ABSTRACTS

Multi-attribute automatic interpretation of salt domes based on deep learning.ZHANG Yuxi1,2,LIU Yang1,2,3,ZHANG Haoran1,2,LYU Wenjie1,and XUE Hao4.Oil Geophysical Prospecting,2020,55(3):475-483.

It is difficult and inefficient for salt dome interpretation using 3D seismic data.A new workflow uses different seismic attributes to automatically inpterpret salt domes based on a small amount of 2D seismic data as training samples and testing models after deep learning.The wrokflow consists of three parts.First,according to the characteristics of a salt dome on seismic data,extract three types of sensitive attributes including chaotic and RMS amplitude,and variance.For each type of attribute,select a small amount of inline and time slices as training samples and use a data augmentation method to automatically generate massive samples.Second,construct a convolutional neural network based on anencoder-decoder architecture,and input two types of samples with different attributes for training and testing models to obtain multiple independent models.Finally,to comprehensively consider the features of all attributes and obtain more accurate classified results,use an ensemble learning method to merge the models and acquire optimized results.The results indicate that the boundaries of salt domes are clear and the classification errors can be significantly removed.This method can efficiently realize automatic segmentation of salt domes in 3D data setand further improve the prediction ability of models.

Keywords:deep learning,automatic interpretation of salt domes,seismic attributes,ensemble learning

1.State Key Laboratory of Petroleum Resources and Prospecting,China University of Petroleum (Beijing),Beijing 102249,China

2.CNPC Key Laboratory of Geophysical Prospectiong,China University of Petroleum (Beijing),Beijing 102249,China

3.Karamay Campus,China University of Petroleum (Beijing),Karamay,Xinjiang 834000,China

4.CNOOC Research Institute Co.,Beijing 100028,China

Prediction of S-wave velocity based on GRU neural network.SUN Yuhang1,2,and LIU Yang1,2,3.Oil Geophysical Prospecting,2020,55(3):484-492,503.

There is a relationship between reservoir parameters and shear wave velocity.But it is too complex to get analytic solutions.This paper proposes a GRU (gated recurrent unit) neural network which includes building a neural network,data preprocessing,samples training and data predication.After approximating the relationship between S-wave velocity and reservoir parameters by training a neural network,S-wave velocity can be predicted directly from P-wave velocity,density,gamma ray,porosity and logarithm of resistivity.The neural network was trained and tested by logging data from 30 wells in Block D.The results show that: ①The P-wave velocity,density and logarithm of resistivity are positively related to the S-wave velocity,while the gamma ray and porosity are negatively related to the S-wave velocity; ②In the case of being trained by more wells and tested by less wells,the relative error between the S-wave velocity and the real one is about 3.00%,and the correlation coefficient is 0.9837 for training data,while they are 3.19% and 0.9805 for tested data.In the case of being trained by less wells and tested by more wellsg,the relative error is 2.49% and the correlation coefficient is 0.9867 for training data,while they are 3.92% and 0.9686 for testing data.The new method has a high prediction accuracy and generalization ability.

Keywords:prediction of S-wave velocity,GRU neural network,reservoir parameters

1.State Key Laboratory of Petroleum Resources and Prospecting,China University of Petroleum (Beijing),Beijing 102249,China

2.CNPC Key Laboratory of Geophysical Prospecting,China University of Petroleum (Beijing),Beijing 102249,China

3.Karamay Campus,China University of Petroleum (Beijing),Karamay,Xinjiang 834000,China

Pre-stack random noise suppression with deep residual network.LI Haishan1,CHEN Dewu1,WU Jie1,and CHANG Dekuan1.Oil Geophysical Prospecting,2020,55(3):493-503.

Deep residual network,as an advanced deep learning algorithm,has received high attention from academic and industrial circles in recent years.To realize intelligent and efficient suppression of random noise in pre-stack seismic records,first,a deep nonlinear denoising network is designed based on the principle of deep residual network,and then the network is trained by the constructed high-quality random noise training sets to automatically learn the features of random noises in a high-dimensional space,so as to fit the nonlinear mapping relationship between noisy seismic records and random noises,and achieve the purpose of automatic suppression of random noises.Both model test and field application have proved the effectiveness of this method.Though the denoising capability of this method is as good as the method used for generating label data,the former has better denoising efficiency and adaptability than the latter.It provides an idea to deal with the denoising problem of TB-level pre-stack seismic data.

Keywords:deep learning,convolutional neural network,deep residual network,random noise,denoising

1.Northwest Branch,Research Institute of Petroleum Exploration & Development,PetroChina,Lanzhou,Gansu 730020,China

Semi-supervised seismic facies analysis based on prestack seismic texture.CAI Hanpeng1,2,HU Hao-yang1,WU Qingping1,WANG Jun3,and LI Zhipeng3.Oil Geophysical Prospecting，2020,55(3):504-509.

A semi-supervised seismic facies analysis algorithm based on prestack seismic texture is proposed for taking full use of subtle information contained in prestack seismic data based on prior knowledge such as drilling and geological data.First,prestack seismic texture is introduced to highlight the variability of tiny space and amplitude with azimuth/offset in prestack seismic data.Second,self-organizing map (SOM) is used to train samples.Finally,constrained by prior drilling knowledge,the semi-supervised clustering of neurons in the output layer of SOM is carried out to generate the mapping relation between the neurons and the seismic facies category.Theoretical model and application demonstrated that the method can improve the accuracy of seismic facies map and enhance the ability to distinguish seismic microfacies.It is a better tool for seismic facies analysis.

Keywords: seismic facies analysis,pre-stack texture,semi-supervised learning,self-organizing map

1.School of Resources and Environment,University of Electronic Science and Technology of China,Chengdu,Sichuan 611731,China

2.Center for Information Geoscience,University of Electronic Science and Technology of China,Chengdu,Sichuan 611731,China

3.Research Institute of Exploration & Production,SINOPEC Shengli Oilfield,Dongying,Shandong 257015,China

Reconstruction of logging traces based on GRU neural network.WANG Jun1,CAO Junxing1,and YOU Jiachun1.Oil Geophysical Prospecting,2020,55(3):510-520.

Formation parameters such as lithology,resistivity,porosity,permeability and saturation can be obtained by logging interpretation.However,it is frequently the case that logging data are missing or incomplete,but it is difficulty and expensive to record again.Present reconstruction of logging data based on traditional linear hypothesis and statistical analysis can not meet the requirements of fine description of reservoir characteristics.Gated Recurrent Unit (GRU) neural network is a new deep learning algorithm which is suitable for solving nonlinear and sequential problems.Based on the latest achievements of deep learning,a logging reconstruction method based on GRU neural network is proposed.The method considers the nonlinear mapping between logging data,how logging data change with depth and the relationship among historical data.Tested with real logging data and compared with the multiple regression method,the GRU network model is effective to reconstruct logging traces.It is a new idea a for logging traces reconstruction.

Keywords: logging trace reconstruction,deep learning,GRU neural network,multiple regression ana-lysis

1.School of Geophysics,Chengdu University of Technology,Chengdu,Sichuan 610059,China

Wavefield prediction with inverse scattering series and 2D blind separation of convolved mixtures for suppressing internal multiples.BI Lifei1,QIN Ning1,LI Zhongxiao2，LIANG Hongxian1,LI Zhenchun3，and DOU Jingying4.Oil Geophysical Prospecting，2020,55(3):521-529.

Multiple suppression is an important step of seismic data processing.The prediction subtraction method based on wave equation is a common one to suppress internal multiples.This method includes two steps: internal multiples prediction and adaptive subtraction.After introducing hypothesis conditions such as step function and finite integral interval,we proposed a simplified formula to simplify the inverse scattering series for predicting internal multiples,and then introducing adaptive subtraction based on 2D blind separation of convolved mixtures to conduct adaptive subtraction with matching filters.Synthetic and field data have de-monstrated that the method is data-driven and independent on a velocity model for predicting internal multiples.Furthermore,compared with the traditional method based on L2-norm and single-channel blind separation of convolved mixtures,the method based on multi-channel blind separation of convolved mixtures can better protect effective primary waves while suppressing internal multiples.

Keywords: internal multiples,inverse scattering series,2D blind separation of convolved mixtures,adaptive subtraction,data-driven

1.Shengli Geophysical Research Institute of Sinopec,Dongying,Shandong 257022,China

2.College of Electronic Information,Qingdao University,Qingdao,Shandong 266071,China

3.School of Geosciences,China University of Petroleum(East China),Qingdao,Shandong 266580,China

4.Research Institute of CNOOC Zhanjiang Branch,Zhanjiang,Guangdong 524057,China

Adaptive multiple subtraction based on a 3D mat-ching filter and pseudo seismic data algorithm.LI Zhongxiao1,GAO Haotian1,CHEN Xinze1,LI Yong-qiang2,and LI Zhenchun2.Oil Geophysical Prospecting，2020,55(3):530-540.

Adaptive multiple subtraction is an important step for suppressing multiples by multiple prediction and subtraction.A 3D matching filter is proposed over a 2D matching filter to reduce residual multiples by combining with several gathers of the predicted multiples to match with original data.To avoid possible primary distortion,the same 3D matching filter is used to fit several original gathers.In addition,we introduce the pseudo seismic data algorithm to solve the optimization problem with the Huber norm minimization constraint on primaries.Without assuming the orthogonality of primaries and multiples,the new method can separate primaries from multiples effectively.During iterating,the pseudo seismic data algorithm only needs to conduct matrix decomposition with Cholesky factorization once,and the computational efficiency is high.Synthetic and field data have proved that the new method can better preserve primaries and remove multiples than a 3D matching filter based on minimization of primary energy and a 2D matching filter based on pseudo seismic data algorithm.

Keywords: adaptive multiple subtraction,3D matching filter,pseudo seismic data algorithm,Huber norm

1.School of Electronic Information,Qingdao University,Qingdao,Shandong 266071,China

2.School of Geosciences,China University of Petroleum (East China),Qingdao,Shandong 266580,China

Spectrum shaping of desired targets based on broadband Ricker wavelets.ZHAO Baoyin1,CHEN Siyuan2,TAO Yu2,XU Jing2,and WANG Hao2.Oil Geophysical Prospecting,2020,55(3):541-547.

Spectrum whitening based on seismic wavelet shaping is a common method for high-resolution post-stack seismic processing,and the desired wavelet shape has always been a much hotter topic.Compared with band-limited wavelet,Ricker wavelet and other commonly used shaping wavelets,broadband Ricker wavelet is widely used in seismic data processing on account of its weaker sidelobe energy and higher resolution at the same main lobe width than Ricker wavelet.By designing the parameter(R) that takes into account resolution and fidelity and searching for the extreme point,the optimal shape of the broadband Ricker wavelet with a limited bandwidth is determined,and then a ba-lanced wavelet based on broadband Ricker wavelet is proposed to guide wavelet shaping of post-stack data.Model and field data have proved that the balanced wavelet provides a new reference to seismic processing with high resolution and high fidelity.

Keywords: broadband Ricker wavelet,Yu’s wavelet,spectrum shaping,balanced wavelet

1.PetroChina Jidong Oilfield Company,Tangshan,Hebei 063004,China

2.CNPC Key Laboratory of Geophysical Exploration,China University of Petroleum(Beijing)，Beijing 102249,China

Using sparse-constrained nonstationary polynomial regression to remove seismic noises and picking up first arrival.LIU Guochang1,CAI Jiaming2,YAN Haiyang3,LI Jieli1,and CHEN Xiaohong1.Oil Geophysical Prospecting,2020,55(3):548-556.

Nonstationary polynomial fitting relates to optimization with L2norm.Although the time-dependent characteristics of signals are considered,the residual is still assumed to be randomly distributed.In the case that there are strong non-random noises in seismic data,conventional nonstationary polynomial fitting based on L2norm is no longer applicable.This study investigated the theory and method of sparse-constrained nonstationary polynomial regression.First,we reviewed the basic principle of non-stationary polynomial regression.Second,to solve the problem related to the complex sparse residual,under the framework of inverse problem regularization theory,we combined non-stationary polynomial regression with L1norm constraint,followed the combined constraint strategy of shaping regularization with L1norm,and solved the multi-constraint inverse problem with conjugate gradient and projection algorithm.In addition,we estimated the coefficient of polynomial regression with time-varying smoothing characteristics and the residual sparsely distributed,which can reduce the influence of sparse strong noises on inversion.Finally,we proposed the basic process and parameter analysis of the algorithm.Synthetic and field data have proved that sparse constrained non-stationary polynomial regression is effective for noise suppression and pick up first arrival.

Keywords: nonstationary polynomial regression,L2norm,L1norm,sparse constraint,pick up first arrival,noise suppression

1.College of Geophysics,China University of Petroleum(Beijing),Beijing 102249,China

2.Geophysical Research Institute,BGP,CNPC,Zhuozhou,Hebei 072751,China

3.Division of Marine Geophysical Exploration,BGP,CNPC,Tianjin 300450,China

Direct numerical calculation of ray Jacobian and its application.LIANG Quan1，2，3,MAO Weijian1，2,OUYANG Wei1，2,LI Wuqun1，2,and ZHANG Jianlei4.Oil Geophysical Prospecting,2020,55(3):557-566.

Ray Jacobian plays a fundamental role in solving transport equation,obtaining the geometrical propagating factor of seismic wavefront and the amplitude of the Green’s function.The ray Jacobian is of great importance for true amplitude migration imaging based on the ray theory and is a part of the weight function of the true amplitude imaging condition of generalized Radon transform (GRT) based inverse scattering migration.In the medium with strong velocity variation,the ray Jacobian computed by dynamic ray tracing may be rather problematic for the unstable second spatial derivatives of the velocity field,which results in generating pseudo-caustic points and missing real caustic points.To compute ray Jacobian accurately,we focus on its definition and calculate it through the ratio of the element cross-sectional area of a ray tube to the element initial angular area.Since the direct method has a clear physical meaning,we can get an accurate Jacobian,and real caustic points can be detected effectively.In addition,the singularity in inverse scattering amplitude-preserved migration caused by the caustics can be avoided by adopting a reasonable ray Jacobian smoothing threshold.Application to a salt dome model has proved the method is effective and applicable.

Keywords: ray Jacobian,cross-sectional area of ray tube,direct numericial calculation,caustic,smoothing threshold,inverse scattering amplitude-preserved migration

1.Center for Computational & Exploration Geophysics，Institute of Geodesy and Geophysics，Chinese Academy of Science,Wuhan,Hubei 430077,China

2.State Key Laboratory of Geodesy and Earth’s Dynamics,Wuhan,Hubei 430077,China

3.University of Chinese Academy of Sciences，Beijing 100049，China

4.Research & Development Center,BGP,CNPC,Zhuozhou,Hebei 072751,China

Numerical simulation of zero-offset VSP wave field based on the acoustic scattering theory.YAN Hongqin1.Oil Geophysical Prospecting,2020,55(3):567-574.

Anumerical method of preconditioned least squares is proposed based on volume integral equation to accurately simulate zero-offset VSP wave field.First,based on the acoustic scattering theory,the one-dimensional acoustic wave equation in frequency domain can be expressed as the Lippman-Schwinger integral equation by the Green function.On this base how to solve the second kind of Fredholm integral equation can be turned to how to sov-le large linear equations through rectangular integral formula,and this lays a foundation for subsequent accurate calculation.Second,the whole wave field can be solved by the least squares method,rather than adding the scattering component to the whole wave field step by step by iteration.After introducing the preconditioned operator,the numerical calculation process is more stable.Finally,taking a stepped model and a real model based on logging data as examples,the VSP wave field can be solved using the new method,and the results were compared with those from the other conventional finite difference methods.The results show that the new method is feasible,effective and accurate in simulating zero-offset VSP wave field.

Keywords: acoustic scattering theory,zero-offset VSP wave field,Green function,pre-conditioned least square method

1.Loudi Vocational and Technical College,Loudi,Hunan 417000,China

Numerical simulation to the influence of fracture parameters on P-wave anisotropy.DUAN Xi1,2,LIU Xiangjun1,LIANG Lixi1,and XIONG Jian1.Oil Geophysical Prospecting,2020,55(3):575-583,590.

In the study of P-wave azimuthal anisotropy of natural fractures using seismic data,systematic sensitivity of P-wave attributes to fractures is not carried out.A non-uniform random fracture medium is produced through adjacent point fusion after setting fracture parameters based on digital image processing technology.Based on the acoustic wave theory,the acoustic wave field of the random discrete fracture model is numerically simulated.After calculating the acoustic velocity and attenuation coefficient of the azimuth of a survey line,the relationship between acoustic parameters and fracture distribution,orientation and density and fluid is analyzed to determine fracture orientation more accurately.Numerical simulation to the acoustic wave field of the random discrete fracture model shows that the influence of fracture parameters on acoustic attenuation coefficient is far greater than that on acoustic velocity,and the azimuth of the survey line corresponding to the minimum acoustic attenuation coefficient can be used to accurately determine fracture orientation.With the increase of fracture density,the relative change of attenuation coefficient decreases.With the increase of water saturation,the relative changes of acoustic velocity and attenuation coefficient increase first and then decrease.

Keywords:random fracture medium,fracture parameters,P-wave anisotropy,acoustic wave theory,numerical simulation

1.State Key Laboratory of Oil and Reservoir Geology and Exploitation (Southwest Petroleum University),Chengdu,Sichuan 610500,China

2.Department of Science,Southwest Petroleum University,Chengdu,Sichuan 610500,China

Seismic scatterring data migration based on scattering wave equation.HU Ziduo1,YONG Xueshan1,LIU Wei1,CHEN Shengchang2,and WANG Yanxiang1.Oil Geophysical Prospecting,2020,55(3):584-590.

Present reverse time migration methods don’t distinguish scattering data from reflection data,and the imaging formula is established with the concept of reflected wave propagation.In this study,starting from the perturbation form of wave equation,we defined the scatterer generating scattering wave and the reflector generating reflected wave according to the relationship between the size of a underground heterogeneous body and the wavelength of seismic wave,and formulated a linear forward equation of primary scattering data,then derived the formula of calculating seismic reverse time migration on the spatial scattering location by the linear inverse theory,and finally tested the method on Sigbee2A model data.Theory and numerical test have demonstrated that the scattering data migration method can provide migration results with more correct phase and higher resolution than those from the conventional reverse time migration method based on the concept of reflection.

Keywords: scatterer,seismic scattering data,forward equation,linear inversion,migration

1.Northwest Branch,Research Institute of Petroleum Exploration & Development,PetroChina,Lanzhou,Gansu 730020,China

2.School of Earth Sciences,Zhejiang University,Hangzhou,Zhejiang 310027,China

Seismic diffraction imaging by reverse time migration in dip angle domain.WANG Tianchi1,2,LIU Shaoyong1,2,GU Hanming1,2,TANG Yongjie1,2,and YAN Zhe1,2.Oil Geophysical Prospecting,2020,55(3):591-598.

The seismic responses of fracture-cavity reservoirs and irregular orebodies are characterized by diffractions.Reflected wave suppression during seismic migration in dip-angle domain is beneficial to imaging these targets.We used Kirchhoff migration and reverse time migration (RTM) to extract dip-angle gathers,and then compared their imaging performances based on diffractions.The results shows that RTM has a better reflections focusing ability than a Kirchhoff migration operator in dip-angle domain,and makes reflections suppression much easier.The Poynting vector method is introduced in our study to produce the dip-angle gather during the RTM process.Among the dip-angle gathers extracted by RTM,the energy of reflections focus on isolated “noises”,but diffraction events appear as a flattened straight line.Based on the difference in their responses,suppressing “noises” with median filtering can separate diffractions from reflections,and consequently RTM diffraction imaging to get good results.Numerical experiments have verified the correctness and effectiveness of the method proposed.

Keywords: diffraction image,reverse time migration,dip-angle gather,median filtering

1.Institute of Geophysics and Geomatics,China University of Geosciences(Wuhan),Wuhan,Hubei 430074,China

2.Hubei Subsurface Multiscale Imaging Key Laboratory,Wuhan,Hubei 430074,China

Research of geological model-constrained FWI and application in complex fault-block zones.ZHANG Ziliang1，2,LI Zhenchun1,ZHANG Kai1,ZHAO Shuo1,WANG Feiyi1,and Wu Dongying3.Oil Geophysical Prospecting,2020,55(3):599-606.

A high-precision velocity model can be built through full waveform inversion,but the inversion result is not ideal in a complex fault-block zone because of low S/N radio and limited frequency band.Based on the conventional FWI theory,this paper proposes a model-constrained objective functional formula which constrains full waveform inversion through seismic interpretation.Model and field data show that the model-constrained FWI can provide the velocity field of a complex fault-block zone with high precision.It can improve the image of a complex fault-block zone.

Keywords: full waveform inversion(FWI),velocity modeling,complex fault,geological model-constrained,seismic imaging

1.School of Geosciences,China University of Petroleum(East China),Qingdao,Shandong 266580,China

2.Key Laboratory of Sedimentary Mineralization and Sedimentary Minerals，Shandong University of Science and Technology,Qingdao,Shandong 266580,China

3.Geological Research Institute of Daqing Yushulin Oilfield Co.,Ltd.Daqing,Helongjiang 163000,China

Offset-domain common image gathers on wave equation assisted by attribute migration.LIU Guofeng1,LI Haibo2，LIU Yu1,and MENG Xiaohong1.Oil Geophysical Prospecting,2020,55(3):607-615.

Seismic migration based on wave equation is more accurate than Kirchhoff-based migration,and the former is more sensitive to the accuracy of a velocity model.Therefore,instead of Kirchhoff depth migration,it is of important practical significance for velocity iteration of offset-domain common image gathers based on wave equation.A computation scheme based on attribute migration is proposed to analyze offset-domain common image gathers through wave equation migration.In other words,migration is done after modulating surface offset,then the ratio of the migration based on modulated offset to the migration based on original data is taken as the surface offset of an imaging point.By giving new migration to its corresponding offset,and calculating and stacking literately,the final result is surface offset gathers.In practice,the two kinds of migration are performed during the same migration circle,and the computing effort is only increased by 30% since only increasing the propagation of receiver wavefield.2D and 3D model and field data have proved the method effective.

Keywords：wave equation migration,common image gathers,attribute migration,modulation

1.School of Geophysics and Information Technology,China University of Geosciences (Beijing),Beijing 100083,China

2.Geophysical Prospecting Team of Shandong Bureau of Geology,Jinan,Shandong 250104,China

Least-squares reverse-time migration based on a fractional Laplacian viscoacoustic wave equation.CHEN Hanming1，2，3，ZHOU Hui1，2，3，and TIAN Yukun4.Oil Geophysical Prospecting，2020,55(3):616-626.

The attenuation-compensated reverse-time migration (Q-RTM) compensates for amplitude loss and phase distortion along the entire wave propagatiing path,so it can enhance seismic imaging accuracy and resolution.However,sinceQ-RTM requires to solve an amplitude-boosted wave equation to simulate exponentially increased amplitude,it is not stable.A novel least-squaresQ-RTM (LSQRTM) is proposed to stably invert a subsurface reflectivity model.The LSQRTM is based on viscoacoustic Born forward modeling and its adjoint operator that are derived from a new fractional Laplacian viscoacoustic wave equation.The wavefield simulated by the fractional Laplacian viscoacoustic wave equation can well match realistic attenuation and dispersion predicted by a constant-Qmodel.The L-BFGS algorithm is used in least-squares inversion to calculate the descent direction to update the reflectivity model.Used for Marmousi model and real data,the LSQRTM can stably compensate for the viscosity of the medium and provide a reflectivity model with high resolution.

Keywords：least-squares reverse-time migration,fractional Laplacian,viscoacoustic,seismic imaging

1.State Key Lab of Petroleum Resources and Prospecting,Beijing 102249,China

2.CNPC Key Lab of Geophysical Prospecting,Beijing 102249,China

3.College of Geophysics,China University of Petroleum (Beijing),Beijing 102249,China

4.Oil & Gas Survey,CGS,Beijing 100083,China

First-order perturbation approximation of azimuth converted wave reflection coefficient of HTI media.LIU Hongying1,WU Guochen1,2,SHAN Junzhen1,and YANG Sen1.Oil Geophysical Prospecting,2020,55(3):627-634.

Depending on separate research purpose,scholars simplified and approximated the Zoeppritz equation.However,most of the approximate equations cannot accurately describe the changing laws of the amplitude of PSV wave at a large incident angle.And due to seismic data changing with azimuth,research on the azimuth reflection coefficient of PSV wave is relatively weak.Based on the precise expression of the reflection / transmission coefficient of HTI media,according to the theory of medium decomposition and perturbation,and assuming weak anisotropy,we derived an approximate expression of the reflection coefficient of PSV wave,analyzed the amplitude azimuth through forward modeling,and discussed the sensitivity of azimuthal PSV reflection coefficient to anisotropic parameters.Forward modeling shows that: ①At a small incident angle,the azimuth of the reflection coefficient is not obvious,and changes in anisotropic parameters have little effect on the reflection coefficient; however,as the increase of the incident angle,the characteristics of the reflection coefficient changing with the azimuth become more obvious,and the reflection coefficient is affected by the change of anisotropic parameters; ②The greater the contrast between the anisotropic parameters of the media above and below an interface,the more obvious the reflection coefficient changes with azimuth; ③γ(V)(related to the velocity difference in vertical and horizontal propagation of S-wave) has the greatest effect on the reflection coefficients,andε(V)(related to the velocity difference invertical and horizontal propagation of qP-wave) has the least effect on the reflection coefficient,and the effect of anisotropic parameters on the reflection coefficient is related to azimuth.

Keywords: HTI media,PSV wave,first-order perturbation approximation,anisotropic parameter,reflection coefficient,azimuthal characteristics,AVO analysis

1.School of Geoscience,China University of Petroleum (East China),Qingdao,Shandong 266580,China

2.Evaluation and Detection Technology Laboratory for Marine Mineral Resources,Qingdao,Shandong 266071,China

Extended anisotropic linear approximation for elastic phase velocity in VTI media.GU Yipeng1,YIN Xingyao1,2,LIANG Kai1,2,and ZHANG Jiajia1,2.Oil Geophysical Prospecting,2020,55(3):635-642.

The condition for linear approximation of weak anisotropy given by Thomsen is that absolute anisotropy is less than 0.2.However,for some highly anisotropic rocks,because the absolute value of anisotropy is large,the phase velocity error calculated by weak anisotropy linear approximation is large.A linear approximation formula for extended anisotropy of elastic wave phase velocity in VTI media is proposed.The formula maintains a linear relationship to anisotropic parameters.Starting from the accurate phase velocity formula of elastic wave in VTI media,a first-order Taylor expansion is performed at the tangent point of the accurate expression function curve and linear approximation function curve of phase velocity,a Jacobi matrix is formed from the first-order partial derivative of the phase velocity at the tangent point to the anisotropic parameter,and the extended anisotropic linear approximation formula is obtained.Theoretical analysis and numerical examples show that,when the absolute value of anisotropy is small,the results obtained by the weak anisotropy linear approximation and the extended anisotropy linear approximation agree well with theoretical values;when the absolute value of anisotropy is larger,the result obtained by the weak anisotropy approximation is larger than the theoretical value,and the result obtained by the linear anisotropy extension is in good agreement with the theoretical value.It’s proved that the extended anisotropic linear approximation is applicable for strong anisotropy by adjusting the tangent point of linear approximation.

Keywords:VTI media,weak anisotropy linear approximation,extended anisotropic linear approximation,first-order Taylor expansion,tangent point

1.School of Geosciences,China University of Petroleum(East China),Qingdao,Shandong 266580,China

2.Evaluation and Detection Technology Laboratory for Marine Mineral Resources,Qingdao,Shandong 266071,China

Tectonic stress prediction method based on curvature attribute.MA Ni1,YIN Xingyao2,3，ZONG Zhaoyun2,3,SUN Chengyu2,3,and WANG Shixing1.Oil Geophy-sical Prospecting,2020,55(3):643-650.

Traditional methods for estimating tectonic stress using curvature attribute calculate stress components through the curvatures inxandydirections and the torsion onxOyplane to obtain tectonic stress.However,no direct relationship can be established between curvatures and tectonic stress,and available commercial software usually cannot directly extract the curvatures inxandydirections and the torsion onxOyplane,but can only directly extract attributes such as maximum positive curvature and minimum negative curvature and so on.On the assumptions of the thin plate bending theory,the relationship among maximum positive curvature,minimum negative curvature and tectonic stress has been established,which is used to estimate tectonic stress based on maximum positive curvature and minimum negative curvature.This method is practical and simple to use.Applications have proved that the method can provide consistent results with traditional methods in predicting tectonic stress.

Keywords：thin plate bending theory,maximum positive curvature,minimum negative curvature,tectonic stress

1.SINOPEC Geophysical Research Institute,Nanjing,Jiangsu 211103,China

2.School of Geosciences,China University of Petroleum(East China),Qingdao,Shandong 266580,China

3.Laboratory for Marine Mineral Resources,Qing-dao National Laboratory for Marine Science and Technology,Qingdao,Shandong 266071,China

Identification on inner structures of fractured fault zones based on geological model：A case study on Well Fan 162 area in Dongying Sag.DU Kai1,LIN Chengyan1，2， MA Cunfei1，2， REN Lihua1，2，LIANG Shuyi3，and ZHANG Zongxuan1.Oil Geophysical Prospecting,2020,55(3):651-660.

A fractured fault zone includes a fault core and surrounding induced fractured zones.Studying its inner structure can help evaluate how underground fluid flows in it.Following the workflow of “from a geological model to seismic forward modeling to multi-attribute comprehensive identification”,we studied the fractured fault zone in Well Fan 162 area.The results show that: ① Both the fault core and the induced fractured zone have impacts on the seismic waveform of the original formation,and seismic data recorded with small trace spacing can better identify the inner structure.The core has responses within the seismic frequency band.The waveform of the induced fractured zone is affected easily by the core at low frequency,while at high frequency,the resolution is higher,and the interference effect is weak,so that the waveform of the induced fracture zone can be distinguished.② The amplitude of the fault core is weak,and the coherence is poor within the seismic frequency band.The amplitude of the fractured zone is weak,and the high-frequency coherence is poor.But the undisturbed formation is well coherent and strong in amplitude; ③ The core can be distinguished from the fractured zone by the difference between low-frequency and high-frequency coherence and instantaneous amplitude.The fractured fault zone can be identified by curvature and high-frequency coherence.

Keywords：fractured fault zone,geological model,instantaneous amplitude,discrete frequency coherency,RGB fusion,seismic forward modeling,curvature attribute

1.School of Geosciences,China University of Petroleum (East China),Qingdao,Shandong 266580,China

2.Reservoir Geology Key Laboratory of Shandong Province,Qingdao,Shandong 266580,China

3.Dongsheng Petroleum Development Group co.Ltd of Shengli Oilfield,Dongying,Shandong 257000,China

Application of spatial-windowed 2D Hilbert transform in volumetric edge detection of 3D seismic data.LYU Bingnan1,CHEN Xuehua1,2,XU He2,LIU Yunfei2,LUO Xin2,and ZHOU Chen2.Oil Geophysical Prospecting,2020,55(3):661-668.

Discontinuous seismic data greatly affect the identification to channels,faults and fracture zones.An edge detection method using 3D seismic data is proposed,which is based on 2D Hilbert transform.By adding time and depth windows into the 3D target layer and adding a 2D Gauss function,an improved 2D Hilbert operator is established.It can update the calculation aperture in 3D space for detecting any discontinuous seismic anomalies and reduce noises.Model and real seismic data have proved that the edge detection method can effectively describe the spatial characteristics of geological anomalies,and highlight the edge and orientation of a fracture zone.It is help improves the accuracy of seismic data interpretation.

Keywords: 2D Hilbert transform,3D seismic data,volumetric edge detection,Gaussian filtering,seismic attribute

1.State Key Laboratory of Oil & Gas Reservoir Geology and Exploitation,Chengdu University of Technology,Chengdu,Sichuan 610059,China

2.Key Lab of Earth Exploration & Information Techniques of Ministry of Education,Chengdu University of Technology,Chengdu,Sichuan 610059,China

Interpretation technique of seismic geomorphological slice and its application.YANG Zhanlong1,2.Oil Geophysical Prospecting，2020,55(3):669-677.

Time slicing，horizon slicing，stratal slicing and horizon flattening are analytical methods commonly used for studying structure,deposition and paleo-geomorphology in seismic interpretation.In this paper,the concept of seismic geomorphological slice and its making method are proposed for the practical needs for studying fine sedimentary system.A seismic geomorphological slice refers to a seismic slice extracted on a point of seismic data that significantly reflects the paleo-geomorphology of a geological period.Usually,it is made by spatial tracking of a geological body and small time-window perspective.This type of slice has low isochronous requirements to constrained horizons,so it is widely applicable in seismic data interpretation.It is easy to understand for geologists based on the analogy of geomorphology and has a strong practicability.It provides reliable prediction based on the idea of using the present as a key to understand the past.In summary,this method is effective for seismic sedimentary analysis.Practical applications have proved the scientificity of the concept,the rationality of the slicing method and the validity of the results.

Keywords：seismic geomorphological slice,depositional system,spatial tracking of geological body,small time-window perspective,horizon flattening

1.Northwest Branch,Research Institute of Petroleum Exploration & Development,PetroChina,Lanzhou,Gansu 730020,China

2.Key Laboratory of Reservoir Description,China National Petroleum Corporation,Lanzhou,Gansu 730020,China

Optimization and application of 3D non-topological consistent blocking.LUO Kaiyun1,FANG Peng2,HE Yongqing1,JIANG Xianyi1,and WU Heng1.Oil Geophysical Prospecting,2020,55(3):678-685.

In seismic exploration,a 3D model is very important for forward numerical simulation,of which blocking is the core.Presently,block tracking technologies are based on the topologic structure of the intersection on the model.When some topologic errors exist on the intersection,such as leaks or crossover,a correct block can’t be established.A new blocking method is proposed,which is totally independent of topology of the intersection curves of the model.The algorithm is simply based on the shape of the curved model itself to build a block and use the Visual Observation Technology to automatically tracking the triangular mesh of the block boundary.In order to reduce the computing redundancy in the original algorithm and enhance tracking efficiency,an optimized method which is based on Octree is proposed.Many modeling tests have proved that the optimization algorithm has not only better blocking performance but also provides a higher blocking success rate.It is practical and effective in field application.

Keywords: Non-topological consistency,blocking of 3D geological block,Octree

1.Acquisition Technique Center,BGP,Zhuozhou,Hebei 072751,China;

2.College of Information Science and Technology,Chengdu University of Technology,Chengdu,Sichuan 610059,China

Simultaneous and joint inversion of gravity and seismic data based on variable density-velocity relation.XIANG Peng1,WANG Jinduo1,TAN Shaoquan1,and CHEN Xueguo1.Oil Geophysical Prospecting,2020,55(3):686-693.

In view of the shortcomings of available simultaneous joint inversion methods,this paper pre-sents a simultaneous inversion method of gravity and seismic data based on variable density-velocity relation,derives the formulas for getting joint solutions of density model and velocity model in details,and presents the updating algorithm of density-velocity relation using higher-order statistics.During inversion,the initial relation is updated according to previous and present density and velocity models.Model applications show that the me-thod reduces the adverse effects resulted by an inaccurate initial relation.Practical applications show that the velocity model established by this method can improve seismic imaging quality,and has a great application potential in the area less explored.

Keywords: joint inversion of gravity and seismic data,simultaneous joint inversion,variable density-velocity relation

1.Research Institute of Exploration & Development,SINOPEC Shengli Oilfield,Dongying,Shandong 257000,China

Seismic techniques for predicting fractures in granite buried hills.JIANG Xiaoyu1,ZHANG Yan1,GAN Lideng1,SONG Tao1,DU Wenhui1,and ZHOU Xiaoyue1.Oil Geophysical Prospecting，2020,55(3):694-704.

After analyzing the preconditions,application,advantages and disadvantages of techniques for predicting fractures in granite buried hill reservoirs on the basis of a large number of references,it is concluded that: ①The structure-oriented filtering technique can improve the signal-to-noise ratio of post-stack seismic data,and the seismic data with WWH (wide frequency,wide azimuth and high density) can significantly improve the imaging accuracy of the top and inside of buried hills; ②Coherent volume is not affected by interpretation errors,but its description accuracy is lower,and it can only predict faults with discontinuous events; ③Curvature attributes can only identify the fault with folded formations on both sides of it; ④Ant bodies attributes can qualitatively identify small faults and large fractures,and combining with prestack azimuth anisotropy,they can quantitatively predict large and medium fracture zones; ⑤Compared with traditional post-stack attributes such as coherence,the maximum likelihood volume is accurate to predict fractures,and it can predict the interior of a fault zone and a dense fracture area,but it can’t provide the information about trunk faults; ⑥The anisotropic strength based on statistical data is more accurate than that based on ellipse fitting.

Keywords: granite buried hills,seismic attributes,fracture prediction,the maximum likelihood,WHH,OVT

1.Research Institute of Petroleum Exploration and Development,CNPC,Beijing 100083,China