基于频陷理论的地震谱反演技术分析

Seismic spectral inversion technology based on frequency trap theory

我国陆相盆地广泛发育薄层砂体,但受地震分辨率限制,从地震剖面上难以直接识别薄层反射特征,为寻找岩性圈闭带来困难。谱反演技术的出现,为薄层砂体识别方法带来新的方向。该技术避开调谐振幅进行薄层识别的方法,采用在频率域中分析薄层频谱中出现的频陷特征,并且把地震信号分解为奇偶分量,进一步提高薄层的识别能力。通过模型分析,发现薄层厚度越小,对应信号频谱中频陷周期越大。地震信号奇偶分解后,其中偶分量能够提高地震资料的视分辨率。通过对谱反演目标函数的剖析,给出了具体实现过程。

Thin layer sandstone is widely developed in continental basin of China.Limited by seismic resolution,it is difficult to directly identify reflection characteristics of thin layer in seismic section and find depositional trap.The emergence of spectral inversion technology brings new direction for the thin layer sand identification.This technology avoids identifying thin layer sand by using tuning amplitude.It adopts the way of analyzing notch characteristic appear in thin layer spectrum in frequency domain,and decomposes seismic signal into odd and even components to further improve the ability to identify thin layer.Through model analysis,the results show that the smaller thickness of thin layer is,the greater notch period of corresponding signal spectrum is.Even component can increase apparent resolution of seismic data after decomposing seismic signal.Specific implementation of spectral inversion method was given by analyze objective function.