Atmospheric pollution migration prediction system under influence of meteorological parameter gain in complex terrain

Abstract

The application relates to the technical field of meteorological services, and discloses a system for predicting migration of atmospheric pollutants under influence of meteorological parameter gain in complex terrain, which comprises the following modules: a terrain topography analysis module that extracts spatial features of a target region and constructs a terrain topography feature matrix; a meteorological background preprocessing module that extracts a potential temperature gradient and a vertical wind shear parameter; an atmospheric layer evaluation module that determines a local Richardson number according to the parameters; a terrain gain correction module that corrects matrix weights according to the local Richardson number and outputs a terrain dynamic correction coefficient; a microscale field reconstruction module that couples the coefficient with background field data to generate a microscale meteorological field; and a diffusion trend prediction module that outputs a migration trajectory of atmospheric pollutants. The application can correct terrain dynamic gain weights in real time through identification of a feedback mechanism of stability of a layer and a forcing action of terrain, eliminate physical distortion in the process of reconstruction of a microscale flow field, and improve physical fitting degree of prediction of pollutant diffusion in a complex terrain region.

Description

Technical Field

[0001] This invention relates to an atmospheric pollutant migration prediction system that considers the influence of meteorological parameter gains under complex terrain, and belongs to the field of meteorological service technology. Background Technology

[0002] Numerical weather forecasting is a common technique for obtaining atmospheric element distributions from background fields, aiming to provide a driving force for microscale environmental evolution analysis. However, when this method is applied to complex terrain areas with significant undulations, existing techniques often use static terrain correction coefficients or spatial linear interpolation to process terrain forcing. Since the physical gain generated by the terrain is constrained by the real-time evolution of atmospheric thermodynamics, this approach, which ignores the dynamic modulation logic of atmospheric stratification stability on the terrain gain effect, makes it impossible for the system to capture the nonlinear gain of airflow caused by terrain topology in microscale environments.

[0003] Simply increasing the resolution of the computational grid or adding more physical observation stations to alleviate the above-mentioned biases not only leads to a surge in computational resource consumption, but also results in physical distortion of the meteorological field reconstruction results under critical operating conditions such as strong temperature inversion or thermal instability due to the lack of a real-time feedback mechanism for dynamic parameters such as topographic fingerprints and atmospheric Richardson numbers. This causes the subsequent fitting process of pollutant migration trajectories to lose its physical reference. In addition to the limitations of topographic analysis accuracy, existing technologies have mechanistic deficiencies in flow field reconstruction algorithms and boundary dynamic coupling. For example, Chinese invention patent application CN120874682A discloses a method suitable for complex terrain... The proposed method and system for simulating pollutant diffusion in mountainous terrain utilizes satellite elevation data to enhance the fineness of the terrain grid and fits boundary conditions using wind speed profile equations. The dynamic logic is anchored to the static correction of underlying surface roughness and slope parameters. However, the physical gain of the terrain flow field in actual complex mountainous conditions does not depend solely on the geometric topology but is modulated in real time by the atmospheric stratification and thermal state. The proposed method does not establish a nonlinear coupling feedback between the terrain dynamic response and local Richardson number thermodynamic parameters. In stable stratification or severe convection conditions, the system cannot adaptively adjust the terrain forcing contribution weight. The reconstructed flow field is prone to non-physical velocity jumps or distortion of the mass source term, limiting the physical fidelity of environmental risk assessment in complex conditions.

[0004] Therefore, how to analyze the mechanical forcing law of airflow on topographic geometry and topology, and couple atmospheric thermodynamic parameters to achieve adaptive reconstruction of meteorological element gains, thereby providing a physically accurate driving field for environmental risk assessment under complex working conditions, has become the technical problem to be solved by this invention. Summary of the Invention

[0005] To address the problems mentioned in the background art, the technical solution of the present invention is as follows: an atmospheric pollutant migration prediction system based on the influence of meteorological parameter gain under complex terrain, comprising:

[0006] The terrain and morphology analysis module is used to extract the surface elevation distribution data of the target area and use the surface elevation distribution data to determine the terrain spatial curvature and effective openness in order to construct a terrain and morphology feature matrix.

[0007] The meteorological background preprocessing module is used to acquire gridded meteorological background field data and extract potential temperature gradient parameters and vertical wind shear parameters at different gridded layers from the gridded meteorological background field data.

[0008] The atmospheric stratification assessment module is used to calculate the local Richardson number corresponding to the gridded meteorological background field data based on the potential temperature gradient parameter and the vertical wind shear parameter. The local Richardson number is used to characterize the stratification stability of the atmosphere when it passes through the target area.

[0009] The terrain gain correction module is used to calculate the correction weight of each component in the terrain feature matrix based on the deviation of the local Richardson number from the preset thermodynamic threshold, and output the terrain dynamic correction coefficient.

[0010] The microscale field reconstruction module is used to spatially couple gridded meteorological background field data with topographic dynamic correction coefficients to reconstruct a microscale meteorological field that satisfies mass conservation constraints and topographic dynamic forcing characteristics.

[0011] The diffusion trend prediction module is used to fit the diffusion path of pollutants within a microscale meteorological field and output the spatiotemporal migration trajectory of atmospheric pollutants.

[0012] Preferably, the terrain feature analysis module constructs the terrain feature matrix through the following sub-steps: Step S11, constructing a topological geometric lattice of the target area based on the digital elevation model; Step S12, calculating the spatial curvature tensor of each sampling point in the topological geometric lattice to describe the local changes in surface slope and aspect; Step S13, determining the effective openness based on the three-dimensional line-of-sight relationship of each sampling point; Step S14, encapsulating the spatial curvature tensor and the effective openness into a matrix to obtain the terrain feature matrix.

[0013] Preferably, when the meteorological background preprocessing module calculates the potential temperature gradient parameter, it obtains the potential temperature difference and height difference between adjacent grid layers, and determines the ratio of the potential temperature difference to the height difference as the potential temperature gradient parameter.

[0014] Preferably, the terrain gain correction module performs the following weight correction based on the local Richardson number: Step S41, when the local Richardson number is greater than 0.25, an exponential decay function is introduced to reduce the gain weight of the terrain feature matrix on near-surface wind speed; Step S42, when the local Richardson number is less than or equal to 0.25, the weight of the funnel effect feature quantity in the terrain feature matrix is ​​increased by the enhancement coefficient.

[0015] Preferably, the microscale field reconstruction module uses the tensor product mapping algorithm to map the terrain dynamic correction coefficients to three-dimensional voxels of the gridded meteorological background field data, thereby achieving downscaling of wind field components and temperature elements.

[0016] Preferably, when generating a microscale meteorological field, the microscale field reconstruction module introduces a three-dimensional divergence constraint operator to adjust the node vectors in the reconstructed flow field, so that the microscale meteorological field satisfies the fluid continuity equation.

[0017] Preferably, the diffusion trend prediction module fits the diffusion path through the following sub-steps: Step S61, injecting tracer particles characterizing the emission intensity of the pollution source into the microscale meteorological field; Step S62, calculating the residence time of the tracer particles on the leeward side of the terrain using the local circulation structure determined by the topographic dynamic correction coefficient; Step S63, correcting the settling displacement of the tracer particles based on the vertical velocity caused by topographic dynamic forcing.

[0018] Preferably, the system is applied to refined meteorological support services for mountainous environments, and the topographic dynamic correction coefficient is updated online by triggering local Richardson numbers.

[0019] Preferably, the system also includes a three-dimensional visual presentation module, which is used to overlay and display the terrain feature matrix, microscale meteorological field and spatiotemporal migration trajectory within the geographic information system.

[0020] Compared with the prior art, the beneficial effects of the present invention are:

[0021] 1. In the prediction of atmospheric pollutant migration, the local Richardson number of the background meteorological field is calculated to characterize the stability of atmospheric stratification. This number is then used as a dynamic adjustment factor for the topographic geometric fingerprint matrix, establishing a nonlinear coupling mechanism between topographic forcing and atmospheric thermodynamic state. This mechanism changes the traditional meteorological service approach of treating topographic gain as a static geometric function. It enables the system to adjust the modulation weight of topographic undulations on wind and temperature / humidity fields according to the real-time evolution of atmospheric thermodynamics. Under stable stratification conditions, the system enhances the physical compression effect of topography on airflow; under unstable convection conditions, the system automatically adjusts the contribution of topographic gain. Thus, even under diurnal or inversion conditions, the system can still reconstruct a microscale meteorological field that conforms to the true dynamic laws of the atmosphere, providing a highly physically accurate data foundation for subsequent pollutant migration prediction.

[0022] 2. The topographic geometric fingerprint, which includes spatial curvature and effective openness, is extracted using the topographic digital mapping module. This fingerprint is then used as a physical constraint template in the nonlinear reconstruction process, ensuring a high degree of fit between the reconstructed refined meteorological element field and the real surface topology. By spatially tensor coupling the background meteorological field and the topographic fingerprint matrix, the microscale flow field generated by the system logically follows the law of mass conservation and topographic dynamic constraints, eliminating non-physical phenomena such as airflow passing through mountains or near-surface airflow suspension that are prone to occur in conventional numerical simulations. This reconstruction method based on topographic features enables the fitting process of pollutant migration trajectories to truly reflect the characteristics of flow around, deceleration, and backflow caused by topography, thereby improving the processing accuracy of meteorological parameters in complex underlying surface environments.

[0023] 3. By constructing a nonlinear reconstruction engine based on tensor product operations, a large-scale downscaling of the background meteorological field can be achieved without introducing high-energy-consuming transient computational fluid dynamics simulation. The reconstruction engine combines offline preprocessing of terrain geometric fingerprints with real-time online correction of meteorological gain coefficients. It uses lightweight nonlinear mapping functions to replace the complex process of solving full dynamic equations. This logical architecture not only reduces the system's dependence on hardware resources, but also shortens the response time from background field input to refined meteorological element field output. This enables the system to perform real-time weather forecasting and path tracking in micro-scale scenarios such as industrial parks or densely built-up urban areas in response to sudden atmospheric environmental risks. Attached Figure Description

[0024] Figure 1 This is a data flow diagram of the atmospheric pollutant prediction system with meteorological parameter gain according to the present invention;

[0025] Figure 2 This is a diagram of the microscale meteorological field and trajectory prediction architecture for multi-module collaboration under complex terrain, as described in this invention.

[0026] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0027] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0028] An atmospheric pollutant migration prediction system based on the influence of meteorological parameter gains under complex terrain includes:

[0029] The terrain and morphology analysis module is used to extract the surface elevation distribution data of the target area and use the surface elevation distribution data to determine the terrain spatial curvature and effective openness in order to construct a terrain and morphology feature matrix.

[0030] The meteorological background preprocessing module is used to acquire gridded meteorological background field data and extract potential temperature gradient parameters and vertical wind shear parameters at different gridded layers from the gridded meteorological background field data.

[0031] The atmospheric stratification assessment module is used to calculate the local Richardson number corresponding to the gridded meteorological background field data based on the potential temperature gradient parameter and the vertical wind shear parameter. The local Richardson number is used to characterize the stratification stability of the atmosphere when it passes through the target area.

[0032] The terrain gain correction module is used to calculate the correction weight of each component in the terrain feature matrix based on the deviation of the local Richardson number from the preset thermodynamic threshold, and output the terrain dynamic correction coefficient.

[0033] The microscale field reconstruction module is used to spatially couple gridded meteorological background field data with topographic dynamic correction coefficients to reconstruct a microscale meteorological field that satisfies mass conservation constraints and topographic dynamic forcing characteristics.

[0034] The diffusion trend prediction module is used to fit the diffusion path of pollutants within a microscale meteorological field and output the spatiotemporal migration trajectory of atmospheric pollutants.

[0035] Preferably, the terrain feature analysis module constructs the terrain feature matrix through the following sub-steps: Step S11, constructing a topological geometric lattice of the target area based on the digital elevation model; Step S12, calculating the spatial curvature tensor of each sampling point in the topological geometric lattice to describe the local changes in surface slope and aspect; Step S13, determining the effective openness based on the three-dimensional line-of-sight relationship of each sampling point; Step S14, encapsulating the spatial curvature tensor and the effective openness into a matrix to obtain the terrain feature matrix.

[0036] Preferably, when the meteorological background preprocessing module calculates the potential temperature gradient parameter, it obtains the potential temperature difference and height difference between adjacent grid layers, and determines the ratio of the potential temperature difference to the height difference as the potential temperature gradient parameter.

[0037] Preferably, when calculating the local Richardson number, the atmospheric stratification assessment module follows the following physical quantification relationship: ,in, The acceleration due to gravity is taken as 9.8 m / s². The average potential temperature of the grid layer. These are potential temperature gradient parameters. These are the vertical wind shear parameters.

[0038] Preferably, the terrain gain correction module performs the following weight correction based on the local Richardson number: Step S41, when the local Richardson number is greater than 0.25, an exponential decay function is introduced to reduce the gain weight of the terrain feature matrix on near-surface wind speed; Step S42, when the local Richardson number is less than or equal to 0.25, the weight of the funnel effect feature quantity in the terrain feature matrix is ​​increased by the enhancement coefficient.

[0039] Preferably, the microscale field reconstruction module uses the tensor product mapping algorithm to map the terrain dynamic correction coefficients to three-dimensional voxels of the gridded meteorological background field data, thereby achieving downscaling of wind field components and temperature elements.

[0040] Preferably, when generating a microscale meteorological field, the microscale field reconstruction module introduces a three-dimensional divergence constraint operator to adjust the node vectors in the reconstructed flow field, so that the microscale meteorological field satisfies the fluid continuity equation.

[0041] Preferably, the diffusion trend prediction module fits the diffusion path through the following sub-steps: Step S61, injecting tracer particles characterizing the emission intensity of the pollution source into the microscale meteorological field; Step S62, calculating the residence time of the tracer particles on the leeward side of the terrain using the local circulation structure determined by the topographic dynamic correction coefficient; Step S63, correcting the settling displacement of the tracer particles based on the vertical velocity caused by topographic dynamic forcing.

[0042] Preferably, the system is applied to refined meteorological support services for mountainous environments, and utilizes local Richardson data... Triggers an online update of the terrain dynamics correction coefficient.

[0043] Preferably, the system also includes a three-dimensional visual presentation module, which is used to overlay and display the terrain feature matrix, microscale meteorological field and spatiotemporal migration trajectory within the geographic information system.

[0044] Example 1: In a mountainous industrial park scenario containing deep canyons and continuous mountain ranges, the background meteorological field is controlled by a weak pressure field, and the prevailing wind speed is below 2 m / s. The static terrain correction coefficient fails to capture the local wind speed nonlinear gain caused by the canyon topology, resulting in a spatial offset in the prediction of atmospheric pollutant migration. The terrain morphology analysis module extracts surface elevation distribution data of the target area, constructs a topological geometric lattice of the target area based on a digital elevation model, calculates the spatial curvature tensor of each sampling point to characterize the local changes in surface slope and aspect, determines the effective openness using the three-dimensional line-of-sight relationship of each sampling point, and encapsulates the spatial curvature tensor and effective openness into a matrix to generate a terrain morphology feature matrix. The meteorological background preprocessing module obtains the potential temperature difference and height difference between adjacent grid layers, and determines the ratio of the potential temperature difference to the height difference as the potential temperature gradient parameter. It then extracts the vertical wind shear parameters for different grid layers from the gridded meteorological background field data. The atmospheric stratification assessment module uses the potential temperature gradient parameter and the vertical wind shear parameter to calculate the local Richardson number, which characterizes the stratification stability as the atmosphere passes through the target area. The calculation formula is as follows: ,in, The acceleration due to gravity is taken as 9.8 m / s². The average potential temperature of the grid layer. These are potential temperature gradient parameters. For vertical wind shear parameters, the terrain gain correction module is based on the local Richardson number. The deviation from the preset thermodynamic threshold is used to calculate the terrain feature matrix. The correction weights of each component are calculated, and the output terrain dynamics correction coefficient and local Richardson number are calculated. When the value is greater than 0.25, the system reduces the terrain feature matrix. The gain weighting for near-surface wind speed avoids flow field distortion caused by forced gain due to static geometry under stable stratification conditions; while at local Richardson numbers... Under convective instability conditions less than or equal to 0.25, the system enhances the terrain feature matrix. The component weights representing the septum effect in the middle make the reconstructed flow field reflect the nonlinear gain characteristics of the atmosphere under the dynamic forcing of complex terrain.

[0045] The microscale field reconstruction module uses the tensor product algorithm to map the terrain dynamic correction coefficients to three-dimensional voxels of the gridded meteorological background field data. It introduces a three-dimensional divergence constraint operator to correct the vectors at each node in the reconstructed flow field, generating a microscale meteorological field that satisfies mass conservation constraints and fits the surface topology. It employs boundary condition decoupling and dimensionality reduction parameterization strategies to replace the energy-intensive transient solution of the all-fluid dynamics momentum equation. It transforms the momentum loss caused by airflow boundary layer friction and geometric drag due to complex terrain into a static calibration of the spatial curvature tensor components in the terrain feature matrix. It also converts the atmospheric stable stratification or unstable convection state into... The nonlinear dissipation and enhancement effects of turbulent momentum flux transport are extracted into dynamic weights of the local Richardson number-dominant dynamic correction coefficient. The three-dimensional momentum evolution mechanism is implicitly mapped to the geometric topological parameter system. Combined with the divergence smoothing cooperative constraint mechanism for solving the Poisson equation, the overall mass conservation law is strictly adhered to. On the microscale grid, the characteristics of topographic flow deceleration and leeward vortex that conform to physical common sense are approximately derived. The diffusion trend prediction module fits the pollutant diffusion trajectory in this microscale meteorological field. The residence time of tracer particles is calculated using the local circulation structure determined by the topographic dynamic correction coefficient, and the spatiotemporal migration trajectory of atmospheric pollutants is output.

[0046] Example 2: In a simulated environment with an altitude of 850m and featuring U-shaped canyon topography, an experimental environment was constructed using a computational fluid dynamics-based simulation platform. This platform obtained the flow field distribution by numerically solving the Navier-Stokes equations. The experimental input data originated from high-resolution gridded meteorological background field data and digital elevation model data with a spatial resolution of 30m; grid spacing... This value is used to describe the degree of discretization of the computational region. Its determination considers the trade-off between terrain feature capture accuracy and computational load. When the elevation change rate between sampling points is greater than 0.15, spatial aliasing of terrain features during reconstruction is avoided. The value range is set to 25m to 50m. Under the specific working conditions of this experiment... The value is 30m. Setting 0.15 as the critical threshold for triggering spatial frequency aliasing is based on the fact that, according to the Shannon-Nyquist spatial sampling theorem, the offline analytical verification results of various mountain profiles using Fast Fourier Transform (FFT) show that when the rate of change of the sampling point elevation exceeds the limit, the proportion of high-frequency harmonic energy contained in the terrain undulations increases sharply. If the original coarse basic grid is maintained, a technical compromise is made to forcibly truncate the key peak and valley characteristic signals. Only by executing trigger control to rigidly compress the upper limit of the sampling resolution to within the 50m limit can the divergence of discretized values ​​be suppressed. The experiment is divided into the sample group of this invention, control group one, and control group two. The sample group of this invention adopts the technical solution provided by this invention, control group one adopts the static terrain correction coefficient, and control group two directly outputs the original meteorological background field. The background wind speed is 1.5m / s and the potential temperature gradient parameter is... Under a stable stratification condition of 0.04 K / m, the atmospheric stratification assessment module determines the local Richardson number. The value is 0.38, which exceeds the preset thermodynamic threshold of 0.25; among which, the local Richardson number... The calculation formula is as follows: ,in, The acceleration due to gravity is taken as 9.8 m / s². The average potential temperature of the grid layer. These are potential temperature gradient parameters. These are the vertical wind shear parameters.

[0047] Richardson's number of authorities When the value is 0.38, the terrain gain correction module reduces the terrain feature matrix. The gain ratio for near-surface wind speed; the measurement results show that the deviation rate between the reconstructed near-surface wind speed value and the measured value of the sample group of this invention is 8.2%, while the wind speed deviation rate of control group one is 26.5% due to maintaining a constant funnel effect weight, and the deviation rate of control group two reaches 41.3%; with the local Richardson number The prediction error increases within the range of 0.15 to 0.45. The sample group of this invention utilizes nonlinear modulation logic to determine the terrain dynamic correction coefficient and adjust the feedback depth of terrain forcing, thus keeping the prediction error stable. In contrast, the prediction error of control group one increases with the local Richardson number. The increase shows a growth trend; regarding the local Richardson number The extreme state was simulated in an experiment with a background wind speed of 0.5 m / s and a surface temperature inversion. The local Richardson number... When the value exceeds 0.65, the terrain dynamic correction factor tends to stabilize, and reducing the weight has less than 2.5% impact on reconstruction accuracy; in the local Richardson number... Under a convection state of 0.08, the terrain gain correction module will adjust the terrain feature matrix. The weights of the components characterizing the canyon narrowing effect were adjusted to 1.8 times the initial weights to restore the dynamic gain of the canyon narrowing section. Pollutant diffusion results showed that the residence time of tracer particles on the leeward side of the terrain was measured at 12.6 minutes, a time deviation of 3.8% compared to the 13.1 minutes observed in the field. This was combined with different local Richardson data. The experimental results under gradient conditions show that the system provided by this invention can control the spatial deviation of the predicted migration trajectory of atmospheric pollutants in complex terrain areas to within 120m; data proves that through local Richardson data... Dynamically adjusting the terrain dynamic gain weights can improve the matching degree between the microscale meteorological field reconstruction and the real physical environment.

[0048] Example 3: In an underlying surface environment containing continuously distributed ridgelines and leeward slope circulation, the topographic feature matrix is ​​determined by the topographic feature analysis module. The effective openness parameter; the terrain topography analysis module establishes a hemispherical spatial search domain at the surface sampling points, with the sampling points as the origin, according to 5 azimuth increment and 2 Discretized ray projection is performed using elevation angle increments. The shortest straight-line distance between the ray and the simulated surface slope in the digital elevation model is calculated. All ray lengths are ranged from 0 to 360 degrees. and elevation angles from 0 to 90 degrees The effective openness, characterizing the terrain occlusion intensity around the sampling point, is obtained by integrating, summing, and normalizing within the spatial range; the terrain gain correction module receives the local Richardson number output by the atmospheric stratification assessment module. And determine the correction parameters, and use a piecewise mapping function to adjust the local Richardson number. When compared with a preset thermodynamic threshold of 0.25, a local Richardson number is detected. When the value is greater than 0.25, the system uses exponential correction logic to calculate the terrain feature matrix. Correction weights for the narrow tube effect component Among them, the correction weight With local Richardson numbers The relationship exhibits an exponential negative correlation, characterizing the thermodynamic inhibition experienced by the atmosphere as it passes through canyon regions under the constraint of stable stratification; while in local Richardson numbers... When the value is less than or equal to 0.25, the system switches to gain compensation logic, based on the local Richardson number. The correction weight increases linearly with the absolute value of the deviation from 0.25. Generate terrain dynamic correction coefficients.

[0049] The numerical algebraic expression for the exponential decay logic triggered under stable stratification conditions is as follows: ,in, To correct the weights after the decay, For the initial reference topological weights of the terrain, To control the convergence rate, a priori empirical decay constant is used; the logical algebraic relationship for triggering gain compensation under unstable convection conditions is as follows: ,in, The corrected weights are for linear compensation. To utilize the enhancement coefficients, the slope of the fitting is determined by performing a univariate linear regression analysis on a sample set of historical extreme gusts and prevailing surface wind speeds of the target mountain. After obtaining the topographic dynamic correction coefficients, the microscale field reconstruction module performs spatial tensor coupling on the three-dimensional voxels of the gridded meteorological background field data. For grid data at non-node locations in the three-dimensional voxels, the system uses a trilinear interpolation operator to calculate the dynamic correction gain at that point based on the coordinates of the eight adjacent nodes and the corresponding topographic dynamic correction coefficient components. Here, the spatial tensor coupling mapping operator is defined as... ,in, This represents the three-dimensional wind speed vector at coordinates i,j,k in the original gridded meteorological background field. This is the microscale three-dimensional wind speed vector after downscaling reconstruction. This is the tensor product operator. To transform the two-dimensional terrain feature matrix based on the terrain height attenuation projection mapping function. According to height level A three-dimensional spatial projection is performed to generate a corrected weight tensor that matches the three-dimensional voxel dimension of the background field. By performing a tensor product operation sequence on a grid-by-grid basis, the mechanical forcing physical gain caused by terrain undulations is quantitatively reduced in dimensionality and precisely coupled into the horizontal and vertical airflow components of the background flow field according to spatial distribution characteristics. The diffusion trend prediction module inputs the corrected three-dimensional wind field data into the Eulerian diffusion model and uses the velocity field that satisfies the mass conservation constraint to drive the spatiotemporal displacement calculation of the tracer particles. By monitoring the concentration gradient distribution of the tracer particles in the near-surface layer of the underlying surface, the migration trajectory of atmospheric pollutants reflecting the influence of terrain dynamic gain is output.

[0050] Example 4: In a mountainous industrial park scenario with uneven distribution of surface vegetation cover and soil moisture, the system determines the stratification stability judgment benchmark through thermodynamic threshold calibration procedures. The atmospheric stratification assessment module retrieves 24 hours of continuous historical meteorological observation data of the target area and calculates the local Richardson number including the entire diurnal cycle. The sequence was analyzed, and a cross-correlation matching relationship was established between it and synchronously acquired near-surface sensible heat flux data. This was then used in local Richardson number analysis. When the cross-correlation coefficient between the sensible heat flux and the trend of change reaches its maximum point, its corresponding value is... The numerical values ​​are confirmed as the preset thermodynamic threshold of the target area, thus making the judgment boundary of the terrain gain correction module correspond to the physical critical point driven by the microclimate characteristics of the target area. Considering the inherent time lag property of the vertical heat transfer process of atmospheric boundary layer turbulence, in order to eliminate the matching misalignment distortion caused by the phase shift of heat flux, before establishing the mathematical cross-correlation matching relationship, the dynamic time warping (DTW) algorithm is used to perform elastic deformation translation operation on the sensible heat flux time series axis. After accurately quantifying and deducting the phase difference experienced by the thermal disturbance signal as it diffuses upward from the surface forced source to the reference meteorological grid layer, the product integral judgment is performed on the aligned two-dimensional data sequence to ensure that the captured maximum point truly maps the critical moment of the surface layer state reversal in causal physical logic.

[0051] To ensure the stability of dynamic parameters under long-term monitoring conditions, an online calibration procedure for preset thermodynamic thresholds is established. The system continuously reads the historical sequence of cross-correlation coefficients within the 24-hour sliding time window preceding the current moment, calculates the variance volatility of the historical sequence, and determines that the initial microclimate reference baseline has decayed when the variance volatility exceeds the 15% threshold for three consecutive evaluation periods. This triggers an update command, extracts newly added near-surface sensible heat flux data from the last 72 hours to reconstruct the cross-correlation matching matrix, and recalculates the maximum point to cover the original preset thermodynamic threshold, maintaining consistency between the underlying thermodynamic judgment boundary and the actual evolving environmental baseline. When the system operates on an underlying surface composed of vegetation and buildings, the terrain morphology analysis module will use aerodynamic roughness length parameters... Injected terrain feature matrix To achieve spatial feature alignment, the terrain morphology analysis module analyzes the sub-grid-scale obstacle height distribution in the digital elevation model (DEM). Based on the 3D point cloud data captured by the airborne remote sensing platform equipped with multispectral lidar, and using the normalized vegetation index (NZVI) inverse algorithm, point-to-point differential stripping is performed on the raster layers of the digital surface model (DSM) and the bare land digital elevation model (DEM) to generate the difference dataset. The difference dataset accurately isolates and characterizes the vertical geometric protrusions of independent industrial buildings and tree canopy communities within a single computational grid at meter-level super-resolution. The dynamic roughness of each sampling point is determined based on the ratio of the obstacle's projected area to the cell area, and the dynamic roughness is used as a weight to apply to the spatial curvature tensor and effective openness to correct for kinetic energy loss caused by non-geometric friction. Under the premise of satisfying the mass conservation constraint, the microscale field reconstruction module maps the terrain dynamic correction coefficients to the roughness-corrected terrain morphology feature matrix. This generates a microscale meteorological field that incorporates topographic dynamic forcing and local friction effects, keeping the spatial deviation between the peak concentration location in the pollutant migration path prediction and the monitoring point within 80m.

[0052] Example 5: In deployment scenarios involving coastal hilly terrain and interference from sea and land breeze circulation, the system uses a terrain influence depth calibration method to determine the terrain feature matrix. The vertical attenuation coefficient is calculated by the terrain topography analysis module using a discretized grid search method to calculate the terrain feature variance within a 5km radius of the sampling point. This terrain feature variance is then used as input into the mapping function to calculate the cutoff threshold of the terrain dynamics in the vertical direction. ;in As the cutoff threshold for the impact of terrain dynamics in the vertical direction, when the height of the sampling grid point... satisfy Less than or equal to hour, The system determines the height of the sampling grid points from the ground based on... The ratio determines the topographic feature matrix. The component weights at this altitude layer cause the topographic dynamic correction coefficient to appear as a topographic forcing gain in the near-surface layer, while converging to the original meteorological background field in the free atmosphere.

[0053] When the microscale field reconstruction module performs tensor product mapping in three-dimensional voxel space, the system corrects the mesh deformation error caused by abrupt changes in terrain slope using an error compensation table, and the microscale field reconstruction module obtains the terrain slope parameters. ;in The local terrain slope angle at the sampling point is used, and the discretized interpolation correction operator under the corresponding slope interval is retrieved. The error compensation table is a discretized inverse mapping reference matrix established in the early stage of research and development through the joint analysis of flow field observation data from ideal terrain wind tunnel experiments and standard parabolic mesh stretching simulation tests. It systematically records the distribution law of residuals calculated by coordinate transformation Jacobian determinant under different slope gradient profiles. When the actual slope angle of the target area exceeds the mesh orthogonality tolerance threshold, the local terrain slope angle is directly used as the independent variable index to accurately match and extract the error from the error compensation table. Corresponding to the anti-aliasing filter correction factor, the risk of anomalous grid stretching and numerical overflow accumulated in the spatial tensor mapping at steep cliffs or deep ravines is forcibly neutralized in the form of algebraic difference. A three-dimensional divergence constraint operator that satisfies the continuity equation is injected into the grid layer of the reconstructed flow field. By iteratively solving the Poisson equation, the reconstructed wind speed field eliminates non-physical mass source terms while maintaining the topographic dynamic characteristics. The prediction results of pollutant migration paths show that the deviation rate between the vertical diffusion deflection angle of tracer particles in the alternating sea-land breeze stage and the lidar observation value is 5.2%.

[0054] In a mountainous diffusion scenario with complex thermodynamic interactions, the diffusion trend prediction module determines the displacement vector of the tracer particles, acquires the microscale meteorological field, and extracts the grid point velocity vector containing three-dimensional wind speed components. The system uses a Lagrange random walk method to calculate the spatial increment of the tracer particle within each integration step and directs it to the velocity vector. Superimposed by local Richardson numbers The system determines turbulence statistics to compensate for diffusion bias caused by atmospheric stratification; it acquires the circulation center coordinates defined by the topographic dynamic correction coefficient, calculates the residence time of the pollution plume due to circulation capture on the leeward side of the terrain by determining the Euclidean distance between the tracer particle coordinates and the circulation center coordinates, and outputs the spatiotemporal migration trajectory of atmospheric pollutants reflecting the depth of topographic dynamic gain. In the integrated deployment procedure of microscale meteorological field reconstruction and pollutant prediction, the system performs quantization alignment of the data interface. The data structure output by the microscale field reconstruction module is a three-dimensional voxel matrix. Each voxel in the three-dimensional voxel matrix contains the air temperature, air pressure, wind speed, and wind direction parameters at that spatial location. The diffusion trend prediction module realizes in-situ transmission of flow field data by reading the memory address of the three-dimensional voxel matrix, avoiding spatial interpolation errors caused by data resampling; the system acquires the initial emission intensity parameters of the pollution source, converts them into the initial release frequency of tracer particles, and performs iterative calculations with a time step of 1 second in the reconstructed microscale meteorological field, outputting pollutant diffusion results that satisfy the constraints of the fluid dynamics continuity equation.

[0055] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A system for predicting atmospheric pollutant migration under complex terrain conditions by influencing meteorological parameter gains, comprising: The terrain and morphology analysis module is used to extract the surface elevation distribution data of the target area and use the surface elevation distribution data to determine the terrain spatial curvature and effective openness in order to construct a terrain and morphology feature matrix. The meteorological background preprocessing module is used to acquire gridded meteorological background field data and extract potential temperature gradient parameters and vertical wind shear parameters at different gridded layers from the gridded meteorological background field data. The atmospheric stratification assessment module is used to calculate the local Richardson number corresponding to the gridded meteorological background field data based on the potential temperature gradient parameter and the vertical wind shear parameter. The local Richardson number is used to characterize the stratification stability of the atmosphere when it passes through the target area. The terrain gain correction module is used to calculate the correction weight of each component in the terrain feature matrix based on the deviation of the local Richardson number from the preset thermodynamic threshold, and output the terrain dynamic correction coefficient. The microscale field reconstruction module is used to spatially couple gridded meteorological background field data with topographic dynamic correction coefficients to reconstruct a microscale meteorological field that satisfies mass conservation constraints and topographic dynamic forcing characteristics. The diffusion trend prediction module is used to fit the diffusion path of pollutants within a microscale meteorological field and output the spatiotemporal migration trajectory of atmospheric pollutants.

2. The atmospheric pollutant migration prediction system based on the influence of meteorological parameter gains under complex terrain as described in claim 1, characterized in that, The terrain feature analysis module constructs the terrain feature matrix through the following sub-steps: Step S11, constructing a topological geometric lattice of the target area based on the digital elevation model; Step S12, calculating the spatial curvature tensor of each sampling point in the topological geometric lattice to describe the local changes in surface slope and aspect; Step S13, determining the effective openness based on the three-dimensional line-of-sight relationship of each sampling point; Step S14, encapsulating the spatial curvature tensor and the effective openness into a matrix to obtain the terrain feature matrix. .

3. The atmospheric pollutant migration prediction system based on the influence of meteorological parameter gains under complex terrain as described in claim 1, characterized in that, When the meteorological background preprocessing module calculates the potential temperature gradient parameter, it obtains the potential temperature difference and height difference between adjacent grid layers, and determines the ratio of the potential temperature difference to the height difference as the potential temperature gradient parameter.

4. The atmospheric pollutant migration prediction system based on the influence of meteorological parameter gains under complex terrain according to claim 1, characterized in that, The terrain gain correction module performs the following weight correction based on the local Richardson number: Step S41, when the local Richardson number is greater than 0.25, an exponential decay function is introduced to reduce the gain weight of the terrain feature matrix on near-surface wind speed; Step S42, when the local Richardson number is less than or equal to 0.25, an enhancement coefficient is used to increase the terrain feature matrix. Weights of the characteristic quantities of the narrow tube effect.

5. The atmospheric pollutant migration prediction system based on the influence of meteorological parameter gains under complex terrain according to claim 1, characterized in that, The microscale field reconstruction module uses the tensor product mapping algorithm to map the terrain dynamic correction coefficients to three-dimensional voxels of the gridded meteorological background field data, thereby achieving downscaling of wind field components and temperature elements.

6. The atmospheric pollutant migration prediction system based on the influence of meteorological parameter gains under complex terrain according to claim 1, characterized in that, When generating a microscale meteorological field, the microscale field reconstruction module introduces a three-dimensional divergence constraint operator to adjust the vectors of each node in the reconstructed flow field, so that the microscale meteorological field satisfies the fluid continuity equation.

7. The atmospheric pollutant migration prediction system based on the influence of meteorological parameter gains under complex terrain according to claim 1, characterized in that, The diffusion trend prediction module fits the diffusion path through the following sub-steps: Step S61, injecting tracer particles characterizing the emission intensity of the pollution source into the microscale meteorological field; Step S62, using the local circulation structure determined by the topographic dynamic correction coefficient, calculating the residence time of the tracer particles on the leeward side of the terrain. Step S63: Correct the sedimentation displacement of the tracer particles based on the vertical velocity caused by the terrain dynamics.

8. The atmospheric pollutant migration prediction system based on the influence of meteorological parameter gain under complex terrain according to claim 1, characterized in that, The system is applied to provide refined meteorological support services for mountainous environments, and utilizes local Richardson data. Triggers an online update of the terrain dynamics correction coefficient.

9. The atmospheric pollutant migration prediction system based on the influence of meteorological parameter gains under complex terrain according to claim 1, characterized in that, The system also includes a 3D visual presentation module, which is used to overlay and display the terrain feature matrix, microscale meteorological field, and spatiotemporal migration trajectory within the geographic information system.

Images