基于Python语言的ObsPy软件包从地震背景噪声中提取瑞利面波经验格林函数的实行方案

Efficient scheme to extract Green functions of Rayleigh wave from seismic noise via a Python library for seismology-ObsPy

针对海量地震数据的高效自动处理是现代地震学重要的研究课题之一.本文基于Python语言的地震学数据处理软件平台ObsPy重构了背景噪声成像数据处理工作流程方案.以大兴安岭地区的不同来源与不同格式的地震数据为例,利用ObsPy软件包实现了背景噪声成像去除仪器响应、重采样、滤波、去均值趋势等预处理过程,以及One-Bit时间域归一化、谱白化、互相关计算等主要工作流程.对比测试表明基于ObsPy软件包的新流程与常用的噪声成像软件包的计算效率具有可比性,但代码更加简洁易用.同时,基于Python语言实现了利用Bootstrap方法对背景噪声成像提取面波频散曲线的误差分析.最后,本文利用归一化背景能量流方法分析了大兴安岭地区相关噪声源的时空分布.分析结果显示,研究区接收的短周期噪声可能与太平洋浪高的季节变化相关,接收的长周期噪声可能与全球海洋浪高的季节变化相关.

Automatic and efficient processing massive seismic data has been one of the most important topics in modern seismology.We rebuilt a new and efficient data processing scheme in ambient noise tomography via a Python seismological library of ObsPy,which execute a set of processes include removing response,resampling,filtering,removing trend,one-bit normalization,spectral whitening and cross correlation computation for seismic data with various format acquired from different seismic arrays deployed in the Great Xing’an Range area.The analysis of computational efficiency indicates that our new procedure which was based on ObsPy runs as fast as traditional software but with lucid and easy-to-use codes.Also,we perform an error analysis of surface wave dispersion measurements for ambient noise tomography using bootstrap method with ObsPy as well as NumPy.Finally,the features of spatial and temporal variations of correlated noises were studied by computing seasonal Normalized Background Energy Flow(NBEF)through the networks.The azimuthal distributions suggested that the shorter period noises might be related to seasonal variations of wave height of the Pacific Ocean,while longer were related to seasonal variations of wave height of the global oceans.