Layered modeling device and method for chromatography model

Abstract

The invention belongs to the technical field of geophysical exploration, and particularly discloses a stratified modeling device and method for a chromatography model. The device comprises a tomography inversion speed module, a speed difference calculation module, an elevation interface extraction module, a grid point surface layer speed calculation module, a correction curve module and a hierarchical model module. The method comprises the following steps: firstly, obtaining a grid velocity model through tomography inversion, then calculating the longitudinal velocity difference of the grid velocity model to obtain all boundary grid points, extracting an elevation interface, obtaining the surface thickness, then calculating the surface velocity at the grid points, and then establishing a thickness correction curve and a velocity correction curve according to the previous results, and continuously correcting the thickness and the speed to obtain a final hierarchical model. According to the method, the accuracy of constant velocity interface extraction can be improved, and control and verification data are provided for static correction of land petroleum seismic exploration data processing and shallow surface layer velocity modeling of pre-stack depth migration imaging.

Description

Technical Field

[0001] The invention belongs to the technical field of geophysical exploration, and in particular relates to a layered modeling method and device for a tomographic model. Background Art

[0002] The main method for finding and exploring oil and gas, the main work includes three steps: seismic data acquisition, processing and interpretation. Seismic data processing mainly provides data results for seismic data interpretation. There are many contents in seismic data processing, mainly including static correction, denoising, deconvolution, dynamic correction, velocity analysis, stacking and migration. In basic seismic theory, it is assumed that the excitation point and the receiving point are on the same horizontal plane, and the seismic wave velocity of the formation is uniform. But the actual surface elevation is undulating, and there is a set of weathered layers or unconsolidated sediments called low velocity reduction layers in the near surface. Due to the different degrees of weathering or consolidation at different locations, the thickness and velocity of the low velocity reduction layer vary greatly relative to the underlying strata. The underlying strata here are generally called high-speed layers or refractive layers. There is an interface between it and the low velocity reduction layer with a physical property change of more than 30%. The physical properties here mainly refer to velocity, lithology and water content. At the same time, the seismic wave velocity is 30% higher than the maximum velocity of the low velocity reduction layer. The lateral change of the low velocity reduction layer distorts the hyperbolic shape of the travel time of the seismic exploration reflection wave, affecting the accuracy of seismic imaging. Therefore, the distortion of the near-surface low velocity reduction layer on the travel time of the reflection wave must be eliminated first when processing seismic data. There are two ways to eliminate the distortion of the reflection wave travel time in seismic data processing. One method is to first peel off the low velocity reduction layer, and then fill it to a horizontal reference plane with the velocity of the pre-weathering or consolidated rock. This is called static correction of seismic data time domain imaging. Another method is to directly use the velocity and thickness of the low velocity reduction layer as the near-surface velocity model, and then merge it with the shallow, medium and deep velocity model as the velocity model for depth domain imaging to perform depth domain migration imaging.

[0003] However, whether it is static correction in the time domain or PSDM (prestack depth migration), shallow surface velocity modeling requires accurate low-velocity reduction thickness and velocity, that is, finding the position of the high-velocity layer top interface. Therefore, in order to improve the accuracy of static correction and PSDM shallow surface velocity modeling, it is necessary to convert the continuous velocity model obtained by first-arrival tomographic inversion into a layered velocity structure.

[0004] First-arrival tomographic inversion is a major method for calculating static correction and shallow surface modeling in recent years. The velocity model obtained has the characteristics of large depth and high reliability. However, the tomographic velocity model is a continuous model, and it is necessary to find an interface between a low-velocity reduction layer and a high-velocity layer when calculating static correction. The usual calculation method is: use a fixed velocity value in the tomographic velocity model to extract an isovelocity interface as the depth value of the interface, and use this as the bottom interface of the low-velocity reduction layer in the entire area. The velocity used when extracting the interface is generally determined based on the average velocity of the high-velocity layer at the surface survey point in a region or based on the high-velocity layer velocity value used by the calculator to extract similar areas in the past. The advantage of this approach is simple operation; the disadvantage is that the high-velocity layer velocity used when extracting the isovelocity interface is subject to artificial intervention, and because the velocity used is unchanged, it cannot reflect the real high-velocity top interface changes in a region, especially when there are different high-velocity layer velocities in a region, the isovelocity interface cannot take both into account when extracting the isovelocity interface. Summary of the invention

[0005] In order to solve the above deficiencies existing in the known technology, the present invention aims to provide a layered modeling device for a tomographic model to solve the problem of inaccurate extraction of isokinetic interfaces.

[0006] Another object of the present invention is to provide a layered modeling method for a tomographic model, which is implemented using the above-mentioned layered modeling device for a tomographic model. The method improves the shortcomings of conventional methods for extracting interface elevations such as high-speed layers in first-arrival tomographic inversion velocity models, improves the calculation accuracy of static correction values ​​and the accuracy of shallow layer modeling, and provides control and verification data for static correction in onshore oil seismic exploration data processing and shallow layer velocity modeling in pre-stack depth migration imaging.

[0007] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0008] A layered modeling device for a tomographic model, the device comprising:

[0009] Tomographic inversion velocity module, velocity difference calculation module, elevation interface extraction module, grid point surface velocity calculation module, correction curve module and layered model module;

[0010] A tomographic inversion velocity module is used to perform tomographic inversion and output the tomographic inversion model as a grid velocity model, which is further output to a velocity difference calculation module;

[0011] The velocity difference calculation module is used to calculate the longitudinal velocity difference of the grid velocity model and obtain the boundary grid points of all high-speed layers and low-velocity reduction layers in the tomographic inversion velocity model of the entire work area, and further output them to the elevation interface extraction module;

[0012] The elevation interface extraction module is used to extract the boundary grid points between the final high-speed layer and the low-speed reduction layer, determine the final selected elevation interface, convert the relative elevation of the grid points in the elevation interface into the grid elevation, obtain the surface thickness of the grid points in the elevation interface, and output them to the grid point surface velocity calculation module and the correction curve module respectively;

[0013] The grid point surface velocity calculation module calculates the surface velocity at the grid point according to the surface thickness of the grid point in the elevation interface and outputs it to the correction curve module;

[0014] The correction curve module establishes a thickness correction curve and a velocity correction curve according to the results obtained by the elevation interface extraction module and the grid point surface velocity calculation module, and outputs them to the layered model module;

[0015] The layered model module obtains the thickness correction formula and the speed correction formula according to the thickness correction curve and the speed correction curve, corrects the surface thickness and the surface speed, and establishes the final layered model.

[0016] The present invention also discloses a layered modeling method for a tomographic model, which is implemented by using the layered modeling device for a tomographic model according to claim 1. The method comprises the following steps performed in sequence:

[0017] S1. Prepare a tomographic inversion velocity model, perform tomographic inversion according to the first arrivals picked up from seismic data through a tomographic inversion velocity module, and output the obtained tomographic inversion model as a grid velocity model;

[0018] S2, using the velocity difference calculation module to calculate the longitudinal velocity difference of the grid velocity model, and obtain the boundary grid points of all high-speed layers and low-speed reduction layers in the tomographic inversion velocity model of the entire work area;

[0019] S3, using the elevation interface extraction module, extracting the boundary grid points of the final high-speed layer and the low-speed reduction layer from the boundary grid points of all high-speed layers and low-speed reduction layers, determining the final selected elevation interface, converting the relative elevation of the grid points in the elevation interface into the grid elevation, and obtaining the thickness of the grid points in the elevation interface from the ground surface, that is, the surface thickness; according to the surface thickness, calculating the surface velocity at the grid point through the grid point surface velocity calculation module;

[0020] S4, repeating steps S1 to S3 to obtain the surface thickness and surface velocity at all grid points, and using a correction curve module to establish a thickness correction curve and a velocity correction curve;

[0021] S5, according to the thickness correction curve and the speed correction curve, using the layered model module to obtain a thickness correction formula and a speed correction formula, and establishing a final layered model after correction by the thickness correction formula and the speed correction formula;

[0022] The thickness correction formula obtained according to the thickness correction curve in step S4 is:

[0023] y1=0.811987x1-10.916624,

[0024] Among them, x1 is the thickness value before correction, and y1 is the thickness value after correction;

[0025] The speed correction formula obtained according to the speed correction curve is:

[0026] y2=1.35066x2-471.75570,

[0027] Among them, x2 is the speed value before correction, and y2 is the speed value after correction.

[0028] As a limitation, the velocity model in step S1 is a data body in segy format or a data body in text format, and the velocity model includes velocity information and grid size information.

[0029] As a second limitation, when calculating the velocity difference between adjacent longitudinal points in step S2, the length of the grid velocity model is first defined as L. x , width is L y , depth is H, define each grid length direction as △x, width direction as △y, vertical length as △z, then the number of grids in the length direction is L x / △x, the number of grids in the width direction is L y / △y, the number of vertical grids is H / △z;

[0030] Then, each grid in the velocity model is represented as (i, j, k), and the value range of i is The value range of j is The value range of k is

[0031] Next, the velocity difference between adjacent longitudinal points is expressed as:

[0032] △v1 i,j+1,k =v i,j+i,k -v i,j,k ,

[0033]

[0034] Among them, v is the speed of each grid, △v1 is the speed difference between two adjacent grids in the longitudinal direction, and △v2 is the enhancement value of the speed difference between two adjacent grids;

[0035] If the point of the enhanced value of the velocity difference between two adjacent grids satisfies: △v2 i,j+1,k >△v2 i,j,k And △v2 i,j+1,k >△v2i,j+2,k , then this point is the boundary grid point between the high-speed layer and the low-speed reduction layer, that is, the grid point that meets the conditions;

[0036] Finally, all grid points that meet the criteria are recorded.

[0037] As a third limitation, the method for extracting the boundary grid points between the final high-speed layer and the low-speed reduction layer in step S3 is: by comparing the existing surface survey point results within the initial arrival data range used in the tomographic inversion, calibrating the high-speed top interface elevation, and determining the grid points corresponding to the final selected elevation interface;

[0038] The conversion formula for converting the relative elevation of the boundary grid point to the grid elevation is:

[0039] ele i,j,k =Ej*△z,

[0040] Where E is the maximum elevation in the grid velocity model;

[0041] The surface thickness is:

[0042] h i,j,k =F i,j,k -ele i,j,k ,

[0043] Among them, F i,j,k is the surface elevation corresponding to the grid point (i, j, k).

[0044] As a fifth limitation, the surface velocity at the grid point in step S3 is expressed as:

[0045] V i,j,k =h i,j,k / t i,j,k ;

[0046]

[0047] t i ' ,j,k =△z / v i,j,k ;

[0048] Among them, t i,j,k is the time from the surface to the grid point (i, j, k), t i ' ,j,k is the time of the grid point (i, j, k).

[0049] As a sixth limitation, the method for establishing the thickness correction curve in step S4 is:

[0050] a1) Establishing a first group of forward models, which are composed of multiple sets of forward models with two layers of shallow surface medium and different thickness of low velocity reduction layer, and recording the surface thickness h' of the low velocity reduction layer of the first group of forward models respectively. i,j,k(p) , where p represents the pth set;

[0051] a2) Perform forward simulation based on the established multiple sets of forward models with different thicknesses, perform first arrival picking on the seismic data, obtain the tomographic inversion velocity model, and repeat steps S1 to S3 to obtain the grid elevation and surface thickness h at all grid points. i,j,k(p) ;

[0052] a3) Compare the surface thickness h obtained by inversion i,j,k(p) And the forward model low velocity reduction layer surface thickness h' i,j,k(p) The relationship between the thickness and the thickness correction curve is obtained by fitting using the least squares method:

[0053]

[0054] Wherein, the superscript T indicates transposition, -1 indicates inversion, k1 is the slope of the thickness correction curve, and c1 is the intercept of the thickness correction curve.

[0055] As a seventh limitation, the method for establishing the speed correction curve in step S4 is:

[0056] b1) Establishing a second group of forward models, which are composed of multiple sets of forward models with different low velocity reduction layers in two layers of shallow surface media, and recording the low velocity reduction layer velocity V' of the second group of forward models respectively. i,j,k(p) ;

[0057] b2) Perform forward simulation based on the established multiple sets of forward models with different velocities of low-velocity layers, perform first arrival picking on seismic data, obtain a tomographic inversion velocity model, and repeat steps S1 to S3 to obtain the model inversion surface velocity V at all grid points. i,j,k(p) ;

[0058] b3) Compare the surface velocity V obtained by inversion i,j,k(p) and the velocity V' of the low velocity reduction layer in the forward model i,j,k(p) The relationship between the speed correction curve is obtained by fitting the least square method:

[0059]

[0060] Wherein, k2 is the slope of the speed correction curve, and c2 is the intercept of the speed correction curve.

[0061] Due to the adoption of the above technical solution, the present invention has the following beneficial effects compared with the prior art:

[0062] (1) The device of the present invention solves the problems that the high-speed layer velocity used in extracting the constant-speed interface is subject to artificial intervention, the used velocity remains unchanged and cannot reflect the real high-speed top interface changes in a region, and when there are different high-speed layer velocities in a region, the constant-speed interface cannot be taken into account when extracting the constant-speed interface;

[0063] (2) The method of the present invention solves the problem of human intervention in the high-speed layer velocity used in extracting the constant-velocity interface by adopting a new method for extracting the elevation of the high-speed layer top interface. The velocity used can more truly reflect the actual high-speed top interface changes in a region. At the same time, when there are different high-speed layer velocities in a region, the high-speed layer velocity can be effectively extracted.

[0064] (3) The method of the present invention effectively improves the calculation accuracy of static correction and shallow surface modeling, and provides control and verification data for static correction of onshore petroleum seismic exploration data processing and shallow surface velocity modeling of prestack depth migration imaging.

[0065] In summary, the present invention can improve the accuracy of isovelocity interface extraction, thereby improving the accuracy of static correction calculation and shallow surface modeling, and providing control and verification data for static correction of onshore petroleum seismic exploration data processing and shallow surface velocity modeling of prestack depth migration imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0067] Figure 1 This is a principle block diagram of Embodiment 1 of the present invention;

[0068] Figure 2 This is a continuous speed model diagram of Example 2 of the present invention;

[0069] Figure 3 This is a longitudinal velocity difference curve diagram of Example 2 of the present invention;

[0070] Figure 4 This is a thickness correction curve diagram of Example 2 of the present invention;

[0071] Figure 5 This is a speed correction curve diagram of Example 2 of the present invention. DETAILED DESCRIPTION

[0072] In order to better explain the present invention and facilitate understanding, the preferred embodiments of the present invention are described in detail below through specific implementation modes in conjunction with the accompanying drawings.

[0073] Example 1 A layered modeling device for a tomographic model

[0074] like Figure 1As shown, this embodiment includes: a tomographic inversion velocity module, a velocity difference calculation module, an elevation interface extraction module, a grid point surface velocity calculation module, a correction curve module and a layered model module.

[0075] The tomographic inversion velocity module is used to perform tomographic inversion and output the tomographic inversion model as a grid velocity model, which is further output to the velocity difference calculation module; the velocity difference calculation module is used to calculate the longitudinal velocity difference of the grid velocity model and obtain the boundary grid points of all high-speed layers and low-speed reduction layers in the tomographic inversion velocity model of the entire work area, which are further output to the elevation interface extraction module; the elevation interface extraction module is used to extract the boundary grid points of the final high-speed layer and the low-speed reduction layer, determine the final selected elevation interface, convert the relative elevation of the grid points in the elevation interface into the grid elevation, and obtain the grid point table in the elevation interface. The layer thickness is calculated and output to the grid point surface velocity calculation module and the correction curve module respectively; the grid point surface velocity calculation module calculates the surface velocity at the grid point according to the surface thickness of the grid point in the elevation interface, and outputs it to the correction curve module; the correction curve module establishes the thickness correction curve and the velocity correction curve according to the results obtained by the elevation interface extraction module and the grid point surface velocity calculation module, and outputs them to the layered model module; the layered model module obtains the thickness correction formula and the velocity correction formula according to the thickness correction curve and the velocity correction curve, corrects the surface thickness and the surface velocity, and establishes the final layered model.

[0076] Example 2 A layered modeling method for a tomographic model

[0077] This embodiment provides a layered modeling method for a tomographic model, which needs to be implemented using the device in Example 1. This embodiment includes the following steps performed in sequence:

[0078] S1. Prepare a tomographic inversion velocity model. Perform tomographic inversion according to the first arrivals picked up from seismic data through a tomographic inversion velocity module. Output the obtained tomographic inversion model as a grid velocity model. The velocity model is a segy format data body or a text format data body. The velocity model includes velocity information and grid size information.

[0079] S2. Use the velocity difference calculation module to calculate the longitudinal velocity difference of the grid velocity model and obtain the boundary grid points of all high-speed layers and low-velocity reduction layers in the tomographic inversion velocity model of the entire work area.

[0080] When calculating the velocity difference between adjacent points in the longitudinal direction, first define the length of the grid velocity model as L x , width is L y , depth is H, define each grid length direction as △x, width direction as △y, vertical length as △z, then the number of grids in the length direction is L x / △x, the number of grids in the width direction is L y / △y, the number of vertical grids is H / △z;

[0081] Then, each grid in the velocity model is represented as (i, j, k), and the value range of i is The value range of j is The value range of k is

[0082] Next, the velocity difference between adjacent longitudinal points is expressed as:

[0083] △v1 i,j+1,k =v i,j+i,k -v i,j,k ,

[0084]

[0085] Among them, v is the speed of each grid, △v1 is the speed difference between two adjacent grids in the longitudinal direction, and △v2 is the enhancement value of the speed difference between two adjacent grids;

[0086] If the point of the enhanced value of the velocity difference between two adjacent grids satisfies: △v2 i,j+1,k >△v2 i,j,k And △v2 i,j+1,k >△v2 i,j+2,k , then this point is the boundary grid point between the high-speed layer and the low-speed reduction layer, that is, the grid point that meets the conditions;

[0087] Finally, all grid points that meet the criteria are recorded.

[0088] S3. Since there is more than one grid that meets the conditions in step S2 in a column of grids, it is necessary to use an elevation interface extraction module to extract the boundary grid points of the final high-speed layer and the low-speed reduction layer from the boundary grid points of all high-speed layers and low-speed reduction layers. At the same time, since the positions of the boundary grid points are relative positions in the tomography inversion model, they cannot be used for modeling without elevation information. Therefore, it is necessary to determine the final selected elevation interface, convert the relative elevation of the grid points in the elevation interface into grid elevation, and obtain the thickness of the grid points in the elevation interface from the ground surface, that is, the surface thickness.

[0089] The method for extracting the boundary grid points between the final high-speed layer and the low-speed reduction layer is as follows: by comparing the existing surface survey point results within the first arrival data range used in the tomographic inversion, calibrating the high-speed top interface elevation, and determining the grid points corresponding to the final selected elevation interface;

[0090] The conversion formula for converting the relative elevation of the boundary grid point to the grid elevation is:

[0091] ele i,j,k =Ej*△z,

[0092] Where E is the maximum elevation in the grid velocity model;

[0093] The surface thickness is:

[0094] h i,j,k =F i,j,k -ele i,j,k ,

[0095] Among them, F i,j,k is the surface elevation corresponding to the grid point (i, j, k).

[0096] According to the surface thickness, the surface velocity at the grid point is calculated by the grid point surface velocity calculation module;

[0097] The surface velocity at the grid point is expressed as:

[0098] V i,j,k =h i,j,k / t i,j,k ;

[0099]

[0100] t' i,j,k =△z / v i,j,k ;

[0101] Among them, t i,j,k is the time from the surface to the grid point (i, j, k), t' i,j,k is the time of the grid point (i, j, k).

[0102] S4, repeating steps S1 to S3 to obtain the surface thickness and surface velocity at all grid points, and using the correction curve module to establish a thickness correction curve and a velocity correction curve.

[0103] The method for establishing a thickness calibration curve comprises the following steps performed in sequence:

[0104] a1) Establishing a first group of forward models, which are composed of multiple sets of forward models with two layers of shallow surface medium and different thickness of low velocity reduction layer, and recording the surface thickness h' of the low velocity reduction layer of the first group of forward models respectively. i,j,k(p) , where the subscript p represents the pth set, p ≥ 1;

[0105] In the first group of forward models of this embodiment, ten sets of forward models with low velocity reduction layer thickness values ​​ranging from 20m to 40m were selected. The velocity of the low velocity reduction layer in each forward model is 750m / s, and the velocity of the high velocity layer is 2000m / s;

[0106] a2) Perform forward simulation based on the established multiple sets of forward models with different thicknesses, perform first arrival picking on the seismic data, obtain the tomographic inversion velocity model, and repeat steps S1 to S3 to obtain the grid elevation and surface thickness h at all grid points. i,j,k(p) ;

[0107] a3) Compare the surface thickness h obtained by inversion i,j,k(p) And the forward model low velocity reduction layer surface thickness h' i,j,k(p) The relationship between the thickness and the thickness correction curve is obtained by fitting using the least squares method:

[0108]

[0109] Wherein, the superscript T indicates transposition, -1 indicates inversion, k1 is the slope of the thickness correction curve, and c1 is the intercept of the thickness correction curve;

[0110] The thickness correction formula obtained from the thickness correction curve is:

[0111] y1=0.811987x1-10.916624,

[0112] Among them, x1 is the thickness value before correction, and y1 is the thickness value after correction.

[0113] The method for establishing a speed correction curve comprises the following steps performed in sequence:

[0114] b1) Establishing a second group of forward models, which are composed of multiple sets of forward models with different low velocity reduction layers in two layers of shallow surface media, and recording the low velocity reduction layer velocity V' of the second group of forward models respectively. i,j,k(p) ;

[0115] In the second group of forward models of this embodiment, three forward models with low velocity reduction layer thickness of 50m, 100m, and 150m and ten forward models with low velocity reduction layer velocities were selected, with a total of thirty sets of forward models. The velocity value of the low velocity reduction layer was selected in the range of 500m / s to 1700m / s; the velocity of the high-speed layer was 2000m / s;

[0116] b2) Perform forward simulation based on the established multiple sets of forward models with different velocities of low-velocity layers, perform first arrival picking on seismic data, obtain a tomographic inversion velocity model, and repeat steps S1 to S3 to obtain the model inversion surface velocity V at all grid points. i,j,k(p) ;

[0117] b3) Compare the surface velocity V obtained by inversion i,j,k(p) and the forward model low velocity layer velocity V i ' ,j,k(p) The relationship between the speed correction curve is obtained by fitting the least square method:

[0118]

[0119] Wherein, k2 is the slope of the speed correction curve, and c2 is the intercept of the speed correction curve;

[0120] The speed correction formula obtained from the speed correction curve is:

[0121] y2=1.35066x2-471.75570,

[0122] Among them, x2 is the speed value before correction, and y2 is the speed value after correction.

[0123] S5. According to the thickness correction curve and the speed correction curve, a thickness correction formula and a speed correction formula are obtained by using a layered model module, and a final layered model is established after correction by the thickness correction formula and the speed correction formula.

[0124] Figure 2 is a continuous velocity model diagram obtained by tomographic inversion obtained in step S1, in which the horizontal axis corresponds to the length of the continuous velocity model and the vertical axis corresponds to the depth of the continuous velocity model, in units of m. Figure 2 On the right side are the speeds corresponding to the colors at different depths, in m / s.

[0125] Figure 3 The difference curve is obtained by calculating the velocity model difference in step S2. The horizontal axis represents the number of tracks and the vertical axis represents the depth. When calibrating the elevation of the high-speed top interface, the survey point results within the data range are compared with Figure 3 The curve with the highest degree of overlap corresponds to the curve in , and the curve with the highest degree of overlap is selected as the elevation of the high-speed top interface.

[0126] Figure 4 It is the thickness correction curve established in step S4, the horizontal axis is the surface thickness of the inversion model, and the vertical axis is the surface thickness of the forward model. In order to ensure the accuracy of the experimental results, this embodiment performs multiple calculations on each model in the first group of forward models, and obtains multiple inversion thickness data for the same forward model thickness. The figure selects the points corresponding to the inversion model thickness where the inversion thickness is stable within a certain range.

[0127] Figure 5 It is the velocity correction curve established in step S4, the horizontal axis is the surface velocity of the inversion model, and the vertical axis is the low velocity reduction layer velocity of the forward model. In order to ensure the accuracy of the experimental results, this embodiment performs multiple calculations on each model in the second group of forward models, and obtains multiple inverted low velocity reduction layer velocity data for the same forward model low velocity reduction layer velocity. The figure selects the points corresponding to the inversion model low velocity reduction layer velocity where the inverted low velocity reduction layer is stable within a certain range.

Claims

1. A layered modeling device for a tomographic model, characterized in that: The device comprises: a tomographic inversion velocity module, a velocity difference calculation module, an elevation interface extraction module, a grid point surface velocity calculation module, a correction curve module and a layered model module; A tomographic inversion velocity module is used to perform tomographic inversion and output the tomographic inversion model as a grid velocity model, which is further output to a velocity difference calculation module; The velocity difference calculation module is used to calculate the longitudinal velocity difference of the grid velocity model and obtain the boundary grid points of all high-speed layers and low-velocity reduction layers in the tomographic inversion velocity model of the entire work area, and further output them to the elevation interface extraction module; The elevation interface extraction module is used to extract the boundary grid points between the final high-speed layer and the low-speed reduction layer, determine the final selected elevation interface, convert the relative elevation of the grid points in the elevation interface into the grid elevation, obtain the surface thickness of the grid points in the elevation interface, and output them to the grid point surface velocity calculation module and the correction curve module respectively; The grid point surface velocity calculation module calculates the surface velocity at the grid point according to the surface thickness of the grid point in the elevation interface and outputs it to the correction curve module; The correction curve module establishes a thickness correction curve and a velocity correction curve according to the results obtained by the elevation interface extraction module and the grid point surface velocity calculation module, and outputs them to the layered model module; The layered model module obtains the thickness correction formula and the speed correction formula according to the thickness correction curve and the speed correction curve, corrects the surface thickness and the surface speed, and establishes the final layered model.

2. A layered modeling method for a tomographic model, implemented by the layered modeling device for a tomographic model according to claim 1, the method comprising the following steps performed in sequence: S1. Prepare a tomographic inversion velocity model, perform tomographic inversion according to the first arrivals picked up from seismic data through a tomographic inversion velocity module, and output the obtained tomographic inversion model as a grid velocity model; S2, using the velocity difference calculation module to calculate the longitudinal velocity difference of the grid velocity model, and obtain the boundary grid points of all high-speed layers and low-speed reduction layers in the tomographic inversion velocity model of the entire work area; S3, using the elevation interface extraction module, extracting the boundary grid points of the final high-speed layer and the low-speed reduction layer from the boundary grid points of all high-speed layers and low-speed reduction layers, determining the final selected elevation interface, converting the relative elevation of the grid points in the elevation interface into the grid elevation, and obtaining the thickness of the grid points in the elevation interface from the ground surface, that is, the surface thickness; according to the surface thickness, calculating the surface velocity at the grid point through the grid point surface velocity calculation module; S4, repeating steps S1 to S3 to obtain the surface thickness and surface velocity at all grid points, and using a correction curve module to establish a thickness correction curve and a velocity correction curve; S5, according to the thickness correction curve and the speed correction curve, using the layered model module to obtain a thickness correction formula and a speed correction formula, and establishing a final layered model after correction by the thickness correction formula and the speed correction formula; The thickness correction formula obtained according to the thickness correction curve in step S4 is: y1=0.811987x1-10.916624, Among them, x1 is the thickness value before correction, and y1 is the thickness value after correction; The speed correction formula obtained according to the speed correction curve is: y2=1.35066x2-471.75570, Among them, x2 is the speed value before correction, and y2 is the speed value after correction.

3. The layered modeling method of a tomographic model according to claim 2, characterized in that: The velocity model in step S1 is a data body in segy format or a data body in text format, and the velocity model includes velocity information and grid size information.

4. The layered modeling method of a tomographic model according to claim 3, characterized in that: When calculating the velocity difference of adjacent points in the longitudinal direction in step S2, the length of the grid velocity model is first defined as L. x , width is L y , depth is H, define each grid length direction as △x, width direction as △y, vertical length as △z, then the number of grids in the length direction is L x / △x, the number of grids in the width direction is L y / △y, the number of vertical grids is H / △z; Then, each grid in the velocity model is represented as (i, j, k), and the value range of i is The value range of j is The value range of k is Next, the velocity difference between adjacent longitudinal points is expressed as: △v1 i,j+1,k =v i,j+i,k -v i,j,k , Among them, v is the speed of each grid, △v1 is the speed difference between two adjacent grids in the longitudinal direction, and △v2 is the enhancement value of the speed difference between two adjacent grids; If the point of the enhanced value of the velocity difference between two adjacent grids satisfies: △v2 i,j+1,k >△v2 i,j,k And △v2 i,j+1,k >△v2 i,j+2,k , then this point is the boundary grid point between the high-speed layer and the low-speed reduction layer, that is, the grid point that meets the conditions; Finally, all grid points that meet the criteria are recorded.

5. The layered modeling method of a tomographic model according to claim 4, characterized in that: The method for extracting the boundary grid points between the final high-speed layer and the low-speed reduction layer in step S3 is: by comparing the existing surface survey point results within the initial arrival data range used in the tomographic inversion, calibrating the high-speed top interface elevation, and determining the grid points corresponding to the final selected elevation interface; The conversion formula for converting the relative elevation of the boundary grid point to the grid elevation is: he i,j,k =Ej*△z, Where E is the maximum elevation in the grid velocity model; The surface thickness is: h i,j,k =F i,j,k -he i,j,k , Among them, F i,j,k is the surface elevation corresponding to the grid point (i, j, k).

6. The layered modeling method of a tomographic model according to claim 5, characterized in that: The surface velocity at the grid point in step S3 is expressed as: V i,j,k =h i,j,k / t i,j,k ; t' i,j,k =△z / v i,j,k ; Among them, t i,j,k is the time from the surface to the grid point (i, j, k), t' i,j,k is the time of the grid point (i, j, k).

7. The layered modeling method of a tomographic model according to claim 6, characterized in that: The method for establishing the thickness calibration curve in step S4 is: a1) Establishing a first group of forward models, which are composed of multiple sets of forward models with two layers of shallow surface medium and different thickness of low velocity reduction layer, and recording the surface thickness h' of the low velocity reduction layer of the first group of forward models respectively. i,j,k(p) , where p represents the pth set; a2) Perform forward simulation based on the established multiple sets of forward models with different thicknesses, perform first arrival picking on the seismic data, obtain the tomographic inversion velocity model, and repeat steps S1 to S3 to obtain the grid elevation and surface thickness h at all grid points. i,j,k(p) ; a3) Compare the surface thickness h obtained by inversion i,j,k(p) And the forward model low velocity reduction layer surface thickness h' i,j,k(p) The relationship between the thickness and the thickness correction curve is obtained by fitting using the least squares method: Wherein, the superscript T indicates transposition, -1 indicates inversion, k1 is the slope of the thickness correction curve, and c1 is the intercept of the thickness correction curve.

8. The layered modeling method of a tomographic model according to claim 6, characterized in that: The method for establishing the speed correction curve in step S4 is: b1) Establishing a second group of forward models, which are composed of multiple sets of forward models with different low velocity reduction layers in two layers of shallow surface media, and recording the low velocity reduction layer velocity V' of the second group of forward models respectively. i,j,k(p) ; b2) Perform forward simulation based on the established multiple sets of forward models with different velocities of low-velocity layers, perform first arrival picking on seismic data, obtain a tomographic inversion velocity model, and repeat steps S1 to S3 to obtain the model inversion surface velocity V at all grid points. i,j,k(p) ; b3) Comparison of the surface velocity V obtained by inversion i,j,k(p) and the velocity V' of the low velocity reduction layer in the forward model i，j，k(p) The relationship between the speed correction curve is obtained by fitting the least square method: Wherein, k2 is the slope of the speed correction curve, and c2 is the intercept of the speed correction curve.

Images