基于波形相似性的远震初至拾取方法进展与对比研究

发布时间：2018-03-17 03:09

本文选题：初至拾取　切入点：波形相似性　出处：《地球物理学进展》2017年04期 　论文类型：期刊论文

【摘要】：远震初至拾取是地震学研究中的关键环节,它在地震定位、核爆炸检测以及远震层析成像中起着重要的作用.基于波形相似性的远震初至拾取方法利用互相关估计各地震道的相对时差,具有抗噪和拾取精度高的特点.目前流行的各种基于波形相似性的远震初至拾取方法的原理各不相同,在时窗参数设定、稳定性和计算效率等方面存在差异,有必要对不同方法进行总结,并对其精度和效率进行评价.本文针对多道互相关方法、迭代互相关叠加方法、自适应叠加方法、台网远震体波处理方法、改进多道互相关方法等五种基于波形相似性的初至拾取典型方法的特点,对比分析了各方法的精度、稳定性及其与时窗参数的关系.实际数据的应用结果显示:在相同数据条件下,改进多道互相关方法精度最高,台网远震体波处理方法次之,再次是多道互相关方法和迭代互相关叠加方法,自适应叠加方法最低.不同方法的稳定性对互相关时窗长度变化的敏感度不同,时窗过短或者过长均会造成拾取精度和稳定性的降低.当初至震相信号的时窗长度取主周期的1~2倍时,互相关时窗恰好覆盖初至震相起跳部分的波形,各种方法拾取初至的精度接近.在实际应用中,建议根据不同的震源子波类型、频率成分、噪声水平等采用不同的初至拾取策略以提高拾取的精度和稳定性.
[Abstract]:Picking up the first break of a distant earthquake is a key link in seismological research, and it is used to locate the earthquake. Nuclear explosion detection and remote seismic tomography play an important role. Using cross-correlation to estimate the relative time difference of seismic traces, the method of first arrival pickup based on waveform similarity is used to estimate the relative time difference of each seismic trace. It has the characteristics of anti-noise and high precision. At present, the principle of various popular methods based on waveform similarity is different, and there are differences in parameter setting, stability and calculation efficiency of time-window, etc. It is necessary to summarize different methods and evaluate their accuracy and efficiency. This paper aims at multi-channel cross-correlation method, iterative cross-correlation superposition method, adaptive superposition method, remote seismic body wave processing method. The characteristics of five typical first break picking methods based on waveform similarity are improved, and the accuracy of each method is compared and analyzed. The application results of the practical data show that under the same data condition, the improved multi-channel cross-correlation method has the highest accuracy, and the remote seismic body wave processing method is the second. Thirdly, the multi-channel cross-correlation method and iterative cross-correlation superposition method, the adaptive superposition method is the lowest. The stability of different methods has different sensitivity to the cross-correlation window length change. If the window is too short or too long, the pickup accuracy and stability will be reduced. When the length of the window of the seismic phase signal is 1 ~ 2 times of the main period, the cross-correlation window exactly covers the waveform from the initial phase to the take-off part of the seismic phase. In practical application, it is suggested that different first arrival picking strategies should be adopted according to different source wavelet types, frequency components and noise levels in order to improve the accuracy and stability of pick-up.
【作者单位】：中国海洋大学海洋地球科学学院;海底科学与探测技术教育部重点实验室;
【基金】：山东省自然科学基金(ZR2014DM006) 国家自然科学基金(41230318) 国家重大专项(2016ZX05006002-001) 教育部留学回国基金中国海洋大学科研专项联合资助
【分类号】：P631.4

【相似文献】