测井数据小波变换用于准层序研究  

RECOGNITION OF PARASEQUENCES BASED ON WAVELET TRANSFORM OF LOG DATA

作　　者：房文静[1] 范宜仁[1] 邓少贵[1] 黄隆基[1] 

机构地区：[1]中国石油大学地球资源与信息学院

出　　处：《地层学杂志》2007年第3期303-308,共6页Journal of Stratigraphy

基　　金：中石油创新基金(W060122);中国石油大学博士创新基金(B2004-02)联合资助

摘　　要：层序界面的识别及其内部特征是层序地层学研究的关键问题。准层序地层单元的分界面上物理性质变化明显,测井曲线表现为突变,测井数据小波变换能够表征这种突变。以东营凹陷某井为例,选取高斯小波对SP测井数据进行小波变换,得到时频色谱图和不同尺度上的小波系数曲线。选定最佳分解尺度后,依据该尺度下小波系数模极值的位置能准确识别出准层序界面。并利用测井多尺度分析方法对不同井段GR测井曲线进行分析,得到不同尺度下的低频和高频信息能够识别出准层序内部的沉积旋回类型,建立了相应的频谱响应特征。这些探索为层序界面的定量划分及其内部特征研究提供了一种新的思路。The principle used to recognize sequence boundary and the anatomy of stratigraphic architecture is the key of sequence stratigraphy. Log curves change abruptly due to variation of physical properties at the interface of parasequences. Such changes can be revealed by wavelet transform of log data. As a case study, SP log curve from a well in the Dongying Sag is processed using Gauss wavelet transformation, and the time-frequency chromatogram and wavelet coefficient curves on different scales are obtained. When the optimal decomposition scale is determined, the interface between parasequences can be recognized exactly according to the position of wavelet modulus. The depositional cycle types can be identified by using the high and low frequency wavelet coefficient of GR after multiscale analysis, and the corresponding frequency spectrum response character was established. This research provides a completely new method to quantitatively demarcate sequence boundaries and stratigraphic architecture.

关 键 词：测井数据 小波变换 层序地层学 准层序 模极值 多尺度分析 

分 类 号：P539.2[天文地球—古生物学与地层学]