ABSTRACTS

Suppressionofseismicrandomnoiseusingdual-channelconvolutionalneuralnetwork.XUYankai1，LIUZengmei1，XUEYaru1，andCAOSiyuan2.OilGeophysicalProspecting,2022,57(4):747-756.

Random noise suppression has always been a focus in seismic data processing，and the traditional methods can hardly balance the removal of noise and the preservation of effective signals.Therefore，a random noise suppression method based on a dual-channel convolutional neural network(SDC-CNN)is proposed in this paper.SDC-CNN is composed of two different sub-networks aiming to extract complementary information in the process of noise suppression.In the lower channel of the network，dilated convolution is introduced to expand the receptive field，which can fully capture the neighborhood information in seismic data and better preserve the details.Moreover，the idea of residual learning and the Swish activation function are adopted to improve the denoising performance of the network.The synthetic and field data expe-riments all demonstrate that the proposed method can effectively suppress random noise while preserving more texture details.

Keywords：seismic data，random noise，dual-channel convolutional neural network，dilated convolution，activation function

1.College of Information Science and Engineering，China University of Petroleum(Beijing)，Beijing 102249，China

2.College of Geopgysics, China University of Petroleum(Beijing)，Beijing 102249，China

Seismicinternalmultiplesuppressionmethodwithencoder-decoderconvolutionalnetworkbasedondataaugmentation.LIUXiaozhou1，HUTianyue1，LIUTao2，WEIZhefeng2，XIEFei2，andANShengpei2.OilGeophysicalProspecting,2022,57(4):757-767.

Internal multiple suppression is a cutting-edge technology challenge in seismic data denoising，which is of great significance for obtaining high-quality data and understanding the real subsurface structure.Current suppression methods of internal multiples are time-consuming and have high requirements for manual parameter tuning，which may lead to internal multiple leakage when processing data with a low signal-to-noise ratio(SNR).Therefore，this paper proposes an internal multiple suppression method with the encoder-decoder convolutional neural network(CNN)based on data augmentation.First，we estimate the primaries and internal multiples from the raw data by the internal multiple suppression method based on virtual events to obtain primary labels.Then，we establish two augmented training datasets.On the one hand，the internal multiple data are augmented by the change in amplitude，polarity，and travel time of internal multiples in the training samples to raise the generalization ability of the internal multiple suppression network.On the other hand，the Gaussian noise augmented datasets are obtained after different levels of Gaussian noise are added to the raw data，which can improve the anti-noise performance of the network.Finally，a deep encoder-decoder network suitable for internal multiple suppression is built for neural network training and prediction by the combination of the advantages of the denoising CNN(DnCNN)and U-shaped fully connolutional network(U-Net).The tests on synthetic and field data indicate that the proposed method can effectively suppress internal multiples and protect primaries and has strong generalization ability and anti-noise performance，which can significantly improve computational efficiency.

Keywords: internal multiple suppression，virtual event，convolutional neural network，encoder-decoder network，data augmentation

1.School of Earth and Space Sciences，Peking University，Beijing 100871，China

2.SINOPEC Petroleum Exploration and Production Research Institute，Beijing 100083，China

Intelligentmatchingmethodbasedondeeplearningformultiwaveseismicsignalsanditsapplication.LINGLiyang1，XUTianji2，3，FENGBo1，XUHongtao1，andWEIShuijian4.OilGeophysicalProspecting,2022,57(4):768-776.

In the field of oil and gas exploration and development，high-precision matching of multiwave seismic signals needs to be conducted to give full play to the technical advantages of multiwave and multi-component seismic exploration.This research proposes an intelligent matching method based on deep learning for multiwave seismic signals.Different from the traditional method that changes such features of the seismic signal as propa-gation time，phase，and frequency to complete the matching，this method uses the powerful feature extraction ability of the convolutional neural network(CNN)to directly extract the waveform features of the seismic signal.Moreover，converted wave(PS)extraction by resampling，longitudinal wave(PP)and converted wave feature loss weighting，and the Adam gradient descent algorithm to update PS wave features are also applied so that the waveform of the PS wave approaches that of the PP wave in the time domain with no overall changes.The dynamic，kinematic，and geometric features，such as the propagation time，phase，and frequency，of multiwave seismic signals are matched automatically through the waveform matching between the PP wave and the PS wave.The application of the 3D3C seismic data from Xinchang in the Western Sichuan Depression shows that this method does not require manual intervention in the matching of multiwave seismic signals and that it has the advantages of high precision，high efficiency，intelligence，and automation.In addition to maintaining its original characteristics，the matched PS wave obtains a dominant frequency，bandwidth，and waveform closer to those of the PP wave.Effectively describing the formation contact relationship and being more conducive to geological interpretations，such as fault identification，formation tracking，and lithological boundary chara-cterization，the proposed method lays a solid foundation for subsequent applications such as multiwave contrast geological interpretation and joint inversion.

Keywords：multiwave matching，high precision，intelligence，automation，deep learning，convolutional neural network(CNN)

1.School of Computer Science and Engineering(School of Cyberspace Security)，University of Electronic Science and Technology of China，Chengdu，Sichuan 611731，China

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

3.Yangtze Delta Region Institute of University of Electronic Science and Technology of China，Huzhou，Huzhou，Zhejiang 313000，China

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

Seismicdataregularizationbasedondeeplearningcombiningwaveletdomain.ZHANGYan1，LIJie1，WANGBin1，LIXinyue1，andDONGHongli2，3.OilGeophysicalProspecting,2022,57(4):777-788.

Data regularization is a fundamental step in seismic data processing，and the conventional me-thod based on physical modeling requires massive computations and is not widely in use.At present，the regularization methods of seismic data based on convolutional neural networks(CNNs)are usually limited in the time domain，which leads to the problems of the excessively smooth reconstructed data and severe loss of texture details，especially at a low sampling rate.Wavelet analysis has the characteristics of multiple scales and multiple directions，which is more suitable to represent the texture characteristics of two-dimensional data and can focus on the details of seismic data signals.Therefore，a CNN model combining the wavelet domain is proposed to learn the joint distribution characteristics of seismic data in the time and wavelet domains and thus approximate the actual data.Specifically，the reconstruction of irregular seismic data is transformed into the wavelet coefficient prediction of components of different directions in each scale under the framework of a CNN to reconstruct regularized seismic data.A joint loss function in the time and wavelet domains is constructed，and by the overall distribution and local details of seismic data，the network model is constrained.The attention of CNN learning can be adjusted by the modification of the weight of the joint loss function to raise the signal-to-noise ratio(SNR)of the reconstructed seismic data.The experiments demonstrate that the proposed method can better preserve details compared with other methods，and it is insensitive to the missing location of seismic data and has good robustness.

Keywords：regularization of seismic data，deep learning，joint loss function，wavelet transform，convolutional neural network

1.School of Computer & Information Technology，Northeast Petroleum University，Daqing，Heilongjiang 163318，China

2.Artificial Intelligence Energy Research Institute，Northeast Petroleum University，Daqing，Heilongjiang 163318，China

3.Heilongjiang Provincial Key Laboratory of Networking and Intelligent Control，Daqing，Heilongjiang 163318，China

Seismicdatareconstructionmethodbasedonunsupervisedresidualnetwork.MENGHongyu1，2，YANGHuachen1，2，andZHANGJianzhong1，2，3.OilGeophysicalProspecting,2022,57(4):789-799.

Seismic data collected in the field usually have the problem of missing seismic traces.Reconstructing such traces has always been a difficult problem in seismic data processing.The deep learning(DL)method currently used mainly adopts a supervised learning approach for seismic data reconstruction，that is，it needs to use complete seismic data as labels to train the network model.Nevertheless，accurate labels for measured field data are difficult to obtain，and the dependence on a large number of training samples affects the application of the depth learning method in seismic data reconstruction.Therefore，this paper proposes a seismic data reconstruction method of unsupervised deep learning based on a residual network.Instead of using complete seismic data as the training set to train the residual network，this method takes random data as the input of the residual network，with the seismic data containing the missing seismic traces as the expected output of the network.Through the back propagation of the error between the network predicted output and the expected output，the network parameters are iteratively optimized to minimize the error，obtain the residual network with the optimal parameters，and use the network to reconstruct the missing seismic data.During network parameter optimization，the local and translation invariant properties of convolution are leveraged to learn the similar features between seismic data neighborhoods at multiple scales with the convolution filter，and the learned prior features are presented in the network output.This method is used to reconstruct the regular and irregular missing traces in the seismic data simulated with the Marmousi model and the measured marine streamer data，and the results are compared with those of the traditional fast projection onto convex set-soft threshold(FPOCS-Soft)method.The comparison shows that the proposed unsupervised residual network method can effectively reconstruct missing seismic traces，offer results with high accuracy and continuity，and outperforms the FPOCS-Soft method in precision.

Keywords：seismic data reconstruction，convolutional neural network，deep learning，residual network，unsupervised learning

1.Key Laboratory of Submarine Geosciences and Prospecting Techniques，MOE China，Qingdao，Shandong 266100，China

2.College of Marine Geosciences，Ocean University of China，Qingdao，Shandong 266100，China

3.Functional Laboratory of Marine Mineral Resources Evaluation and Exploration Technology，Pilot National Laboratory for Marine Science and Technology(Qingdao)，Qingdao，Shandong 266100，China

Real-timeacousticpositioningmethodbyappyingrobustestimationforseabedexploration.FANGYihan1，HUWenkai1，andYUSujun2.OilGeophysicalProspecting,2022,57(4):800-805.

Due to the fact that there are a large number of burst noise and gross errors in the real-time acoustic positioning process of seabed exploration，these unqualified observations will seriously affect the positioning accuracy of seabed transponders if they are not processed correctly.This paper analyzes the application fields and limitations of intersection positioning with depth constraints，differential positioning based on weight selection iteration，and positioning methods considering sound ray bending.Toward better positioning accuracy and reliability，this paper proposes a real-time acoustic positioning method based on robust estimation，which considers the working characteristics of real-time positioning and single-direction navigation in the acoustic positioning data acquisition of seabed exploration.Specifically，the possible range of observation values is given to eliminate burst noise，and the gross errors are detected by the probability and statistical hypothesis method.Moreover，the extreme value function is constructed to determine the weight matrix of the observation values，and the spatial position of the seabed transponder is obtained through iterative calculations.The simulations and field data analysis reveal that the proposed method can effectively eliminate the burst noise and gross errors and obtain the optimal unbia-sed estimation.Compared with the current real-time acoustic positioning software based on least square estimation，the proposed method improves the internal coincidence accuracy of acoustic positioning in seabed exploration and the reliability of positioning results.It meets the needs of real-time route correction of a lofting navigation ship and ensures the lofting accuracy of seabed geophones.

Keywords：seabed exploration，real-time acoustic positioning，robust estimation，extreme value function, route correction

1.School of Automation，China University of Geosciences，Wuhan，Hubei 430074，China

2.Geophysical Equipment Research Center,BGP Inc.，CNPC，Zhuozhou，Hebei 072751，China

SeismicexplorationtechnologiesandapplicationsincomplexmountainareasofKuqaDepression.CHENXueqiang1，WANGYanfeng1，LYUJingfeng1，ZHOUXu1，andWANGJian1.OilGeophysicalProspecting,2022,57(4):806-814.

The surface and underground conditions of Kuqa Depression in Tarim Basin are extremely complex，and thus the previous seismic data cannot meet the needs of oil and gas exploration.This paper summarizes the technical progress and practical application effect of seismic exploration in complex mountain areas of Kuqa Depression in recent years.First，the design technology of 3D recording geometry based on depth-domain imaging of complex structure is formed，and it is clarified that the parameter design should give priority to the selection of small line spacing，followed by wide azimuth，and moderate bin.Second，the physical point layout and geometry-variable design technology based on LiDAR data is developed to realize the accurate construction according to the design，which ensures the uniformity of 3D attributes.Third，the joint excitation and supporting technology of well shots and high-precision vibroseis has been formed to achieve high-efficiency acquisition and construction on the basis of the data with improved quality.Fourth，the field receiving technology based on node instruments is created to adapt to local conditions，which greatly improves the flexibility of the acquisition technology and construction scheme.Fifth，the "two-step" field surface modeling technology based on multi-data is developed to accurately build a near-surface structure model of the piedmont in mountain areas.The application of these technologies improves the accuracy of structural ascertainment.

Keywords：3D observation geometry design，physical point layout，node instrument，surface modeling

1.Tarim Geophysical Division，BGP Inc.，CNPC，Korla，Xinjiang 841000，China

Observationsystemtransformationbasedontheconstructionofequivalentsubsets.ZHANGHua1，WANGMeisheng1，WUKe2，CHUFangdong2，ZHANGHongtao1，andGAOZihan1.OilGeophysicalProspecting，2022,57(4):815-827.

The development of three-dimensional seismic acquisition technology featuring wide azimuth and high density requires a great amount of seismic acquisition equipment.Under the condition of limi-ted resources of seismic acquisition equipment or the condition of high seismic receiving cost and low excitation cost，it is suitable to adopt the “using shots rather than traces” seismic acquisition method to transform the observation system in the originally designed seismic acquisition arrays.The precondition for the observation system transformation is that the cross subsets constructed by the two observation systems are the same to ensure consistent bin attributes，achieving the expected prospecting targets.The results from practical applications show that the push-pull observation system and the large-cross observation system can easily construct the equivalent subset and realize the accurate transformation using shots rather than traces.The study of these two observation systems is the basis for the analysis and design of the complex observation system transformation using shots rather than traces.

Keywords：observation system transformation，cross subsets，shots rather than traces，push-pull observation system，large-cross observation system，parallel observation system，Patch observation system

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

2.New Resources Geophysical Exploration Division，BGP Inc.，CNPC，Zhuozhou，Hebei 072751，China

Shearlet-domainconsistencyprocessingmethodbasedonstatisticaldistributionformulti-surveydata.TIANKun1，WANGDeying2，LIULibin1，WANGYanguang1，WANGChangbo1，andZHANGXuetao1.OilGeophysicalProspecting，2022,57(4):828-837,869.

A relatively accurate time-space-varying consistency correction factor is the key to improving the quality of multi-survey data processing.At present，the commonly used consistency processing methods for multi-survey data mainly include the deconvolution parameter adjustment method and the matched filtering method.Although these methods have achieved certain application results，they still fail to meet the needs of refined consistency processing of multi-survey data due to their lack of time-space-varying processing ability，susceptibility to noise disturbance，and poor stability.To solve this problem，this paper proposes a Shearlet-domain consistency processing method based on the statistical distribution for multi-survey data.Specifically，the multi-survey data are transformed into Shearlet-domain，and the time-varying mean and mean square deviation at different scales and in different directions are estimated by the Alpha-trim mean filtering method.Subsequently，a certain amount of data are selected in the exploration area with relatively high data quality，and the statistical time-varying mean and mean square deviation are calculated in Shearlet-domain to serve as the target model.Then，the temporal and spatial low-frequency trends of the mean and mean square deviation of the data to be processed are extracted in the Shearlet-domain，and the time-space-varying correction factors of the mean and the mean square deviation are calculated with the extracted low-frequency trends and the target model.Finally，the data to be processed are corrected with the time-space-varying correction factors to obtain the results of multi-survey data consistency processing.Model-based tentative calculation and field data processing are conducted to verify this method.The results show that the proposed method deserves practical application as it can effectively eliminate the differences among multi-survey data in amplitude，frequency，and space，and it has favorable stability in noisy environments.

Keywords：statistical distribution，Shearlet-domain，multi-survey data，consistency processing

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

2.College of Earth Science and Engineering，Shandong University of Science and Technology，Qing-dao，Shandong 266590，China

WaveforminversionbasedonadaptivesourcewaveletextractionandcorrectionforRayleighwaves.SHAOGuangzhou1，DUTing1，andWUHua2.OilGeophysicalProspecting,2022,57(4):838-846.

Full waveform inversion(FWI)based on Rayleigh waves has made significant advances in recent years.If the receiver coverage is adequate，FWI offers a high resolution in laterally heterogeneous complex media.Source wavelet is one of the key factors affecting the accuracy of waveform inversion.However，the seismic wavelet for field seismic record is unknown，and an improper source wavelet is bound to have a negative effect on inversion results.The conventional solution is extracting and correcting a wavelet with synthesized data.However，wavelet correction is time-consuming because it requires repeated iterations.To solve this problem，this paper proposes a windowed adaptive wavelet extraction method to extract a source wavelet that is in line with the actual situation from the observation records，use it as the initial wavelet for wavelet correction，and thereby achieve the purpose of further optimizing inversion results.Inversion tests on a fault model constructed are conducted with five wavelets，including the actual wavelet，a wrong wavelet，a wrong wavelet combined with wavelet correction，an extracted wavelet，and an extracted wavelet combined with wavelet correction，respectively.The test results show that the wavelet obtained by the proposed method corresponds to the best inversion result，while the wrong wavelet corresponds to the worst one.Although a favorable inversion effect can also be achieved with the extracted wavelet alone，combining wavelet extraction with wavelet correction can save computation time and further improve the accuracy of Rayleigh waveform inversion.More-over，field data application further verifies this conclusion.

Keywords：Rayleigh wave，waveform inversion，source wavelet，wavelet extraction，wavelet correction

1.School of Geological Engineering and Geomati-cs，Chang’an University，Xi’an，Shaanxi 710054，China

2.School of Science，Chang’an University，Xi’an，Shaanxi 710064，China

Diffractionseparationmethodintheprestackcommonvirtualsourcegather.YANGChengzeng1，ZHANGXuantang1，SHENGTongjie2，LIXiaowei1，MAYunli1,andLIUTao3.OilGeophysicalProspecting,2022,57(4):847-854.

Diffractions carry high-resolution information on underground small-scale heterogeneous geolo-gical bodies，which are of great significance to the recognition of karst reservoirs.The diffractions with weak energy can be easily masked by the reflected wave field，which results in many problems such as unclear migration imaging as well as poor fine-grained recognition and difficult characterization of small-scale anomalies.The traditional diffraction separation method confronts serious energy loss of deep diffractions and difficult separation when diffractions intersect with or are tangential to the reflected wave field.To solve the above problems，we propose a prestack diffraction separation method.Specifically，by the common virtual source transformation，the common shot gather is converted into a common virtual source gather，on which the migration，reflected wave removal，and demigration are performed.Then，the diffraction field of the common shot gather is obtained by inverse transformation of the common virtual source.The application of the numerical model and field data reveals that the method can effectively suppress the reflected waves，keep the energy of diffractions，and improve the recognition accuracy of the small-scale geological bodies on the imaging section.

Keywords：diffraction separation，common virtual source gather，demigration，diffraction imaging

1.Exploration and Development Research Institute，SINOPEC North China Peptroleum Bureau，Zhengzhou，Henan 466100，China

2.School of Geosciences & Surveying Enginee-ring，China University of Mining & Technology-Beijing，Beijing 100083，China

3.SINOPEC Petroleum Exploration and Production Research Institute，Beijing 100083，China

High-precisionimagingtechnologyofdeepreflectionseismicinXiong’anNewAreaanditssurroundings.YUEHangyu1，2，3，4，WANGKai1，2，5，ZHANGJie1，2，LIUJianxun1，2，WANGXiaojiang1，2，andZHANGBaowei1，2，3.OilGeophysicalProspecting,2022,57(4):855-869.

As an effective method to find out the deep geo-logical structure and tectonic features，deep reflection seismic detection can provide detailed geologi-cal data for the overall planning and construction of Xiong’an New Area and 10000-meter “Transparent Xiong’an New Area” basic platform.Therefore，deep reflection seismic is applied to the geological survey of this area and its surroundings.Meanwhile，research on high-precision imaging technology of deep reflection seismic is carried out.Given the long length，wide span，imbalanced energy，weak signals at depth，and complex and variable noise of the deep reflection seismic profile in Xiong’an New Area and its periphery，a targeted high-precision imaging process of deep reflection seismic is designed and formulated.The key applicable technologies are also selected to gradually extract effective signals of deep reflection seismic，which are related to stratum interfaces and tectonic features from the complex seismic wave field.The SNR and resolution of the final results are high，with rich geological information，distinct structural framework，and clear strata relationship extracted from the deep reflection seismic profile.The results can provide data support for stratigraphic identification and comprehensive geology(resource)evaluation in this area and can better serve the basic geological construction of Xiong’an New Area.

Keywords：Xiong’an New Area，deep reflection seismic，high-precision imaging，weak signal extraction at depth

1.Institute of Geophysical and Geochemical Exploration，Chinese Academy of Geological Scien-ces，Langfang，Hebei 065000，China

2.National Center for Geological Exploration Technology，Langfang，Hebei 065000，China

3.Center for Geophysical Survey，China Geological Survey，Langfang，Hebei 065000，China

4.School of Geophysics and Information Techno-logy，China University of Geosciences(Beijing)，Beijing 100083，China

5.College of GeoExploration Science and Techno-logy，Jilin University，Changchun，Jilin 130026，China

Angle-domaincommon-imagegatherextractionbasedonelasticreversetimemigration.XUWeiya1，2，3，ZHUChenghong1，2，3，WEIZhefeng1，2，3，ZHANGXiaoyu4，andDUQizhen5.OilGeophysicalProspecting，2022,57(4):870-878.

Due to the coupling effect of P-and S-waves in the medium，multi-wave angle-domain common-image gathers(ADCIGs)encounter the problems such as energy crosstalk and low wavenumber noise in multicomponent seismic exploration.To solve the stress crosstalk problem in the construction of the Poynting vector of S-waves，this study uses the decoupled wave equation and pseudo-S-wave stress construction method to decouple the particle vibration velocity and stress，and the Poynting vectors of the P-and S-waves are constructed without energy crosstalk.Then，the PP-wave and PS-wave reflected angles are extracted given the Poynting vectors.On this basis，this paper proposes the PP-wave and PS-wave ADCIG extraction method based on the elastic reverse time migration(ERTM).Considering the low wavenumber noise in PP-wave ADCIGs，an angle attenuation superposition strategy for ADCIGs is proposed to improve the migration imaging accuracy.The model test verifies the correctness of the method.

Keywords：decoupled wave equation，pseudo-S-wave stress，Poynting vector，angle-domain common-image gather(ADCIG)

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

2.SINOPEC Key Laboratory of Seismic Elastic Wave Technology，Beijing 100083，China

3.SINOPEC Petroleum Exploration and Production Research Institute，Beijing 100083，China

4.Institute of Oceanographic Instrumentation，Qilu University of Technology(Shandong Academy of Sciences)，Qingdao，Shandong 266100，China

5.Key Laboratory of Deep Oil and Gas，China University of Petroleum(East China)，Qingdao，Shandong 266580，China

Petrophysicalmodelingforoffshorenaturalgashydrates.ZHANGJinqiang1，2，HANLei1，2，3，LIUJunzhou1，2，andLIUXiwu1，2.OilGeophysicalProspecting,2022,57(4):879-887.

As offshore natural gas hydrates occur in unconsolidated sediments，the petrophysical model of unconsolidated sediments should be built before the petrophysical modeling of natural gas hydrates.The traditional petrophysical model of unconsolidated sediments，however，faces the pro-blems of a high error in the estimation of shear wave velocity and a non-self-consistent modeling process.Taking account of the lubrication effect of brine on sediment particles，this study introduces the modified Hertz-Mindlin equation in the mode-ling process to improve accuracy.During the mo-deling of unconsolidated sediments，the modified Hertz-Mindlin equation is used to calculate the elastic moduli of sediments under critical porosity，and then Hashine-Shtrikman bounds are applied to calculate the elastic moduli of the skeleton of unconsolidated sediments.Furthermore，this paper discusses the influence of different occurrence forms of hydrates on the elastic properties of sediments，as well as the applicable models and discrimination criteria.On this basis，for natural hydrates existing as supporting grains and pore-filling fluids，the petrophysical modeling method and process that suits the two occurrence forms are designed.The modeling results are highly consistent with the measured curves of shallow drilling，which verifies the applicability of this method.

Keywords：natural gas hydrate，petrophysical mo-deling，unconsolidated sediments，Hertz-Mindlin model，equivalent medium

1.Sinopec Key Laboratory of Shale Oil/Gas Exploration and Production Technology，Beijing 102206，China

2.Sinopec Petroleum Exploration and Production Research Institute，Beijing 102206，China

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

Time-lapseseismicAVOinversionbasedonadouble-differenceinversionstrategy.FUXin1，ZHANGXiao-long2，3，SUNYunqiang2，DAIShiming4，andJIAJunsheng3.OilGeophysicalProspecting，2022,57(4):888-896.

The time-lapse pre-stack seismic data contains important amplitude variation with offset(AVO)information.The inversion of time-lapse seismic data can yield dynamic changes in elastic parameters before and after oil reservoir development，which is conducive to the exploration of remaining oil distributions，the optimization of injection-production schemes，and the enhanced oil recovery of oilfields.The inversion of time-lapse seismic data in different years，however，is inconsistent，which often leads to inversion artifacts and incorrect interpretation results.To solve the problem，this paper introduces the double-difference inversion strategy in time-lapse full waveform inversion(FWI)into time-lapse AVO inversion.Meanwhile，the improved approximate equation of the seismic P-wave reflection coefficient in the ray parameter domain is employed as the forward modeling equation for time-lapse seismic AVO inversion.The model tests indicate that the time-lapse seismic AVO inversion method based on the double-difference inversion strategy is not sensitive to the influence of the inversion quality of the baseline model as well as the inconsistency in inversion behaviors of the baseline model and the monitoring model.Moreover，it can effectively suppress artifacts outside the target area and concentrate the inversion results in the target area to accurately reflect the changes in the reservoir，which improves the inversion accuracy of the target area and reduces the errors of oil reservoir interpretations.With strong anti-noise performance，this method can also stably and accurately perform the inversion of the changes in the P-wave and S-wave impedance of the reservoir and effectively reflect the slight change in the P-wave and S-wave impedance caused by the pressure change of the overlying formation.

Keywords：double-difference inversion strategy，time-lapse seismic，AVO inversion，S-wave impedance，P-wave impedance

1.Department of Geoscience，University of Calgary，Calgary，Alberta T2N 1N4，Canada

2.College of Transportation and Civil Engineering，Fujian Agriculture and Forestry University，Fuzhou，Fujian 350002，China

3.Shenzhen Osdiot Technology Co.，Ltd.，Hefei，Anhui 230001，China

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

Strippingtechnologyofsub-volumewaveformPCAbasedonthree-dimensionalpolynomialsurfacefittingforstrongigneousrockshielding.YUEYouxi1,2，YANGJiefei3，JIANGLei4，CHENXueguo5，WUJiawei1,2，andCHENYidou1,2.OilGeophysicalProspecting,2022,57(4):897-906.

Igneous rocks have significant shielding and absorbing effects on seismic waves and often generate strong reflections，resulting in poor quality of seismic data on the underlying strata below igneous rocks.For the stripping of the strong reflections from igneous rocks，the corresponding time window width can be set according to the interpreted igneous rock horizon information，and the igneous rock horizon can be accurately obtained by searching longitudinal and lateral seismic waveform features to achieve local horizon flattening.Then，the strong reflections from igneous rocks can be stripped by the principal component analysis(PCA)of the sub-volume waveform.On this basis，three-dimensional polynomial surface fitting，instead of Wheeler transform，is performed to achieve local horizon flattening，and a stripping technology of sub-volume waveform PCA based on three-dimensional polynomial surface fitting is developed for strong igneous rock shielding.This method not only avoids the problems of local distortion of the reconstructed seismic signal caused by the failure of the reflection event of igneous rocks to be flattened when using Wheeler transform, but also saves the need to introduce amplitude threshold control into the stripping process.As for the rapid lateral change in the amplitude of the reflection from igneous rocks，the strong reflection from igneous rocks in the sub-volume window can be extracted and stripped laterally by sliding the sub-volume window as long as the amplitude of the reflection from igneous rocks within this window set does not change significantly，which improves the practicability of the method.The model test and application of practical data show that the proposed method allows for high lateral continuity of the strong igneous rock shielding stripped and that it can accurately eliminate the comprehensive response on the interface of the strong reflection from igneous rocks.In this way，it weakens the shielding effect of igneous rocks on the reflected waves from the underlying strata，promotes the ability to identify weak seismic reflections，and ultimately lays a foundation for energy compensation for weak signals and the subsequent processing of those signals.

Keywords：strong igneous rock shielding，stripping，three-dimensional polynomial surface fitting，principal component analysis，multi-wavelet decomposition，sub-volume waveform

1.China Universty of Petroleum(East China), Qingdao, Shandong 266580, China

2.Shandong Provincial Key Laboratory of Deep Oil & Gas，Qingdao，Shandong 266580，China

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

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

5.Research Institute of Exploration and Development，SINOPEC Shengli Oilfield Company，Dongying，Shangdong 257015，China

Researchandapplicationofcarbonatefracture-cavityboundarycharacterizationmethodbasedongradientstructure-tensor.ZHANGSheng1，LIYalin1，XIAOYoujun1，ZHENGDuoming1，YUANYuan1，andFENGLei1.OilGeophysicalProspecting，2022,57(4):907-915.

The seismic responses of carbonate fracture-cavity，channels，and faults are all discontinuous anomalies；however，the application of gradient structure-tensor(GST)to delineate carbonate fracture-cavity boundary is relatively rare.Therefore，the influence of each calculation link on the result is analyzed based on previous studies and according to the detailed calculation method of the GST attribute and its geometric significance.The results of technical principle analysis and actual seismic data tests are as follows.① The smooth calculation of tensor matrix elements，as the key to the whole technique，makes the structure-tensor matrix contain the features in the maximum gradient energy direction and the changing features in all orthogonal directions.The smooth of gradient elements can suppress background noise to some extent and improve the signal-to-noise ratio of the final calculation.② Both the second and the third eigenvalues have an ideal ability to describe the carbonate fracture cavities，reflecting the three-dimensional morphological characteristics of the fracture cavities from different perspectives.By the selection or combination of the second and third eigenvalue attributes，the carbonate fracture-cavity boundary can be characterized effectively in the suitable range of low-value interception and display.③ With drilling information as a scale to the dimensionless eigenvalues，the linear combination of eigenvalues is essentially a multi-constrained solution，which incorporates the detailed features of the second and third eigenvalues and effectively solves the problem of inconsistent values for different drilling scales.In conclusion，compared with conventional attributes such as amplitude change rate，the GST attribute can describe carbonate fracture-cavity more continuously，thus reflecting the corrosion law to a certain extent.

Keywords：carbonate，seismic attribute，gradient structure-tensor，eigenvalue，fracture-cavity boundary

1.Institute of Tarim Oilfield Company，PetroChina，Korla，Xinjiang 841000，China

SeismicsedimentologyfordevelopmentofDaqingChangyuanOilfield.LICao1，2，JIANGYan2，FANXiaodong2，LIUZongbao1，WANGYuanbo2，andQIJincheng2.OilGeophysicalProspecting,2022,57(4):916-925.

As the research and application of seismic sedimentology continue，a number of problems are encountered during the prediction of thinly interbedded continental sand-mud reservoirs in the development area with a dense well pattern in Daqing Changyuan Oilfield.Such problems include low efficiency of stratal slice optimization，difficult matching analysis of logging and seismic data information，and lack of quantified correlation of stratal slices with reservoir thickness.Therefore，methods such as automatic stratal slice optimization，plane reliability analysis of seismic attributes，and quantitative thickness prediction integrating logging and seismic data are studied after the analysis of wave impedance characteristics and amplitude-preserved seismic data processing based on geological targets.As a result，the basic framework of seismic sedimentology-based technology for Daqing Changyuan Oilfield is established，and reservoir prediction software integrating logging and seismic data is developed to improve the accuracy and efficiency of seismic sedimentology-based reservoir prediction for the development of the Changyuan Oilfield.Practical applications show that the development area with a dense well pattern is still subject to sand body changes that cannot be controlled with logging data and that narrow channels with a width less than half the well spacing are likely to develop between wells.The prediction and analysis of such geological phenomena with seismic sedimentology are of great significance for oilfield development.

Keywords：development area with a dense well pattern，amplitude-preserved seismic data processing，stratal slice optimization，seismic attribute reliability analysis，quantitative thickness prediction integrating logging and seismic data

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

2.Research Institute of Exploration and Development，Daqing Oilfield of CNPC，Daqing，Heilongjiang 163712，China

Physicalpropertyclassificationandevaluationme-thodbasedontheporestructureforconglomeratereservoirs.LIUShiqiong1，LIUXiangjun1，SUNYangsha2，LIUHongqi1，KONGYuhua3，andLIXiansheng4.OilGeophysicalProspecting,2022,57(4):926-936.

Since the discovery of the giant conglomerate oil reservoir in Mahu，Xinjiang，the exploration and development of conglomerate oil and gas reservoirs have attracted increasing attention.Nevertheless，such reservoirs are highly heterogeneous，resulting in difficult reservoir identification and evaluation.Taking the upper Wuerhe Formation in the HM work area as an example，this study classifies conglomerate reservoirs into five categories.After the Sphere-Cylinder model is applied to optimize the inversion of NMR(nuclear magnetic resonance)echo，an optimizedT2spectrum is obtained and then characterized by 12 spectral shape parameters.A spectral shape prediction Vector and a Joint prediction Vector are formed，respectively，by the 12 spectral shape parameters and the spectral eigenvalue.Then，they are trained by SVM(Support Vector Machine)to build a prediction model for reservoir physical property classification parameters.According to the comparison of the results of the prediction model with the reservoir physical property parameters and oil and gas production，the reservoir physical property classification parameters correlate well with the reservoir physical properties and are in good agreement with the test production of the reservoir.The comparison results show that the shape parameters of theT2spectrum obtained by optimized inversion with the Sphere-Cylinder model and the spectral eigenvalue can well characterize the physical properties of conglomerate oil and gas reservoirs.The results of this paper can help improve the quality and reliability of the classification and evaluation of conglomerate oil and gas reservoirs and provide logging technical support for the efficient exploitation of conglomerate oil and gas reservoirs.

Keywords：conglomerate reservoir，nuclear magne-tic resonance，logging evaluation，pore structure，reservoir physical property，echo inversion

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

2.Emergency Response Bureau，Zhongshan District，Liupanshui，Liupanshui，Guizhou 553000，China

3.Research Institute of Exploration and Development of Xinjiang Oilfield Company，Karamay，Xinjiang 834000，China

4.Geological Exploration & Development Research Institute，CNPC Chuanqing Drilling Engineering Company Limited(CCDC)，Chengdu，Sichuan 610051，China

DifferentialsegmentalactivitiesofQianbeifaultzoneinJianghanBasinanditsevolution.LIUAiwu1，TANGDaqing2，GUOLibin1，WENHui1，TANGWenxu1，andZOUShuai2.OilGeophysicalProspecting,2022,57(4):937-949.

The Qianbei fault zone can control sedimentary filling and differential enrichment of oil and gas，and it demonstrates huge structural and tectonic differences in different segments.The previous geo-logical understanding of the Qianbei fault zone and the analysis accuracy of the tectonic style can no longer meet the needs of the current fine oil and gas exploration.Therefore，to finely analyze the segmental differences of the Qianbei fault zone and their cause evolution，this paper divides the Qianbei fault zone into six segments from the west to the east in accordance with the fault types，geometrical characteristics，and cause evolution，namely，Xinnong Segment，Zhongshi Segment，Banghu Segment，Tankounan Segment，Tankoubei Segment，and Sanhechang Segment.Significant differences in tectonic characteristics are identified in certain fault segments of the Qianbei fault zone：Fault terraces and compound grabens are developed in Xinnong Segment，and “Jingsha Red Walls” and complex fault blocks are developed in Zhongshi Segment；the structure in Banghu Segment is represented by synthetic fault terraces，which are relatively simple；armchair-shaped main faults and complex fault blocks are found in Tankounan Segment，and deck chair-shaped main faults and complex fault blocks are developed in Tankoubei Segment；Sanhechang Segment features a combination of gently-dipping slopes，fault terraces，and graben-horst structure.The differential segmental activities of the Qianbei fault zone are controlled by multi-phase and multi-directional regional stress with different properties，and “Jingsha Red Walls”，salt layers，deep magmation，and pre-existing structures also exert a certain influence on those activities.The fault activities have experienced four stages of evolution，i.e.，the rift phase(the deposition period of Shashi Formation-lower 4th member of Qianjiang Formation)，the fault-depression phase(the deposition period of the upper 4th member of Qianjiang Formation-Jinghezhen Formation)，the structural inversion phase(the late deposition period of Jinghezhen Formation and pre-deposition period of Guanghuashi Formation)，and the depression and reform-finalization phase(the deposition period of Guanghuashi Formation-Pingyuan Formation).The results can provide a theoretical reference for systematically understan-ding the differential segmental activity mechanism of the Qianbei fault zone and the practice of oil and gas exploration.

Keywords：faults，differential activity，tectonic evolution，Qianbei fault zone，Jianghan Basin

1.Exploration and Development Research Institute of Jianghan Oilfield Company，SINOPEC，Wuhan，Hubei 430223，China

2.Faculty of Earth Resources，China University of Geosciences，Wuhan，Hubei 430074，China

Characteristicsofsimulatedmarinecontrolled-sourceelectromagneticresponsesofcoupledgashydrateandpetroleumreservoirmodels.GAOYan1，2,MAChao1，2，3,andZHANGXiangyu1.OilGeophysicalProspecting,2022,57(4):950-962.

Marine controlled-source electromagnetic method(MCSEM)can be employed to identify resistivity anomalies on the seafloor and is thus widely used in offshore petroleum exploration.A simulation method of frequency-domain vector finite-element electromagnetic forward modeling based on preconditioned iterative solution is proposed on the basis of a complex seafloor strata model with a complex electrical structure.The preconditioning method based on incomplete lower-upper(ILU)decomposition improves the efficiency of iterative solution.The simulation results of the layered model verify the effectiveness of the method.A geo-electric model of a certain area of the South China Sea is built with arbitrary hexahedral structured ele-ments.A deep-buried petroleum reservoir model，a gas hydrate deposit model，and a geological model with a complex basement are built，respectively，according to the reservoir characteristics of deposits with different structures.The characteristics of MCSEM response are analyzed under different excitation-receiving modes.Specifically，when the offset is small，the influence of topography is ob-vious.The influence is especially salient on shallow-buried gas hydrate deposits when the offset is small，while it is more distinct on deep-buried petroleum reservoirs when the offset is large.

Keywords：marine controlled-source electromagnetic(MCSEM)method，vector finite-element，preconditioned iterative，seafloor topography

1.Guangzhou Marine Geological Survey，Guangzhou，Guangdong 511458，China

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

3.National Engineering Research Center of Gas Hydrate Exploration and Development，Guangzhou，Guangdong 511458，China

Frequency-domainelectromagneticmethodwithgroundmovingsourceanditsdetectioneffect.WANWei1，WANGZhigang2，andLUYao2.OilGeophysicalProspecting,2022,57(4):963-972.

As the traditional ground frequency-domain controlled-source electromagnetic(CSEM)method uses a single transmitter，large source-receiver distances need to be adopted for observation to obtain large detection depths，which is highly likely to cause a low signal-to-noise ratio of the observed data.This paper attempts to use ground moving sources and achieve the purpose of improving the signal-to-noise ratio of the data by narrowing the source-receiver distance and employing transmitters with different offsets for observation.More-over，the effective detection depth is enhanced by comprehensively using geometric sounding and frequency sounding.The typical one-dimensional la-yered model is utilized to comparatively analyze the detection effects of the frequency-domain electromagnetic method with ground moving sources，i.e.，the ground frequency-domain MSEM method，and the traditional ground CSEM method.The results show that the MSEM method has a favorable detection effect on common geoelectric models and that it can make up for the disadvantage of the traditional ground CSEM method that it is not sensitive to underground high-resistance layers.Experiments on the proposed method are conducted in a mining area in northeast Jiangxi Province，and the data processing results verify the applicability of the ground MSEM method to solve practical problems.

Keywords：ground moving source，electromagentic field，frequency sounding，geometric sounding

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

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

CSEMpseudo-randomsignalprocessingmethodbasedonfeatureextractionandclusteringidentification.ZHANGXian1，2，LIDiquan1，2，LIJin3，andHUYanfang1，2.OilGeophysicalProspecting,2022,57(4):973-981，1008.

The controlled-source electromagnetic(CSEM)data is susceptible to noise，which leads to unsatisfactory exploration effects.Human factors will exert a huge influence on traditional CSEM data processing which usually employs frequency point screening，abnormal elimination and other me-thods，and the filtering method cannot retain pseudo-random effective signals.According to the recorded CSEM data in the time domain，we analyze the time-domain statistical characteristics of useful signals and noises in the CSEM data，and quantitatively identify and qualitatively analyze useful CSEM signals to address the above problems.As a result，a CSEM pseudo-random signal processing method based on feature extraction and clustering identification is proposed in this paper.Firstly，the sample library including two kinds of typical noises and pseudo-random signals is established，and features of the time and frequency domains of the sample library signals are analyzed.Then the time-domain statistical features are extracted，and the fuzzy C-means clustering algorithm is adopted to identify and eliminate the noise for retaining useful signals and reconstructing original CSEM data.Finally，the frequency spectrum of effective frequency points is extracted by digital coherence technology.Through the processing of simulated data and measured data，results show that the proposed method can identify and eliminate typical noises accurately and effectively，thereby significantly improving the quality of CSEM data.After being processed by the proposed method，the componentExnormalization electric field curve and wide field electromagnetic(WFEM)resistivity curve are smoother and more continuous，thus effectively increasing the signal-to-noise ratio of CSEM signals.

Keywords：controlled-source electromagnetic me-thod，pseudo-random signal，feature extraction，clustering identification

1.Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Monitoring(Central South University)，Ministry of Education，Changsha，Hunan 410083，China

2.School of Geosciences and Info-Physics，Central South University，Changsha，Hunan 410083，China

3.College of Information Science and Engineering，Hunan Normal University，Changsha，Hunan 410081，China

CharacteristicsofdeepgeologicalstructureofLang-shanmetallogenicbeltinInnerMongolia.ZHANGZhenyu1，2，HUXiangyun3，WANGGang1，2，andLIYongbo1，2.OilGeophysicalProspecting,2022,57(4):982-991.

The special tectonic position of the Langshan metallogenic belt leads to its completely exposed strata，frequent magma activities，and complex structure.Affected by natural conditions and other factors，there is little deep geophysical exploration in Langshan metallogenic belt.The features of the deep geological structure are less studied generally，and the research on regional deep structural characteristics is not deep.To study the deep geological structure of the Langshan metallogenic belt，this paper utilizes the results of two comprehensive geophysical survey profiles of the magnetotelluric sounding and the gravity and magnetic detection perpendicular to the main geological structure strike.In combination with the geological data，the deep geological structure characteristics of the Langshan region are deeply studied and the following knowledge is obtained.First，nine fractures(F1-F9)are identified.Specifically，F1 is the intersection of the Urad Houqi-Huade-Chifeng deep fault zone and Wendur Temple-Xilamulun River deep fault zone，and it is also the suture zone of the North China plate and Siberia plate.Second，many high-resistance and high-magnetic bodies are found，which are magmatic rocks or the comprehensive reflection of magmatic rocks with Paleozoic，Proterozoic，and Archean strata.Third，it features low resistance between F7 and F4，between F2 and the southeast end of section Ⅰ，and between F1 and F3，which is the reflection of Mesozoic and Cenozoic fault basins.Fourth，Baoyintu uplift is located between deep faults F4 and F5 with high-density bulges at both ends，and the inversion resistivity is characterized by high resistance and low resistance on both sides.Influenced by F4 and F5，Baoyintu uplift has developed internal secondary faults，with multiple magmatic intrusions.Fifth，there are obvious wavy density anomalies on both sides of F1，which are caused by the folding structure formed on both sides of the suture zone between the North China plate and the Siberia plate.Sixth，huge low-resistivity anomalies develop at the depth of more than 10km in F1 and 20 km in F1 and F9.The upper part of these anomalies is upright，and the lower part is inclined to the north and can extend downward below the Moho surface.It is inferred that the low-resistivity anomalies are magma soft fluids(or magma channels)upwelling from the mantle，which provides deep geological conditions for magma overflow and deep mineralization.

Keywords：Langshan metallogenic belt，magmatic rock，fault structure，deep geological structure，magnetotelluric sounding，gravity and magnetic detection

1.National Modern Geological Exploration Engineering Technology Research Center，Langfang，Hebei 065000，China

2.Institute of Geophysical and Geochemical Exploration，Chinese Academy of Geological Sciences，Langfang，Hebei 065000，China

3.China University of Geosciences(Wuhan)，Wuhan，Hubei 430074，China

Researchprogressofabsorptionandattenuationmechanismandpetrophysicaltheoryforgashydratereservoir.WUCunzhi1，ZHANGFeng1，andLIXiangyang1.OilGeophysicalProspecting,2022,57(4):992-1008.

The bottom simulating reflection(BSR)characteristics of reflected seismic waves are an important sign of gas hydrate.Although they can indicate the bottom of hydrate，they can hardly be used for quantitative interpretation of hydrate content.The rapid development of the gas hydrate exploration technology in recent years results in an understanding that the “blank”zone of seismic amplitude above BSR，directly related to the absorption and attenuation of seismic waves，can be used as an indicator of gas hydrate distribution and quantification.This paper reviews the seismic wave absorption and attenuation characteristics of various hydrate exploration areas around the world(the Mallik permafrost area in Canada，the Nankai Trough in Japan，the Makran accretionary wedge in the Arabian Sea，the Gulf of Mexico，and the Shenhu area in the South China Sea)and artificial hydrate-bearing rock samples.The results show that for different hydrate exploration areas，hydrate-bearing samples，and data used，seismic waves show diffe-rent attenuation characteristics.Then，the possible attenuation mechanisms and related petrophysical theories are summarized for hydrate reservoirs，mainly including global flow attenuation(the Leclaire model)，squirt flow(the improved Leclaire model，the hydrate effective grain(HEG)model for submicron hydrate particle squirt，or the hydrate-bearing effective sediment(HBES)model for micron flow squirt)，skeleton friction attenuation(the improved Leclaire model).At present，the main problem is that although the hydrate-bearing strata in many areas demonstrate obvious absorption and attenuation characteristics，the relationships of absorption and attenuation variation with hydrate saturation remain unknown due to the varied hydrate formation conditions and geological environments and different occurrence states of hydrate in sediments of different areas.In addition，the frequency band ranges of the current measured observation data and those petrophysical experiments test are limited，so the characteristics of attenuation variation with frequency are not fully reflected.Therefore，petrophysical experimental studies need to be further conducted，and available data from actual exploration areas and the making and experimental measurement results of artificial cores shall be well utilized，thereby studying the additional effect of the microstructure of the hydrate reservoir on the attenuation mechanism in depth.After the reasons of seismic wave attenuation in hydrate reservoirs are clarified，a quantitative seismic interpretation method for hydrate saturation can be developed.

Keywords：gas hydrate，seismic attenuation，petrophysics，microstructure

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