流动台阵地震检测与震相自动拾取研究

发布时间：2018-06-06 00:06

本文选题：流动 + 台阵　；参考：《中国地震局地球物理研究所》2017年硕士论文

【摘要】：地震震相到时信息是地震学研究的重要基础数据,广泛应用于地震定位、走时层析成像和地震预警等多个不同研究领域。随着固定地震台网台站密度的增加及流动地震台阵观测的广泛应用,传统的依靠人工分析拾取震相的工作量变得十分巨大,也无法满足现代社会对地震信息不断增长的实时性要求,发展高效可靠、抗干扰能力强的地震震相自动拾取和地震自动检测技术,成为地震学研究领域的重要发展方向之一。本文以西昌流动地震台阵观测资料为基础,针对观测记录中噪声干扰较强的特点,开展了地震事件自动检测和震相自动拾取方法研究:1、归纳总结了常用的震相拾取方法,对长短时(STA/LTA)、赤池信息准则(AIC)、偏振分析、奇异值分解(SVD)、小波分析和人工神经元等方法进行了重点分析,分析了这些方法的主要优缺点和研究现状。2、提出了采用STA/LTA和VAR-AIC相结合的方法进行震相到时初步拾取,利用Delaunay三角剖分与台间距走时差约束剔除噪声干扰的方法,较好地抑制了噪声干扰的影响,提高地震检测效果。3、开展了STA/LTA方法、VAR-AIC方法和偏振方法拾取震相到时的对比研究,设计了一种根据STA/LTA方法和VAR-AIC方法拾取的震相到时差判定信号强弱,进而自动选择VAR-AIC方法或偏振方法拾取到时的方法。与单一使用VAR-AIC方法或偏振方法相比,新方法具有更高的整体震相拾取精度。4、为了提高震相的拾取率,设计了基于地震检测结果的震相拾取方法。该方法包含四部分:(1)利用信噪比较高的震相到时进行地震定位获得地震的震源位置;(2)根据震源位置,计算所有台站的预测到时;(3)通过设计特定的震相判断标准,在各台站预测到时附近设定时间窗进一步确定是否存在震相;(4)若存在震相,则采用STA/LTA方法、VAR-AIC方法和偏振分析方法综合拾取震相到时。5、根据上述方法,在Matlab环境下开发了一套地震检测和震相识别软件,对西昌流动地震台阵2013年304个事件波形数据进行了分析处理和检验,结果表明:(1)与单纯的Delaunay三角剖分约束相比,本文提出的地震检测方法具有更强的地震检测能力;(2)新的震相拾取方法不仅可以检测到信噪比较高的人工遗漏震相,而且可以检测到一些人工拾取过程中容易忽略的低信噪比震相。
[Abstract]:Seismic phase arrival information is an important basic data in seismology. It is widely used in many different research fields such as seismic location, travel time tomography and earthquake warning. With the increase of the density of fixed seismic network stations and the wide application of mobile seismic array observation, the workload of traditional manual analysis to pick up seismic phases becomes very large. It is also unable to meet the real-time requirements of increasing seismic information in modern society. It has become one of the important development directions in the field of seismology to develop automatic seismic phase pickup and seismic automatic detection technology with high efficiency and reliability and strong anti-interference ability. Based on the observation data of Xichang mobile seismic array, aiming at the strong noise disturbance in the observation records, this paper has carried out the research on automatic detection of seismic events and automatic picking up of seismic phases, and summarized the commonly used seismic phase pick-up methods. The methods such as STAP / LTA, red cell information criterion (AIC), polarization analysis, singular value decomposition (SVD), wavelet analysis and artificial neuron are analyzed in detail. In this paper, the main advantages and disadvantages of these methods and the present research situation are analyzed, and a method of phase arrival preliminary pick-up using the combination of STA-LTA and VAR-AIC is put forward, and the method of eliminating noise interference by Delaunay triangulation and time-difference constraint of station spacing is proposed. The effect of noise interference is well suppressed, and the seismic detection effect is improved. A comparative study is carried out on the phase arrival time of the phase pickup by the STAP / LTA method and the polarization method. A method is designed to determine the signal strength according to the phase to time difference picked up by the STA-LTA method and the VAR-AIC method, and then to automatically select the VAR-AIC method or polarization method to pick up the arrival time. Compared with the single VAR-AIC method or polarization method, the new method has higher precision of phase pickup. In order to improve the picking rate of seismic phase, a phase pickup method based on seismic detection results is designed. The method consists of four parts: 1) the focal location of the earthquake is obtained by using the phase arrival time with a high SNR) according to the source location, the prediction time of all the stations is calculated. Setting a time window near the prediction time at each station to further determine whether there is a phase or not, if there is a phase, then the phase time. 5 is comprehensively picked up by using the STAP / LTA method / VAR-AIC method and the polarization analysis method. According to the above method, A set of software for seismic detection and phase identification is developed in Matlab. The data of 304 event waveforms of Xichang mobile seismic array in 2013 are analyzed and verified. The results show that: 1) compared with the simple Delaunay triangulation constraint, The seismic detection method presented in this paper has a stronger seismic detection ability. The new phase pickup method can not only detect the artificially missing phases with high SNR. Moreover, some low SNR phases which are easy to be ignored during manual pickup can be detected.
【学位授予单位】：中国地震局地球物理研究所
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P315.7

【参考文献】