Wideband Spatiotemporal Variant Seismic Wavelet Extraction Method

Abstract

The present invention provides a method for extracting broadband spatio-temporal variable seismic wavelets. The method for extracting broadband spatio-temporal variable seismic wavelets includes: Step 1, acquiring seismic data and preprocessing the seismic data; Step 2, obtaining seismic interpretation horizons from the preprocessed seismic data and intercepting horizon time windows in the seismic interpretation horizons, where the horizon time windows represent the target area; Step 3, demarcating one or more wavelet spatial variable areas in the target area; Step 4, demarcating one or more time-varying sliding continuation time windows in the one or more wavelet spatial variable areas; Step 5, combining the estimated seismic wavelets obtained from all the time-varying sliding continuation time windows in all the wavelet spatial variable areas to obtain broadband spatio-temporal variable seismic wavelets. The method for extracting broadband spatio-temporal variable seismic wavelets adopts an efficient algorithm and provides time-varying and spatial variable seismic wavelets simultaneously. Compared with the conventional homomorphic domain wavelet extraction algorithm, it has higher stability and accuracy, providing support for high-precision seismic inversion.

Description

Technical Field

[0001] The present invention relates to the technical field of geophysical exploration, and particularly to a method for extracting broadband spatio-temporal variable seismic wavelets. Background Art

[0002] Seismic wavelet estimation is a necessary step in seismic data inversion, forward modeling, deconvolution and other processes, and is also one of the important research topics in the field of seismic data processing. The accuracy of wavelet extraction is one of the key factors restricting the processing accuracy.

[0003] At present, the seismic wavelet extraction methods at home and abroad are mainly based on the convolution model, which are divided into two categories: deterministic and statistical. The deterministic wavelet extraction method means that in the case of having well logging data, using the well logging data and the seismic trace data beside the well, and estimating the seismic wavelet by a mathematical method based on the convolution model. It is one of the most commonly used wavelet extraction methods in inversion research. The advantage of this method is that it does not need to make a certain assumption about the distribution of reflection coefficients. However, due to the need for well logging data to participate, the accuracy of wavelet extraction is closely related to the accuracy of well logging curves and the accuracy of well logging-seismic calibration. The statistical wavelet extraction method is characterized by not requiring the participation of well logging curves, and only using seismic trace data to estimate the wavelet by statistical methods. Usually, certain assumptions need to be made about the seismic data and the distribution of reflection coefficients. The wavelet extraction method in the homomorphic domain is a high-order statistical wavelet extraction method.

[0004] Traditional seismic wavelet extraction methods are based on the assumption that the wavelet is time-invariant. However, due to the absorption and attenuation of seismic wave energy by underground media, the seismic wavelets in actual seismic data have time-varying characteristics; at the same time, the changes in underground geological characteristics and the differences in acquisition and processing also make the seismic wavelets have space-varying characteristics. The accurate extraction of spatio-temporal variable wavelets is a necessary condition for realizing high-precision seismic data inversion and is also an important problem that urgently needs to be solved at present.

[0005] The inventor found that the existing seismic wavelet extraction technologies cannot effectively solve the problem of efficient and high-precision extraction of spatio-temporal variable wavelets. Therefore, it is urgent to develop a method for extracting broadband spatio-temporal variable seismic wavelets.

[0006] In the Chinese patent application with the application number: CN202010013147.X, it involves a method for improving the resolution of thin interbeds in the Gaussian frequency domain based on compressive sensing, including: performing forward modeling of thin interbeds and analyzing their seismic response characteristics; transforming seismic data from the time domain to the frequency domain using the Fourier transform; introducing a Gaussian function that has a denoising and smoothing effect on the signal and transforming the frequency-domain seismic data to the Gaussian frequency domain; extracting seismic wavelets from the original seismic data using the multi-channel complex spectrum technique; performing sparse inversion based on compressive sensing using the extracted seismic wavelets and the Gaussian frequency-domain seismic data to obtain sparse reflection coefficients; performing frequency extension processing based on the extracted seismic wavelets to obtain broadband seismic wavelets, and convolving the broadband seismic wavelets with the sparse reflection coefficients to obtain high-resolution seismic data.

[0007] In the Chinese patent application with the application number: CN201810855372.0, it involves a method for improving the identification ability of thin interbeds based on linear scanning, including: analyzing the dominant frequency band range and phase information of seismic data by performing spectral and waveform feature analysis on the seismic data; extracting wavelets from each trace of seismic data using the complex spectrum and adjusting the frequency band and phase of the obtained wavelets; using the broadband information of the linear scanning signal and sliding the time window to select the optimal linear scanning wavelet; performing wavelet shaping deconvolution on a single trace record using the extracted wavelet and the optimal linear scanning wavelet to obtain the shaping operator for that target trace; convolving the new composite shaping operator with the target seismic trace to obtain a new high-resolution seismic record.

[0008] In the Chinese patent application with the application number: CN201810465628.7, it involves an intelligent time-varying blind deconvolution broadband processing method and device. Among them, the method includes the following steps: extracting an initial seismic wavelet from the attenuated seismic data; randomly generating a group of Q values including multiple formation quality factor Q values; establishing a group of time-varying wavelet matrices according to the initial seismic wavelet and the group of Q values; obtaining multiple reflection coefficients according to the attenuated seismic data and the group of time-varying wavelet matrices; establishing a fitness function; obtaining multiple fitness values according to the fitness function and the multiple reflection coefficients, where one fitness value corresponds to one formation quality factor Q value; determining the optimal formation quality factor Q value according to the multiple fitness values.

[0009] The above existing technologies are all quite different from the present invention and fail to solve the technical problems we want to solve. For this reason, we have invented a new method for extracting broadband spatio-temporal variable seismic wavelets. Summary of the Invention

[0010] The purpose of the present invention is to solve the deficiencies of the existing seismic wavelet extraction technology and provide a method for extracting broadband spatio-temporal variable seismic wavelets to promote the application of broadband spatio-temporal variable seismic wavelets in high-precision seismic data inversion.

[0011] The object of the present invention can be achieved by the following technical measures: a broadband spatio-temporal variable seismic wavelet extraction method, which includes:

[0012] Step 1: Obtain seismic data and preprocess the seismic data.

[0013] Step 2: Obtain seismic interpretation horizons from the preprocessed seismic data, and intercept horizon time windows in the seismic interpretation horizons, where the horizon time windows represent the target area.

[0014] Step 3: Define one or more wavelet variable regions in the target area.

[0015] Step 4: Define one or more time-varying sliding extension time windows in one or more wavelet variable regions.

[0016] Step 5: Combine the estimated seismic wavelets obtained from all time-varying sliding extension time windows in all wavelet variable regions to obtain a broadband spatio-temporal variable seismic wavelet.

[0017] The object of the present invention can also be achieved by the following technical measures:

[0018] In Step 1, the seismic data is post-stack seismic data, which can be fully stacked post-stack seismic data, or can also be partial offset stacked seismic data, partial angle stacked seismic data or partial azimuth stacked seismic data; the preprocessing includes denoising and extracting header information.

[0019] In Step 3, the wavelet variable region refers to dividing the wavelet variable region according to geological characteristics, seismic data acquisition and processing conditions and research needs. If only time-varying wavelets are extracted, the target area is divided into a single variable region; in different variable regions, time-varying wavelets are respectively extracted to obtain the spatio-temporal variable wavelets of the entire target area.

[0020] In Step 4, the time-varying sliding extension time window means that for each variable region, assuming the starting time-varying sampling point is τ 0 , the time-varying sliding interval is Δt, the number of sampling times is N, and the ending time-varying sampling point is τ 0 + N·Δt, and the extension time window length is 2ΔT. Then the set of time-varying sampling points τ is expressed as: τ∈{τ 0 + i·Δt|i = 0, 1, 2…N}, and the time-varying sliding extension time window is expressed as [τ - ΔT, τ + ΔT]. By sliding the time-varying sampling points, the seismic wavelets at the time-varying sampling points are extracted within the range of the time-varying sliding extension time window, and finally the time-varying wavelets are formed.

[0021] In Step 4, for at least one of one or more time-varying sliding extension time windows, the following steps are executed:

[0022] Step 4.1: Use a window function to extract multi-channel random short-time window seismic data from at least one of one or more time-varying sliding extended time windows;

[0023] Step 4.2: Perform frequency-domain filtering processing and homomorphic-domain averaging processing on the multi-channel random short-time window seismic data to obtain homomorphic-domain wavelet data;

[0024] Step 4.3: Transform the homomorphic-domain wavelet data to obtain time-domain wavelet data;

[0025] Step 4.4: Perform wavelet shaping processing on the time-domain wavelet data to obtain an estimated seismic wavelet.

[0026] In Step 4.1, the window function refers to the Hanning window function Win(t). For the time-varying sampling point τ i , its expression is given as follows:

[0027]

[0028] where t 1 , t 2 are the start and end times of the random short-time window respectively, t is the sampling point time,

[0029] In Step 4.1, the multi-channel random short-time window means randomly selecting several channels of data within the space-varying region, randomly selecting several short-time windows within the range of the time-varying sliding extended time window, and using the Hanning window function to obtain short-time window seismic data. The random number of channels and short-time windows are given by the user according to actual needs. The random number of channels is not less than 10% of the total number of channels, and the number of single-channel random short-time windows is not less than 25.

[0030] In Step 4.2, the homomorphic-domain wavelet data means that for a certain time-varying sampling point τ i , S j (ω, τ i ) is the j-th random short-time window seismic data obtained in Step 4.1. S j (ω, τ i ) is represented as S j (ω, τ i ) after Fourier transform and noise reduction filtering to obtain the random short-time window seismic data in the frequency domain. Then the homomorphic-domain wavelet is obtained by the following formula:

[0031]

[0032] where is the homomorphic-domain wavelet, and N is the total number of random short-time windows.

[0033] In step 4.3, the time-domain seismic wavelet data w(t, τ i ) is obtained through the following transformation:

[0034]

[0035] where F -1 is the inverse Fourier transform.

[0036] In step 4.4, the wavelet shaping process includes one or a combination of sidelobe suppression, spectral smoothing, and band-pass filtering, which is used to stabilize the shape of the estimated spatio-temporal varying wavelet.

[0037] The broadband spatio-temporal varying seismic wavelet extraction method in the present invention discloses a broadband spatio-temporal varying seismic wavelet extraction method for addressing the problem of the need for spatio-temporal varying wavelet extraction in high-precision seismic inversion of non-stationary seismic data. It includes: obtaining seismic data and performing preprocessing; obtaining horizon data and intercepting the horizon time window; dividing the wavelet spatial variation region; dividing the time-varying sliding continuation time window; for the seismic data within each time-varying time window in each spatial variation region, performing: intercepting multi-trace random short-time window seismic data, frequency-domain filtering and homomorphic domain averaging, time-domain transformation, and wavelet shaping to obtain the estimated wavelet in this spatial variation region and this time-varying time window; and obtaining the broadband spatio-temporal varying seismic wavelet through wavelet combination. The present invention adopts an efficient algorithm and provides both time-varying and spatially varying seismic wavelets simultaneously. Compared with the conventional homomorphic domain wavelet extraction algorithm, it has higher stability and accuracy, providing support for high-precision seismic inversion.

[0038] Compared with the prior art, the present invention solves the problem of the need for spatio-temporal varying wavelet extraction in high-precision seismic inversion of non-stationary seismic data through the broadband spatio-temporal varying seismic wavelet extraction method. Compared with other time-varying wavelet extraction methods, this method adopts an efficient algorithm and provides both time-varying and spatially varying seismic wavelets simultaneously; in the wavelet extraction of a single time-varying time window in a single spatial variation region, compared with the conventional homomorphic domain wavelet extraction technology, this method extracts wavelets with higher stability and greatly improved accuracy through means such as specific window functions, specific technical processes, and wavelet shaping techniques. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0040] Figure 1 is a flowchart of a specific embodiment of the broadband spatio-temporal varying seismic wavelet extraction method of the present invention;

[0041] Figure 2 Schematic diagram of seismic profile and interpreted horizons in a specific embodiment of the present invention;

[0042] Figure 3 Schematic plan view of seismic data range and variable-offset area division in a specific embodiment of the present invention;

[0043] Figure 4 Schematic diagram of broadband time-varying seismic wavelet extracted from variable-offset area A in a specific embodiment of the present invention;

[0044] Figure 5 Schematic diagram of broadband time-varying seismic wavelet extracted from variable-offset area B in a specific embodiment of the present invention;

[0045] Figure 6 Schematic diagram of broadband seismic wavelet and its spectrogram extracted from a specific time-varying window in a specific embodiment of the present invention;

[0046] Figure 7 Schematic plan view of variable-offset area division in a specific embodiment of the present invention;

[0047] Figure 8 Schematic diagram of broadband time-varying seismic wavelet and its spectrogram extracted from variable-offset area C with near offset in a specific embodiment of the present invention. Detailed implementation manners

[0048] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0049] It should be noted that the terms used herein are only for describing specific implementation manners and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, and / or combinations thereof.

[0050] The method for extracting broadband spatio-temporal variable seismic wavelet of the present invention includes the following steps:

[0051] Step 1, obtaining seismic data and preprocessing the seismic data;

[0052] The seismic data is post-stack seismic data, which can be fully stacked post-stack seismic data, or can be offset-distance stacked seismic data, angle stacked seismic data, or azimuth stacked seismic data; the preprocessing includes denoising and extracting header information.

[0053] Step 2: Obtain seismic interpretation horizons from the preprocessed seismic data, and intercept a horizon time window in the seismic interpretation horizons, where the horizon time window represents the target area;

[0054] Step 3: Define one or more wavelet spatially variant regions in the target area,

[0055] The wavelet spatially variant region refers to the division of the wavelet spatially variant region according to geological characteristics, seismic data acquisition and processing conditions, and research needs. If only time-varying wavelets are extracted, the target area is divided into a single spatially variant region; in different spatially variant regions, time-varying wavelets are extracted respectively to obtain the spatio-temporal variant wavelets of the entire target area.

[0056] Step 4: Define one or more time-varying sliding extension time windows in the one or more wavelet spatially variant regions;

[0057] The time-varying sliding extension time window means that for each spatially variant region, assuming the starting time-varying sampling point is τ 0 , the time-varying sliding interval is Δt, the ending time-varying sampling point is τ 0 +N·Δt, and the length of the extension time window is 2ΔT, then the set of time-varying sampling points τ is expressed as: τ∈{τ 0 +i·Δt|i = 0, 1, 2…N}, and the time-varying sliding extension time window is expressed as [τ - ΔT, τ + ΔT]. By sliding the time-varying sampling points, the seismic wavelets at these time-varying sampling points are extracted within the range of the time-varying sliding extension time window, and finally the time-varying wavelets are formed.

[0058] Among them, for at least one of the one or more time-varying sliding extension time windows, perform the steps, which include:

[0059] Step 4.1: Use a window function to extract multi-channel random short-time window seismic data from at least one of the one or more time-varying sliding extension time windows;

[0060] The window function refers to the Hanning window function. For the time-varying sampling point τ i , its expression is given as follows:

[0061]

[0062] In the formula, t 1 , t 2 are the start and end times of the random short-time window respectively, t is the sampling point time, The multi-channel random short-time window refers to randomly selecting a number of channels of data within the space-variant region, randomly selecting a number of short-time windows within the time-variant sliding extension time window, and obtaining short-time window seismic data using a Hanning window function. The random number of channels and short-time windows are given by the user according to actual needs. The random number of channels is not less than 10% of the total number of channels, and the number of single-channel random short-time windows is not less than 25.

[0063] Step 4.2: Perform frequency-domain filtering and homomorphic-domain averaging on the multi-channel random short-time window seismic data to obtain homomorphic-domain wavelet data.

[0064] The homomorphic-domain wavelet data refers to, for a certain time-variant sampling point τ i , s j (t, τ i ) is the j-th random short-time window seismic data obtained in Step 4.1. s j (t, τ i ) is transformed by Fourier transform and noise reduction filtering to obtain the random short-time window seismic data in the frequency domain, denoted as S j (ω, τ i ). Then, the homomorphic-domain wavelet is obtained by the following formula:

[0065]

[0066] In the formula, is the homomorphic-domain wavelet, and N is the total number of random short-time windows.

[0067] Step 4.3: Transform the homomorphic-domain wavelet data to obtain time-domain wavelet data.

[0068] The time-domain seismic wavelet data w(t, τ i ) is obtained by the following transformation:

[0069]

[0070] In the formula, F -1 is the inverse Fourier transform.

[0071] Step 4.4: Perform wavelet shaping on the time-domain wavelet data to obtain an estimated seismic wavelet.

[0072] The wavelet shaping includes one or a combination of sidelobe suppression, spectral smoothing, and band-pass filtering, so as to obtain a stable estimated spatio-temporal variable wavelet shape.

[0073] Step 5: Combine all the estimated seismic wavelets obtained from all the time-variant sliding extension time windows in all the wavelet space-variant regions to obtain a broadband spatio-temporal variable seismic wavelet.

[0074] Example 1

[0075] In the first specific embodiment of applying the present invention, as Figure 1 shown, the broadband spatio-temporal variable seismic wavelet extraction method includes:

[0076] Step 1: Obtain seismic data and preprocess the seismic data;

[0077] For a certain actual work area, carry out broadband spatio-temporal variable wavelet extraction, obtain the post-stack seismic data in this area, and perform denoising and extract header information.

[0078] Step 2: Obtain seismic interpretation horizons from the preprocessed seismic data, and intercept a horizon time window in the seismic interpretation horizons, where the horizon time window represents the target area;

[0079] Obtain seismic interpretation horizons T0 and T1, as Figure 2 shown, and intercept a horizon time window to represent the target area.

[0080] Step 3: In the target area, delimit one or more wavelet variable regions;

[0081] According to the characteristics of the seismic data in this area, divide this area into two variable regions A and B, as Figure 3 shown; in different variable regions, extract time-varying wavelets respectively to obtain the spatio-temporal variable wavelets of the entire target area.

[0082] Step 4: In the one or more wavelet variable regions, delimit one or more time-varying sliding extension time windows;

[0083] For each variable region, in this embodiment, set the starting time-varying sampling point as T0 + 200ms, the time-varying sliding interval as 100ms, the ending time-varying sampling point as T0 + 500ms, and the extension time window length as 300ms. Then the set of time-varying sampling points τ is expressed as: τ ∈ {T0 + 200ms, T0 + 300ms, T0 + 400ms, T0 + 500ms}, and the time-varying sliding extension time window is expressed as [τ - 150ms, τ + 150ms]. By sliding the time-varying sampling points, extract the seismic wavelets of this time-varying sampling point within the time-varying sliding extension time window range, and finally form time-varying wavelets.

[0084] Among them, for at least one of the one or more time-varying sliding extension time windows, perform the steps, which include:

[0085] Step 4.1: Use a window function to extract multi-channel random short-time window seismic data from at least one of the one or more time-varying sliding extension time windows;

[0086] For a time-varying time window (e.g., the first time-varying time window [T0 + 50ms, T0 + 350ms] of the spatially varying region A), the window function refers to the Hanning window function. For the time-varying sampling point τ i , its expression is given as follows:

[0087]

[0088] where t 1 , t 2 are the start and end times of the random short time window respectively, t is the sampling point time, The multi-channel random short time window refers to randomly selecting several channels of data within the spatially varying region, randomly selecting several short time windows within the time-varying sliding extended time window, and using the Hanning window function to obtain the short time window seismic data. The random number of channels and short time windows are given by the user according to actual needs. In this embodiment, the random number of channels is set to 10% of the total number of channels, and the number of single-channel random short time windows is set to 25.

[0089] Step 4.2, perform frequency domain filtering processing and homomorphic domain averaging processing on the multi-channel random short time window seismic data to obtain homomorphic domain wavelet data;

[0090] The homomorphic domain wavelet data refers to, for a certain time-varying sampling point τ i , s j (t, τ i ) is the jth random short time window seismic data obtained in Step 4.1. s j (t, τ i ) is transformed by Fourier transform and denoising filtering to obtain the random short time window seismic data in the frequency domain, denoted as S j (ω, τ i ). Then the homomorphic domain wavelet is obtained by the following formula:

[0091]

[0092] where is the homomorphic domain wavelet, and N is the total number of random short time windows.

[0093] Step 4.3, transform the homomorphic domain wavelet data to obtain time domain wavelet data;

[0094] The time domain seismic wavelet data w(t, τ i ) is obtained by the following transformation:

[0095]

[0096] where F -1 is the inverse Fourier transform.

[0097] Step 4.4, perform wavelet shaping processing on the time-domain wavelet data to obtain an estimated seismic wavelet;

[0098] In this embodiment, the wavelet shaping processing adopted includes sidelobe suppression, spectral smoothing, and band-pass filtering, which are used to stabilize the shape of the estimated spatio-temporal variable wavelet.

[0099] In this embodiment, for each time-varying sliding extension time window in the spatially variable regions A and B, Steps 4.1 to 4.4 are repeated to obtain the estimated seismic wavelet in this spatially variable region and under this time-varying time window, and finally the wide-band spatio-temporal variable seismic wavelet for the whole area is combined. Figure 4 Shows the wide-band time-varying seismic wavelet of the spatially variable region A in this embodiment, Figure 5 Shows the wide-band time-varying seismic wavelet of the spatially variable region B in this embodiment, and the combination of the two is the wide-band spatio-temporal variable seismic wavelet for the whole area.

[0100] Step 5, combine the estimated seismic wavelets obtained from all the time-varying sliding extension time windows in all the wavelet spatially variable regions to obtain a wide-band spatio-temporal variable seismic wavelet.

[0101] Embodiment Two

[0102] In a specific Embodiment Two of applying the present invention, with reference to Figure 1 , the flowchart of the wide-band spatio-temporal variable seismic wavelet extraction method includes Steps 1 to 5. For a certain actual work area, wide-band time-varying wavelet extraction is carried out at different angles. In Step 1, near, medium, and far angle stack seismic data of this area are obtained, and near, medium, and far wide-band time-varying seismic wavelets are respectively extracted.

[0103] In Step 2, seismic interpretation horizons at the top and bottom of the target layer are obtained, and a layer time window is intercepted to represent the target area.

[0104] In Step 3, because in this embodiment, only time-varying seismic wavelets are extracted in this area, the whole area is divided into a single spatially variable region.

[0105] In Step 4, the starting time-varying sampling point is set as τ 0 , the time-varying sliding interval is 50 ms, the ending time-varying sampling point is τ 0 + 800 ms, and the extension time window length is 300 ms. Then the set of time-varying sampling points τ is expressed as: τ ∈ {τ 0 + i·50 (unit: ms)|i = 0, 1, 2…16}, and the time-varying sliding extension time window is expressed as [τ - 150, τ + 150] (unit: ms). By sliding the time-varying sampling points, the seismic wavelets at this time-varying sampling point are extracted within the time-varying sliding extension time window, and finally the time-varying wavelets are formed.

[0106] In step 4.1, a random short-time window seismic data is obtained using a Hanning window function, where the random number of traces and the number of short-time windows are set to 10% of the total number of traces and 50 respectively in this embodiment.

[0107] In step 4.2, for a certain time-varying sampling point τ i , s j (t, τ i ) is the j-th random short-time window seismic data obtained in step 4.1. s j (t, τ i ) is Fourier-transformed and noise-reducing filtered to obtain the random short-time window seismic data in the frequency domain, denoted as S j (ω, τ i ). Then, the homomorphic domain wavelet is obtained by the following formula:

[0108] In the formula, is the homomorphic domain wavelet, and N is the total number of random short-time windows.

[0109] In step 4.3, the time-domain seismic wavelet data w(t, τ i ) is obtained by the following transformation:

[0110]

[0111] In the formula, F -1 is the inverse Fourier transform.

[0112] In step 4.4, the wavelet shaping process adopted in this embodiment includes sidelobe suppression and band-pass filtering, so that the spatiotemporal variable wavelet morphology estimated stably, Figure 6 shows the broadband seismic wavelet extracted from a certain time-varying time window in this embodiment and its spectrogram.

[0113] In step 5, the estimated seismic wavelets obtained from all the time-varying sliding extended time windows are combined to obtain broadband time-varying seismic wavelets at near, medium, and far angles.

[0114] Embodiment III

[0115] In a specific Embodiment III of applying the present invention, referring to Figure 1 , the flowchart of the broadband spatiotemporal variable seismic wavelet extraction method includes steps 1-step 5. For a certain actual work area, broadband spatiotemporal variable wavelet extraction is carried out by offset. In step 1, the near, medium, and far offset stacked seismic data of this area are obtained, and the broadband spatiotemporal variable seismic wavelets at near, medium, and far offsets are extracted respectively.

[0116] In step 2, the seismic interpretation horizons N1sT and N1sB are obtained, and the horizon time window is intercepted to represent the target area.

[0117] In step 3, according to the geological characteristics and seismic data characteristics of this area, the area is divided into three variable-offset regions A, B, and C, as shown in Figure 7 ; in different variable-offset regions, time-varying wavelets are extracted respectively to obtain the spatio-temporal variable wavelets of the entire target area.

[0118] In step 4, for each variable-offset region, the starting time-varying sampling point is set as N1sT, the time-varying sliding interval is 100 ms, the ending time-varying sampling point is N1sT + 300 ms, and the extension time window length is 200 ms. Then the set of time-varying sampling points τ is expressed as: τ ∈ {N1sT, N1sT + 100 ms, N1sT + 200 ms, N1sT + 300 ms}, and the time-varying sliding extension time window is expressed as [τ - 100 ms, τ + 100 ms]. By sliding the time-varying sampling points, the seismic wavelets at this time-varying sampling point are extracted within the range of the time-varying sliding extension time window, and finally the time-varying wavelets are formed.

[0119] In step 4.1, for a certain time-varying time window (for example, the second time-varying time window in variable-offset region C), the Hanning window function is used to obtain the random short-time window seismic data. Among them, the random number of channels and the number of short-time windows are set to 20% of the total number of channels and 100 respectively in this embodiment.

[0120] In step 4.2, for a certain time-varying sampling point τ i , s j (t, τ i ) is the j-th random short-time window seismic data obtained in step 4.1. s j (t, τ i ) is transformed by Fourier transform and denoised and filtered to obtain the random short-time window seismic data in the frequency domain, which is expressed as S j (ω, τ i ). Then the homomorphic domain wavelet is obtained by the following formula:

[0121] In the formula, is the homomorphic domain wavelet, and N is the total number of random short-time windows.

[0122] In step 4.3, the time-domain seismic wavelet data w(t, τ i ) is obtained by the following transformation:

[0123]

[0124] In the formula, F -1 is the inverse Fourier transform.

[0125] In step 4.4, the wavelet shaping process adopted in this embodiment is sidelobe suppression to obtain a stable estimated spatio-temporal variable wavelet form. For each time-varying sliding extended time window in the spatially variable region A, the spatially variable region B, and the spatially variable region C, steps 4.1 to 4.4 are repeated to obtain the estimated seismic wavelet in the spatially variable region and under the time-varying time window.

[0126] In step 5, the estimated seismic wavelets obtained from the time-varying sliding extended time windows in all the spatially variable regions are combined to obtain broadband time-varying seismic wavelets with near, medium, and far offsets. Figure 8 The broadband time-varying seismic wavelet and its spectrogram in the spatially variable region C with a near offset in this embodiment are shown.

[0127] The foregoing are various preferred embodiments of the present invention. If the preferred implementation manners in each preferred embodiment are not obviously self-contradictory or premised on a certain preferred implementation manner, each preferred implementation manner can be arbitrarily superimposed and combined for use. The embodiments and the specific parameters in the embodiments are only for clearly expressing the inventor's invention verification process and are not used to limit the patent protection scope of the present invention. The patent protection scope of the present invention still takes its claims as the criterion. Any equivalent structural changes made by using the content of the specification and drawings of the present invention should similarly be included in the protection scope of the present invention.

[0128] Except for the technical features described in the specification, they are all known technologies to those skilled in the art.

Claims

1. Wideband spatio-temporal variable seismic wavelet extraction method, Characterized in that, The wideband spatio-temporal variable seismic wavelet extraction method includes: Step 1, obtain seismic data and preprocess the seismic data; Step 2, obtain seismic interpretation horizons from the preprocessed seismic data, and intercept horizon time windows in the seismic interpretation horizons, where the horizon time windows represent the target area; Step 3, in the target area, delimit one or more wavelet variable areas; Step 4, in one or more wavelet variable areas, delimit one or more time-varying sliding extension time windows; Step 5, in all wavelet variable areas, combine the estimated seismic wavelets obtained from all time-varying sliding extension time windows to obtain a wideband spatio-temporal variable seismic wavelet; In step 4, the time-varying sliding extended time window means that for each empty-varying region, assuming the starting time-varying sampling point is τ 0 , the time-varying sliding interval is Δt, the number of samplings is N, and the ending time-varying sampling point is τ 0 +N·Δt. If the length of the extended time window is 2ΔT, then the set of time-varying sampling points τ is expressed as: τ∈{τ 0 +i·Δt|i = 0, 1, 2…N}. The time-varying sliding extended time window is expressed as [τ - ΔT, τ + ΔT]. By sliding the time-varying sampling points, the seismic wavelet of the time-varying sampling point is extracted within the range of the time-varying sliding extended time window, and finally a time-varying wavelet is formed; In Step 4, for at least one of the one or more time-varying sliding extension time windows, perform the following steps: Step 4.1, use a window function to extract multi-channel random short-time window seismic data from at least one of the one or more time-varying sliding extension time windows; Step 4.2, for the multi-channel random short-time window seismic data, perform frequency domain filtering processing and homomorphic domain averaging processing to obtain homomorphic domain wavelet data; Step 4.3, transform the homomorphic domain wavelet data to obtain time domain wavelet data; Step 4.4, perform wavelet shaping processing on the time domain wavelet data to obtain an estimated seismic wavelet; In step 4.2, the homomorphic domain wavelet data refers to a time-varying sampling point τ i , s j (t, τ i ) is the j-th random short-time window seismic data obtained in step 4.1, and s j (t, τ i ) is transformed by Fourier transform and denoised by filtering to obtain the random short-time window seismic data in the frequency domain, denoted as S j (ω, τ i ). Then the homomorphic domain wavelet is obtained by the following formula: In the formula, is the homomorphic domain wavelet, and N is the total number of random short-time windows; In step 4.3, the time-domain seismic wavelet data w(t, τ i ) is obtained by the following transformation: where F -1 is the inverse Fourier transform.

2. The wideband spatio-temporal variable seismic wavelet extraction method according to claim 1, Characterized in that, In Step 1, the seismic data is post-stack seismic data, which can be full-stack post-stack seismic data, or can be offset-stack post-stack seismic data, angle-stack seismic data or azimuth-stack seismic data; the preprocessing includes denoising and extracting header information.

3. The wideband spatio-temporal variable seismic wavelet extraction method according to claim 1, Characterized in that, In Step 3, the wavelet variable area refers to dividing the wavelet variable area according to geological characteristics, seismic data acquisition and processing conditions and research needs. If only time-varying wavelets are extracted, the target area is divided into a single variable area; in different variable areas, time-varying wavelets are extracted separately to obtain the spatio-temporal variable wavelets of the entire target area.

4. The wideband spatio-temporal variable seismic wavelet extraction method according to claim 1, Characterized in that, In step 4.1, the window function refers to the Hanning window function Win(t), for the time-varying sampling point τ 1 , its expression is given as follows: where t 1 and t 2 are the start and end times of a random short time window respectively, t is the sampling point time, 5. The wideband spatio-temporal variable seismic wavelet extraction method according to claim 4, Characterized in that, In Step 4.1, the multi-channel random short-time window means randomly selecting several channels of data in the variable area, randomly selecting several short-time windows within the range of the time-varying sliding extension time window, and using a Hanning window function to obtain short-time window seismic data, where the random number of channels and the number of short-time windows are given by the user according to actual needs, the random number of channels is not less than 10% of the total number of channels, and the number of single-channel random short-time windows is not less than 25.

6. The wideband spatio-temporal variable seismic wavelet extraction method according to claim 1, Characterized in that, In Step 4.4, the wavelet shaping processing includes one or a combination of sidelobe suppression, spectrum smoothing, and band-pass filtering to stabilize the shape of the estimated spatio-temporal variable wavelet.

Images