Earthquake detection and automatic phase picking
-
摘要: 针对微震事件易受噪声干扰等特点，本文将STA/LTA方法和基于方差的AIC方法（var-AIC）相结合，在震相到时初步拾取的基础上，使用台站的德洛内（Delaunay）三角剖分及台站间最大走时差约束来减少噪声干扰的影响. 利用到时进行地震定位之后，根据台站预测到时，在设定的时间窗内对地震震相进行更精细的分析. 特别是针对微震事件信噪比低的特点，设计了基于偏振分析的拾取函数，根据窗内STA/LTA方法和var-AIC方法的拾取结果自动选择合适的值作为震相到时. 最后，对西昌流动地震台阵2013年304个单事件波形数据的分析处理和检验结果表明，本文方法较传统方法具有更高的地震事件检测能力和更高的震相拾取精度.Abstract: With the continuous development of digital seismic networks and the application of mobile seismic array observation technology, arrival picking of seismic phases and detection of seismic events are paid more and more attention. Based on the preliminary results obtained by STA/LTA and AIC method, this paper combines the Delaunay triangulation and the maximum travel-time difference between stations to reduce the influence of noise. The arrival time predicted in every station is obtained by earthquake location on the basis of the preliminary results. Moreover, in accordance with the feature of low signal noise ratio of microseismic events, this paper developed a new phase picker which is based on the polarization. According to the results of STA/LTA method and AIC method in the time window, it automatically selects the appropriate values as arrival times. The analysis and processing of 304 seismic events recorded by Xichang portable seismic array in 2013 demonstrate that the new method proposed in this paper has stronger earthquake detection capability and higher seismic picking accuracy than traditional methods.
-
-
![]()
图 1 西昌地震台阵的德洛内三角剖分图
正方形为快速定位时网格搜索区域，星形为地震实例发生位置，三角形为台站位置
Figure 1. Delaunay triangulation of Xi’chang seismic array
The square represents the search area，the star stands for the hypocentral location，and triangles for stations
![]()
图 2 西昌地震台阵记录的2013年1月28日21时4分ML1.2地震
（a） 垂直分量波形；（b） 短长时窗均值之比；（c） 归一化的var-AIC函数；（d） Pol函数拾取P波；（e） Pol函数拾取S波；（f） 改良后的组合函数
Figure 2. Event of ML1.2 recorded by Xi’chang seimic array at 21:04 on January 28，2013
（a） Vertical component of waveform；（b） STA/LTA；（c） Normalized var-AIC；（d） P-wave picking by Pol-function；（e） S-wave picking by Pol-function；（f） Improved compound function
![]()
图 3 地震观测记录及拾取的震相到时图
（a） STA/LTA方法拾取的各种可能的P波震相到时；（b） 利用本文地震检测方法分析处理后获得的P波震相到时；（c） 用AK135模型计算的理论到时，其中紫线为P波预测到时，黄线为S波预测到时，蓝线为地震检测方法得到的P波震相到时，红线和绿线分别为取窗拾取的P波到时和S波到时，其中剔除了Pol函数小于C4的S波到时
Figure 3. Waveform records and arrival picking in different steps
（a） Probable P-waves picking by STA/LTA；（b） P-waves processing by the earthquake detection method in this paper；（c） Arrival time calculated by AK135 model，where purple lines denote arrival time of P-wave，yellow lines denote arrival time of S-wave，blue lines denote arrival time of P-wave processing by earthquake detection method，red and green lines denote P-wave and S-wave picking by setting windows in waveforms，S-wave of which Pol-function below C4 is deleted
![]()
图 4 实际震源位置与算法计算结果的对比
图中数字表示事件触发顺序
Figure 4. Comparison of hypocentral location with calculation result
The numbers represent the triggered order of events
![]()
图 5 各震相拾取方法精度对比
（a） 波形计算窗口；（b） 垂直分量波形；（c） 短长时比值；（d） var-AIC函数；（e） Pol函数；（f） 改良后的组合函数
Figure 5. Accuracy comparison of different phase picking methods
（a） Calculation window；（b） Vertical component of waveform；（c） STA/LTA；（d） var-AIC；（e） Pol-function；（f） Improved compound function
![]()
图 6 触发台站数不同时所拾取事件数与震级的关系
Figure 6. Relationship between the amount of events with magnitude with different amount of triggered station
![]()
图 7 遗漏事件占比与深度关系图
Figure 7. Relationship between the proportion of omitted events with focal depth
![]()
图 9 本文方法与人工拾取结果对比图
横坐标为相对预测到时的时间，纵坐标已对齐预测到时。红线为本文方法拾取到的P波震相，蓝线为人工拾取P波震相
Figure 9. Comparison of picking results by manual and the method proposed in this paper
Horizontal ordinate is relative time. Vertical ordinate has been lined up with forecasted arrival time. Red lines represent P-waves picking by the method in this paper，and blue lines represent P-waves picking by manual
![]()
图 10 不同方法震相拾取率与信噪比SNR的关系图
（a） 拾取误差小于0.1 s；（b） 拾取误差小于0.2 s；（c） 拾取误差小于0.3 s；（d） 拾取误差小于0.4 s
Figure 10. Relationship between phase picking rate with SNR by different methods
（a） Deviation below 0.1 s；（b） Deviation below 0.2 s；（c） Deviation below 0.3 s；（d） Deviation below 0.4 s
表 1 本文方法采用的相关参数
Table 1 Related parameters used in the method proposed in this paper
参数名 数值 单位 意义 C1 10 长短时方法获得到时的触发阈值 C2 1.5 长短时方法信号恢复时的触发阈值 C3 30 km 多台检测事件台间距的上限 C4 20 偏振分析Pol函数的触发阈值 C5 5 s 定位时的残差阈值 C6 10 震相存在判别的信噪比阈值 C7 1 000 震相存在判别的Pol函数阈值 C8 0.5 s AIC与STA/LTA结果接近的到时差范围 N1 30 s 长时平均值计算的滑动窗长 N2 0.5 s 短时平均值计算的滑动窗长 N3 1 s 为纠正STA/LTA拾取到时的AIC函数计算窗长 N4 4 s 代入AK135后拾取P波和S波到时的窗长 Lb 0.5 s 通过Pol函数计算S波到时设置的长窗长 Ls 0.2 s 通过Pol函数计算S波到时设置的短窗长 vP 5.8 km/s 检测地震事件时所取的P波速度 
下载: 导出CSV
表 2 3种地震检测方法对304个地震事件的检测效果对比
Table 2 Comparison of detection results for 304 events by three event detection methods
台站数 拾取正确事件数 拾取错误事件数 拾取重复事件数 遗漏事件数 拾取事件数 本文
方法台间距 德洛内三
角判别本文
方法台间距 德洛内三
角判别本文
方法台间距 德洛内三
角判别本文
方法台间距 德洛内三
角判别本文
方法台间距 德洛内三
角判别3 239 186 214 19 3 21 32 19 29 65 118 90 290 208 264 4 186 135 167 8 0 7 12 5 14 118 169 137 206 140 188 5 149 107 135 1 0 2 10 3 10 155 197 169 160 110 147
下载: 导出CSV
-
房立华, 吴建平, 明跃红, 王未来, 王长在, 杨婷, 蔡妍. 2014. 安宁河—则木河断裂带地震台阵观测及研究进展[C/OL]//中国地球科学联合学术年会. [2017-03-30]. http://115.29.6.223/files/work/jiyao30.pdf. Fang L H, Wu J P, Ming Y H, Wang W L, Wang C Z, Yang T, Cai Y. 2014. Anning river-Zemu river seismic array observation and research progress[C/OL]//Annual Meeting of Chinese Geoscience Union. [2017-03-30]. http://115.29.6.223/files/work/jiyao30.pdf （in Chinese）.
高淑芳, 李山有, 武东坡, 马强. 2008. 一种改进的STA/LTA震相自动识别方法[J]. 世界地震工程, 24（2）: 37–41. Gao S F, Li S Y, Wu D P, Ma Q. 2008. An automatic identifcation method of seismic phase based on modified STA/LTA algorithm[J]. World Earthquake Engineering, 24（2）: 37–41 （in Chinese）.
李慧婷, 黄文辉. 2000. 三分向数据震相的自动识别[J]. 华南地震, 20（4）: 46-53. Li H T, Huang W H. 2000. Automatic identification of three-component seismic phase[J]. South China Journal of Seismology, 20（4）: 46-53 （in Chinese）.
刘代志, 王仁明, 李夕海, 刘志刚. 2005. 基于小波包分解及AR模型的单通道地震波信号初至点检测[J]. 地球物理学报, 48（5）: 1098-1102. Liu D Z, Wang R M, Li X H, Liu Z G. 2005. Onset point identification of single-channel seismic signal based on wavelet packet and the AR model[J]. Chinese Journal of Geophysics, 48（5）: 1098-1102 （in Chinese）.
刘希强, 周蕙兰, 郑治真, 沈萍, 杨选辉, 马延路. 1998. 基于小波包变换的弱震相识别方法[J]. 地震学报, 20（4）: 373–380. Liu X Q, Zhou H L, Zheng Z Z, Shen P, Yang X H, Ma Y L. 1998. Micro-seismic phase recognition based on wavelet packet transform[J]. Acta Seismologica Sinica, 20（4）: 373–380 （in Chinese）.
刘希强, 周彦文, 曲均浩, 石玉燕, 李铂. 2009. 应用单台垂向记录进行区域地震事件实时检测和直达P波初动自动识别[J]. 地震学报, 31（3）: 260–271. Liu X Q, Zhou Y W, Qu J H, Shi Y Y, Li B. 2009. Real-time detection of regional events and automatic P-phase identification from the vertical component of a single station record[J]. Acta Seismologica Sinica, 31（3）: 260–271 （in Chinese）.
马强, 金星, 李山有, 陈绯雯, 廖诗荣, 韦永祥. 2013. 用于地震预警的P波震相到时自动拾取[J]. 地球物理学报, 56（7）: 2313-2321. Ma Q, Jin X, Li S Y, Chen F W, Liao S R, Wei Y X. 2013. Automatic P-arrival detection for earthquake early warning[J]. Chinese Journal of Geophysics, 56（7）: 2313-2321 （in Chinese）.
赵大鹏, 刘希强, 李红, 周彦文. 2012. 峰度和AIC方法在区域地震事件和直达P波初动自动识别方面的应用[J]. 地震研究, 35（2）: 220-225. Zhao D P, Liu X Q, Li H, Zhou Y W. 2012. Detection of regional seismic events by Kurtosis method and automatic identification of direct P-wave first motion by Kurtosis-AIC method[J]. Journal of Seismological Research, 35（2）: 220-225 （in Chinese）.
Akaike H. 1974. Markovian representation of stochastic processes and its application to the analysis of autoregressive moving average processes[J]. Ann Inst Stat Math, 26（1）: 363-387.
Akansu A N, Serdijn W A, Selesnick I W. 2010. Emerging applications of wavelets: a review[J]. Phys Commun, 3（1）: 1-18.
Allen R V. 1978. Automatic earthquake recognition and timing from single traces[J]. Bull Seismol Soc Am, 68（5）: 1521-1532.
Allen R. 1982. Automatic phase pickers: their present use and future prospects[J]. Bull Seismol Soc Am, 72（6）: S225-S242.
Anant K S, Dowla F U. 1997. Wavelet transform methods for phase identification in Three-Component seismograms[J]. Bull Seismol Soc Am, 87（6）: 1598-1612.
Bai C Y. 2000. Automatic phase-detection and identification by full use of a single three-component broadband seismogram[J]. Bull Seismol Soc Am, 90（1）: 187-198.
Baillard C, Crawford W C, Ballu V, Hibert C, Mangeney A. 2014. An automatic kurtosis-based P-and S-phase picker designed for local seismic networks[J]. Bull Seismol Soc Am, 104（1）: 394-409.
Bogiatzis P, Ishii M. 2015. Continuous wavelet decomposition algorithms for automatic detection of compressional-and shear-wave arrival times[J]. Bull Seismol Soc Am, 105（3）: 1628-1641.
Chen C, Holland A A. 2016. PhasePApy: a robust pure python package for automatic identification of seismic phases[J]. Seismol Res Lett, 87（6）: 1384-1396.
Dai H C, MacBeth C. 2007. Automatic picking of seismic arrivals in local earthquake data using an artificial neural network[J]. Geophys J Int, 120（3）: 758-774.
Flinn E A. 1965. Signal analysis using rectilinearity and direction of particle motion[J]. Proc IEEE, 53（12）: 1874-1876. doi: 10.1109/PROC.1965.4462.
Gentili S, Michelini A. 2006. Automatic picking of P and S phases using a neural tree[J]. J Seismol, 10（1）: 39-63.
Hibert C, Mangeney A, Grandjean G, Baillard C, Rivet D, Shapiro N M, Satriano C, Maggi A, Boissier P, Ferrazzini V, Crawford W. 2014. Automated identification, location, and volume estimation of rockfalls at Piton de la Fournaise volcano[J]. J Geophys Res Earth Surf, 119（5）: 1082-1105.
Jurkevics A. 2002. Polarization analysis of three-component array data[J]. Bull Seismol Soc Am, 78（5）: 1725-1743.
Kitagawa G, Akaike H. 1978. A procedure for the modeling of non-stationary time series[J]. Ann Inst Stat Math, 30（1）: 351-363.
Kurzon I, Vernon F L, Rosenberger A, Ben-Zion Y. 2014. Real-time automatic detectors of P and S waves using singular value decomposition[J]. Bull Seismol Soc Am, 104（4）: 1696-1708.
Leonard M, Kennett B L N. 1999. Multi-component autoregressive techniques for the analysis of seismograms[J]. Phys Earth Planet Int, 113（1/4）: 247-263.
Leonarda M. 2000. Comparison of manual and automatic onset time picking[J]. Bull Seismol Soc Am, 90（6）: 1384-1390.
Maeda N. 1985. A method for reading and checking phase time in auto-processing system of seismic wave data[J]. Zisin, 38（3）: 365-379.
Panagiotakis C, Kokinou E, Vallianatos F. 2008. Automatic P-phase picking based on local-maxima distribution[J]. IEEE Trans Geosci Remote Sens, 46（8）: 2280-2287.
Rosenberger A. 2010. Real-time ground-motion analysis: distinguishing P and S arrivals in a noisy environment[J]. Bull Seismol Soc Am, 100（3）: 1252-1262.
Saragiotis C D, Hadjileontiadis L J, Panas S M. 2002. PAI-S/K: a robust automatic seismic P phase arrival identification scheme[J]. IEEE Trans Geosci Remote Sens, 40（6）: 1395-1404.
Vassallo M, Satriano C, Lomax A. 2012. Automatic picker developments and optimization: a strategy for improving the performances of automatic phase pickers[J]. Seismol Res Lett, 83（3）: 541-554.
Vidale J E. 1986. Complex polarization analysis of particle motion[J]. Bull Seismol Soc Am, 76（5）: 1393-1405.
Wang J, Teng T L. 1997. Identification and picking of S phase using an artificial neural network[J]. Bull Seismol Soc Am, 87（5）: 1140-1149.
Withers M, Aster R, Young C, Beiriger J, Harris M, Moore S, Trujillo J. 1998. A comparison of select trigger algorithms for automated global seismic phase and event detection[J]. Bull Seismol Soc Am, 88（1）: 95-106.
Zhang H J, Thurber C, Rowe C. 2003. Automatic P-wave arrival detection and picking with multiscale wavelet analysis for single-component recordings[J]. Bull Seismol Soc Am, 93（5）: 1904-1912.
计量
- 文章访问数: 2094
- HTML全文浏览量: 1398
- PDF下载量: 157
