地震地质 ›› 2022, Vol. 44 ›› Issue (3): 604-624.DOI: 10.3969/j.issn.0253-4967.2022.03.004

• 研究论文 • 上一篇    下一篇

基于多通道互相关的青藏高原东北缘背景噪声程函层析成像

马小军1,2)(), 吴庆举1),*(), 潘佳铁1), 钟世军3), 徐荟1)   

  1. 1)中国地震局地球物理研究所, 北京 100081
    2)宁夏回族自治区地震局, 银川 750001
    3)北京市地震局, 北京 100080
  • 收稿日期:2021-05-31 修回日期:2021-11-03 出版日期:2022-06-20 发布日期:2022-08-02
  • 通讯作者: 吴庆举
  • 作者简介:马小军, 男, 1982年生, 现为中国地震局地球物理研究所固体地球物理专业在读博士研究生, 高级工程师, 主要从事地震背景噪声成像方面的研究, E-mail: maxiaojun2018@gmail.com
  • 基金资助:
    国家自然科学基金(42030310);中国地震局地震科技星火项目(XH19046Y);宁夏回族自治区重点研发计划项目(2018BFG02011)

AMBIENT NOISE EIKONAL TOMOGRAPHY BASED ON MUTI-CHANNEL CROSS-CORRELATION IN THE NORTHEASTERN MARGIN OF THE TIBETAN PLATEAU

MA Xiao-jun1,2)(), WU Qing-ju1),*(), PAN Jia-tie1), ZHONG Shi-jun3), XU Hui1)   

  1. 1) Institute of Geophysics, China Earthquake Administration, Beijing 100081, China
    2) Earthquake Agency of Ningxia Hui Autonomous Region, Yinchuan 750001, China
    3) Beijing Earthquake Agency, Beijing 100080, China
  • Received:2021-05-31 Revised:2021-11-03 Online:2022-06-20 Published:2022-08-02
  • Contact: WU Qing-ju

摘要:

文中利用布设在青藏高原东北缘及相邻区域的密集台阵1年的连续波形得到了背景噪声面波经验格林函数, 采用多通道互相关噪声面波程函成像方法, 反演得到了8~40s周期的相速度分布。与天然地震面波程函成像结果对比表明所得结果是可靠的, 具有高精度和高分辨率的特点; 与基于频时分析提取频散的传统程函成像方法对比, 文中的结果能够降低多路径散射面波和低信噪比带来的误差, 提高了反演结果的稳定性。研究结果显示, 河套-吉兰泰盆地下地壳—上地幔存在弱低速体, 地幔高温软弱物质的侵入可能造成了下地壳—上地幔的低速特征; 松潘-甘孜地块东北部下地壳的低速层可能暗示存在部分熔融, 结合其他结果推测可能不存在地壳管道流, 地幔软流圈物质的上涌可能是造成下地壳低速异常的原因; 祁连山西部的下地壳存在显著低速层, 南、 北块体挤压下的地壳缩短、 增厚和解耦应当形成了低速层。

关键词: 多通道互相关, 背景噪声, 程函层析成像

Abstract:

The traditional surface wave tomography method is a ray-theoretic travel-time tomography based on the high-frequency approximation, and adopts the regularization method with model smoothing parameters, which is likely to produce false anomalies. The current eikonal tomography is a geometrical ray theoretic method that can obtain the travel time gradient of the wave field by tracking the propagation of the wave front, and then get the slowness vector of wave field gradient. This method has the advantages of high efficiency and high resolution. But both surface wave travel-time tomography and traditional eikonal tomography need to extract dispersion curve. For example, the extraction of dispersion curve with auto frequency-time analysis method usually requires a manual extraction again, which may increase systematic error or human error. The multichannel cross-correlation surface wave eikonal tomography for earthquakes developed in recent years does not need to extract the dispersion curve, but automatically measures the relative phase delay between nearby stations based on waveform cross-correlations by using the far field condition of wave equation, and then inverts the two-dimensional surface wave phase velocity distribution with eikonal tomography method. This method can suppress the random incoherent noise and reduce bias caused by strong multipath scattering.

In this paper, we collected the one-year three-channel continuous waveform data from 676 temporary stations under the project China Array II and calculated the surface wave empirical Green’s function of ambient noise through noise cross-correlation from January to December 2015. The multichannel cross-correlation surface wave eikonal tomography was firstly applied to ambient noise tomography. The first step was to calculate the relative phase delay using the multi-channel cross-correlation, and at the second step, we inverted the Rayleigh wave apparent phase velocity at 8~40s periods based on eikonal equation for the whole study area, with the high resolution of about 40km in the major regions. At last, we compared our results with other results and discussed the tectonic deformation and dynamic process of the study area. The results are as follows:

(1)In contrast to traditional eikonal tomography method in which the dispersion has to be extracted based on frequency analysis, our results can reduce the bias resulting from multi-path scattering wave and low signal-to-noise ratio, and improve the stability of inversion results. Moreover, our results of long-period surface waves have higher accuracy and stability because our method reduces short-wavelength heterogeneity.

(2)There are obvious low-velocity anomalies in the upper crust of Hetao-Jilantai Basin at 18s period, and a weak low-velocity zone in the lower crust and upper mantle, which is associated with the upwelling of hot asthenosphere mantle materials in the “big mantle wedge”.

(3)A weak layer with low S-wave velocity exists in the middle and lower crust of the northeastern Songpan-Garzê block and the western Qilian orogenic belt. Receiver function results indicate that there is high Poisson’s ratio(0.28)and low P wave velocity(less than 6.3km/s)in the northeastern Songpan-Garzê block, which may suggest partial melting in the middle and lower crust of the northeastern Songpan-Garzê block; The radial anisotropy from ambient noise tomography in the western Qilian orogenic belt shows negative radial anisotropy characteristics, which may be associated with the crustal shortening, thickening and coupling under the compression from the north and south blocks.

Key words: multi-channel cross-correlation, ambient seismic noise, eikonal tomography

中图分类号: