CN102590858B - Two-way wave imaging method based on broadband wavelet reconstruction - Google Patents

Abstract

The invention discloses a two-way wave imaging method based on broadband wavelet reconstruction. The method comprises the following steps of: (1) carrying out frequency spectrum analysis on a single shot in seismic acquisition and determining the main frequency and dominant frequency distribution range of the single shot; (2) determining a wave form and an amplitude spectrum of a Ricker wavelet according to the main frequency of seismic data; (3) constructing a broadband source wavelet in combination with the frequency-band range of seismic data; (4) realizing forward modeling of the source wavelet and recording reverse-time extrapolation in a seismic wave field; (5) segmenting the recording time in the forward pass of a shot wave field and storing only two levels of wave field data at the beginning in each segment; and (6) carrying out reverse-time extrapolation and relatively imaging seismic wave fields. According to the method, the best matching of the source wavelet and the seismic recording and the precise extrapolation of the full waves are realized, the difficulty of precise imaging of complex construction and complex seismic wave fields is solved well, the amplitude and phase information of deep seismic imaging is protected effectively and the reliability of precisely imaging earthquake in a deep domain is guaranteed.

Description

Round trip wave imaging method based on the reconstruct of wideband wavelet

Technical field

The invention belongs to oil seismic exploration technical field, relate in particular to a kind of new round trip wave imaging method based on the reconstruct of wideband wavelet being applicable to polytype complicated seismic wave field, complex structure accurately image.

Background technology

In recent years, along with deepening continuously of oil seismic exploration work, seismic prospecting develops to complex structural area, and the problem that seismic survey work faces is more complicated, wants to obtain the accurate structural feature of underground medium more difficult.Imaging technique has developed for many years, is also the effective means of obtaining underground structure.In prestack depth domain imaging method, the method that at present industry member adopts comprises Integral Solution of Wave Equation method based on ray theory and the differential wave equation one way ripple solution based on wave theory, and these two class methods are difficult to process the high steep dip structure imaging problem that seismic event lateral speed change is violent.

Recently, the geophysical survey field higher reverse-time migration formation method of a kind of precision that grows up, the method adopts round trip ripple to solve differential wave equation, be not subject to impact, the imaging precision of medium lateral speed change and high steep dip structure high, can utilize the advantages such as the correct imaging of bow-tie, made up theoretically the image defects that common seismic skew faces.But in the process of application, due to the impact of difference, difference scheme inverse time wave field extrapolation algorithm and the image-forming condition of source wavelet and seismologic record, imaging results is undesirable aspect the accuracy of frequency, image space.Source wavelet has a certain impact to prestack round trip ripple imaging tool, and it affects imaging phase place and imaging resolution. conventionally choose single-frequency Ricker wavelet.Therefore, source wavelet and seismologic record wavelet there are differences, and seismic imaging resolution is lower, affects seismic imaging effect.

Summary of the invention

Technical matters to be solved by this invention is to have overcome conventional round trip ripple imaging technique to cause seismic imaging frequency band narrow because of single wavelet frequency, the false defect of frequency content; Through repeatedly studying and repetition test, and a kind of wideband source wavelet of reconstruct, the reverse-time migration high precision formation method based on multiple dimensioned wideband source wavelet; The method can make source wavelet more approach seismologic record wavelet, the land geological data accurately image that realization theory model data, marine seismic data and wavelet thereof are changeable; There is seismic imaging phase place accurate, the feature that resolution is higher and image space is true and reliable.

The technical scheme that technical solution problem of the present invention is taked is: a kind of round trip wave imaging method based on the reconstruct of wideband wavelet, is characterized in that: the reconstruct of wideband source wavelet and inverse time wave field extrapolation, and concrete steps comprise:

A, single big gun of earthquake-capturing is done to spectrum analysis, determine the dominant frequency of single big gun, visual dominant frequency scope is between 20～40Hz, and dominant frequency distribution range can be between 5～70Hz;

B, according to geological data dominant frequency, utilize following formula to determine waveform and the spectral amplitude of Ricker wavelet;

r(t)＝[1-2(πwt) ²]exp[1-(πwt) ²] (1)

$R\left(f\right)=\frac{{2f}^2}{g^3\sqrt{\mathrm{\π}}}\exp [-{\left(\frac{f}{w}\right)}^2]---\left(2\right)$

Formula (1) represents Ricker wavelet waveform, and in formula, w is frequency; Formula (2) represents Ricker wavelet spectral amplitude, and in formula, g is the constant relevant with frequency;

C, on step B basis, utilize formula (3)～(6), in conjunction with earthquake frequency ranges of data scope, build wideband source wavelet;

$w\left(t\right)=\frac{1}{w_2-w_1}{\∫}_{w_1}^{w_2}r\left(t\right)\mathrm{dw}---\left(3\right)$

$w\left(t\right)=\frac{1}{w_2-w_1}{\∫}_{w_1}^{w_2}[1-2{\left(\mathrm{\πwt}\right)}^2]\exp [-{\left(\mathrm{\πwt}\right)}^2]\mathrm{dw}---\left(4\right)$

$w\left(t\right)=\frac{1}{w_2-w_1}\left\{w_2\exp \right[-{\left({\mathrm{\πw}}_{2}t\right)}^2]-w_1\exp [-{\left({\mathrm{\πw}}_{1}t\right)}^2\left]\right\}---\left(5\right)$

$Y\left(f\right)=\frac{1}{(w_2-w_1)\sqrt{\mathrm{\π}}}\left\{\exp \right[-{\left(\frac{f}{w_2}\right)}^2]-\exp [-{\left(\frac{f}{w_1}\right)}^2\left]\right\}---\left(6\right)$

W in formula ₁, w ₂for seismic data minimax frequency, conventionally according to the result of spectrum analysis, to determine, parameter area is generally 4～80H _z;

D, according to three-dimensional medium round trip ACOUSTIC WAVE EQUATION, utilize time second order centered finite difference approximate utilize Spatial higher order centered finite difference approximate releasing truncation error is o (Δ _x ^m, Δ _y ^m, Δ _z ^m, Δ _t ²) three-dimensional high-order difference wave field extrapolation equation:

$\begin{array}{c}{u}_{i,j,k}^{n\±1}={2u}_{i,j,k}^n-u_{i,j,k}^{n\stackrel{\‾}{+}1}+\frac{1}{2}{\left(\frac{\mathrm{v\Δt}}{\mathrm{\Δx}}\right)}^2[{\mathrm{\ω}}_0{u}_{i,j,k}^n+\underset{m=1}{\overset{\frac{M}{2}}{\mathrm{\Σ}}}{\mathrm{\ω}}_m(u_{i+m,j,k}^n+u_{i-m,j,k}^n)]\\ +\frac{1}{2}{\left(\frac{\mathrm{v\Δt}}{\mathrm{\Δy}}\right)}^2[{\mathrm{\ω}}_0{u}_{i,j,k}^n+\underset{m=1}{\overset{\frac{M}{2}}{\mathrm{\Σ}}}{\mathrm{\ω}}_m(u_{i,j+m,k}^n+u_{\mathrm{i\; j}-m,k}^n)]\\ +\frac{1}{2}{\left(\frac{\mathrm{v\Δt}}{\mathrm{\Δz}}\right)}^2[{\mathrm{\ω}}_0{u}_{i,j,k}^n+\underset{m=1}{\overset{\frac{M}{2}}{\mathrm{\Σ}}}{\mathrm{\ω}}_m(u_{i,j,k+m}^n+u_{i,j,k-m}^n)]\end{array}---\left(7\right)$

In formula, P is seismic wave field, and x, y, z is coordinate, and M is space difference order, and u is space wave field function, and i, j, k, n are the net point sequence number of x, y, z direction, ω ₀, ω _mfor difference coefficient;

Adopt above formula equation, can realize source wavelet and just drill and recording the extrapolation of seismic wave field inverse time;

E, according to step D, in big gun wave field is just pushing through journey, whole T segmentation writing time, add absorbing boundary condition simultaneously, during extrapolation by storage every section of initial time two-layer wave field data in machine, store;

F, after step e completes, start inverse time extrapolation imaging; Two-layer wave field while utilizing this section of step e storage initial in this process recalculates whole wave field value of this section, and be kept in machine internal memory, in this section, record time step of the every backstepping of wave field, just to the corresponding big gun wave field constantly of preserving in internal memory carry out relevant, by formula (8) to obtain the imaging results in this moment;

$\mathrm{Image}(x,y)={\∫}_0^{T\max }{P}_D(x,z,t)P_H(x,z,t)\mathrm{dt}---\left(8\right)$

Integrand p wherein _d(x, z, t) p _h(x, z, t) represents that t constantly does Polaroid computing to whole wave field, and in integration explanation image space Image (x, z) similarly is the stack of a time step imaging; So simple crosscorrelation image-forming condition takes full advantage of image-forming information, in Enhanced Imaging signal, also effectively suppressed imaging noise; According to above step, the wave field inverse time extrapolation of calculated amount maximum is realized by GPU; And utilize random velocity border to improve the concurrency of wave field extrapolation algorithm, solved the I/O problem of Mass storage.

Described structure wideband source wavelet adopts difference method to realize source wavefield and just drills and be specially: on single shot record corresponding to geological data, carry out spectrum analysis and determine frequency range, according to earthquake recording frequency, build source wavelet, utilize high-order limited difference method just drilling source wavefield to the maximum record time, wave field data are just being drilled in storage simultaneously.

Remarkable result of the present invention is:

(1), the inventive method realized the optimum matching of source wavelet and seismologic record, realized all-wave field and accurately extrapolated; Solve well an accurately image difficult problem for complicated seismic wave field, effectively protected amplitude and the phase information of deep earthquake imaging, guaranteed the reliability of degree of depth accurately image.

(2), the present invention builds source wavelet by the frequency parameter of seismologic record, guaranteed the optimum matching of source wavelet and seismologic record.

(3), the simple crosscorrelation image-forming principle that adopts of the present invention, simple crosscorrelation image-forming condition takes full advantage of image-forming information, has also effectively suppressed imaging noise in Enhanced Imaging signal.

(4), the present invention adopts wideband wavelet can effectively preserve the frequency band in geological data, because the spectral amplitude of wideband wavelet skew is than Ricker wavelet amplitude of deflection spectrum bandwidth, is more conducive to the analysis to migration result.

Accompanying drawing explanation

Fig. 1 is wideband wavelet reconstruct round trip ripple image-forming principle block diagram of the present invention

Fig. 2 a is BLH seismologic record source map of the present invention

Fig. 2 b is BLH seismologic record data spectrogram of the present invention

Fig. 3 a is Ricker wavelet oscillogram of the present invention

Fig. 3 b is Ricker wavelet spectrogram of the present invention

Fig. 4 a is reconstruct wideband wavelet oscillogram of the present invention

Fig. 4 b is reconstruct wideband wavelet spectrum figure of the present invention

Fig. 5 a is 40Hz Ricker wavelet migration result figure of the present invention (pressure make an uproar before)

Fig. 5 b is 40Hz Ricker wavelet migration result figure of the present invention (pressure make an uproar after)

Fig. 6 a is wideband wavelet migration result figure of the present invention (pressure make an uproar before)

Fig. 6 b is wideband wavelet migration result figure of the present invention (pressure make an uproar after)

Fig. 7 a is Ricker wavelet migrated section amplitude spectrogram of the present invention

Fig. 7 b is wideband wavelet migrated section amplitude spectrogram of the present invention

Fig. 8 a is that the present invention extracts Ricker wavelet imaging results oscillogram

Fig. 8 b is reconstruct wideband wavelet imaging results oscillogram of the present invention

Fig. 9 is one way ripple imaging results figure of the present invention

Figure 10 is wideband wavelet round trip ripple imaging results figure of the present invention

Embodiment

Below in conjunction with drawings and Examples, the present invention is described in further detail

Embodiment 1

As shown in Figure 1, the embodiment of the present invention provides wideband wavelet reconstruct round trip ripple imaging flow process:

First according to e, determine frequency trend scope; when need to process e, protect wave field, amplitude and frequency information; according to the parameter that e is definite, determine waveform and the spectral amplitude of Ricker wavelet according to geological data dominant frequency, build wideband source wavelet and obtain a, e can obtain h according to step D, E.By a → d, can be completed by step D, E, and preserve wave field value.In the process of practical application, due to the restriction of machine internal memory and storage, do not preserve whole wave field information.By a to d source wavefield along in time forward extrapolation process, only store this section of two-layer wave field when initial and add absorbing boundary condition.By e, extrapolated in imaging process to the h inverse time, the two-layer wave field while utilizing this section of storage initial recalculates whole wave field value of this section, and is kept in internal memory, also can be stored on hard disk, when the imaging of i place, reads out.Record time step of the every backstepping of wave field, just carry out relevantly to the corresponding big gun wave field constantly of preserving in internal memory, obtain the imaging results in this moment, until every section of imaging is complete, the imaging processing of other sections repetition said process.

Its principle of method of the present invention is: take seismologic record and Ricker wavelet as basis, first carry out seismologic record analysis and wavelet reconstruct, the more approaching wideband source wavelet of structure and real seismic record carries out wave field just drills, adopt advanced wave field extrapolation operator to carry out the inverse time extrapolation of seismologic record wave field, by wave field related realization Depth Domain seismic imaging simultaneously.

Embodiment 2

For application example, hills, grassland be take as main in earth's surface, basin, the Caspian Sea, shore, has river, swampland, clay ground simultaneously, and middle low, two ends are high, and the northern cliff of displacement in work area is more, ravines and guillies criss-cross, and relief is larger; Work area surface structure major part is three layers (weathering zone, reduction of speed band, high-velocity beds), and the variation of weathering zone speed and thickness is milder, and its speed is 350-900 meter per second, and its one-tenth-value thickness 1/10 is 10-15 rice, and relevant to earth's surface elevation; Adopt three-dimensional data (gathering for 2005,2006,2007,2008) in flakes, it is all the 3-D seismics record that vibroseis gathers, from Fig. 7 a, seismic data is shallow, in, deep reflex ripple group is more complete, energy is moderate, signal to noise ratio (S/N ratio) is higher, overlaps reflection wave groups clear continuously more, illustrates that the quality of source book is pretty good.But huge thick salt dome is grown by this area, seismic wave field is very complicated, between different masses three-dimensional, exist the differences such as seismic wave waveform, amplitude, because the complexity of earth's surface and underground structure causes this area's seismic imaging problem serious, through repetition test repeatedly, find to adopt common seismic imaging technique cannot solve this district's accurately image problem.

We utilize the round trip wave imaging method that the present invention is based on the reconstruct of wideband wavelet to carry out seismic imaging processing in BLH basin, have obtained very desirable earthquake effect; Operation steps is as follows:

1), the earthquake big gun record of field acquisition is carried out to conventional processing, for high-quality big gun collection is prepared in the skew of round trip ripple;

2), at the enterprising line frequency analysis of spectrum of single big gun seismologic record, determine frequency range (as shown in Fig. 2 a, Fig. 2 b), from frequency analysis, 500-1200ms shallow-layer significant wave frequency mainly concentrates between 8-75Hz, depending on dominant frequency in about 40Hz; 1600-2400ms objective interval significant wave frequency mainly concentrates between 6-70Hz, depending on dominant frequency in about 30Hz; 3600-4400ms deep layer significant wave frequency mainly concentrates between 5-50Hz, depending on dominant frequency in about 20Hz.

3) utilize frequency parameter to build wideband source wavelet

Referring to Fig. 3 a, be Ricker wavelet oscillogram, Fig. 3 b is Ricker wavelet spectrogram; Fig. 4 a is reconstruct wideband wavelet oscillogram, and Fig. 4 b is reconstruct wideband wavelet spectrum figure; Frequency parameter by seismologic record builds source wavelet, guarantees the optimum matching of source wavelet and seismologic record.

4), according to step D, E, can realize source wavelet and just drill, the forward that forms focal point is push wave field, and according to above-mentioned implementation storage wave field.

5), seismologic record is read in to internal memory, according to given imaging space parameter, according to step D, E, carry out the inverse time continuation of all-wave wave equation, extension equation is as follows:

$\begin{array}{c}{u}_{i,j,k}^{n\±1}={2u}_{i,j,k}^n-u_{i,j,k}^{n\stackrel{\‾}{+}1}+\frac{1}{2}{\left(\frac{\mathrm{v\Δt}}{\mathrm{\Δx}}\right)}^2[{\mathrm{\β}}_0{u}_{i,j,k}^n+\underset{m=1}{\overset{\frac{M}{2}}{\mathrm{\Σ}}}{\mathrm{\β}}_m(u_{i+m,j,k}^n+u_{i-m,j,k}^n)]\\ +\frac{1}{2}{\left(\frac{\mathrm{v\Δt}}{\mathrm{\Δy}}\right)}^2[{\mathrm{\ω\β}}_0{u}_{i,j,k}^n+\underset{m=1}{\overset{\frac{M}{2}}{\mathrm{\Σ}}}{\mathrm{\β}}_m(u_{i,j+m,k}^n+u_{\mathrm{i\; j}-m,k}^n)]\\ +\frac{1}{2}{\left(\frac{\mathrm{v\Δt}}{\mathrm{\Δz}}\right)}^2[{\mathrm{\ω}}_0{u}_{i,j,k}^n+\underset{m=1}{\overset{\frac{M}{2}}{\mathrm{\Σ}}}{\mathrm{\ω}}_m(u_{i,j,k+m}^n+u_{i,j,k-m}^n)]\end{array}$

Calculating parameter:

When M=4, β ₀=-5.0, β ₁=2.666667, β ₂=-0.1666667

When M=6, β ₀=-5.444444, β ₁=3.000000,

β ₂＝-0.3000003，β ₃＝0.0222225，

When M=8, β ₀=-2.847222054, β ₁=3.20000000, β ₂=-0.4000002,

β ₃＝0.05079369，β ₄＝-0.003571436

When M=10, β ₀=-5.8544445, β ₁=3.333333, β ₂=-0.4761901,

β ₃＝0.07936513，β ₄＝-0.009920621，β ₅＝0.0006349185

6), adopt dependent imaging condition to carry out wave field dependent imaging;

Consider the realizability of complex wave field, the simple crosscorrelation image-forming principle that we adopt, the imaging formula of simple crosscorrelation image-forming condition is (8) formula, and simple crosscorrelation image-forming condition takes full advantage of image-forming information, has also effectively suppressed imaging noise in Enhanced Imaging signal.

7), according to the round trip wave imaging method that the present invention is based on the reconstruct of wideband wavelet, theoretical model is carried out to numerical experimentation;

Referring to Fig. 5 a, be 40Hz Ricker wavelet migration result figure (pressure make an uproar before), Fig. 5 b is 40Hz Ricker wavelet migration result figure (pressure make an uproar after); Fig. 6 a is wideband wavelet migration result figure (pressure make an uproar before), and Fig. 6 b is wideband wavelet migration result figure (pressure make an uproar after).No matter two kinds of wavelets are to contrast after the front contrast of denoising or denoising, visually all can not find out obvious difference, in order further to analyze this two kinds of wavelets impact on migration result, the spectral amplitude that we produce section to two kinds of wavelets is analyzed, as shown in Fig. 7 a, Fig. 7 b.Simultaneously also analyzed the waveform character that two kinds of wavelets obtain imaging results, as shown in Fig. 8 a, Fig. 8 b.By figure, found out, the spectral amplitude of wideband wavelet skew, than Ricker wavelet amplitude of deflection spectrum bandwidth, therefore, is used broadband wavelet more can preserve the effective band in geological data, is conducive to migration result analysis.

8), according to the technology of the present invention, BIH actual seismic data are tested

Referring to Fig. 9, be one way ripple imaging results, Figure 10 is wideband wavelet round trip ripple imaging results; From comparison diagram, can find out, wideband wavelet round trip ripple wave equation Depth Domain imaging technique has been realized the accurate playback to salt dome border and salt dome flank, make interface, salt brae, salt flank border, be greater than structure energy accurately image under 90 ° of special salt of salt flank, imaging section predominant frequency is higher simultaneously, frequency information abundant, more be conducive to seismic reservoir study, by theoretical model tentative calculation and real data test, proved the advantage that the round trip wave imaging method based on the reconstruct of wideband wavelet has.

The above, as preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of doing, is equal to replacement, within improvement waits and all should be included in protection scope of the present invention.

Claims (2)

1. the round trip wave imaging method based on the reconstruct of wideband wavelet, is characterized in that: the reconstruct of wideband source wavelet and inverse time wave field extrapolation, and concrete steps comprise:

A, single big gun of earthquake-capturing is done to spectrum analysis, determine the dominant frequency of single big gun, visual dominant frequency scope is between 20～40Hz, and dominant frequency distribution range can be between 5～70Hz;

B, according to geological data dominant frequency, utilize following formula to determine waveform and the spectral amplitude of Ricker wavelet;

r(t)＝[1-2(πwt) ²]exp[1-(πwt) ²] (1)

$R\left(f\right)=\frac{{2f}^2}{g^3\sqrt{\mathrm{\π}}}\exp [-{\left(\frac{f}{w}\right)}^2]---\left(2\right)$

Formula (1) represents Ricker wavelet waveform, and in formula, w is frequency; Formula (2) represents Ricker wavelet spectral amplitude, and in formula, g is the constant relevant with frequency;

C, on step B basis, utilize formula (3)～(6), in conjunction with earthquake frequency ranges of data scope, build wideband source wavelet;

$w\left(t\right)=\frac{1}{w_2-w_1}{\∫}_{w_1}^{w_2}r\left(t\right)\mathrm{dw}---\left(3\right)$

$w\left(t\right)=\frac{1}{w_2-w_1}{\∫}_{w_1}^{w_2}[1-2{\left(\mathrm{\πwt}\right)}^2]\exp [-{\left(\mathrm{\πwt}\right)}^2]\mathrm{dw}---\left(4\right)$

$w\left(t\right)=\frac{1}{w_2-w_1}\left\{w_2\exp \right[-{\left({\mathrm{\πw}}_{2}t\right)}^2]-w_1\exp [-{\left({\mathrm{\πw}}_{1}t\right)}^2\left]\right\}---\left(5\right)$

$Y\left(f\right)=\frac{1}{(w_2-w_1)\sqrt{\mathrm{\π}}}\left\{\exp \right[-{\left(\frac{f}{w_2}\right)}^2]-\exp [-{\left(\frac{f}{w_1}\right)}^2\left]\right\}---\left(6\right)$

W in formula ₁, w ₂for seismic data minimax frequency, conventionally according to the result of spectrum analysis, to determine, parameter area is generally 4～80H _z;

D, according to three-dimensional medium round trip ACOUSTIC WAVE EQUATION, utilize time second order centered finite difference approximate utilize Spatial higher order centered finite difference approximate releasing truncation error is o (Δ _x ^m, Δ _y ^m, Δ _z ^m, Δ _t ²) three-dimensional high-order difference wave field extrapolation equation:

$\begin{array}{c}{u}_{i,j,k}^{n\±1}={2u}_{i,j,k}^n-u_{i,j,k}^{n\stackrel{\‾}{+}1}+\frac{1}{2}{\left(\frac{\mathrm{v\Δt}}{\mathrm{\Δx}}\right)}^2[{\mathrm{\ω}}_0{u}_{i,j,k}^n+\underset{m=1}{\overset{\frac{M}{2}}{\mathrm{\Σ}}}{\mathrm{\ω}}_m(u_{i+m,j,k}^n+u_{i-m,j,k}^n)]\\ +\frac{1}{2}{\left(\frac{\mathrm{v\Δt}}{\mathrm{\Δy}}\right)}^2[{\mathrm{\ω}}_0{u}_{i,j,k}^n+\underset{m=1}{\overset{\frac{M}{2}}{\mathrm{\Σ}}}{\mathrm{\ω}}_m(u_{i,j+m,k}^n+u_{\mathrm{i\; j}-m,k}^n)]\\ +\frac{1}{2}{\left(\frac{\mathrm{v\Δt}}{\mathrm{\Δz}}\right)}^2[{\mathrm{\ω}}_0{u}_{i,j,k}^n+\underset{m=1}{\overset{\frac{M}{2}}{\mathrm{\Σ}}}{\mathrm{\ω}}_m(u_{i,j,k+m}^n+u_{i,j,k-m}^n)]\end{array}---\left(7\right)$

In formula, P is seismic wave field, and x, y, z is coordinate, and M is space difference order, and u is space wave field function, and i, j, k, n are the net point sequence number of x, y, z direction, ω ₀, ω _mfor difference coefficient;

Adopt above formula equation, can realize source wavelet and just drill and recording the extrapolation of seismic wave field inverse time;

E, according to step D, in big gun wave field is just pushing through journey, whole T segmentation writing time, add absorbing boundary condition simultaneously, during extrapolation by storage every section of initial time two-layer wave field data in machine, store;

F, after step e completes, start inverse time extrapolation imaging; Two-layer wave field while utilizing this section of step e storage initial in this process recalculates whole wave field value of this section, and be kept in machine internal memory, in this section, record time step of the every backstepping of wave field, just to the corresponding big gun wave field constantly of preserving in internal memory carry out relevant, by formula (8) to obtain the imaging results in this moment;

$\mathrm{Image}(x,y)={\∫}_0^{T\max }{P}_D(x,z,t)P_H(x,z,t)\mathrm{dt}---\left(8\right)$

Integrand p wherein _d(x, z, t) p _h(x, z, t) represents that t constantly does Polaroid computing to whole wave field, and in integration explanation image space Image (x, z) similarly is the stack of a time step imaging.

2. the round trip wave imaging method based on the reconstruct of wideband wavelet according to claim 1, it is characterized in that, described structure wideband source wavelet adopts difference method to realize source wavefield and just drills and be specially: on single shot record corresponding to geological data, carry out spectrum analysis and determine frequency range, according to earthquake recording frequency, build source wavelet, utilize high-order limited difference method just drilling source wavefield to the maximum record time, wave field data are just being drilled in storage simultaneously.