西南石油大学学报(自然科学版)
西南石油大學學報(自然科學版)
서남석유대학학보(자연과학판)
JOURNAL OF SOUTHWEST PETROLEUM UNIVERSITY(SEIENCE & TECHNOLOGY EDITION)
2015年
1期
1-10
,共10页
广义S变换%振幅谱%动态反褶积%地震分辨率%吸收衰减
廣義S變換%振幅譜%動態反褶積%地震分辨率%吸收衰減
엄의S변환%진폭보%동태반습적%지진분변솔%흡수쇠감
generalized S transform%amplitude spectrum%dynamic deconvolution%seismic resolution%absorption attenuation
大多数提高地震资料分辨率的反褶积方法都基于传统静态褶积模型,这与非均匀介质中的地震子波实际传播规律不符。因此,在动态褶积模型基础上,结合地震波在地层中的吸收衰减特性,提出一种基于时频域的动态反褶积方法。该方法将广义S变换良好的多分辨特性引入到地震资料动态反褶积处理中,首先将非平稳地震记录的广义S变换近似表示为静态震源子波的傅里叶变换、复数值时频衰减函数与反射系数广义S变换的乘积,再采用多项式拟合的方法对非平稳地震记录时频谱进行平滑处理,从而可估算动态传播子波和地层反射系数。本文方法不需要直接求取Q值,能够适用于变Q值情况。理论模型和三维实际资料处理结果表明,该方法不仅能改善地震资料的分辨率,还能有效补偿深部地层吸收所引起的能量衰减。
大多數提高地震資料分辨率的反褶積方法都基于傳統靜態褶積模型,這與非均勻介質中的地震子波實際傳播規律不符。因此,在動態褶積模型基礎上,結閤地震波在地層中的吸收衰減特性,提齣一種基于時頻域的動態反褶積方法。該方法將廣義S變換良好的多分辨特性引入到地震資料動態反褶積處理中,首先將非平穩地震記錄的廣義S變換近似錶示為靜態震源子波的傅裏葉變換、複數值時頻衰減函數與反射繫數廣義S變換的乘積,再採用多項式擬閤的方法對非平穩地震記錄時頻譜進行平滑處理,從而可估算動態傳播子波和地層反射繫數。本文方法不需要直接求取Q值,能夠適用于變Q值情況。理論模型和三維實際資料處理結果錶明,該方法不僅能改善地震資料的分辨率,還能有效補償深部地層吸收所引起的能量衰減。
대다수제고지진자료분변솔적반습적방법도기우전통정태습적모형,저여비균균개질중적지진자파실제전파규률불부。인차,재동태습적모형기출상,결합지진파재지층중적흡수쇠감특성,제출일충기우시빈역적동태반습적방법。해방법장엄의S변환량호적다분변특성인입도지진자료동태반습적처리중,수선장비평은지진기록적엄의S변환근사표시위정태진원자파적부리협변환、복수치시빈쇠감함수여반사계수엄의S변환적승적,재채용다항식의합적방법대비평은지진기록시빈보진행평활처리,종이가고산동태전파자파화지층반사계수。본문방법불수요직접구취Q치,능구괄용우변Q치정황。이론모형화삼유실제자료처리결과표명,해방법불부능개선지진자료적분변솔,환능유효보상심부지층흡수소인기적능량쇠감。
Most methods for improving the resolution of seismic data are usually based on the traditional static convolution model,which are inconsistent with the actual propagation law of seismic wavelet in inhomogeneous media. Therefore,based on the dynamic convolution model,this paper combines attenuation and absorption characteristics of seismic wave,and proposes a dynamic deconvolution method based on time-frequency domain. This method introduces the good multi-resolution character-istic of GST(generalized S transform)into the processing procedure of dynamic deconvolution for seismic data. Firstly,GST of non-stationary seismic records can be approximately presented as the product of Fourier transform of static source wavelet, time-frequency attenuation function and GST of the reflection coefficient. And then we use polynomial fitting to smooth the time-frequency spectral of non-stationary seismic records,thus dynamic propagation wavelet and the reflection coefficient of formation can be estimated. The method of this paper does not need calculate the value of Q directly,and it is applicable in the situation that the value of Q is changed. The processed results for both theoretical model and real 3D seismic data validate that this method not only can improve the resolution of the seismic data,but also can compensate efficiently the energy attenuation caused by deep formation absorption.