Clustering seismic activities using linear and nonlinear discriminant analysis

Küyük, H. Serdar; Yildirim, Eray; Doğan, Emrah; Horasan, Gündüz

doi:10.1007/s12583-014-0406-x

Cited by 29 publications

(11 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The blast waveform exhibit much higher amplitudes for the first few seconds of the trace, which leads to an overestimation of the magnitude by ElarmS. Most of the existing algorithms make use of the ratios between P-and S-wave properties to discriminate between earthquakes and quarry blasts (e.g., Kuyuk, Yildirim, et al, 2014, and references therein), and are not suitable for EEW due to the delay needed for processing the S wave. One of our observations is that future versions of ElarmS should include a filter for quarry blasts to reduce low-magnitude false alerts of this kind.…”

Section: Real-time Resultsmentioning

confidence: 99%

Implementing the ElarmS Earthquake Early Warning Algorithm on the Israeli Seismic Network

Nof

Allen

2016

Bulletin of the Seismological Society of America

View full text Add to dashboard Cite

Earthquake early warning systems (EEWS) are being operated and tested increasingly around the globe in recent years. Following the Israeli government's decision to build an EEWS in Israel, and as the Californian EEWS (ShakeAlert) moves toward its operational phase, we demonstrate implementation of one of its three algorithms, ElarmS, to the Israel region. We provide new tools and approaches for implementing and assessing ElarmS outside of California. The main challenges of this research are to identify, verify, and adjust the embedded location-dependent parameters in ElarmS to the Israeli region, utilizing an unoptimized seismic network and low seismicity rate. To this end, we run ElarmS in three different modes: (1) historical playbacks, (2) real-time continuous data processing, and (3) simulated data playbacks. These modes enable us to overcome the limitations of low seismicity rates in the region and evaluate the performance of ElarmS with the network that is currently available. We use historical playbacks to adjust the magnitude estimation equations of ElarmS. We then analyze real-time processing results and provide detailed analysis of two significant events in the region (M D 5.5 and 4.4). Finally, we provide the first case of how to use synthetic data to evaluate the performance of ElarmS. We find that alert times are mostly affected by the network geometry and also by data delays. Alerts are typically issued within 80 ms after the arrival of the required four P-wave triggers data to the system. Magnitude estimations are reliable for events with M D > 3:5 within 100 km of the Israeli network using a locally adjusted magnitude relation equation.

show abstract

Section: Real-time Resultsmentioning

confidence: 99%

Implementing the ElarmS Earthquake Early Warning Algorithm on the Israeli Seismic Network

Nof

Allen

2016

Bulletin of the Seismological Society of America

View full text Add to dashboard Cite

show abstract

“…To foresee the classes of new data, the created classifier finds the class with the smallest misclassification. LDF analysis estimates one covariance matrix for all classes and can be demonstrated as (Kuyuk et al 2014;Yavuz et al 2019):…”

Section: -Spectral Ratio (Sr) Is the Ratio Of Integrated Spectralmentioning

confidence: 99%

Decontamination of earthquake catalog from quarry blast events in northern Egypt

et al. 2019

View full text Add to dashboard Cite

Earthquake catalogs are often contaminated by man-made crustal activities such as quarry blasts. That is the case in northern Egypt, where many of cement quarries were working so long time ago since the beginning of the twentieth century (1911). Therefore, it is necessary and essential to separate these quarry blast events from the earthquakes for real seismicity and hazard analysis. Generally, discrimination of quarry blasts from tectonic events often done through different techniques and methods. In this study, the time and frequency-domain analyses of the Egyptian National Seismological Network (ENSN) digital seismograms have performed using three techniques and methods (S/P wave amplitude peak ratio, complexity, and spectral ratio). Moreover, the power of event (Pe) analysis comprising three techniques (S/P wave amplitude peak ratio, complexity, spectral ratio) was also used together to extend the discrimination accuracy as well. Finally, we have used a linear discrimination function (LDF) analysis. A total of 1473 seismic event waveforms (M L ≤ 3.0) from the ENSN digital database have been investigated. The obtained results indicate that for West of Nile Dislocation (zone no. 1) 78.1% of the investigated seismic events are tectonic events; however, 21.1% are quarry events and 0.8% are misclassified events. For the East Nile Dislocation (zone no. 2) 36.2% and 63.7% are earthquakes and quarry blast events, respectively and 0.1% are misclassified events. This means that the earthquake catalog for zone no. 2 in northern Egypt is highly contaminated and needs very high concern during the further treats especially with microearthquakes dataset for actually real seismicity and hazard analysis.

show abstract

“…Interestingly, the discrimination between the earthquakes and artificial seismic sources using simple methods, i.e., satellite images, occurrence times, is probably not efficient if the artificial seismic sources are located along active fault zones [1], [11]. Consequently, the discrimination can be done utilizing variant techniques for both the time, and frequency domains, e.g., the peak amplitude ratio of the seismic phases [12], [13], the spectral ratio of seismic phases, or average amplitude in low-and high-frequency bands for a specific phase [9], [14]- [16].…”

Section: Introductionmentioning

confidence: 99%

Comparative Performance Assessments of Machine-Learning Methods for Artificial Seismic Sources Discrimination

et al. 2021

View full text Add to dashboard Cite

Mankind is vulnerable to artificial seismic sources and accompanying explosions' consequences. Recently, seismicity catalog contamination is among the main problems faced by seismologists. Since identifying artificial seismic sources is the first and always challenging stage, it is imperative to develop an automated control system that will discriminate tectonic from non-tectonic events. Detection and removal of the artificial seismic sources have become urgent. Early treatments and cleaning of contaminated seismicity catalogs are crucial to assist in accurate seismic hazard identification and enhance the planning of future urban developments. With the advancement of machine learning (ML) techniques, artificial seismic source detection accuracy has been improved. Today, there are different kinds of methods, ML techniques, and diverse processes like knowledge discovery are developed for discriminating artificial seismic sources and earthquakes. ML techniques offer various probabilistic and statistical methods that allow intelligent systems to learn from reoccurring experiences to detect and identify patterns from a dataset. This study aims to build an automated system that is able to detect the existence of artificial seismic sources in seismicity catalogs. More concretely, we classify seismic activity reports into two classes using classical and ensemble ML algorithms. Classical seismicity parameters or features are supplied to linear and nonlinear ML classifiers. The proposed scheme based on the four features (Latitude, Longitude, depth, and Magnitude) can enhance the performance. To assure the enhanced performance, we have examined the proposed scheme by both the accuracy of each model, ROC curves, Precision-Recall, and Calibration. The obtained results prove that the ensemble learning algorithms exhibit better results compared to other classical ML algorithms by having 98.14% testing accuracy.

show abstract

Clustering seismic activities using linear and nonlinear discriminant analysis

Cited by 29 publications

References 12 publications

Implementing the ElarmS Earthquake Early Warning Algorithm on the Israeli Seismic Network

Implementing the ElarmS Earthquake Early Warning Algorithm on the Israeli Seismic Network

Decontamination of earthquake catalog from quarry blast events in northern Egypt

Comparative Performance Assessments of Machine-Learning Methods for Artificial Seismic Sources Discrimination

Contact Info

Product

Resources

About