System for assessing the risk of pollutant release from sediment aggregates based on three-dimensional microstructure inversion prediction

Abstract

The present application relates to the technical field of water environment monitoring and ecological risk assessment, solves the technical problems of the existing endogenous pollutant release risk assessment technology, and especially relates to a kind of bottom mud aggregate pollutant release risk assessment system based on three-dimensional microstructure inversion prediction, including high-resolution three-dimensional structure acquisition module, multi-physical field disturbance-diffusion simulation module, structure-function parameter inversion module, in-situ verification and online correction module and risk index generation and visualization module.The present application obtains the critical disturbance intensity of pollutant release according to the flux-disturbance relationship, and establishes a functional mapping with the above structural characteristic parameters to obtain a functional relationship;Combine field monitoring automatic calibration model;The ratio of environmental disturbance intensity and critical intensity is converted into a risk spatial distribution map of geographic information system.The system realizes structure quantization, threshold unification and model online correction, and is suitable for precise assessment and intuitive display in different water areas and scenarios.

Description

Technical Field

[0001] This invention relates to the field of water environment monitoring and ecological risk assessment technology, and in particular to a risk assessment system for pollutant release from sediment aggregates based on three-dimensional microstructure inversion prediction. Background Technology

[0002] Sediments in rivers, lakes, reservoirs, and wetlands play a crucial role in the "accumulation-re-release" of pollutants. Aggregates formed by the flocculation of fine particulate matter, organic matter, and microorganisms constitute complex pore-surface systems capable of adsorbing / encapsulating various pollutants, including nitrogen, phosphorus, organic carbon, and heavy metals. When disturbed by external forces such as wind, waves, ship wakes, sudden changes in water level, or human activities, the aggregate structure may be disrupted or rearranged, leading to the release of pollutants across the sediment-water interface and creating a considerable endogenous load. In eutrophic waters, this process is closely related to algal blooms, benthic habitat degradation, and long-term water quality deterioration. Therefore, quantitatively assessing aggregate release behavior and ecological risks under real-world disturbance scenarios is a crucial prerequisite for watershed management and aquatic ecological restoration.

[0003] I. Scale mismatch and mechanism gaps in existing methods.

[0004] Current assessment methods mainly fall into two categories: one is laboratory perturbation experiments (stirring, oscillation, steady flow, etc.) to determine the release per unit time; the other is statistical / empirical regression models based on monitoring data. The former is convenient for comparing different perturbation intensities under controllable conditions, but the experimental perturbation spectrum is singular, making it difficult to reproduce the hydrodynamic processes in natural water bodies that are multi-form superposition of steady-state, periodic, and pulse modes and change over time. More importantly, its results are often simplified to "apparent diffusion" parameters, making it difficult to reflect the true constraints of the three-dimensional pore network of aggregates on material migration. The latter, although it can correlate perturbation with water quality changes to a certain extent, often suffers from poor extrapolation and non-transferability due to the lack of quantitative characterization of structural quantities such as porosity (ψ), network connectivity (ρ), channel tortuosity (κ), surface roughness (σ), specific surface area (C), and morphological complexity (A). Some works have introduced two-dimensional or three-dimensional macroscopic hydrodynamic models, but these models focus on the overall flow field characteristics and usually equate pore-scale processes to constant terms, failing to reveal the true convection-diffusion-reaction pathways of pollutants in microscale channels.

[0005] II. Lack of pore size control relationship and critical behavior.

[0006] Mechanistically, solute transport within aggregates is controlled by a convection-diffusion-reaction coupling, with net flux determined by both diffusion and convection terms. This process is highly sensitive to microstructural parameters (e.g., effective diffusion is often determined by porosity and tortuosity). When external disturbances intensify to a certain extent, previously confined channels may become open, shifting the transport mechanism from diffusion-dominated to convection-dominated, resulting in a threshold-like jump in interfacial flux. This "critical disturbance intensity" is not a constant but is influenced by structural quantities such as ρ, κ, σ, C, and A, and also drifts with time-varying factors such as aggregate cementation state, microbial film growth, and seasonal compaction. Existing methods generally lack the means to calculate and quantify this critical threshold on the actual pore geometry, and also lack a technical path to establish a clear functional relationship between structural characterization results and the threshold.

[0007] III. Faults in Numerical Simulation and Structural Characterization.

[0008] In recent years, tools such as X-ray micro-CT, laser confocal microscopy (CLSM), and cryo-scanning electron microscopy (Cryo-SEM) have enabled the acquisition of three-dimensional voxel structures of aggregates at the micrometer scale. Based on three-dimensional segmentation and skeletonization, the topology and connectivity paths of pore networks can be further obtained. However, in engineering practice, these three-dimensional structural data are rarely used directly to establish multiphysics coupling models (coupling of incompressible Navier-Stokes and multi-component convection-diffusion-reaction) at the pore scale, let alone to deduce the critical perturbation intensity based on them. Even though some studies have performed calculations on ideal pore geometry (regular spheres, tube bundles), it is still difficult to transfer these calculations to the heterogeneous and irregular structures of natural aggregates.

[0009] IV. The "static" and "non-spatial" nature of the evaluation system.

[0010] At the risk assessment level, existing technologies mostly rely on point or profile indicators for threshold determination, lacking a unified and comparable risk index to integrate "environmental disturbance intensity" and "structurally determined critical intensity." Furthermore, there is a lack of mature practices for spatially classifying and visualizing risk results on GIS platforms, making it difficult to directly serve zoning control, engineering site selection, and operation and maintenance scheduling. Simultaneously, existing models generally lack online correction capabilities: when hydrodynamic scenarios or aggregate structures change over time, model parameters are difficult to update in a timely manner, resulting in insufficient predictive stability and portability.

[0011] In summary, current technology still has significant gaps in the chain of "3D microstructure acquisition and topology reconstruction – pore-scale multiphysics simulation – structural-functional parameter inversion to obtain critical perturbation intensity – in-situ verification and online correction – spatialized risk classification based on a unified index." Engineering urgently needs a system capable of: realistically reconstructing pore networks at the micron-voxel level and quantifying key structural parameters; establishing multiphysics coupling on this geometry and calculating flux-perturbation response; establishing a clear and calibrable functional mapping between structural characteristic parameters and critical perturbation intensity; performing online correction of model parameters based on in-situ monitoring data; and achieving low / medium / high risk classification and spatial display on GIS using a unified risk index. This would enable a quantifiable, portable, and spatially applicable risk assessment system for the release of endogenous pollutants. Summary of the Invention

[0012] To address the shortcomings of existing technologies, this invention provides a risk assessment system for pollutant release from sediment aggregates based on three-dimensional microstructure inversion prediction, solving several technical problems inherent in existing endogenous pollutant release risk assessment technologies. Specifically:

[0013] First, there is a lack of high-precision quantitative descriptions of key structural features of sediment aggregates, such as three-dimensional pore networks, connectivity, and surface roughness, making it difficult for structural information to be directly used in risk prediction. Second, disturbance-diffusion processes are mostly simulated at the macroscopic hydrodynamic level, without being combined with the multi-physics coupling effect of microscopic structures, making it difficult to reveal the true migration patterns of pollutants at the microscale. Third, there is a lack of methods for inverting critical disturbance intensity based on structure-function relationships, and risk classification standards rely on empirical settings, lacking a unified and comparable threshold system. Fourth, assessment models are mostly static and cannot be corrected online by combining on-site monitoring data, making it difficult to adapt to the spatiotemporal changes in water conditions.

[0014] To address the aforementioned technical problems, this invention provides a risk assessment system for pollutant release from sediment aggregates based on three-dimensional microstructure inversion prediction. The system includes: a high-resolution three-dimensional structure acquisition module, a multiphysics perturbation-diffusion simulation module, a structure-function parameter inversion module, an in-situ verification and online correction module, and a risk index generation and visualization module. Each module forms a closed-loop workflow through a data interface, from microstructure information acquisition to perturbation-diffusion numerical simulation, critical perturbation intensity calculation, in-situ verification and dynamic correction, and risk visualization output.

[0015] The high-resolution three-dimensional structure acquisition module is used to perform micron-level voxel imaging on river and lake sediment aggregates to obtain structural feature parameters of their three-dimensional pore network. .in, Porosity For the connectivity of pore networks, For surface roughness, Specific surface area For morphological complexity, The pore channel tortuosity is defined as the length of the curve along the channel centerline. The straight-line distance between the start and end points of the passage The ratio, that is .

[0016] Preferably, the micron-scale voxel imaging method may include laser confocal microscopy, X-ray computed tomography, cryo-scanning electron microscopy, or other imaging techniques capable of acquiring micron-scale voxel structures. A single method or a combination of methods may be selected based on the aggregate particle size and composition to meet the structural acquisition requirements of aggregates with different particle size ranges.

[0017] Preferably, after image acquisition, the original slice data is voxelized to generate a three-dimensional digital model containing pore spatial distribution, pore size variation and channel morphology, i.e., a voxelized three-dimensional image, to ensure the authenticity and integrity of the geometric data.

[0018] Preferably, the data processing and parameter extraction submodule is used to perform solid-liquid phase differentiation on the voxelized 3D image based on 3D Otsu threshold segmentation, extract pore connectivity paths using a 3D skeletonization algorithm, and generate a topological connectivity matrix. Based on this, the connectivity of the pore network is calculated using the following formula:

[0019]

[0020] in, This represents the total number of nodes in the three-dimensional porous network. , representing a node With nodes A boolean value indicating whether the two connections are connected, allowing direct or indirect reachability, and Structural characteristic parameters also include porosity. ,in The pore phase volume is the number of voxels that are labeled as pores / liquid phase after three-dimensional segmentation. With monomer volume The product of ; The total volume of the selected 3D reconstructed region of interest (ROI), i.e., the total number of voxels in the region. and The product of ,in Determined by the size of the imaging voxel. Surface roughness. Specific surface area morphological complexity and pore channel tortuosity ,in Defined as the length of the channel centerline curve The straight-line distance between the start and end points of the passage The ratio, that is For example, to characterize the effect of a structure on hydrodynamics, an effective permeability coefficient can be introduced, i.e.:

[0021]

[0022] in, The effective permeability coefficient under disturbed conditions; Used as a reference permeability coefficient; All results were obtained through experimental fitting or simulation calibration.

[0023] The multiphysics perturbation-diffusion simulation module is used to establish a coupled model of the Navier-Stokes equations and the multicomponent convection-diffusion-reaction equations on the aforementioned three-dimensional porous network. The expression is as follows:

[0024]

[0025]

[0026] in, For velocity field; For fluid density; For fluid viscosity; For pressure; This refers to the concentration of pollutant components; The diffusion coefficient is denoted as . For the reaction term; This is a volume force term; It is a time variable.

[0027] Preferably, the coupled model is solved using one of the following methods: finite volume method, finite element method, lattice Boltzmann method, or other numerical methods capable of multiphysics coupling calculation, to obtain simulation results.

[0028] Preferably, it can be based on the connectivity of the porous network. and pore channel tortuosity The spatial distribution of the local mesh is adaptively refined to improve the calculation accuracy of the velocity and concentration fields in high-gradient regions. The net pollutant flux is calculated using the following formula:

[0029]

[0030] Among them, the first item For diffusion flux, the second term The flux-disturbance intensity curve is used to determine the critical disturbance conditions for pollutant release.

[0031] The structure-function parameter inversion module calculates the critical perturbation intensity for pollutant release based on simulation results. and structural characteristic parameters The calibration yields the functional relationship, namely:

[0032]

[0033] in, These are model parameters obtained through experimental calibration or simulation fitting. If necessary, the input variables can be normalized or derived from the model parameters. The dimension of absorption is not a limitation. For example, in a physical interpretation, the relationship between the perturbation input energy density and effective penetration can also be introduced, namely:

[0034]

[0035]

[0036] in, Energy is input to the per unit volume disturbance; The critical flow rate that leads to release; The volume for effective perturbation.

[0037] Preferably, the critical disturbance intensity The methods for obtaining the regression prediction result include numerical inversion methods and machine learning-based regression prediction methods. The machine learning-based regression prediction method is at least selected from random forest regression, gradient boosting tree regression, deep neural network regression, or other algorithms suitable for regression prediction, and is obtained through structural feature parameters. By combining the feature inputs with the perturbation conditions, the generalization accuracy of the model under conditions of multiple structural differences can be improved.

[0038] The in-situ verification and online correction module includes a hydrodynamic disturbance unit and a multi-parameter online monitoring unit installed on the controllable disturbance reaction tank. The hydrodynamic disturbance unit applies disturbance conditions, including constant flow velocity, periodic fluctuations, or pulsed flow, to the controllable disturbance reaction tank. The multi-parameter online monitoring unit simultaneously measures changes in water quality parameters, including dissolved oxygen, oxidation-reduction potential, ammonia nitrogen, phosphate concentration, and dissolved organic carbon, and bases these measurements on the measured disturbance intensity. With the predicted critical disturbance strength The deviation automatically adjusts the model parameters in the functional relationship. and This enables dynamic adaptive online correction of functional relationships.

[0039] The risk index generation and visualization module also includes a risk grading submodule that classifies risk levels into low, medium, and high based on the risk index RRI and a preset threshold range, and that calculates or predicts the critical disturbance intensity based on historical measured data. The preset threshold range is dynamically adjusted to adapt to the risk characteristics of different water areas. This includes the risk index. The calculation formula is:

[0040]

[0041] in, This represents the total number of pollutants. For the first The risk weights of various pollutants can be normalized. This reflects its proportion of contribution to overall risk; For the first The critical disturbance intensity of a pollutant.

[0042] Preferably, the risk grading submodule will use the risk index Compared with the preset threshold range, it is divided into three levels: low, medium and high, and the threshold range can be dynamically adjusted according to the statistical distribution of historical measured data or the model prediction results.

[0043] Preferably, GIS technology is used to output a color-coded (or grayscale-coded) risk spatial distribution map, generating risk assessment results that include structural parameter contribution rates, critical disturbance intensity, and hydrodynamic condition analysis.

[0044] By employing the above technical solution, the present invention provides a risk assessment system for pollutant release from sediment aggregates based on three-dimensional microstructure inversion prediction, which has at least the following beneficial effects:

[0045] This invention deeply integrates the three-dimensional microstructure of sediment aggregates with pollutant release risk assessment: quantification is achieved through high-resolution voxel imaging and topological analysis. Multi-physics coupled models are constructed on the real pore geometry using multi-dimensional parameters, and the critical perturbation intensity is accurately determined by structure-function inversion. This will enable the standardization and portability of risk classification thresholds.

[0046] It is also equipped with in-situ verification and online correction mechanisms, which can dynamically update parameters based on on-site measured data to ensure the accuracy and stability of prediction under varying hydrodynamic and structural conditions.

[0047] Ultimately, through the RRI index and GIS visualization, the complex microstructural effects are transformed into intuitive risk levels and spatial distributions, providing directly applicable quantitative data for water environment governance, pollution prevention and control, and ecological restoration. Attached Figure Description

[0048] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0049] Figure 1 This is a structural block diagram of the bottom sediment agglomerate pollutant release risk assessment system of the present invention;

[0050] Figure 2 This is a schematic diagram of the three-dimensional structure acquisition and structural feature parameter extraction in this invention;

[0051] Figure 3 This is a schematic diagram of the multiphysics perturbation-diffusion simulation model and boundary conditions in this invention;

[0052] Figure 4 This is a spatial heatmap of the Risk Index (RRI) in Embodiment 1 of the present invention;

[0053] Figure 5 RRI in Embodiment 2 of the present invention N Spatial heat map;

[0054] Figure 6 RRI in Embodiment 2 of the present invention P Spatial heat map;

[0055] Figure 7 The weighted RRI in Embodiment 2 of the present invention total Spatial heat map;

[0056] Figure 8 This is a spatial heatmap of the Risk Index (RRI) in Embodiment 3 of the present invention;

[0057] Figure 9 The critical disturbance intensity D in Embodiment 3 of the present invention c Spatial distribution map. Detailed Implementation

[0058] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. This will allow for a full understanding of how the present application uses technical means to solve technical problems and achieve technical effects, and to facilitate its implementation.

[0059] This application proposes a risk assessment system for pollutant release from sediment aggregates based on three-dimensional microstructure inversion prediction, applicable to the management of endogenous pollution in rivers, lakes, and reservoirs. For example... Figure 1As shown, the system comprises five parts: a high-resolution 3D structure acquisition module, a multiphysics perturbation-diffusion simulation module, a structure-function parameter inversion module, an in-situ verification and online correction module, and a risk index generation and visualization module. It reconstructs a 3D pore network using micron-sized voxels and quantifies porosity. Pore ​​network connectivity Surface roughness Specific surface area morphological complexity With channel curvature ,like Figure 2 The diagram shown illustrates the 3D structure acquisition and parameter extraction process, including Otsu segmentation, skeletonization, and parameter extraction. Definition. A coupled model that solves incompressible flow and multi-component convection-diffusion-reaction on real pore geometry, such as... Figure 3 The multiphysics disturbance-diffusion simulation model and boundary condition diagram shown include inlet / outlet / solid wall and local mesh refinement sections. The critical disturbance intensity for pollutant release is determined based on the flux-disturbance relationship, and a functional mapping is established with the aforementioned structural characteristic parameters to obtain the functional relationship. Combined with on-site monitoring and automatic calibration models, the ratio of environmental disturbance intensity to critical intensity is converted into a risk spatial distribution map in a geographic information system. This system achieves quantifiable structure, standardized thresholds, and online model correction, making it suitable for accurate assessment and intuitive display in different water areas and scenarios.

[0060] Example 1: Three-dimensional structural characterization of sediment aggregates in lake A – calculation of critical disturbance strength – in-situ verification – risk classification.

[0061] This embodiment uses a short columnar sediment sample (sampling depth 0-10 cm) collected from the nearshore zone of a eutrophic lake A as an example to demonstrate the complete application process of the system of the present invention on real samples. This embodiment is for illustrative purposes only and is not intended to limit the scope of protection of the present invention.

[0062] 1. Sample and Imaging

[0063] Samples were collected and stored in a sealed container at 4°C. Intact aggregates with particle sizes ranging from 500 μm to 2 mm were selected. X-CT imaging (2 μm voxels) was used, and a three-dimensional pore-solid model was obtained after planar correction and voxel reconstruction.

[0064] 2. Structural feature parameter extraction (input data)

[0065] Perform 3D Otsu segmentation and 3D skeletonization on the voxel data to form a topological connectivity matrix. T ij The structural feature parameters (example values) obtained in this embodiment are as follows:

[0066] Porosity ψ = 0.41;

[0067] Connectivity ρ ( , T ij (∈{0,1}) = 0.72;

[0068] Surface roughness σ (characterized by surface fractal dimension) = 2.32;

[0069] Specific surface area C = 12.8 mm² -1 ;

[0070] The morphological complexity A (curvature statistic) = 0.67;

[0071] Channel curvature κ (κ=L) p / L 直 =1.84;

[0072] 3. Structure → Permeability Characteristic Mapping

[0073] To characterize the overall effect of microstructure on flow, the effective permeability coefficient is used, i.e.:

[0074]

[0075] Pick K 0 = 1.0 × 10 -12 m 2 , m 1 = 2.0 m 2 = 1.5 m 3 = 1.0. Substituting the values ​​from the previous section, we get:

[0076] =5.58×10 -14 m 2

[0077] 4. Multiphysics Disturbance – Diffusion Simulation (Model and Constants)

[0078] Solving in the three-dimensional pore domain, i.e.:

[0079]

[0080]

[0081] in, =998 kg·m -3 , μ =1.0×10 -3 Pa·s, above water cover Representative components D i = 7.0 × 10 -10 m 2 / s, the reaction term is analyzed using first-order analytical methods. R i = k des ( C s - C i ), k des = 2.0 × 10 -6 s -1 .

[0082] The inlet boundary is scanned using three operating conditions: constant flow, periodic fluctuation, and pulse; the local mesh is based on the connectivity of the pore network. and pore channel tortuosity Adaptive encryption. Net pollutant flux is calculated using the following formula:

[0083]

[0084] Among them, the first item For diffusion flux, the second term The flux-disturbance intensity curve is used to determine the critical disturbance conditions for pollutant release.

[0085] Simulation results: The flux-disturbance curve shows a sudden change, corresponding to the critical velocity. U c =0.045m / s.

[0086] 5. Inversion function and scaling (mapping dimensionless output to physical dimensions)

[0087] To map the structure-function inversion output to physical dimensions (m / s), a scaling coefficient γ is introduced before the functional relationship, as follows:

[0088]

[0089] This embodiment uses parameters obtained by fitting from a calibration dataset (n=12) of the same water area:

[0090] α 1 = 0.38, β 1 = 1.20; α 2 = 0.44, β 2 = 1.05;

[0091] α 3 = 0.21, β 3 = 0.92; α 4 = 0.17, β 4 = 1.10;

[0092] α5 = 0.15, β 5 = 0.88; α 6 = 0.30, β 6 = -0.75;

[0093] Substituting the structural parameters of this sample, we obtain the dimensionless output:

[0094] D c * = α 1ψ β1 +⋯+ α 6κ β6 =4.000701

[0095] The critical flow velocity obtained from the simulation U c =0.045m / s was used as the physical anchor point for this type of sample, and the scaling factor was determined:

[0096]

[0097] The critical perturbation intensity (physical dimension) of this sample is then obtained:

[0098] D c =γ. D c * =0.0450m / s

[0099] Among them, γ is obtained through calibration; if the water body is changed or the composition is significantly altered, it can be recalibrated using the same method to maintain dimensional consistency and portability.

[0100] 6. Energy density expression

[0101] Input energy using per unit volume perturbation:

[0102]

[0103] Substitution ρ f =998kg·m -3 , U c =0.045m / s:

[0104] e input =1.0105J / m 3

[0105] To obtain the total energy, multiply by the effective volume. V eff In this embodiment, energy density is used as the expression.

[0106] 7. In-situ verification and online correction

[0107] Under controlled disturbance reaction conditions, online monitoring of DO / ORP / / / DOC. Above water cover The net increase rate and the sudden change in flux at the bottom interface were used as threshold criteria to obtain the experimental critical intensity. D c,exp The initial prediction deviated from the actual measurement by +6.3%, which was corrected by online least squares. α i , β i}and K eff middle{ m 1, m 2, m After step 3, the deviation in the second round was less than 3%.

[0108] 8. Risk Index and Classification

[0109] The equivalent environmental disturbance intensity was measured at the on-site cross-section. D env =0.062m / s.

[0110] Risk Index:

[0111]

[0112] Based on the risk classification thresholds of this invention (preferably RRI < 0.8 for low, 0.8–1.0 for medium, and RRI > 1.0 for high, which can be dynamically adjusted based on historical data), RRI > 1.0 is judged as high risk. For example... Figure 4 As shown, a spatial heat map is output on the GIS with a 50m×50m grid, IDWp=2, and the parameter contribution rates are ranked as follows: ρ>ψ>κ>C>σ>A.

[0113] Example 2: Periodic disturbance assessment of fine-grained aggregates in the B channel area of ​​the river - multi-component threshold inversion and weighted risk classification.

[0114] This embodiment focuses on the navigation area of ​​urban river B, with the sampling point located in the slow-flow zone outside the main channel. This area is significantly affected by the periodic wake of ships, and the bottom sediment is mainly composed of fine-grained colloidal aggregates (representative particle size 50–300 μm). This embodiment fully demonstrates the process of "microstructure characterization → multiphysics (periodic) perturbation simulation → component-specific critical intensity inversion and scaling → in-situ verification and online correction → weighted RRI classification," and provides recalcifiable key values. This embodiment is for illustrative purposes only and not for limiting the scope of protection of this invention.

[0115] 1. Sample and Imaging

[0116] Samples were obtained using a columnar sampler (0–8 cm) and stored in a sealed container at 4°C. To characterize the microstructure of fine-particle aggregates, the high-resolution three-dimensional structure acquisition module used laser confocal microscopy (CLSM) to acquire fluorescently labeled multilayer sections (axial step 1 μm), and cryo-scanning electron microscopy (Cryo-SEM) was used to supplement the surface micromorphological details. The sections were registered and voxelized to reconstruct a three-dimensional pore-solid phase model.

[0117] 2. Structural feature parameter extraction (input data)

[0118] By performing 3D Otsu thresholding (solid / liquid phase separation), connected component labeling, and 3D skeletonization on the voxel data, a topological connectivity matrix is ​​formed. T ij The structural characteristics of this sample were calculated (example values):

[0119] Porosity ψ = 0.55;

[0120] Connectivity ρ ( , T ij (∈{0,1}) = 0.65;

[0121] Surface roughness σ (characterized by surface fractal dimension) = 2.45;

[0122] Specific surface area C = 18.5 mm² -1 (In this embodiment, to be consistent with the calibration set, the numerical values ​​are directly substituted into the inversion function, and the units are uniformly absorbed by the subsequent scaling coefficients γ.)

[0123] The morphological complexity A (curvature statistics index) = 0.74 (curvature statistics index);

[0124] Channel curvature κ (κ=L) p / L 直 =2.30;

[0125] To characterize the overall impact of microstructure on permeability, the effective permeability coefficient is used:

[0126] , K 0 = 1.0 × 10 -12 m 2 , m 1 = 2.0, m 2 = 1.5, m 3 = 1.0

[0127] Substituting into the above equation, we get K eff =6.85×10 -14 m 2 .

[0128] 3. Multiphysics perturbation – diffusion simulation (periodic boundary)

[0129] Solving the coupled model in a three-dimensional porous domain:

[0130]

[0131]

[0132] in =998 kg·m -3 , μ =1.0×10 -3 Pa·s, simulated using two pollutants:

[0133] : D P =7.0×10 -10 m 2 / s, R P = k des (Cs-C), k des =2.0×10 -6 s -1

[0134] DN=1.5×10 -9 m 2 / s, R N = k nit C, k nit =1.0×10 -5 s -1 (Approximate first-order nitrification under constant oxygen conditions).

[0135] Setting the period boundary for the inlet velocity: u (t)= U 0 [1+ a sin(2πft)], take a =0.6, f =1.2Hz; the outlet is a pressure boundary with no slippage on the solid wall. The mesh is in ρ High and large κ channel local adaptive encryption. Net pollutant flux:

[0136]

[0137] The first term is diffusion flux, and the second term is convection flux. Through J iThe abrupt change point of the disturbance curve determines the critical conditions.

[0138] Simulation results (cyclic operating conditions): Critical flow velocity U c,P =0.032m / s; Critical flow velocity U c,N =0.029 m / s. Corresponding per unit volume disturbance energy density:

[0139]

[0140] 4. Structure-function inversion and scaling (component-specific)

[0141] The same inversion structure as in Example 1 is used (coefficients are from the calibration set of this water area, n=12):

[0142]

[0143] ( α 1, β 1)=(0.38,1.20),( α 2, β 2) = (0.44, 1.05), ( α 3, β 3) = (0.21, 0.92),

[0144] ( α 4, β 4) = (0.17, 1.10), ( α 5, β 5) = (0.15, 0.88), ( α 6, β 6) = (0.30, -0.75)

[0145] Substitute the structural parameters of this sample (including C=18.5mm) -1 )have to:

[0146]

[0147] To make dimensionless Mapped to a critical perturbation intensity with units (m / s), a component-specific scaling coefficient γ is introduced:

[0148]

[0149]

[0150] Therefore:

[0151]

[0152] The final value matches the simulation critical value, ensuring dimensional and numerical closed-loop consistency.

[0153] 5. In-situ verification and online correction

[0154] Reproducing the stern spectrum in a controlled disturbance reaction tank ( a =0.6, f =1.2Hz), online monitoring of DO / ORP / / / DOC. Above water cover and The net increase rate and interface flux mutation were used as threshold criteria; the deviation between the first round of actual measurement and prediction: +7.8%, +9.1%. The system performs online least squares correction. α i , β i} and {γ P ,γ N With fine-tuning, the deviation in each round was reduced to <3%.

[0155] 6. Equivalent environmental disturbance and weighted risk index

[0156] Channel area flow velocity time series u (t)=U0[1+0.6sin(2π ft The equivalent characteristic velocity of )] is taken as the root mean square. Monitoring data from this section obtained... U 0 = 0.0377 m / s, based on this D env = U rms =0.0410m / s.

[0157] Component-based risk index:

[0158]

[0159]

[0160] Weighted total risk (example weights, considering eutrophication sensitivity) w P =0.6, w N =0.4):

[0161]

[0162] According to the classification of this invention (adaptive to historical data): RRI total A value >1.0 is considered high-risk. For example... Figure 5 , Figure 6 , Figure 7 As shown, the measurement points are interpolated into a 50m×50m grid, and the results are output respectively. , and The spatial heat map is provided, and the parameter contribution rates are ranked as ρ>ψ>κ>C>σ>A in the report.

[0163] Example 3: Multimodal Imaging Resolution Consistency and Storm Pulse Disturbance Assessment – ​​Cross-resolution Calibration and Consistency Check

[0164] This embodiment uses sediment aggregates (0-8cm) in the nearshore shallows of Reservoir C as the object to verify the consistency of structural characteristic parameters, critical disturbance intensity, and risk classification under different spatial resolutions and multimodal imaging conditions; it also assesses the risk under storm pulse-type disturbances (short-term strong winds / sudden increases in inflow). This embodiment is for illustrative purposes only and is not intended to limit the scope of protection of this invention.

[0165] 1. Sample and multimodal imaging

[0166] The samples were stored in a sealed container at 4°C after collection. To cover structures at different scales, X-CT (5μm voxel) was used to obtain the overall pore network, and CLSM (1μm axial step) was used to characterize the surface micro-morphology and fine channels; key interfaces were locally verified using Cryo-SEM. After registration based on mutual information, two independent reconstructions were output:

[0167] Coarse resolution (5μm) model (for "rapid engineering evaluation");

[0168] Fine-resolution (1μm) model (for "high-precision evaluation");

[0169] A multimodal fusion model (for cross-resolution calibration) is obtained by fusing voxel-consistent masks.

[0170] 2. Structural parameter extraction and results (input data, three models)

[0171] Perform 3D Otsu segmentation, connected component labeling, and skeletonization on 3D voxels to form a topological connectivity matrix. T ij Calculate: porosity ψ, connectivity Surface roughness σ (fractal dimension), specific surface area C (in this embodiment, the value in mm⁻¹ is directly substituted into the inversion function, and the dimensions are uniformly absorbed by the scaling coefficients), morphological complexity A, and channel tortuosity κ=L p / L 直 .get:

[0172] Coarse resolution: ψ=0.36, ρ=0.66, σ=2.28, C=10.5mm-1 A=0.62, κ=1.95

[0173] Fine resolution: ψ=0.39, ρ=0.70, σ=2.36, C=13.9mm -1 A=0.66, κ=1.88

[0174] Fusion model: ψ=0.385, ρ=0.69, σ=2.33, C=12.2mm -1 A=0.64, κ=1.90

[0175] Further, calculate the effective permeability coefficient:

[0176]

[0177] K 0 = 1.0 × 10 -12 m 2 m1=2.0, m2=1.5, m3=1.0

[0178] Get: coarse K eff =3.56×10 -14 m 2 Fine 4.74×10 -14 m 2 4.47×10 -14 m 2 .

[0179] 3. Multiphysics simulation of storm pulse disturbance (LBM preferred)

[0180] Solving in the three-dimensional pore domain:

[0181]

[0182]

[0183] in =998 kg·m -3 , μ =1.0×10 -3 Pa·s, above water cover Representative components D i =7.0×10 -10 m 2 / s, R i = k des ( C s - C i ),k des =2.0×10 -6 s -1 The Lattice Boltzmann Method (LBM) is preferred for explicit solution over complex pore geometry. Pulsed flow with high velocity is used at the boundary. U high =0.09m / s, low flow velocity U low =0.02m / s, duty cycle 20%, outlet is pressure boundary, solid wall with no slippage. Net pollutant flux is:

[0184]

[0185] The first term is diffusion flux, and the second term is convection flux. Criticality is determined by the abrupt change point of the flux-disturbance curve.

[0186] Simulation anchor point (fusion model): The critical flow velocity is obtained through simulation. U c =0.038m / s. The corresponding per unit volume disturbance energy density is:

[0187]

[0188]

[0189] ( α 1, β 1)=(0.38,1.20),( α 2, β 2) = (0.44, 1.05), ( α 3, β 3) = (0.21, 0.92)

[0190] ( α 4, β 4) = (0.17, 1.10), ( α 5, β 5) = (0.15, 0.88), ( α 6, β 6) = (0.30, -0.75)

[0191] Substituting the three sets of structural parameters, we get:

[0192] coarse resolution

[0193] Fine resolution

[0194] Fusion Model

[0195] Simulation anchor point of the fusion modelU c =0.038m / s calibration scaling factor

[0196]

[0197] Based on this, the critical strength of physical dimensions at different resolutions was obtained:

[0198] coarse resolution

[0199] Fine resolution

[0200] Fusion Model (Consistent with simulation anchor points)

[0201] Therefore, the initial difference across resolutions is approximately -11.6% (coarse) and +11.1% (fine); the online correction module of this invention is used to correct this difference. α i , β i The least squares fine-tuning of} and γ (combined with in-situ threshold experiments) eventually converged to a range that differed from the fusion model by less than 3%, meeting the engineering consistency requirements.

[0202] 5. In-situ verification and online correction

[0203] By reproducing the storm pulse sequence (U) in a controlled perturbation reaction tank high / U low Duty cycle 20%), online monitoring of DO / ORP / / DOC. Above water cover The net increase rate and the sudden change in interface flux were used as threshold criteria to obtain the experimental critical intensity. D c,exp The system will D c,exp The results were compared with predictions from three models, and online parameter adjustments were performed. After a second round of validation, the three models were compared with... D c,exp The deviations were all <3%.

[0204] 6. Environmental disturbance and risk classification

[0205] By converting the pulse sequence to the equivalent characteristic velocity (RMS):

[0206]

[0207] Risk index (in fusion model) D c =0.038m / s):

[0208]

[0209] According to the classification of this invention (adaptive to historical data): RRI A value >1.0 is considered high-risk. Figure 8 As shown, the measurement points are interpolated to a 50m×50m grid to output a spatial heatmap. If the initial coarse / fine resolution is used directly... D c Calculations show that RRI coarse = 1.311 and RRI fine = 1.043. Figure 9 As shown. After online correction, the three converged (deviation <5%), and using the fusion model as the standard for grading is more robust.

[0210] The core of this invention lies in constructing a formula based on GIS spatial visualization and quantitative evaluation. RRI=D env / D c Risk identification and zoning methods (or multi-component weighted methods) are used to obtain the environmental disturbance distance of the target area. D env Critical distance from substrate characteristics D c This method calculates and maps low, medium, and high risk indices, enabling a clear and intuitive classification of ecological risks to aquatic sediments. By seamlessly integrating numerical calculations with a geographic information system, it not only improves the accuracy and spatial representation of risk assessment but also provides a scientific and actionable basis for decision-making in water environment management, sediment restoration, and ecological regulation.

[0211] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Moreover, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0212] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. Since the above embodiments are substantially similar to the method embodiments, their descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0213] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A risk assessment system for pollutant release from sediment aggregates based on three-dimensional microstructure inversion prediction, characterized in that, include: A high-resolution 3D structure acquisition module is used to perform micron-level voxel imaging of river and lake sediment aggregates and acquire the structural feature parameters of their 3D pore network. The multiphysics perturbation-diffusion simulation module, connected to the high-resolution three-dimensional structure acquisition module, is used to establish a coupled model of the Navier-Stokes equations and the multi-component convection-diffusion-reaction equations on the three-dimensional pore network, and to obtain simulation results by coupled solving the incompressible flow and multi-component convection-diffusion-reaction on the real pore geometry. The structure-function parameter inversion module calculates the critical perturbation intensity for pollutant release based on the simulation results. And establish a calibrable functional relationship with the structural characteristic parameters; the critical disturbance intensity The methods for obtaining these values ​​include numerical inversion methods and regression prediction methods based on machine learning. The regression prediction methods include, but are not limited to, random forest regression algorithm, gradient boosting tree regression algorithm, and deep neural network regression algorithm; The expression for the functional relationship is: ; in, For structural characteristic parameters, Porosity For the connectivity of pore networks, For surface roughness, Specific surface area For morphological complexity, The pore channel tortuosity is defined as the length of the curve along the channel centerline. Straight-line distance between the start and end points of the passage The ratio, that is ; These are the model parameters obtained through experimental calibration or simulation fitting; The in-situ verification and online correction module is used to verify the critical perturbation intensity in a controlled perturbation reaction tank. And update the model parameters in the functional relationship based on real-time monitoring data. and ; The risk index generation and visualization module is used to generate and visualize the risk index based on the critical disturbance intensity. Calculate the risk index of pollutant release It outputs risk levels and spatial distribution maps, enabling a quantifiable, transferable, and spatially quantifiable assessment of the risk of endogenous pollutant release.

2. The sediment agglomerate pollutant release risk assessment system according to claim 1, characterized in that, The micron-level voxel imaging methods include laser confocal microscopy, X-ray computed tomography, or cryo-scanning electron microscopy, and a combination of one or more methods can be selected based on the particle size and composition of the river and lake sediment aggregates.

3. The sediment agglomerate pollutant release risk assessment system according to claim 1, characterized in that, The high-resolution three-dimensional structure acquisition module specifically includes: After image acquisition, the original slice data is voxelized to generate a voxelized 3D image containing pore spatial distribution, pore size variation and channel morphology. Solid-liquid phase differentiation is performed on the voxelized 3D image based on 3D Otsu thresholding, and a 3D skeletonization algorithm is used to extract pore connectivity paths and generate a topological connectivity matrix. Based on this, the connectivity of the pore network is calculated. The calculation formula is: ; in, This represents the total number of nodes in the three-dimensional porous network. Represents a node With nodes A Boolean value indicating whether the two connections are connected.

4. The sediment agglomerate pollutant release risk assessment system according to claim 1, characterized in that, The multiphysics perturbation-diffusion simulation module specifically includes: A coupled model of the Navier-Stokes equations and the multi-component convection-diffusion-reaction equations is established on the three-dimensional porous network, with the following expression: ; ; in, For velocity field; For fluid density; For fluid viscosity; For pressure; This refers to the concentration of pollutant components; The diffusion coefficient is denoted as . For the reaction term; This is a volume force term; It is a time variable; The coupled model is solved using the finite volume method, finite element method, or lattice Boltzmann method, and the connectivity of the pore network is considered during the iteration process. and pore channel tortuosity The spatial distribution adaptively refines the local mesh to improve the calculation accuracy of the velocity and concentration fields in the high gradient region, and the boundary conditions are set as constant velocity boundary, periodic fluctuation boundary or pulse flow boundary according to the actual hydrodynamic conditions.

5. The sediment agglomerate pollutant release risk assessment system according to claim 1, characterized in that, Surface roughness in the structural characteristic parameters The morphological complexity was calculated using the surface fractal dimension method. It was calculated using the three-dimensional surface mesh curvature distribution method.

6. The sediment agglomerate pollutant release risk assessment system according to claim 1, characterized in that, The in-situ verification and online correction module includes a hydrodynamic disturbance unit and a multi-parameter online monitoring unit installed on the controllable disturbance reaction tank; The hydrodynamic disturbance unit is used to apply disturbance conditions, including constant flow velocity, periodic fluctuation or pulse flow, to the controllable disturbance reaction tank. The multi-parameter online monitoring unit is used to simultaneously measure changes in water quality parameters including dissolved oxygen, oxidation-reduction potential, ammonia nitrogen, phosphate concentration, and dissolved organic carbon, and to measure changes based on the measured disturbance intensity. With the predicted critical disturbance strength The deviation automatically adjusts the model parameters in the functional relationship. and This enables dynamic adaptive online correction of functional relationships.

7. The sediment agglomerate pollutant release risk assessment system according to claim 1, characterized in that, The risk index generation and visualization module also includes a risk grading submodule that classifies risk levels into low, medium, and high based on the risk index RRI and a preset threshold range, and that calculates or predicts the critical disturbance intensity based on historical measured data. Dynamically adjust the preset threshold range to adapt to the risk characteristics of different water areas; The risk index The calculation formula is: ; in, The measured disturbance intensity of the target water area; Based on risk index Using GIS technology, output color-coded or grayscale-coded risk spatial distribution maps, and generate risk assessment results that include the contribution rate of structural characteristic parameters, critical disturbance intensity, and hydrodynamic condition analysis.

8. An application of a sediment aggregate pollutant release risk assessment system, employing the sediment aggregate pollutant release risk assessment system as described in any one of claims 1-7, characterized in that, It can be applied to the analysis of the structural characteristics of the bottom sediment of water bodies including rivers, lakes, reservoirs or wetlands, the prediction of pollutant release behavior and the classification and assessment of ecological safety risks.

Images