ground motion. Indeed, the ground motion above the rupture volume is mainly influenced by the epistemic uncertainties of rupture nucleation and slip distribution. The computed broadband seismograms are consistent with the near-source shaking recorded during the 2016 MW 6.6 Norcia earthquake and constrain the upper bound of the simulated ground motion at specific sites. Our modelling reinforces the importance of considering vertical ground motion near the source in seismic design. It could reach shaking values comparable to or larger than those of the horizontal components. This approach can be applied in other areas with high seismic hazard to evaluate the seismic safety of existing critical facilities.

Near-fault broadband seismograms synthesis in a stratified transversely isotropic half-space using a semi-analytical frequency-wavenumber method

Anisotropic behavior occurs in natural soils and rocks, owing to sedimentation and weathering. However, the semi-analytical methods of near-fault broadband seismogram synthesis in the anisotropic medium are still rare. The physical-based frequency-wavenumber method was constructed for broadband seismogram synthesis due to a kinematic hybrid finite-fault source in a stratified transversely isotropic (TI) half-space for the first time. As the crucial foundation of the seismogram synthesis, the Green's function is established by the dynamic stiffness matrix method, effectively and stably modeling the broadband seismic waves propagation in a TI crust. A hybrid method of the kinematic fault source is applied to reasonably combine the low-frequency deterministic parts with the high-frequency stochastic parts, achieving a broadband waves excitation. The theoretical formulations are well-programmed by the FORTRAN code with OpenMP-based parallel technology to implement the fast solution. Compared with the two published results, the correctness of the methodology is well verified. A numerical example of an Mw7.0 scenario strike-slip earthquake is simulated to discuss the difference between TI and isotropic medium in the aspects: (i) attenuation characteristics near the fault area, (ii) rupture directional effect and velocity pulse, and (iii) response spectrum characteristics.

Testing Site Amplification Curves in Hybrid Broadband Ground Motion Simulations of M6.0, 24 August 2016 Amatrice Earthquake, Italy

This research focuses on predicting and assessing earthquake impact due to future scenarios regarding the ground motion seismic hazard by accounting mainly for site effect in the Central Apennines. To this end, we produced synthetic broadband seismograms by adopting a hybrid simulation technique for the Mw6.0 Amatrice earthquake, Central Italy, on 24 August 2016, accounting for site conditions by means of amplification curves, computed with different approaches. Simulations were validated by comparing with data recorded at 57 strong-motion stations, the majority installed in urban areas. This station sample was selected among stations recording the Amatrice earthquake within an epicentral distance of 150 km and potentially prone to experience site amplification effects because of lying in particular site conditions (sedimentary basins, topographic irregularities, and fault zones). The evaluation of amplification curves best suited to describe local effects is of great importance because many towns and villages in central Italy are built in very different geomorphological conditions, from valleys and sedimentary basins to topographies. In order to well reproduce observed ground motions, we accounted for the site amplification effect by testing various generic and empirical amplification curves such as horizontal-to-vertical spectral ratios (calculated from Fourier spectra using both earthquake, HVSR, and ambient noise, HVNSR, recordings) and those derived from the generalized inversion technique (GIT). The site amplifications emanated from GIT improve the match between observed and simulated data, especially in the case of stations installed in sedimentary basins, where the empirical amplification curve effectively reproduces spectral peaks. On the contrary, the worst performances are for the spectral ratios between components, even compared to the generic site amplification, although the latter ignores the strong bedrock/soil seismic impedance contrasts. At sites on topography, we did not observe any systematic behavior, the use of empirical curves ameliorating the fit only in a small percentage of cases. These results may provide a valuable framework for developing ground motion models for earthquake seismic hazard assessment and risk mitigation, especially in urban areas located in the seismically active central Italy region.

Source parameters inversion of the global large earthquakes using 3-D SEM Green's functions: strain Green's function calculation and validation

SUMMARYPrecisely constraining the source parameters of large earthquakes is one of the primary objectives of seismology. However, the quality of the results relies on the quality of synthetic earth response. Although earth structure is laterally heterogeneous, particularly at shallow depth, most earthquake source studies at the global scale rely on the Green's functions calculated with radially symmetric (1-D) earth structure. To avoid the impact of inaccurate Green's functions, these conventional source studies use a limited set of seismic phases, such as long-period seismic waves, broad-band P and S waves in teleseismic distances (30° < ∆ < 90°), and strong ground motion records at close-fault stations. The enriched information embedded in the broad-band seismograms recorded by global and regional networks is largely ignored, limiting the spatiotemporal resolution. Here we calculate 3-D strain Green's functions at 30 GSN stations for source regions of 9 selected global earthquakes and one earthquake-prone area (California), with frequency up to 67 mHz (15 s), using SPECFEM3D_GLOBE and the reciprocity theorem. The 3-D SEM mesh model is composed of mantle model S40RTS, crustal model CRUST2.0 and surface topography ETOPO2. We surround each target event with grids in horizontal spacing of 5 km and vertical spacing of 2.0–3.0 km, allowing us to investigate not only the main shock but also the background seismicity. In total, the response at over 210 000 source points is calculated in simulation. The number of earthquakes, including different focal mechanisms, centroid depth range and tectonic background, could further increase without additional computational cost if they were properly selected to avoid overloading individual CPUs. The storage requirement can be reduced by two orders of magnitude if the output strain Green's functions are stored for periods over 15 s. We quantitatively evaluate the quality of these 3-D synthetic seismograms, which are frequency and phase dependent, for each source region using nearby aftershocks, before using them to constrain the focal mechanisms and slip distribution. Case studies show that using a 3-D earth model significantly improves the waveform similarity, agreement in amplitude and arrival time of seismic phases with the observations. The limitations of current 3-D models are still notable, dependent on seismic phases and frequency range. The 3-D synthetic seismograms cannot well match the high frequency (>40 mHz) S wave and (>20 mHz) Rayleigh wave yet. Though the mean time-shifts are close to zero, the standard deviations are notable. Careful calibration using the records of nearby better located earthquakes is still recommended to take full advantage of better waveform similarity due to the use of 3-D models. Our results indicate that it is now feasible to systematically study global large earthquakes using full 3-D earth response in a global scale.

The 2021 South Sandwich Island Mw 8.2 Earthquake: A Slow Event Sandwiched Between Regular Ruptures

AbstractWe determined the rupture sequence of the 12 August 2021 Mw 8.2 South Sandwich Island earthquake which appears to be a complex sequence in both time and space. Notable tsunamis were recorded by tide gauges at global distances. Given the complexity of this event, we conducted a multiple subevent inversion on broadband seismograms, to resolve its complex variations of fault geometry, location, depth, and temporal characteristics. We found that the rupture initiated as a regular deep thrust earthquake; it then ruptured shallower and triggered a silent and dominantly slow subevent extending ∼200 km to the south, and ended with two other regular subevents. The total duration is ∼260 s, unusually long for an Mw 8.2 event. Our result is qualitatively consistent with other moment tensor solutions and the deviant mB–Mw and MS–Mw relations, and provides a more quantitative space‐temporal pattern of this unusual sequence.

Seismic Source Processes of 25 Earthquakes (Mw>5) in the Gulf of California

ABSTRACT The Gulf of California (GoC) is a complex tectonic boundary that has been instrumented in the past several decades to record broadband seismograms. This volume of data has allowed us to study several source parameters systematically. Before, only a few source parameters of earthquakes greater than magnitude five had been studied in the GoC area. We re-examined the focal mechanisms of several earthquakes in the southern GoC that occurred over the last 20 yr using local–regional distance broadband seismograms. These focal mechanisms were then used as input data to retrieve the time–space history of the rupture for each earthquake. This work contributes to the study of 25 rupture-process models computed with the method proposed by Yagi et al. (1999). To investigate more about the nature of the seismicity in the GoC, we also calculated the non-double-couple component of moment tensors for 45 earthquakes. Previous studies (e.g., Ortega et al., 2013, 2016) have shown that non-double-couple components from moment tensors in this region are associated with complex faulting, suggesting that oblique faults or several parallel faults are interacting simultaneously. Our results show that, at least for moderate earthquakes (5 < M < 6), rupture processes in the GoC show a complex interaction between fault systems. It is revealed on the important contribution of non-double-couple component obtained in the full moment tensor analysis.

A Seismological and Geophysical Database of Kos Island: 10 Years of Digital Records

A ten-year complete seismological database is evaluated in the present work. These data are provided for events with magnitude ML ≥ 1.0, which have occurred in Eastern Greece and, more specifically, in Kos Island. Several selection criteria are applied and, hence, a catalog of the seismological records is compiled. Detailed big data (seismological, tide gauge, geodetic stations, accelerometers, etc.) of this region are used and processed in this work. The database consists of approximately 35,000 three-component broadband seismograms from 1198 digitally recorded events. It covers the last ten years of measurements, including records from the 20 July 2017, ML 6.2, Kos Island, Greece event. The seismological communities can either use this database to conduct new research or improve already existing seismic hazard studies in the region.

On the inverse problem of reflection Green’s function seismograms over stratified acoustic media: Estimation of background models

A theory is presented for estimating the background velocity and density of an acoustic stratified medium by iterative least-squares waveform inversion in the frequency-horizontal slowness domain of low-frequency precritical reflection incidence seismograms of time length [Formula: see text]. The initial model is constant. The prerequisites for the method are that the reflection seismograms should be Green’s function seismograms and that the fundamental frequency component [Formula: see text] is present. Then, the gradients of the objective function provide the low-wavenumber trend of the medium. A matrix formulation for the model update is expressed mathematically by the classic seismogram residual, Jacobian, gradient, and Hessian in the Levenberg-Marquardt approximation. The first iteration, which is equal to a constant-parameter migration inversion (CPMI), is thoroughly analyzed, and expressions for band-limited gradients and block Hessians are found. For primary precritical reflection incidence seismograms of infinite bandwidth, it is shown theoretically that the partial gradients in the CPMI model become a reflection strength-weighted sum of shifted discrete sign functions, typical of step or staircase functions, which provide interface locations in Born depth and amplitudes that can be mapped to velocity and density information. For frequency-band-limited primary reflection seismograms, the partial gradients become a reflection strength-weighted sum of wavenumber-band-limited discrete sign functions. When the fundamental frequency component in the seismograms is present, the band-limited discrete sign functions are oscillatory but keep the information of the step function characteristic of the partial gradient. When the fundamental frequency component in the seismograms is absent, the band-limited discrete sign functions keep information of where the steps are located but lose the information of the amplitudes of the steps. The Hessian elements are nonstandard with the Hessian modeled over a broader frequency range than the frequencies of the observed low-frequency seismogram to avoid it becoming close to singular. The main mathematical findings are illustrated by a simple model and seismograms, for which the background models are found after two iterations. For the sake of completeness, the background models are classically used as initial models in a Levenberg-Marquardt least-squares inversion scheme to estimate the layer velocities and densities from broadband seismograms.

Focal mechanism inversion of the 2018 MW7.1 Anchorage earthquake based on high-rate GPS observation

The MW7.1 Anchorage earthquake is the most destructive earthquake since the 1964 MW9.2 great Alaska earthquake in the United States. In this study, high-rate GPS data and near-field broadband seismograms are used in separate and joint inversions by the generalized Cut-and-Paste (gCAP) method to estimate the focal mechanism. In order to investigate the influence of crustal velocity structure on the focal mechanism inversion results, two velocity models (Crust1.0 and Alaska Earthquake Center (AEC)) are used for detailed comparison and analysis. The results show that: (1) The two nodal planes of the optimal double-couple solution are nearly north-south striking, with dip angles of about 30° and 60°respectively, and the centroid focal depth is 54–55 km, which is an intraplate normal fault event. (2) The inversion results for the two types of data and the two velocity models are consistent with some previous studies, which indicates that the results are stable and reliable. The more accurate velocity structure model is helpful for focal mechanism inversion of the complex earthquake. (3) The inclusion of high-rate GPS data in joint inversion provides a more effective constraint on centroid depth.

Hybrid broadband strong-motion simulation to investigate the near-source characteristics of the M6.5, 30 October 2016 Norcia, Italy earthquake

During the 2016–2017 Central Italy earthquake sequence, a series of moderate to large earthquakes M > 5 occurred near the Amatrice and Norcia towns. These events are recorded on a dense seismic network, providing relevant observational evidence of complex earthquakes in time and space. In this work, we used this substantial data set to study the ground-motion characteristics of the Norcia earthquake M6.5 on October 30, 2016, through a broadband ground-motion simulation. Three-component broadband seismograms are generated to cover the entire frequency band of engineering interest. Low and high frequencies are computed considering the heterogeneous slip rupture model of Scognamiglio et al. (2018) [1]. High frequencies are calculated using a stochastic approach including P, SV, and SH waves, while low frequencies are obtained through a forward simulation of the kinematic model at the various stations. To predict earthquake-induced ground-motions in the area, we adopted region-specific attenuation and source scaling parameters derived by Malagnini et al. (2011) [2]. Ground-motion parameters, including peak ground acceleration (PGA), peak ground velocity (PGV) and spectral amplitudes, are calculated at the selected sites adopting physics-based parameters to understand better the earthquake fault rupture, the wave propagation, and their impacts on the seismic hazard assessment in the region. We showed that combining the fault rupture history over the entire frequency spectrum of engineering interest, the attenuation characteristics of the seismic wave propagation and the properly defined site responses can improve the prediction of ground-motions and time histories, especially in near seismic sources.

Rupture directivity and fault plane determination from analysis of rupture duration using broadband P-waves of the September 2018 M w 7.5 Palu, Indonesia earthquake

The aim of this study is to determine earthquake rupture directivity and fault plane of the 28 September 2018 M w 7.5 event in Palu, Indonesia. The data used in the current study were all acquired from broadband seismograms recorded by a network of 56 seismic stations. The method included analysis of rupture duration, where T dur was calculated using the vertical component of P-waveforms downloaded from IRIS site at http://ds.iris.edu/wilber3/find_event. For all stations involved in the T dur calculations, the largest distance between the epicenter and the corresponding station was estimated to be 40° from the azimuth station. The direction of the rupture indicated fault plane geometry of the source. The rupture directivity was found to be at azimuth 179°, in good agreement with the Palu Koro Fault azimuth. The finding implies that the direction of the earthquake rupture passed below the sea surface of Palu Bay, which could be the main source of the tsunami.

Complex dynamics of repeating and river-blocking landslides in Jiangda during 2018

Two large landslides successively blocked the Jinsha River at the same location in Jiangda Village on October 10 and November 3, 2018, respectively. The dynamic processes and possible interactions of the two landslides need to be studied to better understand the physical processes involved, and to provide information on future disaster mitigation. We investigated their force histories and sliding directions by inverting regional broadband seismograms. The scale of the October landslide was approximately three times that of the November event. The October event revealed a particularly strong deceleration force, which may have been caused by a collision between the sliding mass and ground surface. In contrast, the November event had a relatively weaker deceleration force, indicating that it may have been gradually stopped by the landslide dam formed during the October landslide. The sliding directions of the two landslides differed significantly in terms of both horizontal and vertical directions, indicating a change in their sliding surfaces. We conclude that unconsolidated materials at the top of the October landslide continued sliding along a curved slope during the November event. From our seismic models of dynamic processes, both the October landslide and local background may have affected and even changed a subsequent landslide’s mechanism.

Ensemble ShakeMaps for Magnitude 9 Earthquakes on the Cascadia Subduction Zone

AbstractWe develop ensemble ShakeMaps for various magnitude 9 (M 9) earthquakes on the Cascadia megathrust. Ground-shaking estimates are based on 30 M 9 Cascadia earthquake scenarios, which were selected using a logic-tree approach that varied the hypocenter location, down-dip rupture limit, slip distribution, and location of strong-motion-generating subevents. In a previous work, Frankel et al. (2018) used a hybrid approach (i.e., 3D deterministic simulations for frequencies <1 Hz and stochastic synthetics for frequencies >1 Hz) and uniform site amplification factors to create broadband seismograms from this set of 30 earthquake scenarios. Here, we expand on this work by computing site-specific amplification factors for the Pacific Northwest and applying these factors to the ground-motion estimates derived from Frankel et al. (2018). In addition, we use empirical ground-motion models (GMMs) to expand the ground-shaking estimates beyond the original model extent of Frankel et al. (2018) to cover all of Washington State, Oregon, northern California, and southern British Columbia to facilitate the use of these ensemble ShakeMaps in region-wide risk assessments and scenario planning exercises. Using this updated set of 30 M 9 Cascadia earthquake scenarios, we present ensemble ShakeMaps for the median, 2nd, 16th, 84th, and 98th percentile ground-motion intensity measures. Whereas traditional scenario ShakeMaps are based on a single hypothetical earthquake rupture, our ensemble ShakeMaps take advantage of a logic-tree approach to estimating ground motions from multiple earthquake rupture scenarios. In addition, 3D earthquake simulations capture important features such as strong ground-motion amplification in the Pacific Northwest’s sedimentary basins, which are not well represented in the empirical GMMs that compose traditional scenario ShakeMaps. Overall, our results highlight the importance of strong-motion-generating subevents for coastal sites, as well as the amplification of long-period ground shaking in deep sedimentary basins, compared with previous scenario ShakeMaps for Cascadia.

Broad-band ground-motion simulation of 2016 Amatrice earthquake, Central Italy

SUMMARYOn 24 August 2016 at 01:36 UTC a ML6.0 earthquake struck several villages in central Italy, among which Accumoli, Amatrice and Arquata del Tronto. The earthquake was recorded by about 350 seismic stations, causing 299 fatalities and damage with macroseismic intensities up to 11. The maximum acceleration was observed at Amatrice station (AMT) reaching 916 cm s–2 on E–W component, with epicentral distance of 15 km and Joyner and Boore distance to the fault surface (RJB) of less than a kilometre. Motivated by the high levels of observed ground motion and damage, we generate broad-band seismograms for engineering purposes by adopting a hybrid method. To infer the low frequency seismograms, we considered the kinematic slip model by Tinti et al . The high frequency seismograms were produced using a stochastic finite-fault model approach based on dynamic corner-frequency. Broad-band synthetic time-series were therefore obtained by merging the low and high frequency seismograms. Simulated hybrid ground motions were compared both with the observed ground motions and the ground-motion prediction equations (GMPEs), to explore their performance and to retrieve the region-specific parameters endorsed for the simulations. In the near-fault area we observed that hybrid simulations have a higher capability to detect near source effects and to reproduce the source complexity than the use of GMPEs. Indeed, the general good consistency found between synthetic and observed ground motion (both in the time and frequency domain), suggests that the use of regional-specific source scaling and attenuation parameters together with the source complexity in hybrid simulations improves ground motion estimations. To include the site effect in stochastic simulations at selected stations, we tested the use of amplification curves derived from HVRSs (horizontal-to-vertical response spectra) and from HVSRs (horizontal-to-vertical spectral ratios) rather than the use of generic curves according to NTC18 Italian seismic design code. We generally found a further reduction of residuals between observed and simulated both in terms of time histories and spectra.

Simulation of Broadband Ground Motion by Superposing High-Frequency Empirical Green’s Function Synthetics on Low-Frequency Spectral-Element Synthetics

Seismic wave-propagation simulations are limited in their frequency content by two main factors: (1) the resolution of the seismic wave-speed structure of the region in which the seismic waves are propagated through; and (2) the extent of our understanding of the rupture process, mainly on the short length scales. For this reason, high-frequency content in the ground motion must be simulated through other means. Toward this end, we adopt a variant of the classical empirical Green’s function (EGF) approach of summing, with suitable time shift, recorded seismograms from small earthquakes in the past to generate high-frequency seismograms (0.5–5.0 Hz) for engineering applications. We superimpose these seismograms on low-frequency seismograms, computed from kinematic source models using the spectral element method, to produce broadband seismograms. The non-uniform time- shift scheme used in this work alleviates the over-estimation of high-frequency content of the ground motions observed. We validate the methodology by simulating broadband motions from the 1999 Hector Mine and the 2006 Parkfield earthquakes and comparing them against recorded seismograms.

High-Accuracy Discrimination of Blasts and Earthquakes Using Neural Networks With Multiwindow Spectral Data

AbstractThe manual separation of natural earthquakes from mine blasts in data sets recorded by local or regional seismic networks can be a labor-intensive process. An artificial neural network (ANN) applied to automate discriminating earthquakes from quarry and mining blasts in eastern Kentucky suggests that the analyst effort in this task can be significantly reduced. Based on a dataset of 152 local and regional earthquake and 4192 blast recordings over a three-year period in and around eastern Kentucky, ANNs of different configurations were trained and tested on amplitude spectra parameters. The parameters were extracted from different time windows of three-component broadband seismograms to learn the general characteristics of analyst-classified regional earthquake and blast signals.There was little variation in the accuracies and precisions of various models and ANN configurations. The best result used a network with two hidden layers of 256 neurons, trained on an input set of 132 spectral amplitudes and extracted from the P-wave time window and three overlapping time windows from the global maximum amplitude on all three components through the coda. For this configuration and input feature set, 97% of all recordings were accurately classified by our trained model. Furthermore, 96.7% of earthquakes in our data set were correctly classified with mean-event probabilities greater than 0.7. Almost all blasts (98.2%) were correctly classified by mean-event probabilities of at least 0.7. Our technique should greatly reduce the time required for manual inspection of blast recordings. Additionally, our technique circumvents the need for an analyst, or automatic locator, to locate the event ahead of time, a task that is difficult due to the emergent nature of P-wave arrivals induced by delay-fire mine blasts.

Source parameters for small-moderate earthquakes in Marmara Region (Turkey)

The main aim of this study is to investigate the self-relation and self-similarity of earthquakes in and around the Marmara Sea (NW Turkey) by using these obtained source parameters. With this purpose, spectral source parameters for 77 small to moderate earthquake (3.5 ≤ ML ≤ 5.2) that occurred during 2004–2018 have been determined from P and S wave spectra according to Brune’s source model by using regional broadband seismograms. The average ratio of P/S wave corner frequency is found to be 1.3 that suggesting higher corner frequency for P wave. The static stress drops range from 0.1 and 136 MPa with a median value of 9.8 MPa (98 bars). The high stress drops for these events can indicate high frictional strength and low strain-rate of the faults. Similarly, the low values of the stress drop can indicate a general weakness of the faults in the study area. There is no clear dependence between the seismic moment and the static stress drop in the analyzed events but some events which have lower seismic moment value also have lower stress drop. Obtaining results indicated the corner frequency decreases with increasing of the seismic moment. Also, a slight depth dependence of the corner frequency has been observed for the analyzed events. Shallower events have larger corner frequency value than deeper events. Also, a clear depth dependence of the stress drop values has not been observed. However, the depth dependence of the seismic moment is seen more clearly. Our results indicated that the deeper events have larger seismic moment values in the study area. In spite of scattering in small events, a linear relationship between local magnitude (ML) and moment magnitude (MW) could be obtained as MW = 1.4261(± 0.31) + 0.6399(± 0.08)ML from P waves spectra and MW = 0.0136(± 0.21) + 0.9883(± 0.05)ML from S wave spectra and calculated MW values are consistent with waveform inversion (centroid moment tensor — CMT) results. These relationships may be useful for seismic hazard studies in the study area.