Frequency domain gather construction method based on post-stack seismic data

Abstract

The invention relates to a frequency domain gather construction method based on post-stack seismic data, which belongs to the technical field of oil-gas exploration geophysics, and comprises the following steps: S1, post-stack seismic data trace header information analysis, loading and data preprocessing; s2, determining an effective frequency interval of a target layer section and designing a broadband center frequency energy focusing filter; s3, performing frequency division processing on the pre-processed post-stack seismic data by using a broadband center frequency energy focusing filter to generate a single frequency body; s4, defining a center frequency recording position, and reconstructing a frequency domain gather header parameter; s5, combining the single frequency bodies according to a center frequency sequence to form a frequency domain gather; and S6, extracting the characteristics of the amplitude changing along with the frequency, and quantifying the change rule of the amplitude along with the frequency. According to the method, post-stack data frequency division, trace header reconstruction and trace gather combination are carried out, systematic representation of characteristics of seismic amplitude changing along with frequency is achieved, and the data processing efficiency and the understanding and explaining degree of the data processing efficiency are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of geophysical technology for oil and gas exploration, and in particular to a method for constructing frequency domain gathers based on post-stack seismic data. Background Technology

[0002] In oil and gas exploration, seismic gathers are a type of data that characterizes the changes in seismic response with certain parameters. Among them, AVO (amplitude variation with offset) gathers use the relationship between amplitude and offset to identify the fluidity of reservoirs. However, physical properties of the subsurface medium, such as lithology, porosity, and fluid properties, can affect the offset characteristics of seismic wave amplitude and have a stronger effect on the absorption and attenuation of seismic wave frequency. For example, fluid-bearing reservoirs will exhibit the characteristics of enhanced low-frequency amplitude and attenuated high-frequency amplitude in the frequency domain. This type of frequency domain response cannot be directly observed in conventional post-stack seismic data.

[0003] Currently, research on seismic data frequencies mainly focuses on quantitative studies of single frequency attributes or methods for local time-frequency analysis. There is a lack of a method like the AVO gather that can perform continuous and multi-faceted quantitative analysis in the frequency domain. Furthermore, in practice, frequency division may lead to overlapping frequency bandwidths and inconsistent single-frequency volume amplitudes, which are detrimental to using frequency domain information for reservoir prediction. Summary of the Invention

[0004] To address the aforementioned problems in the prior art, this invention provides a method for constructing frequency domain gathers based on post-stack seismic data.

[0005] A method for constructing frequency domain gathers based on post-stack seismic data includes the following steps: Step S1: Parsing, loading, and preprocessing of post-stack seismic data trace head information; Step S2: Determining the effective frequency range of the target layer and designing a wideband center frequency energy focusing filter; Step S3: Use a broadband center frequency energy focusing filter to perform frequency division processing on the preprocessed post-stack seismic data to generate a single-frequency volume; Step S4: Define the center frequency recording position and reconstruct the frequency domain gather head parameters; Step S5: Combine the single-frequency units in order of their center frequencies to form a frequency domain gather; Step S6: Extract the amplitude variation characteristics with frequency and quantify the amplitude variation law with frequency.

[0006] Furthermore, step S1 specifically includes: Post-stack seismic data are stored in the standard SEGY format and analyzed to obtain key spatial and temporal reference information, including main survey line number, connecting survey line number, spatial coordinates, start time of temporal sampling points, total number of sampling points, and sampling interval. After analysis, the data is loaded using seismic interpretation or inversion industrial software to extract the three-dimensional mesh parameters of the post-stack seismic data; After loading is complete, data validity checks, labeling, and data preprocessing are performed.

[0007] Furthermore, the validity check and marking specifically involve: checking and identifying the track head, identifying empty tracks and bad tracks and marking them; the marked data will not participate in subsequent processing; the data preprocessing includes multi-domain joint denoising, specifically: using adaptive median filtering in the time domain to remove random noise, using tilt filtering in the fk domain to remove linear interference, and using bandpass filtering in the frequency domain to remove low-frequency and high-frequency noise from the interference wave.

[0008] Furthermore, step S2 specifically includes: The seismic trace spectrum of the target layer is extracted from the preprocessed data, and the peak distribution of the spectrum is calculated using short-time Fourier transform. Design a broadband center-frequency energy focusing filter with low-cutoff, low-pass, high-pass, and high-cutoff frequencies of [f]. min -5,f center , f center , f max +5], f center Indicates the center frequency.

[0009] Furthermore, step S3 specifically includes: based on the MPI parallel framework, dividing the three-dimensional data volume into blocks according to the Line direction, and performing frequency division processing of filtering channel by channel using a wideband center frequency energy focusing filter to obtain multiple single-frequency volumes.

[0010] Furthermore, the frequency division process specifically involves: convolving each seismic trace with a Gaussian window operator in the time domain, converting it to the frequency domain using a fast Fourier transform, extracting the frequency band energy corresponding to the center frequency, and then converting it back to the time domain using an inverse Fourier transform to obtain single-channel data; traversing all traces to generate a single-frequency volume.

[0011] Furthermore, step S4 specifically includes: Key spatial and temporal reference information is inherited from the trace header of the original post-stack seismic data, and a center frequency item indicating the single-frequency volume is added to a specified byte in the trace header. Export the generated single-frequency body into a track header editing script. While reading the data files of each single-frequency body in a loop, extract the center frequency from the file name and modify the specified byte position of the track header to the center frequency value.

[0012] Furthermore, the specified byte is an empty byte or an unused byte, specifically the 37th byte.

[0013] Furthermore, step S5 specifically includes: Using the original post-stack data volume's three-dimensional spatial grid as a reference, all single-frequency volumes are spatially resampled; by applying a bilinear interpolation algorithm to each single-frequency volume, the horizontal and vertical directions of each single-frequency volume are made to perfectly conform to the reference grid. For each CDP point, the seismic traces at the corresponding location are extracted from all single-frequency volumes that have completed spatial alignment, and a trace gather element composed of multi-frequency information is constructed. Each trace gather element contains multiple seismic traces from low frequency to high frequency, sorted in ascending order of the center frequency value. After the gather elements corresponding to different spatial locations are sorted and placed together according to the original Line-Trac, a data volume with a four-dimensional structure of main survey line, connecting survey line, frequency, and time is formed. The gather elements of all spatial locations are organized and stored using the extended SEGY format, and the seismic traces of different frequencies on the same CDP point are distinguished by the frequency values ​​stored in the specified bytes of the trace header.

[0014] Furthermore, step S6 specifically includes: First, for each seismic trace with a center frequency of f in the trace element, the instantaneous amplitude envelope is obtained by applying Hilbert transform in the time domain to eliminate the phase influence of the seismic wavelet and obtain the envelope amplitude sequence E(f,t) that truly represents the energy magnitude, where f is the center frequency and t is the time sampling point. Then, the envelope amplitude sequence of the frequency channels is standardized to represent the relative change of amplitude with frequency. Specifically, the envelope amplitude value corresponding to the lowest center frequency in the channel set is selected as a reference, and the envelope amplitude values ​​corresponding to all frequency channels at the same sampling point are divided to calculate the relative amplitude ratio R(f,t) = E(f,t) / E(f min ,t); Finally, the relative amplitude ratio is modeled to obtain quantifiable AFV properties. For each time sampling point t, the R(f,t) values ​​of all frequencies thereunder are an observation sample group. The least squares method is used to perform linear fitting to obtain the slope k(t) of the straight line, which is used to characterize the rate of change of amplitude with respect to frequency. The fitting intercept b(t) contains a certain amount of background value. Finally, the k(t) and b(t) calculated at each time point of the entire data volume are summarized into a new feature volume.

[0015] The beneficial effects of this invention are reflected in: (1) By frequency division, the accuracy and availability of data are improved. By using a wideband filter, the energy of the main frequency is focused and the information in the entire frequency band is retained, which solves the problems of narrow frequency band and serious Gibbs effect in conventional frequency division technology; (2) A new frequency domain gather architecture standard is proposed. The frequency domain structured four-dimensional gather is innovatively constructed to continuously convert the change of seismic amplitude with frequency into gather representation, so that the frequency domain information is concentrated and can realize the quantitative study of the system, and realize the application of frequency attributes in reservoir research; (3) By defining the center frequency information in a specific byte in the SEGY header, the data body of multiple frequencies is standardized and can be organically combined, which can achieve compatibility with mainstream industrial software, greatly improving the efficiency of data processing and the degree of understanding and interpretation. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method of the present invention; Figure 2 The standard roadway information represents the intended meaning, in which Figure 2 'a' represents 5-77 bytes of information. Figure 2 b represents 103-117 bytes of information; Figure 3 A three-dimensional display of post-stack seismic data; Figure 4 Preprocessed post-stack seismic data profiles and target segment maps; Figure 5 For the spectrum diagram; Figure 6 This is a schematic diagram of the amplitude-frequency response curve of a conventional frequency divider filter. Figure 6 The center frequency of a is 20Hz. Figure 6 The center frequency of b is 60Hz; Figure 7 This is a schematic diagram of the amplitude-frequency response curve of a broadband center-frequency energy focusing filter. Figure 7 The center frequency of a is 20Hz. Figure 7 b. The center frequency is 60Hz; Figure 8 This is the original seismic data profile; Figure 9 This is a profile of seismic data obtained by filtering with a conventional frequency division filter. Figure 9 The center frequency of a is 20Hz. Figure 9 The center frequency of b is 60Hz; Figure 10 This is a seismic data profile obtained by filtering with a broadband center frequency energy focusing filter. Figure 10 The center frequency of a is 20Hz. Figure 10 b. The center frequency is 60Hz; Figure 11 A schematic diagram of frequency domain gather head parameters; Figure 12 This is a schematic diagram of a frequency domain gather. Figure 13 This is a characteristic graph of amplitude versus frequency (AFV). Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] In this embodiment: Refer to Figure 1 As shown, an embodiment of the present invention provides a method for constructing frequency domain gathers based on post-stack seismic data, comprising the following steps: S1: Post-stack seismic data header information parsing, loading, and data preprocessing, specifically: Head information was analyzed from post-stack seismic data (standard SEGY format). Figure 2 The standard trace information table shows that: the main line number, connecting line trace number, X coordinate, and Y coordinate of the post-stack seismic data are recorded in 4 bytes, with the main line number recorded in bytes 9-12 and the connecting line trace number recorded in bytes 21-25. The X coordinate is recorded in bytes 73-76 and the Y coordinate in bytes 77-80. The start time, total number of time sampling points, and sampling interval of the post-stack seismic data are recorded in 2 bytes, with the start time recorded in bytes 105-106, the total number of time sampling points recorded in bytes 115-116, and the sampling interval recorded in bytes 117-118.

[0019] After completing the analysis of the post-stack seismic data trace head information, the data is loaded using seismic interpretation or inversion industrial software to extract the three-dimensional mesh parameters of the post-stack seismic data. Figure 3 The three-dimensional display results of the post-stack seismic data show that: Line range 1~142, Trace range 1~110, and time sampling Z range 500~1500ms.

[0020] After the post-stack seismic data is loaded, the seismic data undergoes data validity checks, labeling, and data preprocessing.

[0021] Regarding data validity checks and labeling, empty channels and bad channels are first identified and labeled (empty channels are defined as having an amplitude value of 0, and bad channels are defined as having an amplitude value that exceeds 10 times the range of the entire data volume). After identifying empty channels and bad channels, these channels are labeled and do not participate in subsequent data preprocessing and frequency domain channel set construction, etc.

[0022] Data preprocessing includes noise removal using multi-domain joint denoising techniques: in the time domain, adaptive median filtering is used to remove random noise; in the fk domain, tilt filtering (with a filtering threshold of 1.2 times the apparent velocity of the surface wave) is used to remove linear interference (mainly surface waves); and in the frequency domain, bandpass filtering is used to remove low-frequency and high-frequency noise from the interference wave (the filtering parameters are selected based on the spectrum scanning results).

[0023] S2: Determination of the effective frequency range of the target segment and design of the broadband center frequency energy focusing filter, specifically: Extract the target segment from the preprocessed data ( Figure 4 The seismic trace spectrum from the top to the bottom boundary of the middle reservoir (time range T1~T3) was calculated, with the time window length set to approximately three times the target segment. Then, the peak distribution of the spectrum was calculated using Short-Time Fourier Transform (STFT), as shown below. Figure 5 As shown, with an amplitude of -24dB, the effective frequency range is determined. f min ~ f max [6~80Hz].

[0024] Conventional frequency divider filters typically select a range of 10Hz around the center frequency for filtering. The filter can be represented as: low-cutoff, low-pass, high-pass, and high-cutoff frequencies respectively […]. f center -5Hz, f center , f center , f center +5Hz], f center The center frequency. Figure 6 a and Figure 6 b represents conventional frequency divider filters with center frequencies of 20Hz and 60Hz, respectively. To overcome the narrow bandwidth limitation of conventional frequency divider filters, this invention redesigns a wideband center frequency energy focusing filter: low cutoff, low-pass, high-pass, and high-cutoff frequencies are respectively […]. f min -5, f center , f center , f max+5], for example, wideband center frequency energy focusing filters with center frequencies of 20Hz and 60Hz are [1, 20, 20, 85] and [1, 60, 60, 85], respectively, as Figure 7 a and Figure 7 As shown in b. The advantage of the wideband center frequency energy focusing filter is that the main frequency is focused to the center frequency, which achieves the purpose of frequency division, but at the same time each filter retains the full-band information, overcoming the narrow bandwidth defect of conventional frequency division technology.

[0025] like Figure 8 The image shows the original seismic data profile. Figure 9 a and Figure 9 b represents seismic data profiles obtained by conventional frequency division filters with center frequencies of 20Hz and 60Hz, respectively. It can be seen that the conventional frequency division filter produces a narrow-band single-frequency volume. This results in a significant Gibbs effect, and the seismic data is severely distorted at particularly low or high frequencies, differing greatly from the original seismic data. Figure 10 a and Figure 10 b represents the seismic data profiles obtained by filtering with broadband center frequency energy focusing filters at center frequencies of 20Hz and 60Hz, respectively. It can be seen that the profiles obtained by filtering at 20Hz and 60Hz are relatively close to the original seismic data profiles, but they can still reflect both low-frequency and high-frequency information.

[0026] S3: The preprocessed post-stack seismic data is frequency-divided using a broadband center frequency energy focusing filter to generate a single-frequency volume. Specifically: Based on the MPI parallel framework, the 3D data volume is divided into blocks along the Line direction, and frequency division processing based on a broadband center frequency energy focusing filter is performed on each trace: for each seismic trace, convolution is performed in the time domain with a Gaussian window operator, and then converted to the frequency domain by a Fast Fourier Transform. The energy of the frequency band corresponding to the center frequency is extracted, and then converted back to the time domain by an Inverse Fourier Transform to obtain single-channel data; all traces are traversed to generate a single-frequency volume. From the previous step, the effective frequency range is known […]. f min ~ f max The frequency range is [6~80Hz]. Single-frequency bodies with center frequencies of 6Hz, 10Hz, 20Hz, 30Hz, 40Hz, 50Hz, 60Hz, 70Hz, and 80Hz are generated respectively. In order to make the most of the effective information, the frequency is extended outward by 10Hz to form an additional single-frequency body with a frequency of 90Hz, thus obtaining 10 single-frequency bodies.

[0027] For each single-frequency body, 1000 seismic traces are randomly selected, and its actual center frequency is calculated using Fast Fourier Transform, with a deviation from the design value of <1Hz. Five typical profiles (including faults and stratigraphic interfaces) are selected, and the original data is compared with the single-frequency body superposition results to ensure that there is no distortion in the geological structure (phase axis offset <1 sampling point). The proportion of clutter energy in the single-frequency body is statistically analyzed (calculated by energy spectrum integration), with a requirement of <10%.

[0028] S4: Define the center frequency recording position and reconstruct the frequency domain gather head parameters, specifically: After generating a single-frequency body, each multi-frequency single-frequency body needs to be systematically combined into a frequency domain gather according to certain rules in order to complete the joint analysis of multiple frequencies in the frequency domain. Therefore, it is necessary to standardize and unify the data of each single-frequency body's head, use unified head parameters, ensure that the frequency points correspond one-to-one, and ensure that the spatial positions of the head completely overlap.

[0029] First, key spatial and temporal reference information, such as the main line number, trace number, spatial coordinates (X, Y), time sampling start time, total number of sampling points, and sampling interval, is inherited from the trace header of the original post-stack seismic data. The storage location and method of these parameters in the original trace header remain unchanged, ensuring compatibility with the existing seismic interpretation system and providing a basis for subsequent spatial alignment. To distinguish the frequency information corresponding to different single-frequency bodies, a center frequency item indicating the single-frequency body is added to the trace header. This invention uses the 37th byte of the trace header (originally an empty or unused byte) as the "center frequency (Hz)" and records it as an integer. For example, for a single-frequency body with a center frequency of 6Hz, the storage value of the 37th byte of the trace header is "6" (…). Figure 11 As shown in the figure, after each seismic trace is loaded, its frequency can be directly identified from the trace head.

[0030] The generated 10 single-frequency cells (center frequencies of 6Hz, 10Hz, 20Hz, 30Hz, 40Hz, 50Hz, 60Hz, 70Hz, 80Hz, and 90Hz) are exported as a trackhead editing script. When reading the data files of each single-frequency cell in a loop, the center frequency is extracted from the file name, and the 37th byte of the trackhead is modified to the center frequency value. When there are multiple input files, the trackhead parallel processing method is used to improve processing efficiency.

[0031] S5: Combine the single-frequency units in order of their center frequencies to form frequency domain gathers, specifically: Based on the 3D mesh of the original post-stack data volume, spatial interpolation is performed on each single-frequency volume. After generating each single-frequency volume and reconstructing the standardized trace head, it is integrated into a complete set of frequency domain gathers, such as... Figure 12 As shown.

[0032] First, based on the original post-stack data volume's three-dimensional spatial grid (Line range, Trace range, cell size, and time sampling rate) as a reference, all single-frequency volumes are spatially resampled. By using a bilinear interpolation algorithm on each single-frequency volume, the horizontal and vertical directions of each single-frequency volume are made to fully conform to the reference grid, eliminating the subtle spatial errors caused by frequency division and ensuring that data from different frequencies under the same ground rectangular coordinates reflect the position and frequency response of the same underground point.

[0033] Then, within each CDP, seismic traces at corresponding locations are extracted from all spatially aligned single-frequency volumes, constructing a "trace gather element" composed of multi-frequency information. Each "trace gather element" contains 10 seismic traces ranging from low frequency (6Hz) to high frequency (90Hz), sorted in ascending order of their center frequency values. After the "trace gather elements" corresponding to different spatial locations are sorted according to their original Line-Trac order and placed together, a data volume with a four-dimensional structure of main line, trace, frequency, and time is formed. Each trace gather element is organized in a strict frequency order, thus forming a clear frequency coordinate axis. The data organization method in this file is a one-dimensional time series + three-dimensional gridded spatial organization method. The trace gather elements of all spatial locations are organized and stored using the extended SEGY format. Seismic traces of different frequencies at the same CDP point, distinguished by the frequency value stored in the 37th byte of the trace header, can be directly recognized and loaded by commonly used geophysical software such as Petrel and Landmark, and interpreted and analyzed with frequency as an independent dimension.

[0034] S6: Extract amplitude versus frequency (AFV) features and quantify the amplitude variation with frequency. Specifically: Extracting quantitative amplitude variations with frequency from trace assemblies in the frequency domain is crucial for identifying reservoir fluids and lithology. For each trace element, the frequency is... f The instantaneous amplitude envelope of the seismic trace corresponding to the center frequency is obtained by applying Hilbert transform in the time domain. This eliminates the phase influence of the seismic wavelet, resulting in an envelope amplitude sequence that truly represents the energy magnitude. E ( f , t ),in f For the center frequency, t These are the time sampling points. This transformation can convert the time amplitude changes of each frequency channel into a relatively stable energy curve, providing a consistent basis for quantitative comparison between frequencies.

[0035] Then, the envelope amplitude sequence of the frequency channels is standardized to represent the relative change of amplitude with frequency. Specifically, the envelope amplitude value corresponding to the lowest center frequency in the channel set is selected as a reference, and the envelope amplitude values ​​corresponding to all frequency channels at the same sampling point are divided to calculate the relative amplitude ratio. R ( f , t )= E ( f , t ) / E ( f min , t ).

[0036] Finally, the relative amplitude ratio can be modeled to obtain quantifiable AFV properties. For each time sampling point t All frequencies below it R ( f , t The value is taken from an observed sample group. A linear fit is performed using the least squares method, and the slope of the resulting straight line is... k ( t The slope is an important parameter that characterizes the rate of change of amplitude with respect to frequency: a positive slope indicates that the amplitude increases with increasing frequency (high-frequency energy is relatively enriched); a negative slope indicates that the amplitude decreases with increasing frequency (high-frequency energy is relatively attenuated). The fitting intercept... b ( t The dataset contains a certain amount of background values, and the final result is calculated from the entire dataset at each time point. k ( t )and b ( t ) are summarized into a new feature body (e.g. Figure 13 (As shown).

[0037] In the description of embodiments of the present invention, the terms "first," "second," "third," and "fourth" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first," "second," "third," or "fourth" may explicitly or implicitly include one or more of that feature. In the description of the present invention, unless otherwise stated, "a plurality of" means two or more.

[0038] In the description of embodiments of the present invention, the term "and / or" is used only to describe the relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Additionally, the character " / " generally indicates that the preceding and following associated objects are in an "or" relationship.

[0039] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for constructing frequency domain gathers based on post-stack seismic data, characterized in that, Includes the following steps: Step S1: Parsing, loading, and preprocessing of post-stack seismic data trace head information; Step S2: Determining the effective frequency range of the target layer and designing a wideband center frequency energy focusing filter; Step S3: Use a broadband center frequency energy focusing filter to perform frequency division processing on the preprocessed post-stack seismic data to generate a single-frequency volume; Step S4: Define the center frequency recording position and reconstruct the frequency domain gather head parameters; Step S5: Combine the single-frequency units in order of their center frequencies to form a frequency domain gather; Step S6: Extract the amplitude variation characteristics with frequency and quantify the amplitude variation law with frequency.

2. The method for constructing frequency domain gathers based on post-stack seismic data according to claim 1, characterized in that, Step S1 specifically includes: Post-stack seismic data are stored in the standard SEGY format and analyzed to obtain key spatial and temporal reference information, including main survey line number, connecting survey line number, spatial coordinates, start time of temporal sampling points, total number of sampling points, and sampling interval. After analysis, the data is loaded using seismic interpretation or inversion industrial software to extract the three-dimensional mesh parameters of the post-stack seismic data; After loading is complete, data validity checks, labeling, and data preprocessing are performed.

3. The method for constructing frequency domain gathers based on post-stack seismic data according to claim 2, characterized in that, The validity check and marking specifically involve: checking and identifying the track head, identifying empty tracks and bad tracks and marking them; the marked data will not participate in subsequent processing; the data preprocessing includes multi-domain joint denoising, specifically: using adaptive median filtering to remove random noise in the time domain, using tilt filtering to remove linear interference in the fk domain, and using bandpass filtering to remove low-frequency and high-frequency noise from the interference wave in the frequency domain.

4. The method for constructing frequency domain gathers based on post-stack seismic data according to claim 1, characterized in that, Step S2 specifically includes: The seismic trace spectrum of the target layer is extracted from the preprocessed data, and the peak distribution of the spectrum is calculated using short-time Fourier transform. Design a broadband center-frequency energy focusing filter with low-cutoff, low-pass, high-pass, and high-cutoff frequencies of [f]. min -5, f center ,f center , f max +5], f center Indicates the center frequency.

5. The method for constructing frequency domain gathers based on post-stack seismic data according to claim 1, characterized in that, Step S3 specifically includes: based on the MPI parallel framework, the three-dimensional data volume is divided into blocks according to the Line direction, and a wideband center frequency energy focusing filter is used to perform frequency division processing for filtering channel by channel to obtain multiple single-frequency volumes.

6. The method for constructing frequency domain gathers based on post-stack seismic data according to claim 5, characterized in that, The frequency division process specifically involves: convolving each seismic trace with a Gaussian window operator in the time domain, converting it to the frequency domain using a fast Fourier transform, extracting the energy of the frequency band corresponding to the center frequency, and then converting it back to the time domain using an inverse Fourier transform to obtain single-channel data; traversing all traces to generate a single-frequency volume.

7. The method for constructing frequency domain gathers based on post-stack seismic data according to claim 1, characterized in that, Step S4 specifically includes: Key spatial and temporal reference information is inherited from the trace header of the original post-stack seismic data, and a center frequency item indicating the single-frequency volume is added to a specified byte in the trace header. Export the generated single-frequency body into a track header editing script. While reading the data files of each single-frequency body in a loop, extract the center frequency from the file name and modify the specified byte position of the track header to the center frequency value.

8. The method for constructing frequency domain gathers based on post-stack seismic data according to claim 1, characterized in that, The specified byte is an empty byte or an unused byte, specifically the 37th byte.

9. A method for constructing frequency domain gathers based on post-stack seismic data according to claim 1, characterized in that, Step S5 specifically includes: Using the original post-stack data volume's three-dimensional spatial grid as a reference, all single-frequency volumes are spatially resampled; by applying a bilinear interpolation algorithm to each single-frequency volume, the horizontal and vertical directions of each single-frequency volume are made to perfectly conform to the reference grid. For each CDP point, the seismic traces at the corresponding location are extracted from all single-frequency volumes that have completed spatial alignment, and a trace gather element composed of multi-frequency information is constructed. Each trace gather element contains multiple seismic traces from low frequency to high frequency, sorted in ascending order of the center frequency value. After the gather elements corresponding to different spatial locations are sorted and placed together according to the original Line-Trac, a data volume with a four-dimensional structure of main survey line, connecting survey line, frequency, and time is formed. The gather elements of all spatial locations are organized and stored using the extended SEGY format, and the seismic traces of different frequencies on the same CDP point are distinguished by the frequency values ​​stored in the specified bytes of the trace header.

10. A method for constructing frequency domain gathers based on post-stack seismic data according to claim 1, characterized in that, Step S6 specifically includes: First, for each seismic trace with a center frequency of f in each trace element, the instantaneous amplitude envelope is obtained by applying Hilbert transform in the time domain. This eliminates the phase influence of the seismic wavelet and obtains the envelope amplitude sequence E(f,t) that truly represents the energy magnitude, where f is the center frequency and t is the time sampling point. Then, the envelope amplitude sequence of the frequency channels is standardized to represent the relative change of amplitude with frequency. Specifically, the envelope amplitude value corresponding to the lowest center frequency in the channel set is selected as a reference, and the envelope amplitude values ​​corresponding to all frequency channels at the same sampling point are divided to calculate the relative amplitude ratio R(f,t) = E(f,t) / E(f min ,t); Finally, the relative amplitude ratio is modeled to obtain quantifiable AFV properties. For each time sampling point t, the R(f,t) values ​​of all frequencies thereunder are an observation sample group. The least squares method is used to perform linear fitting to obtain the slope k(t) of the straight line, which is used to characterize the rate of change of amplitude with respect to frequency. The fitting intercept b(t) contains a certain amount of background value. Finally, the k(t) and b(t) calculated at each time point of the entire data volume are summarized into a new feature volume.

Images