Method for extracting sea surface reflection coefficient based on air gun source far-field observation record

Abstract

The invention discloses a method for extracting a sea surface reflection coefficient based on an air gun focus far-field observation record, a far-field hydrophone is arranged near the seabed and is used for recording a focus wavelet emitted by an air gun, the focus wavelet recorded by the far-field hydrophone is a wave field superposition result of a first broadcast and a ghost wave, and a trend fitting line is constructed to obtain the sea surface reflection coefficient. A head wave and a ghost wave are separated from an original seismic source wavelet, and then a sea surface reflection coefficient is obtained through inversion solution by using head wave and ghost wave expressions. According to the invention, the accurate sea surface reflection coefficient can be obtained.

Description

Technical Field

[0001] The present invention belongs to the technical field of marine seismic exploration, and particularly relates to a method for extracting sea surface reflection coefficient based on far-field observation records of air gun sources. Background Art

[0002] In marine seismic exploration, the source and the acquisition cable are generally located at a certain depth underwater. The existence of the sea surface will cause reflection of the incident wave field, which has a certain impact on seismic data processing and interpretation. The accurate extraction of the sea surface reflection coefficient has an important impact on source wavelet simulation and inversion, multiple wave prediction and suppression, and ghost wave removal at the source and acquisition ends. At present, the sea surface reflection coefficient is mostly assumed to be -1, that is, a horizontal sea surface, which has a significant error from the undulating sea conditions in the actual acquisition process. Summary of the Invention

[0003] The purpose of the present invention is to provide a method for extracting sea surface reflection coefficient based on far-field observation records of air gun sources, which can obtain an accurate sea surface reflection coefficient.

[0004] To achieve the above purpose, the present invention provides a method for extracting sea surface reflection coefficient based on far-field observation records of air gun sources, including the following steps:

[0005] Place a far-field hydrophone near the seabed to record the source wavelet emitted by the air gun;

[0006] The source wavelet S(t) recorded by the far-field hydrophone is the superposition result of the primary wave field and the ghost wave field. The primary wave p(t) and the ghost wave g(t) can be respectively expressed as:

[0007] S i (t) = p i (t) + g i (t) (1)

[0008]

[0009] Wherein, t represents time, i represents the shot number, s represents the imaginary source, r p and r g are the propagation distances when the primary wave and the ghost wave reach the far-field hydrophone, v w is the sound speed in water, and R is the sea surface reflection coefficient;

[0010] Solve the sea surface reflection coefficient R through the primary wave and the ghost wave corresponding to the source wavelet;

[0011] R i (f) = G i (f) * r gi / P i (f) * r pi (4)

[0012] Among them, G i (f), P i (f) Frequency domain expressions corresponding to the first wave and the ghost wave, respectively.

[0013] As a further solution of the present invention: the positive peak in the source wavelet contains only the first wave, while the negative peak contains the ghost wave and part of the first wave. The ghost wave is obtained by the trend fitting method. The mathematical expression is as follows:

[0014]

[0015] Where f is the frequency, f n is the selected upper frequency limit, S ref is the first wave content in the negative peak, E is the square error, A pos and A neg Corresponding to positive and negative peaks, S is removed ref The amplitude spectrum after .

[0016] As a further solution of the present invention: the trend fitting method comprises the following steps:

[0017] S1, intercept the positive peak of the source wavelet as the matching target. At this time, since the ghost wave has not arrived yet, all of them are first wave components;

[0018] S2, intercepting the negative peak part of the source wavelet as the initial matching input. Due to the periodic oscillation of the bubble, this part contains part of the tail wave oscillation following the positive peak of the first wave;

[0019] S3, keep the amplitude unchanged to simplify the tail wave oscillation to different reference bases and remove it from the negative peak;

[0020] S4, perform spectrum analysis on S1 and S3, take low frequency as target window, and calculate the square error between the two within the target window;

[0021] S5. The ghost wave obtained by using the reference benchmark corresponding to the minimum square error is the best matching result.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] Based on the consistent energy characteristics of the ultra-low frequency parts of the first wave and the ghost wave, a trend fitting line is constructed to separate the first wave and the ghost wave from the original source wavelet. The ghost wave reflection is obtained by the trend fitting method, and on this basis, the accurate sea surface reflection coefficient and its variation with frequency and incident angle are inverted to serve the subsequent data processing (ghost wave and multiple wave suppression, wavelet deconvolution, etc.). BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic diagram of far-field wavelet observation of airgun seismic source of the present invention.

[0025] Figure 2 This is the air gun source wavelet profile after peak flattening in the embodiment of the present invention.

[0026] Figure 3 This is the propagation distance (a) and incident angle (b) of the first arrival wave and ghost wave when the same shot is fired in the embodiment of the present invention.

[0027] Figure 4 This is the process diagram of ghost wave fitting based on trend prediction in the embodiment of the present invention.

[0028] Figure 5 This is the result of the variation of the sea surface reflection coefficient with the incident angle and frequency in the embodiment of the present invention. Detailed implementation mode

[0029] The present invention will be further described below through embodiments.

[0030] A method for extracting the sea surface reflection coefficient based on the far - field observation record of an air gun source includes the following steps:

[0031] Place the far - field hydrophone near the seabed to record the source wavelet emitted by the air gun; the air gun and the far - field hydrophone are arranged as shown, where α and β respectively represent the incident angles of the first arrival wave and the ghost wave corresponding to the far - field hydrophone; Figure 1 as shown, Figure 1 in which α and β respectively represent the incident angles of the first arrival wave and the ghost wave corresponding to the far - field hydrophone;

[0032] The source wavelet S(t) recorded by the far - field hydrophone is the superposition result of the wave fields of the first arrival wave and the ghost wave. Among them, the first arrival wave p(t) and the ghost wave g(t) can be respectively expressed as:

[0033] S i (t) = p i (t)+g i (t) (1)

[0034]

[0035] where t represents time, i represents the shot number, s represents the imaginary source, r p and r g are the propagation distances when the first arrival wave and the ghost wave reach the far - field hydrophone, v w is the sound speed in water, and R is the sea surface reflection coefficient;

[0036] Solve the sea surface reflection coefficient R through the first arrival wave and the ghost wave corresponding to the source wavelet;

[0037] R i (f) = G i (f)*r gi / P i (f)*r pi (4)

[0038] Among them, G i (f), P i (f) Frequency domain expressions corresponding to the first wave and the ghost wave, respectively.

[0039] Furthermore, the positive peak of the source wavelet only contains the first wave, while the negative peak contains ghost waves and part of the first wave. The ghost wave is obtained by trend fitting. The main idea is to use the positive peak of the source wavelet as the matching target and the negative peak as the matching input. By adjusting the first wave component in the negative peak and calculating the low-frequency error between it and the matching target, when the low-frequency error is the lowest, the components in the corresponding negative peak are all ghost waves. The mathematical expression is as follows:

[0040]

[0041] Where f is the frequency, f n is the selected upper frequency limit, S ref is the first wave content in the negative peak, E is the square error, A pos and A neg Corresponding to positive and negative peaks, S is removed ref The amplitude spectrum after .

[0042] Further, the trend fitting method steps include:

[0043] S1, intercept the positive peak of the source wavelet as the matching target. At this time, since the ghost wave has not arrived yet, all of them are first wave components;

[0044] S2, intercepting the negative peak part of the source wavelet as the initial matching input. Due to the periodic oscillation of the bubble, this part contains part of the tail wave oscillation following the positive peak of the first wave;

[0045] S3, keep the amplitude unchanged to simplify the tail wave oscillation to different reference bases and remove it from the negative peak;

[0046] S4, perform spectrum analysis on S1 and S3, take low frequency as target window, and calculate the square error between the two within the target window;

[0047] S5. The ghost wave obtained by using the reference benchmark corresponding to the minimum square error is the best matching result.

[0048] In order to illustrate the effectiveness of the present invention, actual measured data are used for verification. The embodiments are as follows:

[0049] During the actual acquisition process, the far-field hydrophone was 280 m away from the seabed and 1500 m from the sea surface. The airgun source excitation depth was 5 m, and the source ship towed the airgun source over the far-field hydrophone. Figure 2 This is the airgun source wavelet profile after the peak is flattened (around 250ms).

[0050] Comparing formulas (2) and (3), it is found that the source wavelet results are directly related to the propagation distance, incident angle, etc. When the propagation distance is large enough relative to the source excitation depth, the propagation distances and incident angles of the first arrival and ghost wave of the same shot are approximately the same, as Figure 3 shown.

[0051] Figure 4 Figure is a schematic diagram of the trend fitting process based on low-frequency consistency. Its main basis is that the sea surface reflection is stable for the ultra-low frequency part, that is, the reflection coefficient is assumed to be -1 at the low-frequency end. Using this rule, after fitting the low frequencies of the first arrival and the ghost wave, the ghost wave part can be separated from the original source wavelet ( Figure 3 a). Figure 3 The red curve in b). Figure 4 In c, the black line corresponds to the first arrival spectrum, the light blue line corresponds to the fitting curve, and the red line corresponds to the best-fitting ghost wave spectrum curve.

[0052] With the separated first arrival and ghost wave parts, the Fourier transforms of the two waveforms can be performed using formulas (2) and (3), and the spectra can be compared to obtain the sea surface reflection coefficient R. Figure 5 The reflection coefficient during this data acquisition process is obtained using the data. The analysis shows that the overall range of the reflection coefficient is between -1 and -0.2. As the incident angle increases, R approaches -1 more closely. At the same time, affected by the frequency, the higher the frequency, the lower the sea surface reflection coefficient (absolute value), indicating that more energy is scattered away.

Claims

1. A method for extracting sea surface reflection coefficient based on airgun seismic source far-field observation records, characterized in that: The following steps are involved: A far-field hydrophone is placed near the seafloor to record the source wavelets emitted by the airgun; The source wavelet S(t) recorded by the far-field hydrophone is the wave field superposition result of the first wave and the ghost wave, where the first wave p(t) and the ghost wave g(t) can be expressed as: S i (t)=p i (t)+g i (t) (1) Among them, t represents time, i represents the shot number, s represents the hypothetical earthquake source, and r p and r g is the propagation distance of the first wave and the ghost wave when they reach the far-field hydrophone, v w is the speed of sound in water, R is the reflection coefficient of the sea surface; The sea surface reflection coefficient R is solved by the head wave and ghost wave corresponding to the source wavelet; R i (f)=G i (f)*r gi / P i (f)*r pi (4) Among them, G i (f), P i (f) Frequency domain expressions corresponding to the first wave and the ghost wave, respectively.

2. The method for extracting sea surface reflection coefficient based on airgun seismic source far-field observation records according to claim 1 is characterized in that: The positive peak value in the source wavelet only contains the first wave, while the negative peak value contains the ghost wave and part of the first wave. The ghost wave is obtained by the trend fitting method. The mathematical expression is as follows: Where f is the frequency, f n is the selected upper frequency limit, S ref is the first wave content in the negative peak, E is the square error, A pos and A neg Corresponding to positive and negative peaks, S is removed ref The amplitude spectrum after .

3. The method for extracting sea surface reflection coefficient based on airgun seismic source far-field observation records according to claim 2 is characterized in that: The trend fitting method steps include: S1, intercept the positive peak of the source wavelet as the matching target. At this time, since the ghost wave has not arrived yet, all of them are first wave components; S2, intercepting the negative peak part of the source wavelet as the initial matching input. Due to the periodic oscillation of the bubble, this part contains part of the tail wave oscillation following the positive peak of the first wave; S3, keep the amplitude unchanged to simplify the tail wave oscillation to different reference bases and remove it from the negative peak; S4, perform spectrum analysis on S1 and S3, take low frequency as target window, and calculate the square error between the two within the target window; S5. The ghost wave obtained by using the reference benchmark corresponding to the minimum square error is the best matching result.

Images