Dextral-Slip Thrust Faulting and Seismic Events of the Ms 8.0 Wenchuan Earthquake,Longmenshan Mountains,Eastern Margin of the Tibetan Plateau

Dextral-slip thrust movement of the Songpan-Garze terrain over the Sichuan block caused the Ms 8.0 Wenchuan earthquake of May 12, 2008 and offset the Central Longmenshan Fault （CLF） along a distance of -250 km. Displacement along the CLF changes from Yingxiu to Qingchuan. The total oblique slip of up to 7.6 m in Yingxiu near the epicenter of the earthquake, decreases northeastward to 5.3 m, 6.6 m, 4.4 m, 2.5 m and 1.1 m in Hongkou, Beichuan, Pingtong, Nanba and Qingchuan, respectively. This offset apparently occurred during a sequence of four reported seismic events, EQ1-EQ4, which were identified by seismic inversion of the source mechanism. These events occurred in rapid succession as the fault break propagated northeastward during the earthquake. Variations in the plunge of slickensides along the CLF appear to match these events. The Mw 7.5 EQ1 event occurred during the first 0-10 s along the Yingxiu-Hongkou section of the CLF and is characterized by 1.7 m vertical slip and vertical slickensides. The Mw 8.0 EQ2 event, which occurred during the next 10-42 s along the Yingxiu-Yanziyan section of the CLF, is marked by major dextralslip with minor thrust and slickensides plunging 25°-35° southwestward. The Mw 7.5 EQ3 event occurred during the following 42-60 s and resulted in dextral-slip and slickensides plunging 10° southwestward in Beichuan and plunging 73° southwestward in Hongkou. The Mw 7.7 EQ4 event, which occurred during the final 60-95 s along the Beichuan-Qingchuan section of the CLF, is characterized by nearly equal values of dextral and vertical slips with slickensides plunging 45°-50° southwestward. These seismic events match and evidently controlled the concentrations of landslide dams caused by the Wenchuan earthquake in Longmenshan Mountains.

Calibration of P/S amplitude ratios for seismic events in Xinjiang and its adjacent areas based on a Bayesian Kriging method

Correction maps of P/S amplitude ratios for seismic events distributed in Xinjiang, China and its adjacent areas were established using a Bayesian Kriging method for the two seismic stations WMQ and MAK. The relationship between correction maps and variations of along-path features was analyzed and the validity of applying the correction maps to improve performances of P/S discriminants for seismic discrimination was investigated. Results show that obtained correction maps can generally reflect event-station path effects upon corresponding P/S discriminants; and the correction of these effects could further reduce scatters of distance-corrected P/S measurements within earthquake and explosion populations as well as improve their discriminating performances if path effects are a significant factor of such scatters. For example, as corresponding Kriging correction map was applied, the misidentification rate of earthquakes by Pn（2-4 Hz）/Lg（2-4 Hz） at MAK was reduced from 16.3% to 5.2%.

Classification of clustered microseismic events in a coal mine using machine learning

Discrimination of seismicity distributed in different areas is essential for reliable seismic risk assessment in mines.Although machine learning has been widely applied in seismic data processing,feasibility and reliability of applying this technique to classify spatially clustered seismic events in underground mines are yet to be investigated.In this research,two groups of seismic events with a minimum local magnitude(ML) of-3 were observed in an underground coal mine.They were respectively located around a dyke and the longwall face.Additionally,two types of undesired signals were also recorded.Four machine learning methods,i.e.random forest(RF),support vector machine(SVM),deep convolutional neural network(DCNN),and residual neural network(ResNN),were used for classifying these signals.The results obtained based on a primary dataset showed that these seismic events could be classified with at least 91% accuracy.The DCNN using seismogram images as the inputs reached the best performance with more than 94% accuracy.As mining is a dynamic progress which could change the characteristics of seismic signals,the temporal variance in the prediction performance of DCNN was also investigated to assess the reliability of this classifier during mining.A cascaded workflow consisting of database update,model training,signal prediction,and results review was established.By progressively calibrating the DCNN model,it achieved up to 99% prediction accuracy.The results demonstrated that machine learning is a reliable tool for the automatic discrimination of spatially clustered seismicity in underground mining.

Seismic evaluation of the destress blasting efficiency

In this paper, selected methods of destress blasting efficiency assessment are presented, and novel quantitative methods based on in situ seismic measurements are proposed. The newly formulated solution combines two different approaches. The first, which is useful mostly for the near-field seismic analyses, is based on the analysis of seismic amplitude characteristics, and the second, relevant for farfield evaluation, is extended by the duration and frequency of the seismic wave. Both approaches are based on the seismic analyses of the waveforms generated by blasting recorded by the local seismic network. The proposed solutions are tested and validated in deep underground mines in Poland in which the room-and-pillar mining method is applied. Based on performed analysis, it is shown that both methods may be used as a rockburst hazard control in underground mines. However, developed methods may also be successfully implemented in other engineering practices, including the assessment of seismic vibrations in open pits and quarries.

Multivariate discrimination technique based on the Bayesian theory

A multivariate discrimination technique was established based on the Bayesian theory. Using this technique, P/S ratios of different types （e.g., Pn/Sn, Pn/Lg, Pg/Sn or Pg/Lg） measured within different frequency bands and from different stations were combined together to discriminate seismic events in Central Asia. Major advantages of the Bayesian approach are that the probability to be an explosion for any unknown event can be directly calculated given the measurements of a group of discriminants, and at the same time correlations among these discriminants can be fully taken into account. It was proved theoretically that the Bayesian technique would be optimal and its discriminating performance would be better than that of any individual discriminant as well as better than that yielded by the linear combination approach ignoring correlations among discriminants. This conclusion was also validated in this paper by applying the Bayesian approach to the above-mentioned observed data.

Suitable triggering algorithms for detecting strong ground motions using MEMS accelerometers

With the recent development of digital Micro Electro Mechanical System （MEMS） sensors, the cost of monitoring and detecting seismic events in real time can be greatly reduced. Ability of MEMS accelerograph to record a seismic event depends upon the efficiency of triggering algorithm, apart from the sensor＇s sensitivity. There are several classic triggering algorithms developed to detect seismic events, ranging from basic amplitude threshold to more sophisticated pattern recognition. Algorithms based on STA/LTA are reported to be computationally efficient for real time monitoring. In this paper, we analyzed several STA/LTA algorithms to check their efficiency and suitability using data obtained from the Quake Catcher Network （network of MEMS accelerometer stations）. We found that most of the STA/LTA algorithms are suitable for use with MEMS accelerometer data to accurately detect seismic events. However, the efficiency of any particular algorithm is found to be dependent on the parameter set used （i.e., window width of STA, LTA and threshold level）.

Fast magnitude determination using a single seismological station record implementing machine learning techniques

In this work a Support Vector Machine Regression（SVMR） algorithm is used to calculate local magnitude（MI） using only five seconds of signal after the P wave onset of one three component seismic station. This algorithm was trained with 863 records of historical earthquakes, where the input regression parameters were an exponential function of the waveform envelope estimated by least squares and the maximum value of the observed waveform for each component in a single station. Ten-fold cross validation was applied for a normalized polynomial kernel obtaining the mean absolute error for different exponents and complexity parameters. The local magnitude（MI） could be estimated with 0.19 units of mean absolute error. The proposed algorithm is easy to implement in hardware and may be used directly after the field seismological sensor to generate fast decisions at seismological control centers, increasing the possibility of having an effective reaction.