统计F-X谱估计方法及应用

Statistical estimation of F-X spectrum and its application

原始地震数据的频率成分受多种因素的影响,在处理过程中数据的频率成分也在发生变化,因此,频率特征分析是地震数据处理中重要的分析工作之一。简要介绍生产中常用的频率分析方法,即频谱分析和频率扫描,说明了其特点和局限性。提出了统计F-X谱估计方法,该方法基于地表一致性假设前提,可进行多域分析和统计,在傅里叶变换之前,采用随机滑动时窗的方式提高谱估计的精度,增强谱估计的稳定性,并在F X域中提取特征参数,便于量化分析地震数据的频率变化规律。

The frequency content of raw seismic data may suffer from many factors.It may be changed during data processing.Therefore,characteristics analysis of frequency is an important work to be carried out in seismic data processing.Based on an analysis of the commonly used methods,that is the spectrum analysis and frequency sweeping for frequency analysis,for their characters and limitations,the author put forward a method called statistical estimation of F-X spectrum.This method is based on surface consistency assumption,and can be used in multiple domains.Before Fourier transform,a randomly sliding window is adopted to improve the accuracy and stability of spectrum estimation.Characteristic parameters were extracted on F-X plane to quantitatively determine the variability of frequency of seismic data.