scholarly journals Microseismic event detection in large heterogeneous velocity models using Bayesian multimodal nested sampling

2021 ◽  
Vol 2 ◽  
Author(s):  
Saptarshi Das ◽  
Michael P. Hobson ◽  
Farhan Feroz ◽  
Xi Chen ◽  
Suhas Phadke ◽  
...  
Keyword(s):  
Event Detection ◽  
Marginal Likelihood ◽  
Microseismic Monitoring ◽  
Origin Time ◽  
Nested Sampling ◽  
Velocity Models ◽  
Microseismic Events ◽  
Spatio Temporal ◽  
Microseismic Event ◽  
False Event

Abstract In passive seismic and microseismic monitoring, identifying and characterizing events in a strong noisy background is a challenging task. Most of the established methods for geophysical inversion are likely to yield many false event detections. The most advanced of these schemes require thousands of computationally demanding forward elastic-wave propagation simulations. Here we train and use an ensemble of Gaussian process surrogate meta-models, or proxy emulators, to accelerate the generation of accurate template seismograms from random microseismic event locations. In the presence of multiple microseismic events occurring at different spatial locations with arbitrary amplitude and origin time, and in the presence of noise, an inference algorithm needs to navigate an objective function or likelihood landscape of highly complex shape, perhaps with multiple modes and narrow curving degeneracies. This is a challenging computational task even for state-of-the-art Bayesian sampling algorithms. In this paper, we propose a novel method for detecting multiple microseismic events in a strong noise background using Bayesian inference, in particular, the Multimodal Nested Sampling (MultiNest) algorithm. The method not only provides the posterior samples for the 5D spatio-temporal-amplitude inference for the real microseismic events, by inverting the seismic traces in multiple surface receivers, but also computes the Bayesian evidence or the marginal likelihood that permits hypothesis testing for discriminating true vs. false event detection.

Evolution of a block cave from time-lapse passive source body-wave traveltime tomography

Geophysics ◽  
2015 ◽  
Vol 80 (2) ◽  
pp. WA85-WA97 ◽  
Author(s):  
Jean-Philippe Mercier ◽  
Willem de Beer ◽  
Jean-Pascal Mercier ◽  
Simon Morris
Keyword(s):  
Rock Mass ◽  
Velocity Distribution ◽  
Body Wave ◽  
Time Lapse ◽  
Microseismic Monitoring ◽  
Monitoring Systems ◽  
Simultaneous Estimation ◽  
Velocity Models ◽  
Microseismic Events ◽  
Microseismic Event

Most underground mines are equipped with microseismic monitoring systems that allow the detection, location, and characterization of microseismic events. Microseismic events can be exploited to understand the rock mass response to mining. However, seismicity provides information only for regions that are seismically active. Although some information on nonseismically active regions can be obtained from point measurements and numerical modeling, these methods suffer from limitations of their own. Passive source traveltime body-wave tomography (passive source tomography [PST]) uses information readily collected by microseismic monitoring systems, namely, the P- and/or S-wave traveltimes and microseismic event hypocenter locations. This technique allowes the simultaneous estimation of the velocity distribution between sensors and microseismic events and the correction of microseismic event hypocenter locations. In this paper, we present an application of time-lapse PST to the Northparkes Mines E26 Lift 2 block cave showing that PST can be used to obtain information on evolution and distribution of seismic velocities, leading to a better understanding of stress distribution and redistribution and of rock mass behavior during the development and production phases. In particular, we found that (1) the magnitude of the velocity perturbation varied through time and appeared to be strongly correlated with the intensity of microseismic activity, the mining rate, and the nature of the mining activity, (2) the velocity models provided information that allowed for the inference of the cave geometry and its evolution through time, (3) the stress distributions inferred from the velocity model were not fully consistent with a widely accepted conceptual stress redistribution model, which may reflect the significant influence of rock mass inhomogeneities and the mining sequence, (4) seismicity was found in regions in which velocity was higher and lower than the background velocity, and (5) there was no obvious correlation between geology and velocity distribution and evolution.

scholarly journals Automatic high-resolution microseismic event detection via supervised machine learning

2020 ◽  
Vol 222 (3) ◽  
pp. 1881-1895 ◽  
Author(s):  
Shan Qu ◽  
Zhe Guan ◽  
Eric Verschuur ◽  
Yangkang Chen
Keyword(s):  
Feature Selection ◽  
High Resolution ◽  
Event Detection ◽  
Training Data ◽  
Supervised Machine Learning ◽  
Gaussian Kernel ◽  
Selection Strategy ◽  
Support Vector ◽  
Microseismic Events ◽  
Microseismic Event

SUMMARY Microseismic methods are crucial for real-time monitoring of the hydraulic fracturing dynamic status during the development of unconventional reservoirs. However, unlike the active-source seismic events, the microseismic events usually have low signal-to-noise ratio (SNR), which makes its data processing challenging. To overcome the noise issue of the weak microseismic events, we propose a new workflow for high-resolution microseismic event detection. For the preprocessing, fix-sized segmentation with a length of 2*wavelength is used to divide the data into segments. Later on, 191 features have been extracted and used as the input data to train the support vector machine (SVM) model. These features include 63 1-D time/spectral-domain features, and 128 2-D texture features, which indicate the continuity, smoothness, and irregularity of the events/noise. The proposed feature extraction maximally exploits the limited information of each segment. Afterward, we use a combination of univariate feature selection and random-forest-based recursive feature elimination for feature selection to avoid overfitting. This feature selection strategy not only finds the best features, but also decides the optimal number of features that are needed for the best accuracy. Regarding the training process, SVM with a Gaussian kernel is used. In addition, a cross-validation (CV) process is implemented for automatic parameter setting. In the end, a group of synthetic and field microseismic data with different levels of complexity show that the proposed workflow is much more robust than the state-of-the-art short-term-average over long-term-average ratio (STA/LTA) method and also performs better than the convolutional-neural-networks (CNN), for this case where the amount of training data sets is limited. A demo for the synthetic example is available: https://github.com/shanqu91/ML_event_detection_microseismic.

Pitfalls locating microseismic events from borehole measurements — Practical observations from field applications

2013 ◽  
Vol 1 (2) ◽  
pp. A11-A17 ◽  
Author(s):  
Carlos Cabarcas
Keyword(s):  
Hydraulic Fracturing ◽  
Hydraulic Fracture ◽  
Microseismic Monitoring ◽  
Reservoir Stimulation ◽  
Velocity Models ◽  
Unconventional Resources ◽  
Geophysical Techniques ◽  
Microseismic Events ◽  
The Impact

Borehole microseismic monitoring of hydraulic fracturing is among the best tools for reservoir stimulation evaluation. After decades of research and execution, the technique has gained a well-deserved place within the engineering toolbox. Moreover, in recent years, its popularity has increased exponentially, together with the development of unconventional resources. However, while involved with a significant number of borehole microseismic monitoring campaigns, I noticed that it is a common practice to overlook fundamental principles during the location of microseismic events. This may lead to potentially erroneous hydraulic fracturing assessments. Examples of microseismic results qualitatively illustrate this assertion showing poor recording, velocity models, processing constraints, and display. They also underscore the interpreter’s role in ensuring the most reasonable outcome from a microseismic hydraulic fracture evaluation. In this respect, any conclusion derived from a microseismic experiment should be fully supported by a thorough understanding of the impact that multiple acquisition and processing assumptions have on the interpretation, as is the case for all other geophysical techniques. Ultimately, my intent is to raise awareness of some common pitfalls while also providing recommendations to increase the value of a microseismic monitoring exercise.

scholarly journals Research on the Law of Large-Scale Deformation and Failure of Soft Rock Based on Microseismic Monitoring

2018 ◽  
Vol 2018 ◽  
pp. 1-8 ◽  
Author(s):  
Feng Chen ◽  
Tianhui Ma ◽  
Chun’an Tang ◽  
Yanhong Du ◽  
Zhichao Li ◽  
...  
Keyword(s):  
Monitoring System ◽  
Large Scale ◽  
Failure Process ◽  
Soft Rock ◽  
Microseismic Monitoring ◽  
Surrounding Rock ◽  
Deformation And Failure ◽  
Microseismic Events ◽  
Rock Tunnel ◽  
Microseismic Event

Based on the existing Canadian ESG microseismic monitoring system, a mobile microseismic monitoring system for a soft rock tunnel has been successfully constructed through continuous exploration and improvement to study the large-scale nucleation and development of microfractures in the soft rock of the Yangshan Tunnel. All-weather, continuous real-time monitoring is conducted while the tunnel is excavated through drilling and blasting, and the waveform characteristics of microseismic events are analysed. Through the recorded microseismic monitoring data, the variation characteristics of various parameters (e.g., the temporal, spatial, and magnitude distributions of the microseismic events, the frequency of microseismic events, and the microseismic event density and energy) are separately studied during the process of large-scale deformation instability and failure of the soft rock tunnel. The relationship between the deterioration of the rock mass and the microseismic activity during this failure process is consequently discussed. The research results show that a microseismic monitoring system can be used to detect precursors; namely, the microseismic event frequency and energy both will appear “lull” and “active” periods during the whole failure process of soft rock tunnel. Two peaks are observed during the evolution of failure. When the second peak occurs, it is accompanied by the destruction of the surrounding rock. The extent and strength of the damage within the surrounding rock can be delineated by the spatial, temporal, and magnitude distributions of the microseismic events and a microseismic event density nephogram. The results of microseismic analysis confirm that a microseismic monitoring system can be used to monitor the large-scale deformation and failure processes of a soft rock tunnel and provide early warning for on-site construction workers to ensure the smooth development of the project.

Full-waveform inversion for microseismic events using sparsity constraints

Geophysics ◽  
2019 ◽  
Vol 84 (2) ◽  
pp. KS1-KS12 ◽  
Author(s):  
Bharath Shekar ◽  
Harpreet Singh Sethi
Keyword(s):  
Waveform Inversion ◽  
Synthetic Data ◽  
Velocity Model ◽  
Full Waveform Inversion ◽  
Limited Memory ◽  
Full Waveform ◽  
Sparsity Constraints ◽  
Microseismic Events ◽  
Spatio Temporal ◽  
Microseismic Event

Full-waveform inversion (FWI) is a powerful tool that can be used to invert for microseismic event locations and the source signature because it can exploit the complete waveform information. We have developed an algorithm to invert for a spatio-temporal source function that encapsulates microseismic events with spatially localized or distributed locations and source signatures. The algorithm does not require assumptions to be made about the number or type of sources; however, it does require that the velocity model is close to the true subsurface velocity. We reformulate the conventional FWI algorithm based on the [Formula: see text]-norm data-misfit function by adding sparsity constraints using a sparsity promoting [Formula: see text]-norm as an additional regularization term to get more focused and less noise-sensitive event locations. The Orthant-Wise Limited-memory quasi-Newton algorithm is used to solve the optimization problem. It inherits the advantageous (fast convergence) properties of the limited memory Broyden-Fletcher-Goldfarb-Shanno method and can easily overcome the nondifferentiability of [Formula: see text]-norm at null positions. We determine the performance of the algorithm on noise-free and noisy synthetic data from the SEG/EAGE overthrust model.

Image-domain velocity inversion and event location for microseismic monitoring

Geophysics ◽  
2017 ◽  
Vol 82 (5) ◽  
pp. KS71-KS83 ◽  
Author(s):  
Ben Witten ◽  
Jeffrey Shragge
Keyword(s):  
Oil And Gas ◽  
Random Noise ◽  
Microseismic Monitoring ◽  
Seismic Events ◽  
S Wave ◽  
Origin Time ◽  
Image Domain ◽  
Velocity Inversion ◽  
Velocity Models ◽  
Event Location

Microseismic event locations obtained from seismic monitoring data sets are often a primary means of determining the success of fluid-injection programs, such as hydraulic fracturing for oil and gas extraction, geothermal projects, and wastewater injection. Event locations help the decision makers to evaluate whether operations conform to expectations or parameters need to be changed and may be used to help assess and reduce the risk of induced seismicity. However, obtaining accurate event location estimates requires an accurate velocity model, which is not available at most injection sites. Common velocity updating techniques require picking arrivals on individual seismograms. This can be problematic in microseismic monitoring, particularly for surface acquisition, due to the low signal-to-noise ratio of the arrivals. We have developed a full-wavefield adjoint-state method for locating seismic events while inverting for P- and S-wave velocity models that optimally focus multiple complementary images of recorded seismic events. This method requires neither picking nor initial estimates of event location or origin time. Because the inversion relies on (image domain) residuals that satisfy the differential semblance criterion, there is no requirement that the starting model be close to the true velocity. We determine synthetic results derived from a model with conditions similar to a field-acquisition scenario in terms of the number and spatial sampling of receivers and recorded coherent and random noise levels. The results indicate the effectiveness of the methodology by demonstrating a significantly enhanced focusing of event images and a reduction of 95% in event location error from a reasonable initial model.

scholarly journals Toward Real-time Microseismic Event Detection using the YOLOv3 Algorithm

2022 ◽  
Author(s):  
Xiaoyu Zhu ◽  
Jeffrey Shragge
Keyword(s):  
Real Time ◽  
Event Detection ◽  
High Speed ◽  
Detection System ◽  
Gaussian White Noise ◽  
Detection Methods ◽  
Support Vector ◽  
Signal To Noise ◽  
Microseismic Events ◽  
Microseismic Event

Real-time microseismic monitoring is essential for understanding fractures associated with underground fluid injection in unconventional reservoirs. However, microseismic events recorded on monitoring arrays are usually contaminated with strong noise. With a low signal-to-noise ratio (S/R), the detection of microseismic events is challenging using conventional detection methods such as the short-term average/long-term average (STA/LTA) technique. Common machine learning methods, e.g., feature extraction plus support vector machine (SVM) and convolutional neural networks (CNNs), can achieve higher accuracy with strong noise, but they are usually time-consuming and memory-intensive to run. We propose the use of YOLOv3, a state-of-art real-time object detection system in microseismic event detection. YOLOv3 is a one-stage deep CNN detector that predicts class confidence and bounding boxes for images at high speed and with great precision. With pre-trained weights from the ImageNet 1000-class competition dataset, physics-based training of the YOLOv3 algorithm is performed on a group of forward modeled synthetic microseismic data with varying S/R. We also add randomized forward-modeled surface seismic events and Gaussian white noise to generate ``semi-realistic'' training and testing datasets. YOLOv3 is able to detect weaker microseismic event signals with low signal-to-noise ratios (e.g., S/N=0.1) and achieves a mean average precision of 88.71\% in near real time. Further work is required to test YOLOv3 in field production settings.

scholarly journals Real-time spatio-temporal event detection on geotagged social media

2021 ◽  
Vol 8 (1) ◽  
Author(s):  
Yasmeen George ◽  
Shanika Karunasekera ◽  
Aaron Harwood ◽  
Kwan Hui Lim
Keyword(s):  
New York ◽  
Social Media ◽  
Event Detection ◽  
Detection System ◽  
Time And Space ◽  
Social Media Data ◽  
Event Time ◽  
Spatio Temporal ◽  
Geographical Space ◽  
Media Data

AbstractA key challenge in mining social media data streams is to identify events which are actively discussed by a group of people in a specific local or global area. Such events are useful for early warning for accident, protest, election or breaking news. However, neither the list of events nor the resolution of both event time and space is fixed or known beforehand. In this work, we propose an online spatio-temporal event detection system using social media that is able to detect events at different time and space resolutions. First, to address the challenge related to the unknown spatial resolution of events, a quad-tree method is exploited in order to split the geographical space into multiscale regions based on the density of social media data. Then, a statistical unsupervised approach is performed that involves Poisson distribution and a smoothing method for highlighting regions with unexpected density of social posts. Further, event duration is precisely estimated by merging events happening in the same region at consecutive time intervals. A post processing stage is introduced to filter out events that are spam, fake or wrong. Finally, we incorporate simple semantics by using social media entities to assess the integrity, and accuracy of detected events. The proposed method is evaluated using different social media datasets: Twitter and Flickr for different cities: Melbourne, London, Paris and New York. To verify the effectiveness of the proposed method, we compare our results with two baseline algorithms based on fixed split of geographical space and clustering method. For performance evaluation, we manually compute recall and precision. We also propose a new quality measure named strength index, which automatically measures how accurate the reported event is.

Spatio-Temporal Gaussianization Flows for Extreme Event Detection

2021 ◽  
Author(s):  
Miguel-Ángel Fernández-Torres ◽  
J. Emmanuel Johnson ◽  
María Piles ◽  
Gustau Camps-Valls
Keyword(s):  
Event Detection ◽  
Land Surface ◽  
Extreme Event ◽  
Learning Approaches ◽  
Proposed Model ◽  
Individual Contributions ◽  
Current Context ◽  
Spatio Temporal ◽  
The Individual ◽  
Anomaly Score

<p>Automatic anticipation and detection of extreme events constitute a major challenge in the current context of climate change. Machine learning approaches have excelled in detection of extremes and anomalies in Earth data cubes recently, but are typically both computationally costly and supervised, which hamper their wide adoption. We alternatively present here an unsupervised, efficient, generative approach for extreme event detection, whose performance is illustrated for drought detection in Europe during the severe Russian heat wave in 2010. The core architecture of the model is generic and could naturally be extended to the detection of other kinds of anomalies. First, it computes hierarchical appearance (spatial) and motion (temporal) representations of several informative Essential Climate Variables (ECVs), including soil moisture, land surface temperature, as well as features describing vegetation health. Then, these representations are combined using Gaussianization Flows that yield a spatio-temporal anomaly score. This allows the proposed model not only to detect droughts areas, but also to explain why they were produced, monitoring the individual contributions of each of the ECVs to the indicator at its output.</p>

Sign in / Sign up

Export Citation Format

Share Document