A fault phased modeling method based on focal mechanism constraint

Abstract

The present application relates to the technical field of geophysical microseismic monitoring, and particularly relates to a fault phased modeling method based on a focal mechanism constraint, which comprises the following steps: obtaining microseismic catalog and focal mechanism solution data; identifying linear bodies, clustering the linear bodies, fitting and extracting planar faults, screening out planar faults with similar focal mechanisms, and constructing a three-dimensional planar fault model; eliminating events corresponding to planar faults with similar focal mechanisms from the microseismic catalog to generate a remaining microseismic catalog; clustering the remaining microseismic events, merging clusters with similar planar solutions, and fitting to obtain a three-dimensional non-planar fault model; and fusing the two fault models to obtain a final three-dimensional fault model based on phased modeling. The phased modeling strategy of first constructing planar faults and then constructing non-planar faults effectively overcomes the problems of over-planarization and false merging of a single modeling method.

Description

Technical Field

[0001] This invention relates to the field of geophysical microseismic monitoring technology, and in particular to a fault stage modeling method based on source mechanism constraints. Background Technology

[0002] Precise characterization of fault geometry is crucial for understanding earthquake gestation and rupture processes, assessing seismic activity, and conducting earthquake hazard risk assessments. In recent years, advancements in high-density seismic network deployment and AI-powered seismic detection technologies have significantly improved the completeness and location accuracy of microseismic catalogs, making it possible to effectively extract fault geometric information from the spatial distribution of seismic events.

[0003] Fault modeling methods based on microseismic catalogs include planar fault fitting methods and non-planar (surface) reconstruction methods. Planar fault fitting methods can quickly provide fault strike, dip angle, and spatial extent, but the single-planar assumption can easily lead to oversimplification. Non-planar fault reconstruction methods can obtain more complex three-dimensional fault surfaces, but directly performing non-planar modeling on all microseismic events can easily complicate fault segments that should be represented by simple planes. In addition, existing methods mostly rely solely on spatial location for event clustering, without considering focal mechanisms as constraints for fault modeling.

[0004] Therefore, it is necessary to provide a fault staged modeling method based on focal mechanism constraints to solve the above-mentioned technical problems. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a phased fault modeling method based on focal mechanism constraints. This method first performs planar fault modeling, followed by non-planar fault modeling. During the modeling process, this invention introduces focal mechanism constraints to resolve fault modeling, overcoming the problem of fault mis-merging that may be caused by relying solely on spatial location clustering, and utilizing focal mechanism information to improve the geometric rationality and geological interpretability of the fault model.

[0006] This invention provides a staged fault modeling method based on focal mechanism constraints, comprising: S1: Obtain the microseismic catalog and focal mechanism solution data, match the focal mechanism solution data to the corresponding events in the microseismic catalog, and convert the latitude, longitude and depth coordinate systems to the local Cartesian coordinate system; S2: Linear bodies are identified using three-dimensional Hough transform, and the linear bodies are clustered using the DBSCAN algorithm. Three-dimensional plane faults and their geometric parameters are extracted by principal component analysis, and plane faults with similar source mechanisms are selected to construct a three-dimensional plane fault model. S3: Remove events corresponding to plane faults with similar focal mechanisms from the microseismic catalog to generate the remaining microseismic catalog; S4: The HDBSCAN algorithm is used to cluster the remaining microseismic events. Clusters are merged based on the similarity of fault plane solutions between adjacent clusters. Then, the support vector regression algorithm is used to fit and obtain a three-dimensional non-planar fault model. S5: Spatially superimpose and fuse the planar fault model and the non-planar fault model to obtain the final three-dimensional fault model based on phased modeling.

[0007] Optionally, in S1, the microseismic catalog includes microseismic event number, occurrence time, latitude and longitude, depth, magnitude and location error, and the focal mechanism solution data includes latitude and longitude, depth, strike, dip and slip angle.

[0008] Optionally, in S2, the DBSCAN algorithm process includes: Set the neighborhood radius and minimum number of events, and calculate the mixed distance between any two events; Starting from any point, events that are less than the neighborhood radius of that point are grouped into the same cluster, and the number of events in each cluster is greater than or equal to the minimum number of events.

[0009] Optionally, in S2, after fitting and extracting the three-dimensional planar faults, a planar standard judgment is also performed. If the fitting result meets the planar standard, the geometric parameters are retained and calculated. If it does not meet the standard, the corresponding clusters are re-clustered and fitted.

[0010] Optionally, in S2, a plane fault with similar focal mechanisms is defined as a fault with μ ≤ 30° and containing two or more events with focal mechanism solutions; the value of μ is calculated through the following steps: (1) Extract all available focal mechanism solutions within the cluster containing the plane fault; (2) Solve for the nodal surface normal vector of each focal mechanism solution; (3) Combine all nodal surface normal vectors within the cluster by weighting them according to magnitude, with a weight value of 10. aM Where M is the magnitude and a is a preset constant; (4) Calculate the angle between the composite vector of all the above nodal surface normal vectors and the normal vector of the plane fault surface, which is the μ value of the plane fault.

[0011] Optionally, in S4, the HDBSCAN algorithm process includes: Set a minimum number of events and calculate the blending distance between any two events; Construct a minimum spanning tree, break edges according to the mixed distance from largest to smallest, and obtain the cluster.

[0012] Optionally, both the DBSCAN algorithm in S2 and the HDBSCAN algorithm in S4 use a mixed distance as a measure of similarity between events. The expression for the mixed distance is as follows: ; in, Indicates the mixing distance. For normalized Euclidean distance, To account for differences in focal mechanisms, when both events have focal mechanism solutions, we take the normalized Kagan angle, which is the minimum three-dimensional rotation angle between the two focal mechanism solutions. , Let each represent a corresponding weight coefficient, satisfying the following conditions: + =1, The larger the magnitude, the stronger the constraint of the focal mechanism on the clustering results.

[0013] Optionally, in S4, the support vector regression algorithm maps the input data to a high-dimensional space through a kernel function, performs linear regression in the high-dimensional space, and fits the non-planar tomography to obtain the result.

[0014] Optionally, in S4, an improved inverse power-law relationship is used to analyze the variation of microseismic event density with distance from the fault plane; for each fault and its microseismic events, an inverse power-law formula is fitted, as shown in the following expression: ; Among them, distance Logarithmically spaced bins are used. The number of microseismic events within each distance interval. The number of microseismic events at the fault plane. For fault distances applicable to the inverse power law, The decay index; The parameters obtained by fitting Used to assist in judging the degree to which a model fault plane represents seismic activity, when When the error is significantly greater than the source location error, there is a true fault complexity that is not represented by the model. In this case, the cluster to which the fault belongs should be re-clustered and fitted.

[0015] The application of the technical solution of the present invention has the following beneficial effects: This invention provides a staged fault modeling method based on focal mechanism constraints. The method employs staged modeling, first extracting planar faults with similar focal mechanisms, and then reconstructing the remaining events using non-planar surfaces. This preserves fault segments with good planarity while avoiding the overcomplication of simple planar structures by non-planar modeling, thus balancing model simplicity and precision. Furthermore, by introducing focal mechanism solution constraints, this method integrates focal mechanism solution information throughout the entire process of planar fault modeling, selection, and non-planar fault modeling. This effectively distinguishes between parallel or conjugate fault groups, overcoming the fault malmerging problem caused by relying solely on spatial location clustering, and improving the geometric rationality and geological interpretability of the fault model. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0017] Figure 1 This is a flowchart of the steps of the fault staged modeling method based on source mechanism constraints in a preferred embodiment of the present invention; Figure 2 This is a schematic diagram of a three-dimensional planar fault model in a preferred embodiment of the present invention; Figure 3 This is a schematic diagram of a three-dimensional non-planar fault model in a preferred embodiment of the present invention; Figure 4 This is a schematic diagram of a three-dimensional fault model based on phased modeling in a preferred embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1: like Figure 1 As shown, this embodiment provides a staged fault modeling method based on focal mechanism constraints, including (steps S1-S5): S1: Obtain the microseismic catalog and focal mechanism solution data, match the focal mechanism solution data to the corresponding events in the microseismic catalog, and convert the latitude, longitude and depth coordinate systems to the local Cartesian coordinate system.

[0020] In this embodiment, the microseismic catalog includes microseismic event number, occurrence time, latitude and longitude, depth, magnitude and location error, and focal mechanism solution data includes latitude and longitude, depth, strike, dip and slip angle.

[0021] S2: Linear bodies are identified using the three-dimensional Hough transform, and the linear bodies are clustered using the DBSCAN algorithm. Three-dimensional plane faults and their geometric parameters are extracted by principal component analysis, and plane faults with similar source mechanisms are selected to construct a three-dimensional plane fault model.

[0022] Specifically, based on the three-dimensional spatial coordinates of microseismic events, the three-dimensional Hough transform is used to identify linear bodies. The identified linear bodies are classified into acceptable linear bodies and unacceptable linear bodies according to their orientation. Events in unacceptable linear bodies are reassigned to the nearest acceptable linear bodies. The DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm is a density-based clustering algorithm. The calculation process for clustering each linear entity using this algorithm is as follows: Set the neighborhood radius to 1 and the minimum number of events to 10. Calculate the normalized Euclidean distance and normalized Kagan angle between any two events to obtain the mixture distance between events. Starting from any point, group events whose distance to that point is less than the neighborhood radius into the same cluster. The number of events in each cluster is greater than or equal to the minimum number of events. The expression for the mixture distance is as follows: ; in, Indicates the mixing distance. For normalized Euclidean distance, To account for differences in focal mechanisms, when both events have focal mechanism solutions, a normalized Kagan angle is used. The Kagan angle is the minimum three-dimensional rotation angle between the two focal mechanism solutions. In this step, weighting coefficients are set. =0.5、 =0.5.

[0023] Three-dimensional planar faults and their geometric parameters are extracted by fitting using principal component analysis. Planar standard judgment is then performed. If the fitting result meets the planar standard, the geometric parameters are retained and calculated. If not, the corresponding clusters are re-clustered and fitted. The geometric parameters include the length, width, orientation, and dip of the three-dimensional planar faults.

[0024] Planar faults with similar focal mechanisms are selected, and three-dimensional planar fault models are constructed. A planar fault with similar focal mechanisms is defined as a fault with μ ≤ 30° and containing two or more events with focal mechanism solutions; the value of μ is calculated through the following steps: (1) Extract all available focal mechanism solutions within the cluster containing the plane fault; (2) Solve for the nodal surface normal vector of each focal mechanism solution; (3) Combine all nodal surface normal vectors within the cluster by weighting them according to magnitude, with a weight value of 10. aM Where M is the magnitude and a is a preset constant, and in this embodiment a value of 1; (4) Calculate the angle between the composite vector of all the above nodal surface normal vectors and the normal vector of the plane fault surface, which is the μ value of the plane fault.

[0025] The three-dimensional planar fault model constructed in this embodiment is as follows: Figure 2 As shown.

[0026] S3: Remove events corresponding to plane faults with similar focal mechanisms from the microseismic catalog to generate the remaining microseismic catalog.

[0027] S4: The HDBSCAN algorithm is used to cluster the remaining microseismic events. Clusters are merged based on the similarity of fault plane solutions between adjacent clusters. Then, the support vector regression algorithm is used to fit and obtain a three-dimensional non-planar fault model.

[0028] Specifically, the HDBSCAN (Hierarchical Density-Based Spatial Clustering of Applications with Noise) algorithm is a hierarchical density clustering algorithm. The algorithm's calculation process is as follows: A minimum number of events is set to 40; the normalized Euclidean distance and normalized Kagan angle between any two events are calculated; and weighting coefficients are set... =0.7、 =0.3, obtain the mixed distance between events; construct the minimum spanning tree, break the edges according to the mixed distance from large to small, and obtain the cluster.

[0029] Support vector regression algorithm maps input data to a high-dimensional space through kernel function, performs linear regression in the high-dimensional space, and fits the non-planar fault to obtain the result. Furthermore, an improved inverse power-law relationship was used to analyze the variation of microseismic event density with distance from the fault plane; for each fault and its microseismic events, an inverse power-law was fitted, the expression of which is as follows: ; Among them, distance Logarithmically spaced bins are used. The number of microseismic events within each distance interval. The number of microseismic events at the fault plane. For fault distances applicable to the inverse power law, The decay exponent. The parameters obtained from the fitting. It can help determine the degree to which the model fault plane represents seismic activity. In this embodiment, when When the error is greater than twice the source location error, the cluster to which the fault belongs is re-clustered and refitted to obtain, as shown below. Figure 3 The three-dimensional non-planar fault model shown.

[0030] S5: Spatially superimpose and fuse the planar fault model and the non-planar fault model to obtain the final three-dimensional fault model based on phased modeling. The three-dimensional fault model based on phased modeling is as follows: Figure 4 As shown.

[0031] This embodiment provides a staged fault modeling method based on focal mechanism constraints. By first constructing planar faults and then non-planar faults, it effectively overcomes the problems of over-planarization and erroneous merging inherent in single modeling methods. The three-dimensional planar fault model constructed using this method clearly depicts fault segments with similar focal mechanisms, while the three-dimensional non-planar fault model meticulously describes fault segments with complex geometries. The fused three-dimensional fault model ensures both overall consistency of fault geometry and representation of local structural details.

[0032] The above description is only a preferred embodiment of the present invention and does not limit the scope of the present invention. All equivalent structural transformations made under the inventive concept of the present invention using the contents of the present invention specification and drawings, or direct / indirect applications in other related technical fields, are included within the protection scope of the present invention.

Claims

1. A staged fault modeling method based on focal mechanism constraints, characterized in that, include: S1: Obtain the microseismic catalog and focal mechanism solution data, match the focal mechanism solution data to the corresponding events in the microseismic catalog, and convert the latitude, longitude and depth coordinate systems to the local Cartesian coordinate system; S2: Linear bodies are identified using three-dimensional Hough transform, and the linear bodies are clustered using the DBSCAN algorithm. Three-dimensional plane faults and their geometric parameters are extracted by principal component analysis, and plane faults with similar source mechanisms are selected to construct a three-dimensional plane fault model. S3: Remove events corresponding to plane faults with similar focal mechanisms from the microseismic catalog to generate the remaining microseismic catalog; S4: The HDBSCAN algorithm is used to cluster the remaining microseismic events. Clusters are merged based on the similarity of fault plane solutions between adjacent clusters. Then, the support vector regression algorithm is used to fit and obtain a three-dimensional non-planar fault model. S5: Spatially superimpose and fuse the planar fault model and the non-planar fault model to obtain the final three-dimensional fault model based on phased modeling.

2. The fault staged modeling method based on focal mechanism constraints according to claim 1, characterized in that, In S1, the microseismic catalog includes microseismic event number, occurrence time, latitude and longitude, depth, magnitude and location error, and focal mechanism solution data includes latitude and longitude, depth, strike, dip and slip angle.

3. The fault staged modeling method based on focal mechanism constraints according to claim 1, characterized in that, In S2, the DBSCAN algorithm process includes: Set the neighborhood radius and minimum number of events, and calculate the mixed distance between any two events; Starting from any point, events that are less than the neighborhood radius of that point are grouped into the same cluster, and the number of events in each cluster is greater than or equal to the minimum number of events.

4. The fault staged modeling method based on focal mechanism constraints according to claim 1, characterized in that, In S2, after fitting and extracting the three-dimensional planar faults, a planar standard judgment is also performed. If the fitting result meets the planar standard, the geometric parameters are retained and calculated. If it does not meet the standard, the corresponding clusters are re-clustered and fitted.

5. The fault staged modeling method based on focal mechanism constraints according to claim 1, characterized in that, In S2, a plane fault with similar focal mechanisms is defined as a fault with μ ≤ 30° and containing two or more events with focal mechanism solutions; the value of μ is calculated through the following steps: (1) Extract all available focal mechanism solutions within the cluster containing the plane fault; (2) Solve for the nodal surface normal vector of each focal mechanism solution; (3) Combine all nodal surface normal vectors within the cluster by weighting them according to magnitude, with a weight value of 10. aM Where M is the magnitude and a is a preset constant; (4) Calculate the angle between the composite vector of all the above nodal surface normal vectors and the normal vector of the plane fault surface, which is the μ value of the plane fault.

6. The fault staged modeling method based on focal mechanism constraints according to claim 1, characterized in that, In S4, the HDBSCAN algorithm process includes: Set a minimum number of events and calculate the blending distance between any two events; Construct a minimum spanning tree, break edges according to the mixed distance from largest to smallest, and obtain the cluster.

7. The fault staged modeling method based on focal mechanism constraints according to claim 1, characterized in that, Both the DBSCAN algorithm in S2 and the HDBSCAN algorithm in S4 use mixed distance as a measure of similarity between events. The expression for the mixed distance is as follows: ； in, Indicates the mixed distance. For normalized Euclidean distance, To account for differences in focal mechanisms, when both events have focal mechanism solutions, we take the normalized Kagan angle, which is the minimum three-dimensional rotation angle between the two focal mechanism solutions. , Let each represent the corresponding weight coefficient, satisfying... + =1.

8. The fault staged modeling method based on focal mechanism constraints according to claim 1, characterized in that, In S4, the support vector regression algorithm maps the input data to a high-dimensional space through a kernel function, performs linear regression in the high-dimensional space, and fits the non-planar fault to obtain the result.

9. The fault staged modeling method based on focal mechanism constraints according to claim 1, characterized in that, In S4, an improved inverse power law relationship is used to analyze the variation of microseismic event density with distance from the fault plane; for each fault and its microseismic events, an inverse power law formula is fitted, as shown in the following expression: ； Among them, distance Logarithmically spaced bins are used. The number of microseismic events within each distance interval. The number of microseismic events at the fault plane. For fault distances applicable to the inverse power law, The decay index; The parameters obtained by fitting Used to assist in judging the degree to which a model fault plane represents seismic activity, when When the error is significantly greater than the source location error, there is a true fault complexity that is not represented by the model. In this case, the cluster to which the fault belongs should be re-clustered and fitted.

Images