ABSTRACTS

Optimization of parameter design for high-productivity marine blending acquisition.LI Peiming1,SONG Jiawen2,LIU Xinggang3,MA Zhu3,MA Yuanming2,and WANG Wenchuang2.Oil Geophysical Prospecting,2020,55（4）：707-715.

The high-productivity blending acquisition（HPBA）technology can greatly improve seismic acquisition，it is currently widely used in high-precision and high-density seismic survey.However，serious crosstalk noises exist in HPBA seismic data which should be removed in subsequent seismic processing.Parameters used in field blending acquisition directly affect the distribution of cross-talk noises and effectiveness noises removal.Unfortunately，there hasn’t been an effective method for optimizing such parameters that can ensure adequate noise removal.We propose a method based on raw and simulated data to design and optimize parameters such as simultaneous source spacing and vibrating time.The main steps are as follows：divide shooting zones and plan the navigating route according to available shooting and receiving equipment；simulate shooting time；simulate de-blended data to blended data；optimize acquisition parameters according to the difference between simulated blended data and real data.Tests using modeling and application to marine raw data have proved the method effective and practicable.Hence，it is a feasible method for optimizing field HPBA parameters.

Keywords：high-productivity blending acquisition（HPBA），simulation of blending acquisition，acquisition parameters analysis，simultaneous source spacing，vibrating time

1.BGP，CNPC，Zhuozhou，Hebei 072751，China

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

3.Acquisition Technique Center，BGP，CNPC，Zhuozhou，Hebei 072751，China

Reconstructing regularly missing seismic data based on anti-aliasing POCS algorithm.DIAO Su1,ZHANG Hua1,ZHANG Hengqi1,ZHANG Bohong1,YU Zheng1,and PANG Yang1.Oil Geophysical Prospecting,2020,55（4）：716-724.

Affected by complex terrains in the work area，raw seismic data may be incomplete and irregular，and spatial aliasing is serious.When collecting data on uniform spatial grids，spatial aliasing is similar to the real spectra of missing seismic data.However，anti-aliasing reconstruction can’t be realized based on Fourier transform and projection onto convex sets（POCS）.To make up the missing linear events，first we pick up effective waves by dip-scanning inf-kfrequency domain，and then establish a mask function and use it to the convex set projection algorithm to reconstruct the missing data.Numerical simulation has proved that the proposed method can reconstruct missing seismic data while effectively eliminating aliasing.

Keywords：Fourier transform，projection onto convex sets（POCS），mask function，anti-aliasing，threshold

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

Denoising and reconstruction of 3Dseismic data on a datadriven tight frame.CHEN Jie1,NIU Cong2,LI Yong1,3,HUANG Rao2,CHEN Lixin1,and MA Zechuan1.Oil Geophysical Prospecting,2020,55（4）：725-732.

Constrained by economic and geological conditions and other factors，generally seismic data are undersampled and noises are serious，which severely impact data processing and interpretation.We studied how to denoise and reconstruct 3D seismic data on a data-driven tight frame（DDTF）.The DDTF theory defines the learning dictionary to be a translation-invariant redundant wavelet tight frame.By further controlling the degree of freedom of the dictionary，it makes the DDTF algorithm have good robustness，and the fine features of seismic data can be better preserved by making use of perfect reconstruction on the wavelet tight frame.Model and real data have proved that the DDTF algorithm works well in denoising and reconstructing 3D synthetic seismic data with simple structures and real 3D seismic data with complex structures，but the computational efficiency is low and needs to be further improved.In addition，curvelet transform is less effective for denoising and reconstructing real data；the denoised and reconstructed results from block matched four-dimensional cooperative filtering are too smooth to preserve some structural features.

Keywords：3D seismic data，denoising，reconstruction，data-driven tight frame

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

2.CNOOC Research Institute Co.Ltd.，Beijing 100027，China

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

Low-rank one-step wave extrapolation for pure qP-wave forward modeling in viscoacoustic anisotropic media.GU Hanming1,ZHANG Kuitao1,LIU Chuncheng2,and WANG Jianhua2.Oil Geophysical Prospecting,2020,55（4）：733-746.

Forward modelling and reverse time migration（RTM）of pure quasi-P（qP）waves in anisotropic media have attracted a lot of attentions in recent years.However，the results have serious noises and low resolution because of neglecting the viscosity although considering the anisotropy of the media.The application of the conventional quasi-acoustic equation is limited due to the interference from quasi-shear waves，limitation of model parameters and instability of the simulating process.This paper introduces a one-step wave field continuation method，deduces the expression for viscoelastic media in space-wavenumber domain，and constructs a pure qP wave field continuation operator in space-wavenumber domain for viscoacoustic anisotropic media，and finally realizes the forward modeling of pure qP waves in viscoacoustic anisotropic media based on the Low-Rank decomposition algorithm.The modelling results indicate that：a.the seismic wavefield can simulate the anisotropy and viscosity of underground media，so it is consistent with underground conditions；the Low-Ran method is not limited by the weak points of the quasi-acoustic equation which is interfered by quasishear waves，restricted by model parameters and with instable simulating process；c.no numerical aliasing appears when the time step is appropriately increased.By improving computational efficiency while ensuring accuracy，the Low-Ran method is stable and effective in calculation.It provides a theoretical basis for reverse time migration for anisotropic media based on Q compensation.

Keywords：viscoacoustic anisotropy，pure qP wave，Low-Rank decomposition，one-step wave extrapolation，forward modeling

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

2.CNOOC Research Institute，Beijing 100028，China

Research on direct simulation to seismic record of prestack depth migration.WU Han1,LIAN Ximeng2,SUN Chengyu1,RUI Yongjun2,CAI Ruiqian1,and DENG Xiaofan1.Oil Geophysical Prospecting,2020,55（4）：747-753.

Pre-stack depth migration（PSDM）imaging is an important step for high-resolution and high-precision seismic exploration.In order to get good PSDM imaging results，and avoid the cumbersome processes of source record forward modeling，normal processing and migration，this paper proposes a direct simulation method for PSDM record.Without wavefield forward modeling and migration，this method uses wavelets，recording parameters and ray path to construct a point spread operator in wavenumber domain，and produces theoretical imaging results based the velocity model and the point spread operator.The method is simple and effective，and delivers precise imaging results.Based on the imaging results，problems in the recording geometry and the quality of seismic data can be detected.It is a reference to optimizing geometry parameters and evaluating imaging quality which are basic data for seismic interpretation.The method can meet the theoretical research requirements of seismic acquisition，processing and interpretation.

Keywords：simulation to PSDM record，generalized convolution，scattering tomography，point spread operator

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

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

Reflection waveform inversion based on cross-correlation misfit function.LI Qingyang1,2,WU Guochen1,2,DUAN Peiran1,LIANG Zhanyuan1,2.Oil Geophysical Prospecting,2020,55（4）：754-765.

On a velocity and density model，the density from conventional FWI（Full Waveform Inversion）is seriously migrated，namely low wavenumber information in the model hasn’t been recovered.Since RWI（Reflection Waveform Inversion）can estimate low wavenumber component，the combination of RWI and FWI can invert the density model more accurately.A velocitydensity RWI model based on crosscorrelation-based misfit function is proposed.This paper first reviews the two-parameter inversion based on the conventional acoustic wave equation，and analyze the migration characteristics of density inversion using a radiation mode；then synchronously updates velocity and density by the Gardner formula based on the high trade-off between velocity density model when running RWI，and input the RWI results as the initial model for traditional velocity-density FWI；and finally builds a velocity-density RWI＋FWI inversion workflow.Application on a simple two-layer model and a resampled Sigsbee 2A model has proved that RWI could avoid inversion to be trapped in local minima and suppress density migration；and in the process of RWI，low wavenumber information of velocity and density could be accurately recovered by using the Gardner formula to update velocity and density at the same time.

Keywords：acoustic wave equation with variable density，velocity，density，full waveform inversion（FWI），reflection waveform inversion（RWI），cross-correlation misfit function

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

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

Velocity modeling of the low-relief structure in the Matouying area of Jidong Oilfield.WANG Peng1,WANG Xiaowei1,YONG Yundong1,LIU Wei1,and QIEShuhai1.Oil Geophysical Prospecting,2020,55（4）：766-773.

The low-relief structure of the Guantao Formation in the Matouying area of Jidong Oilfield is very potential in exploration.However，with complex terrains and developed faults，and limited by less accurate depth-domain velocity models，it is difficult to identify the lowrelief structural traps.To this end，a new and comprehensive velocity modeling scheme is proposed as follows：Considering the influence of faults，first establish an accurate horizon interpretation model in time domain，and based on the model，effectively eliminate the impact of dramatic changes in lateral velocity caused by faults；second，considering that error accumulation of the shallow velocity model in prestack depth migration affects the imaging quality of deep target layers，invert the accurate near-surface velocity field by diving wave tomography for shallow layers，build the velocity model by reflected wave tomography for middle layers，and combine the two parts to obtain a merged velocity field；finally improve the velocity model on the basis of the merged velocity field and model-based high-precision grid tomography under the constraint of geological horizons.The high-precision velocity model established by the above scheme has significantly improved the imaging accuracy of the low-relief structure in the Matouying area in the Jidong Depression，and the result is consistent with drilling data.This modeling scheme is a reference to imaging low-relief structures with similar geological conditions.

Keywords：velocity modeling，low-relief structure，horizon interpretation model，diving wave tomography，grid tomography

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

Research on seismic processing technologies for improving RTM imaging.ZHANG Yishan1,GUO Jiuying2,ZHANG Mingyu3,WANG Yong4,WEI Zhengda4,and LI Wandong4.Oil Geophysical Prospecting,2020,55（4）：774-781.

Reverse-time migration（RTM）has been widely applied for imaging complex structures，and it is playing an important role in exploring complex structures.RTM needs high-quality raw seismic data，precise velocity models and RTM algorithm.In order to reduce to the effects of near surface anomaly，avoid the cyclic skip arising out of conventional full waveform inversion（FWI），and improve SNR，we developed several preprocessing and FWI technologies，and built a new RTM workflow including surface-consistent phase correction of spatial wavelets，near-surface Q compensation，pattern-based adaptive ground roll attenuation and pattern-based adaptive waveform inversion（PAWI）.These technologies can provide high-quality seismic data，high-precision and high-resolution velocity models，and consequently improve the image of geological structures.Data from several wells have proved these technologies effective to improve RTM imaging.

Keywords：reverse-time migration（RTM），patternbased，full waveform inversion，self-adaptive，dispersive ground-roll，cross-spread，phase matching，absorption compensation

1.Institute of Earth and Space Science，Peking University，Beijing 100871，China

2.GeoApex Technology（Beijing）Inc.，Beijing 100176，China

3.PetroChina Xinjiang Oilfield Company，Karamay，Xinjiang 834000，China

4.BGP Inc.，CNPC，Zhuozhou，Hebei 072751，China

Parallel optimization of reverse-time migration on staggered grids in pseudo-depth domain.JIN Zongwei1,2,HUANG Jinqiang1,2,WANG Ganlu1,2,XIA Peng1,2,and MU Yuliang1,2.Oil Geophysical Prospecting,2020,55（4）：782-792.

Reverse time migration based on the acoustic wave equation is one of the most important imaging methods at present，but it is constrained by computer speed and memory use.To reduce model grids for improving calculating efficiency and reduce memory use，this paper introduces curvilinear coordinate transform into forward modeling and reverse time migration，derives the first-order acoustic wave equation and its discrete form in the pseudo-depth domain，and gets the discrete equation with PML absorption boundaries，and then optimizes the algorithm of forward modeling and reverse time migration on the C＋＋AMP parallel architecture.The algorithm can greatly improve calculating efficiency while reducing memory occupation.Model test has proved the algorithm and parallel architecture correct and effective.

Keywords：reverse-time migration，pseudo-depth domain，C＋＋AMP parallel optimization，staggered grids

1.College of Resources and Environmental Engineering，Guizhou University，Guiyang，Guizhou 550025，China

2.Petroleum Society of Guizhou Province，Guiyang，Guizhou 550000，China

Viscoacoustic reverse time migration of prismatic wave for steeply dipped structures.QU Yingming1,2,3,WEI Zhefeng1,2,4,LIU Chang3,LI Zhenchun3,XU Kai4,and LI Runze5.Oil Geophysical Prospecting,2020,55（4）：793-803.

Prismatic waves carry more information about steeply dipped structures which primaries can not reflect.Therefore，prismatic waves are separately used in some migration methods to improve the illumination and imaging effect of steeply dipped structures.However，attenuation leads to amplitude loss and phase distortion of seismic waves，which seriously affects the accurate imaging of prismatic waves.A viscoacoustic RTM（reverse time migration）of prismatic waves is proposed for imaging steeply dipped structures.First aQ-compensated forward-propagating operator and a back-propagating adjoint operator are derived，and thenQattenuation is compensated along the propagating path of prismatic waves.Application on two typical steeply dipped structural models has proved that：①viscoacoustic RTM of prismatic wave can produce better images of steeply dipped structures with higher signalto-noise ratio，higher resolution and more balanced amplitude than conventional viscoacoustic RTM and acoustic RTM of prismactic wave，but losses some resolution of less dipped structures，therefore underground structures should be interpreted on viscoacoustic RTM and acoustic RTM of prismatic wave；②the time used for viscoacoustic RTM of prismatic wave is about 2 times that of viscoacoustic RTM，so the former needs a larger memory to save wavefields because it cannot reconstruct multiple wavefields at the same time.

Keywords：viscoelasticity，attenuation compensation，prismatic wave，reverse time migration（RTM），adjoint operator，steeply dipped structure

1.State Key Laboratory of Shale Oil and Gas Enrichment Mechanisms and Effective Development，Beijing 100083，China

2.SINOPEC Key Laboratory of Theory and Survey Technology for Seismic Elastic Waves，Beijing 100083，China

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

4.SINOPEC Research Institute of Petroleum Exploration＆Production，Beijing 100083，China

5.PetroChina Beijing Oil＆Gas Pipeline Control Center，Beijing 100007，China

Elastic Gaussian beam migration in anisotropic media.DU Xiangdong1,HAN Wenming2,CAO Xiangyang2,ZHANG Yingde2,ZHANG Shixin2,and LIU Qiang3.Oil Geophysical Prospecting,2020,55（4）：804-812,830.

Elastic wave Gaussian beam migration is a multicomponent seismic imaging method with higher imaging accuracy and computational efficiency.By now more researches have focused on acoustic and elastic imaging of isotropic media and acoustic imaging of anisotropic media，but less on elastic wave migration of anisotropic media.This paper proposes the ray tracing equation of elastic wave in anisotropic media，derives the forward and backward continuation formulas of the elastic wave field in anisotropic media based on the Kirchhoff-Helmholtz integration of 2D anisotropic elastic wave and using the Green tensor characterized by elastic dynamics，and establishes an elastic Gaussian beam migration method for multicomponent seismic data of anisotropic media.In addition，the weight function is introduced into the imaging formula to suppress the crosstalk caused by different waveforms；and the sign function is introduced to facilitate the polarity inversion during converted wave imaging.Applications on a TTI sag model and a TTI multilayer model show that the imaging results of anisotropic elastic Gaussian beam migration have higher SNR，more continuous seismic events，and more accurate structural position than the similar isotropic migration.With multicomponent seismic data，and information of P-wave and S-wave fields，the method provides the imaging result with higher resolution.Moreover，it has better adaptability than the conventional imaging method based on scalar wave field.

Keywords：anisotropic media，ray tracing，elastic wave，polarity correction，crosstalk，Gaussian beam migration

1.CNOOC Research Institute Co.，Ltd.，Beijing 100027，China

2.CNOOC International Co.，Ltd.，Beijing 100027，China

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

Time-domain Gaussian beam prestack depth migration for acoustic anisotropic media.QIN Ning1.Oil Geophysical Prospecting,2020,55（4）：813-820.

As an improved ray migration method，Gaussian beam migration not only inherits the efficiency and flexibility of ray methods，but also has the imaging accuracy close to wave equation migration methods.Compared with traditional Gaussian beam migration in frequency domain，the method in time domain greatly improves computational efficiency by adjusting the imaging formula and converting frequency integral into the function of imaging time.As seismic azimuth becomes larger and recording spread becomes longer，the influence of anisotropy on seismic migration can’t be ignored any longer.This paper introduces the anisotropic ray tracing algorithm based on phase velocity to timedomain Gaussian beam migration and optimizes the dynamic ray tracing correlation coefficient，and finally proposes a more efficient method for acoustic anisotropic media.Model tests have demonstrated that the method proposed has higher computational efficiency and better applicability while maintaining higher imaging accuracy than similar anisotropic algorithms，especially it can be used for imaging complex anisotropic structures.

Keywords：time domain，Gaussian beam，depth migration，TI media，ray tracing

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

Numerical analysis of frequency-dependent seismic responses from fractured-porous rock saturated with twophase immiscible fluids.LIU Yunfei1,CHEN Xuehua1,LUO Xin2,ZHANG Jie2,and NI Hui2.Oil Geophysical Prospecting,2020,55（4）：821-830.

Types and distribution of fluids in fractured-porous rock are important factors affecting the elastic parameters of the rock.In most cases，fluids in actual reservoirs are not single components，and the influences of their saturation states are rarely considered in numerical simulation.Based on the Chapman model，this paper discusses how saturation and capillary pressure affect the dispersion attenuation and seismic response of fractured-porous sandstone reservoirs saturated with two-phase immiscible fluids.The numerical results show that with the increase of oil or gas saturation，the characteristic frequency is different from that of the medium saturated with a single fluid，and it decreases first and then increases.Especially in the intermediate frequency band，such variation is the most outstanding，and the seismic responses are suffered from travel-time delay and distortion of the seismic waveform.When there is gas in the two-phase fluids，the dispersion，attenuation and seismic response are more sensitive to changes in saturation.As the parameter q measuring the capillary pressure between fluids increases，the state of two-phase fluids changes from“patchy saturation＂to“uniform saturation＂，and the characteristic frequency moves toward the direction with“high frequency and low saturation＂.The results of numerical simulation provide an theoretical basis for reservoir fluid identification.

Keywords：two-phase immiscible fluids，saturation，capillary pressure，dispersion and attenuation，seismic response

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

2.Key Laboratory of Earth Exploration and Information Technology，Ministry of Education，Chengdu University of Technology，Chengdu，Sichuan 610059，China

Locating borehole microseismic events based on the tracking component scanning.CUI Qinghui1,PAN Shulin2,DIAO Rui1,SHANG Xinmin1,and RUI Yongjun1.Oil Geophysical Prospecting,2020,55（4）：831-838,863.

Conventionally，locating borehole microseismic events is based on first arrivals and polarized information.However，if the SNR is low，they can’t work well.Theoretical studies show that theXandYcomponents of a borehole detector can be synthesized into a new vector in any direction on the horizontal plane of the detector.When the vector is the tracking component of microseismic events，it has the strongest energy and the best waveform consistency among all detectors.According to the nature of three-component microseismic record，this paper introduces the stacked energy scanning method commonly used in surface microseismic location into borehole microseismic location.First compose theXandYcomponents of a detector into a new component in the horizontal plane along the direction from an imaginary source to the detector.Second，stack the waveforms and calculate energyE1after NMO correcting the vectors of all detectors，then taking the sum of the energy of the synthetic vector of all detectors to getE2，and finally，taking the sum ofE1andE2（E1＋E2）as an objective function，find the grid with the largest cumulative energy，which is the location of the source point.It is not necessary to pick up first arrivals.Model and real data have proved the method is accurate to locate microseismic events and strong to resist noises.It can be used to locate microseismic events in the environment where the SNR is low.

Keywords：tracking component，borehole microseismic location，composite vector，stacked energy scanning

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

2.School of Geosciences and Technology，Southwest Petroleum University，Chengdu，Sichuan 610500，China

Prestack seismic inversion driven by mixture probabilistic models.LI Kun1,2 and YIN Xingyao1,2.Oil Geophysical Prospecting,2020,55（4）：839-853.

Prestack seismic inversion is the most important method for quantitatively evaluate the elastic，lithologic and fluid properties of subsurface media.Conventional seismic inversion often seperately predicts the‘elastic modulus’，‘discrete lithology’and‘fluid factor’while ignoring the influence of lithology on model parameters，so that the errors of prior knowledge introduced will seriously affect the accuracy of seismic inversion and lithology prediction.Considering that the prior probability density function（PDF）of model parameters follows the distributon of mixture probabilistic density，this paper derives the explicit solution to the posterior PDF under the constraints of four conditional datasets including time-domain seismic data，frequencydomain seismic data，composite low-frequency prior informaton and known model points on the Bayesian framework.After introducing non-linear boundary constraints into prestack elastic inverson，the solution to model inversion becomes stable.The posterior PDF is randomly sampled by the sequential simulation algorithm，and the simulation results of different posterior probability components are classified，consequently a prestack simultaneous prediction method is established for continuous‘elastic parameters’，‘discrete lithology’and‘fluid factor’.Model and real data have proved the method effectivee and practical.

Keywords：prestack seismic inversion，probabilistic inversion，mixture probabilitstic model，lithology classification，hydrocarbon identification

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

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

Application of pure P-wave seismic data for studying Dn4 sub-member reservoirs in Gaoshiti area.DING Bozhao1,SUN Zandong1,CAI Jiaming1,WANG Yajing1,ZHU Xinghui1,and XU Lijun2.Oil Geophysical Prospecting,2020,55（4）：854-863.

Compared with conventional stacking data，pure Pwave seismic data can provide information from an approximate zero incident angle，so that AVO effect can be eliminated，and amplitude distortion caused by multiple stacking can be avoided.Furthermore，pure Pwave seismic data can reduce the effect caused by NMO stretching，restore seismic resolution，and reflect real sedimentary phenomena.Taking the Dn4 member in the Gaoshiti area，the central Sichuan Basin as a case，we chose pure P-wave from the seismic data to study the Dn4 sub-member reservoir.The results show the pure P-wave data can：①correctly reflect the impedance contrast between the reservoir and the caprock；②describe the geological phenomena of bi-directional overlapping in the Qiongzhusi Formation，and accurately divide the high and low acoustic velocity in the caprock and identify the underlying reservoir；③describe the boundary between the upper and lower Dn4 submembers，and the lithologic difference between the transgressive system tract and the high water system tract，and the thinning bio-flat sediments of the upper Dn4 sub-member from the margin to center of the platform.These results provide new ideas for wells deployment in the Gaoshiti area.

Keywords：pure P-wave seismic data，Gaoshiti area，Dn4 member，caprock layers，sequence boundary

1.GRI，BGP Inc.，CNPC，Zhuozhou，Hebei 072751，China

2.BGP Offshore，CNPC，Tianjin 300457，China

3.Acquisition Technology Center，BGP Inc.，CNPC，Zhuozhou，Hebei 072751，China

Prediction of relative permeability and calculation of water cut of tight sandstone reservoir based on radial basis function neural network.WANG Qian1,TAN Maojin1,SHI Yujiang2,LI Gaoren2,CHENG Xiangzhi3,and LUO Weiping4.Oil Geophysical Prospecting,2020,55（4）：864-872.

Tight sandstone reservoir is characterized by poor physical properties，complex pore structures and strong heterogeneity.It is difficult for conventional methods to predict or estimate the relative permeability and water cut.This paper proposes to use the radial basis function（RBF）neural network to predict the relative permeability of tight sandstone reservoir.Based on the RBF neural network，we choose the Gaussian function and the nearest neighbor algorithm to build a network model，and take water saturation，nuclear magnetic irreducible water saturation，porosity and permeability as inputs and relative oil and water permeability as output to define the best relative permeability network model and parameters after error analysis，and finally calculate the water cut using the split flow equation.For the Chang 8 reservoir of the Yanchang Formation in the Longdong area of the Ordos Basin，the relative oil and water permeability predicted by this method is consistent with the experimental results，and the water cut calculated is consistent with the measured value too.

Keywords：tight sandstone，radial basis function（RBF），relative permeability，split flow equation，water cut

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

2.Research Institute of Exploration and Development，PetroChina Changqing Oilfield Company，Xi’an，Shaanxi 710021，China

3.PetroChina Research Institute of Petroleum Exploration＆Development，Beijing 100083，China

4.Research Institute of Exploration and Development，PetroChina Tarim Oilfield Company，Korla，Xinjiang 841000，China

Recognition of single well reservoir architecture in alluvial fan based on K-means clustering and Bayes discrimination.ZHANG Yang1,2,ZHAO Pingqi1,LU Fengming1,LI Ji3,GUO Zhiqiao1,and WANG Rui1.Oil Geophysical Prospecting,2020,55（4）：873-883.

The division of single well reservoir architecture is a basic research.The problems are less effective，human factors and different standards to divide reservoir architecture using traditional methods.Taking the alluvial fan in the Cangdong sag as a case，we propose a single well reservoir recognition method for alluvial fan based on core description，K-means clustering and Bayes discrimination with core and well data.Applied for the well not cored，the reservoir architecture in the well was recognized accurately，and the cause why previous recognization was not accurately was found.The new method works in four steps：divide the reservoir architecture in the well cored，build the dividing standard，define the discrimining formula，and identify the reservoir architecture in the well not cored.It is important to define the standard and identification formula for the well not cored based on the reservoir architecture divided in the well cored.Due to low resolution of logging data and lithologic changes in the transition belt，the accuracy of identifying the reservoir architecture in the well not cored positively depends on the thickness of the reservoir architecture unit.Generally，the recognition accuracy of seventh-，eighth-，ninthorder reservoir architecture units reduces in turn.

Keywords：reservoir architecture recognition in single well，K-means clustering，Bayes discrimination，Cangdong sag，alluvial fan

1.PetroChina Dagang Oilfield Company，Tianjin 300280，China

2.Karamay Campus，Faculty of Petroleum，China Unirersity of Petroleum（Beijing），Karamay，Xinjiang 834000，China

3.Experiment Testing Institute，PetroChina Xinjiang Oilfield Company，Karamay，Xinjiang 834000，China

Segmented evolution of Borjianghaizi faults and their control on Taiyuan-Shanxi Formations.XU Litao1,SHI Wanzhong1,2,WU Rui1,WANG Ren1,XU Qinghai3,and ZHANG Xiaoming1.Oil Geophysical Prospecting,2020,55（4）：884-891.

The Sanyanjing faults，Wulanjilinmiao faults and Borjianghaizi faults are distributed in an echelon form from west to east.They are composed of a dominant fault zone in the Hangjinqi area in the northern Ordos Basin.The formations and traps on the two sides of the fault zone are different.The reservoirs on the south side are closely related to fault activities besides the control of sedimentary facies.How fault activities control sedimentation is not clear.We studied the structural characteristics of the Borjianghaizi faults by using 3D seismic data，and analyzed the evolution of the faults according to regional dynamics and the control of fault activities on the reservoir distribution.The following conclusions have been drawn：①Today’s Borjianghaizi fault zone is composed of five secondary isolated faults.These faults were segmented and thrusting during the Hercynian period，then they were interconnected with the large-scale thrusting movement of the Late Jurassic during the Yanshanian Period，and underwent structural inversion and transformation in the multi-stage extensional-compressional cycles from the Early Cretaceous to the Neogene，and consequently today’s structural pattern is presented.②During the Taiyuan-Shanxi period，the structural transfer zones between the faults were well corresponding to the distribution of sediments on the south side，and the transfer zones or inactive faults controlled the development and distribution of large sedimentary structures in the south.

Keywords：Boerjianghaizi fault，segmented evolution，transfer zone，sedimentary control

1.Faculty of Earth Resources，China University of Geosciences（Wuhan），Wuhan，Hubei 430074，China

2.Key Laboratory of Tectonics and Petroleum Resources of the Ministry of Education，China University of Geosciences（Wuhan），Wuhan，Hubei 430074，China

3.College of Earth Science，Yangtze University，Wuhan，Hubei 430100，China

Formation classification of Juyanhai Depression in Yin-E Basin.ZHANG Jinxue1,LI Fulei1,SUN Bo1,YANG Shengquan1,ZHANG Boen1,and MA Yuzhu1.Oil Geophysical Prospecting,2020,55（4）：892-897.

There are different opinions on how to classify the formations of Lower Cretaceous in the Juyanhai Depression in the Yin’e Basin.The abrupt change in vitrinite reflectance is used to identify the unconformities.The Estherian fossils found in well Mengedi 1 is used to analyze the formations.The lithological and electrical characteristics shown in wells such as Tian1 prove that the Permian formation is characterized by low AC and high Rt.Finally，according to the unconformity on the seismic profiles，the Lower Cretaceous Bayingobi Formation identified by previous studies is classified into the Permian.This finding suggests a great oil and gas potential in the Permian exploration in the Juyanhai Depression.

Keywords：Yin-E Basin，Juyanhai Depression，formation classification，Permian，vitrinite reflectance，seismic reflection characteristics

1.Urumqi Branch，GRI，BGP Inc.，CNPC，Urumqi，Xinjiang 830016，China

Time-frequency electromagnetic（TFEM）technique：stepby-step constraint inversion based on artificial fish swarm algorithm.HE Zhanxiang1,2,3,HU Zuzhi4,WANG Zhigang4,ZHAO Yunsheng1,SHEN Yibin1,3,and LIU Xuejun4.Oil Geophysical Prospecting,2020,55（4）：898-905.

In oil and gas exploration，time-frequency electromagnetic（TFEM）inversion can use more known data.It is of great significance to make full use of known data and effectively reduce the non-uniqueness of inversion.Based on the artificial fish swarm algorithm，we developed an objective function for TFEM inversion under joint constraints of horizontal electric field and vertical magnetic field，designed three kinds of model spaces including layer thickness，resistivity and polarization for suitably constraining horizon parameters，proposed the methods and rules for setting model space with the purpose of making full use of known data，and developed step-by-step constrained inversion schemes for three types of models.These measures effectively reduce the non-uniqueness of inversion and improve inversion accuracy.Model and real data tests have proved that the artificial fish swarm algorithm and step-by-step constrained inversion schemes can reduce the non-uniqueness and ensure exploration accuracy and effect，so it can be used widely.

Keywords：time-frequency electromagnetic（TFEM），artificial fish swarm，model space，step-by-step constraint inversion

1.SUSTech Academy for Advanced Interdisciplinary Studies，Shenzhen，Guangdong 518055，China

2.Southern Marine Science and Engineering Guangdong Laboratory（Guangzhou），Guangzhou，Guangdong 511458，China

3.Shenzhen Key Laboratory of Deep Offshore Oil and Gas Exploration Technology，Department of Earth and Space Science，SUSTech，Shenzhen，Guangdong 518055，China

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

Staged adaptive regularized inversion of magnetotelluric data.GUO Yihao1,2,CHEN Xiao1,2,YANG Haiyan1,2,ZHANG Zhiyong1,2,ZENG Zhiwen1,2,and ZHOU Yong1,2.Oil Geophysical Prospecting,2020,55（4）：906-914.

How to reasonably choose a regularization factor has always been a hotspot and difficulty in geophysical inversion.Choosing the regularization factor from the perspective of improving the stability of inversion is a new idea.In addition，the randomness of nonlinear optimization algorithm is not considered enough in previous studies on regularized inversion.Based on the adaptive algorithm proposed by Zhdanov，we propose a self-adaptive regularized algorithm，also called“Staged Adaptive Algorithm＂，which adjusts the regularization factor stage by stage.We use the algorithm for magnetotelluric（MT）conjugate gradient inversion and differential evolution inversion.Model tests show that the new adaptive algorithm can improve the stability of inversion，reduce the influence of the attenuation factor to some extent，and it is applicable for linear and nonlinear optimization algorithms.

Keywords：staged adaptive algorithm，regularization factor，magnetotelluric（MT），conjugate gradient，differential evolution algorithm

1.Engineering Research Center of Nuclear Technology Application（East China University of Technology）Ministry of Education，Nanchang，Jiangxi 330013，China

2.School of Geophysics and Measurement-control Technology，East China University of Technology，Nanchang，Jiangxi 330013，China

Cubic-spline-interpolation-based FFT and its application in forward modeling of gravity and magnetic fields.ZHOU Yinming1,2,DAI Shikun1,3,LI Kun1,3,HE Zhanxiang4,HU Xiaoying2,and WANG Jinhai5.Oil Geophysical Prospecting,2020,55（4）：915-922,930.

Discrete Fourier transform（DFT）plays an important role in signal processing，spectral analysis and numerical solution to partial differential equation（PDE）.In the process of discretization，conventional FFT causes spectrum confusion and boundary effects.FFT based on Gauss integral weakens truncation effects，but reduces calculation efficiency.A Fourier transform method based on cubic-spline interpolation is proposed.First it discretizes Fourier transform integral into multiple units，and uses cubic spline interpolation to represent the function change in every unit，and finally adds the integrals of all units and gets the results of Fourier transform.This method makes full use of the high-order continuity and sampling flexibility of spline interpolation integration，and analytic calculation of unit integral.It provides a new idea for efficient and precise Fourier transform.Using the Gauss function designed，we compared 1D and 2D forward and inverse transforms with analytical solutions.It is verified that the theory of the method is correct，and its calculation accuracy is high.A continuum model is designed，then we applied fast Fourier transform based on cubic-spline interpolation to 3D numerical simulation of gravity and magnetic fields.The results show that the method can reduce the influence of truncation effect，and has high calculation accuracy and efficiency.Also，this has been proved by real data.

Keywords：cubic-spline interpolation，FFT，flexible sampling，3D modeling of gravity and magnetic fields

1.School of Geosciences and Info-physciences，Central South University，Changsha，Hunan 410083，China

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

3.Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Monitoring，Ministry of Education，Changsha，Hunan 410083，China

4.SUSTech Academy for Advanced Interdisciplinary Studies，Shenzhen，Guangdong 518055，China

5.The 3rd Geological Prospecting Institute of Qinghai Province，Xining，Qinghai 810029，China

Fourth-order spectral moment analysis for extracting the information of crustal arc structures.FU Lihua1,ZENG Cheng2,YANG Wencai3,and LI Yanmei4.Oil Geophysical Prospecting,2020,55（4）：923-930.

The geometric feature analysis of potential field based on spectral moment is a new and automatic edge detection method.It can identify more even and convergent geometric features of geological bodies than conventional methods.Second-order spectral moment can describe linear scratches，but it cannot well recognize other kinds of scratches.Fourth-order spectral moment can recognize arc and other scratches，so we propose an arc coefficient.Theoretical experiments show that the coefficient can reflect the geometric features of circular arc or intersecting linear scratches，and has low sensitivity to depth，so it can extract weak signals of curvature arc scratches in the potential field.Applied in the Tarim Basin，this method has successfully recognized isometric structures like paleo-volcanic group and other geological bodies based on magnetic anomalies，indicating that fourth-order spectral moment can extract necessary information and effectively recognize special crustal structures.

Keywords：geometric feature analysis of potential field，fourth-order spectral moment，edge detection，curvature arc coefficient，crustal structure

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

2.School of Computer Science and Technology，Xidian University，Xi’an，Shaanxi 710071，China

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

4.School of Mathematics，Harbin Institute of Technology，Harbin，Heilongjiang 150001，China

Mechanism analysis and function positioning of magnetic RTP-L methods in wave-number domain.CHAI Yupu1,and WAN Haizhen1.Oil Geophysical Prospecting,2020,55（4）：931-937.

Using the Reduction-to-the-pole（RTP）Algorithm-Error（A-E）equation as a theoretical tool，The authors conduct mechanism analysis and function positioning of three wave-number-domain RTP methods for magnetic anomalies at low latitude（RTP-L）.The suppression RTP method is ineffective for quantitative interpretation，but effective for qualitative interpretation of magnetic anomalies at low latitude.Theoretically，the inversion RTP method is applicable for quantitative interpretation of magnetic anomalies at low latitude，but it demands that the unknown data outside the calculation window should be well modeled，so it is not practical because measured data are not satisfying.The shift sampling RTP method may be applicable for quantitative interpretation of magnetic anomalies at low latitude，but the data（grids）in the magnetic declination direction can’t be more than 64 and the data accuracy should not be less than 0.2%.

Keywords：applied Reduction-to-the-pole Algorithm-error（RTP-AE）equation，wave-number domain，suppression RTP method，inversion RTP method，shift sampling RTP method

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