CN102156296A - Elastic reverse time migration imaging method by combining seismic multi-component - Google Patents

Abstract

The invention discloses an elastic reverse time migration imaging method by combining seismic multi-component and belongs to the field of exploration geophysics. The method comprises the following steps: directly inputting seismic multi-component data; without performing the wave field separation on the input data, extrapolating forwards at inverse time based on an elastic wave equation; building an underground elastic vector seismic wave field by combining with multi-component; acquiring four elastic wave imaging results having specific physical significance, such as longitudinal wave, transverse wave, transition longitudinal wave and transition transverse wave, by applying an elastic wave cross-correlation imaging condition of lighting compensation; and suppressing the low wave number noise in the elastic reverse time migration by performing low-pass angle filtering, thereby acquiring an imaging result of final elastic wave migration. By directly inputting the seismic multi-component data and building the underground elastic vector seismic wave field by combining with multi-component, the method can be used for accurately imaging for complex earth medium.

Description

The many component associatings of earthquake elasticity reverse-time migration formation method

One, technical field:

The present invention relates to a kind of exploration geophysics field, the many component associatings of particularly a kind of earthquake elasticity reverse-time migration formation method.

Two, background technology:

Along with the deep development of seismic prospecting, the target of oil-gas exploration turns to complicated earth surface, complex structure and complicated reservoirs direction gradually, and the precision of seismic imaging is had higher requirement.At present in oil, seismic gas exploration, still based on the longitudinal wave exploration method.Yet the wave field of propagating in ball medium not only is confined to compressional wave, also has wave modes such as shear wave and transformed wave, that is to say that seismic wave field is the elasticity wave field, so could portray ball medium more accurately based on the migration and imaging techniques of equations for elastic waves.Just be based on this, the elasticity seismic migration is regarded as significant improvement to conventional earthquake imaging technique by the geophysics worker.How underground elasticity wave field is handled the multiwave multicomponent earthquake data by ground multi-component seismic wave detector record so the elasticity seismic migration must relate to.But conventional multi-component seismic data treatment technology has more directly been copied the ripe compressional wave treatment technology flow process of industry member mechanically, multi-component earthquake data is carried out the processing that is similar to compressional wave of " separate type ", " scalar formula ".The multi-component seismic data processing mode of this " separate type ", " scalar formula " can not make full use of the elastic wave field information, but on the inspection surface of the face of land, multi-component earthquake data is carried out wave field separation earlier, only utilize a subclass (compressional wave or shear wave) of elastic wave field information, handle and another one subclass (shear wave or compressional wave) is used as noise.The problem that this processing mode is brought is as follows: the first, and ground wave field separation is faulty.Its reason be above-mentioned way according to the face of land be the hypothesis of weathering zone, think that ray all is the vertical outgoing face of land, vertical component as being the P ripple, horizontal component as being the S ripple.But in fact ray can not be fully perpendicular to the face of land, so be that vertical component or horizontal component all comprise P ripple and S ripple simultaneously.Therefore this simple processing will inevitably bring noise and illusion.If structure distant shock source wave field P ripple and S wave component need be known P ripple and S wave polarization information (being the polarization direction) in advance.Therefore more accurate carry out wave field separation and should adopt polarization filtering, can estimate P ripple and S wave polarization information, on P ripple and S wave polarization direction, carry out projection respectively then, isolate P ripple and S ripple on the face of land.But it is not complete in the ground wave field information, the wave field of disappearance depth direction is asked for the polarization direction and must be adopted certain hypothesis to be similar to ask for, can not accomplish to ask for accurately the polarization direction, therefore general separation algorithm based on the face of land all is similar to, and is not perfect.Separation method should carry out wave field extrapolation based on wave equation accurately, makes up to separate required underground whole wave fields, and then utilizes the result after separating the wave equation continuation to go back to ground.The second, ignored the elasticity vector characteristic of many component datas.Utilize scalar P ripple and S wave datum scalar equation to make up underground wave field and carry out pre-stack depth migration, no matter whether wave field separation is perfect, all abandoned the vector characteristic of underground wave field inevitably.Because actual underground medium is flexible, seismic wave field is a vector wave, and what still use in processing is scalar wave equation, and the wave field of Gou Jianing can not keep the characteristics of vector wave and elastic wave thus.The characteristics of vector wave are with three component statement seismic events, the direction of vibration of seismic wave definitely.The characteristics of elastic wave are that seismic event is flexible, allow the mutual conversion of wave mode and the distribution of energy.Use the scalar wave treatment technology, must lose above characteristics.Therefore can't solve energy conversion between the different subclass signals (compressional wave and shear wave) based on the disposal route of this routine, more can not guarantee the accuracy and the physical significance of imaging amplitude.For this reason, must set up a cover novel be " the many components of earthquake are united the elastic wave imaging " technology of characteristics with " elastic wave, vector wave ".

Three, summary of the invention:

Purpose of the present invention is exactly the above-mentioned defective that exists at prior art, the many component associatings of a kind of earthquake elasticity reverse-time migration formation method is provided, by direct input multi-component seismic data, unite and use three components to make up underground three dimensions earthquake elasticity vector wave field, be used for the accurate imaging of complicated ball medium.

Its technical scheme is:

The invention has the beneficial effects as follows: at the input phase of data, directly utilize multi-component seismic data to make up underground full elasticity wave field, avoided the seismologic record of input is carried out wave field separation, under the situation of not introducing noise, fully, nondestructively utilized multi-component seismic data to the full extent; In the structure stage of wave field, based on equations for elastic waves forward and the inverse time extrapolation of setting right, keep the elasticity and the vector characteristic of seismic wave field, correctly solve the energy conversion problem of seismic event in underground elastic medium communication process and the maintenance problem of elasticity seismic amplitude; In the imaging stage, the elastic wave simple crosscorrelation image-forming condition of using illumination compensation obtains four kinds of elastic waves (compressional wave, shear wave, conversion compressional wave and converted shear wave, brief note is PP, SS, SP and PS ripple) the imaging results of explicit physical meaning, and utilize lower wave number illusion in the low pass angle domain filtering compacting reverse-time migration, obtain final migration imaging result.

Four, description of drawings:

Fig. 1 (a) is a Marmousi2 model PP ripple imaging results;

Fig. 1 (b) is a Marmousi2 model PS ripple imaging results;

Fig. 1 (c) is a Marmousi2 model SP ripple imaging results;

Fig. 1 (d) is a Marmousi2 model SS ripple imaging results;

Fig. 2 (a) is the final imaging results of Marmousi2 model PP ripple;

Fig. 2 (b) is the final imaging results of Marmousi2 model PS ripple;

Fig. 2 (c) is the final imaging results of Marmousi2 model SP ripple;

Fig. 2 (d) is the final imaging results of Marmousi2 model SS ripple;

Fig. 3 is the process flow diagram of multi-component seismic data associating elasticity reverse-time migration formation method.

Five, embodiment:

Embodiment 1: the present invention realizes that the technical scheme of being taked is directly to utilize multi-component seismic data as input, do not carrying out under the situation of wave field separation to input data, carry out forward and inverse time extrapolation based on equations for elastic waves, unite the underground elasticity vector wave field of the common structure of many components, the elastic wave simple crosscorrelation image-forming condition of using illumination compensation obtains four kinds of elastic wave (compressional waves of explicit physical meaning, shear wave, conversion compressional wave and converted shear wave, brief note is PP, SS, SP and PS ripple) imaging results, and carry out lower wave number noise in the reverse-time migration of low pass angle domain filtering compacting elasticity, obtain final elastic wave migration imaging result.Concrete steps are as follows:

1, the structure of elastic wave propagation operator and underground vector wave field

Single order elastic wave velocity---stress equation:

$\frac{\∂V(x,z,t)}{\∂t}=(A^x\frac{\∂}{\∂x}+A^z\frac{\∂}{\∂z})V(x,z,t)---\left(1\right)$

Wherein,

$V=\left[\begin{array}{c}{v}_x\\ v_z\\ {\mathrm{\τ}}_{\mathrm{xx}}\\ {\mathrm{\τ}}_{\mathrm{zz}}\\ {\mathrm{\τ}}_{\mathrm{xz}}\end{array}\right],$ $A^x=\left[\begin{array}{ccccc}0& 0& \frac{1}{\mathrm{\ρ}}& 0& 0\\ 0& 0& 0& 0& \frac{1}{\mathrm{\ρ}}\\ C_{11}& 0& 0& 0& 0\\ C_{13}& 0& 0& 0& 0\\ 0& C_{44}& 0& 0& 0\end{array}\right],$ $A^z=\left[\begin{array}{ccccc}0& 0& 0& 0& \frac{1}{\mathrm{\ρ}}\\ 0& 0& 0& \frac{1}{\mathrm{\ρ}}& 0\\ 0& C_{13}& 0& 0& 0\\ 0& C_{33}& 0& 0& 0\\ C_{44}& 0& 0& 0& 0\end{array}\right],---\left(2\right)$

v _x, v _zBe respectively the horizontal component and the vertical component of speed, τ _IjBe the components of stress, C _IjBe elastic constant, ρ is a Media density.

Utilize high-order staggered-mesh finite difference scheme that (1) formula is dispersed, obtain elasticity and just drilling propagation operator and inverse time extrapolation operator.

Elasticity is just being drilled propagation operator:

$V_{i,j}^n=V_{i,j}^{n+1}-\mathrm{\Δ}{\mathrm{tA}}^x{L}_x\left(V_{\mathrm{ai},j}^{n+\frac{1}{2}}\right)-\mathrm{\Δ}{\mathrm{tA}}^z{L}_z\left(V_{\mathrm{bi},j}^{n+\frac{1}{2}}\right)---\left(3\right)$

Elasticity inverse time extrapolation operator:

$V_{i,j}^{n+1}=V_{i,j}^n+\mathrm{\Δ}{\mathrm{tA}}^x{L}_x\left(V_{\mathrm{ai},j}^{n+\frac{1}{2}}\right)+\mathrm{\Δ}{\mathrm{tA}}^z{L}_z\left(V_{\mathrm{bi},j}^{n+\frac{1}{2}}\right)---\left(4\right)$

Wherein,

$\left\{\begin{array}{c}{V}_{\mathrm{ai},j}^n={[v_{\mathrm{xi}+\frac{1}{2},j}^n,v_{\mathrm{zi},j+\frac{1}{2}}^n,{\mathrm{\τ}}_{\mathrm{xxi},j}^n,{\mathrm{\τ}}_{\mathrm{zzi},j}^n,{\mathrm{\τ}}_{\mathrm{xzi}+\frac{1}{2},j+\frac{1}{2}}^n]}^T\\ V_{\mathrm{bi},j}^n={[v_{\mathrm{xi},j+\frac{1}{2}}^n,v_{\mathrm{zi}+\frac{1}{2},j}^n,{\mathrm{\τ}}_{\mathrm{xxi},j}^n,{\mathrm{\τ}}_{\mathrm{zzi}+\frac{1}{2},j+\frac{1}{2}}^n,{\mathrm{\τ}}_{\mathrm{xzi},j}^n]}^T\end{array}\right.,---\left(5\right)$

L _x, L _zBe space higher difference operator.The 2L jump divides operator to be expressed as follows:

$\left\{\begin{array}{c}{L}_x\left(V_{i,j}^n\right)=\frac{1}{\mathrm{\Δx}}\underset{m=1}{\overset{L}{\mathrm{\Σ}}}{a}_m(V_{i+\frac{2m-1}{2},j}^n-V_{i-\frac{2m-1}{2},j}^n)\\ L_z\left(V_{i,j}^n\right)=\frac{1}{\mathrm{\Δz}}\underset{m=1}{\overset{L}{\mathrm{\Σ}}}{a}_m(V_{i,j+\frac{2m-1}{2}}^n-V_{i,j-\frac{2m-1}{2}}^n)\end{array}\right.,---\left(6\right)$

a _mBe the higher difference coefficient, can determine by following equation:

$\left[\begin{array}{cccc}1& 3^1& L& {(2L-1)}^1\\ 1& 3^2& L& {(2L-1)}^3\\ M& \mathrm{<figure-callout\; id="1"\; label="M\; O\; M"\; filenames="HSA00000486866200012.png"\; state="\{\{state\}\}">M</figure-callout>}& O& M\\ 1& 3^{2L-1}& L& {(2L-1)}^{2L-1}\end{array}\right]\left[\begin{array}{c}{a}_1\\ a_2\\ M\\ a_L\end{array}\right]=\left[\begin{array}{c}1\\ 0\\ M\\ 0\end{array}\right].---\left(7\right)$

Carry out the reverse-time migration of many components associating elasticity, need utilize elasticity just drilling propagation operator and extrapolation operator makes up source wavefield and detection wave field respectively.For source wavefield, load given seismic wavelet, utilize propagation operator (3) formula of just drilling, make up underground all focus vector wave fields along the forward time; For the detection wave field, as boundary condition, utilize inverse time extrapolation operator (4) formula with many component datas of ground record, extrapolate the underground detection vector of reconstruct wave field along reversion time.

2, based on the foundation of the elastic wave simple crosscorrelation image-forming condition of illumination compensation

In order flexibly to shake the accurate imaging of vector wave field, must use image-forming condition to underground.Owing to use the elastic wave imaging, must consider the combination of all seismic event wave mode patterns, and this combination also should have clear physical meaning.In the disposal route of traditional multi-component earthquake data, P ripple or PS ripple have only been used in the wave field structure stage, therefore imaging separately in imaging.Exactly because and multi-component seismic data associating elasticity reverse-time migration formation method adopts the mode of " associating of many components " to make up underground all vector wave fields in the wave field structure stage, thereby can carry out the simple crosscorrelation combination to multiple wave mode pattern in the imaging stage and carry out imaging, obtain PP, SS, SP and the PS ripple imaging results of explicit physical meaning.

Finish said process and must use elastic wave simple crosscorrelation image-forming condition based on illumination compensation.Simple crosscorrelation image-forming condition in the reverse-time migration is actually for having the summation that superposes of same spatial location, the focus of same propagation time and detection wave field.For the reverse-time migration of many component associating elasticity, need carry out the simple crosscorrelation combination to multiple wave mode pattern, therefore before utilizing the simple crosscorrelation image-forming condition, should at first decompose focus and detection vector wave field, obtain corresponding P ripple and S wave component, focus S with identical time, same position is carried out relevant summation with detection R wave field form reflecting interface, final imaging results is that all big guns, institute are sued for peace if having time.For example the PS imaging results is the P wave component S with source wavefield _p(x is t) with the S wave component R of detection wave field _s(x, t) simple crosscorrelation obtains; The imaging results of other wave mode is carried out simple crosscorrelation and is tried to achieve by choosing corresponding source wavefield and detection wave field.And, in the simple crosscorrelation imaging,, should carry out illumination compensation in order better to keep the characteristic of amplitude.With the auto-correlation of all imaging results, guarantee the accuracy of imaging amplitude divided by source wavefield.Therefore, the present invention has utilized the elastic wave simple crosscorrelation image-forming condition based on illumination compensation:

$I_{\mathrm{pp}}=\underset{0}{\overset{t_{\max }}{\&Integral;}}{S}_p(x,t)R_p(x,t)/\underset{0}{\overset{t_{\max }}{\&Integral;}}{S}_p(x,t)S_p(x,t)$

$I_{\mathrm{ps}}=\underset{0}{\overset{t_{\max }}{\&Integral;}}{S}_p(x,t)R_s(x,t)/\underset{0}{\overset{t_{\max }}{\&Integral;}}{S}_p(x,t)S_p(x,t)$

(8)

$I_{\mathrm{sp}}=\underset{0}{\overset{t_{\max }}{\&Integral;}}{S}_s(x,t)R_p(x,t)/\underset{0}{\overset{t_{\max }}{\&Integral;}}{S}_s(x,t)S_s(x,t)$

$I_{\mathrm{ss}}=\underset{0}{\overset{t_{\max }}{\&Integral;}}{S}_s(x,t)R_s(x,t)/\underset{0}{\overset{t_{\max }}{\&Integral;}}{S}_s(x,t)S_s(x,t)$

Wherein, S (x, t) and R (x t) represents source wavefield and detection wave field respectively, and subscript P and S represent associated P ripple and S wave component respectively; X represents the locus of imaging, t _MacRepresent the maximum time in the wave field structure; I _Pp, I _Ps, I _SpAnd I _SsRepresent PP, PS, SP, SS ripple imaging results respectively.

3, carry out low pass angle domain filtering compacting lower wave number noise

Because the round trip wave equation is adopted in reverse-time migration, therefore in source wavefield and detection wave field were reconstructed, running into the wave impedance interphase will reflect.In imaging, the detection wave field of the source wavefield of forward-propagating and back-reflection carries out simple crosscorrelation, and perhaps the detection wave field of the source wavefield of back-reflection and forward-propagating carries out simple crosscorrelation, all can produce the lower wave number illusion of not expecting.The angle that these lower wave number illusions show as the wave vector of source wavefield and detection wave field can produce the imaging results that does not have physical significance when being the super large angle, especially show shallow-layer and high-amplitude wave impedance boundary, these lower wave number illusion severe contaminations the result of migration imaging.Carry out the filtering of low pass angle domain during the simple crosscorrelation imaging and can suppress the lower wave number illusion.For these noises are suppressed, the present invention uses Laplace filtering with cos after imaging ²θ advances super large angle imaging results is decayed:

${\&dtri;}^{2}f=\frac{{\&PartialD;}^{2}f}{{\&PartialD;x}^2}+\frac{{\&PartialD;}^{2}f}{{\&PartialD;y}^2}+\frac{{\&PartialD;}^{2}f}{{\&PartialD;z}^2}---\left(9\right)$

Embodiment 2: the present invention is an example with the Marmousi2 model, and in conjunction with the accompanying drawings content of the present invention and realization principle is described:

For adopting the Marmousi2 model, the correctness of testing elastic wave imaging method carries out test analysis.

1, input multi-component seismic data

The input multi-component seismic data does not carry out any type of wave field separation, under the situation of not introducing noise, fully, nondestructively utilizes multi-component seismic data to the full extent.

2, make up underground vector wave field

Utilize the elastic wave propagation operator to make up focus and detection vector wave field respectively.For source wavefield, load given seismic wavelet, utilize propagation operator (3) formula of just drilling, make up underground all focus vector wave fields along the forward time.For the detection wave field, as boundary condition, utilize inverse time propagation operator (4) formula with many component datas of ground record, extrapolate the underground detection vector of reconstruct wave field along reversion time.

3, utilize elastic wave simple crosscorrelation image-forming condition to carry out imaging based on illumination compensation

In to single big gun processing procedure, on the basis of source wavefield that makes up and detection wave field, decompose P ripple and the S wave component that obtains source wavefield and detection wave field respectively by wave field; Application is carried out imaging based on elastic wave simple crosscorrelation image-forming condition (8) formula of illumination compensation to each big gun, obtains single big gun PP, PS, SP, SS ripple imaging results respectively; Final imaging results is all single big gun migration imaging results' stack (accompanying drawing 1).Accompanying drawing 1 (a)～(d) is respectively Marmousi2 model PP ripple, PS ripple, SP ripple and SS ripple imaging results.Can see that from accompanying drawing 1 (a) and 1 (d) imaging results seriously disturbed by the lower wave number illusion, and accompanying drawing 1 (b) and 1 (c) also there is part lower wave number illusion at strong impedance interphase.

4, carry out low pass angle domain filtering compacting lower wave number noise

To the imaging results of all big guns, use (9) formula, carry out the Laplace denoising, final imaging results is as shown in Figure 2.Accompanying drawing 2 (a)～(d) is respectively final Marmousi2 model PP ripple, PS ripple, SP ripple and SS ripple imaging results.Can see that the lower wave number illusion is eliminated fully.Excite and obtain because the multi-component seismic data of input is a p-wave source, so the main energy of source wavefield is the P ripple, and the S wave component S ripple that to be the interphase conversion of energy by place, seabed in the Marmousi2 model of P ripple epicenter excitation form, so PP ripple and PS ripple imaging results are better.As seen from the figure, not only for level, inclination reflecting interface, three major rupture and two accurate imagings of anticline, and the hydrocarbon-bearing pool in left side also portray clear accurately.4 imaging section (PP ripple, PS ripple, SP ripple and SS ripple) have clear physical meaning, help further GEOPHYSICAL EXPLANATION and petrophysical parameter inverting.

Accompanying drawing 3 is the many component associatings of earthquake of the present invention elasticity reverse-time migration formation method process flow diagram.Compare with conventional multi-component seismic data image processing method, the present invention uses the underground elasticity vector seismic wave field of the common structure of many components by directly importing multi-component seismic data, uniting, and is used for the accurate imaging of complicated ball medium.

Claims (4)

1. the many components of earthquake are united elasticity reverse-time migration formation method, it is characterized in that concrete image-forming step is as follows:

A. import the multi-component seismic common-shot-gather of handling without wave field separation;

B. make up underground elasticity vector wave field based on many components equations for elastic waves;

C. many component associating elastic wave simple crosscorrelation image-forming conditions of using illumination compensation obtain four kinds of elastic wave imaging results of explicit physical meaning;

D. carry out the lower wave number noise in the reverse-time migration of low pass angle domain filtering compacting elasticity.

2. the many component associatings of earthquake according to claim 1 elasticity reverse-time migration formation method, it is characterized in that: described step b comprises the steps:

B1. load given seismic wavelet, utilize the elasticity vector just drilling propagation operator, make up subsurface source vector wave field along the forward time;

B2. with many component datas of ground record as boundary condition, utilize elasticity vector inverse time extrapolation operator, extrapolate along reversion time, make up underground detection vector wave field.

3. the many component associatings of earthquake according to claim 1 elasticity reverse-time migration formation method, it is characterized in that: described step c comprises the steps:

C1. decompose P ripple, the S wave component that obtains source wavefield and detection wave field respectively by wave field;

C2. use the many components associating elastic wave simple crosscorrelation image-forming conditions of illumination compensation, obtain explicit physical meaning, amplitude characteristic many ripples joint imaging result accurately;

C3. each big gun is carried out imaging, obtain single big gun PP, PS, SP, SS ripple imaging results respectively, final imaging results is all single big gun migration imaging results' stack.

4. the many component associatings of earthquake according to claim 3 elasticity reverse-time migration formation method, it is characterized in that: many component associating elastic wave simple crosscorrelation image-forming conditions of described illumination compensation, imaging results by four kinds of explicit physical meaning is formed, wherein the imaging results of PP ripple is to carry out simple crosscorrelation by the P wave component of source wavefield and the P wave component of detection wave field, obtains divided by the auto-correlation of source wavefield P wave component again; The imaging results of PS ripple is that the P wave component by source wavefield becomes to carry out the branch simple crosscorrelation with the S ripple of detection wave field, obtains divided by the auto-correlation of source wavefield P wave component again; The imaging results of SP ripple is that the P wave component by the S wave component of source wavefield and detection wave field carries out simple crosscorrelation, obtains divided by the auto-correlation of source wavefield S wave component again; The imaging results of SS ripple is that the S wave component by source wavefield becomes to carry out the branch simple crosscorrelation with the S ripple of detection wave field, obtains divided by the auto-correlation of source wavefield S wave component again.

Images