A multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution

Abstract

This invention discloses a multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution, belonging to the field of water conservancy engineering monitoring technology, including the following steps: Data acquisition: acquiring gravity / gravity gradient data, surface geometric deformation data, and spectral data. Spatial constraint: constructing a dynamic mask using geometric deformation data to constrain the spatial boundary of gravity inversion. Physical enhancement: performing gradient processing or fusion on gravity data to obtain high-resolution mass distribution. Density correction: calculating the dynamic density of sediments over time based on a sedimentary evolution model, and correcting the background fluid density based on spectral data. Coupled solution: substituting the above parameters into the coupling equation to solve for the volume or mass of sediments. This invention can achieve long-term, continuous, non-contact, and quantitative accurate monitoring of reservoir capacity changes and sediment deposition volume through the synergistic fusion of satellite gravity observation, airborne gravity gradient tensor observation, optical remote sensing, and sedimentary physical evolution models.

Description

Technical Field

[0001] This invention relates to the field of water conservancy engineering monitoring technology, specifically to a multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution. Background Technology

[0002] Currently, with the aging of global water conservancy facilities, the problem of siltation in reservoirs and dams is becoming increasingly serious. Siltation not only reduces effective reservoir capacity, lowering power generation and flood control benefits, but may also threaten the structural safety of dams. Existing reservoir siltation monitoring technologies are mainly divided into two categories: contact and non-contact, but both have insurmountable technical bottlenecks.

[0003] 1. Underwater Topographic Sonar Mapping Technology (Contact Type): This method utilizes a survey vessel equipped with a multibeam or single-beam sonar system to construct an underwater topographic model (DEM). Limitations: This method suffers from numerous blind spots (drawdown zones, shoals), and is costly and infrequent (once every 3-5 years), failing to meet the monthly dynamic monitoring needs during the flood season. For deep-water high dams (>100m) or reservoirs with high sediment content, sonar signal attenuation is severe, significantly reducing measurement accuracy.

[0004] 2. Geometric Optics-Based Remote Sensing Monitoring Technology (Non-Contact): This method uses optical satellites to extract water surface area and combines this with the "water level-reservoir capacity curve" to infer reservoir capacity. Limitations: This method only monitors two-dimensional information of the "water surface" and assumes the reservoir bottom topography remains unchanged. Once siltation occurs, the original EAV curve becomes invalid, and the mass of sediment cannot be directly quantified.

[0005] 3. Traditional satellite gravity monitoring technology (GRACE / GRACE-FO): Monitors terrestrial water storage by inverting time-varying signals of the Earth's gravity field. Limitations: Spatial resolution and signal leakage: The effective spatial resolution of the GRACE satellite is approximately 300 km, while most reservoirs are only a few kilometers in size. The weak gravity signal from the reservoir area can be drowned out by surrounding soil water and groundwater signals (leakage error can reach 30%-50%). Furthermore, this method has weak vertical resolution: traditional gravity anomalies ( The decay rate decreases with the square of the distance. It has low sensitivity to changes in shallow sediment quality and is difficult to distinguish between shallow sediments and deep crustal tectonic responses.

[0006] In summary, there is an urgent need to solve the core problems of existing reservoir sedimentation monitoring technologies in deep water, high sediment content, and large-scale complex environments, which are caused by the lack of closed physical mechanisms and the single observation dimension, resulting in "inaccurate calculation, inability to separate, and inability to measure". Summary of the Invention

[0007] To address the shortcomings of existing technologies, this invention aims to provide a multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution. This method can achieve long-term, continuous, non-contact, and quantitative accurate monitoring of reservoir capacity changes and sediment deposition volume through the synergistic fusion of satellite gravity observation, airborne gravity gradient tensor observation, optical remote sensing, and sedimentary physical evolution models.

[0008] This invention is achieved through the following technical solution:

[0009] A multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution includes the following steps:

[0010] S1: Acquire satellite gravity data, airborne gravity gradient data, surface geometric deformation data, and spectral data;

[0011] S2: Use airborne gravity gradient data to supplement satellite gravity data and construct a satellite-aircraft gradient fusion model based on Wiener filtering;

[0012] S3: Based on the physical forward modeling constraints of the gravity gradient tensor and the fused gravity field data, a joint inversion objective function is constructed to convert the gravity anomalies and gravity gradient physical quantities on the observation surface into the mass distribution of the reservoir bottom sediment body, and accurately lock the boundary of the reservoir bottom sediment body.

[0013] S4: Within the boundary of the sedimentary body at the bottom of the reservoir, the sedimentary mass is obtained by filtering out the signal related to the water level height through the time response function of the total mass change of water mass and sediment mass, based on the time series filtering.

[0014] S5: Based on the soil mechanics consolidation model, the density and volume are transformed and calculated to obtain the dynamic change equation of the density of the sedimentary body at the bottom of the reservoir;

[0015] S6: Finally, multi-temporal interferometry is used to constrain the geometric boundaries and volume of the water body, and the fluid background density is corrected based on hyperspectral inversion, so as to finally calculate and output the sediment volume distribution.

[0016] Further optimization, step S2 specifically includes the following steps:

[0017] S21: Obtain the spherical harmonic coefficients of the time-varying gravity field of the satellite, calculate the gravitational potential, and derive the gravity gradient components at the satellite scale.

[0018] S22: Utilize UAVs to fly over the reservoir area to observe the gravity gradient tensor and its vertical component, and use the Fourier domain downward extension operator to reduce the airborne observation data to the ground surface;

[0019] S23: Based on gravity gradient components and airborne observation data, a Wiener filter-type fusion operator is constructed to obtain a computational model of the fused gradient field, namely the star-aircraft gradient fusion model.

[0020] Further optimization, the calculation model for the gradient field is as follows:

[0021] ;

[0022] In the formula, These are satellite observations. These are airborne observations. This represents the satellite noise power spectrum. For satellite noise variance, The attenuation factor is highly correlated with satellite observations. This is the airborne noise power spectrum. This represents the variance of airborne noise.

[0023] Further optimization, in step S3, the specific steps for the physical forward modeling constraint of the gravity gradient tensor include:

[0024] S31: For any volume element at the bottom of the reservoir At the observation point The resulting vertical gradient response follows Decay law:

[0025] ;

[0026] In the formula, Due to density difference, For geocentric radial, Here are the coordinates of the observation point. The coordinates of the source point (center of the bottom grid cell) are... It is a volume element.

[0027] Further optimization, in step S3, the specific steps for constructing a joint inversion objective function to convert the physical quantities of gravity anomalies and gravity gradients on the observation surface into the mass distribution of sediment bodies at the bottom of the reservoir include:

[0028] S32: Define and construct the gravity kernel function matrix and gradient kernel function matrix And based on potential field theory, the gradient kernel function matrix is ​​obtained. Vertical component in Kernel function expression; where, The gravitational anomaly response per unit mass at the observation point. The gravitational gradient response per unit mass at the observation point;

[0029] S33: Obtain the gravity anomaly and gravity gradient of the observed data vector, and construct a weighted least squares objective functional related to the spatial continuous distribution of gravity.

[0030] S34: The weighted least squares objective functional is derived, and the sediment mass distribution is obtained by solving the preconditional conjugate gradient method.

[0031] Further optimization yields the following formula for the weighted least squares objective functional:

[0032] ;

[0033] In the formula, For the gravity data fitting term, For gradient data fitting term, For model regularization terms, Let be the parameter vector to be determined. This is the vector of observed surface gravity anomalies. This is the vector of observed gravity gradient values. and For the data weight matrix, As the gradient balancing factor, For regularization parameters, This is the roughness operator.

[0034] Further optimization, step S4 specifically includes the following steps:

[0035] S41: At any time Total mass change Decomposed into water mass and sediment quality : ;

[0036] S42: Based on water level respectively and sediment transport flux The time integral is used to separately calculate the water mass and sediment mass: ; In the formula, For water density, This represents the water surface area corresponding to the water level. For water level changes, The deposition efficiency coefficient;

[0037] S43: Use time series filtering to correlate with water level Highly correlated signal stripping yields pure sediment quality. .

[0038] Further optimization, step S5 specifically includes the following steps:

[0039] S51: Based on the Terzaghi effective stress principle of the soil consolidation model, and calculating the evolution of void ratio according to compression and creep effects. ;in, For time ,depth The porosity at that location;

[0040] S52: Based on the average dry density of sediments throughout the reservoir area The saturated bulk density was calculated. : In the formula, This refers to the specific gravity of the sediment particles. This represents the density of the water.

[0041] Further optimization, step S6 specifically includes the following steps:

[0042] S61: Define the mask matrix Used to define the integration space for gravity inversion:

[0043] ;

[0044] S62: Reservoir bank settlement rate obtained using multi-temporal interferometry technique The original topographic model (DEM) from the initial stage of database construction is corrected, and the current geometric apparent volume is calculated. That is, the total space of water and sediment;

[0045] S63: Utilizing the multi-band reflectivity of hyperspectral satellites Inverting the instantaneous suspended matter mass concentration in the total space ;

[0046] S64: According to the law of conservation of mass, the mixed density of sediment-laden turbid water... Adjusted according to volume percentage and organized into a summary of... The linear correction formula;

[0047] S65: The final derivation yields... The final coupled solution equation after hyperspectral constraint is used to calculate and output the sediment volume distribution.

[0048] Further optimization yields the following final coupled solution equation:

[0049] ;

[0050] In the formula, The volume of the sediment. This represents the total mass change obtained from the gravity gradient inversion. For real-time mixed fluid background density, This represents the InSAR-corrected geometric volume. This represents the dynamic saturated packing density.

[0051] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0052] 1. The present invention provides a multi-source fusion monitoring method based on gravity gradient tensor and sediment evolution. By using deep physical coupling of multi-source satellite data and high-order gravity gradient inversion technology, it overcomes the limitations of traditional reservoir sedimentation monitoring schemes in deep water, high sediment content and complex terrain areas.

[0053] 2. This invention provides a multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution, which can significantly improve detection accuracy and vertical resolution: because traditional gravity anomaly monitoring is limited by... The slow decay characteristics of sediment make it difficult to distinguish weak shallow sedimentation signals. Therefore, this invention introduces the gravity gradient tensor (GGT) and utilizes its... The rapid attenuation characteristics effectively suppress long-wave background noise from deep crustal structures. (1) Improved spatial resolution: By integrating the satellite-machine gradient spectral domain, the monitoring resolution is improved from 300km at the satellite scale to the level of hundreds of meters, which can accurately locate the position of local deposition pits in the reservoir area. (2) Enhanced vertical resolution: Gradient data is extremely sensitive to the vertical distribution of mass changes. Combined with the depth-weighted algorithm of this invention, it can accurately distinguish the sediment layer at the bottom of the water body with a thickness of only centimeters.

[0054] 3. The present invention provides a multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution, realizing a fully dynamic physical closed loop of mass-to-volume conversion: Since existing technologies usually treat sediment density as a fixed constant, there is a systematic deviation of 15%-30% between the inverted volume and the actual volume. Therefore, this scheme can achieve: (1) sedimentary compaction correction: The present invention introduces a soil mechanics self-weight consolidation evolution model for the first time, considering the porosity decay effect of sediment with deposition time and burial depth. This makes the inverted sediment density ( (1) No longer empirical values, but dynamic variables with physical evolution logic, which significantly improves the robustness of long-term sedimentation estimation. (2) Mixed fluid density correction: The background fluid density was corrected in real time using suspended solids concentration (SSC) retrieved from hyperspectral satellites. This fundamentally solves the false interference of "high-density turbid water" on sedimentation signals during flood season, and achieves accurate decoupling of "water" and "sand" in the mass dimension.

[0055] 4. The present invention provides a multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution, which enhances the signal recovery capability under complex geometric boundaries: Since gravity signals have omnidirectional radiation characteristics, the undulation of reservoir bank topography often produces serious "signal leakage" errors. Therefore, this scheme can achieve: (1) Geometric consistency constraint: The present invention uses high-precision dynamic water mask and small deformation of reservoir bank obtained by SAR satellite (InSAR) to construct a rigorous geometric constraint operator for gravity inversion. (2) Deformation error compensation: SAR technology can remove false mass changes caused by geological subsidence or reservoir bank collapse, ensuring that all monitored gravity signals originate from the water and sediment mass allocation inside the reservoir.

[0056] 5. The present invention provides a multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution, which realizes high-frequency closed-loop monitoring in all time and space and non-contact: The "star-aircraft-ground" system constructed by the present invention completely gets rid of the dependence of traditional sonar mapping on ship entry and navigation conditions in the reservoir area. (1) All-weather operation capability: The combination of the ability of SAR satellite to penetrate clouds and fog and the non-optical characteristics of gravity observation enables this method to maintain a monthly monitoring frequency even in extreme weather conditions such as flood season and rainstorm when sonar and optical remote sensing cannot operate. (2) Cost and timeliness advantages: Through non-contact means, the coverage area of ​​a single monitoring can reach thousands of square kilometers, the monitoring efficiency is increased by more than 10 times, and the comprehensive operating cost is only less than 30% of the traditional manual ship survey scheme.

[0057] 6. The multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution provided by this invention has the core advantage of constructing a three-in-one inversion equation of "mass-geometry-evolution" based on physical mechanisms. It is not just a simple superposition of multi-source data, but rather a mathematical joint objective functional that couples the physical characteristics of gravity gradient, the geometric accuracy of SAR, the compositional perception of hyperspectral data, and the dynamic evolution of soil mechanics together. This solves the long-standing technical pain points of "undetectable, inaccurate, and inseparable" sedimentation monitoring in reservoirs, and has extremely high industrial application value and patent protection barriers. Attached Figure Description

[0058] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0059] Figure 1 This is a schematic diagram of the overall process of the present invention;

[0060] Figure 2This is a schematic diagram of the data fusion processing flow for step S2 provided by the present invention;

[0061] Figure 3 This is a schematic diagram of the joint inversion modeling process for steps S3-S4 provided by the present invention;

[0062] Figure 4 The schematic diagram of the multi-source constraint and solution process for steps S5-S6 provided by the present invention is shown. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0064] Example 1: This Example 1 provides a multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution, such as... Figures 1-4 As shown, this method constructs a rigorous monitoring system with potential field theory as its core, soil mechanics principles as its constraint, and multi-source remote sensing as its supplement:

[0065] 1. Geometric and Deformation Constraints: Utilizing the all-weather imaging and high-precision interferometry (InSAR) capabilities of multi-temporal synthetic aperture radar (SAR) satellites, millimeter-level deformations of the reservoir drawdown zone and banks are extracted in real time, and dynamic water boundaries under complex meteorological conditions are accurately delineated, providing rigorous spatial geometric priors and initial volume constraints for gravity inversion.

[0066] 2. Dynamic Correction of Fluid Density: Using broadband remote sensing data acquired by hyperspectral satellites, a suspended solids concentration (SSC) inversion model is established based on the spectral scattering characteristics of sediment, enabling real-time inversion of the water body's mass concentration distribution. This process overcomes the drawback of traditional methods that treat water density as a constant, achieving dynamic correction of "fluid background density (SSC)." The dynamic correction of ") ensures the physical rigor of the quality difference calculation.

[0067] The three-dimensional mapping and analytical solution of this invention are as follows:

[0068] Spatial level: The gravity gradient tensor (GGT) is introduced and its high sensitivity to shallow mass is utilized to peel away the background noise of the deep crust layer by layer, thus mathematically realizing a rigorous mapping from "planar observation data" to "three-dimensional sediment distribution".

[0069] At the physical level: a sedimentary physical compaction evolution model is introduced to correct the density deviation caused by sediment consolidation over time.

[0070] Ultimately, this method overcomes the bottlenecks of "insufficient spatial resolution" and "lack of physical mechanisms" in traditional monitoring through the collaborative inversion of multi-source heterogeneous data, achieving full-parameter and high-precision analytical calculation of sediment deposition, and significantly improving the physical reliability of water conservancy project safety monitoring.

[0071] The specific implementation steps and detailed derivation of this invention are as follows:

[0072] Step 1: Construct a star-machine gradient fusion model based on Wiener filtering;

[0073] This step aims to address the issues of insufficient spatial resolution and spectral aliasing in the gravity field. Airborne gravity gradient data (high-frequency components) is used to supplement satellite data (low-frequency components), and Wiener filtering is employed to achieve optimal signal-to-noise ratio (SNR) fusion.

[0074] 1.1 Spherical harmonic synthesis and gradient transformation of satellite gravity field (low-frequency reference);

[0075] Obtain the spherical harmonic coefficients of the time-varying gravitational field in GRACE-FO Level-2 ( First, calculate the gravitational potential. This leads to the derivation of the gravity gradient components at the satellite scale. (Vertical gradient).

[0076] According to the potential theory, the geocentric radial... The expression for the second derivative of the direction is: ;

[0077] In the formula, As the geocentric coplanarity, Longitude of the Earth's core The average radius of the Earth The gravitational constant is... For Earth mass, The order of the spherical harmonic coefficients, The order of the spherical harmonic coefficients. The maximum order of the spherical harmonic coefficient expansion. For a fully normalized consecutive Legendre function, and For the fully normalized time-varying gravitational field spherical harmonic coefficients, where, The unit is Eotvos ( This physical quantity is more sensitive to changes in shallow mass than gravity anomalies. One order of magnitude higher.

[0078] 1.2 Observation and extension of airborne gravity gradient tensor (high frequency enhancement);

[0079] Using drones at low altitude (flight altitude) in the reservoir area Direct observation of the gravity gradient tensor The focus is on utilizing the vertical component. .

[0080] Using the Fourier domain downward extension operator, airborne observation data are reduced to the Earth's surface. ):

[0081] ;

[0082] In the formula, Spatial frequency in the wavenumber domain For flight altitude, This is a regularized low-pass filter. It is a regularized low-pass filter used to suppress high-frequency noise amplification during the downward extension process.

[0083] 1.3 Optimal Spectral Fusion Based on Wiener Filtering;

[0084] To minimize the variance of the fused estimation error (Minimum Mean Square Error, MMSE), a fusion operator in the form of a Wiener filter is constructed.

[0085] In the wavenumber domain, let the satellite observation value be... Airborne observations fused gradient field for:

[0086] ;

[0087] in: , For satellite noise variance, The satellite noise power spectrum increases exponentially with order as the satellite observation is highly correlated with the attenuation factor. , The variance of the airborne noise is given, and the power spectrum of the airborne noise is approximately white noise within the observation frequency band.

[0088] The above formula ensures that the system automatically weights satellite data in the low-frequency band and automatically weights airborne data in the high-frequency band, achieving optimal complementarity at the physical level.

[0089] Step 2: Spatial constraints based on the joint inversion of gravity and gradient;

[0090] This step utilizes the gravity gradient tensor. The physical properties of the sediments allow for precise identification of the boundaries of the sedimentary bodies at the bottom of the reservoir.

[0091] 2.1 Physical forward modeling constraints of gradient tensors;

[0092] For any volume element at the bottom of the reservoir (density difference is) ), at the observation point The resulting vertical gradient response follows Decay law:

[0093] ;

[0094] In the formula, Here are the coordinates of the observation point. The coordinates of the source point (center of the bottom grid cell) are... It is a volume element. Compared to gravity anomalies ( The narrower half-wave width of the gradient signal means that it can physically distinguish adjacent sediment pits, thus solving the "resolution" problem.

[0095] 2.2 Construct the joint inversion objective function;

[0096] After obtaining the fused gravity field data, the core of this step lies in using mathematical inversion to transform the gravity anomalies on the observation surface (…). ) and gravity gradient ( Physical quantities converted into reservoir bottom sediment mass distribution ).

[0097] (1) Physical prerequisites: Definition of kernel function;

[0098] To establish a linear relationship between the observed data and the quality to be determined, we first define the kernel matrix:

[0099] Gravity kernel function matrix : Describes the gravitational anomaly response produced by a unit mass at the observation point, which follows the inverse square law.

[0100] Gradient kernel function matrix This describes the gravitational gradient response per unit mass at the observation point. According to potential field theory, its vertical component... The kernel function expression is:

[0101] ;

[0102] in, Here are the coordinates of the observation point. The coordinates of the center of the bottom grid cell are the coordinates of the bottom grid cell. is the gravitational constant.

[0103] (2) Constructing a joint objective functional ;

[0104] To simultaneously leverage the advantages of gravity anomalies (controlling the global total) and gravity gradients (controlling local boundaries), the following weighted least squares objective functional is constructed:

[0105] ;

[0106] Detailed explanation of parameter meanings:

[0107] For the gravity data fitting term, For gradient data fitting term, For model regularization terms;

[0108] (Parameter vector to be determined): Divide the sedimentary layer at the bottom of the reservoir into... A grid of equal volume Density anomaly values ​​for each grid cell. This is achieved through calculation. This allows for the direct acquisition of spatially continuous sediment distribution maps.

[0109] and (Observation Vector): These represent the surface gravity anomaly observation vector and gravity gradient observation vector, respectively, after the first step of fusion processing.

[0110] and (Data weight matrix): Usually, it is the inverse of the observation noise covariance matrix. Function: To automatically balance observation errors. High-precision measurement points have higher weights in the inversion process to prevent inversion oscillations caused by individual abnormal measurement points.

[0111] Special Note – Depth Weighting: Due to the extremely rapid decay of the gravity gradient with depth ( Inversion results often tend to be superficial. Therefore, in Introducing depth-weighted operators (generally This forces the inversion energy to shift to deeper layers in order to restore the true thickness of the sedimentary layer.

[0112] (Gradient balance factor): Because the gravitational anomaly (mGal) and the gravitational gradient (E) have different dimensions, Used to adjust the relative importance of the two types of data. Derivation suggestion: Through analysis... and Given the singular value spectrum distribution, select values ​​that make the orders of magnitude of the two residuals consistent.

[0113] (regularization parameter) and (Roughness operator): The inversion problem is highly ill-posed. The second-order difference operator (Laplace operator) is usually used to constrain the smoothness of the density between adjacent grids; The L-curve method is used to determine the optimal balance between minimizing residuals and smoothing the model.

[0114] (3) Derivation process and solution logic;

[0115] To obtain the target functional Minimum optimal solution ,right Find the partial derivative and set it to zero:

[0116] ;

[0117] The resulting system of normal equations is obtained by rearranging the equations as follows:

[0118] ;

[0119] The solution steps are as follows:

[0120] 1) Matrix construction: Utilize the water bottom water mask constraint to construct the kernel matrix only for the meshes that may have siltation, thereby reducing memory usage.

[0121] 2) Preconditional conjugate gradient method (PCG): Since the left-hand matrix is ​​huge and sparse, the PCG algorithm is used to solve large-scale linear equation systems.

[0122] 3) Result transformation: Solving for... Then, by combining the sedimentary evolution density model in subsequent steps, the mass distribution can be transformed into an accurate volumetric thickness distribution.

[0123] Step 3: Introduce quality separation with time causal consistency constraints;

[0124] To address the signal aliasing caused by the mixing of "water" and "sand," this invention introduces a causal constraint in the time dimension.

[0125] 3.1 Time response function to mass change;

[0126] at any time Total mass change Decomposed into water mass and sediment quality : .

[0127] 3.2 Logic of Causal Separation;

[0128] Water body response (transient): with water level It exhibits instantaneous linear correlation: In the formula For water density, This represents the water surface area corresponding to the water level. For water level changes;

[0129] Sedimentary response (cumulative): manifested as sediment transport flux The time integral is irreversible: , The deposition efficiency coefficient;

[0130] Utilizing this characteristic, time series filtering can be used to correlate water levels. After stripping away highly correlated signals, the remaining long-term trend term represents the pure sediment quality. .

[0131] Step 4: Density-volume evolution and calculation based on the soil mechanics consolidation model;

[0132] Traditional methods assume a constant sediment density, leading to significant errors in volume inversion. This invention introduces the Terzaghi effective stress principle to establish a time-depth-density evolution model.

[0133] 4.1 Equation of sediment porosity evolution;

[0134] Porosity of sediments at the bottom of the reservoir (time ,depth The porosity at a given location dynamically changes with burial depth (self-weight stress) and time (creep):

[0135] ;

[0136] in, The initial void ratio, For reference only; The compressibility index (primary consolidation) reflects the compaction that occurs as the sediment thickness increases. The secondary consolidation coefficient reflects the skeletal creep over time; For depth Effective overburden stress at the location.

[0137] 4.2 Time-varying equivalent density integral;

[0138] Average dry density of sediments in the entire reservoir area The integral function for depth:

[0139] ;

[0140] In the formula, For the thickness of the sedimentary layer, This refers to the specific gravity of the sediment particles. This represents the density of the water.

[0141] This leads to the saturated packing density. :

[0142] ;

[0143] This model ensures that the density parameter is automatically corrected as the project's operating years increase, which conforms to objective physical laws.

[0144] Step 5: Coupled solution of "gradient-volume-evolution";

[0145] In the joint inversion system, the SAR satellite provides the "geometric shell" constraint, and the hyperspectral satellite provides the "fluid composition" correction. Together, they provide the boundary conditions for gravity gradient inversion.

[0146] 1. Geometric boundaries and volume constraints of water bodies based on InSAR;

[0147] SAR satellites (such as Sentinel-1 or Gaofen-3) use multi-temporal interferometry (InSAR) technology to obtain minute deformations of the reservoir banks and drawdown zones, and construct real-time water area masks.

[0148] 1.1 Dynamic Geometric Boundary Constraint Matrix ;

[0149] Define mask matrix This is used to define the integration space for gravity inversion and prevent the signal from "leaking" to areas outside the reservoir bank:

[0150] ;

[0151] This matrix directly acts on the kernel function in the aforementioned objective functional. , so that: .

[0152] 1.2: Geometric Apparent Volume InSAR correction;

[0153] Reservoir bank subsidence rate obtained using InSAR The original topography (DEM) at the initial stage of reservoir construction is corrected, and the current geometric apparent volume (total space of water + sediment) is calculated:

[0154] ;

[0155] in, Real-time water level obtained from satellite altimetry. for At the initial dam base elevation, Let be an infinitesimal area element. This equation provides an accurate geometric denominator for the mass-volume transformation.

[0156] 2. Fluid background density correction based on hyperspectral inversion;

[0157] In gravity inversion, the observed mass change It refers to the residual mass after "siphoning off an equal volume of water with sediment". Traditional methods assume that the density of the water is constant. However, during flood season, the density of "muddy water" with high sediment content is significantly greater than that of pure water, and if this is not corrected, the amount of sediment will be seriously underestimated.

[0158] 2.1 Hyperspectral suspended solids concentration (SSC) inversion;

[0159] Utilizing the multi-band reflectivity of hyperspectral satellites (such as Zhuhai-1 or Huanjing-2) (Especially the ratio of near-infrared to red light bands), to retrieve instantaneous suspended matter mass concentration. :

[0160] ;

[0161] in These are regional parameters calibrated based on ground-measured spectra. and Reflectivity for a specific wavelength band.

[0162] 2.2 Background density of mixed fluid The physical correction derivation;

[0163] According to the law of conservation of mass, the mixed density of turbid water containing sand... It's not a simple addition; it needs to be corrected for volume percentage.

[0164] ;

[0165] Organize it into a document about The linear correction form:

[0166] ;

[0167] in This refers to the density of sediment particles. This will be directly substituted into the final volume analytical formula.

[0168] 3. The final coupled solution equation after embedding SAR and hyperspectral constraints;

[0169] Based on the above derivation, the core solution logic of this invention is transformed from "blind inversion" to "constrained analytical solution":

[0170] ;

[0171] In the formula, The volume of the sediment. This represents the total mass change obtained from the gravity gradient inversion. For real-time mixed fluid background density, This represents the InSAR-corrected geometric volume. This represents the dynamic saturated packing density.

[0172] Among them, the denominator term is: Provided by soil compaction model, Provided by a hyperspectral satellite, the dynamic nature of matter properties was addressed; molecular term: Provided by joint inversion of gravity gradients, The accuracy of spatial integration is solved by the geometric constraints provided by SAR satellites.

[0173] The core of this invention lies in breaking the limitations of traditional single physical quantity monitoring by constructing a four-dimensional coupled solution system of "mass (gravity) - geometry (SAR) - evolution (soil mechanics) - composition (hyperspectral)".

[0174] In the steps of this invention: (1) a coupled solution equation based on the physical mechanism of "gradient-volume-evolution"; this step is the core of this invention. Existing technologies usually only perform simple subtraction (total mass - water mass) and assume that the density is constant. This invention establishes a closed-loop analytical equation that integrates the mass obtained from gravity inversion, the volume measured by SAR, the water density obtained from hyperspectral inversion, and the sediment density calculated by soil mechanics. Thus, "blind inversion" is transformed into "constrained analytical solution", solving the problem of The problem involves calculations where both the numerator (mass) and denominator (density) change dynamically with time / environment.

[0175] (2) Gravity gradient tensor (GGT) and spectral domain fusion technology; this step solves the problem of "not being able to see clearly". Utilizing the gravity gradient signal... The attenuation characteristics (compared to gravitational anomalies) To suppress deep background noise, a satellite-aircraft gradient fusion model based on Wiener filtering was constructed to improve the spatial resolution of the monitoring from 300km (satellite) to 100m (airborne enhancement), enabling the identification of local sediment pits.

[0176] (3) Correction of "time-varying density" based on soil consolidation model; this step solves the problem of "inaccurate calculation". Because traditional methods assume that the density of sediment is constant (leading to an error of 15%-30%), this invention introduces the Terzaghi effective stress principle to calculate the dynamic density that varies with burial depth and time. This step introduces the time evolution parameter (compression index) of geotechnical engineering into physical remote sensing monitoring. Secondary consolidation coefficient This allows for interdisciplinary revisions.

[0177] (4) Dynamic removal of "fluid background density" based on hyperspectral analysis; this step solves the problem of "inseparability". For the situation where the density of "turbid water" with high sediment content is much greater than that of pure water during flood season, the background fluid density is corrected in real time by using hyperspectral inversion of suspended solids concentration (SSC). This step eliminates the physical interference of "turbid water" on the quality signal of "sedimentary sediment" at the source, ensuring that the mass difference obtained from gravity inversion is the true mass of the sediment.

[0178] This invention sequentially employs the following steps: 1) Data acquisition: acquiring gravity / gravity gradient data, surface geometric deformation data (SAR), and spectral data; 2) Spatial constraint: constructing a dynamic mask using the geometric deformation data to constrain the spatial boundaries of gravity inversion; 3) Physical enhancement: performing gradient processing or fusion on the gravity data to obtain a high-resolution mass distribution; 4) Density correction: calculating the dynamic density of sediments over time based on a sedimentary evolution model and correcting the background fluid density based on spectral data; 5) Coupled solution: substituting the above parameters into a coupling equation to calculate the volume or mass of the sediments. This enables non-contact monitoring of changes in the volume and mass of underwater / subsurface materials under variable-density fluid cover.

[0179] Therefore, it can also be applied to: 1) Monitoring of near-shore ports and waterways: monitoring siltation in ports, the principle is exactly the same (seawater replaces freshwater). 2) Safety monitoring of tailings dams: the sediments in tailings dams also have a consolidation effect, and there is usually a cover water layer above them. 3) Monitoring of groundwater and land subsidence: although there is no "underwater" concept, it involves the coupled process of "fluid extraction (mass reduction)" leading to "soil compaction (geometric deformation)," and the soil mechanics + gravity coupling approach of this invention is also applicable.

[0180] In summary, the multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution provided by this invention can solve the following problems:

[0181] 1. This invention addresses the problems of ambiguous spatial boundaries and lack of geometric constraints in complex environments: Existing remote sensing monitoring largely relies on geometric optics, which is susceptible to cloud shielding and interference from complex terrain in the drawdown zone, making it difficult to obtain accurate water area boundaries; furthermore, traditional gravity inversion ignores the impact of reservoir bank deformation on volume. This invention aims to introduce SAR satellites (Synthetic Aperture Radar), utilizing their all-weather imaging and high-precision interferometry (InSAR) characteristics to provide millimeter-level reservoir bank deformation monitoring and precise dynamic water area mask constraints, thus solving the problems of signal "leakage" to the shore and geometric volume constraints in gravity inversion. The problem is that the calculation benchmark is inaccurate.

[0182] 2. It can solve the problems of weak signals and insufficient vertical resolution in shallow sedimentation: traditional satellite gravity anomalies ( The signal varies with the square of the distance ( This method suffers from low sensitivity to shallow mass changes and is easily affected by long-wave background noise such as deep crustal tectonic movements. The present invention aims to introduce "gravity gradient tensor (GGT)" observations and utilize its variation with distance cubic (GGT)... The attenuation characteristics significantly suppress the deep background field, achieving a physical "lock" on the mass distribution of shallow sediments at the bottom of the reservoir, and improving the spatial resolution from the hundred-kilometer level to the hundred-meter level.

[0183] 3. It can solve the signal interference problem caused by background density fluctuations in multiphase fluids: Existing gravity inversion methods usually assume that the water density in the reservoir area is constant ( However, during flood season or under high sediment content conditions, water density fluctuates dramatically with suspended solids concentration (SSC), causing sediment deposition signals to be masked by turbid water quality anomalies. This invention aims to introduce hyperspectral satellites to invert instantaneous suspended solids concentration through spectral feature inversion, thereby achieving the determination of "mixed fluid density (…)" The real-time correction of "muddy water" eliminates the physical interference of "muddy water" on the inversion of "sedimentary sediment" from the source.

[0184] 4. This invention addresses the volume conversion distortion caused by the dynamic compaction effect of sediment: Existing technologies generally use fixed empirical density constants to convert mass into volume, neglecting the compaction and consolidation process of sediments under their own weight with increasing burial depth and time. This invention aims to introduce the effective stress principle of soil mechanics and a consolidation evolution model to establish time-varying porosity and dynamic density functions, correcting the nonlinear volume deviation caused by sediment "self-weight consolidation," and resolving the theoretical dilemma of inversion results deviating from engineering measured values ​​in long-term monitoring.

[0185] Example 2: Based on Example 1, Example 2 provides a specific implementation case. This example selects a deep-water canyon reservoir (dam height 200m+) in Southwest China as the object, and the monitoring time is September 2023.

[0186] The detailed implementation steps are as follows:

[0187] 1. Data Acquisition: Acquiring Level-2 data from the GRACE-FO satellite; data was also acquired via a UAV carrying a cold atom gravimeter. The components are more accurate than 5E; the Sentinel-2 satellite extracts water surface masks, and the InSAR satellite monitors reservoir bank deformation.

[0188] 2. Fusion and Inversion: Utilizing Wiener filtering in The system integrates satellite and machine data at different stages; the reservoir quality anomalies are calculated by solving the joint inversion equation. .

[0189] 3. Parameter evolution calculation: based on reservoir area soil and rock parameters ( ), calculate the average void ratio during the 5th year of operation. Corresponding saturation density Hyperspectral inversion of mixed water body density .

[0190] 4. Final Solution: Geometric Apparent Volume Corrected by InSAR Given. Substitute into the core formula: .

[0191] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution, characterized in that, Includes the following steps: S1: Acquire satellite gravity data, airborne gravity gradient data, surface geometric deformation data, and spectral data; S2: Use airborne gravity gradient data to supplement satellite gravity data and construct a satellite-aircraft gradient fusion model based on Wiener filtering; S3: Based on the physical forward modeling constraints of the gravity gradient tensor and the fused gravity field data, a joint inversion objective function is constructed to convert the gravity anomalies and gravity gradient physical quantities on the observation surface into the mass distribution of the reservoir bottom sediment body, and accurately lock the boundary of the reservoir bottom sediment body. S4: Within the boundary of the sedimentary body at the bottom of the reservoir, the sedimentary mass is obtained by filtering out the signal related to the water level height through the time response function of the total mass change of water mass and sediment mass, based on the time series filtering. S5: Based on the soil mechanics consolidation model, the density and volume are transformed and calculated to obtain the dynamic change equation of the density of the sedimentary body at the bottom of the reservoir; S6: Finally, multi-temporal interferometry is used to constrain the geometric boundaries and volume of the water body, and the fluid background density is corrected based on hyperspectral inversion, so as to finally calculate and output the sediment volume distribution. The specific steps of step S6 include: S61: Define the mask matrix Used to define the integration space for gravity inversion: ； S62: Reservoir bank settlement rate obtained using multi-temporal interferometry technique The original topographic model (DEM) from the initial stage of database construction is corrected, and the current geometric apparent volume is calculated. That is, the total space of water and sediment; S63: Utilizing the multi-band reflectivity of hyperspectral satellites Inverting the instantaneous suspended matter mass concentration in the total space ; S64: According to the law of conservation of mass, the mixed density of sediment-laden turbid water... Adjusted according to volume percentage and organized into a summary of... The linear correction formula; S65: The final derivation yields... The final coupled solution equation after hyperspectral constraint is used to calculate and output the sediment volume distribution based on the final coupled solution equation. The final coupled solution equation is: ； In the formula, The volume of the sediment. This represents the total mass change obtained from the gravity gradient inversion. For real-time mixed fluid background density, This represents the InSAR-corrected geometric volume. This represents the dynamic saturated packing density.

2. The multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution according to claim 1, characterized in that, The specific steps of step S2 include: S21: Obtain the spherical harmonic coefficients of the time-varying gravity field of the satellite, calculate the gravitational potential, and derive the gravity gradient components at the satellite scale. S22: Utilize UAVs to fly over the reservoir area to observe the gravity gradient tensor and its vertical component, and use the Fourier domain downward extension operator to reduce the airborne observation data to the ground surface; S23: Based on gravity gradient components and airborne observation data, a Wiener filter-type fusion operator is constructed to obtain a computational model of the fused gradient field, namely the star-aircraft gradient fusion model.

3. The multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution according to claim 2, characterized in that, The calculation model for the gradient field is as follows: ； In the formula, These are satellite observations. These are airborne observations. This represents the satellite noise power spectrum. For satellite noise variance, The attenuation factor is the high correlation of satellite observations. This is the airborne noise power spectrum. This represents the variance of airborne noise.

4. The multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution according to claim 1, characterized in that, In step S3, the specific steps of the physical forward modeling constraint of the gravity gradient tensor include: S31: For any volume element at the bottom of the reservoir At the observation point The resulting vertical gradient response follows Decay law: ； In the formula, Due to density difference, For geocentric radial, Here are the coordinates of the observation point. The coordinates of the center of the bottom grid cell are the coordinates of the bottom grid cell. It is a volume element.

5. The multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution according to claim 4, characterized in that, In step S3, the specific steps for constructing the joint inversion objective function and converting the gravity anomalies and gravity gradient physical quantities on the observation surface into the mass distribution of the sedimentary body at the bottom of the reservoir include: S32: Define and construct the gravity kernel function matrix and gradient kernel function matrix And based on potential field theory, the gradient kernel function matrix is ​​obtained. Vertical component in Kernel function expression; where, The gravitational anomaly response per unit mass at the observation point. The gravitational gradient response per unit mass at the observation point; S33: Obtain the gravity anomaly and gravity gradient of the observed data vector, and construct a weighted least squares objective functional related to the spatial continuous distribution of gravity. S34: The weighted least squares objective functional is derived, and the sediment mass distribution is obtained by solving the preconditional conjugate gradient method.

6. The multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution according to claim 5, characterized in that, The formula for the weighted least squares objective functional is: ； In the formula, For the gravity data fitting term, For gradient data fitting term, For model regularization terms, Let be the parameter vector to be determined. This is the vector of observed surface gravity anomalies. This is the vector of observed gravity gradient values. and For the data weight matrix, As the gradient balancing factor, For regularization parameters, This is the roughness operator.

7. The multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution according to claim 1, characterized in that, The specific steps of step S4 include: S41: At any time Total mass change Decomposed into water mass and sediment quality : ; S42: Based on water level respectively and sediment transport flux The time integral is used to separately calculate the water mass and sediment mass: ; In the formula, For water density, This represents the water surface area corresponding to the water level. For water level changes, The deposition efficiency coefficient; S43: Use time series filtering to correlate with water level Highly correlated signal stripping yields pure sediment quality. .

8. The multi-source fusion monitoring method based on gravity gradient tensor and sedimentary evolution according to claim 1, characterized in that, The specific steps of step S5 include: S51: Based on the Terzaghi effective stress principle of the soil consolidation model, and calculating the evolution of void ratio according to compression and creep effects. ;in, For time ,depth The porosity at that location; S52: Based on the average dry density of sediments throughout the reservoir area The saturated bulk density was calculated. : In the formula, This refers to the specific gravity of the sediment particles. This represents the density of the water.

Images