广东地震台网近震定位精度的经验性GT准则

利用广东地震台网的固定台和流动台记录的2次人工爆破事件,通过随机模拟台网几何形态,获得了该台网近震定位精度的经验性GT（Ground Truth）准则。结果表明：地震定位精度与台网几何形态密切相关;当台网满足以下条件时近震震中定位精度优于2km,震源深度精度优于3km,具有95%的置信度：（1）震中距小于150km的台站数≥7;（2）一级方位角空区<180°;（3）震中距小于10km的台站数≥1。在评估震源深度精度时,发现对于壳内地震,可靠的近台数据能够提高震源深度的精度,而定位时震中距较大的台站数量的增加对震源深度的精度影响很小。最后,应用该GT准则对广东地震台网产出的地震观测报告进行了评估,获得了1160个GT2参考事件。     

应急流动观测组网技术在康定6.3级地震中的应用

以2014年11月22日四川康定发生6.3级地震后开展应急流动观测为例,描述了在应急流动观测组网中,如何应对震区高原环境,完成4个流动台选址架台、快速组网、运维保障;并通过对流动台址地动噪声功率谱密度分析计算,获得的台基达Ⅱ或Ⅲ级环境地动噪声水平,余震微小震检测能力与定位精度得到极大改善。流动台网运行期间,提高了震区监测台站覆盖面,记录到了大量余震微小震,加密台站对震区定位精度提升了量级,震源深度、震级更加准确。结果表明:一旦大震发生后,及时组织开展应急流动观测是必要的,而完成好一次大震后快速有效的应急流动监测,必须综合考虑震区地域环境、人员配备、预案选择、通信条件、设备可靠诸多因素与运维保障,采用固定与流动台联合组网方式,应急流动发挥效益最佳。     

Some evidences for earthquake occurrence in the shallow crust in some areas of North China and Southwest China

IntroductionThe hypocentral location parameters (the longitude and latitude of epicenter and the focal depth) are important fundamental data for the study of seismology and the earth interior physics. To locate the seismic source precisely is the basis for the study of seismicity patterns, relation between earthquake and active structures, engineering seismology, etc (WANG, et al, 1995). Among these parameters, the precise measurement of focal depth is closely related to the correct understa…     

华北、西南一些地区地震发生在地壳浅部的证据

依据区域小孔径数字地震台网对河北省唐山、张北、怀来，山西省大同和云南省剑川、禄劝、姚安等７次地震的定位或重定位结果，并结合相应区域个别大震用远场资料反演的震源深度结果，以及山西省临汾遥测地震台网重新测定的震源深度结果，与以往地震目录中给出的震源深度测定结果进行对比分析，发现小孔径台网的震源深度测定结果基本分布在地表至地下十几千米的范围内．而以往给出的测定结果中，震源深度大致分布在地表至地下30ｋｍ 的深度上，与其平均值相差一半至一倍，两者之间的差别非常显著．分析认为，小孔径数字地震台网的测定结果可能更接近震源深度分布的真实情况，即上述地区的地震应是发生在地壳浅层内．对造成上述情况的原因，进行了初步的讨论，并提出改善震源深度测定精度的建议和措施．      

基于模板匹配定位法的江苏盐城附近微震检测和构造分析

利用基于GPU加速的匹配定位法和双差定位法,对江苏盐城及邻区18个台站记录的2009~2018年共10年的连续地震资料进行分析。首先从台网目录中挑选211个地震事件作为模板事件,使用匹配定位技术对江苏盐城附近连续10年的地震进行检测和识别,共识别出1349个地震事件,约为台网目录地震事件的3倍,最小完备震级由台网目录的M_L1.9降为M_L1.2。然后利用双差定位法对检测到的地震事件进行精定位,精定位的结果揭示：建湖地区的地震密集带与洪泽-沟墩断裂有关,震源深度优势分布为5~20km,断裂两侧震源深度有显著差异,断裂带倾向NW;射阳震群震源深度比建湖震群有所加深,优势分布为10~25km,震源深度由南东向西北逐渐变浅;宝应地区地震丛集分布;东台地区由于模板事件相对较少,扫描定位后,地震事件在陈家堡-小海断裂带附近零星分布。研究结果为研究盐城地区的地震活动性、发震断层的深部构造提供了基础数据支撑。     

Location Performance and Detection Threshold of the Spanish National Seismic Network

Spain is a low-to-moderate seismicity area with relatively low seismic hazard. However, several strong shallow earthquakes have shaken the country causing casualties and extensive damage. Regional seismicity is monitored and surveyed by means of the Spanish National Seismic Network, maintenance and control of which are entrusted to the Instituto Geográfico Nacional. This array currently comprises 120 seismic stations distributed throughout Spanish territory (mainland and islands). Basically, we are interested in checking the noise conditions, reliability, and seismic detection capability of the Spanish network by analyzing the background noise level affecting the array stations, errors in hypocentral location, and detection threshold, which provides knowledge about network performance. It also enables testing of the suitability of the velocity model used in the routine process of earthquake location. To perform this study we use a method that relies on P and S wave travel times, which are computed by simulation of seismic rays from virtual seismic sources placed at the nodes of a regular grid covering the study area. Given the characteristics of the seismicity of Spain, we drew maps for M _L magnitudes 2.0, 2.5, and 3.0, at a focal depth of 10 km and a confidence level 95 %. The results relate to the number of stations involved in the hypocentral location process, how these stations are distributed spatially, and the uncertainties of focal data (errors in origin time, longitude, latitude, and depth). To assess the extent to which principal seismogenic areas are well monitored by the network, we estimated the average error in the location of a seismic source from the semiaxes of the ellipsoid of confidence by calculating the radius of the equivalent sphere. Finally, the detection threshold was determined as the magnitude of the smallest seismic event detected at least by four stations. The northwest of the peninsula, the Pyrenees, especially the westernmost segment, the Betic Cordillera, and Tenerife Island are the best-monitored zones. Origin time and focal depth are data that are far from being constrained by regional events. The two Iberian areas with moderate seismicity and the highest seismic hazard, the Pyrenees and Betic Cordillera, and the northwestern quadrant of the peninsula, are the areas wherein the focus of an earthquake is determined with an approximate error of 3 km. For M _L 2.5 and M _L 3.0 this error is common for almost the whole peninsula and the Canary Islands. In general, errors in epicenter latitude and longitude are small for near-surface earthquakes, increasing gradually as the depth increases, but remaining close to 5 km even at a depth of 60 km. The hypocentral depth seems to be well constrained to a depth of 40 km beneath the zones with the highest density of stations, with an error of less than 5 km. The M _L magnitude detection threshold of the network is approximately 2.0 for most of Spain and still less, almost 1.0, for the western sector of the Pyrenean region and the Canary Islands.     

Local seismic network for monitoring of a potential nuclear power plant area

This study presents a plan for seismic monitoring of a region around a potential nuclear power plant. Seismic monitoring is needed to evaluate seismic risk. The International Atomic Energy Agency has set guidelines on seismic hazard evaluation and monitoring of such areas. According to these guidelines, we have made a plan for a local network of seismic stations to collect data for seismic source characterization and seismotectonic interpretations, as well as to monitor seismic activity and natural hazards. The detection and location capability of the network were simulated using different station configurations by computing spatial azimuthal coverages and detection threshold magnitudes. Background noise conditions around Pyhäjoki were analyzed by comparing data from different stations. The annual number of microearthquakes that should be detected with a dense local network centered around Pyhäjoki was estimated. The network should be dense enough to fulfill the requirements of azimuthal coverage better than 180° and automatic event location capability down to ML?～?0 within a distance of 25 km from the site. A network of 10 stations should be enough to reach these goals. With this setup, the detection threshold magnitudes are estimated to be ML?=??0.1 and ML?=?0.1 within a radius of 25 and 50 km from Pyhäjoki, respectively. The annual number of earthquakes detected by the network is estimated to be 2 (ML?≥?～ ?0.1) within 25 km radius and 5 (ML?≥?～?0.1 to ～0.1) within 50 km radius. The location accuracy within 25 km radius is estimated to be 1–2 and 4 km for horizontal coordinates and depth, respectively. Thus, the network is dense enough to map out capable faults with horizontal accuracy of 1–2 km within 25 km radius of the site. The estimation is based on the location accuracies of five existing networks in northern Europe. Local factors, such as seismic noise sources, geology and infrastructure might limit the station configuration and detection and location capability of the network.     

福建台网接入台湾台站对定位台湾中深源地震精度影响分析

福建台网负责监测中国台湾地区地震。对于中深源地震使用何种定位方法能获得较好的地震参数,这直接影响到地震定位精度。利用JOPENS系统中交互分析软件MSDP提供的定位方法,对同一地震进行两次定位,即不使用和使用接入的台湾台站,将福建台网得出的两次结果与中国台湾公布的地震参数进行对比,分析定位精度,进而找出适用于台湾地区中深源地震的定位方法,以便进一步判断在地震速报中使用这些台站进行辅助定位的可行性,并给出相关的操作方法及建议。     

An algorithm for the optimum distribution of a regional seismic network—I

Summary An algorithm for the optimum distribution of any number of regional seismic stations is formulated. The information needed from a given region is the relative distribution of seismic intensity and the travel times of seismic waves used in earthquake location procedure. To define the optimum distribution of a regional network, some generalization of D-planning (minimization of the ellipsoid volume of earthquake location errors) was applied. The D-optimum criterion is generalized for a case where equations describing the process are nonlinear and when the most probable values of the sought parameters (origin time and earthquake focus coordinates) are not known.As an example of the algorithm application, the optimum distribution of seismic stations in the central district of the Lublin Coal Basin in Poland is given.     

Detection of earthquake clusters on the basis of waveform similarity: An application in the monferrato region (Piedmont, Italy)

In August 2000 and July 2001 two seismic sequences, characterized by mainshocks with Ml (local magnitude) respectively 5.1 and 4.8, occurred in the Monferrato region (Italy). The regional seismic network of North-western Italy (RSNI) recorded more than 250 foreshocks and aftershocks. The routine locations, obtained from the Hypoellipse code, show a seismic activity concentrated in a circular area, of a radius of about 15 km, located near Acqui Terme, and randomly distributed in depth. Location errors, due to an uneven azimuthal station distribution of the regional seismic network, prevent recognition of the geometry of the active zone. Waveform analysis revealed the presence of several multiplets. In order to discriminate and successively relocate them, an automatic cross-correlation procedure was applied. Normalized cross-correlation matrixes, for the RSNI stations which recorded almost 90% of considered events, on the basis of a signal to noise ratio analysis, were computed using only S wave time windows. The use of a relocation procedure, based on the double-difference method which incorporates ordinary absolute travel-time measurements and/or cross-correlation differential travel-times, allowed us to overcome the constraints of the uneven distribution of stations giving a quite different frame of seismicity. The improved locations showed that the seismic activity is mainly arranged along a NE-SW oriented volume, at a depth range of 8–20 km, involving the basement crystalline units. This orientation is confirmed by the analysis of the focal mechanisms: most focal solutions show a strike slip component with one of the nodal planes consistent with the main orientation of the seismic events.     

2021年11月17日江苏大丰海域MS5.0地震震源深度对比分析

选取江苏测震台网震中距2°以内18个地震台站记录清晰的波形资料,使用单纯形、LocSAT、HypoSAT地震定位方法,分别采用华南速度模型、郯庐断裂带中南段速度模型以及IASP91全球速度模型,重新测定江苏大丰海域M_S5.0地震震源深度,对获取的深度值进行对比分析。结果表明:3种地震定位方法测定的震源深度结果接近,约为11 ± 2 km;HypoSAT定位方法计算结果较为稳定,IASP91、郯庐断裂带中南段模型均适用于该区域震源深度计算。      

2009年4月19日新疆阿合奇5.4级地震

利用俄罗斯科学院比什凯克科学站地震台网中3个台站资料和新疆地震台网12个台站资料,分别用不同地震定位方法,对2009年4月19日中国新疆阿合奇5.4级地震和其3个余震震源参数进行重新定位。并且将其结果与新疆地震目录、中国地震台网统一地震目录进行了比较,对其宏观震中、发震构造、震源机制及地震类型等进行了分析与讨论。     

The 14 April 2012 Koyna Earthquake of M<Subscript>w</Subscript> 4.8: insights into active tectonics of the Koyna region

The 14 April 2012 earthquake of M_w 4.8 is the best monitored event in the Koyna region, a globally significant site of reservoir triggered seismicity in western India. Hence, investigation of this event assumes great importance, also considering its epicentral location close to that of the 1967 Koyna earthquake of M 6.3, the world’s largest reservoir triggered earthquake. Inversion of P-wave amplitude data along with the first motion polarities at 30 digital seismic stations provides a well-constrained strike-slip type focal mechanism solution, similar to that of the 1967 earthquake. The mechanism is further confirmed by moment tensor inversion of 3-component waveform data recorded at the three nearest broadband stations. The depth distribution of the aftershocks clearly delineates a NNE-SSW trending fault plane dipping about 78° to the WNW and coinciding with the trend of the Donachiwada fault, as well as the left-lateral fault plane of the focal mechanism solution obtained. The precise location, focal mechanism and the seismicity distribution from our dense network indicate that the activity in the Koyna region is mainly controlled by the NNE-SSW trending Donachiwada (D) fault zone rather than the Koyna River Fault Zone (KRFZ) on the west as suggested previously.     

基于双差定位方法的内蒙古中西部地区地震精确定位

针对测震台网定位存在偏差的问题,采用双差定位方法,对2008—2019年9月内蒙古测震台网及周边台站记录的内蒙古中西部地区4410个地震事件进行重新定位.研究表明,利用双差定位方法重新定位后地震分布更集中,更符合内蒙古中西部地区的地质构造特征,定位精度显著提高.     

Distribution Characteristics of the Seismicity of Zipingpu Reservoir Region after the Wenchuan Earthquake

815 earthquakes recorded by 12 seismic stations of the Zipingpu reservoir seismic network in 2009 were relocated using the double difference algorithm to analyze the seismic activity of the Zipingpu reservoir.Relocation results show that the earthquakes are concentrated relatively in three zones.The distribution characteristics of focal depth are obviously different among different concentration zones.This means earthquakes in different concentration zones may have different causes.Compared to relocation of...     

用单纯形法和PTD法对2016年新疆阿克陶M_S6.7地震序列的震源深度测定分析

用单纯形法计算了阿克陶M_S6.7主震的震源深度,结果为7.5 km,但单纯形法对设定的初始深度依赖值过高,震源深度值不一定可靠。用PTD法计算了阿克陶M_S6.7主震的震源深度,结果为9.4 km,且用了44个台站的震相数据共组成99个有效样本数参与了计算,不仅样本数多,而且震源深度样本数的集中度也比较符合高斯分布的特征,所以震源深度值可信度高,也非常接近国内和国外地震研究机构给出的此次地震的震源深度。当主震的震源深度已知且适合单纯形法设定的初值时,用单纯形法来计算地震序列的震源深度,不仅残差小,而且软件实现简单,确保了地震序列的按时提交。再用PTD法来检验该地震序列震源深度的准确性,结果显示,2种方法得到的震源深度差值不超过10 km,符合中国地震台网对震源深度误差的要求。     

稀疏台网反演区域地震偏量矩张量解的稳定性

使用区域地震台网的波形资料,研究了2010年2月26日台湾以东海域Mw5.4地震的偏量矩张量,所得结果与Havard CMT目录给出的解之间的Kagan角为9.5°,两者较为一致;反演得到这次地震的震源深度为41 km,接近USGS给出的定位结果,可能更接近真实的震源深度.选择不同的台站组合,分析了台站布局对矩张量解的...     

Application of an improved algorithm to high precision relocation of ISC test events

New location features for possible implementation by the International Seismological Centre in its standard location procedures are tested using a set of 156 well-located and geographically well-distributed earthquakes and explosions. The tests are performed using the Engdahl et al. ([Engdahl, E.R., Van der Hilst, R.D., Buland, R.P., 1998. Global teleseismic earthquake relocation with improved travel times and procedures for depth determination. Bull. Seism. Soc. Am. 88, 3295-3314]; EHB) location algorithm with the 1-D reference Earth model ak135 [Kennett, B.L.N., Engdahl, E.R., Buland, R., 1995. Constraints on seismic velocities in the Earth from travel times, Geophys. J. Int. 122, 108-124]. Weighting by phase variance as a function of distance improves location accuracy by 7%. Use of later phase arrival times does not result in a significant improvement in location or depth for events with observing stations well distributed in azimuth. However, with application of an improved phase identification technique, depth phases provide significantly better estimates of focal depth.     

利用sPn震相计算新源、和静交界区MS 6.6地震余震震源深度

以新疆地震台网波形数据为基础,研究2012年6月30日新源、和静交界区M_S 6.6地震序列中10个M_L 4.1—4.8余震。利用sPn震相计算10个余震的震源深度,将计算结果与CAP反演深度、单纯型法定位深度及Hyposat定位深度等结果进行对比发现,利用sPn震相所得震源深度与CAP反演深度一致性最好,与Hyposat定位结果也有较好的一致性,与单纯型法定位结果的一致性稍差。研究认为,对于新源、和静交界区地震利用sPn震相计算震源深度结果可靠,故该方法可以作为新源、和静交界区地震震源深度计算方法。     

2008年5月12日汶川MW 7.9地震的震源位置与发震时刻

综合运用四川省地震台网与紫坪铺水库地震台网的观测资料,精确地测定了2008年5月12日汶川MW7.9地震的震源位置与发震时刻．指出,对汶川地震这样的大地震精准定位,必须克服或尽量减少远台观测对地震精确定位的局限性、地壳介质模型的不完善性以及识别与检测初至波震相的不一致性等因素的影响．通过分析对比、反复试验,从上述台网中精心选取了方位分布均匀、具有近震源台站约束、直达P波震相确系由初始破裂辐射出的15个地震台的直达P波到时数据,反演得出精确度比区域性地震台网常规测定的精确度高一个数量级的汶川大地震的定位结果,即:发震时刻（北京时间）:2008年5月12日14:27:57.59plusmn;0.03 s;震中位置:31.018deg;Nplusmn;0.3 km,103.365deg;Eplusmn;0.3 km;震源深度:15.5 kmplusmn;0.3 km．