Live fault finite fault source model construction method and system

Abstract

The invention provides an active fault finite fault source model construction method and system, and relates to the technical field of seismic engineering, and the method comprises the steps: collecting seismic observation data, tremor monitoring data and borehole wave velocity test data; geometric parameters such as fault length, width, burial depth, trend and inclination angle are extracted; calculating scalar seismic moments and moment magnitudes; fault plane sliding distribution is established by adopting a concave-convex body model and a k-square sliding model; constructing a three-dimensional fault dynamic model by adopting a double couple point source superposition method; according to the method, a coupling mechanism of geometric parameters and kinetic parameters of the fault is established, and complete seismic source input parameters are provided for near-fault strong ground motion simulation.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of earthquake engineering and seismology, specifically to a method and system for constructing a finite fault source model of an active fault. Background Technology

[0002] Active faults are faults that were active during the Quaternary period and may shift again in the future. Assessing their potential seismic hazard is crucial for the site selection and seismic fortification of major engineering projects. my country is located at the intersection of the Circum-Pacific Seismic Belt and the Eurasian Seismic Belt, with a wide distribution of active faults and a history of numerous destructive earthquakes. With the rapid development of the national economy, an increasing number of major infrastructure projects are being built near or across active faults, making seismic hazard assessment of active faults increasingly urgent.

[0003] Near-fault ground motion simulation is a key technique for assessing the seismic hazard of active faults. Unlike far-field ground motions, near-fault ground motions exhibit significant directional effects, hanging wall effects, and velocity pulse characteristics. These characteristics are primarily controlled by the fault's geometry, rupture process, and slip distribution. Traditional empirical attenuation relationships are insufficient to accurately describe the complex characteristics of near-fault ground motions, while physical model-based numerical simulation methods can more realistically reproduce the generation mechanism of near-fault ground motions. Establishing a source model is a prerequisite and core element for conducting ground motion numerical simulations; the accuracy of the source model directly determines the reliability of the simulation results.

[0004] The finite-fault source model is a standard method for describing the rupture process of large earthquakes. This model discretizes the fault plane into multiple sub-faults, each with independent parameters such as slip magnitude, slip direction, rupture time, and rise time. By superimposing the contributions of all sub-faults, the ground motion time history at any point can be calculated. Constructing a finite-fault source model requires determining the geometric and dynamic parameters of the fault. The former includes fault length, width, depth, strike, and dip, while the latter includes slip distribution, rupture velocity, and source time function.

[0005] In the prior art, Chinese invention CN120065335A discloses a method and system for constructing an earthquake rupture fault model and a coseismic displacement field. This method acquires GNSS data, strong-motion meter data, and accelerometer data from a preset area, fuses the multi-source data to obtain multi-source earthquake data, then performs seismic waveform inversion to obtain crustal parameters, calculates the Green's function using a layered planar dislocation model to obtain the earthquake rupture fault model, and finally obtains the earthquake rupture surface through gradient algorithm inversion and analytically derives the coseismic displacement field. This method has certain technical value in the inversion of source parameters of already occurred earthquakes.

[0006] However, the aforementioned existing technologies have the following technical problems: First, this method is mainly for post-event analysis of earthquakes that have already occurred, relying on actual observation data for inversion, and cannot be directly applied to the source model prediction of future scenario earthquakes, while the construction of source models for scenario earthquakes is a core requirement for seismic hazard assessment of engineering sites; Second, this method does not consider the coupling mechanism between fault geometric parameters and dynamic parameters. There is an inherent physical constraint relationship between the geometric characteristics of the fault and the slip distribution. Ignoring this coupling relationship will lead to physical inconsistencies in the source model; Third, the coseismic displacement field output by this method is mainly used for static deformation analysis, and does not provide key dynamic parameters required for seismic motion simulation, such as the source time function, making it difficult to directly apply to the numerical calculation of seismic motion time histories; Fourth, this method does not involve the stochastic modeling of fault slip distribution. The slip distribution of actual earthquakes has significant non-uniformity and stochastic characteristics, and deterministic slip distribution models cannot reflect this complexity and uncertainty.

[0007] Therefore, existing technologies have significant shortcomings in constructing source models for earthquakes occurring on active faults. They struggle to provide accurate and complete source input parameters for simulating strong ground motion near faults, thus limiting the accuracy and reliability of active fault earthquake hazard assessment. There is an urgent need to develop a finite fault source model construction method that can comprehensively utilize multi-source seismic geological data, establish the coupling relationship between fault geometry and dynamic parameters, and output complete source parameters. Summary of the Invention

[0008] The technical problem solved by this invention is to provide a method and system for constructing a finite fault source model of an active fault, so as to solve the problems in the prior art of difficulty in establishing a source model, unclear coupling mechanism of fault geometry and dynamic parameters, and inability to provide complete source parameters for earthquake motion numerical simulation.

[0009] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0010] The method for constructing a finite-source seismic model of an active fault includes: a multi-source seismic geological data acquisition step, which acquires seismic observation data, ground motion monitoring data, and borehole wave velocity test data of the target active fault area; performing spatiotemporal benchmark unification processing and quality verification on the acquired data to generate a multi-source seismic geological dataset; a fault geometric parameter extraction step, which, based on the multi-source seismic geological dataset, uses a joint inversion method to extract the fault length, fault width, fault depth, fault strike, and fault dip angle of the target active fault, and determines the site medium parameters based on the borehole wave velocity test data; a scalar seismic moment calculation step, which calculates the fault area and shear modulus based on the fault length, fault width, and site medium parameters, calculates the scalar seismic moment by combining the average slip determined by empirical statistical relationships, and determines the moment magnitude based on the scalar seismic moment; and a fault plane slip distribution modeling step, which models the fault plane slip distribution based on the moment... The magnitude and fault geometry parameters discretize the fault plane into multiple sub-faults. A concave-convex body model is used to determine the location and extent of high-slip regions. A k-squared slip model is used to generate random slip components for each sub-fault. The concave-convex body slip components are superimposed with the random slip components to generate the fault plane slip distribution matrix. In the dual-couple point source superposition step, each sub-fault is equivalent to a dual-couple point source. The moment tensor components of each dual-couple point source are calculated based on the fault strike, dip angle, and slip angle to determine the rupture initiation time and rise time of each dual-couple point source. A three-dimensional fault dynamics model is constructed by spatially superimposing the dual-couple point sources. In the source parameter output step, the time release process of each dual-couple point source in the three-dimensional fault dynamics model is integrated over time to generate a source time function. The fault plane slip distribution matrix is ​​then converted into a spatial slip distribution data format for output.

[0011] Preferably, the seismic observation data includes historical earthquake catalog data and seismic waveform record data, the ground motion monitoring data includes site micromotion observation data and horizontal and vertical spectral ratio data, and the borehole wave velocity test data includes shear wave velocity profile data and compression wave velocity profile data.

[0012] Preferably, the fault length ranges from 1km to 300km, the fault width ranges from 1km to 50km, and the fault depth ranges from 0km to 30km.

[0013] Preferably, the size of the sub-fault is adaptively determined based on the moment magnitude, and the area of ​​the concave-convex body accounts for 15% to 25% of the total fault area.

[0014] A system for constructing a finite fault source model of an active fault includes: a multi-source seismic geological data acquisition module, a fault geometric parameter extraction module, a scalar seismic moment calculation module, a fault plane slip distribution modeling module, a dual-couple point source superposition module, and a source parameter output module. These modules are connected in sequence to form a data processing link.

[0015] The technical effects of this invention include: First, by integrating multi-source data such as seismic observation, ground vibration monitoring, and borehole wave velocity testing, a joint constraint mechanism for fault geometric parameters and site medium parameters is established, improving the reliability of the source model parameters; Second, a hybrid modeling method combining the concave-convex body model and the k-squared slip model is adopted, which ensures both the deterministic and stochastic characteristics of the slip distribution, improving the model's ability to characterize the complexity of actual earthquake rupture; Third, a three-dimensional fault dynamic model is constructed using the dual-couple point source superposition method, outputting source time functions and spatial slip distribution data, providing complete source input parameters for earthquake motion numerical simulation; Fourth, a complete technical link from multi-source data to the source model is established, which can be directly applied to earthquake motion prediction in active fault scenarios, providing technical support for seismic fortification of major engineering projects. Attached Figure Description

[0016] Figure 1 This is a flowchart of the method for constructing a finite fault source model for active faults according to an embodiment of the present invention. Figure 1 .

[0017] Figure 2 This is a system architecture diagram for constructing a finite fault source model of a live fault, according to an embodiment of the present invention.

[0018] Figure 3 This is a flowchart of the method for constructing a finite fault source model for an active fault according to an embodiment of the present invention. Figure 2 . Detailed Implementation

[0019] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only for illustrating the present invention and are not intended to limit the scope of protection of the present invention.

[0020] Reference Figure 1 and Figure 3 This invention provides a method for constructing a finite fault source model for active faults. This method addresses the technical problems of difficulty in establishing source models and unclear coupling mechanisms between fault geometry and dynamic parameters in near-fault strong ground motion simulations. It constructs a finite fault source model by integrating multi-source seismic geological data. This embodiment uses a specific active fault as an example to explain the detailed implementation process of this method.

[0021] Step S1: Multi-source seismic geological data acquisition steps.

[0022] The multi-source seismic geological data acquisition step is used to obtain basic data on the target active fault area, providing data support for subsequent parameter extraction and model construction. The core of this step lies in integrating seismic geological data from different sources and of different types, improving the reliability of source model parameter estimation through data fusion.

[0023] In this embodiment, the study area of ​​the target active fault is first determined. The determination of the study area needs to comprehensively consider the fault's extension length, influence range, and data availability. Typically, the study area boundary is defined as extending 20km to 50km to both sides of the fault trace. For strike-slip faults, the study area can be appropriately extended along the fault strike direction; for thrust or normal faults, it should be appropriately extended along the fault dip direction to cover the hanging wall and footwall regions.

[0024] The acquisition of seismic observation data involves two aspects. The first aspect is historical earthquake catalog data, obtained from the National Earthquake Science Data Center or regional seismic network centers. This catalog includes basic parameters such as earthquake occurrence time, epicenter location, focal depth, and magnitude. The completeness and reliability of the historical earthquake catalog directly affect the accuracy of fault activity assessment; therefore, earthquake records spanning a longer time period should be collected whenever possible. In this embodiment, a complete earthquake catalog of the target fault region since 1900 was acquired, recording over 1200 earthquake events, including 58 earthquakes with a moment magnitude of 4.0 or higher, 12 earthquakes with a moment magnitude of 5.0 or higher, and 3 earthquakes with a moment magnitude of 6.0 or higher. For earthquakes with a moment magnitude of 6.0 or higher, focal mechanism solutions also need to be collected to constrain the fault slip type.

[0025] The second aspect involves seismic waveform recordings. Three-component waveform records of historical moderate to strong earthquakes within the study area were acquired from broadband seismic stations. The waveform data sampling rate was no less than 100Hz, and the recording duration covered the entire earthquake process. The quality of the waveform data directly affects the accuracy of source parameter inversion; therefore, preprocessing of the raw waveforms is necessary, including instrument response removal, baseline correction, and bandpass filtering. In this embodiment, waveform records of 12 earthquakes with a moment magnitude of 5.0 or higher within the study area were collected, with the number of stations ranging from 8 to 35. This provides important evidence for subsequent source mechanism solutions and fault geometric parameter constraints.

[0026] Ground vibration monitoring data is acquired through on-site observation. Ground vibrations are low-amplitude ground vibrations, mainly excited by environmental factors such as ocean waves, wind, and traffic. Their spectral characteristics are closely related to the underground structure of the site. By analyzing the horizontal and vertical spectral ratios of ground vibrations, the dominant frequencies and amplification factors of the site can be extracted, thereby inverting the shallow shear wave velocity structure. Ground vibration observation points are set up along the target fault and on both sides, with the spacing between observation points determined according to topographic and geological conditions, typically ranging from 500m to 2km. Each observation point uses a three-component velocity sensor to record environmental micro-motion signals for a recording time of no less than 30 minutes and a sampling rate of no less than 100Hz. The frequency response range of the sensor should cover 0.1Hz to 50Hz to ensure that complete ground vibration signals can be recorded.

[0027] In this embodiment, 45 ground vibration observation points were deployed along the target fault to acquire complete site micromotion observation data. Observation periods were chosen at night or on weekends to reduce interference from human-induced noise. For each observation point, the horizontal and vertical spectral ratio method was used to extract the site response characteristics. Specifically, the three-component waveform data was first segmented, with each segment ranging from 30 to 60 seconds in length and overlapping by 50%. Then, the Fourier amplitude spectrum of each segment was calculated, and the geometric mean of the two horizontal components was taken to obtain the horizontal spectrum. Finally, the ratio of the horizontal spectrum to the vertical spectrum was calculated to obtain the horizontal and vertical spectral ratio curve. The peak frequency of the spectral ratio curve is the dominant frequency of the site, and the peak amplitude reflects the strength of the site amplification effect.

[0028] Borehole wave velocity data are obtained through on-site drilling and testing. Wave velocity testing is the most direct method for obtaining the mechanical parameters of the subsurface medium, and the test results play a decisive role in the calculation of the shear modulus. Representative locations within the target fault area are selected for wave velocity testing boreholes. The borehole depth is determined based on site conditions, typically ranging from 30m to 100m. The borehole layout principle is to cover as many different geological units on both sides of the fault as possible, while also considering accessibility and construction conditions.

[0029] Shear and compression wave velocities at various depths were determined using either suspended wave velocity logging or cross-hole wave velocity testing. Suspended wave velocity logging involves suspending a transducer inside the borehole and calculating the wave velocity by measuring the propagation time of seismic waves at different depths. Cross-hole wave velocity testing measures the wave velocity across two or more boreholes, offering higher accuracy but also higher cost. In this embodiment, eight wave velocity testing boreholes were deployed at depths ranging from 50m to 80m, with a test spacing of 1m. Shear and compression wave velocity profiles were obtained from each borehole. The test results show that the shear wave velocity in the study area is 200m / s to 400m / s near the surface, gradually increasing with depth, reaching 2500m / s to 3500m / s below the bedrock surface.

[0030] Unifying the spatiotemporal references of data is a crucial step in ensuring the quality of multi-source data fusion. Regarding time reference unification, all data timestamps are converted to Coordinated Universal Time (UTC), with a time accuracy better than 0.01 seconds. For historical earthquake catalogs, a unified conversion of earthquake occurrence times is required; for waveform data, clock drift correction is necessary. Regarding spatial reference unification, the spatial coordinates of all data are converted to a unified geographic coordinate system, with horizontal coordinates using the WGS84 coordinate system and elevation using the 1985 National Height Datum. For spatial data from different sources, coordinate system transformation and projection transformation are required to ensure spatial consistency.

[0031] In terms of quality verification, integrity checks, outlier removal, and consistency verification are performed on various types of data. Integrity checks include confirming the completeness of data records and the absence of missing values; outlier removal includes identifying and deleting obviously unreasonable data points; and consistency verification includes checking for inconsistencies between different data sources. For example, comparing wave velocity test results from adjacent boreholes can reveal potential testing errors; comparing different records of the same earthquake in historical earthquake catalogs can help identify more reliable source parameters.

[0032] After the above processing, a multi-source seismic geological dataset is generated. This dataset contains quality-checked seismic observation data, ground motion monitoring data, and borehole wave velocity test data. The data format is uniform and the spatiotemporal reference is consistent, making it directly usable for subsequent analysis and processing. The dataset is organized using a hierarchical storage structure, with raw data, preprocessed data, and fused data stored separately, facilitating data traceability and updates.

[0033] Step S2: Fault geometry parameter extraction step.

[0034] The fault geometry parameter extraction step is used to extract the geometric parameters and site medium parameters of the target active fault from a multi-source seismic geological dataset. Fault geometry parameters are the fundamental input for finite fault source models, and their accuracy directly affects the reliability of subsequent slip distribution modeling and ground motion simulation. This step employs a multi-source data joint constraint method, comprehensively utilizing various data such as seismic activity, focal mechanism solutions, geological surveys, and geophysical exploration to improve the accuracy and stability of parameter estimation.

[0035] The determination of fault length employs a combination of seismic activity analysis and geological survey. First, based on the epicenter distribution in the historical earthquake catalog, the least squares method is used to fit the distribution direction and length of seismic activity bands, serving as a preliminary estimate of the fault length. Identification of seismic activity bands requires setting reasonable spatial windows and magnitude thresholds to exclude seismic events unrelated to the target fault. Typically, earthquakes with a moment magnitude of 2.0 or higher are selected for analysis, with the spatial window width set at 10 to 20 km on each side of the fault strike. Then, the length of the fault surface traces is verified and corrected using geological survey data, including regional geological maps, fault activity identification reports, and geological trench exposures. These data provide evidence of fault surface exposure length and Quaternary activity.

[0036] In this embodiment, the seismic activity strip of the target fault is approximately 85 km long, obtained by fitting the epicenter distribution of earthquakes with a moment magnitude of 2.5 or higher since 1970. Geological surveys indicate that the fault has an exposed length of 78 km at the surface, and both ends of the fault can be traced back to Quaternary sediment-covered areas where they are concealed. Considering the possibility that the seismic activity range is slightly larger than the surface exposed range of the fault, and the possibility of concealed segments at the fault ends, the final fault length L is determined to be 80 km.

[0037] The fault width is determined using a focal depth distribution analysis method. The focal depths of each earthquake event in the historical earthquake catalog are statistically analyzed to determine the depth range of seismic activity. Fault width refers to the distance the fault plane extends in the dip direction and has a geometric relationship with the depth range of seismic activity and the fault dip angle. For faults with a certain dip angle, the fault width is related to both the focal depth range and the fault dip angle; the calculation formula is as follows:

[0038] ,

[0039] Where W is the fault width, in km; The maximum focal depth is expressed in km. Minimum focal depth, in km; The fault dip angle is expressed in degrees. The physical meaning of this formula is that when a fault dips at an angle... As it extends underground, it represents the width of the fault plane corresponding to the shallowest to the deepest focal depth. It is important to note that... It is usually not taken as zero, but as the shallowest depth above the brittle-ductile transition zone where brittle fracture can occur, typically 2km to 5km.

[0040] In this embodiment, the focal depths of historical earthquakes range from 3 km to 18 km, indicating that seismic activity along this fault primarily occurs within the shallow, brittle layer of the crust. The fault dip angle is initially estimated at 70°, determined based on the focal mechanism solution and geophysical exploration data. Substituting these values ​​into the above formula, the calculated fault width W is approximately 16 km. This width value coincides with the thickness of the brittle layer in the region, indicating that the fault activity penetrates the entire brittle layer.

[0041] Determining fault depth employs a combination of shallow seismic exploration and borehole verification. Fault depth refers to the vertical distance from the fault crest to the Earth's surface. For exposed faults, the depth is zero; for concealed faults, geophysical exploration is required to determine the depth. Shallow seismic reflection profiles are an effective means of detecting fault depth. By identifying fault discontinuities in the reflected waves, the deep location and attitude of the fault can be determined. Shallow seismic lines are deployed at key locations where the fault may cross, and reflected wave data is collected, processed, and interpreted to obtain the fault's location and depth on the profile.

[0042] In this embodiment, five shallow seismic survey lines were laid out perpendicular to the fault strike, with line lengths ranging from 2 km to 5 km, geophone spacing of 5 m, and excitation point spacing of 10 m. Data processing employed a standard reflection seismic processing workflow, including static correction, velocity analysis, dynamic correction, stacking, and migration. Interpretation results showed that the fault depth at each survey line location ranged from 1.5 km to 2.5 km, with an average of approximately 2 km. Based on geological borehole findings, the fault depth d was determined to be 2 km.

[0043] The determination of fault strike employs a method combining the direction of seismic activity strips with geological surveys. Fault strike is the azimuth of the intersection of the fault plane and the horizontal plane, usually expressed as the angle of clockwise rotation from true north. First, a preliminary value for the fault strike is determined based on the direction of the strips fitted from the epicenter distribution. The direction of the epicenter strips is usually consistent with the fault strike, but due to epicenter location errors, the fitting results may have some deviation. Then, the direction of the fault surface traces is checked in conjunction with the geological survey. The direction of the fault surface traces can be directly measured from geological maps or remote sensing images, offering high accuracy.

[0044] In this embodiment, the direction of the seismic activity strip is N45°E, obtained through least squares fitting. The fault strike from the geological survey is N42°E, measured from a 1:50000 geological map. The strike values ​​obtained by the two methods differ by 3°, which is within a reasonable error range. Combining the two results, a weighted average method is used to determine the fault strike. It is N43°E, which is 43°.

[0045] The fault dip angle is determined using focal mechanism analysis. The fault dip angle is the angle between the fault plane and the horizontal plane, ranging from 0° to 90°. Focal mechanism solution data for moderate to strong earthquakes within the study area are collected. Each focal mechanism solution provides two possible fault plane solutions; the correct fault plane must be selected based on other constraints. The dip angle distribution in the fault plane solutions is statistically analyzed, and the dominant dip direction is taken as the fault dip angle. For strike-slip faults, the dip angle is typically close to 90°; for thrust faults, the dip angle is typically 30° to 60°; and for normal faults, the dip angle is typically 50° to 70°.

[0046] In this embodiment, focal mechanism solutions for 12 earthquakes with a moment magnitude of 5.0 or higher were collected. These earthquakes all occurred near the target fault and their focal mechanisms were consistent with the fault characteristics. The focal mechanism solutions were obtained from the global moment center moment tensor catalog and the focal mechanism solution catalog published by the China Earthquake Administration. By analyzing the fault plane parameters of each focal mechanism solution, fault plane solutions consistent with the strike of the target fault were selected, with dip angles ranging from 65° to 80° and a mode of 72°. Considering that this fault is a strike-slip fault, the relatively steep dip angle is consistent with the general characteristics of strike-slip faults, and the final fault dip angle was determined. It is 72°.

[0047] The determination of site medium parameters is based on borehole wave velocity test data. The main site medium parameters include shear modulus, density, and wave velocity, among which shear modulus is a key parameter for calculating scalar seismic moment. The formula for calculating shear modulus is:

[0048] ,

[0049] in, Shear modulus, in Pa; This refers to the density of the medium, expressed in kg / m³. 3 ; The shear wave velocity is expressed in m / s. Since the medium parameters vary with depth within the fault depth range, representative equivalent parameters need to be selected. Typically, a weighted average value within the focal depth range is used as the equivalent parameter, and the weights can be determined based on the slip distribution or depth distribution.

[0050] In this embodiment, a velocity structure model of the study area was established based on wave velocity test data from eight boreholes. The borehole test depths ranged from 50m to 80m. The test results showed that the shear wave velocity varied in a stepwise manner with increasing depth, with a significant velocity jump at the bedrock surface. For the velocity structure at deeper depths, extrapolation was performed using regional crustal velocity models and deep geophysical exploration results. The shear modulus at each depth was calculated, and the average value within the source depth range (3km to 18km) was taken as the equivalent shear modulus.

[0051] The density of the medium was determined using a combination of empirical formulas and measured data. For shallow areas with borehole data, the density value could be obtained from laboratory measurements of core samples; for deep areas without measured data, the density-wave velocity empirical formula was used for estimation. In this embodiment, the density of the shallow medium was determined to be 2400 kg / m³ based on core sample measurements. 3 Up to 2600 kg / m 3 The density of the deep medium is estimated to be 2700 kg / m³ based on Gardner's empirical formula. 3 Up to 2900 kg / m 3 The average density within the focal depth range is taken as 2700 kg / m³. 3 .

[0052] Based on the combined wave velocity and density data, the average shear wave velocity within the focal depth range was calculated to be 3200 m / s, corresponding to a shear modulus of... =2700×3200²=2.76×10 10 Pa. This value is within the reasonable range of typical crustal rock shear modulus (1×10⁻⁶ Pa). 10 Pa to 5×10 11Pa can be used as an input parameter for subsequent scalar seismic moment calculations.

[0053] Based on the above analysis, the geometric parameters of the target active fault in this embodiment are: fault length L = 80 km, fault width W = 16 km, fault depth d = 2 km, and fault strike. =43°, fault dip angle =72°. Site medium parameters are: shear modulus =2.76×10 10 Pa, medium density =2700kg / m 3 Average shear wave velocity =3200m / s. These parameters will serve as inputs for subsequent steps, including scalar moment calculation, slip distribution modeling, and dual-couple point source stacking.

[0054] Step S3: Calculation steps for scalar seismic moment

[0055] The scalar seismic moment calculation steps are used to calculate the scalar seismic moment and moment magnitude based on fault geometry parameters and site medium parameters.

[0056] The fault area is calculated by multiplying the fault length by the fault width:

[0057] ,

[0058] Where A is the fault area, in km² 2 L represents the fault length in km; W represents the fault width in km. In this embodiment, the fault area A = 80 × 16 = 1280 km². 2 Converted to SI units, it is 1.28 × 10⁻⁶. 9 m 2 .

[0059] The mean slip volume is determined using empirical statistical relationships. Based on the empirical relationship between magnitude and fault parameters proposed by Wells and Coppersmith, the mean slip volume and fault area have the following statistical relationship:

[0060] ,

[0061] Where D is the average slip, in meters (m); and A is the fault area, in kilometers (km²). 2 In this embodiment, based on a fault area of ​​1280 km² 2 The calculated average sliding amount D is approximately 2.1m.

[0062] The scalar seismic moment is calculated using the standard seismological formula:

[0063] ,

[0064] in, The scalar seismic moment is expressed in N·m. Here, A is the shear modulus, in Pa; and A is the fault area, in m³. 2 D represents the average slip, in meters (m). In this embodiment, the scalar seismic moment... =2.76×10 10 ×1.28×10 9 ×2.1=7.42×10 19 N·m.

[0065] The moment magnitude is calculated using the standard relation proposed by Hanks and Kanamori:

[0066] ,

[0067] in, Moment magnitude; The moment is a scalar seismic moment, expressed in N·m. In this embodiment, the moment magnitude is... =2 / 3×(log 10 (7.42×10 19 )-9.1)=2 / 3×(19.87-9.1)=7.18, take the integer part as level 7.2.

[0068] Step S4: Fault plane slip distribution modeling steps.

[0069] The fault-plane slip distribution modeling step is used to establish a spatial distribution model of slip on the fault plane and is a core component of finite fault source models. The fault-plane slip distribution reflects the displacement at various points along the fault plane during earthquake rupture, and its spatial distribution characteristics directly affect the amplitude and frequency characteristics of near-field ground motions. This step employs a hybrid modeling method combining a concave-convex body model and a k-squared slip model, ensuring both the deterministic and stochastic characteristics of the slip distribution.

[0070] The sub-fault discretization employs an adaptive size determination method. Discretizing a continuous fault plane into multiple rectangular sub-faults is a fundamental operation in finite fault source models. The selection of sub-fault sizes requires a balance between computational efficiency and model accuracy: excessively large sizes will lose spatial details of the slip distribution, while excessively small sizes will significantly increase computational complexity. This invention employs an adaptive size determination criterion based on moment magnitude to match the sub-fault sizes with the earthquake scale.

[0071] when At that time, sub-fault size km, applicable to medium-scale earthquakes; when At that time, sub-fault size km, applicable to large earthquakes; when At that time, sub-fault size km, applicable to major earthquakes.

[0072] The physical basis for the above criteria is that larger earthquakes have larger fault planes and a larger spatial scale of rupture processes, allowing for the use of larger sub-fault sizes without losing the main slip distribution characteristics. At the same time, the sub-fault size should be smaller than the typical size of the concave and convex bodies during earthquake rupture to ensure that the main spatial variations of the slip distribution can be described.

[0073] In this embodiment, the moment magnitude is 7.2, and the sub-fault size is 2km × 2km. Discretization along the fault strike direction is... =80 / 2=40 sub-faults, discretized along the fault dip direction. =16 / 2=8 sub-faults, totaling 320 sub-faults. The area of ​​each sub-fault is 4 km². 2 This can effectively describe the spatial variation characteristics of the slip distribution in earthquakes of this scale.

[0074] Concave-convex body models are used to determine the location and extent of high-slip regions. Concave-convex bodies are areas on the fault plane where the slip is significantly higher than the average, reflecting the non-uniformity of fault strength. Seismological studies show that the slip distribution of most earthquakes exhibits significant non-uniformity, with one or more high-slip regions. These high-slip regions typically correspond to weak zones or stress concentration areas on the fault plane, releasing most of the seismic energy during rupture. Concave-convex body models are an effective method for describing this non-uniform slip distribution.

[0075] Number of protrusions and depressions The determination based on moment magnitude is grounded in the physical principle that larger-scale earthquakes typically involve the cascading rupture of multiple concave and convex bodies, while smaller-scale earthquakes may involve only the rupture of a single concave or convex body. This invention uses the following criteria to determine the number of concave or convex bodies:

[0076] when hour, ; when hour, ; when hour, .

[0077] In this embodiment, the moment magnitude is 7.2, and the number of concave-convex bodies is 2. The total area of ​​the concave-convex bodies is... proportion of the total fault area Take 0.20, that is km 2This ratio is based on statistical analysis of historical earthquake slip distribution inversion results, and the area ratio of the concave and convex bodies is typically between 15% and 25%. The area distribution of the two concave and convex bodies is unequal, with the primary concave and convex body having a larger area and the secondary concave and convex body having a smaller area, with an area ratio of approximately 5:3. In this embodiment, =160km 2 , =96km 2 .

[0078] The locations of the concave and convex bodies were determined using a random method, but certain constraints were met. The center of the first concave and convex body (the primary concave and convex body) was located within 0.3 to 0.5 times the length of the fault, i.e., 24 km to 40 km from one end of the fault. The center of the second concave and convex body (the secondary concave and convex body) was located within 0.6 to 0.8 times the length of the fault, i.e., 48 km to 64 km from one end of the fault. This arrangement ensured that the two concave and convex bodies did not overlap and were distributed in different parts of the fault plane. The amount of slip within the concave and convex bodies... Average slip times, Take 2.0, that is =2.0×2.1=4.2m. The value typically ranges from 1.5 to 3.0; in this embodiment, the midpoint value of 2.0 is used.

[0079] The k-squared slip model is used to generate the random components of subfault slip. This model is based on the wavenumber spectrum characteristics of slip distribution found in seismological studies, namely, the spatial power spectrum of slip distribution follows a certain pattern in high wavenumber regions. Regular decay. This characteristic reflects the self-similarity and randomness of the slip distribution during earthquake rupture, and is an important supplement to deterministic concave-convex body models.

[0080] The steps for generating random sliding volume components are as follows:

[0081] First, generating in the wavenumber domain with Random fields with spectral properties. Wavenumber spectrum. The form is:

[0082] ,

[0083] in, wave number Spectral amplitude at the location; The zero-wavenumber amplitude of the spectrum is determined by the total seismic moment constraint; The angular wavenumber, related to fault size, reflects the characteristic spatial scale of the slip distribution. This spectral form is constant in the low wavenumber region (…). In the high wavenumber region, according to Attenuation, turning wave number is In this embodiment, Take the reciprocal of the fault characteristic size, that is km⁻¹.

[0084] Then, a two-dimensional inverse Fourier transform is performed on the wavenumber domain random field to obtain the random slip distribution in the spatial domain. Specifically, a complex random field is generated in the wavenumber domain, whose amplitude is determined according to... The distribution and phase are uniformly distributed random numbers. A two-dimensional inverse Fourier transform is performed on this complex random field, and the real part is taken as the random sliding distribution in the spatial domain. The random sliding amount is normalized so that its mean is zero and its standard deviation is 0.5 times the average sliding amount, i.e. m.

[0085] Finally, the slip components of the concave-convex body and the random slip components are superimposed to obtain the slip distribution matrix of the fault plane. :

[0086] ,

[0087] in, For the first Line 1 The slippage of the Liezi fault, in meters (m). For the sliding component of the concave-convex body, take within the range of the concave-convex body. Take outside the area of ​​the concave and convex body ; This represents the random slip component. The superimposed slip is subject to non-negativity constraints, truncating negative values ​​to zero to ensure that the slip of all sub-faults is greater than or equal to zero. Simultaneously, the slip distribution matrix is ​​normalized to ensure that the total seismic moment equals the scalar seismic moment calculated in step S3. The normalization coefficient is calculated using the following formula:

[0088] ,

[0089] use Multiply by the slip of all sub-faults to obtain the final slip distribution matrix.

[0090] Step S5: Superposition of Two-Force-Couple Point Sources

[0091] The double-couple point source superposition step is used to transform discrete sub-faults into equivalent double-couple point sources and construct a complete three-dimensional fault dynamics model.

[0092] The double-couple point source is a standard mechanical model in seismology describing fault slip. For each sub-fault, the moment tensor of its equivalent double-couple point source is... It can be represented as:

[0093] ,

[0094] in, For the moment tensor Quantity; For the first Line 1 The seismic moment of the Liezi fault, ; and The first and second normal vectors of the fault plane are respectively and the Quantity; and The first and second are the sliding direction vectors respectively. and the Quantity.

[0095] Normal vector of the fault plane and sliding direction vector According to the fault strike Fault dip angle and sliding angle Calculation. In this embodiment, the sliding angle... Based on the regional tectonic stress field and historical earthquake focal mechanism solutions, a value of -15° is used to represent a slip type dominated by left-lateral strike-slip with a small amount of normal fault component. The formulas for calculating the fault plane normal vector and slip direction vector are as follows:

[0096] ,

[0097] ,

[0098] Substituting the above vector components into the moment tensor formula, we obtain the six independent moment tensor components of the dual-couple point source of each sub-fault. , , , , , The moment tensor is a symmetric tensor with a trace of zero, i.e. .

[0099] The rupture initiation time was determined using an isotropic rupture propagation model. It was assumed that the rupture started at the hypocenter and propagated outwards at a constant rupture velocity. Line 1 time of rupture initiation of the Liezi fault The calculation formula is:

[0100] ,

[0101] in, This represents the rupture initiation time of the sub-fault, in seconds. This represents the distance from the center of the sub-fault to the hypocenter, in km. The fracture velocity is expressed in km / s. The fracture velocity is typically taken as 0.7 to 0.9 times the shear wave velocity; in this embodiment, it is taken as... km / s. The location of the epicenter is taken as the center of the main concave-convex body, which in this embodiment is located at 0.4 times the fault length and 0.5 times the fault width.

[0102] The rise time is related to the size and rupture velocity of the sub-fault. Defined as the time required for a sub-fault to reach its maximum slip from the start of slippage, the calculation formula is:

[0103] ,

[0104] in, Rise time, in seconds; and These represent the length and width of the sub-fault, respectively, in km. In this embodiment, the rise time... s.

[0105] A three-dimensional fault dynamics model is constructed by spatially superimposing the dual-couple point sources of all sub-faults. This model contains 320 dual-couple point sources, each with a defined spatial location, moment tensor, rupture initiation time, and rise time, which can fully describe the spatiotemporal evolution of fault rupture.

[0106] Step S6: Source parameter output step.

[0107] The source parameter output step is used to extract the source time function and spatial slip distribution data from the three-dimensional fault dynamics model, providing input parameters for earthquake motion numerical simulation.

[0108] The source time function is generated using a sub-fault contribution superposition method. (Source time function) The release rate of the scalar seismic moment over time is expressed by the following formula:

[0109] ,

[0110] in, The seismic moment release rate is expressed in N·m / s. For the first Line 1 The seismic moment of the Liezi fault; This is the rupture initiation time of the sub-fault; The seismic moment release time function for a single sub-fault is usually represented by an isosceles triangular function or a trapezoidal function.

[0111] In this embodiment, the seismic moment release time function for a single sub-fault adopts an isosceles triangular function:

[0112] ,

[0113] in, The rise time is given. This function satisfies the normalization condition. Ensure that the total seismic moment released by a single sub-fault is equal to .

[0114] The source time function was sampled at 0.01 s intervals, covering the entire rupture duration. The total duration of the source time function is... The estimate is:

[0115] ,

[0116] In this embodiment, s. Integrate the sampled source time function to verify that the integral value equals the scalar seismic moment. .

[0117] The spatial slip distribution data is output in the standard format of the finite fault model. The output includes: fault geometry parameters (length, width, depth, strike, dip), sub-fault mesh parameters (number of sub-faults, sub-fault dimensions), and parameters for each sub-fault (center coordinates, slip amount, slip angle, rupture initiation time, rise time). The data format is compatible with the USGS finite fault model database and the SRCMOD database, and can be directly used in various seismic motion simulation software.

[0118] Reference Figure 2 This invention also provides a system for constructing a finite fault source model of an active fault. This system is used to execute the steps in the above-described method embodiments. The system adopts a modular architecture design, including six functional modules: a multi-source seismic geological data acquisition module, a fault geometric parameter extraction module, a scalar seismic moment calculation module, a fault plane slip distribution modeling module, a dual-couple point source stacking module, and a source parameter output module. The modules exchange data through a standard data interface, forming a complete data processing chain.

[0119] The multi-source seismic geological data acquisition module is the system's data entry module, responsible for interfacing with external data sources and completing data preprocessing. This module comprises three sub-units: the seismic data interface sub-unit, which connects to seismic station data interfaces to acquire historical earthquake catalogs and seismic waveform data, supporting access to various data sources such as the China Earthquake Science Data Center and regional seismic networks; the ground vibration data interface sub-unit, which connects to ground vibration monitoring equipment to acquire ground vibration data observed in the field, supporting data formats from mainstream seismic instrument manufacturers; and the borehole data interface sub-unit, which connects to the borehole test database to acquire wave velocity test data, supporting the import of various data formats.

[0120] This module also includes a data preprocessing subunit for performing spatiotemporal benchmark unification and quality verification functions. The spatiotemporal benchmark unification function converts timestamps from data from different sources to Coordinated Universal Time (UTC) and converts spatial coordinates to the WGS84 geographic coordinate system. The quality verification function performs data integrity checks, outlier identification, and consistency verification, generating a data quality report. The preprocessed data forms a multi-source seismic geological dataset, which is then transferred to the fault geometry parameter extraction module via a standard interface.

[0121] The fault geometry parameter extraction module receives multi-source seismic geological datasets and uses a joint inversion algorithm to extract geometric parameters such as fault length, fault width, fault depth, fault strike, and fault dip angle. Simultaneously, it calculates site medium parameters based on borehole wave velocity test data. This module comprises two main sub-units: a geometry parameter calculation sub-unit and a medium parameter calculation sub-unit.

[0122] The geometric parameter calculation subunit incorporates multiple parameter estimation algorithms, including a seismic activity analysis algorithm for fault length estimation, a focal depth statistical algorithm for fault width estimation, a seismic exploration interpretation algorithm for fault depth estimation, a strip fitting algorithm for fault strike estimation, and a focal mechanism solution statistical algorithm for fault dip estimation. The medium parameter calculation subunit calculates parameters such as shear modulus, compressive modulus, and medium density based on borehole wave velocity test data, supporting the establishment of one-dimensional and three-dimensional velocity models. The fault geometric parameters and site medium parameters output by this module are simultaneously transmitted to the scalar seismic moment calculation module and the fault plane slip distribution modeling module, achieving parallel parameter transfer.

[0123] The scalar seismic moment calculation module receives fault geometric parameters and site medium parameters, and calculates the fault area, shear modulus, mean slip, scalar seismic moment, and moment magnitude according to the calculation formula in step S3 of the method embodiment. This module contains three sub-units: a fault area calculation sub-unit, a mean slip estimation sub-unit, and a seismic moment calculation sub-unit.

[0124] The fault area calculation subunit calculates the fault area based on the fault length and width. The mean slip estimation subunit incorporates various empirical statistical formulas, including Wells and Coppersmith formulas, Hanks and Bakun formulas, etc., and can automatically select an appropriate formula to estimate the mean slip based on user selection or data quality. The seismic moment calculation subunit calculates the scalar seismic moment based on the shear modulus, fault area, and mean slip, and converts it into moment magnitude. The scalar seismic moment and moment magnitude output by this module are transmitted to the fault plane slip distribution modeling module as slip distribution constraints.

[0125] The fault plane slip distribution modeling module receives fault geometric parameters, scalar seismic moments, and moment magnitudes. Following the method described in step S4 of the implementation method, it performs sub-fault discretization, concave-convex body modeling, and k-squared slip model random field generation, ultimately outputting the fault plane slip distribution matrix. This module is the core computational module of the system and comprises four sub-units: sub-fault discretization sub-unit, concave-convex body modeling sub-unit, random slip generation sub-unit, and slip distribution synthesis sub-unit.

[0126] The sub-fault discretization sub-unit adaptively determines the sub-fault size based on the moment magnitude and divides the fault plane into regular grid units. The concave-convex body modeling sub-unit determines the number and area of ​​concave-convex bodies based on the moment magnitude, uses a random method to determine the location of concave-convex bodies, and generates concave-convex body slip components. The random slip generation sub-unit generates a random field in the wavenumber domain using a k-squared slip model, and converts it into random slip components in the spatial domain through inverse Fourier transform. The slip distribution synthesis sub-unit superimposes the concave-convex body slip components and random slip components, performs non-negativity constraints and normalization processing, and generates the final fault plane slip distribution matrix. The slip distribution matrix output by this module is transmitted to the dual-couple point source superposition module.

[0127] The dual-couple point source superposition module receives fault geometric parameters and slip distribution matrix, and calculates the moment tensor components, rupture initiation time, and rise time of each sub-fault dual-couple point source according to the method in step S5 of the method embodiment. A three-dimensional fault dynamic model is then constructed through spatial superposition. This module comprises three sub-units: a moment tensor calculation sub-unit, a rupture time calculation sub-unit, and a model superposition sub-unit.

[0128] The moment tensor calculation sub-unit calculates the fault plane normal vector and slip direction vector based on the fault strike, dip angle, and slip angle, and then calculates the six independent moment tensor components of each sub-fault. The rupture time calculation sub-unit calculates the rupture initiation time of each sub-fault based on the source location and rupture velocity, and the rise time based on the sub-fault size and rupture velocity. The model stacking sub-unit integrates the dual-couple point source parameters of all sub-faults to construct a complete three-dimensional fault dynamic model. The three-dimensional fault dynamic model output by this module is transmitted to the source parameter output module.

[0129] The source parameter output module receives the three-dimensional fault dynamics model, performs time integration on the seismic moment release process of each dual-couple point source according to the method in step S6 of the method embodiment to generate the source time function, and converts the fault plane slip distribution matrix into standard format spatial slip distribution data for output. This module contains two sub-units: a source time function generation sub-unit and a data format conversion sub-unit.

[0130] The source time function generation sub-unit employs a sub-fault contribution stacking method to calculate the overall seismic moment release rate over time, supporting various sub-fault time function forms such as isosceles triangles and trapezoids. The data format conversion sub-unit converts source model parameters into multiple standard output formats, including the USGS finite fault model format, SRCMOD database format, and source time function ASCII format, meeting the input requirements of different seismic motion simulation software. The output of this module can be directly used in mainstream seismic motion numerical simulation software such as SPECFEM3D, OpenSWPC, and COMPSYN.

[0131] The system in this embodiment of the invention adopts a modular design, with modules connected through standard data interfaces, facilitating system deployment, maintenance, and upgrades. The system supports both stand-alone and distributed computing modes, and parallel computing can be used to accelerate the generation of large-scale seismic source models. The system also provides a graphical user interface for convenient parameter setting, process monitoring, and result visualization.

[0132] To verify the effectiveness of the method of this invention, the source model constructed using this method was compared with a source model derived from actual earthquake records. A past earthquake was selected as the verification object. This earthquake had a magnitude of Mw7.1, a fault length of approximately 65 km, and a width of approximately 18 km. This earthquake had dense near-field strong motion observation records, and several research institutions have published source models derived from observation data, which can be used as a comparison benchmark.

[0133] The method of this invention is used to construct a predicted source model for this earthquake. The input data includes a catalog of historical earthquakes before the earthquake, ground motion observation data, and a regional wave velocity structure model. Following the steps of the method embodiment, multi-source data acquisition, fault geometric parameter extraction, scalar seismic moment calculation, slip distribution modeling, dual-couple point source superposition, and source parameter output are performed sequentially to finally obtain the predicted source model.

[0134] The predicted source model was compared with the source model derived from the actual earthquake observation data. The comparison included multiple aspects such as fault geometry, scalar moment, slip distribution, and source time function. The comparison results showed that the predicted source model had an 8% deviation in fault length, a 12% deviation in fault width, a 5° deviation in fault strike, and a 4° deviation in fault dip angle compared to the derived source model; the overall deviation of the geometric parameters was within 10%; the scalar moment deviation was 15%, within the typical uncertainty range of moment estimation; the spatial correlation coefficient of the slip distribution reached 0.72, indicating good similarity between the predicted and actual slip distributions; and the duration deviation of the source time function was 18%, with a peak time deviation of 2 seconds.

[0135] Further seismic motion simulations were conducted using both predicted and inverted source models, and the prediction accuracy of near-field ground motion parameters was compared. Fifteen near-field strong-motion stations recording the earthquake were selected as validation points, with the closest distance to the fault ranging from 5 km to 50 km. Peak ground acceleration and response spectral acceleration were calculated for each station and compared with actual observation records. The results show that the geometric mean deviation between the peak ground acceleration simulated by the predicted source model and the observed values ​​is 0.25 (logarithmic standard deviation), and the deviation of the response spectral acceleration within the period range of 0.2 s to 2 s is 0.28, both within the acceptable range for engineering applications. The overall prediction deviation of near-field ground motion parameters is within 30%, meeting the accuracy requirements for engineering seismic motion prediction.

[0136] The above verification results demonstrate that the method of this invention can effectively construct a finite fault source model of an active fault. The constructed source model can provide reliable input parameters for simulating strong ground motion near faults, and has significant engineering application value. This method does not rely on actual earthquake observation data and can predict the source characteristics of scenario-based earthquakes before they occur, providing technical support for seismic fortification and seismic hazard assessment of major engineering projects.

[0137] The embodiments of the present invention are not limited to the specific embodiments described above. Those skilled in the art can make various equivalent changes or substitutions based on the technical solutions of the present invention, and all such changes or substitutions should be included within the protection scope of the present invention.

Claims

1. A method for constructing a finite fault source model of an active fault, characterized in that, include: Multi-source seismic geological data acquisition steps: acquire seismic observation data, ground vibration monitoring data and borehole wave velocity test data of the target active fault area, perform spatiotemporal benchmark unified processing and quality verification on the acquired data, and generate a multi-source seismic geological dataset; Fault geometric parameter extraction steps: Based on the multi-source seismic geological dataset, the fault length, fault width, fault depth, fault strike and fault dip angle of the target active fault are extracted using a joint inversion method, and the site medium parameters are determined based on the borehole wave velocity test data; The steps for calculating the scalar seismic moment are as follows: Calculate the fault area and shear modulus based on the fault length, the fault width, and the site medium parameters; calculate the scalar seismic moment by combining the average slip determined by empirical statistical relationships; and determine the moment magnitude based on the scalar seismic moment. Fault plane slip distribution modeling steps: Discretize the fault plane into multiple sub-faults according to the moment magnitude and the fault geometric parameters; use the concave-convex body model to determine the location and range of the high slip area; use the k-square slip model to generate the random slip component of each sub-fault; and superimpose the concave-convex body slip component and the random slip component to generate the fault plane slip distribution matrix. The steps for superimposing dual-couple point sources are as follows: Each sub-fault is equivalent to a dual-couple point source. The moment tensor components of each dual-couple point source are calculated based on the fault strike, the fault dip angle, and the slip angle. The rupture initiation time and rise time of each dual-couple point source are determined. A three-dimensional fault dynamics model is constructed by spatially superimposing each dual-couple point source. Source parameter output steps: Perform time integration on the seismic moment release process of each dual-couple point source in the three-dimensional fault dynamics model to generate the source time function, and convert the fault plane slip distribution matrix into a spatial slip distribution data format for output.

2. The method for constructing a finite fault source model of an active fault according to claim 1, characterized in that, In the multi-source seismic geological data acquisition step, the seismic observation data includes historical earthquake catalog data and seismic waveform record data; the ground motion monitoring data includes site micromotion observation data and horizontal and vertical spectral ratio data; and the borehole wave velocity test data includes shear wave velocity profile data and compression wave velocity profile data.

3. The method for constructing a finite fault source model of an active fault according to claim 1, characterized in that, In the fault geometry parameter extraction step, the fault length ranges from 1km to 300km, the fault width ranges from 1km to 50km, the fault burial depth ranges from 0km to 30km, the fault strike ranges from 0° to 360°, and the fault dip angle ranges from 0° to 90°.

4. The method for constructing a finite fault source model of an active fault according to claim 1, characterized in that, In the scalar seismic moment calculation step, the shear modulus is calculated based on the shear wave velocity and medium density from the borehole wave velocity test data, and the value range of the shear modulus is 1×10⁻⁶. 10 Pa to 5×10 11 Pa.

5. The method for constructing a finite fault source model of an active fault according to claim 1, characterized in that, In the fault plane slip distribution modeling step, the size of the sub-fault is adaptively determined according to the moment magnitude. When the moment magnitude is less than 6.5, the size of the sub-fault is 1km×1km; when the moment magnitude is between 6.5 and 7.5, the size of the sub-fault is 2km×2km; and when the moment magnitude is greater than 7.5, the size of the sub-fault is 5km×5km.

6. The method for constructing a finite fault source model of an active fault according to claim 1, characterized in that, In the fault plane slip distribution modeling step, the number of concave and convex bodies in the concave-convex body model is determined according to the moment magnitude, the area of ​​the concave-convex body accounts for 15% to 25% of the total fault area, and the slip amount in the concave-convex body is 1.5 to 3 times the average slip amount.

7. The method for constructing a finite fault source model of an active fault according to claim 1, characterized in that, In the dual-couple point source superposition step, the calculation of the moment tensor components is based on the fault plane normal vector and the slip direction vector determined by the fault strike, the fault dip angle and the slip angle. The moment tensor is a symmetric tensor with a trace of zero.

8. The method for constructing a finite fault source model of an active fault according to claim 1, characterized in that, In the dual-couple point source superposition step, the rupture initiation time is calculated based on the distance from the center of each sub-fault to the rupture initiation point and the rupture velocity. The rupture velocity ranges from 0.7 to 0.9 times the shear wave velocity, and the rise time is related to the sub-fault size and rupture velocity.

9. The method for constructing a finite fault source model of an active fault according to claim 1, characterized in that, In the source parameter output step, the source time function is normalized so that its integral value is equal to the scalar seismic moment, and the output format of the spatial slip distribution data is compatible with the standard format of the finite fault model.

10. A system for constructing a finite fault source model of an active fault, characterized in that, include: The multi-source seismic geological data acquisition module is used to acquire seismic observation data, ground vibration monitoring data and borehole wave velocity test data of the target active fault area, and to perform spatiotemporal benchmark unified processing and quality verification on the acquired data to generate a multi-source seismic geological dataset. The fault geometry parameter extraction module is used to extract the fault length, fault width, fault depth, fault strike and fault dip angle of the target active fault based on the multi-source seismic geological dataset using a joint inversion method, and to determine the site medium parameters based on the borehole wave velocity test data. The scalar seismic moment calculation module is used to calculate the fault area and shear modulus based on the fault length, the fault width and the site medium parameters, calculate the scalar seismic moment by combining the average slip determined by empirical statistical relationships, and determine the moment magnitude based on the scalar seismic moment; The fault plane slip distribution modeling module is used to discretize the fault plane into multiple sub-faults based on the moment magnitude and fault geometric parameters, and generate the fault plane slip distribution matrix using the concave-convex body model and the k-square slip model. The dual-couple point source superposition module is used to treat each of the sub-faults as equivalent to a dual-couple point source, calculate the moment tensor components, rupture initiation time and rise time of each dual-couple point source, and construct a three-dimensional fault dynamics model through spatial superposition. The source parameter output module is used to perform time integration on the three-dimensional fault dynamics model to generate a source time function and output spatial slip distribution data.

Images