Classification Of Seismic Signals Research Articles

Uncertainty-Aware Seismic Signal Discrimination using Bayesian Convolutional Neural Networks

Seismic signal classification plays a crucial role in mitigating the impact of seismic events on human lives and infrastructure. Traditional methods in seismic hazard assessment often overlook the inherent uncertainties associated with the prediction of this complex geological phenomenon. This work introduces a probabilistic framework that leverages Bayesian principles to model and quantify uncertainty in seismic signal classification by applying a Bayesian Convolutional Neural Network (BCNN). The BCNN was trained on a dataset that comprises waveforms detected in the Southern California region and achieved an accuracy of 99.1%. Monte Carlo Sampling subsequently creates a 95% prediction interval for probabilities that considers epistemic and aleatoric uncertainties. The ability to visualize both aleatoric and epistemic uncertainties provides decision-makers with information to determine the reliability of seismic signal classifications. Further, the use of Bayesian CNN for seismic signal classification provides a more robust foundation for decision-making and risk assessment in earthquake-prone regions.

Vehicle Acoustic and Seismic Synchronization Signal Classification Using Long-Term Features

Field vehicle type recognition plays an essential role in border protection tasks. Acoustic and seismic sensors can effectively collect the signal of field vehicle targets in real time. Most vehicle temporal signal classification algorithms are based on extracting and identifying handcrafted features. These algorithms focus on the signal’s frequency domain characteristics and despise the signal’s time domain characteristics. In order to extract appropriate features, this paper proposes a Long-Term Correlation Feature Network (LTCFN) to perform field vehicle acoustic and seismic signal classification. The model includes AlexNet-type feature extractor and an overall classifier implemented by a Long-Short Term Memory(LSTM) network. We presents an intra-frame network and fusion method for extracting feature vector from signals. Meanwhile, an inter-frame classifier is proposed firstly for analysing the time-correlation of the feature map and overall classification. The experiments illustrate that the LTCFN has excellent recognition performance and anti-noise ability. The classification accuracy of the LTCFN can be increased to 96%. This paper also provides a new idea for ground target classification through inter-frame feature measurement.

Seismic Signal Analysis Based on Variational Mode Decomposition and Hilbert Transform for Ground Intrusion Activity Classification.

The identification of ground intrusion is a key and important technology in the national public security field. In this paper, a novel variational mode decomposition (VMD) and Hilbert transform (HT) is proposed for the classification of seismic signals generated by ground intrusion activities using a seismic sensing system. Firstly, the representative seismic data, including bicycles, vehicles, footsteps, excavations, and environmental noises, were collected through the designed experiment. Secondly, each original datum is decomposed through VMD and five Band-limited intrinsic mode functions (BIMF) are obtained, respectively, which will be used to generate a corresponding marginal spectrum that can reflect the actual frequency component of the signal accurately by HT. Then, three features related to the marginal spectrum, including marginal spectrum energy, marginal spectrum entropy, and marginal spectrum dominant frequency, are extracted for the analysis of the multi-classification using the support vector machine (SVM) classifier with the LIBSVM library. For the sake of testing and verifying the effectiveness of the proposed variational mode decomposition and Hilbert transform (VMD-HT) technique, the evaluation indicators including accuracy, precision, recall, and F1-Score are used and the results are compared with the time domain, frequency domain, ensemble empirical mode decomposition (EEMD), and empirical wavelet transform (EWT) combined with the HT analysis method. The performance of the VMD-HT method for ground intrusion activity classification provides an average value of 99.50%, 98.76%, 98.76%, and 98.75% for the four evaluation indicators, which are higher than all the other contrasted methods.

Most complicated lock pattern-based seismological signal framework for automated earthquake detection

BackgroundSeismic signals record earthquakes and also noise from different sources. The influence of noise makes it difficult to interpret seismograph signals correctly. This study aims to develop a computationally lightweight, accurate, and explainable machine learning model for the automated detection of seismogram signals that could serve as an effective warning system for earthquake prediction. Material and methodWe developed a handcrafted model for earthquake detection using a balanced dataset of 5001 earthquakes and 5001 non-earthquake signal samples. The model included multilevel feature extraction, selector-based feature selection, classification, and post-processing. Input signals were decomposed using tunable Q wave transform and fed to a statistical and textural feature extractor based on the most complicated lock pattern (MCLP). Four feature selectors were used to choose the most valuable features for the support vector machine classifier. Additionally, voted vectors were generated using iterative hard majority voting. Finally, the best model was chosen using a greedy algorithm. ResultsThe presented self-organized MCLP-based feature engineering model yielded 96.82% classification accuracy with 10-fold cross-validation using the seismic signal dataset. ConclusionsOur model attained high seismological signal detection performance comparable with more computationally expensive deep learning models. Our handcrafted explainable feature engineering model is computationally less expensive and can be easily implemented. Furthermore, we have introduced a competitive feature engineering model to the deep learning models for the seismic signal classification model.

Automated classification of seismic signals recorded on the Åknes rock slope, Western Norway, using a convolutional neural network

Abstract. A convolutional neural network (CNN) was implemented to automatically classify 15 years of seismic signals recorded by an eight-geophone network installed around the back scarp of the Åknes rock slope in Norway. Eight event classes could be identified and are adapted from the typology proposed by Provost et al. (2018), of which five could be directly related to movements on the slope. Almost 60 000 events were classified automatically based on their spectrogram images. The performance of the classifier is estimated to be near 80 %. The statistical analysis of the results shows a strong seasonality of the microseismic activity at Åknes with an annual increase in springtime when snow melts and the temperature oscillates around the freezing point, mainly caused by events within classes of low-frequency slope quakes and tremors. The clear link between annual temperature variations and microseismic activity could be confirmed, supporting thawing and freezing processes as the origins. Other events such as high-frequency and successive slope quakes occur throughout the year and are potentially related to the steady creep of the sliding plane. The huge variability in the annual event number cannot be solely explained by average temperatures or varying detectability of the network. Groundwater recharge processes and their response to precipitation episodes are known to be a major factor of sliding at Åknes, but the relationship with microseismic activity is less obvious and could not be demonstrated.

Domain Knowledge Informed Multitask Learning for Landslide-Induced Seismic Classification

Automatic seismic signal classification methods are extensively investigated to reduce or replace manual interpretation, with great potential in previous research. Discriminative seismic wave propagation physical characteristics, such as velocities and accelerations, are rarely considered for classification. A multitask learning scheme is proposed that utilises the seismic wave equation and three-dimensional (3D) P-wave velocity <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Vp</i> model for signal representation learning. The classifier uses the obtained latent feature maps on a convolutional neural network architecture for classification of rockfall, slide quake, earthquake, and natural/anthropogenic noise events, recorded at an ongoing landslide. Our experimental results show that our approach outperforms state-of-the-art methods.

Graph-Based Feature Weight Optimisation and Classification of Continuous Seismic Sensor Array Recordings

Slope instabilities caused by heavy rainfall, man-made activity or earthquakes can be characterised by seismic events. To minimise mortality and infrastructure damage, a good understanding of seismic signal properties characterising slope failures is therefore crucial to classify seismic events recorded from continuous recordings effectively. However, there are limited contributions towards understanding the importance of feature selection for the classification of seismic signals from continuous noisy recordings from multiple channels/sensors. This paper first proposes a novel multi-channel event-detection scheme based on Neyman-Pearson lemma and Multi-channel Coherency Migration (MCM) on the stacked signal across multi-channels. Furthermore, this paper adapts graph-based feature weight optimisation as feature selection, exploiting the signal's physical characteristics, to improve signal classification. Specifically, we alternatively optimise the feature weight and classification label with graph smoothness and semidefinite programming (SDP). Experimental results show that with expert interpretation, compared with the conventional short-time average/long-time average (STA/LTA) detection approach, our detection method identified 614 more seismic events in five days. Furthermore, feature selection, especially via graph-based feature weight optimisation, provides more focused feature sets with less than half of the original number of features, at the same time enhancing the classification performance; for example, with feature selection, the Graph Laplacian Regularisation classifier (GLR) raised the rockfall and slide quake sensitivities to 92% and 88% from 89% and 85%, respectively.

Ensemble and Self-supervised Learning for Improved Classification of Seismic Signals from the Åknes Rockslope

A case study with seismic geophone data from the unstable Åknes rock slope in Norway is considered. This rock slope is monitored because there is a risk of severe flooding if the massive-size rock falls into the fjord. The geophone data is highly valuable because it provides 1000 Hz sampling rates data which are streamed to a web resource for real-time analysis. The focus here is on building a classifier for these data to distinguish different types of microseismic events which are in turn indicative of the various processes occurring on the slope. There are 24 time series from eight 3-component geophone data for about 3500 events in total, and each of the event time series has a length of 16 s. For the classification task, novel machine learning methods such as deep convolutional neural networks are leveraged. Ensemble prediction is used to extract information from all time series, and this is seen to give large improvements compared with doing immediate aggregation of the data. Further, self-supervised learning is evaluated to give added value here, in particular for the case with very limited training data.

Algorithms and Measuring Complex for Classification of Seismic Signal Sources, Determination of Distance and Azimuth to the Point of Excitation of Surface Waves

Machine learning and digital signal processing methods are used in various industries, including in the analysis and classification of seismic signals from surface sources. The developed wave type analysis algorithm makes it possible to automatically identify and, accordingly, separate incoming seismic waves based on their characteristics. To distinguish the types of waves, a seismic measuring complex is used that determines the characteristics of the boundary waves of surface sources using special molecular electronic sensors of angular and linear oscillations. The results of the algorithm for processing data obtained by the method of seismic observations using spectral analysis based on the Morlet wavelet are presented. The paper also describes an algorithm for classifying signal sources, determining the distance and azimuth to the point of excitation of surface waves, considers the use of statistical characteristics and MFCC (Mel-frequency cepstral coefficients) parameters, as well as their joint application. At the same time, the following were used as statistical characteristics of the signal: variance, kurtosis coefficient, entropy and average value, and gradient boosting was chosen as a machine learning method; a machine learning method based on gradient boosting using statistical and MFCC parameters was used as a method for determining the distance to the signal source. The training was conducted on test data based on the selected special parameters of signals from sources of seismic excitation of surface waves. From a practical point of view, new methods of seismic observations and analysis of boundary waves make it possible to solve the problem of ensuring a dense arrangement of sensors in hard-to-reach places, eliminate the lack of knowledge in algorithms for processing data from seismic sensors of angular movements, classify and systematize sources, improve prediction accuracy, implement algorithms for locating and tracking sources. The aim of the work was to create algorithms for processing seismic data for classifying signal sources, determining the distance and azimuth to the point of excitation of surface waves.

A Novel Wavelet Selection Method for Seismic Signal Intelligent Processing

Wavelet transform is a widespread and effective method in seismic waveform analysis and processing. Choosing a suitable wavelet has also aroused many scholars’ research interest and produced many effective strategies. However, with the convenience of seismic data acquisition, the existing wavelet selection methods are unsuitable for the big dataset. Therefore, we proposed a novel wavelet selection method considering the big dataset for seismic signal intelligent processing. The relevance r is calculated using the seismic waveform’s correlation coefficient and variance contribution rate. Then values of r are calculated from all seismic signals in the dataset to form a set. Furthermore, with a mean value μ and variance value σ2 of that set, we define the decomposition stability w as μ/σ2. Then, the wavelet that maximizes w for this dataset is considered to be the optimal wavelet. We applied this method in automatic mining-induced seismic signal classification and automatic seismic P arrival picking. In classification experiments, the mean accuracy is 93.13% using the selected wavelet, 2.22% more accurate than other wavelets generated. Additionally, in the picking experiments, the mean picking error is 0.59 s using the selected wavelet, but is 0.71 s using others. Moreover, the wavelet packet decomposition level does not affect the selection of wavelets. These results indicate that our method can really enhance the intelligent processing of seismic signals.

Research on Seismic Signal Classification Based on Time-Frequency Map and Deep Learning

Research on Seismic Signal Classification Based on Time-Frequency Map and Deep Learning

Near-real-time automated classification of seismic signals of slope failures with continuous random forests

Abstract. In mountainous areas, rockfalls, rock avalanches, and debris flows constitute a risk to human life and property. Seismology has proven a useful tool to monitor such mass movements, while increasing data volumes and availability of real-time data streams demand new solutions for automatic signal classification. Ideally, seismic monitoring arrays have large apertures and record a significant number of mass movements to train detection algorithms. However, this is rarely the case, as a result of cost and time constraints and the rare occurrence of catastrophic mass movements. Here, we use the supervised random forest algorithm to classify windowed seismic data on a continuous data stream. We investigate algorithm performance for signal classification into noise (NO), slope failure (SF), and earthquake (EQ) classes and explore the influence of non-ideal though commonly encountered conditions: poor network coverage, imbalanced data sets, and low signal-to-noise ratios (SNRs). To this end we use data from two separate locations in the Swiss Alps: data set (i), recorded at Illgraben, contains signals of several dozen slope failures with low SNR; data set (ii), recorded at Pizzo Cengalo, contains only five slope failure events albeit with higher SNR. The low SNR of slope failure events in data set (i) leads to a classification accuracy of 70 % for SF, with the largest confusion between NO and SF. Although data set (ii) is highly imbalanced, lowering the prediction threshold for slope failures leads to a prediction accuracy of 80 % for SF, with the largest confusion between SF and EQ. Standard techniques to mitigate training data imbalance do not increase prediction accuracy. The classifier of data set (ii) is then used to train a model for the classification of 176 d of continuous seismic recordings containing four slope failure events. The model classifies eight events as slope failures, of which two are snow avalanches, and one is a rock-slope failure. The other events are local or regional earthquakes. By including earthquake detection of a permanent seismic station at 131 km distance to the test site into the decision-making process, all earthquakes falsely classified as slope failures can be excluded. Our study shows that, even for limited training data and non-optimal network geometry, machine learning algorithms applied to high-quality seismic records can be used to monitor mass movements automatically.

Identifying Different Classes of Seismic Noise Signals Using Unsupervised Learning

AbstractProper classification of nontectonic seismic signals is critical for detecting microearthquakes and developing an improved understanding of ongoing weak ground motions. We use unsupervised machine learning to label five classes of nonstationary seismic noise common in continuous waveforms. Temporal and spectral features describing the data are clustered to identify separable types of emergent and impulsive waveforms. The trained clustering model is used to classify every 1 s of continuous seismic records from a dense seismic array with 10–30 m station spacing. We show that dominate noise signals can be highly localized and vary on length scales of hundreds of meters. The methodology demonstrates the complexity of weak ground motions and improves the standard of analyzing seismic waveforms with a low signal‐to‐noise ratio. Application of this technique will improve the ability to detect genuine microseismic events in noisy environments where seismic sensors record earthquake‐like signals originating from nontectonic sources.

PICOSS: Python Interface for the Classification of Seismic Signals

Over the last decade machine learning has become increasingly popular for the analysis and characterization of volcano-seismic data. One of the requirements for the application of machine learning methods to the problem of classifying seismic time series is the availability of a training dataset; that is a suite of reference signals, with known classification used for initial validation of the machine outcome. Here, we present PICOSS (Python Interface for the Classification of Seismic Signals), a modular data-curator platform for volcano-seismic data analysis, including detection, segmentation and classification. PICOSS has exportability and standardization at its core; users can select automatic or manual workflows to select and label seismic data from a comprehensive suite of tools, including deep neural networks. The modular implementation of PICOSS includes a portable and intuitive graphical user interface to facilitate essential data labelling tasks for large-scale volcano seismic studies.

Seismic Signal Classification Using Perceptron With Different Learning Rules

Perceptron is adopted to classify the Ricker wavelets and to detect the seismic anomaly in a seismogram. Three learning rules are used in the training of perceptron to solve the decision boundary. The optimal learning-rate parameter is derived. The lower and upper bounds of the learning-rate parameter are derived. It can provide that the learning can converge when the parameter is within the range. The normalized learning rule is derived also. Combining learning rules, a fusion learning rule is proposed. In the experiments, these rules are applied to the detection of a seismic anomaly in the simulated seismogram and to compare the convergence speed. The fusion learning rule has the fastest convergence and is applied to the real seismogram. The seismic anomaly can be detected successfully. It can improve the seismic interpretation.

A Fast Algorithm for Classifying Seismic Events Using Distributed Computations in Apache Spark Framework

The main ideas of the development of the software implementation of an algorithm for the fast automatic classification of seismic signals based on diagnostic patterns are described. The process of adaptation and integration of this implementation into the distributed computations system Apache Spark is described in detail. A software solution for the preliminary processing of the signals and optimization of the mathematical model for parallel computations using broadcast variables is presented. Performance tests for the classification algorithm on a set of day-long signals are carried out. The execution time of the algorithm in the context of massively parallel computations was reduced tenfold compared with the sequential execution.

CLASSIFICATION OF MOUNT KELUD VOLCANO SEISMIC SIGNALS DURING ERUPTION IN NOVEMBER 2007

Classification of mount Kelud volcano seismic signals can be divided into several types, VT-A, VT-B, low frequency, multiphase, and tremor. The analysis by classifying the volcano seismic signal from eruption of Mount Kelud based on the shape of the signal. Data analyzed were starting on 16 to 27 October 2007. The results of volcano seismic signal which recording showed different shapes and were classified according to the type of signal. Each seismic signal recording is grouped according to its shape and the classification results are signaled into VT-A, VT-B, multiphase, low frequency, and tremor. Keywords: Kelud mountain, low frequency, multiphase, signal classification, tremor, VT-A, and VT-B.

Towards better volcanic risk-assessment systems by applying ensemble classification methods to triaxial seismic-volcanic signals

Analysis of seismic data is the most important method for volcano monitoring. Such data typically consists in digital signals acquired with an arrangement of triaxial seismic sensors which are strategically deployed on the volcano and its surrounding areas. Very rich measurements of the underlying phenomena are obtained from the arrangement because each sensor, through their corresponding three sensing axes, uninterruptedly acquires data at a relatively high sampling rate. Such an uninterrupted acquisition, however, turns manual classification of seismic signals into an inefficient and often error-prone task. As a solution, several systems for automated classification of seismic-volcanic signals have been proposed. All these systems, however, are limited to the usage of only one direction of acquisition; typically the vertical one. In this paper we make a step forward, exploring the potential benefit of using information from the three axes of the signals gathered by a single sensor. Integration is performed by classifier combining techniques, applied at different levels: this permits to take into account in the classification all the three orthogonal orientations —vertical, East-West and North-South— of the phenomenon. Preliminary experimental results on a set of volcanic signals gathered at Nevado del Ruiz volcano in Colombia confirmed the richness of this information.

Towards a standard typology of endogenous landslide seismic sources

Abstract. The objective of this work is to propose a standard classification of seismic signals generated by gravitational processes and detected at close distances (&lt;1 km). We review the studies where seismic instruments have been installed on unstable slopes and discuss the choice of the seismic instruments and the network geometries. Seismic observations acquired at 13 unstable slopes are analyzed in order to construct the proposed typology. The selected slopes are affected by various landslide types (slide, fall, topple and flow) triggered in various material (from unconsolidated soils to consolidated rocks). We investigate high-frequency bands (&gt;1 Hz) where most of the seismic energy is recorded at the 1 km sensor to source distances. Several signal properties (duration, spectral content and spectrogram shape) are used to describe the sources. We observe that similar gravitational processes generate similar signals at different slopes. Three main classes can be differentiated mainly from the length of the signals, the number of peaks and the duration of the autocorrelation. The classes are the “slopequake” class, which corresponds to sources potentially occurring within the landslide body; the “rockfall” class, which corresponds to signals generated by rock block impacts; and the “granular flow” class, which corresponds to signals generated by wet or dry debris/rock flows. Subclasses are further proposed to differentiate specific signal properties (frequency content, resonance, precursory signal). The signal properties of each class and subclass are described and several signals of the same class recorded at different slopes are presented. Their potential origins are discussed. The typology aims to serve as a standard for further comparisons of the endogenous microseismicity recorded on landslides.

Classification of acoustic and seismic signals based on the statistics of their wavelet sub-band coefficients

Classification of acoustic and seismic signals based on the statistics of their wavelet sub-band coefficients