Comparison of four techniques for estimating temporal change of seismic velocity with passive image interferometry

Passive image interferometry (PII) is becoming a powerful tool for detecting the temporal variations in the Earth's structure,which applies coda wave interferometry to the waveforms from the cross-correlation of seismic ambient noise.There are four techniques for estimating temporal change of seismic velocity with PII:moving-window cross-correlation technique (MWCCT),moving-window cross-spectrum technique (MWCST),stretching technique (ST) and moving-window stretching technique (MWST).In this paper,we use the continuous seismic records from a typical station pair near the Wenchuan MS8.0 earthquake fault zone and generate three sets of waveforms by stacking cross-correlation function of ambient noise with different numbers of days,and then apply four techniques to processing the three sets of waveforms and compare their results.Our results indicate that the techniques based on moving-window (MWCCT,MWCST and MWST) are superior in detecting the change of seismic velocity,and the MWCST can give a better estimate of velocity change than the other moving-window techniques due to measurement error.We also investigate the clock errors and their influences on measuring velocity change.We find that when the clock errors are not very large,they have limited impact on the estimate of the velocity change with the moving-window techniques.     

Identification of Local Seismic Events Using the Mikhnevo Seismic Array

A large number of events with sources in the immediate vicinity of an array are usually detected during seismological observations with seismic arrays. These events should be detected and correctly interpreted during processing of seismic array records in order to avoid clogging up the event catalog. This problem can be solved by classifying records of local events by genetic features and creating a databank with the most representative samples. The present paper considers local events recorded using a unique scientific setup, the Mikhnevo small aperture seismic array. Epicenters of local seismic events are located less than 5 km from the center of the array. Seismic responses of acoustic shock waves are also examined. Seismic events caused by anthropogenic sources are identified and classified using cluster, cross-correlation, and wavelet analysis. Events accompanied only by the arrival of surface waves, as well as events represented by body, surface, and acoustic waves, are identified. Shock wave events are classified as a separate category. A small group of supposedly natural weak events is also found. As a result, a databank of waveforms of local seismic events for the Mikhnevo seismic array is established. In the future, this will make it possible to automate their identification when investigating the seismicity of the East European Platform.     

天津滨海爆炸事件检测与相对定位研究

2015年8月12日发生在天津滨海新区的化学品爆炸事故造成了严重的人员伤亡和财产损失。 基于区域地震台网记录的数字化连续地震波形资料, 对爆炸发生过程的精细分析是深入调查爆炸事故发生原因、 快速估计事故所造成灾害与损失的重要手段之一。 本文首先利用震相到时和质点运动轨迹信息对记录波形中能量较强的震相性质进行分析, 再通过模板匹配算法检测爆炸过程中是否存在两次大爆炸以外的小爆炸事件, 并对检测方法的效能进行检验, 最后依据波形互相关得到的震相到时差估计两次较大爆炸的相对位置。 计算结果显示区域台网记录波形中能量最大的体波震相可能为沿沉积层顶部传播的Pg震相, 垂直向能量最强的面波震相为Rg震相, 而能量较弱的初至震相可能为结晶基底首波; 爆炸过程中除了两次能量较大的爆炸以外还能检测到两次较小的爆炸事件, 其当量相当于约M_L0.5的地震; 基于两次大爆炸Pg震相相对到时信息得到第二次大爆炸发生在第一次大爆炸的北西侧, 两次爆炸位置距离约50~55 m。 本文研究结果为爆炸事故的调查分析提供了地震学依据。     

一种地震P波和S波初至时间自动拾取的新方法

地震P波、S波初至时间的拾取是地震波分析的一项基础性工作.本文提出了一种新的地震波初至时间自动拾取的方法:首先,把地震波的三分量时程曲线变换为一组空间向的能量变化率时程曲线;然后对能量变化率时程曲线进行STA/LTA(Short Time Average/Long Time Average,短时间的均值/长时间的均值)处理,拾取地震P波和S波的大致初至时间;最后提出采用一种二次方自回归模型对初至附近的能量变化率曲线进行二次方自回归处理,精确拾取出P波和S波的初至时间.本文采用了10组芦山地震的记录数据和150组汶川地震的记录数据对此方法的可靠性进行了检验.以人工拾取结果为参考,此方法具有很高的准确率和稳定性,同时,相比于常用的STA/LTA方法和AIC(Akaike Information Criterion,Akaike信息准则)方法,此方法在计算时间效率方面稍微逊色,但是对S波初至时间的拾取精度和可靠性更高.此方法丰富了地震P波、S波初至时间的自动拾取方法.     

龙滩水库地震精定位及活动特征研究

本文利用结合了波形互相关技术的双差定位法对龙滩库区2006年至2007年发生的地震进行了精确定位,并对其活动特征进行了分析.通过定位结果的比较,证明利用波形互相关技术提取的地震对的P、S波走时差数据及双差精定位法显著地提高了定位的精度和质量.对龙滩库区地震活动特征的研究结果表明,龙滩库区的地震活动与水库蓄水过程密切相关,成丛分布的地震活动分别表现出对水库蓄水过程不同的响应过程和活动特征,反应了龙滩库区在岩性特征、渗透条件、地质构造及应力场等方面存在局部性差异.     

利用波形互相关方法识别分析灌县—安县断裂重复地震

以中国地震台网中心地震目录中的事件为模板地震,通过滑动窗口的波形互相关方法对布设在灌县—安县断裂周边17个流动地震台的连续地震记录进行处理,识别M_L0.0以上的重复地震. 然后使用结合波形互相关技术的双差算法对这些地震进行重定位,获得了243次地震的重定位结果. 结果表明: 在研究时段内,灌县—安县断裂的地震活动性呈减弱趋势; 地震震源的优势分布深度为5—15 km,震源深度剖面显示地震呈高角度向西倾斜分布; 地震震中沿NE向分布,与龙门山前山断裂的走向基本一致; 研究区内南、 北两段的地震活动性及b值存在差异,这可能与龙门山断裂带中段区域应力方向由南到北发生的WNW向到ENE向转换的构造作用密切相关.      

Location and moment tensor inversion of small earthquakes using 3D Green’s functions in models with rugged topography: application to the Longmenshan fault zone

With dense seismic arrays and advanced imaging methods, regional three-dimensional(3D) Earth models have become more accurate. It is now increasingly feasible and advantageous to use a 3D Earth model to better locate earthquakes and invert their source mechanisms by fitting synthetics to observed waveforms. In this study, we develop an approach to determine both the earthquake location and source mechanism from waveform information. The observed waveforms are filtered in different frequency bands and separated into windows for the individual phases. Instead of picking the arrival times, the traveltime differences are measured by cross-correlation between synthetic waveforms based on the 3D Earth model and observed waveforms. The earthquake location is determined by minimizing the cross-correlation traveltime differences. We then fix the horizontal location of the earthquake and perform a grid search in depth to determine the source mechanism at each point by fitting the synthetic and observed waveforms. This new method is verified by a synthetic test with noise added to the synthetic waveforms and a realistic station distribution. We apply this method to a series of M_W3.4–5.6 earthquakes in the Longmenshan fault(LMSF) zone, a region with rugged topography between the eastern margin of the Tibetan plateau and the western part of the Sichuan basin. The results show that our solutions result in improved waveform fits compared to the source parameters from the catalogs we used and the location can be better constrained than the amplitude-only approach. Furthermore, the source solutions with realistic topography provide a better fit to the observed waveforms than those without the topography, indicating the need to take the topography into account in regions with rugged topography.     

The distribution of earthquake multiplets beneath the southwest Pacific

Earthquakes beneath the southwest Pacific occur from the surface down to 700 km depth. Teleseismic waveforms created by some of these earthquakes are almost identical. We investigate Tonga–Kermadec and Vanuatu subduction zone earthquake P-coda waveforms using a cross-correlation technique and hierarchical clustering algorithm in order to determine the origin of waveform similarity and the distribution of earthquakes producing similar waveforms.We show that scatterers forming the majority of power in the P-wave coda are localised around the receiver. As a result, waveform similarity provides a much weaker constraint on source separation than in local studies. Waveform similarity can provide stronger constraints on focal mechanism.Most earthquake multiplets within the Tonga–Fiji–Kermadec Wadati–Benioff zone are found at depths between 0–60 km and 520–620 km. A significant proportion of all deep-focus events in south Pacific subduction zones have waveforms similar to those of at least one other event. Relative relocation of events within the largest identified multiplet reveals a planar zone of seismicity sub-parallel to the nodal plane of a related centroid moment tensor solution.Groups of earthquakes with similar waveforms remain active on at least the 14-year recording timescale. We equate this to repeated rupture on single or closely related shear systems within the subducting slabs.     

Seismic Imaging of a Bimaterial Interface Along the Hayward Fault,CA, with Fault Zone Head Waves and Direct P Arrivals

We observe fault zone head waves (FZHW) that are generated by and propagate along a roughly 80 km section of the Hayward fault in the San Francisco Bay area. Moveout values between the arrival times of FZHW and direct P waves are used to obtain average P-wave velocity contrasts across different sections of the fault. The results are based on waveforms generated by more than 5,800 earthquakes and recorded at up to 12 stations of the Berkeley digital seismic network (BDSN) and the Northern California seismic network (NCSN). Robust identification of FZHW requires the combination of multiple techniques due to the diverse instrumentation of the BDSN and NCSN. For single-component short-period instruments, FZHW are identified by examining sets of waveforms from both sides of the fault, and finding on one (the slow) side emergent reversed-polarity arrivals before the direct P waves. For three-component broadband and strong-motion instruments, the FZHW are identified with polarization analysis that detects early arrivals from the fault direction before the regular body waves which have polarizations along the source-receiver backazimuth. The results indicate average velocity contrasts of 3–8 % along the Hayward fault, with the southwest side having faster P wave velocities in agreement with tomographic images. A systematic moveout between the FZHW and direct P waves for about a 80 km long fault section suggests a single continuous interface in the seismogenic zone over that distance. We observe some complexities near the junction with the Calaveras fault in the SE-most portion and near the city of Oakland. Regions giving rise to variable FZHW arrival times can be correlated to first order with the presence of lithological complexity such as slivers of high-velocity metamorphic serpentinized rocks and relatively distributed seismicity. The seismic velocity contrast and geological complexity have important implications for earthquake and rupture dynamics of the Hayward fault, including a statistically preferred propagation direction of earthquake ruptures to the SE.     

Active Source Seismic Identification and Automatic Picking of the P-wave First Arrival Using a Convolutional Neural Network

In seismic data processing, picking of the P-wave first arrivals takes up plenty of time and labor, and its accuracy plays a key role in imaging seismic structures. Based on the convolution neural network (CNN), we propose a new method to pick up the P-wave first arrivals automatically. Emitted from MINI28 vibroseis in the Jingdezhen seismic experiment, the vertical component of seismic waveforms recorded by EPS 32-bit portable seismometers are used for manually picking up the first arrivals (a total of 7242). Based on these arrivals, we establish the training and testing sets, including 25,290 event samples and 710,616 noise samples (length of each sample:2s). After 3,000 steps of training, we obtain a convergent CNN model, which can automatically classify seismic events and noise samples with high accuracy (> 99%). With the trained CNN model, we scan continuous seismic records and take the maximum output (probability of a seismic event) as the P-wave first arrival time. Compared with STA/LTA (short time average/long time average), our method shows higher precision and stronger anti-noise ability, especially with the low SNR seismic data. This CNN method is of great significance for promoting the intellectualization of seismic data processing, improving the resolution of seismic imaging, and promoting the joint inversion of active and passive sources.     

Low signal-to-noise event detection based on waveform stacking and cross-correlation: application to a stimulation experiment

We study the microseismicity (M _L ?<?2) in the region of Landau, SW Germany. Here, due to thick sediments (~3?km) and high cultural seismic noise, the signal-to-noise ratio is in general very low for microearthquakes. To gain new insights into the occurrence of very small seismic events, we apply a three-step detection approach and are able to identify 207 microseismic events (?1?<?M _L ?<?~1) with signal-to-noise ratios smaller than 3. Recordings from a temporary broadband network are used with station distances of approximately 10?km. First, we apply a short-term to long-term average detection algorithm for data reduction. The detection algorithm is affected severely by transient noise signals. Therefore, the most promising detections, selected by coinciding triggers and high-amplitude measures, are reviewed manually. Thirteen seismic events are identified in this way. Finally, we conduct a cross-correlation analysis. As master template, we use the stacked waveforms of five manually detected seismic events with a repeating waveform. This search reveals additional 194 events with a cross-correlation coefficient exceeding 0.65 which ensures a stable identification. Our analysis shows that the repeating events occurred during the stimulation of a geothermal reservoir within a source region of only about 0.5?km³. Natural background seismicity exceeding our detection level of M _L ?~?0.7 is not found in the region of Landau by our analysis.     

Application of Machine Learning Methods in Arrival Time Picking of P Waves from Reservoir Earthquakes

Reservoir earthquake characteristics such as small magnitude and large quantity may result in low monitoring efficiency when using traditional methods. However, methods based on deep learning can discriminate the seismic phases of small earthquakes in a reservoir and ensure rapid processing of arrival time picking. The present study establishes a deep learning network model combining a convolutional neural network (CNN) and recurrent neural network (RNN). The neural network training uses the waveforms of 60 000 small earthquakes within a magnitude range of 0.8-1.2 recorded by 73 stations near the Dagangshan Reservoir in Sichuan Province as well as the data of the manually picked P-wave arrival time. The neural network automatically picks the P-wave arrival time, providing a strong constraint for small earthquake positioning. The model is shown to achieve an accuracy rate of 90.7% in picking P waves of microseisms in the reservoir area, with a recall rate reaching 92.6% and an error rate lower than 2%. The results indicate that the relevant network structure has high accuracy for picking the P-wave arrival times of small earthquakes, thus providing new technical measures for subsequent microseismic monitoring in the reservoir area.     

基于深度学习的地震检测模型在区域台网的泛化性研究

将识别地震的深度学习算法PhaseNet应用于四川台网和首都圈台网,对该模型的泛化能力进行了测试和评估.首先利用2010年1月至2018年10月首都圈台网199个地震台站记录的29 328个事件(ML0～ML4)所对应的126761段事件波形,以及2019年4-9月四川及邻省部分台网227个地震台站记录的16595个事...     

Study of Local Seismic Events in Lithuania and Adjacent Areas Using Data from the PASSEQ Experiment

The territory of Lithuania and adjacent areas of the East European Craton have always been considered a region of low seismicity. Two recent earthquakes with magnitudes of more than 5 in the Kaliningrad District (Russian Federation) on 21 September 2004 motivated re-evaluation of the seismic hazard in Lithuania and adjacent territories. A new opportunity to study seismicity in the region is provided by the PASSEQ (Pasive Seismic Experiment) project that aimed to study the lithosphere–asthenosphere structure around the Trans-European Suture Zone. Twenty-six seismic stations of the PASSEQ temporary seismic array were installed in the territory of Lithuania. The stations recorded a number of local and regional seismic events originating from Lithuania and adjacent areas. This data can be used to answer the question of whether there exist seismically active tectonic zones in Lithuania that could be potentially hazardous for critical industrial facilities. Therefore, the aim of this paper is to find any natural tectonic seismic events in Lithuania and to obtain more general view of seismicity in the region. In order to do this, we make a manual review of the continuous data recorded by the PASSEQ seismic stations in Lithuania. From the good quality data, we select and relocate 45 local seismic events using the well-known LocSAT and VELEST location algortithms. In order to discriminate between possible natural events, underwater explosions and on-shore blasts, we analyse spatial distribution of epicenters and temporal distribution of origin times and perform both visual analysis of waveforms and spectral analysis of recordings. We show that the relocated seismic events can be grouped into five clusters (groups) according to their epicenter coordinates and origin and that several seismic events might be of tectonic origin. We also show that several events from the off-shore region in the Baltic Sea (at the coasts of the Kaliningrad District of the Russian Federation) are non-volcanic tremors, although the origin of these tremor-type events is not clear.     

Study of Seismic Clusters at Bahía de Banderas Region,Mexico

The coast in the state of Jalisco and south of Nayarit is located within a region of high seismic potential, increasing population, and tourism development. This motivated Civil Defense authorities of Jalisco and the Universidad de Guadalajara to launch in the year 2000 the assessment of the seismic risk of the region. This work focuses in the seismicity study of the area of Bahía de Banderas and northern coast of Jalisco, which is actually a seismic gap. We perform an analysis of available seismograms to characterize active crustal structures, their relationship to surface morphology, and possible extent of these structures into the bay shallow parts. The data consist of waveforms recorded during 2003 when the seismograph network spanned the region. Our method is based on the identification of seismic clusters or families using cross-correlation of waveforms, earthquake relocation and modeling of fault planes. From an initial data set of 404 located earthquakes, 96 earthquakes with M_L < 3.6 are related to 17 potentially active continental structures. We present fault plane model for 11 structures. A subgroup of 7 structures is aligned parallel to the Middle America Trench, as a possible consequence of oblique subduction. The foci of the earthquakes were grouped into clusters corresponding to fault dimensions of hundred of meters, may be considered as asperities or barriers in tectonic structures with lengths between 10 and 30 km. These structures could generate shallow earthquakes with magnitudes between 5.0 and 6.0 and represent an additional seismic threat to the region.     

利用全波形匹配方法确定水力压裂诱发地震震源机制

准确确定水力压裂诱发地震的震源机制对于描述裂缝破裂类型以及工区地应力状态十分重要.本文采用全波形匹配的方法确定诱发地震震源机制解,在拟合波形的同时对实际数据P波初动极性和纵横波振幅比也进行匹配,并在此基础上发展了一种基于邻域算法分级优化确定震源机制解的新方法.新方法的优势在于它对解具有更强的约束,并且能够解决现有方法中存在的权值系数选择问题.合成数据测试结果表明新方法能够得到可靠的震源机制解.本文将该方法应用于国内某页岩气田水力压裂诱发地震监测数据,反演结果表明诱发地震的震源类型以走滑破裂为主,表明该地区的最大水平主应力大于垂直主应力.     

Teleseismic body waves extracted from ambient noise cross correlation between F-net and ChinArray phase II

The vertical-vertical noise cross-correlation functions (NCFs) between two seismic arrays, the Japan F-net and ChinArray phase II, are calculated using continuous recordings during 2013–2016. After array interferometry to obtain bin stacked NCFs, clear body waves are retrieved at different period bands. Teleseismic direct P waves for distance 15–40 degrees are observed between short period 3–10 ?s while core reflected PcP/ScS waves are more obvious for longer period 30–60s. The signal-to-noise-ratio (SNR) of the short period P waves reaches its highest point with bin widths around 20 ?km while SNRs of PcP and ScS increase slowly with bin width. All those body waves demonstrate clear directivity with strong signals traveling from the east. The time-lapse SNR variations for the PcP and ScS show correlation with the occurrence of major earthquakes, while the P-wave SNR demonstrates seasonal variations with additional contribution from major earthquakes. The present results suggest teleseismic body waves can be retrieved through bin stacking, though further processing is still necessary to obtain finer waveforms such as P wave triplications.     

渭河断陷盆地及邻区的地震活动性和深部地球物理背景

探讨了渭河断陷盆地及邻近地区的速度场、介质品质因子的分布特征及其与该区地质构造的关系，分析了该地区的地震活动性．结果表明该区域速度场与介质品质因子分布之间存在着较好的对应关系，并与该区域的地震活动性密切相关．     

河南平顶山平煤矿区天然地震和人工爆破的波形特征及识别判据

基于河南测震台网记录的平顶山平煤矿区及周边发生的天然地震和人工爆破资料,按照直观、可快速识别的要求,选用发震时刻、P波初动方向、振幅比、振幅与尾波持续时间比等方法,对该地区天然地震和人工爆破进行对比分析。结果表明,P波与S波最大振幅之比（Pm/Sm）是识别该地区地震类型的最有效判据;P波初动振幅与S波最大振幅之比（Pc/Sm）和S波最大振幅与尾波持续时间之比（Sm/T）2种方法可以作为辅助判据,以提高识别地震类型的准确性。     

Precursory seismic quiescence: Past,present, and future

Precursory seismic quiescence has played a major role in most of the succesful earthquake predictions made to date. In addition to these successes, the number of detailed post-mainshock documentations of precursory quiescence is steadily growing. These facts suggest that precursory quiescence will play an important role in earthquake prediction programs of the future. For this reason it is important to critically evaluate the present state of knowledge concerning this phenomenon. The history of observations of precursory seismic quiescence includes work on seismic gaps and seismic preconditions as well as actual studies of temporal quiescence. These papers demonstrated the importance of quantitative evaluation of seismicity rates and the benefits of systematic analysis. During the early 1980's the impact of man-made effects on seismicity rates was demonstrated for the first time. Despite progress in catalog understanding, the identification and correction of man-made seismicity changes remains as the major barrier to earthquake prediction using these data. Effects of man-made changes are apparent in many past studies of seismicity patterns, making the results difficult to evaluate. Recent experience with real-time anomalies has demonstrated the necessity of determining the false alarm rates associated with quiescence precursors. Determination of false alarm rates depends on quantitative definitions of anomalies and statistical evaluations of their significance. A number of successful predictions, which have been made on the basis of seismic quiescence, provide important lessons for present and future work. There are many presently unanswered questions regarding seismic quiescence which must be answered before we can determine the reliability of this phenomena as a precursor.