Real Seismograms Research Articles

Discrimination of doubled acoustic emission events using neural networks

In observatory seismology, the effective automatic processing of seismograms is a time-consuming task. A contemporary approach for seismogram processing is based on the Deep Neural Network formalism, which has been successfully applied in many fields. Here, we present a 4D network, based on U-net architecture, that simultaneously processes seismograms from an entire network. We also interpret Acoustic Emission data based on a laboratory loading experiment. The obtained data was a very good testing set, similar to real seismograms. Our Neural network is designed to detect multiple events. Input data are created by augmentation from previously interpreted single events. The advantage of the approach is that the positions of (multiple) events are exactly known, thus, the efficiency of detection can be evaluated. Even if the method reaches an average efficiency of only around 30% for the onset of individual tracks, average efficiency for the detection of double events was approximately 97% for a maximum target, with a prediction difference of 20 samples. Such is the main benefit of simultaneous network signal processing.

Poro-Acoustic Impedance (PAI) as a new and robust seismic inversion attribute for porosity prediction and reservoir characterization

For the purpose of reservoir modelling, precise porosity estimation is vital as it directly influences the storage capacity, fluid flow dynamics, and overall productivity of the reservoir. The computation of porosity is a key component of reservoir characterization. The Poro-Acoustic Impedance (PAI), a seismic inversion attribute, has proven to be effective for porosity estimation in hydrocarbon reservoirs. PAI, an extended version of Acoustic Impedance (AI), incorporates porosity information directly, enhancing its utility in forward modelling and seismic data inversion. In this study, offshore oil resources in Iran were examined, focusing on two components: sandstone and carbonate. The results of AI and PAI were compared, indicating that PAI is a suitable attribute for estimating porosity. The correlation between porosity and AI was −45%, while it was −74% with PAI. Moreover, the synthetic seismogram created using PAI aligns more closely with real seismograms. Porosity was estimated using both AI and PAI, with a 72% correlation between the porosity estimated using AI and the actual porosity. However, the correlation increased to 78% when using PAI. Furthermore, the porosity section was calculated using both AI and PAI, concluding that the PAI porosity section aligns more closely with the porosity log and provides a greater contrast in low porosity zones compared to AI. Given that porosity is incorporated into the PAI formula, the PAI porosity section and inversion results can be used as an indicator for evaluating the hydrocarbon capacity of the reservoir. The findings of this research suggest that PAI is an effective attribute for porosity estimation, bridging the gap between seismic data and porosity estimation, thereby enhancing our understanding and exploration of the reservoir.

Calculation of the stress-strain state of monolithic bridges on the action of real seismic impacts

To improve the transport infrastructure of the Republic of Uzbekistan, there is a tendency to build monolithic structures of bridges and overpasses. The article presents the calculation of a monolithic bridge on the 1083rd km of the M-39 highway passing through the city of Samarkand. The results of the monolithic bridge calculation on dynamic loads are presented, based on the records of two real seismograms of two dangerous earthquakes registered for a system with a spectral composition: the Gazli (Uzbekistan) and Manjil (Iran) earthquakes. The results of calculations of longitudinal and vertical displacements, and normal stress in the upper and lower parts of the span along the length of the bridge were analyzed. The calculations performed, show that the span structure and bridge supports have an overestimated margin of safety for an 8-point earthquake according to MSK-64. To ensure the guaranteed seismic safety of bridge structures, it is required to conduct design calculations based on a set of records of earthquake events that are close in dominant frequencies to the characteristics of the construction site. This provision must be included in the relevant regulatory documents.

Calculation of monolithic bridges taking into account seismic conditions of Republic of Uzbekistan

To improve the transport infrastructure of the Republic of Uzbekistan, the monolithic structure of bridges and overpasses began to be used. The article presents the calculation of a monolithic overpass on the 1083rd km of the M-39 highway passing through Samarkand. The results of the calculation of a monolithic overpass from dynamic load are presented based on the records of two real seismograms for two dangerous earthquakes for the system with spectral composition: Gazli (Uzbekistan) and Manjil (Iran). The calculations show that to ensure guaranteed seismic safety of transport structures, it is necessary to carry out design calculations based on sets of records of earthquakes that occurred, close in dominant frequencies to the characteristics of the construction site. Such a provision should be introduced into the relevant regulatory documents.

Improved seismic well tie by integrating variable-size window resampling with well-tie net

Accurate seismic well tie is essential for seismic inversion and reservoir characterization. The procedure of seismic well tie involves shifting, stretching and squeezing the synthetic seismogram computed from well logs to match the seismic traces at or near the borehole location. Numerous methods have been proposed for nonlinear alignment between synthetic and real seismograms. However, most well-tie methods are prone to over-stretching and the alignment result is sensitive to the chosen window size. To solve those problems, we propose a variable-size window resampling (VWR) algorithm and integrate with convolutional neural network (CNN) for automatic seismic well tie. Using VWR algorithm to reconstruct the waveforms in synthetic seismogram can simulate the variety of subsurface velocity. CNN can learn the characteristic of different waveforms and recognize the most correlated waveforms between synthetic and real seismograms for sequence alignment. We first use VWR algorithm to reconstruct a large number of synthetic seismograms for train set generation. We then build an CNN model that named well-tie net for training to learn the feature of different resampled synthetic seismograms. Finally, we use the well trained CNN model to segment the real seismogram and align with the synthetic seismogram for seismic well tie. We apply our method into the synthetic test and real seismic data with well logs and obtain high correlated seismic-well tie. We also compare with the conventional method dynamic time warping (DTW) to illustrate the effectiveness and robustness of our proposed method. Our proposed method can avoid the problem of over-stretching by using the variable-size window resampling algorithm and automatically tying the well to seismic trace using well-tie net. In addition, the train set for our method is generated automatically.

Early Forecast of Maximum Amplitude due to Aftershocks by Applying Extreme Value Statistics to a Single Continuous Seismogram

ABSTRACT To evaluate the exceedance probability of maximum amplitude (EPMA) due to aftershocks, I developed a forecasting scheme based on the extreme value statistics applied to a single continuous seismogram of early aftershocks. By combining the general laws of aftershock activity (Gutenberg–Richter and Omori–Utsu laws) and a ground-motion prediction equation (including source, path, and site factors), I verified that the interval maximum amplitude of a continuous seismogram of aftershocks follows the non-stationary Frechet distribution (NFD), which is one of the extreme value distributions. The parameters of NFD are written explicitly from the parameters commonly used in seismology. By optimizing the NFD parameters through the maximum-likelihood method and using the maximum-likelihood estimates and their covariance values, I derived the EPMA due to aftershocks based on the Bayesian approach. The performance of the EPMA was examined by Monte Carlo simulations and real seismograms. The numerically generated maximum amplitude was predicted well from the EPMA, which was evident even in the period of intense seismicity in which many waveforms overlap in a seismogram. This performance was also robust for real seismograms of aftershocks for the 2008 Iwate–Miyagi Nairiku, Japan, earthquake. The maximum amplitudes observed for four days were mostly within the 10% and 90% EPMA curves issued within 3 hr of the mainshock. The proposed method does not need to evaluate source, path, and site factors because these factors are included in the estimated NFD parameters. Given that the proposed method allows single-station processing, a seismic “network” is not required. Therefore, the proposed algorithm will be easily implementable in a seismic observation system installed at important facilities. Also, the NFD parameters estimated robustly in the early lapse times may provide important knowledge regarding early aftershocks.

Analyzing the Vibrations of the Critical Components in the Power Generating Unit of 200MW under Kinematic Loadings

An important problem that should be resolved during the power plant operation is related to the reliability control for all modes of operation including extreme external dynamic effects of a natural character, in particular the earthquakes. The vibrations of the elements of power equipment exposed to seismic forces occur due to kinematic perturbations that are transmitted from the ground to the bottom base plate. In this connection, the following requirements are set up to the seismic stability of the power equipment of power plants that include its strength, impermeability and its serviceability when exposed to seismic forces. The loading of the fasteners of the turbine generating unit of 200MW arranged on the base was analyzed in the case of exposure to seismic forces. Calculation methods and the software were used that were developed using the finite element method. The computational model of the dynamic system, in particular the turbine generating unit- bed and base consisted of arbitrary- oriented rods and lumped masses that are rigidly interconnected with each other or these can also be interconnected by linear and nonlinear elastically-damped elements. The rod finite elements with distributed parameters allow us to take into consideration all the types of deformations that occur during rod vibrations. The computational method allows us to take into consideration the structural peculiarities of the system and the peculiarities of its behavior during possible earthquakes. The kinematic perturbation that simulates the seismic action transmitted from the ground to the bottom bed plate is assigned by the accelerogram that includes acceleration in three mutually perpendicular directions. The computations were done for the synthesized seismogram of the earthquake of wind force 7 that results in the broadest spectrum of the loading in comparison with other real seismograms. The analysis of the computational data obtained for the seismic stability of the turbine generating unit of 200MW showed that the displacement of rotors towards the turbine generating unit axis relatively the cylinder bodies and the displacement of the upper bed plate relatively the bottom bed plate is lower than permissible values in the case of the earthquake of wind force 7.The keyed joints are the most loaded system assemblies though the strains that arise in them during the kinematic perturbations in question also fail to attain permissible values. The obtained research data are of practical importance and can be used for the assessment of the seismic stability of turbine generating units of a different power.

Mathematical modeling of the transmission pipeline under the influence of earthquake seismic waves

The stress-strain state of an underground transmission pipeline under the action of earthquake seismic waves is considered. The main objective of this study is the construction of a mathematical model with the analytical solution for lower estimation of the influence of earthquake seismic waves on the transmission pipeline. The spectra of the design seismic actions of the earthquake are determined on the basis of parameters of real seismograms of past earthquakes. The generalized functions of seismic impulses are constructed. The transmission pipeline is modeled by a thin bar. Equations of pipeline motion using the Laplace transformation in time are solved, stresses are determined. The numerical example is considered. The results of calculations on the theories of Kirchhoff and Timoshenko are compared. Graphs of stress changes depending on the earthquake’s strength in points are given. The limiting value of the earthquake magnitude resulting in large stresses and deformations is determined. As a result, the transmission pipeline is destroyed.

SH-waveform modeling of small local seismic events in Ladoga lake

A lot of seismic phenomena with small magnitude has been taking place in the Ladoga lake region (northwestern Russia). One of them is the earthquake which was recorded on the 31st of July, 2010 followed by the swarm of microearthquakes near Nikonovsky cape. There is only one station in the region that is why it is difficult to process the data. In this work, we propose a new approach to construct synthetic seismograms of the main event with minimum data. Also this approach applies us to calculate parameters of the source. The approximate magnitude of the main event is $$M_L=-0.8$$ . For swarm events, we have magnitude values from $$-\,1.0$$ to $$-\,2.6$$ . The data of swarm events (microearthquakes) is used to test the technique of constructing synthetic seismograms. The approach we propose is based on Green’s function method with using two temporal $$\delta$$ -sequences (the Lorentzian function and the continuous Dirichlet kernel). A comparison of synthetic and real seismograms shows that the Lorentzian function is better to describe small local events, and the continuous Dirichlet kernel is more appropriate for earthquakes. This approach can be generalized to various problems of different nature.

Behaviour and Calculation of Polymer Pipelines Under Real Earthquake Records

The effect of real records of several earthquakes on polymer pipeline vibrations are studied and analyzed in the paper. Longitudinal vibrations of straight end section of the underground pipeline are considered using the Kelvin-Voigt model. To solve the equation, the method of finite differences of the second order of accuracy is used. An algorithm has been compiled and on its basis a software product has been developed for the numerical calculation of an underground polymer pipeline under the effect of real seismograms recorded in different cities of the world. The stress-strain state (SSS) of polymer pipelines is determined under various boundary conditions.

Observation and explanation of spurious seismic signals emerging in teleseismic noise correlations

Abstract. Deep body waves have been reconstructed from seismic noise correlations in recent studies. The authors note their great potential for deep-Earth imaging. In addition to the expected physical seismic phases, some spurious arrivals having no correspondence in earthquake seismograms are observed from the noise correlations. The origins of the noise-derived body waves have not been well understood. Traditionally, the reconstruction of seismic phases from inter-receiver noise correlations is attributed to the interference between waves from noise sources in the stationary-phase regions. The interfering waves emanating from a stationary-phase location have a common ray path from the source to the first receiver. The correlation operator cancels the common path and extracts a signal corresponding to the inter-receiver ray path. In this study, with seismic noise records from two networks at teleseismic distance, we show that noise-derived spurious seismic signals without correspondence in real seismograms can arise from the interference between waves without a common ray path or common slowness. These noise-derived signals cannot be explained by traditional stationary-phase arguments. Numerical experiments reproduce the observed spurious signals. These signals still emerge for uniformly distributed noise sources, and thus are not caused by localized sources. We interpret the presence of the spurious signals with a less restrictive condition of quasi-stationary phase: providing the time delays between interfering waves from spatially distributed noise sources are close enough, the stack of correlation functions over the distributed sources can still be constructive as an effect of finite frequencies, and thereby noise-derived signals emerge from the source averaging.

Syntactic Pattern Recognition for Wavelet Clustering in Seismogram

In a seismogram, there exist many kinds of wavelets. The reflected wavelet from the gas sand zone has a different shape with other layers. Usually, the information of each wavelet is weak and unknown, and the unsupervised classification method is applied to the clustering of the wavelets. Using the shape structure of the wavelet, syntactic pattern recognition is applied to the clustering. The extracted wavelets can be represented as strings of symbols. Levenshtein distance is used to calculate the distance between the two strings. Bottom-up and top-down hierarchical clustering methods are used in the construction of the dendrogram. The top-down hierarchical clustering by the recursive method is proposed. A new pseudo F-statistics is proposed to decide the optimal number of clusters. From the experimental results in simulated and real seismograms, the wavelets on the gas sand zone can be detected successfully. It can improve the seismic interpretation.

Validated artificial neural networks in determining petrophysical properties: A case study from Colombia

We have applied instantaneous seismic attributes to a stacked P-wave reflected seismic section in the Tenerife field located in the Middle Magdalena Valley Basin in Colombia to estimate the volume of clay [Formula: see text] and the density [Formula: see text] at seismic scale. The well logs and the seismic attributes associated to the seismic trace closer to one of the available wells (Tenerife-2) is the information used to train some multilayered artificial neural networks (ANN). We perform data analysis via the gamma test, a mathematically nonparametric nonlinear smooth modeling tool, to choose the best input combination of seismic attributes to train ANNs to estimate [Formula: see text] and [Formula: see text]. Once the ANNs are trained, they are applied to predict these parameters along the seismic line. From the continuous estimations of [Formula: see text], we distinguish two facies: sands for [Formula: see text] and shales when [Formula: see text]. These estimations confirm the production of the Mugrosa C-Sands zone, and we draw the brown shale that correlates with the high-amplitude attributes and the yellow sand that correlates with the low-amplitude attributes. Using the well-log information for [Formula: see text] and the facies classification (also in the well log), two cubic polynomials that depend on time (or depth) are obtained, one for sands and the other for shales, to fit the [Formula: see text]. These two cubic polynomials and the facies classification obtained from the [Formula: see text] at the seismic scale enable us to estimate [Formula: see text] at the seismic scale. To validate the 2D [Formula: see text] and [Formula: see text] predicted data, a forward-modeling software (the Kennett reflectivity algorithm) is used. This model calculates synthetic seismograms that are compared with the real seismograms. This comparison indicates a small misfit that suggests that the [Formula: see text] and [Formula: see text] images are representing the reservoir description characteristics and the ANN method is accurate to map these parameters.

Physically based ground motion prediction and validation: a case study medium-size magnitude Marmara Sea earthquakes

The evaluation of realistic time histories for various locations around Marmara Region is aimed to provide reliable input for performance-based seismic design, hazard or risk management studies and developing new seismic standards. The applicability of empirical Green’s functions methodology and physics-based solution of earthquake rupture have been assessed in terms of modeling complex geologic structures. This paper has two main objectives. The first one is to simulate five medium-size magnitude earthquakes (M w ≈ 5.0) recorded in the Marmara Region. A series of synthetic ground motion waveforms for three components are evaluated with a ‘physics-based’ solution of earthquake rupture. The simulation methodology is based on the studies of Hutchings and Wu (J Geophys Res 95:1187–1214, 1990), Hutchings (Seismol Soc Am 81:88–121, 1991; Seismol Soc Am 84:1028–1050, 1994), Hutchings et al. (Geophys J Int 168:569–680, 2007), and Scognamiglio and Hutchings (Tectonophysics 476:145–158, 2009). For each earthquake, we calculate synthetic seismograms by using 500 different rupture scenarios that are generated by Monte Carlo method for a selection of parameters within a range based on prior knowledge of where the earthquakes will occur. The second objective is to validate synthetic seismograms with real seismograms. To improve the credibility of synthetic seismograms from an engineering point of view, the methodology presented by Anderson (in 13th world conference on earthquake engineering, Vancouver, 2003) is followed. This methodology proposes a similarity score based on averages of the quality of fit measuring ground motion characteristics and uses a suite of measurements. In order to compute goodness of fit, ten different ground motion parameters were compared on a scale from 0 to 100, where 100 means perfect agreement. Because the methodology produces source- and site-specific synthetic ground motion time histories, and the goodness-of-fit scores of obtained synthetics are between ‘good’ and ‘excellent’ range (61.128–82.164) based on the Anderson’s score, we conclude that it can be used to produce reliable ground motion time histories for seismic risk regions to develop or improve seismic codes and standards.

Simultaneous multiple well-seismic ties using flattened synthetic and real seismograms

Well-seismic ties allow rock properties measured at well locations to be compared with seismic data and are therefore useful for seismic interpretation. Numerous methods have been proposed to compute well-seismic ties by correlating real seismograms with synthetic seismograms computed from velocity and density logs. However, most methods tie multiple wells to seismic data one by one; hence, they do not guarantee lateral consistency among multiple well ties. We therefore propose a method to simultaneously tie multiple wells to seismic data. In this method, we first flatten synthetic and corresponding real seismograms so that all seismic reflectors are horizontally aligned. By doing this, we turn multiple well-seismic tying into a 1D correlation problem. We then compute only vertically variant but laterally constant shifts to correlate these horizontally aligned (flattened) synthetic and real seismograms. This two-step correlation method maintains lateral consistency among multiple well ties by computing a laterally and vertically optimized correlation of all synthetic and real seismograms. We applied our method to a 3D real seismic image with multiple wells and obtained laterally consistent well-seismic ties.

Evaluating orthotropic continuum analysis of stress wave propagation through a jointed rock mass

Continuum modeling and discontinuum modeling are two approaches that are used to study the problem of stress wave propagation in jointed rock masses. In this study, a numerical analysis of blast wave propagation through a rock mass of three orthogonal joint sets was conducted by replacing the discontinuous media with orthotropic continua. The features of the simulated blast waveforms were fitted to those of the real seismograms (recorded at a distance of 300 m from a production blast at the Gol-e-Gohar iron ore mine) by iterative adjustment of the variable parameters. In order to account for the effect of rock joint dilation using the orthotropic elastic constitutive model, a search range was considered for the nonzero off-diagonal terms of the compliance matrix of the equivalent media, within which these three parameters were back-calculated independently. A simulated annealing search algorithm was used in conjunction with the numerical modeling to supervise the parameter adjustment. Finally, the response of the equivalent continuous model to dynamic loading was compared with the in situ records and the response observed in previously performed discrete modeling. The results demonstrated that the values of the peak particle velocities and the predominant frequencies in the model and those from site monitoring were in good accord. However, eliminating joints led to wider response Fourier spectra than seen for the real records and the jointed model results, and the numerical model required increased additional damping to compensate for this simplification.

Unraveling overtone interferences in Love-wave phase velocity measurements by radon transform

SUMMARY Surface waves contain fundamental mode and higher modes, which could interfere with each other. If different modes are not properly separated, the inverted Earth structures using surface waves could be biased. In this study, we apply linear radon transform (LRT) to synthetic seismograms and real seismograms from the USArray to demonstrate the effectiveness of LRT in separating fundamental-mode Love waves from higher modes. Analysis on synthetic seismograms shows that two-station measurements on reconstructed data obtained after mode separation can completely retrieve the fundamental-mode Love-wave phase velocities. Results on USArray data show that higher mode contamination effects reach up to ∼10 per cent for two-station measurements of Love waves, while two-station measurements on mode-separated data obtained by LRT are very close to the predicted values from a global dispersion model of GDM52, demonstrating that the contamination of overtones on fundamental-mode Lovewave phase velocity measurements is effectively mitigated by the LRT method and accurate fundamental-mode Love-wave phase velocities can be measured.

A Systematic Comparison of the Transverse Energy Minimization and Splitting Intensity Techniques for Measuring Shear‐Wave Splitting Parameters

Over the past several decades, shear-wave splitting (SWS) analyses have been increasingly utilized to delineate mantle structure and probe mantle dynamics. However, the reported splitting parameters (fast polarization orientations and splitting times) are frequently inconsistent among different studies, partially due to the different techniques used to estimate the splitting parameters. Here, we report results from a sys- tematic comparison of the transverse minimization (TM) and the splitting intensity (SI) techniques. The study was motivated by the fact that recent comparative studies led to conflicting conclusions, which include the suggestion that TM, which is arguably the most widely used SWS-measuring technique, performs significantly poorly relative to SI under most circumstances in terms of stability and reliability of the resulting splitting parameters. We use both synthetic and real seismograms to evaluate the performance of the techniques for noise resistance, dominant period dependence, and complex aniso- tropy recognition. For one-layer anisotropy models with a horizontal axis of symmetry, our results show the two techniques can provide measurements with similar reliability. The testing confirms conclusions from previous studies that, although SI cannot distin- guish between simple and complex anisotropy models with a horizontal axis of sym- metry, TM can serve as a powerful tool in recognizing the existence of complex anisotropy, which is characterized by a systematic dependence of the splitting param- eters on the back azimuth of the events. Therefore, when the existence of complex anisotropy beneath a study area is unknown, TM is a better choice. Ⓔ A FORTRAN program for the calculation of Wiener-filtered wavelet and splitting intensity using SI technique is provided as an electronic supplement to this article. Online Material: FORTRAN code to calculate splitting intensity.

Kinematically excited parametric vibration of a tall building model with a TMD—Part 2: Experimental analyses

This paper constitutes the second part of the article Kinematically excited parametric vibration of a tall building model with a TMD. Part 1: numerical analyses (ACME, in press) by K. Majcher and Z. Wójcicki, which presents the results of theoretical research. This paper presents the experimental verification of those results. The experimental studies were carried out with the use of an especially designed physical model of a tall building, which rested on an earthquake simulator – a shaking table – created for this project. The simulator was used to generate several types of kinematic excitations: harmonic ones, superpositions of harmonic ones and, finally, ones generated on the basis of real seismograms. Vibrations were kinematically excited in the horizontal and vertical directions independently and simultaneously.The vertical component of the earthquake causes the pendulum suspension point to vibrate, thus exciting the pendulum parametrically. The theoretical study indicated a significant influence of this parametric excitation (parametric resonance) on the effectiveness of the Pendulum Tuned Mass Damper (PTMD). Therefore, the experimental analyses were especially focused on the parametric effects' impact on:•the PTMD's ability to reduce the building's vibration, and•the possibility of parametric resonance of the building due to parametric resonance of the PTMD.

Using Real and Synthetic Waveforms of the Matata Swarm to Assess the Performance of New Zealand GeoNet Phase Pickers

Abstract We assess the performance of the P ‐ and S ‐phase pickers of the New Zealand Seismograph Network (NZSN) by comparing their outputs for real seismograms and corresponding synthetic seismograms of the local earthquake sequence in the Matata region, New Zealand, during 2008. We study the effect of the radiation pattern on the pickers’ performance by computing synthetic seismograms at 85 hypothetical stations. These are located at distances of 10, 20, 30, 40, and 50 km from an earthquake source represented by a double‐couple point source. The picked S ‐phase onset has larger deviations from the mean picked onset at each distance compared to the P ‐phase onset. Deviation from the mean P ‐phase onset at each point does not affect the corresponding deviation from the mean S ‐phase onset. We confirm that none of the key stations in 2008 are located where the picking errors are large or where the phases are missed due to the radiation pattern. We also study the effect of noise on the pickers’ performance at three different signal‐to‐noise ratio levels (SNR=20, 10, and 3 dB). No P ‐phase onset time was picked at 50 km distance when SNR was 3 dB. Both pickers had a consistent performance at stations with 20 km distance from the earthquake source. At larger distances, three sea‐based observatory points with azimuth angles 0°, 21°, and 339° show a consistent performance in the presence of the increased noise level. Online Material: Tables describing the performance of the pickers with noise free synthetic data.