三峡地区上地壳结构的远震虚震源反射地震成像

远震虚震源反射成像方法利用远震初至波在台网之下地表与地下界面间形成的反射波(PPdp震相)波形资料进行台网地区地震反射结构研究.此方法先用台网各台站的平均初至波形求取震源信号,再用该震源信号与各道地震记录作反褶积从而取得反射剖面.本文介绍了远震虚震源反射成像的基本原理和实现步骤,并以三峡地区的观测资料为例,得到三峡库首区上地壳的反射地震剖面.成像剖面中解释出的四川盆地和秭归盆地的底界面位置和形态与地表地质观测和大地构造背景吻合.为验证虚震源成像的能力,本文使用弹性波正演模拟合成地震数据,经过处理实际资料一样的步骤获得虚震源成像结果.正演模拟表明,叠加多个远震的反射地震剖面可以有效地改善成像的信噪比和连续性;虚震源成像需要选择特定的震源频率范围,以减轻高频噪音以及低频造成的成像问题.在研究上地壳结构时,建议在不适合主动源采集和缺乏低频信号的地区尝试远震虚震源方法.

Imaging upper crustal structure of the Three Gorges region via teleseismic virtual reflection profiling

Former studies suggest that it is possible to establish the seismic reflection image of the upper crust using the high-frequency teleseismic waves. Here we introduce the principles and basic procedure of the teleseismic virtual-source reflection (TVR) profiling method, and use it to image the upper crustal structure of the Three Gorges (TG) region, China. On the basis of the reflector images and the existing velocity model we establish a layered velocity model, which is used in elastic forward modeling to generate the synthetic teleseismic record. By use of the synthetic data the accuracy and the stability of the TVR are analyzed. The method of TVR profiling uses free-surface reverberations of the teleseismic P-wave (PPdp phase) data to produce a seismic reflection image of the subsurface. The TVR method first estimates the source wavelets by averaging the teleseismic first-arrival waveforms of all stations in an array, and then constructs the reflection images by deconvolving the averaged source wavelet from the teleseismic P-wave coda recorded at each station. In the summer of 2011 we deployed a seismic network in the TG region. The recorded teleseismic data were used as the input of TVR method to generate the upper crustal images of the TG region. By controlling the signal-to-noise ratio, frequency content, and the incidence angle the seismogram of teleseismic events with difference hypocentral distances are synthesized and used for analyzing the impact of those factors to the reflection images established by TVR.#br#By comparing the TVR profile and the surface geologic units we interpret the upper crustal structure of the TG region. The interpreted Sichuan basin and Zigui basin correlate well with the surface geology and the tectonic background. We found that the sediments of Zigui basin thicken toward east, and reach to 2.6 km at the thickest point. However, the Sichuan basin becomes thinner toward east. The Sichuan basin is truncated by the Huangling anticlineand is exposed on surface to the west of Huangling anticline. The TVR profile also shows that there is a low-velocity zone at the bottom of the Sichuan basin, supporting the existence of a strong seismic velocity gradient zone. We suspect that the velocity gradient zone is a fracture zone caused by the eastward movement of Sichuan basin. Further geophysical evidences are required to support this interpretation. To test the TVR algorithm we use the elastic forward modeling. The results of forward modeling show that the TVR method can be applied to image the upper crustal structure of the TG region. It also shows that stacking images of multiple-offset teleseismic event can improve the signal-to-noise ratio and the continuation of reflectors. By analyzing the relation between seismic frequency and the quality of TVR images we found that there is an optimal frequency range for TVR. In seismic record the high frequency is usually dominated by the environmental noise, while the low frequency will hurt the resolvability of deconvolution for shallow crustal structure. By use of the teleseismic data and the TVR method we constructed the upper crustal seismic profile of the TG region. The imaged bottoms of the Zigui Basin and the eastern Sichuan basin show the interaction between the sedimentary basins and the crystalline basement. The imaging accuracy and stability can be evaluated by the forward modeling. It is suggested using the teleseismic signal within the frequency band of 1~2 Hz for TVR imaging to reduce the impact of low frequency on the imaging resolution and the influence of high-frequency noise. The TVR method can be applied to those regions that are unsuitable for active seismic surveys and/or lack low frequency data.