A method for data purification and correction of active source bottom seismograph in shallow sea environment

Abstract

The present application belongs to the technical field of seismic data processing, and particularly relates to a method for purifying and correcting data of an active source bottom seismograph in a shallow sea environment, which specifically comprises: on the basis of preprocessing of original data of the active source bottom seismograph in the shallow sea area, including decoding, data clipping, position setting and format conversion, performing system travel time difference correction; suppressing swell interference noise by using a method combining direct current component removal and high-pass filtering; performing amplitude compensation on the data by using a travel time-offset joint amplitude compensation method; performing seismic wave shaping processing on the processed bottom seismograph data, so as to ensure consistent wavelet waveforms of seismic traces and output high-quality bottom seismograph preprocessing data. The present application improves the recording quality of far-offset stations by using new time delay correction, swell noise suppression and amplitude compensation methods, and realizes effective utilization of the recording traces of the bottom seismograph stations.

Description

Technical Field

[0001] This invention belongs to the field of earthquake data processing technology, specifically relating to a method for data purification and correction of active source seafloor seismometers in shallow sea environments. Background Technology

[0002] In the field of exploration geophysics, the detection and detailed characterization of deep-sea structures are the primary objectives of deep-sea seismic exploration. Active-source seafloor seismometers, which are submerged on the seabed and receive seismic wave signals excited by a high-capacity airgun source, are crucial for obtaining detailed velocity structures of the Earth's crust through data processing. Existing active-source seafloor seismometer data purification and correction schemes are traditional and primarily designed for deep-water environments. The general process includes: 1. Amplitude compensation: For standard format seismic data obtained after data decoding and formatting, navigation and coordinate insertion, time exponential function amplitude compensation is used to restore the energy of seismic waves that has attenuated with increasing propagation time due to geometric diffusion and earth absorption.

[0003] 2. Conventional filtering: Apply bandpass or highpass filters to suppress noise in specific frequency bands of the data and improve the signal-to-noise ratio.

[0004] However, when applied to shallow sea (water depth < 100 meters) environments and active source seafloor seismograph data with ultra-large offsets (> 100 kilometers), existing technical solutions exhibit significant inadequacies and poor processing results, specifically in the following three aspects: I. Environmental and Noise Characteristics: Due to the shallow water in shallow sea areas, wind, waves, and ocean currents directly affect the seabed, generating high-energy, low-frequency swell noise. This noise spectrum often overlaps with the low-frequency portion of the effective signal, leading to a dilemma in conventional filtering used in traditional preprocessing processes when suppressing this noise: if the cutoff frequency is set too high, it will damage the effective low-frequency signal; if it is set too low, the suppression effect is limited. Therefore, traditional filtering methods are incomplete in suppressing the high-energy, low-frequency swell noise unique to shallow seas and easily damage the effective low-frequency components, resulting in insufficient improvement in the data signal-to-noise ratio, and the effective phase is still masked by noise.

[0005] II. Acquisition Methods and Systematic Errors: Deep-sea exploration using active-source seafloor seismometers often employs a long-array, multi-ship acquisition method with different excitation directions. Due to factors such as instrument clock synchronization errors, positioning accuracy limitations, and sea state variations, systematic travel time errors occur between seismic traces recorded from different directions at the same offset from the same station. Within the common receiver gather, discontinuous "jitter" or "jumps" appear in the phase axes from different directions, disrupting the smoothness and continuity of the travel time curves. Traditional preprocessing lacks a dedicated correction step for such systematic travel time errors, leading to inaccurate subsequent travel time picking and directly reducing the accuracy and reliability of the velocity inversion model.

[0006] III. Ultra-Large Offsets and Seismic Wave Propagation Attenuation: To probe deep structures, active-source seafloor seismographs typically employ ultra-large offsets. Seismic waves travel extremely long paths, undergoing both geometric diffusion and Earth absorption attenuation, with the degree of attenuation closely related to the propagation path (offset). Traditional amplitude compensation methods only consider energy attenuation over propagation time, neglecting the influence of offset, resulting in severely insufficient energy compensation for large-offset seismic traces, especially for deep-seated reflected waves. This leads to the inability of traditional amplitude compensation methods to effectively recover the signal energy of far-offset seismic traces, making it difficult to identify weak deep-seated reflected signals, resulting in poor amplitude fidelity and limiting the imaging capabilities of deep structures. Summary of the Invention

[0007] This invention overcomes the above-mentioned defects and provides a method for data purification and correction of active source seafloor seismometers in shallow sea environments. It solves the problems in the prior art, such as the effective signal being masked by strong energy low-frequency swell noise in shallow seas, systematic travel time errors caused by the acquisition method, and the inability to identify deep weak reflection signals due to the energy attenuation of seismic wave signals caused by ultra-large offset.

[0008] To achieve the above objectives, the present invention provides a method for data purification and correction of active source seafloor seismometers in shallow sea environments, comprising the following steps: S1. Based on the preprocessing of the original seabed seismograph data in the shallow sea area, including decompilation, data truncation, location insertion, and format conversion, system travel time difference correction processing is performed to correct the first arrival wave travel time differences in the station records caused by different time periods and different excitation directions. S2. Suppress low-frequency interference noise from surge waves by combining DC component removal and high-pass filtering. S3. The amplitude compensation method of travel time-offset joint amplitude compensation is used to compensate the data; S4. Perform seismic wave shaping on the seabed seismograph data processed in steps S1-S3 to ensure consistent wavelet waveforms in the seismic traces and output high-quality preprocessed seabed seismograph data.

[0009] Further, in step S1, the system timing difference correction process specifically includes: For records from submarine seismograph stations, based on the tidal patterns at different times, the time differences caused by the sinking depth of the excitation source are uniformly corrected to the sea level reference using the seawater velocity data measured in the exploration area. Using seismic cross-correlation processing technology, starting from the zero-offset seismic trace, cross-correlation processing is performed on each trace with its adjacent traces, and the time delay of the maximum amplitude obtained by cross-correlation is used as the first arrival time delay between traces. Statistical analysis was performed on all the aforementioned inter-channel first arrival delays. When the correlation of all delay values ​​in the direction of a seismic source or within a fixed time period is above 95%, it is determined to be a system delay. In the records of the seabed seismograph station, the system delay is subtracted from the first arrival time of the seismic traces according to different acquisition directions or acquisition periods to obtain the seabed seismograph data after system travel time difference correction.

[0010] Furthermore, in step S2, the method of combining DC component removal and high-pass filtering to suppress surge interference noise is as follows: S21 calculates the mean amplitude of the sample point amplitude values ​​recorded by each seismic trace from the seabed seismograph station. The mean amplitude is an estimate of the DC component, and the formula is as follows: , in, Indicates the mean amplitude. Indicates the amplitude value of the sample point. N Indicates the number of sampling points in the seismic trace; S22 removes the mean amplitude from the sample amplitude values ​​of the earthquake to obtain the earthquake amplitude value after removing the DC component. The calculation is as follows: ; S23 uses a high-pass filter to remove low-frequency noise from the signal after removing the DC component. The design of the low cutoff parameter in the high-pass filter is based on the low cutoff frequency of the effective signal in the spectrum.

[0011] Furthermore, before executing steps S1 and S2, a data quality determination is also included. If the noise interference of the raw data of the seabed seismometer is severe, making it impossible to identify the first arrival wave, then step S2 is executed first. If the timing jitter of the raw data of the seabed seismometer makes it impossible to accurately track the first arrival wave, then step S1 is executed first.

[0012] Further, in step S3, the shallow seabed seismograph station records with a very large offset. Based on the seismic wave time exponential function amplitude compensation, the influence of the offset on seismic wave attenuation is added to compensate for the amplitude. The amplitude compensation formula is as follows: , in, This is the earthquake signal after amplitude compensation. The initial amplitude of the seismic wave. For the travel time of seismic waves, This represents the earthquake offset.

[0013] Further, in step S4, the seismic wave shaping process is as follows: S41 performs spectral analysis on the receiver point domain gather data to obtain the frequency domain characteristics of the seismic data. The spectral analysis window contains a complete effective seismic waveform. S42 determines the dominant frequency of the seismic wave based on the spectral analysis results and calculates the time window length of the deconvolution operator; , in, The time window length for the deconvolution operator. This refers to the dominant frequency of the seismic wave. S43 establishes a seismic convolution model and performs frequency domain transformation: The seismic convolution model: , in, For seismic reflection trace data, For the reflection coefficient sequence, For wavelets; After Fourier transform, it can be expressed in the frequency domain as: , in, , They are respectively , , Fourier transform; S44 Extraction Wavelet Shaping Deconvolution Operator; Based on the stated time window length, local data segments of the original seismic trace data are extracted, and the statistical characteristics of the original wavelet are calculated to extract the deconvolution operator of the original wavelet. ; For the far-field wavelet of a known air gun array source Perform a Fourier transform to obtain the frequency domain expression. ; use Replace the original wavelet Seismic trace data after wavelet processing were obtained. : , in, Wavelet shaping deconvolution operator; right Perform an inverse Fourier transform to obtain the time-domain deconvolution operator for wavelet processing; S45 performs a convolution operation between the time-domain deconvolution operator and the original seismic trace data to complete wavelet replacement and obtain the shaped seismic trace data.

[0014] Compared with the prior art, the advantages of the present invention are as follows: This invention employs a processing technique that combines DC component removal with high-pass filtering, which solves the problem that traditional filtering methods struggle to effectively suppress low-frequency, high-energy surge noise in shallow sea areas without damaging the effective signal, thereby effectively improving the data signal-to-noise ratio and highlighting the effective seismic phase. This invention proposes a system travel time difference correction technique, which solves the problem of significant systematic travel time errors between seismic traces from the same excitation point (same offset) but different excitation times received by the same seabed seismograph station. This enables the effective fusion and utilization of seismic data from different times and different acquisition directions, providing a more flexible and efficient solution for field data acquisition. This invention employs a combined travel time-offset distance (shot-receiver distance) amplitude compensation processing method, which solves the problem of rapid and non-uniform energy attenuation caused by the dual effects of ground absorption and geometric diffusion during the propagation of seismic waves over such long distances. This lays the foundation for using seismic amplitude to invert formation velocity, density and other physical parameters. Furthermore, the dispersion effect of ground filtering intensifies with increasing propagation distance, leading to waveform distortion and significant differences in frequency characteristics in seismic traces at different offsets. This inconsistency between near and far traces severely disrupts the lateral continuity and comparability of seismic wave group characteristics, making subsequent waveform consistency-based processing and interpretation difficult. To address this issue, wavelet deconvolution processing technology is used to reshape seismic waves, eliminating the ground filtering effect during propagation and ensuring waveform consistency between far-field (especially over 100 km) and near-field seismic traces. Attached Figure Description

[0015] Figure 1 This is a flowchart of a method for data purification and correction of an active source seafloor seismograph in a shallow sea environment, according to the present invention. Figure 2 These are comparison diagrams of the first arrival wave phase before and after time delay correction in an embodiment of the present invention, wherein (a) is the first arrival wave phase diagram before time delay correction, and (b) is the first arrival wave phase diagram after time delay correction according to the present invention. Figure 3 This is a comparison chart of the processing effects of the preprocessing method of the present invention and the conventional method on the records of K03 station in a certain sea area according to an embodiment of the present invention. Among them, (a) is the conventional method and (b) is the preprocessing method of the present invention. Figure 4This is a comparison chart of the processing effects of the preprocessing method of the present invention and the conventional method on the records of C07 station in a certain sea area according to an embodiment of the present invention. In this chart, (a) is the conventional method and (b) is the preprocessing method of the present invention. Figure 5 This is a comparison chart of the processing effects of the preprocessing method of the present invention and the conventional method on the records of station K07 in a certain sea area according to an embodiment of the present invention. In the chart, (a) is the conventional method and (b) is the preprocessing method of the present invention. Detailed Implementation

[0016] The technical solutions in 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.

[0017] This invention proposes a method for data purification and correction of active-source seafloor seismometers in shallow sea environments, such as... Figure 1 As shown, this invention improves upon the traditional seafloor seismograph data preprocessing workflow by enhancing noise suppression. Addressing the high energy of low-frequency noise in shallow sea areas, it employs a combination of DC component removal and high-pass filtering for suppression. Instead of traditional time-exponential amplitude compensation, it uses travel time-offset (shot-receiver distance) combined amplitude compensation, effectively compensating for signal attenuation at large offsets. Furthermore, it adds system travel time error correction and wavelet deconvolution processes to eliminate first-arrival jumps caused by system travel time and differences in seismic waveforms at different offsets, resulting in high-quality seismic station data. Details are as follows: 1. After basic processing such as decompilation, data truncation, location insertion and format conversion of the raw data of the active source seafloor seismograph in the shallow sea area, system travel time difference correction processing is performed to correct the difference in first arrival wave travel time caused by different time periods and different excitation directions in the station records. The data is decompiled from the original ASCII format. With the excitation time as zero, each shot record is truncated into equal-length record channels according to the record length and formed into Segy format station records. The time length must be greater than the Mohorovičić discontinuity reflection wave travel time of the seismic channel record with the maximum shot-receiver distance. Based on this, the geodetic coordinates of the recording station and the excitation point are placed into the trace head of the Segy format station record. The recording position strictly follows the standard Segy format specifications.

[0018] In seabed seismograph data acquisition, to increase the density of seismic beam illumination, it is necessary to reduce the excitation distance between seismic sources. However, because the source vessel must maintain a certain speed during operation to ensure the safety of towing the seismic source, a common practice is to excite the seismic source along one direction of the survey line to the end point, then turn around and start exciting the seismic source in the opposite direction from the end point to the starting point. This results in significant differences in the excitation times of adjacent seismic channels arranged by offset. Furthermore, factors such as sea conditions, avoidance of other vessels, and equipment malfunctions can all lead to data collection at different times along the same survey line, sometimes with considerable intervals between the time periods.

[0019] Under these acquisition conditions, due to the difference between the timing system of the source control and the timing system of the seabed seismograph, there are systematic differences in timing at different times. This results in a large systematic time delay in seismic traces with the same offset distance at different acquisition directions and time periods. In the station records arranged in order of offset distance, the waveform of the first arrival wave phase jumps up and down and jitters.

[0020] Traditional seafloor seismograph data processing does not consider the correction of system time delay, resulting in the use of data from only one source direction for processing. This means that nearly half of the seismic data cannot be effectively utilized, reducing the accuracy of velocity inversion.

[0021] To address the time delay issue in seabed seismograph data, this invention employs the following method for correction: For records from submarine seismograph stations, based on the tidal patterns at different times, the time differences caused by the sinking depth of the excitation source are uniformly corrected to the sea level reference using the seawater velocity data measured in the exploration area. Using seismic cross-correlation processing technology, starting from the zero offset seismic trace, each trace is cross-correlated with its adjacent traces, and the time delay of the maximum amplitude obtained by cross-correlation is used as the first arrival time delay between traces. Statistical analysis was performed on all inter-channel first arrival delays. When the correlation of all delay values ​​within a fixed time period in the direction of a seismic source is above 95%, it is determined to be a system delay.

[0022] For records from seabed seismograph stations, the system delay is subtracted from the first arrival time according to the seismic traces of different acquisition directions / acquisition periods, thus completing the delay correction process.

[0023] Figure 2 This is a comparison diagram of the first arrival seismic phases before and after time delay correction according to an embodiment of the present invention. As can be seen from the diagram, due to the existence of time system delay, the first arrival seismic phases on the seismic profile exhibit a significant vertical jump phenomenon. Figure 2(a) causes strong interference to travel time picking, which seriously affects the accuracy of phase travel time picking and thus affects the inversion accuracy of the final velocity structure results; after the time delay correction processing of this invention, the phenomenon of the first arrival wave phase jumping is eliminated, and the phase waveform is smooth and has good continuity. Figure 2 (b) ensures the accuracy of seismic phase travel time picking, thus laying a good foundation for obtaining accurate final velocity-structure inversion results.

[0024] II. Suppressing surge interference noise by combining DC component removal and high-pass filtering; Because deep-sea seismic exploration by seabed seismometers begins in the deep ocean, the data acquisition environment is quiet, with very little constant amplitude shift at zero (or near-zero) frequency or slowly changing amplitude baseline drift. Therefore, the DC component removal process is not included in the preprocessing workflow for seabed seismic data. However, as deep-sea seismic exploration extends to shallower waters, low-frequency and very low-frequency interference waves caused by swells and other disturbances develop. Conventional high-pass filters, due to their limitations in edge frequency energy leakage, are insufficient to completely suppress swell interference. Therefore, there is an urgent need to develop a targeted DC component removal + high-pass filter combination to suppress low-frequency interference caused by swells and other disturbances.

[0025] The combined processing of DC component removal and high-pass filtering for suppressing low-frequency interference caused by surges, etc., is characterized by the following steps: a) Statistically calculate the constant amplitude offset or slowly changing amplitude baseline drift (DC component drift amplitude value) for each seismic trace recorded by the seabed seismograph station: The basic mathematical model for expressing the DC component is: Assume a seismic record signal is It contains valid signals. and DC offset (It can be a constant or a slowly changing function) ).

[0026] The original signal is: , The goal of removing the DC component is to improve the performance of actual seismic signals. The estimated DC is removed from the signal to obtain the effective signal. .

[0027] Due to effective seismic signals Since the data itself has zero mean (symmetric positive and negative amplitudes), the most common and effective method is to calculate the arithmetic mean of the trace records and use this mean as an estimate of the DC component. Considering the characteristics of seafloor seismograph data in shallow sea areas, a global mean subtraction method is used to obtain the DC component. This involves calculating the arithmetic mean of all sampling points in the entire seismic record and then subtracting this mean from each sampling point in that trace.

[0028] The amplitude values ​​of sample points in each seismic trace recorded by the station Find the mean amplitude The mean amplitude is the DC component, and the formula is as follows: , Where N represents the number of sampling points in the seismic trace; b) By removing the mean amplitude from the sample amplitude values ​​of the earthquake, the earthquake amplitude value after removing the DC component can be obtained. The calculation is as follows: ; This completes the removal of the DC component. c) Even after removing the DC component, there may still be very slowly changing low-frequency trends or noise in the seismic signal; these low-frequency noise components, after Fourier transform, are used to obtain the distribution frequency band of low-frequency noise by using the energy distribution trend that is much higher than that of the effective wave on the spectrum; further, bandpass filtering is used to remove the noise in the low-frequency band.

[0029] In addition, before data processing, data quality can be assessed. The order of steps one and two above can be determined based on data characteristics. If the original seismic data has severe noise interference, making it difficult to identify and track the first arrival wave, then a combination of DC component removal and bandpass filtering can be used to suppress low-frequency interference noise such as swells. Conversely, if the system travel time error has a large impact and waveform jumps make it difficult to identify the first arrival wave, then system travel time difference correction should be performed first to obtain seafloor seismograph data with a higher signal-to-noise ratio.

[0030] 3. Amplitude compensation of the data is performed using the travel time-offset combined amplitude compensation method; Traditional time-exponential function amplitude compensation methods are inadequate for large-offset seismic traces. Seafloor seismograph stations record offsets ranging from several kilometers to a maximum of 200 kilometers, falling into the category of ultra-long offsets. Therefore, for ultra-long offset seafloor seismograph station records, this paper considers the influence of offset on seismic wave attenuation, building upon traditional time-exponential function amplitude compensation. The amplitude compensation equation is as follows: , in, This is the earthquake signal after amplitude compensation. The amplitude of the seismic waves actually received (i.e., the original seismic record). For the travel time of seismic waves, This represents the earthquake offset.

[0031] After performing compensation calculations on the seismic waves of the original seismic records using the above formula, the station records with amplitude compensation are obtained.

[0032] Fourth, the above-processed seabed seismograph data is subjected to seismic wave shaping and waveform homogenization to compensate for the problem of waveform widening with increasing propagation distance caused by ground absorption attenuation of seismic wavelength, ensuring consistent seismic trace wavelet waveforms and outputting high-quality seabed seismograph preprocessed data.

[0033] 1) Perform spectral analysis on the receiver point domain gather data to obtain the frequency domain characteristics of the seismic data. The spectral analysis time window contains a complete effective seismic waveform. 2) Determine the dominant frequency of the seismic wave based on the spectral analysis results, and calculate the time window length of the deconvolution operator; , in, The time window length for the deconvolution operator. This refers to the dominant frequency of the seismic wave. 3) Establish a seismic convolution model and perform frequency domain transformation; The seismic convolution model: , in, For seismic reflection trace data, For the reflection coefficient sequence, For wavelets; After Fourier transform, it can be expressed in the frequency domain as: , in, , They are respectively , , Fourier transform; 4) Extract the wavelet shaping deconvolution operator; Based on the stated time window length, local data segments of the original seismic trace data are extracted, and the statistical characteristics of the original wavelet are calculated to extract the deconvolution operator of the original wavelet. ; For the far-field wavelet of a known air gun array source Perform a Fourier transform to obtain the frequency domain expression. ; use Replace the original wavelet Seismic trace data after wavelet processing were obtained. : , in, Wavelet shaping deconvolution operator; right Perform an inverse Fourier transform to obtain the time-domain deconvolution operator for wavelet processing; 5) Perform a convolution operation between the time-domain deconvolution operator and the original seismic trace data to complete the wavelet replacement and obtain the shaped seismic trace data.

[0034] Figure 3 The preprocessing method of this invention was used to record data from station K03 in a certain sea area in an embodiment of this invention. Figure 3 (b) and conventional pretreatment methods ( Figure 3 The comparison of the processing effects in (a) shows that after preprocessing by the method of the present invention, the imaging quality of the profile is greatly improved, the effective seismic phases are highlighted, the signal-to-noise ratio is improved, the amplitude and wavelet tend to be consistent, and the range of identifiable seismic phases is greatly increased. This processing has also been applied to seismic records from other stations, and all have achieved good preprocessing results. Figure 4 , Figure 5 ).

[0035] The embodiments of the present invention described above do not constitute a limitation on the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A method for data cleaning and correction of active source ocean bottom seismometer in shallow water environment, characterized in that, Includes the following steps: S1. Based on the preprocessing of the original seabed seismograph data in the shallow sea area, including decompilation, data truncation, location insertion, and format conversion, system travel time difference correction processing is performed to correct the first arrival wave travel time differences in the station records caused by different time periods and different excitation directions. S2. Suppress low-frequency interference noise from surge waves by combining DC component removal and high-pass filtering. The method of combining DC component removal and high-pass filtering to suppress surge interference noise is as follows: S21 calculates the mean amplitude of the sample point amplitude values ​​recorded by each seismic trace from the seabed seismograph station. The mean amplitude is an estimate of the DC component, and the formula is as follows: ， in, Indicates the mean amplitude. This represents the amplitude value of the sample point, and N represents the number of sampling points in the seismic trace; S22 removes the mean amplitude from the sample amplitude values ​​of the earthquake to obtain the earthquake amplitude value after removing the DC component. The calculation is as follows: ； S23 uses a high-pass filter to remove low-frequency noise from the signal after removing the DC component. The design of the low cutoff parameter in the high-pass filter is based on the low cutoff frequency of the effective signal in the spectrum. S3. The amplitude compensation method of travel time-offset joint amplitude compensation is used to compensate the data; Shallow seabed seismograph stations record data with extremely large offsets. Based on the seismic wave time exponential function amplitude compensation, the influence of offset on seismic wave attenuation is added to compensate for the amplitude. The amplitude compensation formula is as follows: ， in, This is the seismic signal after amplitude compensation. This represents the initial amplitude of the seismic wave. For the travel time of seismic waves, This is the earthquake offset distance; S4. Perform seismic wave shaping on the seabed seismograph data processed by steps S1-S3 to ensure consistent wavelet waveforms in the seismic traces and output high-quality preprocessed seabed seismograph data. Before executing steps S1 and S2, a data quality determination is also included. If the noise interference of the raw data of the seabed seismometer is severe and the first arrival wave cannot be identified, then step S2 is executed first. If the timing jitter of the raw data of the seabed seismometer causes the first arrival wave to be unable to be accurately tracked, then step S1 is executed first.

2. The method for data purification and correction of active source seafloor seismometers in shallow sea environments according to claim 1, characterized in that, In step S1, the system timing difference correction process specifically includes: For records from submarine seismograph stations, based on the tidal patterns at different times, the time differences caused by the sinking depth of the excitation source are uniformly corrected to the sea level reference using the seawater velocity data measured in the exploration area. Using seismic cross-correlation processing technology, starting from the zero-offset seismic trace, cross-correlation processing is performed on each trace with its adjacent traces, and the time delay of the maximum amplitude obtained by cross-correlation is used as the first arrival time delay between traces. Statistical analysis was performed on all the aforementioned inter-channel first arrival delays. When the correlation of all delay values ​​in the direction of a seismic source or within a fixed time period is above 95%, it is determined to be a system delay. In the records of the seabed seismograph station, the system delay is subtracted from the first arrival time of the seismic traces according to different acquisition directions or acquisition periods to obtain the seabed seismograph data after system travel time difference correction.

3. The method for data purification and correction of active source seafloor seismometers in shallow sea environments according to claim 1, characterized in that, In step S4, the seismic wave shaping process is as follows: S41 performs spectral analysis on the receiver point domain gather data to obtain the frequency domain characteristics of the seismic data. The spectral analysis window contains a complete effective seismic waveform. S42 determines the dominant frequency of the seismic wave based on the spectral analysis results and calculates the time window length of the deconvolution operator: ， in, The time window length for the deconvolution operator. This refers to the dominant frequency of the seismic wave. S43 establishes a seismic convolution model and performs frequency domain transformation: The seismic convolution model: , in, For seismic reflection trace data, It is a sequence of reflection coefficients. For wavelets; After Fourier transform, it can be expressed in the frequency domain as: , in, , They are respectively , , Fourier transform; S44 Extraction Wavelet Shaping Deconvolution Operator: Based on the stated time window length, local data segments of the original seismic trace data are extracted, and the statistical characteristics of the original wavelet are calculated to extract the deconvolution operator of the original wavelet. ; For the far-field wavelet of a known air gun array source Perform a Fourier transform to obtain the frequency domain expression. ; use Replace the original wavelet Seismic trace data after wavelet processing were obtained. : ， in, Wavelet shaping deconvolution operator; right Perform an inverse Fourier transform to obtain the time-domain deconvolution operator for wavelet processing; S45 performs a convolution operation between the time-domain deconvolution operator and the original seismic trace data to complete wavelet replacement and obtain the shaped seismic trace data.

Images