| 预测多次波的逆散射级数方法与SRME方法及比较 |
| |
| 引用本文： | 陈小宏,刘华锋.预测多次波的逆散射级数方法与SRME方法及比较[J].地球物理学进展,2012,27(3):1040-1050. |
| |
| 作者姓名： | 陈小宏 刘华锋 |
| |
| 作者单位： | 海洋石油勘探国家工程实验室,中国石油大学(北京),北京102249;油气资源与探测国家重点实验室,中国石油大学(北京),北京102249 |
| |
| 摘 要： | 逆散射级数方法和SRME方法是典型的两类基于波动方程的无需地下结构信息的自由表面多次波预测方法,本文详细讨论了这两类方法的基本思想、方法原理和实现思路,并对2维逆散射级数方法进行了降维简化,推导给出了1.5维逆散射级数衰减自由表面多次波的方法,减少了计算量,并降低了对3维规则观测系统的要求.利用经典SMAART模型数据测试比较了2维逆散射级数方法、1.5维逆散射级数方法及2维SRME方法预测自由表面多次波的效果及优缺点,比较了三者在方法实现、并行计算效率及对数据要求上的异同,并结合分析比较结果,给出了实际应用中方法选择的建议.
|
| 关 键 词： | 多次波衰减 自由表面多次波 逆散射级数 SRME |
Comparison between inverse scattering series method and SRME method in free surface related multiple prediction |
| |
| CHEN Xiao-hong
,
LIU Hua-feng.Comparison between inverse scattering series method and SRME method in free surface related multiple prediction[J].Progress in Geophysics,2012,27(3):1040-1050. |
| |
| Authors: | CHEN Xiao-hong LIU Hua-feng |
| |
| Institution: | 1,2(1.National Engineering Laboratory for Offshore Oil Exploration,China University of Petroleum(Beijing),Beijing 102249,China. 2.State Key Laboratory of Petroleum Resource and Prospecting,China University of Petroleum(Beijing),Beijing 102249,China.) |
| |
| Abstract: | Multiple interferes velocity analysis,image accuracy and seismic section interpretation in seismic exploration.Therefore,multiple attenuation is one of the key processing steps in marine seismic data processing.Many multiple attenuation methods have been proposed based on the differential between primary and multiple in different domains.Multiple attenuation method based on wave equation,which always be divided into prediction and subtraction phases,can predict and attenuate multiple effectively just using seismic data itself.The latter is more efficient for complex subsurface objective,where the differential assumption of the former will not be met well and the subsurface information can not be obtained easily.Inverse Scattering Series(ISS) method and Surface Related Multiple Elimination(SRME) method are two classical kinds of wave-equation-based multiple attenuation method,which both don’t need any subsurface structure or velocity information.ISS method,based on scattering theory,can predict multiple model data using an inverse scattering sub-series which contribute to multiple generation in forward modeling from original seismic data directly.SRME method assumes any multiple can be constructed by summing a group of multiplications of two records which have special spatial location relationship in seismic data.We discussed the key ideas,principles,basic requirement for seismic data and detail steps of calculation of ISS and SRME method in this paper.For some survey area which has simpler substructures,1.5D ISS method was proposed based on 2D ISS method,which can save lots of calculation and is efficient in PC.It can also degrade the requirement of 3D regular geometry in 2D ISS and SRME method into 2D regular geometry.After that,we discussed the effects,advantages and disadvantages in multiple predictions of 2D ISS,1.5D ISS and 2D SRME method.In order to reduce calculating time and improve efficiency,Message Passing Interface(MPI) technique was introduced into these programs design for PC cluster.For comparing the similarities and differentials in actualization,efficiency and requirement for seismic data of 2D ISS,1.5D ISS and 2D SRME method,we applied these three methods to the SMAART model data,a typical synthetic multiple data,as a demonstration.According to the application example and comparison of the results,2D ISS method is the most accuracy and expensive method for predicting multiple model data in these three methods.If subsurface structure is not very complicated,2D SRME and 1.5D ISS method are suggested to be selected to predict multiple model data.1.5D ISS method is the cheapest method for predicting multiple model data in these three methods and it can meet the requirement of multiple attenuation in relative simple survey areas. |
| |
| Keywords: | multiple attenuation free surface related multiple inverse scattering series SRME |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
