ABSTRACTS

3Dseismicdatareconstructionbasedonafaststructuredictionarylearningmethod.LANNanying1,ZHANGFanchang1,ZHANGYiming2，QINGuangsheng3，andDINGJicai2.OilGeophysicalProspecting，2020，55(1)：1-9.

Currently，the 3D seismic data reconstruction methods based on dictionary learning usually reconstruct the data slice by slice.This strategy neglects the correlation between slices，and doesn’t make full use of the continuity constraints in various directions of seismic data.To solve this problem，a 3D joint reconstruction method based on fast structure dictionary learning was proposed.Under the framework of compressive sensing theory，the method uses fast structure dictionary learning algorithm to train the training set in order to generate a 3D adaptive dictionary，and then reconstruct the data with high precision using 3D adaptive dictionary，observation matrix and regularized orthogonal matching pursuit algorithm.The reconstruction results of model data and real data demonstrated that the method can recover the detailed characteristics of seismic data with high precision and good performance on amplitude preservation.

Key words：fast structure dictionary learning，3D seismic data reconstruction，compressive sensing，amplitude-preserved processing

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

2.CNOOC Research Institute Co.Ltd，Beijing 100028，China

3.SINOPEC Zhongyuan Oilfield，Puyang，Henan 457099，China

KeytechniquesforbroadbandprocessingofplanestreamerdatainBohaiSea.WANGYandong1,2,WANGXiaoliu1,2,SANGShuyun1,2,WANGJianhua1,2,ZHANGJinmiao1,2,andWENGBin1，2.OilGeophysicalProspecting,2020,55(1)：10-16.

In recent years，few progresses have been made in broadband seismic exploration technology in shallow water.Especially in shallow water area such as Bohai Sea，where the average water depth is not greater than 30m，broadband acquisition is difficult to be implemented，and only can be realized through conventional acquisition and broadband processing.Ghost and shallow water multiple are the two main factors which restrict the broadband processing of shallow water data.In this paper，an adaptive deghosting method in τ-p domain was proposed to solving the problem of fierce time-space variation of ghost waves in shallow water.An adaptive shallow water multiple suppression method in τ-p domain was proposed to solve the problem that seabed model of Bohai Sea is difficult to be established.The application results in real seismic data in Bohai Sea demonstrated that the two key techniques proposed in this paper can broaden the frequency bandwidth of seismic data effectively，so as to realize the broadband processing of conventional plane streamer data in Bohai Sea.

Keywords：Bohai Sea，ghost wave，shallow water multiple，adaptive parameter estimation，broadband processing

1.CNOOC Research Institute Co.，Ltd.，Beijing 100028，China

2.National Engineering Laboratory for Offshore Oil Exploration，Beijing 100028，China

Adirectinversionmethodfordeblendingsimultaneous-sourcedata.WANGKunxi1，2，3，MAOWeijian1,2,ZHANGQingchen1,2,LIWuqun1,2,ZHANYi4，andSUNYunsong4.OilGeophysicalProspecting,2020,55(1)：17-28.

In recent years，simultaneous-source acquisition technique has shown a broad application prospect in high-density and wide-azimuth seismic exploration.Based on the limited spatial bandwidth of seismic data，a direct inversion method for deblending simultaneous-source data was discussed in this paper.Different from conventional iterative algorithms with constraints，iteration is not required in this method.A basic point-spread matrix was added to the pseudo inverse of time-delay operator，and thus the direct separation of simultaneous-source data without iteration was realized in frequency domain.By modifying the basic point-spread matrix，this method was further extended to the separation of simultaneous-source data acquired with irregular receiver and shot arrays geometry.Theoretical model tests demonstrated that the method achieves good separation effect with high accuracy and efficiency when the deblending condition given in the paper is satisfied.

Keywords：simultaneous source，direct inversion，irregular array，point-spread matrix，deblending

1.Center for Computational and Exploration Geophysics，Institute of Geodesy and Geophysics，Chinese Academy of Sciences，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

SeismicdatareconstructionusingdiscreteorthonormalS-Transformbasedoncompressivesensing．ZHAOZiyue1,LIZhenchun1,andZHANGMin1．OilGeophysicalProspecting,2020,55(1)：29-35．

The reconstruction effect and computational efficiency of different sparse transform methods in compressive sensing are different.Therefore，a seismic data reconstruction method using discrete orthonormal S-transform (DOST) based on compressive sensing technology was proposed in this paper.By taking the inner product of a set of orthogonal basis functions and time series，the time-frequency matrix was obtained，in order to make original signals more sparse and thus improve the compressed sensing reconstruction effect of seismic data.This method makes up for the limitation that S transform cannot be used as the sparse transform in compressive sensing，and a new sparse transform method was introduced to the theoretical system of compressive sensing.Theoretical model test and real data application achieved overall satisfactory reconstruction effect，and demonstrated the fast iteration speed and stable convergency of the method proposed in this paper.

Keywords：compressive sensing，sparse transform，discrete orthonormal S-transform，data reconstruction

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

3Dseismicdatareconstructionbasedonsparselear-ningviaiterativeminimization.DAIZhigang1,LIUZhihui1,andWANGJinyan1．OilGeophysicalPro-specting,2020,55(1)：36-45．

In seismic exploration，affected by the factors such as acquisition environment，technology and cost，some shots or traces can be missing in field data.The incompleteness of seismic data will have adverse effect on later seismic data processing and imaging，thus the reconstruction of these missing data is essential.In this paper，a sparse learning via iterative minimization (SLIM) method was proposed to reconstruct random missing 3D seismic data.It reconstructs 3D missing seismic data based on the 2D harmonic structure of frequency slice.Firstly，apply Fourier transform to 3D seismic data along time direction.Secondly，use cyclic minimization (CM) algorithm to solve the 2D harmonic spectrum of frequency slice iteratively.Finally，apply inverse Fourier transform to the estimated spectrum，and thus reconstruct the missing data.Besides，conjugate gradient least squares(CGLS) is applied to calculate the inverse in data reconstruction，in order to speed up the reconstruction.Test results indicate that the proposed SLIM method achieved good performance on both synthetic and real 3D seismic data，and it performed better than multi-channel singular spectrum analysis(MSSA) method using singular Hankel matrix based on frequency slice.

Keywords：3D seismic data reconstruction，cyclic minimization，spectrum estimation，conjugate gradient least squares

1.School of Mathematics and Physics，China University of Geosciences (Wuhan)，Wuhan，Hubei 430074，China

Seismicwavefielddecompositionmethodbasedonvectorrotationinτ-pdomain.ZHANGJing1,ZHANGWendong1,ZHANGTieqiang1,SUNPengyuan1,YUANYijun2，andLiJianfeng1．OilGeophysicalProspecting,2020,55(1)：46-56.

High-fidelity elastic wave field decomposition has become a key processing step in multi-wave and multi-component seismic exploration.Conventional wave filed decomposition methods can be classified into two categories，kinematic methods and dynamic methods.Kinematic methods decompose the wave filed into P-P wave field and P-SV wave field inτ-pdomain or in spatial domain.Dynamic methods focus on obtaining polarization characteristics of different types of wave.How-ever，neither of them can achieve ideal decomposition results because of their inherent limitations.Motivated by the idea of combining both kinematic and dynamic characteristics of the wave field，a wave field decomposition method based on vector rotation inτ-pdomain was proposed in this paper.It takes near-surface velocity as the key parameter of wave field decomposition，and applies polarity rotation to the vertical and horizontal components of seismic wave field inτ-pdomain.According to the polarization characteristics of different types of wave，it decomposes seismic wave field into P-P wave field and P-SV wave field to the correspon-ding polarization directions respectively.Model tests were carried out based on layered medium model and Marmousi Ⅱ model，and the method was also applied to real seismic data.The results demonstrated that the method can separate P-P wave and P-SV wave more accurately compared with conventional methods.It avoided energy anomaly，wave field aliasing and spatial aliasing，and preserved amplitude well after wave field decomposition.

Keywords：multi-component，wave field decomposition，τ-ptransform，vector rotation

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

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

Adeghostingmethodforvariable-depthstreamerdatabasedonnon-Gaussianmaximization．FENGQiang1,HANLiguo1,andYANGFan1．OilGeophy-sicalProspecting,2020,55(1)：57-63．

The variable-depth streamer acquisition technology suppresses ghost effectively and obtains wide-band data based on notch diversity.Sea surface reflection coefficient and the time delay of ghost wave are the two most important parameters affecting ghost suppression effect，and cannot be obtained directly.In this paper，the mirrorred record generation formula and the joint deconvolution deghosting formula with these two parameters were derived.Based on non-Gaussian maximization，taken the negative entropy as the non-Gaussian measure，the sliding spatio-temporal data window was used to solve the ghost wave parameter variation problem.The optimized sea surface reflection coefficient and the time delay of ghost wave were obtained，and then the ghost was suppressed by joint deconvolution.Model test and the application in real data processing demonstrated that the method suppressed the ghost in variable-depth streamer data effectively，widened frequency band and improved seismic resolution.

Keywords：variable-depth streamer acquisition，non-Gaussian，joint deconvolution，deghosting

1.College of Geo-exploration Science and Technology，Jilin University，Changchun，Jilin 130026，China

Free-surface-relatedmultiplepredictionforcomplexseafloor．LIXiaozhang1,DENGYong1,HEJianwei1,RENTing1,andGUHanming2．OilGeophysicalProspecting,2020,55(1)：64-70.

For complex seafloor with complex wavepath，diffracted wave and diffracted multiples，multiple model predicted by SRME method has low signal-to-noise ratio，incorrect dynamic and kinematic information.Multiple prediction method based on the extrapolation of one-way and two-way wave equation was proposed in this paper.Based on wave equation and SRME method，it uses rock physics model to extrapolate the received wavefield，to derive the predicted multiple model.It also uses adaptive subtraction to improve multiple elimination effect for complex seafloor.Model tests and the application in real data demonstrate that multiple model predicted by the method proposed in this paper has higher signal-to-noise ratio and matches better with real multiples in dynamic and kinematic characteristics，compared with that predicted by individual SRME method.Theoretically，it is nece-ssary to know accurate rock physics model when using this method.For the majority of multiples in marine seismic data are free-surface-related multiples，given seawater and seafloor rock physics parameters，the method proposed in this paper can predict the majority of free-surface-related multiples.

Keywords：complex seafloor，multiple prediction，surface-related multiple elimination (SRME)，two-way wave equation，one-way wave equation

1.Zhanjiang Branch，CNOOC，Zhanjiang，Guangdong 524057，China

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

Seismicfirst-breakpickingbasedonBPneuralnetworkintegratedwithmomentummethodandadaptivelearningratemethod.CAOXiaoli1,LIUBin1,WANGShurong1,WANXuejuan2，ZHANGTingting1,andZHANGHaixin1.OilGeophysicalProspecting,2020,55(1)：71-79.

A seismic first-break picking method based on BP neural network integrated with momentum method and adaptive learning rate method was proposed in this paper.It improves the network weight updating process.If mean square error is not within the given error range，the weight update is cancelled.Otherwise，the weight is updated，and the learning rate changes accordingly.Through the analysis on the feasibility of first-break identification using different seismic attributes，four typical attributes，including RMS amplitude ratio，curve length ratio，amplitude and frequency，were chosen for model test.Model test results indicated that the improved method performs better than conventional BP neural network method.The application in real data demonstrated that the improved BP neural network algorithm has simple network structure，few parameters，fast convergence speed，good performance in stability and anti-noising and high first-break picking precision.

Keywords：first-break picking，momentum method，adaptive learning rate method，neural network algorithm，anti-noising performance

1.Shengli Branch，Geophysical Company，SINOPEC，Dongying，Shandong 257086，China

2.Huabei Branch，GRI，BGP，CNPC，Renqiu，Hebei 062552，China

CompressivesensingmethodwithHubernormminimizationconstraintonreconstructionerror.LIZhong-xiao1,LIYongqiang2，3，GUBingluo2，3，andLIZhenchun2，3.OilGeophysicalProspecting，2020，55(1)：80-91,135.

Seismic data contain strong noise outliers.The compressive sensing(CS) method based on L2norm minimization constraint on reconstruction error assumes that the reconstruction error satisfies Gaussian distribution.Therefore，the CS method above cannot remove super-Gaussian noise outliers.To better remove outliers and improve interpolation accuracy，Huber norm was used instead of L2norm to implement the minimization constraint on reconstruction error.The minimization constraint of Huber norm is equivalent to the L1norm minimization constraint on large reconstruction error (noise outlier) and the L2norm minimization constraint on small reconstruction error (Gaussian random noise).Therefore，the proposed method is robust when dealing with noise outlier.Furthermore，theoretical pseudo seismic data were introduced to convert the Huber norm minimization problem to the L2norm minimization problem，in order to solve the Huber-L0minimization problem of the proposed CS method based on Huber norm minimization constraint on construction error.Additionally，the affection of Gaussian noise intensity，noise outlier intensity and parameter selection on interpolation accuracy is tested.The processing results of synthetic and field data demonstrated that the proposed CS method based on the Huber norm minimization constraint of the construction error can better remove the noise outliers and preserve effective signals compared with the CS method based on the L2norm minimization constraint of the construction error.

Keywords：compressive sensing method，interpolation，Huber norm，pseudo seismic data，noise outlier,elimination

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

3.Laboratory for Marine Resources,Qingdao National Laboratory for Marine Science and Techno-logy,Qingdao,Shandong 266071,China

AdaptivefocusedbeammigrationinVTImedia.LIShengya1,2,LYUQingda1,2,HUANGJianping1,2,LIZhenchun1,2,HUZiduo3，andLIUDingjin4.OilGeophysicalProspecting,2020,55(1)：92-100.

Adaptive focused beam is an improvement of Gaussian beam.Considering the affection of local velocity field on beam width，it focuses multiple times to control the effective energy of seismic beam within one wavelength.Different from Gaussian beam，adaptive focused beam selects initial beam parameters dynamically.For adaptive focused beam is advantageous in dealing with the i-maging problem in the media with strong horizontal velocity variation，the adaptive focused beam migration method was applied to VTI media in this paper.Based on the anisotropic ray tracing equation system represented by classical elastic parameters，the adaptive focused beam migration in VTI media was realized.The comparison between the beam shape of traditional Gaussian beam and that of adaptive focused beam revealed the advantage of adaptive focused beam in beam shape control.Sub-Sag model test demonstrared the accuracy and stablity of adaptive focused beam migration method in anisotropic media.Anisotropic SEG/Hess model test demonstrated the feasibility of the method proposed in this paper on complex geological model.The application in real data processing demonstrated that the imaging effect of the method proposed in this paper is better than conventional Gaussian beam migration methods in anisotropic media.

Keywords：Gaussian beam，adaptive focused beam，anisotropy，ray tracing，beam shape

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

2.Pilot National Laboratory for Marine Science and Technology (Qingdao)，Qingdao，Shandong 266580，China

3.Northwest Branch，Research Insititute of Petroleum Exploration and Development，PetroChina，Lanzhou，Gansu 730022，China

4.SINOPEC Geophysical Research Insititute，Nanjing，Jiangsu 211103，China

VSP-CDPstackimagingbasedontheweightfunctionofnormaldistribution.YANGFeilong1,2,LIHuifeng1,SUNHui3，4，ZHANGXue1,LUOHao1,andZHAOChi1.OilGeophysicalProspecting,2020,55(1)：101-110.

Limited by the geometry，the imaging sections based on migration imaging technique have serious “arc phenomenon” at the edge.VSP-CDP stack imaging method based on ray tracing can realize amplitude-preserved VSP imaging.However，it is difficult to obtain accurate structural imaging and velocity model in complex structure situation.Therefore，a non-zero-offset VSP stack imaging method was proposed in this paper，based on dynamic ray tracing effective neighborhood wave field approximation theory.By studying the nature and characteristics of normal distribution，the weight function based on normal distribution stack was derived and used in VSP stack imaging.All the sampling points in depth-time domain were converted into multiple sampling points of reflection points in offset-depth domain，in order to uniform the fold times of reflection points.Model test and the application in real data demonstrated that the VSP-CDP stack imaging method based on normal distribution weight function improved the VSP imaging precision.

Keywords：VSP，dynamic ray tracing，normal distribution，weight function，stack imaging

1.School of Earth Sciences and Engineering，Xi’an Shiyou University，Xi’an，Shaanxi 710065，China

2.Key Laboratory of Shaanxi Province Hydrocarbon Geology TIBET，Xi’an，Shaanxi 710065，China

3.Shandong Key Laboratory of Depositional Mineralization and Sedimentary Minerals，College of Earth Science and Engineering，Shandong University of Science and Technology，Qingdao，Shandong 266590，China

4.Faculty of Geosciences and Environmental Engineering，Southwest Jiaotong University，Chengdu，Sichuan 611756，China

AnoptimizedmethodforextractinganisotropicparametersinTTImedia.YANGZongqing1,LIHongwei1,OUJugang1,CHANGMeisi1,andLINYang1.OilGeophysicalProspecting,2020,55(1)：111-116.

When formation thickness varies greatly，the inverted parameter model derived by traditional anisotropic parameter modeling based on well-seismic joint grid tomography in TTI media can be distorted.To solve this problem，an optimized anisotropic parameter modeling method in TTI media was proposed in this paper.It introduces formation thickness information into anisotropic parameter inversion through theδ-based well-seismic error interpolation，and thus makes the inversion result more accurate.Real data processing results demonstrated that the optimized method can avoid the anisotropic parameter distortion caused by the abrupt change of formation thickness effectively，and the obtained model parameters are more accurate and reasonable for they are not related to formation thickness variation.

Keywords：TTI media，anisotropic parameter inversion，well-seismic joint，parameter modeling，grid tomography

1.Southwest Geophysical Research Institute，BGP，CNPC，Chengdu，Sichuan 610000，China

GashydrateS-wavevelocitypredictionmethodbasedoneffectivemediummodel.MENGDajiang1,2,WENPengfei1,2,ZHANGRuwei1,2,ZHAOBin1,2,andLIYan1，2.OilGeophysicalProspecting,2020,55(1)：117-125.

Gas hydrate has different filling modes.It can be a part of pore fillings or a component of solid matrix in strata.So，the rock physical model of gas hydrate is quite different from conventional rock physical models for oil and gas，and conventional S-wave velocity prediction method is not suitable for gas hydrate.Pertinently，a method to predict S-wave velocity of gas hydrate based on effective medium model was proposed in this paper.Firstly，the influence of mineral composition，porosity and hydrate saturation on P-wave and S-wave velocities was analyzed.Secondly，according to the effective medium model，a constrained optimization equation was established based on the conventional well log data such as P-wave slowness，density，shale content，porosity and saturation.The optimization equation was constrained by conventional log data such as P-wave velocity and density，porosity and saturation were used as optimization variables to find the optimal solution.The trust-domain algorithm was applied to solve the constrained optimization equation with fast convergence rate and reliable calculation results.Finally，the efficiency of the proposed method was proved by the drilling data of gas hydrate in Shenhu area，Northern South China Sea.The results of S-wave velocity and saturation match real data well.

Keywords：effective medium model，gas hydrate，rock physics，S-wave velocity prediction，constrained optimization

1.Key Laboratory of Submarine Mineral Resources，Ministry of Natural Resources，Guangzhou，Guangdong 510075，China

2.Guangzhou Marine Geological Survey，Guangzhou，Guangdong 510075，China

Momenttensorinversionmethodfromboreholedataconstrainedbyshear-tensilesourcemodel.TANGJie1,LICong1,LIUYingchang1,andCHENXueguo2.OilGeophysicalProspecting,2020,55(1)：126-135.

Hydraulic fracturing microseismic data can be used to estimate moment tensor and study the azimuth and detailed characteristics of fractures，which have a great effect on depicting the fracture system inside the reservoir.Moment tensor inversion is a linear inversion method.It is not sufficient to invert all moment tensors from single well data.Therefore，additional constraints，near-field data or simplified source models should be considered.Pertinently，the shear-tensile source constrained inversion was studied in this paper.It uses nonlinear inversion method to limit the moment tensor to describing the shear-tensile source，in order to reduce inversion parameters and enhance robust.Firstly，the principle of trust-domain microseismic source mechanism inversion constrained by shear-tensile source was introduced.And then，the application effect of the method was tested using the theoretical model records of single well，double wells and three wells under different tensile angle conditions.The following conclusions were drawn.The moment tensor inversion and the shear-tensile source constrained inversion based on the data of three wells achieved good results when the test data didn’t contain any noise.When the data of few wells were used，the moment tensor inversion couldn’t obtain reasonable results.When the test data contained noise，the shear-tensile source constrained inversion achieved better anti-noise performance，and the inversion effect based on the data of three wells is better than that based on single-well data.The amplitudes derived by shear-tensile source inversion are more selective.

Keywords：microseismic，source mechanism inversion，shear-tensile source，moment tensor，tensile angle

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

2.Exploration and Development Research Institute，SENOPEC Shengli Oilfield，Dongying，Shangdong 257015，China

Sensitivityanalysisofmulti-modeRayleighandLovewavephase-velocitydispersioncurvesinhorizontallayeredmodels.YINXiaofei1,XUHongrui2，HAOXiaohan3，SUNShida4，andWANGPeng1.OilGeophysicalProspecting,2020,55(1)：136-146.

High-frequency surface-wave method with Rayleigh wave and Love wave as the main study objects is widely applied in various fields such as underground water，environment and engineering.For horizontal layered models，S-wave velocities are the most important parameters for estimating multi-mode Rayleigh wave and Love wave phase-velocity dispersion curves.In this paper，the Jacobian matrix was used to infer the sensitivities of multi-mode Rayleigh wave and Love wave phase-velocity dispersion curves to S-wave velocities of the formations at different depths.The following conclusions were drawn.①The sensitivities of surface waves of different modes to the S-wave velocity at a certain depth are different.For both Rayleigh wave and Love wave，low-frequency surface wave is more sensitive to the S-wave velocities in deep formations，compared with high-frequency surface wave.Moreover，multi-mode phase velocities with a broad high-frequency range are sensitive to surface S-wave velocity.②Based on a velocity-increasing layered model and two layered models with velocity anomalies (containing low-velocity interlayer and high-velocity interlayer)，the analysis results suggested that the phase-velocity dispersion curves of both Rayleigh wave and Love wave are sensitive to S-wave velocity in low-velocity layer，and they are both not sensitive to the S-wave velocity in high-velocity layer or below velocity-anomaly layer (low-velocity layer or high-velocity layer).③For surface wave of a certain mode，the frequency band of Love wave，in which phase-velocity dispersion curves are sensitive to S-wave velocity of a specific layer，is wider than that of Rayleigh wave.In addition，the sensitivity peaks of Rayleigh wave and Love wave phase-velocity dispersion curves to S-wave velocity of a certain layer are different.Therefore，joint inversion of multi-mode Rayleigh wave and Love wave phase-velocity dispersion curves can be applied to obtain high-precision shallow subsurface S-wave velocity.

Keywords：Rayleigh wave，Love wave，multi-mode，phase-velocity dispersion curves，Jacobian matrix，sensitivity analysis

1.Institute of Earthquake Forcasting，China Earthquake Administration，Beijing 100036，China

2.Subsurface Imaging and Sensing Laboratory，China Unversity of Geosciences (Wuhan)，Wuhan，Hubei 430074，China

3.Zhejiang Provincial Institute of Communications Planning，Design and Research，Hangzhou，Zhejiang 310000，China

4.MOE Key Laboratory of Fundamental Physical Quantities Measurements，School of Physics，Huazhong University of Science and Technology，Wuhan，Hubei 430074，China

Eliminationofstrongreflectioninfluencebasedonoptimizedvariationalmodedecompositionmethod：acasestudyofthetargetprocessingofbeachbarsandofEs4inDongyingSag.JIANGYu1,ZHANGJunhua1,HANHongwei2，F(xiàn)ENGDeyong2，andYUJingqiang2.OilGeophysicalProspecting,2020,55(1)：147-152,166.

Affected by the shielding effect of the interface between Chunshang submember and Chunxia submember of Es4 in Dongying Sag，the seismic reflections of the beach bar sand reservoir in Chunxia submember are weak.Optimized variational mode decomposition technology was used to decompose the target layer seismic signals into multi-band components with different waveform features to suppress the influence of the strong reflection shielding effectively.Strong reflection information with strong energy and low frequency，mostly concentrate in the first component.Through stripping strong signal components and reconstructing resi-dual information，the effective signals of underlying reservoirs can be strengthened.Theoretical model test and the application in real data demonstrated that the method has higher resolution，more complete decomposition and avoids modal aliasing effect compared with conventional empirical mode decomposition method(EMD).It is worth applying to the regions with similar conditions.

Keywords：empirical mode decomposition，variational mode decomposition，Chunxia submember，strong shielding，beach bar sand，thin interbed

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

2．Geophysical Research Institute，SINOPEC Shengli Oilfield，Dongying，Shandong 257022，China

Acompositeseismicattributeusedtoestimatethesandthicknessforthinbedandthininterbed.WANGYanguang1,LIHao2，LIGuofa2，LIULibin1,CAOGuoming3，andZHANGHuiqing3.OilGeophysicalProspecting,2020,55(1)：153-160.

Amplitude and frequency，the two basic seismic attributes，are often used to predict formation thickness when the thickness is smaller than a quarter of the wavelength.However，these two attributes have a non-linear relationship with formation thickness，and it reduces the quantitative prediction accuracy of sand body thickness.In addition，the function that relates the two attributes to sand body thickness is derived from single sand body wedge model.When it is applied to thin interbed model，the estimation error is inevitable.To solve this problem，a composite seismic attribute composed of amplitude and frequency was proposed in this paper.Firstly，a single sand body wedge model was used to test its precision in sand thickness estimation.Secondly，based on an interbed model superimposed by two sand wedges，the composite seismic attribute was used to estimate the cumulative sand thickness.Finally，the composite seismic attribute was used in real data to estimate the cumulative thickness of thin interbed sand bodies.The results showed that the composite seismic attribute not only improves the estimation accuracy of single thin sand body thickness，but also can be applied to the quantitative prediction of the cumulative thickness of thin interbed sand bodies.

Keywords：seismic attribute，wedge model，thin interbed，sand thickness，reservoir prediction

1.Geophysical Research Institute，SINOPEC Shengli Oilfield，Dongying，Shandong 257022，China

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

3.Research Institute of Exploration and Development，Dagang Oilfield Company，PetroChina，Tianjin 300280，China

Fracturezonepredictionbasedonrandomforestalgorithm.HEJian1,2,WENXiaotao1,2,NIEWen-liang1，3，LILeihao1,andYANGJixin1.OilGeophy-sicalProspecting,2020,55(1)：161-166.

Fracture zone prediction and characterization are of great significance for the exploration and development of fractured oil and gas reservoirs.In order to solve the multi-solution problem of the prediction methods using single attribute，multiple seismic attributes were used comprehensively.The relationships between fracture development degree and seismic attributes are often non-linear.Therefore，random forest algorithm was used to learn the correspondence between seismic attribute characteristics and fracture development degree，and then the fracture development degree in the study area was determined comprehensively according to the learning results，aiming to improve the prediction precision of fracture zone.The application in real data demonstrated that random forest algorithm achieved fracture zone prediction results with high accuracy，and the method is universal generally.

Keywords：fracture zone，comprehensive prediction，random forest，seismic attribute

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

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

3.School of Electronic and Information Engineering，Chongqing Three Gorges University，Wanzhou，Chongqing 404000，China

Theapplicationofprestackgeostatisticalinversioninthepredictionofshalesweetspotsandthininterbeds：acasestudyofBlockWinWesternCanadaBasin.GUOTongcui1,JIANGMingjun2，JIYingzhang2，WANGHongjun1,MAWenji3，andKONGXiangwen1.OilGeophysicalProspecting,2020,55(1)：167-175.

Shale gas is a kind of unconventional natural gas.The prediction of shale “sweet spots” with high TOC，high brittleness and developed fractures between wells is always an important step in shale gas exploration and development in the regions with few wells.The target shale formation in Block W in Western Canada Basin is relatively thin (the maximal seismic reflection time is 20ms)，and contains multiple thin limestone interlayers (0.5～3m).There are overlapping intervals between the P-wave impedances of brittle shales and ductile mudstones.It is difficult to predict shale sweet spots and thin limestone interbeds using seismic data.In this paper，a rock physics template for evaluating shale “sweet spots” was established based on the analysis on the rock-physical elastic parameters，and the elastic characteristics of shale “sweet spots” and interbeds were clarified.The 3D lithology-constrained pre-stack geostatistical inversion technique was used to predict the shale with high TOC and high brittleness，as well as limestone interbeds.Firstly，the rock physics parameters of the shale reservoir were analyzed.The prospective shale objectives have low Poisson’s ratio，median Young’s modulus，high TOC and high brittleness，while the limestone interbeds have high Poisson’s ratio，high Young’s modulus，low TOC and high density.Secondly，1D constrained geostatistical inversion was applied to predict and test the parameters of prospective shale area and interbeds.Thirdly，3D constrained prestack geostatistical inversion was applied to predict prospective shale area and interbeds.The results showed that the overlapping areas of high-TOC and high-brittleness shale and developed fractures are the target shale “sweet spots”.The developed shale “sweet spots” area with thin limestone interbeds are chosen as the prospecting well location area，in order to ensure a high probability that horizontal well drills to high-quality reservoirs with high gas production..

Keywords：shale gas，rock physics，geostatistical inversion，sweet spots，fracture prediction

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

2.CNODC，CNPC，Beijing 100034，China

3.School of Electronics Engineering and Computer Science,Peking University，Beijing 100054，China

TectonicevolutionprocessoftroughsanditscontroloneffectivesourcerockdistributioninTananSag.JIWenting1,SUNYonghe1,SUNXu1,andLIULu1.OilGeophysicalProspecting,2020,55(1)：176-186.

In order to study the evolution process of the troughs in the central fault depression belt in Hailaer-Tamuchage Basin and its control on effective source rock distribution，Tanan Sag was chosen as the target area.Based on 3D seismic interpretation data，fault throw-distance curves，the fault throw-depth curves，the stratigraphic sedimentary characteristics of trough section and the characteristics of transverse anticline between troughs of the major boundary faults were analyzed.The maximum fault throw back-stripping method was applied to restore the evolution history of faults，thus ancient trough distribution was defined，and the structural evolution law of troughs in Tanan Sag was discussed.Combined with the achievements of former researchers，the control of the troughs on effective source rock distribution was studied.The following conclusions were drawn.Based on the fault throw-depth curves and growth indexes，the active periods of the major boundary faults in Tanan Sag were determined，and the stratigraphic deposition periods of the upper part of Nantun1 Formation and the Nantun2 Formation were determined as the main evolution periods of troughs.Combined with the major boundary fault growth mechanisms (isolated fault growth mechanism and fault segmentation growth connection mechanism)，the trough evolution model in Tanan Sag were divided into two patterns：troughs controlled by isolated fault and troughs controlled by segmental growth faults.According to the different trough evolution patterns，the through-controlled distribution of hydrocarbon source rocks was divided into two patterns：source rocks controlled by isolated troughs and sources rocks controlled by segmental growth troughs.Meanwhile，the activity of the major boundary faults controlled the formation of the accommodation space and evolution of effective source rocks.According to the oil and gas distribution and vertical oil-bearing area in Tanan Sag，the oil-bearing zones distribute like “potatos” horizontally in the effective source rock distribution area and along the major boundary faults，and mainly distribute in Nantun Formation and Tongbomiao Formation vertically.

Keywords：Tanan Sag，trough，fault growth mechanism，fault throw-distance curve，fault throw-depth curve，trough evolution pattern，source rock distribution

1.School of Earth Sciences，Northeast Petroleum University，Daqing，Heilongjiang 163318，China

Numericalsimulationofarraylaterologresponsesinanisotropicformationwithmudinvasion.SIZhaowei1,DENGShaogui2，3，LINFawu1,YUANXiyong2，3，LIHaitao2，3,andTIANChaoguo1.OilGeophysicalProspecting,2020,55(1)：187-196.

Mud invasion and formation anisotropy can cause the separation of array laterolog curves with different investigation depths.The curve separation is the key to the inversion of formation parameters.Based on 3D finite element method(3D-FEM)，the array laterolog responses in anisotropic formation with mud invasion were simulated and the affection of anisotropy，mud invasion，wellbore deviation and surrounding rock on the curve separation was studied.The following conclusions were drawn.①In vertical wells，array laterolog curves are mainly affected by the resistivity along bedding direction.With the increment of stratigraphic dip angle，the contribution of vertical resistity increases.In infinite thick anisotropic formation，there is a negative separation among the curves (RLA1～RLA5) with different investigation depths，when the angle is small.When the angle is large，the separation is positive.The critical angle is around 60°.②The influence of mud invasion on curve separation is obviously greater than that of anisotropy.③For anisotropic reservoirs with mud invasion，the difference characteristics among the curves with different investigation depths are messy，which makes it difficult to identify reservoir fluids (oil or water) using the differences among the curves.④For isotropic formations with layered mud invasion，the curve separation degree decreases with the increment of well deviation angle，affected by low-resistivity surrounding rocks.When the mud invasion is not deep (0.1～0.4m)，the curve separation degree increases with the increment of well deviation angle.The achievements of the study on array laterolog responses in anisotropic formation with mud invasion have referential significance for data processing and reservoir evaluation.

Keywords：array laterolog，anisotropy，3D FEM，mud invasion，curve separation

1.Exploration and Development Research Institute，Jidong Oilfield Company，PetroChina，Tangshan，Hebei 063004，China

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

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

Simultaneousinterpolation，edgepaddinganddenoisingmethodforgravitydatabasedontheprojectionontoconvexsets.ZENGXiaoniu1,LIXihai1,HOUWeijun1,andLIUJihao1．OilGeophysicalProspecting,2020,55(1)：197-205.

Gravity exploration blocks are often irregular，which causes the vacancy in the acquired gravity data.Before the processing and transformation of gravity data in wavenumber domain，the original data must be interpolated and processed with edge padding.High-frequency noise in gravity data is the main factor which causes the instability in later processing.Conventional gravity data processing generally performs interpolation，denoising and edge padding independently.These three issues were considered integrally，and an iterative method for simultaneous interpolation，denoising and edge padding of gravity data based on the projection onto convex sets was proposed.Firstly，the final cutoff wavenumber of the iterative method was determined through calculating and fitting the radial average power spectrum of the gravity data.Secondly，interpolation，edge padding and denoising were applied to gravidy data using the spectrum with the wavenumber lower than the cutoff wavenumber，until the preset iteration number was reached.Theoretical gravity model test and the application in real isostatic gravity data acquired in Afghanistan showed that the method proposed is simple theoretically and convenient to be applied.The splicing of the interpolation and edge padding results is smooth without distortion，and the method achieved good interpolation and denoising effect.The results of the method proposed in this paper are better，compared with those of the conventional joint processing method based on minimum curvature，Kriging interpolation，wavelet denosing and cosine edge padding.

Keywords：gravity data，projection onto convex sets，denoising，interpolation，edge padding

1.Rocket Force University of Engineering，Xi’an，Shaanxi 710025，China

Gravityinterpretationusingimprovedsmallsubdomainfilteringofenhancedanomaly.WANGYanguo1,HUANGYuansheng1,andZHANGJin1.OilGeophysicalProspecting,2020,55(1)：206-216．

Small subdomain filtering(SSF) is widely used for picking anomaly boundaries of gravity data，but the method could make anomaly curve distorted because of the unreasonable subdomain division.In this paper，the subdomain division pattern of SSF was improved to reflect the structures with diffe-rent strikes.Moreover，a stable anomaly-enhanced filtering method based on iterative differential was proposed to solve the problem that small subdomain filtering is easily affected by high-frequency disturbance and enhance anomaly boundary identification.Compared with the conventional SSF，in model tests，the improved SSF of enhanced anomaly is more stable，and the detected edges are more consistent with boundaries of field sources.In the application to the gravity data of Yalu River Basin，the improved SSF of enhanced anomaly output more detailed results than conventional SSF.The detected boundaries match lithologic contact zones well，and the negative value area can generally reflect the distribution of low-density rocks in research area.It demonstrated the effectiveness and practicability of the method proposed in this paper，and the results facilitate coalfield exploration and geological-geophysical comprehensive interpretation.

Keywords：small subdomain filtering，anomaly boundary，random disturbance，enhanced anomaly，Yalu River Basin

1.Fundamental Science on Radioactive Geology and Exploration Technology Laboratory，East China University of Technology，Nanchang，Jiangxi 330013，China

MTdatainversionbasedonimprovedcuckoosearchalgorithm.WANGPengfei1,andWANGShuming1.OilGeophysicalProspecting,2020,55(1)：217-225.

Due to the high non-linearity of the magnetotelluric(MT) data inversion，conventional global optimization algorithms converge slowly and easily to local optimum.To solve this problem，an improved cuckoo search(ICS) algorithm combined with simplex method was proposed in this paper to realize MT data inversion.For cuckoo search(CS) algorithm is advantageous in exploration but dis-advantageous in development，the global optimal solution in particle swarm optimization was introduced to improve the local search performance.Simplex method was also used to improve the bird’s nest，in order to enhance optimization precision further.The inversion results of theoretical model and real data demonstrated that ICS has higher stability，faster convergence rate and higher accuracy than CS.

Keywords：magnetotelluric，improved cuckoo search algorithm，particle swarm optimization，simplex method

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

NonlinearconstrainedjointinversionofMTandgravitydata.HUZuzhi1,SHIYanling1,LIUYun-xiang1,LIUXuejun1,SUNWeibin1,andHEZhan-xiang2，3．OilGeophysicalProspecting,2020,55(1)：226-232．

It’s an important way to improve the resolution of the gravity,magnetic and electromagnetic exploration to apply the constrained and joint inversion of electromagnetic and gravity data using the known seismic，geological and logging data.The joint inversion in this study is mainly based on the relationship between resistivity and density by logging data statistics.Nonlinear artificial fish swarm inversion algorithm and parallel design were used，combined with the constraints of priori information such as logging data，seismic and geological interpretation sections，to realize the parallel joint inversion of magnetotelluric(MT) and gravity data，and the resolution of MT and gravity data was improved.The inversion results of model and field data demonstrated the feasibility of the proposed nonlinear constrained joint inversion.

Keywords：magnetotelluric，gravity，artificial fish swarm，constrained inversion，joint inversion

1.GME & Geochemical Surveys of BGP，CNPC，Zhuozhou,Hebei 072751，China

2.Academy for Advanced Interdisciplinary Stu-dies，Southern University of Science and Technology，Shenzhen，Guangdong 518055，China

3.Department of Earth and Space Sciences，Sou-thern University of Science and Technology，Shen-zhen，Guangdong 518055，China