3种远震P波到时拾取方法的比较及其参数优化

分析了自回归赤池信息准则（AR-AIC）、高阶统计量（HOS）和累积和（CUSUM）等3种到时拾取方法中参数对远震P波到时估计的影响,以450个远震P波信号为样本集,参考人工拾取到时,以网格搜索方式确定了每种方法的最优参数。之后重新选取信噪比处于[2,20]区间的100个远震P波信号,用确定最优参数后的3种到时拾取方法估计其P波初至时间,并比较了3种方法对低信噪比远震P波的拾取准确度。结果表明,AR-AIC方法和CUSUM方法对低信噪比远震P波的拾取准确度要优于HOS方法,CUSUM方法的计算速度最快,HOS方法由于其原理的限制更适用于信噪比较大、初动较尖锐的信号。      

远震相对走时数据快速计算方法及应用

远震层析成像技术已成为研究地球深部速度结构最为主要的方法,而其所用数据通常为初至震相的相对走时残差值.本文针对信噪比差的远震波形资料提出一种计算远震相对走时残差数据的快速方法,整个计算过程全部由计算机自动完成.该方法主要有两大特点:(1)利用互相关方法计算波形间的偏移时间时引入波形相位信息,有效压制噪声,增强有用震相的识别;(2)改进常规方法中计算相对走时残差的理论公式,避免求解超定方程组,提高计算精度.为验证新方法的可行性和实用性,以长江中下游地区远震波形资料为例,采用常规方法和新方法分别计算相对走时残差数据,对比结果显示:新方法比常规方法的工作效率和数据精度都高.     

由地方台网记录确定远震信号的方位角和慢度

介绍一种新的适合于地方遥测台网数据处理的方法，估算远震信号到达台站的方位角和慢度．该方法是根据信号在各台站上的到时与台站位置矢量在信号传播方向上的投影之间的相关性的原理．实际分析结果表明，应用该方法对地方台网的记录进行处理时可以准确地计算出信号的方位角和慢度，并能准确快捷地对地方台网记录的远震信号进行解释．      

远震层析成像在柴西狮子沟构造背景研究中的应用

利用2004——2007年在柴达木盆地西部狮子沟地区布设的地震台站的连续观测资料,挑选远震P波到时,采用远震体波走时层析成像方法进行反演处理,获得了该地区地下三维速度结构图象．反演结果表明,在研究观测地区西南侧存在一近北西走向的低速异常体,以较大倾角向北东方向插入;在该低速异常体的北部、东北部和东部环绕着相对高速的异常,其走向和分布范围基本与狮子沟背斜构造一致．      

数字化地震记录震相自动识别的方法研究

针对目前震相自动识别方法不能自动给出震相识别区间,以及不能确定识别出震相名称的问题,运用震相的运动学特征,由平均速度模型和J-B走时表数据,自动计算近震、远震和极远震的震相走时及震中距。对多尺度小波分解进行单支重构作为识别不同震相的分析信号。先求出初至震相和最大面波到时,估算出震中距,然后找到S波或PP波到时,求出准确的震中距,即可自动给出各震相的识别区间,采用线性偏振法在给定区间中识别出震相的初至时刻。由于该方法采用的是先明确要找什么震相,再由该震相的走时确定寻找区间,所以找出的初至就是要寻找的震相,自然解决了识别出震相的名称问题,从而实现了对震相的全程自动识别。     

利用地方震、远震走时和面波数据联合反演日本俯冲带P波和S波层析成像

通过最新收集的大量高质量的地方震和远震事件的到时数据进行联合反演,我们确定了日本俯冲带约700km深度的P波和S波速度层析成像。我们还使用远震瑞利波的振幅和相速度,确定了日本及其附近海域下方20~150s周期基阶瑞利波的二维相速度图像。研究区精细三维S波层析成像可通过地方震和远震事件的S波到时,及瑞利波相速度数据进行联合反演得到。我们的反演结果揭示:一维原始速度模型中,俯冲太平洋板块和菲律宾海板块呈现明显的高速区。在板块上方的地幔楔和太平洋板块下方的地幔中存在显著的低速异常。俯冲板块和周围地幔之间速度有明显的差异,表明温度、水含量和/或部分熔融程度有显著的横向变化。地幔楔低速异常是由板块脱水作用和地幔楔拐角流造成。在日本东北太平洋板块下方显示片状的低速区,这可能反映了地幔深部热上涌以及地幔柱软流圈的俯冲作用。我们的结果表明不同的地震数据联合反演,对于得到地壳和地幔可靠的层析成像图像是非常有效和重要的。     

论远震分析（讲座）

本文是１９９３年大震速报培训班课的一部份。主要讨论了仅用某上个地震台的地图分析远震和极远震震相方法问题，包括地震波和震相的异同，地震图的复杂性，震相特征和分析方法。     

云南省区域数字地震遥测台网远震参数测定

利用云南省区域数字地震遥测台网记录到的资料对远震的定位方法进行了研究。采用“分离变量法”,将震源参数分为发震时间、震中位置和震源深度三类变量,相对独立进行计算,这就减少了Geiger法中,各类变量因不同权重而产生的互相干扰以及因初始解不好引起的迭代不收敛的问题,改善了定位精度。由于大地震的地震波初动清楚,可精确测定初动到时,仅用经过台站高程走时改正和台站综合走时改正后的初动到时即可精确测定零深度时的震中位置。再根据测定的零深度震中位置,查出台网最小到时子台的S波到时S0它与该台的实际S波到时Sh的差可有效判定地震是否深源地震和计算震源深度。为了保持地震震级定义的一致性,采用数字地震仪系统的传递函数(包括幅频特性和相频特性)对实测数字地震记录进行仿真,得到真正的地动速度,再由计算机自动检测定地动速度最大面波振幅并计算震级Ms。     

山西地区远震P波到时残差分布

收集2008-2016年山西地震台网记录的震中距30°-90°范围内1 253个远震事件波形,拾取7 600余条高质量P波初至到时,使用IASP91模型计算相对到时残差,分析残差水平分布特征,结果显示：①以山西地区中部的山西断裂带为界,西部地震台站记录的P波初至主要表现为早到时,东部位于大同火山区的地震台站记录则主要表现为晚到时;②位于山西断裂带内部的地震台站记录的P波初至主要表现为早到时,残差水平显著低于西部地震台站;③研究区P波到时整体呈现自西向东逐渐由早到晚的分布特征。推测山西断裂带西部地区下方可能存在高速异常结构,山西断裂带内部及大同火山区下方可能存在低速异常结构。     

山西地区远震P波到时残差分布

收集2008—2016年山西地震台网记录的震中距30°—90°范围内1 253个远震事件波形,拾取7 600余条高质量P波初至到时,使用IASP91模型计算相对到时残差,分析残差水平分布特征,结果显示:①以山西地区中部的山西断裂带为界,西部地震台站记录的P波初至主要表现为早到时,东部位于大同火山区的地震台站记录则主要表现为晚到时;②位于山西断裂带内部的地震台站记录的P波初至主要表现为早到时,残差水平显著低于西部地震台站;③研究区P波到时整体呈现自西向东逐渐由早到晚的分布特征。推测山西断裂带西部地区下方可能存在高速异常结构,山西断裂带内部及大同火山区下方可能存在低速异常结构。     

Arrival time measurements of first arrival phases P and PKIKP using the method of fixed scale wavelet transformation ratio

Introduction Usually there are abundantly transient high-frequency compositions and weak amplitudes at the onset of a gradually changing signal. Using the method of the fixed scale wavelet transforma- tion ratio, we tried to display and measure the onset time or first arrival time. Unfortunately, the ratio method is lack of the stability. To reduce the noise influence on the ratio, reveal the arrival time information and improve the measurement precision of the arrival time of the gradually ch…     

区域震相初至估计

本在地震数据自动化处理中，给出一种基于自回归模型的Akaike information criteria(AIC)算法和信号平均幅值比的混合方法来估计地震信号的初至．用信号的AIC曲线和平均幅值比曲线构造一种叠加曲线，再进行类似于坐标旋转的校正，可以准确估计低信噪比记录中信号的初至，尤其对于震相类型比较复杂的后续震相(如S波、Lg波)的初至估计结果很好．通过对中国数字地震台网乌鲁木齐台记录到的23次天然地震中P波、S波和Lg波的初至估计，与人工分析结果相比，P波初至估计的均方误差为0．71s，后续震相(S波、Lg波)的均方误差为1．64s，优于传统AIC算法的估计结果．     

首都圈地区莫霍面起伏及沉积层厚度

利用首都圈数字地震台网宽频带和短周期记录提取了接收函数,用H\|Kappa叠加方法反演得到了台站下方地壳厚度和泊松比.反演结果显示首都圈地区莫霍面的区域构造方向呈北东或北北东向展布,地壳由东南向西北方向逐渐增厚,平均厚度为34 km.地壳泊松比分布呈现出分块的特征,泊松比高值区对应于岩石比较破碎的多条活动断裂带交汇区.而某些地区堆积有较厚的低速松散沉积层,其下的结晶基底介质速度相对较高,因此,该界面造成径向直达P波能量非常弱,而紧随其后的转换波能量较强,称其为首到波峰.通过正演计算,建立首到波峰和直达P波到时差与沉积层厚度的定量关系,从而可根据首到波峰相对于直达P波的时间延迟来快速判定沉积层的大致厚度.     

Estimation of azimuth and slowness of tele-seismic signals recorded by a local seismic network

Introduction The azimuth and slowness are two major features of seismic signals. The accurate estimation of them is quite important for both phase identification and event location. Generally, there are two types of seismic stations, i.e. 3-component stations (3C) and arrays. To estimate the two direc-tional parameters, the polarization analysis (Jurkevics, 1988) is commonly used for 3C stations and the frequency-wavenumber spectrum analysis ( f-k) (Capon, 1969; Kvaerna, Doornbos, 1986) is …     

Tomographic imaging of P wave velocity structure beneath the region around Beijing

The three-dimension crustal and upper mantle structures in the region around Beijing were studied by seismic tomography. We used the P wave arrival times from local and teleseismic events. These events were recorded by 250 stations of the North China Seismic Array and 108 stations of the Beijing Telemetry Seismic Network. 118 869 P wave arrivals from 10 285 local events and 12 189 P wave arrivals from 107 teleseismic events were used in the inversion. We obtained the 3-D P wave velocity structure of the cru...     

Travel time source-specific station corrections related to lithospheric structures in the Mediterranean region

We compare the locations obtained from arrival times collected by the International Seismological Centre from a network of regional and teleseismic stations for a cluster of Italian earthquakes with the locations of the same events obtained by the dense national seismic network operated by the Istituto Nazionale di Geofisica e Vulcanologia. We find mislocations on the order of 15 km for epicentral coordinates and on the order of 25 km for depths calculated from the regional and teleseismic network and using the standard IASP91 travel times. These mislocations are generally larger than the sizes of the respective error ellipse semi-axes. We then show that systematic shifts of hypocentral coordinates can be substantially reduced by applying source-specific station corrections. Moreover, we find that the size of error ellipses characterizing the teleseismic locations is significantly reduced by the application of such corrections. Our travel time corrections are compared and found fairly consistent with information available in the literature on tomographic studies on the crust and upper mantle in the European-Mediterranean region.     

VARIATION OF THE TRAVEL TIME DIFFERENCE BETWEEN TELESEISMIC PS CONVERTED WAVE AND PP TRANSMITTED WAVE IN FOCAL REGION BEFORE AND AFTER WENCHUAN MS8.0 EARTHQUAKE

When P waves from distant earthquakes meet a velocity discontinuity in the earth's crust and upper mantle, they give rise to a series of converted PS waves besides PP refracted waves. It is possible to monitor the variation of the physical properties of the medium in the limited formation space above the transition zone of the seismogenic zone by measuring the time difference between the teleseismic PS converted wave and the first arrival P wave, that is, time-variation Δt_PS=t_PS-t_P. The advantage of this method is that the transition point of the teleseismic source with similar source is relatively stable at the transition interface, and the accuracy of the measured relative time is high, and the change of the medium in a small range of the seismogenic zone above the conversion interface can be monitored. This paper studies the variation of the travel time difference Δt_PS in focal region before and after Wenchuan M_S8.0 earthquake. We select 2001 to 2012 as the research period, use teleseismic waveforms which occurred in the southern region of Sumatra and Hindu Kush area recorded by Sichuan station YZP and JJS. These teleseisms satisfy 5.0 ≤ M ≤ 6.5, and their waveform signal-to-noise ratio is high with clear initial P-wave motion. The epicentral distance of teleseisms is less than 3 degrees. Then we obtain the variation of the travel time difference Δt_PS between teleseismic PS converted wave and PP transmitted wave recorded during the study period in the two stations. The results show that there is a slow increasing trend of Δt_PS before 2006, and an obvious low value process of Δt_PS appeared in the period about 2 years before the Wenchuan earthquake. The maximum decline was about 0.2~0.3s, more than 4~5 times the measurement error. The low value has a certain degree of return about 2~3 months before the earthquake. The change of arrival time difference indicates that the medium is in different states in different periods of seismogenic process. The sharp decrease in Δt_PS from 2006 to January 2008 may be due to the strong disturbance caused by the stress accumulation of the medium. At this stage, the velocity of P wave and S wave increases with the increase of stress, and the increase of S wave velocity will result in the decrease of Δt_PS. The change of Δt_PS is greatly affected by S wave velocity, so Δt_PS appears to decrease rapidly. Regarding the low value that has a certain degree of return about 2~3 months before the Wenchuan earthquake, the possible reason is that the release of stress is much higher than the accumulation of stress in meta-instability stage. At this stage, the velocity of S wave decreases and the decrease of S wave causes Δt_PS to increase. Then, the Wenchuan earthquake of magnitude 8.0 occurred. It is shown that the teleseismic converted wave method in this paper can monitor the variation of medium's wave velocity before large earthquakes, and it has a good prospect in seismic monitoring and worth further experimental study.