Some Processing Basics for Seismic Interpretation地震解释的一些处理基础.ppt

Some Processing Basics for Seismic Interpretation地震解释的一些处理基础,Basic definitions 基本定义Some processing basics 一些处理基础Seismic processing artifacts 地震处理假象Summary 结语,Agenda,I.Basic definitions 基本定义,Goals&objectives of seismic processing 地震处理的战略目标与战术目标Spectral:Amplitude&phase issues 谱域：振幅和相位问题Temporal:Analytical signal 时域：解析信号Resolution&fidelity 分辨率与保真度,I.Basic definitions,Goals&objectives of seismic processing 地震处理的战略目标,To meet scientific&business needs with quality seismic imageries&attributes in a timely manner.为满足科学与商业需求及时提供高质量地震图片和属性成果。To transfer raw seismic data into credible&interpretable forms.将原始地震资料转换为可信、可解释之形式。,I.Basic definitions,Signal recognition 信号辩别Noise suppression 噪音压制Velocity model building 波速建模Interface imaging界面成像(Seismic imaging地震偏移)Interpretation-guided processing 以解释为指导的处理,I.Basic definitions,Goals&objectives of seismic processing 地震处理的战术目标,Spectral:Amplitude&phase issues 谱域：振幅和相位问题,Amplitude&phase spectra of a data trace,I.Basic definitions,I.Basic definitions,Amplitude&phase spectra of a data trace,Decompose it into many monochromatic waves,(Yilmaz,1987),Spectral:Amplitude&phase issues 谱域：振幅和相位问题,The phase angle f of a monochromatic wave depends on where is time zero.,I.Basic definitions,The phase angle f of a monochromatic wave depends on where is time zero.,I.Basic definitions,The phase angle f of a monochromatic wave depends on where is time zero.,I.Basic definitions,a.A data trace(solid)compared with a 20 Hz monochromatic wave(dashed)with 175 of phase.b.Correlation between data trace and the monochromatic wave as a function of the phase angle.These are the Fourier amplitude and phase for this trace at 20 Hz.,(Barnes,1998),I.Basic definitions,I.Basic definitions,(Yilmaz,1987),Phase spectrum,I.Basic definitions,(Yilmaz,1987),I.Basic definitions,(Yilmaz,1987),Phase spectrum,Amplitude&phase issues 振幅和相位问题Amplitude&phase spectraZero phase,minimum phase&maximum phase,I.Basic definitions,Amplitude&phase issues 振幅和相位问题Amplitude&phase spectraZero phase,minimum phase&maximum phase,The phase of a monochromatic wave depends on where is time zero.,I.Basic definitions,Amplitude&phase issues 振幅和相位问题Amplitude&phase spectraZero phase,minimum phase&maximum phase,Zero phaseminimum phasemaximum phase The phase of a wavelet depends on its shape.,I.Basic definitions,Amplitude&phase issues 振幅和相位问题Amplitude&phase spectraZero phase,minimum phase&maximum phaseFactors affecting amplitude&phase,Amplitude in stratigraphic exploration(e.g.,AVO)Amplitude attenuation in nature Amplitude alteration in data processing,I.Basic definitions,Factors affecting amplitude&phase,Amplitude in stratigraphic exploration(e.g.,AVO),I.Basic definitions,(Hilterman,1990),Factors affecting amplitude&phase,Amplitude in stratigraphic exploration(e.g.,AVO)Amplitude attenuation in nature,I.Basic definitions,(Quan and Harris,1997),Intrinsic attenuation,Factors affecting amplitude&phase,Amplitude in stratigraphic exploration(e.g.,AVO)Amplitude attenuation in nature,I.Basic definitions,Extrinsic attenuation,(Stolk&Symes,2004),Factors affecting amplitude&phase,Amplitude in stratigraphic exploration(e.g.,AVO)Amplitude attenuation in nature Amplitude alteration in data processing,I.Basic definitions,Sections without&with amplitude correction(Deal et al.,2002),I.Basic definitions,Temporal:Analytical signal 时域：解析信号,Most seismic attributes are given in time domain and require the use of analytical signal,input signal(top)and three edge-detection methods(Luo et al.,2003).,I.Basic definitions,Temporal:Analytical signal 时域：解析信号,Most seismic attributes are given in time domain and require the use of analytical signal,Amplitude time slices from Luo et al.(2003).a)Input data;b)An interpreted channel on the input;c)Hilbert Transform(HT)time slice;d)L2-order Generalized Hilbert Transform(GHT)time slice.,I.Basic definitions,Temporal:Analytical signal 时域：解析信号,Temporal:Analytical signal 时域：解析信号,Complex trace,or analytical signal of x(t):in timex(t)=x(t)-i H x(t)in frequencyX(w)=X(w)1+sgn(w)Hilbert transform H x(t)=x(t)*(-1/pt)Instantaneous propertiesEnvelop of x(t):e(t)=|x(t)|Instantaneous phase of x(t):f(t)=atan Im x(t)/x(t),I.Basic definitions,Examples of Hilbert transform 5,Instantaneous phase.,I.Basic definitions,Examples of Hilbert transform 5,Instantaneous phase.,I.Basic definitions,Examples of Hilbert transform 5,Instantaneous phase.,I.Basic definitions,Resolution&fidelity 分辨率与保真度,Horizontal resolutionVertical resolutionFidelityPreservation of resolutionPreservation of fidelity,I.Basic definitions,Resolution&fidelity 分辨率与保真度Horizontal resolution,The horizontal resolution of unmigrated seismic data is given by the Fresnel zone.(First)Fresnel zone:The part of a reflecting interface which returns energy within cycle of the dominant waves.Due to constructive and destructive interference,events within a Fresnel zone cannot be resolved.,I.Basic definitions,(Pratt,2005),Resolution&fidelity 分辨率与保真度Horizontal resolution,I.Basic definitions,(Pratt,2005),Resolution&fidelity 分辨率与保真度Horizontal resolution,I.Basic definitions,(Pratt,2005),Resolution&fidelity 分辨率与保真度Horizontal resolution,I.Basic definitions,I.Basic definitions,In principle,horizontal resolution of migrated seismic data equals to the spatial Nyquist wavenumber,or twice the CMP trace spacing(typically 10 m).In practice,this horizontal resolution is never achieved,and a resolution of about 3-4 traces(typically 30-40 m)is more realistic.,Resolution&fidelity 分辨率与保真度Horizontal resolution,Resolution&fidelity 分辨率与保真度Horizontal resolutionVertical resolution(Sheriff,1991),I.Basic definitions,The smallest change in input that will produce a detectable change in output.The ability to separate two features that are close together.(Widess)The ability to localize an event seen through a window,usually taken as the half width of the major lobe.(Ricker),1.The Rayleigh limit,or Rayleigh criterion-bed thickness(h)is 1/4 of the seismic wavelength(or,two-way time thickness is 1/2 of the dominant seismic period).This is the tuning limit-i.e.,maximum constructive interference between the top and bottom of the bed takes place for this value of bed thickness.,I.Basic definitions,Practical estimation of vertical resolution,I.Basic definitions,2.Widess limit-bed thickness is 1/8 of the seismic wavelength.This is the resolution limit.For example,if the velocity is 4 km/s and the dominant frequency is 50 Hz,then the seismic wavelength is 80 m.Hence the resolution limit is 10 m-this is the thinnest resolvable bed thickness for these parameters.,Practical estimation of vertical resolution,I.Basic definitions,3.Rickers limit-Limit of resolution occurs when the time separation of events is equal to the time separation between wavelet main lobe inflection points,TR.,Practical estimation of vertical resolution,(Ricker,1953),I.Basic definitions,3.Rickers limit-Limit of resolution occurs when the time separation of events is equal to the time separation between wavelet main lobe inflection points,TR.,Practical estimation of vertical resolution,(Ricker,1953),The Rayleigh limit-bed thickness is 1/4 of the seismic wavelength.Widess limit-bed thickness is 1/8 of the seismic wavelength.Rickers limit-time separation of events is equal to the time separation between wavelet main lobe inflection points,TR.,I.Basic definitions,Practical estimation of vertical resolution,I.Basic definitions,reflects the truthfulness,or how accurate an,Checkerboard resolution test of a crosswell tomography study(Zhou et al.,1993),I.Basic definitions,reflects the truthfulness,or how accurate an,PSDM results using(a)isotropic&(b)anisotropic tomography.(H.B.Zhou et al.,2004),Resolution&fidelity 分辨率与保真度Horizontal resolutionVertical resolution(Sheriff,1991)FidelityPreservation of resolution 维护分辨率,I.Basic definitions,Bandwidth,I.Basic definitions,A:(a)frequency bandwidth unchanged;(b)lose of high frequencies;(c)lose of low frequencies.,Q:Here are three cases of amplitude spectra of input(green solid curve)and output(blue dashed curve)through processing.Describe the frequency change for each case.,小测验：,I.Basic definitions,A:Resolution is proportional to bandwidth.One octave of bandwidth covers an interval when frequency is doubled:2(#of octave)=fmax/fminHaving low frequency is the key for higher resolution!,Q:Which output has lowest resolution?,小测验：,I.Basic definitions,Having low frequency is the key for higher resolution!,Resolution&fidelity 分辨率与保真度Horizontal resolutionVertical resolution(Sheriff,1991)FidelityPreservation of resolution 维护分辨率,I.Basic definitions,Bandwidth Phase,(Schoenberger,1974),Bandwidth Phase,Resolution&fidelityHorizontal resolutionVertical resolution(Sheriff,1991)FidelityPreservation of resolution,I.Basic definitions,(Schoenberger,1974),Zero-phase has the best resolution for interpretation!,Resolution&fidelity 分辨率与保真度Horizontal resolutionVertical resolution(Sheriff,1991)FidelityPreservation of resolution 维护分辨率,I.Basic definitions,Bandwidth Phase Recognize processing artifacts,Resolution&fidelity 分辨率与保真度Horizontal resolutionVertical resolution(Sheriff,1991)FidelityPreservation of resolutionPreservation of fidelity 维护保真度,I.Basic definitions,SignalVelocity model buildingInterface imaging(seismic migration),I.Basic definitions,Some processing basics 一些处理基础Interpretation objectives 地震解释目标,Producing structural imageriesDelivering lithologic&fluid indicatorsIncreasing resolutionImproving seismic-well tieProviding seismic attributes,II.Seismic processing basics,Seismic processing basics 地震处理基础Processing objectives 地震处理目标,Signal recognition&enhancementNoise suppressionInterface imaging(seismic migration)Velocity model buildingInterpretation-guided processing(constraints),II.Seismic processing basics,Seismic processing basics 地震处理基础Processing objectives 地震处理目标,II.Seismic processing basics,II.Seismic processing basics,Signal recognition&enhancement,Model space?,Data space,II.Seismic processing basics,Signal recognition&enhancement,小测验：猜一猜图示共炮集所对应的是什么模型？,Model space?,Data space,Model space,Data space,II.Seismic processing basics,Signal recognition&enhancement,II.Seismic processing basics,Signal recognition&enhancement,Challenge:Most seismic migration methods assume that all data events are primary reflections!,Data space,II.Seismic processing basics,Deconvolution:stationary,Signal recognition&enhancement,(Margrave,1998),II.Seismic processing basics,Deconvolution:non-stationary,Signal recognition&enhancement,(Margrave,1998),II.Seismic processing basics,Noise suppression,Common shot gather of x(red)and y(blue)components of a VSP data before and after polarity correction.,Original,Filtered,II.Seismic processing basics,Noise suppression,Time slices of footprint suppression using 2D wavelet transform(Cvetkovic et al.,2007),II.Seismic processing basics,Noise suppression,Original,Filtered,Difference,Vertical slices of footprint suppression using 2D wavelet transform(Cvetkovic et al.,2007),Seismic EarthSeismically detectable Earth properties,Real EarthGeologic strata of rocks&fluids,Two complementing imaging tools:,II.Seismic processing basics,Interface imaging&Velocity model building,Seismic EarthSeismically detectable Earth properties,Real EarthGeologic strata of rocks&fluids,Two complementing imaging tools:,Internal propertiesVp,Vs,r,Q,.,Boundary propertiesdVp,dVs,dr,dQ,.,II.Seismic processing basics,Interface imaging&Velocity model building,Two complementing imaging tools:,II.Seismic processing basics,Interface imaging&Velocity model building,(Youn&Zhou,2001),Interface imaging(seismic migration),II.Seismic processing basics,Marmousi model,Migration assumptions:All seismic migration methods require velocity model.Most seismic migration methods assume that all data events are primary reflections!All 2D seismic migration methods assume all reflections are 2D!,CMP stack works well with low-dip reflectors&small lateral velocity variations does not work in case of lateral velocity variation,(Youn&Zhou,2001),II.Seismic processing basics,Interface imaging(seismic migration),(Youn&Zhou,2001),CMP stack works well with low-dip reflectors&small lateral velocity variations does not work in case of lateral velocity variationTime migration does a better job than CMP stack,using an average stacking velocity functiondoes not work in areas of high complexity,such as sub-thrust,sub-salt&sub-basalt,II.Seismic processing basics,Interface imaging(seismic migration),(Youn&Zhou,2001),CMP stack works well with low-dip reflectors&small lateral velocity variations does not work in case of lateral velocity variationTime migration does a better job than CMP stack,using an average stacking velocity functiondoes not work in areas of high complexity,such as sub-thrust,sub-salt&sub-basaltPrestack depth migration best way to image complex structure in the presence of lateral velocity variations two-way PDM can use primaries,multiples,converted waves,&refraction data,II.Seismic processing basics,Interface imaging(seismic migration),Time and depth migrations differ in their ability to handle lateral velocity variations,rather than the definition of the vertical axis(time or depth).Depth migration treats lateral velocity variation vigorously.Time migration ignores lateral velocity variation.The above difference dictates the dependency of these migrations on the velocity model.,小测验：时间偏移与空间偏移方法有何不同?,II.Seismic processing basics,Interface imaging(seismic migration),Time migration versus depth migration,II.Seismic processing basics,Interface imaging(seismic migration),(Biondi,2004),II.Seismic processing basics,Interface imaging(seismic migration),Time migration versus depth migration,(Biondi,2004),(after Liner,1999),II.Seismic processing basics,Interface imaging(seismic migration),II.Seismic processing basics,Velocity model buildingCMP Semblance,CMP gather:(a)original CMP gather,(b)CMP gather after NMO,(c)velocity spectrum.(Biondi,2004),II.Seismic processing basics,Velocity model buildingCMP Semblance,Interval velocity of in-line section(CMP Y=8km).,(Biondi,2004),II.Seismic processing basics,Velocity model buildingCMP Semblance,RMS velocity of in-line section(CMP Y=8km).,(Biondi,2004),II.Seismic processing basics,Velocity model buildingCMP Semblance,(Biondi,2004),II.Seismic processing basics,Velocity model buildingCMP Semblance,(Biondi,2004),II.Seismic processing basics,Velocity model buildingCMP Semblance,II.Seismic processing basics,Velocity model buildingCMP SemblanceMigration velocity analysis,II.Seismic processing basics,Velocity model buildingCMP SemblanceMigration velocity analysis,CIG after migration with 8%slower velocity.(Biondi,2004),II.Seismic processing basics,Velocity model buildingCMP SemblanceMigration velocity analysis,CIG after migration with the correct velocity.(Biondi,2004),II.Seismic processing basics,Velocity model buildingCMP SemblanceMigration velocity analysis,CIG after migration with 8%faster velocity.(Biondi,2004),II.Seismic processing basics,Velocity model buildingCMP SemblanceMigr