Self-adaptive pollution diffusion forecasting method and system based on FLEXPART-WRF

Abstract

The embodiment of the invention provides a self-adaptive pollution diffusion forecasting method and system based on FLEXPART-WRF, and is applied to the technical field of pollution detection. The method comprises the following steps: generating a self-adaptive WRF simulation grid of a pollution accident; determining a research area, and performing three-dimensional meteorological field simulation on the research area by using the downscaling meteorological diagnosis model to obtain refined three-dimensional meteorological field data; utilizing an FLEXPART-WRF model to determine a simulated migration track of the pollutants, simulated particle concentration of a self-adaptive WRF simulation grid and simulated dry-wet sedimentation; and determining simulated weather forecast and simulated pollutant diffusion forecast. In this way, through adaptive grid generation, accurate meteorological field simulation and real-time prediction of pollutant diffusion, accurate pollutant diffusion information, efficient automatic simulation and fine pollution source analysis can be provided, and a quantitative and practical decision support tool is provided for environment management.

Description

Technical Field

[0001] The present disclosure relates to the technical field of pollution detection, and in particular to a FLEXPART-WRF-based adaptive pollution diffusion prediction method and system. Background Art

[0002] In recent years, with the rapid development of social economy and industry, a series of sudden air pollution incidents caused by industrial pollution sources have occurred frequently, causing great damage to the ecological environment and public safety.

[0003] Today, monitoring stations are first deployed and established around pollution sources and potentially affected areas. These stations then automatically or manually collect environmental samples, such as air and water, at predetermined intervals (e.g., hourly, daily, or weekly). These samples are then sent to a laboratory, where specialized instruments conduct detailed analysis of pollutant concentrations. This process directly captures real-time or periodic data on pollutant concentrations, providing a key basis for assessing pollution status and spread trends.

[0004] However, although the method of collecting samples and analyzing pollutant concentrations by setting up monitoring stations near pollution sources and in potentially affected areas can directly reflect the pollution situation, it is limited by the layout of monitoring stations, insufficient data representativeness and timeliness challenges, and may not be able to fully and timely capture the overall picture and latest dynamics of pollution spread. Summary of the Invention

[0005] The present disclosure provides an adaptive pollution diffusion prediction method based on FLEXPART-WRF, which solves the problem that the existing technology for assessing pollution status and diffusion trends is to collect samples and analyze pollutant concentrations by setting up monitoring stations near pollution sources and potentially affected areas. Although the method can directly reflect the pollution status, it is limited by the layout of monitoring stations, insufficient data representativeness and timeliness challenges, and cannot comprehensively and timely capture the overall picture and latest dynamics of pollution diffusion.

[0006] In a first aspect, an embodiment of the present application provides an adaptive pollution diffusion prediction method based on FLEXPART-WRF, the method comprising:

[0007] Obtaining the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution accident, and the amount of pollution released by the pollution accident; determining the pollution type according to the time of occurrence of the pollution accident, the time of end of the pollution accident, the type of pollution accident, and the amount of pollution released by the pollution accident; and generating an adaptive WRF simulation grid according to the longitude and latitude information of the pollution release point and a preset adaptive WRF simulation grid generation method;

[0008] Determine the study area based on the latitude and longitude information of the pollution release point and the adaptive WRF simulation grid, and use the downscaling meteorological diagnostic model to perform a three-dimensional meteorological field simulation on the study area to obtain refined three-dimensional meteorological field data;

[0009] Using the FLEXPART-WRF model, the simulated migration trajectory of pollutants is determined based on the refined three-dimensional meteorological field data, the time when the pollution accident occurred, the time when the pollution accident ended, the longitude and latitude information of the pollution release point, the pollution type, and the amount of pollution accident released, as well as the simulated particle concentration and simulated dry and wet deposition of the adaptive WRF simulation grid.

[0010] A simulated meteorological forecast is determined based on the refined three-dimensional meteorological field data, and a simulated pollutant diffusion forecast is determined based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition.

[0011] According to a second aspect of the present application, there is provided an adaptive pollution diffusion prediction system based on FLEXPART-WRF, the system comprising:

[0012] A data acquisition module is used to obtain the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution accident, and the amount of pollution released by the pollution accident; determine the pollution type based on the time of occurrence of the pollution accident, the time of end of the pollution accident, the type of pollution accident, and the amount of pollution released by the pollution accident; and generate an adaptive WRF simulation grid based on the longitude and latitude information of the pollution release point and a preset adaptive WRF simulation grid generation method;

[0013] A refined three-dimensional meteorological field data generation module is used to determine the study area based on the latitude and longitude information of the pollution release point and the adaptive WRF simulation grid, and to perform a three-dimensional meteorological field simulation on the study area using a downscaling meteorological diagnostic model to obtain refined three-dimensional meteorological field data;

[0014] A simulation module is used to determine the simulated migration trajectory of pollutants based on the refined three-dimensional meteorological field data, the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the pollution type, and the amount of pollution accident release using the FLEXPART-WRF model, as well as to determine the simulated particle concentration of the adaptive WRF simulation grid and simulated dry and wet deposition;

[0015] A forecast generation module is used to determine a simulated meteorological forecast based on the refined three-dimensional meteorological field data, and to determine a simulated pollutant diffusion forecast based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition.

[0016] In a third aspect, an embodiment of the present application provides an electronic device comprising a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the method described in the first aspect.

[0017] In an embodiment of the present application, the time of occurrence of a pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution accident, and the amount of pollution released are obtained; the pollution type is determined based on the time of occurrence of the pollution accident, the time of end of the pollution accident, the type of pollution accident, and the amount of pollution released; and an adaptive WRF simulation grid is generated based on the longitude and latitude information of the pollution release point and a preset adaptive WRF simulation grid generation method; a study area is determined based on the longitude and latitude information of the pollution release point and the adaptive WRF simulation grid, and a downscaling meteorological diagnostic model is used to perform a three-dimensional meteorological field simulation on the study area to obtain refined three-dimensional meteorological field data; a FLEXPART-WRF model is used to determine the simulated migration trajectory of pollutants based on the refined three-dimensional meteorological field data, the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution, and the amount of pollution released, as well as the simulated particle concentration and simulated dry and wet deposition of the adaptive WRF simulation grid; a simulated meteorological forecast is determined based on the refined three-dimensional meteorological field data, and a simulated pollutant diffusion forecast is determined based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition. The FLEXPART-WRF-based adaptive pollution dispersion forecasting method, with its adaptive grid generation, precise meteorological field simulation, and real-time pollutant dispersion forecasting, provides precise pollutant dispersion information. Efficient automated simulation and sophisticated pollution source analysis provide quantitative and practical decision-making support tools for environmental management.

[0018] It should be understood that the contents described in the Summary of the Invention section are not intended to limit the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are provided for a better understanding of the present disclosure and do not constitute a limitation of the present disclosure. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, among which:

[0020] Figure 1 1 is a flow chart of the adaptive pollution diffusion prediction method based on FLEXPART-WRF provided in Example 1 of the present application;

[0021] Figure 2This is a schematic diagram of point source adaptive grid generation provided in Example 1 of the present application;

[0022] Figure 3 This is a schematic diagram of line source adaptive grid generation provided by an embodiment of the present application;

[0023] Figure 4 This is a schematic diagram of generating a matching grid for a multi-pollution source simulation area provided in Example 1 of the present application;

[0024] Figure 5 Schematic diagram of the process of the adaptive pollution diffusion prediction method based on FLEXPART-WRF provided in Example 2 of the present application;

[0025] Figure 6 Schematic diagram of the structure of the adaptive pollution diffusion prediction system based on FLEXPART-WRF provided in Example 3 of the present application;

[0026] Figure 7 This is a schematic diagram of the structure of the electronic device provided in Example 4 of the present application. DETAILED DESCRIPTION

[0027] To make the purpose, technical solutions, and advantages of the embodiments of the present disclosure more clear, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present disclosure.

[0028] In this document, the term "and / or" simply describes a relationship between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document generally indicates that the related objects are in an "or" relationship.

[0029] The following describes in detail the adaptive pollution diffusion prediction method based on FLEXPART-WRF provided in the embodiment of the present application through specific embodiments and application scenarios in conjunction with the accompanying drawings.

[0030] Example 1

[0031] Figure 1 This is a flow chart of the adaptive pollution diffusion prediction method based on FLEXPART-WRF provided in Example 1 of this application. Figure 1 As shown, the specific steps include:

[0032] S101, obtaining the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution accident, and the amount of pollution released by the pollution accident, determining the pollution type based on the time of occurrence of the pollution accident, the time of end of the pollution accident, the type of pollution accident, and the amount of pollution released by the pollution accident, and generating an adaptive WRF simulation grid based on the longitude and latitude information of the pollution release point and a preset adaptive WRF simulation grid generation method.

[0033] First, this solution can be used to collect pollution incident information, identify pollution types, and generate an adaptive WRF simulation grid. Next, a downscaled meteorological model is used to simulate the three-dimensional meteorological field in the study area. Subsequently, the FLEXPART-WRF model is used to simulate pollutant migration trajectories, particle concentrations, and dry and wet deposition. Finally, a weather forecast and pollutant dispersion forecast scenario is generated.

[0034] Based on the above usage scenarios, it can be understood that the execution subject of this application can be an adaptive pollution diffusion forecasting system based on FLEXPART-WRF, and no excessive limitations are made here.

[0035] In this embodiment, the time when the pollution accident occurs may be the time when pollutants begin to be discharged, usually the time when the pollution source begins to leak or discharge.

[0036] The end time of a pollution accident may be the time when pollutant emissions end, usually the time when the pollution source stops discharging or leaking.

[0037] The longitude and latitude information of the pollution release point can be the geographical location of the pollution source, usually expressed in the form of longitude and latitude. This is the source location of the pollutant emission.

[0038] Pollution accident types describe different types of sudden pollution incidents, primarily categorized by the nature of the pollution source, the type of pollutant, the method of pollutant release, and the impact of the pollution incident on the environment and human health. Specifically, these include toxic and hazardous substance pollution accidents: pollution accidents caused by the leakage or abnormal discharge of toxic and hazardous chemicals during production and daily life due to improper production, use, storage, transportation, or discharge. Toxic gas pollution accidents are actually a type of this type of accident; common toxic and hazardous gases include carbon monoxide, hydrogen sulfide, chlorine, and ammonia. Explosion accidents: explosions and fires caused by flammable and explosive substances. Pesticide pollution accidents: pollution accidents caused by the leakage of highly toxic pesticides due to accidental or improper use during production, storage, and transportation. Radioactive pollution accidents: pollution accidents caused by improper operation during the production, use, storage, and transportation of radioactive materials, resulting in nuclear radiation hazards. Oil pollution accidents: pollution accidents caused by the leakage of crude oil, fuel oil, and various oil products due to accidental or improper use during production, storage, transportation, and use. Abnormal wastewater discharge pollution accidents: Due to improper operation or accidents, a large amount of high-concentration water is suddenly discharged into surface water bodies, causing the water quality to suddenly deteriorate.

[0039] The amount of pollution released by a pollution accident can be the total amount of pollutants released by a pollution source within a certain period of time, usually expressed in mass units (such as kilograms, tons, etc.).

[0040] Pollution types are a classification of pollution sources and pollutants, which determine the pattern of pollutant diffusion and the scope of impact. Specifically, they can include point source pollution: pollutants are released from a fixed source location, usually with a slow diffusion rate and greatly affected by wind direction and meteorological conditions; line source pollution: pollutants are released from a linear source, with a relatively concentrated diffusion path but greatly affected by meteorological conditions; and non-point source pollution: pollutants are evenly diffused over a larger area, with pollutant concentrations gradually decreasing and the diffusion range being wider.

[0041] The preset adaptive WRF simulation grid generation method can use high resolution in key simulation areas, while the parent domain uses coarse resolution to provide boundary conditions. This not only avoids the use of fine grids across the country that leads to the consumption of a large amount of computing resources and affects the timeliness of forecasts, but also ensures the simulation accuracy of key areas. The system sets a three-dimensional meteorological field simulation grid suitable for key research areas based on the latitude and longitude coordinates of the nuclear pollution occurrence point. Specifically, it can include point source adaptive grid generation method, line source adaptive grid generation method, surface source adaptive grid generation method and multi-pollution source simulation area matching, Figure 2 This is a schematic diagram of point source adaptive grid generation provided by an embodiment of the present application, such as Figure 2As shown in the figure, when the pollution source is a point source, assuming the release point is C, the release point's location (i, j) in the WRF parent domain d01 grid is automatically calculated. Based on the parent domain's grid resolution dx and the user's desired diffusion radius R, the coordinates of the southwest ws(I, J) and northeast en(I, J) points are expanded by (R / dx) grids in each direction. The resulting grid information is then converted into a format recognizable by the WRF model. To account for the model's boundary conditions, the grid boundary must be at least 5 grid intervals (d01) from the outer grid boundary; otherwise, the program terminates. Here, d01 is the first layer of the WRF simulation grid, domain1, and d02 is the nested grid layer, domain2.

[0042] Figure 3 This is a schematic diagram of line source adaptive grid generation provided by an embodiment of the present application, such as Figure 3 As shown in the figure, when the pollution source is a line source, first locate the row and column positions C1 and C2 of the two points in the WRF grid (d01). Then, using the point source calculation method, calculate the grid areas of the two points and merge them into a single grid. d01 is the first layer of the WRF simulation grid, domain1, and d02 is the nested grid layer, domain2.

[0043] When the pollution source is a surface source, first locate the row and column positions C1, C2, C3, and C4 of the four points in the WRF grid (d01), calculate the grid areas of the four points according to the point source calculation method, and then merge them into one grid.

[0044] Figure 4 This is a schematic diagram of the grid generation for the multi-pollution source simulation area matching provided by the embodiment of the present application. Figure 4 As shown in the figure, when there are multiple pollution sources, the corresponding grid areas are first generated for each pollution source, and the grid spacing (d01) between each grid is calculated. If the grid spacing is less than 5 or there is overlap between grids, the corresponding grids are merged. If there are five release points, the five grid areas are first calculated, and then the grids with closer distances are merged into a large grid. Here, d01 is the first layer of the WRF simulation grid, domain1; d02 is the nested grid layer, domain2; d03 is the nested grid layer, domain3; and d04 is the nested grid layer, domain4.

[0045] Adaptive WRF simulation grids can be a way to dynamically adjust the grid size based on the pollutant diffusion process. It refines the grid around the pollution source, using higher spatial resolution near the pollution source and in high-concentration areas, and lower resolution in more distant areas, thereby improving computational efficiency and accuracy.

[0046] The following five core parameters can be obtained, specifically, the time of occurrence of the pollution accident: the starting time of the pollution event, which is usually obtained by the pollution accident detection system, accident report or monitoring system.

[0047] Pollution incident end time: The end time of the pollution incident, which may be obtained from site cleanup and recovery conditions, sensor data or incident end report.

[0048] Pollution release point latitude and longitude information: The geographic location of the pollution source. This can usually be obtained from the location of the accident or sensor data (such as GPS positioning).

[0049] Type of pollution accident: such as toxic and hazardous substance pollution accidents, explosion accidents, pesticide pollution accidents, etc. This information is usually determined by the accident-related units or emergency response teams through preliminary investigation and testing of the accident.

[0050] Pollution accident release volume: refers to the amount of pollutants released into the environment by a pollution source during an accident. This can be obtained through monitoring equipment, calculation of flow rates, and pollutant concentrations.

[0051] The most important parameter is the pollution accident type, which directly determines the characteristics of the pollution source. Common pollution accident types include:

[0052] Pollution accidents involving toxic and hazardous substances: such as toxic gas leaks, may be point source pollution.

[0053] Explosion accidents: usually result in the instantaneous release of large amounts of pollutants, which can also be regarded as point source pollution, but the way it spreads may be more violent.

[0054] Pesticide pollution accidents: usually related to agricultural spraying and are non-point source pollution.

[0055] Radioactive contamination accident: generally caused by leakage of radioactive materials, usually a point source pollution.

[0056] Oil pollution accidents: such as oil spills, may be point source pollution, but can also extend to non-point source pollution (such as oily water surface).

[0057] Abnormal wastewater discharge pollution accidents: generally concentrated discharge at a certain location, usually point source pollution or non-point source pollution.

[0058] The longitude and latitude of the pollution release point help determine the spatial location of the pollution source. Through longitude and latitude, the regional scope of the pollution source and its relationship with the surrounding environment can be determined.

[0059] If the release points are relatively concentrated, it can be inferred that it is point source pollution.

[0060] If the release points are distributed within a certain area and involve pollution over a larger area, it may be line source pollution or surface source pollution.

[0061] The occurrence and end time of a pollution incident will affect the duration of pollutant release and its diffusion process. Factors such as the duration and continuity of the pollution also affect the determination of the pollution type:

[0062] Accidents involving large releases in a relatively short period of time, such as explosions, may be point source pollution.

[0063] Accidents that last for a long time (such as several hours or days), such as oil spills, may be non-point source pollution because the pollution may spread more widely.

[0064] The amount of pollution released during a pollution incident usually reflects the scale of the pollution. Large-scale releases usually lead to widespread pollution and affect the surrounding environment.

[0065] Small-scale, high-concentration pollution releases usually indicate a more concentrated pollution source (point source pollution).

[0066] Large-scale, low-concentration pollution releases may cause regional pollution and are suitable to be considered as non-point source pollution.

[0067] Point source pollution: If the pollution accident type involves concentrated point release such as toxic gas leakage, explosion accident, oil pollution accident, etc.

[0068] If the pollution source is located at a concentrated point (such as a storage tank or pipeline leakage point).

[0069] If the release amount is relatively small and the leakage time is short.

[0070] Line source pollution: If the accident involves the release of pollutants along the line, such as pollutants leaking through a transportation pipeline or pollutants spreading along railways or roads.

[0071] The area where the accident occurred is distributed along a certain line, and the pollution range may be limited, which belongs to the line source type.

[0072] Non-point source pollution: If pollutants are widely distributed over a large area, such as pesticide spraying or pollutants leaking over a large area.

[0073] If the release amount is large and the area where the accident occurs is relatively wide, it is usually considered non-point source pollution.

[0074] Then, the preset adaptive WRF simulation grid generation method is read, and the adaptive WRF simulation grid is generated according to the preset adaptive WRF simulation grid generation method and the latitude and longitude information.

[0075] Based on the above technical solutions, optional, preset adaptive WRF simulation grid generation methods include:

[0076] Obtain the duration of the accident and the current wind speed, and calculate the diffusion radius of the pollution accident based on the pollution accident release amount, the duration of the accident, the current wind speed, and a preset diffusion radius calculation formula;

[0077] Determine the pollutant area based on the pollution accident diffusion radius and the longitude and latitude information of the pollution release point, determine the central area of the pollutant area based on a preset minimum radius threshold, and determine whether there is an area outside the maximum radius in the pollutant area based on a preset maximum radius threshold and the pollution accident diffusion radius;

[0078] If there is an area outside the maximum radius, a preset highest resolution grid is used in the central area, and a preset lowest resolution grid is used in the area outside the maximum radius;

[0079] Setting other areas within the pollutant area as dynamic transition areas, calculating the dynamic resolution of each grid in the dynamic transition area according to a preset dynamic resolution calculation formula, and setting dynamic resolution grids in the dynamic transition area according to the dynamic resolution;

[0080] Determine an adaptive WRF simulation grid according to the maximum resolution grid, the minimum resolution grid and the dynamic resolution grid;

[0081] Accordingly, after determining whether there is an area outside the maximum radius according to the preset maximum radius threshold and the pollution accident diffusion radius, the method further includes:

[0082] If there is no area outside the maximum radius, the preset highest resolution grid is used in the central area;

[0083] Setting other areas within the pollutant area as dynamic transition areas, calculating the dynamic resolution of each grid in the dynamic transition area according to a preset dynamic resolution calculation formula, and setting dynamic resolution grids in the dynamic transition area according to the dynamic resolution;

[0084] An adaptive WRF simulation grid is determined according to the minimum resolution grid and the dynamic resolution grid.

[0085] In this scenario, the accident duration can be the length of time the pollution incident has occurred, typically from the time the pollution source releases pollutants to the current system time. This duration affects the extent of pollutant diffusion, as the time it takes for pollutants to diffuse in the air is directly related to the duration of the accident.

[0086] The current wind speed can be the wind speed at the specific time when the pollution incident occurs. Wind speed is a key meteorological factor affecting the dispersion of pollutants. Higher wind speeds generally increase the speed at which pollutants disperse, and the range of their spread also increases.

[0087] The diffusion radius of a pollution accident can be the maximum impact range of pollutants calculated based on the release amount of the pollution accident, the duration of the accident, and the current wind speed. This radius determines the diffusion area of the pollutants.

[0088] The pollutant area may refer to the affected area after the spread of a pollution accident, and is usually determined based on the calculated diffusion radius.

[0089] The preset minimum radius threshold refers to a fixed, pre-set radius used to determine the center of the pollutant dispersion area during a pollutant dispersion simulation. In practice, it is used to delineate areas with higher pollutant concentrations and determine the use of a higher simulation resolution within these areas to ensure highly accurate pollution dispersion predictions.

[0090] The central area can be the area where the concentration of pollutants is high and the simulation accuracy requirements are high during the diffusion process.

[0091] The preset maximum radius threshold may be a maximum diffusion range set in a pollutant diffusion simulation, and areas beyond this radius are considered to be areas with less impact from pollutants.

[0092] The areas outside the maximum radius can be areas beyond the maximum radius threshold. The pollutant concentration in these areas is low, and the diffusion process can be simulated with a lower resolution grid.

[0093] The preset minimum resolution grid may be a preset grid resolution (ie, simulation accuracy) for low concentration areas in pollutant diffusion simulation.

[0094] The preset minimum resolution grid may be the highest resolution grid used for the central area (area with higher concentration) during the simulation of pollutant diffusion.

[0095] The dynamic transition zone can be the area between the central area and the area outside the maximum radius, where the pollution concentration gradually weakens but has not yet reached the low concentration level.

[0096] The preset dynamic resolution calculation formula may be a formula for calculating the dynamic resolution of different grids in the dynamic transition region, and the formula determines the resolution of each grid in the region based on factors such as the concentration change and diffusion rate of the pollutant.

[0097] Dynamic resolution refers to the process of dynamically adjusting the grid resolution based on the actual needs of a region during pollutant dispersion simulations. This means that different regions use different resolutions to ensure a balance between simulation accuracy and computational efficiency.

[0098] The dynamic resolution grid can be a grid resolution set based on the dynamic resolution formula within the dynamic transition region. It uses different resolutions for simulation based on factors such as pollutant concentration and diffusion rate.

[0099] The time span from the start of the pollution incident (i.e., the moment pollutants began to be released) to the current moment can be used. The duration of the incident can be calculated by recording the difference between the start time of the pollution incident and the current time. The current wind speed is determined using real-time wind speed data provided by the weather station. The amount of pollution released, the duration of the incident, and the current wind speed are then substituted into the preset diffusion radius calculation formula to calculate the diffusion radius of the pollution incident.

[0100] The longitude and latitude of the pollution release point can be obtained. The pollutant area can be determined, with the release point as the center and the diffusion radius as the radius. Concentrations within this area may be high and require detailed monitoring. The pollution accident diffusion radius and minimum radius threshold are obtained. The pollutant diffusion area is divided by radius, with the area within the minimum radius threshold being designated the central area. The pollution accident diffusion radius and maximum radius threshold are then obtained. If the diffusion radius exceeds the maximum radius threshold, the area beyond the maximum radius is designated the area outside the maximum radius. The central area is simulated using the highest-resolution mesh to ensure sufficient computational accuracy in areas with higher concentrations. Areas outside the maximum radius are simulated using the lowest-resolution mesh because lower pollutant concentrations can tolerate lower computational accuracy. The extent of the dynamic transition region is determined based on the actual pollutant diffusion situation. The resolution of the dynamic transition region should be adjusted based on the rate of change of pollutant concentration in that area. The dynamic resolution of each mesh is then calculated using a preset dynamic resolution calculation formula. Within the dynamic transition region, an appropriate mesh resolution is set based on the calculated results. Based on the highest-resolution, lowest-resolution, and dynamic-resolution meshes obtained in the above steps and the regional distribution of pollutant diffusion, an adaptive WRF simulation grid is set. The highest resolution mesh is used in the center area, the lowest resolution mesh is used outside the maximum radius, and a dynamic resolution mesh is used in the dynamic transition area.

[0101] If the diffusion radius is less than the maximum radius threshold, there is no area outside the maximum radius. The central area is simulated using the highest resolution grid to ensure sufficient calculation accuracy in areas with higher concentrations. Determine the scope of the dynamic transition area based on the actual situation of pollutant diffusion. The resolution of the dynamic transition area should be adjusted according to the rate of change of pollutant concentration in the area. Then use the preset dynamic resolution calculation formula to calculate the dynamic resolution of each grid. In the dynamic transition area, set the appropriate grid resolution based on the calculation results. According to the highest resolution grid and dynamic resolution grid obtained in the above steps, combined with the regional distribution of pollutant diffusion, set the adaptive WRF simulation grid. Use the highest resolution grid in the central area, the lowest resolution grid in the area outside the maximum radius, and the dynamic resolution grid in the dynamic transition area.

[0102] In this solution, an adaptive grid setting method is used to adjust the grid resolution according to the actual situation of pollutant diffusion, which not only improves the accuracy and efficiency of the simulation, but also better copes with complex pollution diffusion scenarios and emergency response needs.

[0103] Based on the above technical solution, an optional, preset diffusion radius calculation formula is:

[0104]

[0105] Among them, R(t) is the diffusion radius of the pollution accident; Q is the release amount of the pollution accident; C is the preset pollution accident diffusion constant; U(t) is the current wind speed; t is the duration of the accident.

[0106] In this scenario, C typically refers to the diffusion constant of a pollution incident, which can also be considered a preset value used to describe the diffusion rate of pollutants in the environment. The specific value of this constant depends on many factors and is usually determined through field data analysis.

[0107] The determination of C usually depends on the following methods:

[0108] (1) Estimation based on historical data

[0109] Historical data: By reviewing historical pollution incident data and combining it with measured pollutant diffusion paths and concentrations, the diffusion constant of pollutants can be estimated. Data sources may include past pollution incidents and long-term data records from environmental monitoring stations. For example, by tracking the diffusion path of pollutants from a leak and combining it with parameters such as wind speed and atmospheric temperature, scientists or engineers can infer the diffusion constant C for that incident.

[0110] (2) Calculation based on meteorological models

[0111] Atmospheric Diffusion Models: Pollutant diffusion is typically estimated using atmospheric diffusion models (such as the Gaussian diffusion model or numerical weather prediction models) to estimate the diffusion constant, C. These models consider meteorological factors such as wind speed, atmospheric stability, temperature, and humidity to derive the diffusion constant of the pollutant. For example, models such as the Numerical Weather Forecast (WRF) model and FLEXPART can calculate the diffusion of pollutants within a specific area using meteorological data and parameters of the pollutant release point.

[0112] (3) Laboratory or field experiments

[0113] Field experiments: By conducting pollutant release experiments in a specific area and tracking changes in pollutant concentrations, researchers can directly measure the rate of pollutant diffusion and thus estimate the diffusion constant, C. These experiments often simulate different environmental conditions and calculate specific values for pollutant diffusion. In the laboratory, by controlling the amount of pollutant released and environmental variables (such as wind speed and temperature), the pollutant diffusion process can be simulated, thereby obtaining a baseline value.

[0114] (4) Meteorological factor adjustment method

[0115] Based on real-time meteorological conditions (such as wind speed, temperature, atmospheric pressure, etc.), C may be dynamically determined by adjusting factors. For example, in an environment with high wind speed, the diffusion constant C will increase, while in an environment with low wind speed, the diffusion constant will decrease.

[0116] Typically, the pollution accident diffusion constant C is set at the initial stage and depends on the following factors:

[0117] Type and characteristics of pollutants: Different types of pollutants (such as gas, dust, liquid, etc.) have different diffusion characteristics, so C may vary depending on the type of pollutant.

[0118] Environmental conditions: Meteorological factors such as wind speed, atmospheric stability, and temperature can affect the diffusion constant, but in simple models, C is an average or baseline value based on long-term statistics.

[0119] The pollution incident diffusion constant, C, is typically a fixed value determined based on past incident data, experiments, or meteorological models to describe the diffusion characteristics of pollutants. In some sophisticated simulations, the pollutant diffusion constant, C, may vary with environmental conditions (such as real-time wind speed and temperature), thereby dynamically adjusting the pollutant's diffusion rate.

[0120] Based on the above technical solution, an optional, preset dynamic resolution calculation formula is:

[0121]

[0122] Where Δx(r) is the dynamic resolution; Δx min is the preset minimum resolution; Δx max is the preset maximum resolution; r is the distance from each grid point to the pollution release point; R(t) is the diffusion radius of the pollution accident.

[0123] In this solution, the simulation results can be verified by experiments and historical data, and the grid resolution can be adjusted to ensure the accuracy of the simulation results. For example, by adjusting Δx min and Δx max To match the simulation results with the actual observation results, and thus determine the most appropriate grid resolution range.

[0124] In a gridded simulation, the distance r can also be defined based on the grid divisions. The center point of each grid cell can be used as a reference point to calculate the distance from that grid point to the contamination release point. For example, assuming the grid sizes are Δx and Δy, r can be calculated by calculating the straight-line distance from the center of the corresponding grid cell to the contamination release point.

[0125] On the basis of the above technical solution, optionally, the preset first processing method further includes:

[0126] Update the accident occurrence time and current wind speed according to a preset update step size, and update the pollution accident diffusion radius according to the pollution accident release amount, accident occurrence time, current wind speed and a preset diffusion radius calculation formula;

[0127] Update the pollutant area according to the pollution accident diffusion radius and the longitude and latitude information of the pollution release point, determine the center area of the pollutant area according to a preset minimum radius threshold, and determine whether there is an area outside the maximum radius in the pollutant area according to a preset maximum radius threshold and the pollution accident diffusion radius;

[0128] If there is an area outside the maximum radius, a preset highest resolution grid is used in the central area, and a preset lowest resolution grid is used in the area outside the maximum radius;

[0129] Setting other areas within the pollutant area as dynamic transition areas, calculating the dynamic resolution of each grid in the dynamic transition area according to a preset dynamic resolution calculation formula, and setting dynamic resolution grids in the dynamic transition area according to the dynamic resolution;

[0130] Determining an adaptive WRF simulation grid according to the preset highest resolution grid, the preset lowest resolution grid, and the dynamic resolution grid;

[0131] Accordingly, after determining whether there is an area outside the maximum radius according to the preset maximum radius threshold and the pollution accident diffusion radius, the method further includes:

[0132] If there is no area outside the maximum radius, the preset highest resolution grid is used in the central area;

[0133] Setting other areas within the pollutant area as dynamic transition areas, calculating the dynamic resolution of each grid in the dynamic transition area according to a preset dynamic resolution calculation formula, and setting dynamic resolution grids in the dynamic transition area according to the dynamic resolution;

[0134] An adaptive WRF simulation grid is determined according to the preset highest resolution grid and the dynamic resolution grid.

[0135] In this solution, based on the preset update step size, the system calculates the diffusion radius of the pollution accident by continuously updating the accident time and current wind speed, and determines the pollutant area in combination with the latitude and longitude information of the pollution release point. According to the preset minimum and maximum radius thresholds, the system delineates the central area and the area outside the maximum radius of the pollutant area, and sets different grid resolutions for each area. The highest resolution grid is used in the central area, the lowest resolution grid is used in the area outside the maximum radius, and other areas are set as dynamic transition areas, and their grid resolutions are adjusted according to the dynamic resolution calculation formula. Finally, based on these grids of different resolutions, an adaptive WRF simulation grid is determined to improve computational efficiency while ensuring simulation accuracy.

[0136] This solution reduces unnecessary computation by using different resolutions in the center of the polluted area and outside the maximum radius. Fine-grained computations in the central area and lower resolutions in areas farther from the pollution source avoid unnecessary high-precision computations and improve overall computational efficiency.

[0137] S102: determining a study area based on the latitude and longitude information of the pollution release point and the adaptive WRF simulation grid, and performing a three-dimensional meteorological field simulation on the study area using a downscaling meteorological diagnostic model to obtain refined three-dimensional meteorological field data.

[0138] The study area can refer to a certain spatial range selected during the pollutant dispersion prediction and meteorological simulation process. This range is determined based on the latitude and longitude information of the pollution release point. Specifically, the study area includes:

[0139] The area surrounding the pollution release point: The spread of pollutants will be affected by factors such as wind speed and weather, so a range needs to be defined to simulate the spread of pollutants.

[0140] Consider the impact of the spread and type of pollution: for example, for point source pollution, the study area may be relatively small, while for area or line source pollution, the study area may need to cover a wider geographic area.

[0141] Dynamic Update: Over time, the scope of the study area may also be dynamically adjusted as the spread of pollution changes.

[0142] The definition of the research area needs to be determined based on the following factors:

[0143] Location of pollution release point: Determine the center point based on the latitude and longitude information of the pollution source.

[0144] Diffusion radius: Based on factors such as wind speed and release amount, the range of pollutant diffusion is predicted and the boundaries of the study area are determined.

[0145] Meteorological conditions: Factors such as atmospheric stability and wind speed will affect the diffusion range of pollutants.

[0146] A downscaling meteorological diagnostic model can be a model that converts large-scale meteorological data (such as data from numerical weather prediction models or global climate models) into meteorological data at a smaller spatial scale. Its functions are:

[0147] Providing high-resolution meteorological field data: Since global climate models and atmospheric models usually have low resolutions (such as tens to hundreds of kilometers), these large-scale data can be converted into high-resolution meteorological data suitable for local or regional scales through downscaling methods, which are usually used for simulating local meteorology and pollutant diffusion.

[0148] Improved local meteorological simulations: For applications requiring detailed weather forecasts and pollutant dispersion analysis, downscaling models provide more detailed information by fine-tuning local meteorological conditions (e.g., wind speed, temperature, humidity, etc.).

[0149] These models are usually adjusted based on historical meteorological data or real-time weather forecast data, taking into account factors such as local topography, climate characteristics, and urban effects, thereby outputting more accurate meteorological data. There are several common methods for downscaling meteorological models:

[0150] Statistical downscaling method: Based on historical observation data and the output of large-scale models, large-scale data are converted into local data through statistical relationships.

[0151] Dynamic downscaling method: Downscaling is achieved through high-resolution meteorological simulation. Common methods include regional climate models (RCMs).

[0152] Refined three-dimensional meteorological field data can be meteorological data calculated and predicted by meteorological models (such as downscaling models or WRF models), covering various meteorological elements in the area. Specifically, it can include:

[0153] Wind speed and direction: These are key factors that affect the diffusion of pollutants. Refined three-dimensional meteorological field data will provide changes in wind speed and direction at each point in the entire region.

[0154] Temperature: Temperature is a key meteorological factor that affects air density, stability, and the rise or deposition of pollutants.

[0155] Humidity: Humidity affects air density, wind speed, and pollutant transport under meteorological conditions.

[0156] Air pressure: The air pressure field is closely related to the wind field, affecting the direction and speed of air flow.

[0157] Precipitation amount and type: Precipitation affects the wet deposition of pollutants and affects the diffusion and deposition process of pollutants.

[0158] Atmospheric stability: reflects the vertical flow ability of air, affecting the rise or diffusion of pollutants.

[0159] This data can be three-dimensional meteorological field data, representing the changes in meteorological elements (such as temperature, humidity, and wind speed) at different points in time and space. Refined three-dimensional meteorological field data, calculated using numerical weather prediction models or downscaling models, can support pollutant diffusion models.

[0160] The existing adaptive WRF simulation grid can be used, which has been assigned different resolutions according to the diffusion requirements of different areas. The grid resolution is smaller near the pollution source, and the grid resolution is larger in areas far away from the pollution source. For point source pollution, the study area is usually centered on the pollution source and extends to the maximum range where the pollutant may spread. The size of the area can be determined by estimating the diffusion radius of the pollutant. For line source and surface source pollution, the shape and size of the study area need to be determined according to the specific location and form of the pollution source (for example, the line source extends along a certain line, and the surface source may need to cover a wider area). Combined with the diffusion area of the pollution source, the adaptive grid system is used to determine the boundary of the study area. The grid division of this area should ensure that the maximum range of pollutant diffusion can be covered. Under the adaptive grid framework, the grid coverage of the study area is delineated according to the specific location of the pollution source and the range of the diffusion area. At this time, the grid resolution will be smaller in areas close to the pollution source and gradually increase in areas farther away from the pollution source. Then collect and input the data required for the study area. The input data includes:

[0161] Large-scale meteorological data (such as the output of global climate models or numerical weather prediction models) usually contain basic meteorological elements such as temperature, air pressure, humidity, wind speed, and precipitation.

[0162] Topography and land use information of the study area: for example, terrain undulation, vegetation type, etc., which will affect the local meteorology.

[0163] The latitude and longitude information of the pollution release point: used to determine the center and boundary of the simulation area.

[0164] Then, a downscaling meteorological diagnostic model is used to simulate the local meteorological field, and higher-resolution meteorological field data is generated according to the scope of the study area and the adaptive grid resolution. During the simulation process, the model calculates the spatial and temporal changes in meteorological elements such as wind speed, temperature, humidity, and precipitation. The refined three-dimensional meteorological field data includes meteorological elements at each grid point in the study area, which is usually three-dimensional data, involving the latitude, longitude and altitude of space as well as time. Specifically, it can include: Wind speed and wind direction: the main meteorological factors affecting the diffusion of pollutants. Temperature: affects the stability of the air and the rise and fall of pollutants. Humidity: affects air density, wet deposition, etc. Air pressure: closely related to factors such as wind speed and airflow. Precipitation amount and precipitation type: affects the wet deposition of pollutants.

[0165] The differences between the downscaling meteorological diagnostic model of this scheme and the existing downscaling meteorological diagnostic models are as follows:

[0166] 1. This proposal's downscaling meteorological diagnostic model is designed specifically for pollutant dispersion forecasting scenarios, focusing on the location of pollution sources, the type of pollution incident, the amount of pollution released, and its impact on local meteorological conditions. The model aims to accurately simulate the meteorological field in specific pollution incident areas to support subsequent pollutant dispersion forecasts.

[0167] Existing models: Existing downscaling meteorological models are typically used to provide general weather forecasts, focusing on predicting weather conditions (such as temperature, precipitation, and wind speed) over large areas. They are not specifically optimized for factors influencing pollutant dispersion. As a result, the resolution and meteorological factor processing of existing models may not be sufficiently refined to meet the needs of pollutant dispersion prediction.

[0168] 2. This solution uses an adaptive grid generation method, dynamically adjusting the grid resolution based on the location of the pollution release point and the pollutant's diffusion radius. A smaller grid resolution (e.g., 20 meters) is used near the pollution source, while a larger grid resolution (e.g., 100 meters) can be used in areas farther away. This adaptive grid approach more accurately simulates the diffusion of pollutants in a local area, particularly the meteorological changes surrounding the pollution source.

[0169] Existing models: Many existing downscaling meteorological models typically use a fixed grid resolution and are unable to adaptively adjust resolution based on local meteorological variations and pollution source characteristics. As a result, existing models may be inaccurate in simulating fine-grained local meteorological conditions and predicting pollution dispersion.

[0170] 3. The downscaling meteorological diagnostic model of this scheme can be dynamically adjusted according to real-time pollution accidents and meteorological data. When the pollution accident situation changes, the system can update the simulation results according to the pollution accident situation and respond to changes in sudden pollution accidents.

[0171] Existing models: Existing models are usually based on fixed meteorological data and parameters for forecasting, and are unable to respond quickly to sudden changes in pollution events, especially in terms of emergency response and pollution spread prediction.

[0172] Training process of downscaling meteorological diagnostic model:

[0173] Meteorological data for the region, including temperature, humidity, wind speed, and air pressure, is collected and combined with historical pollution event data to determine the correlation between historical meteorological data and pollutant dispersion results. Historical data on different pollution incident types (such as toxic spills, oil contamination, and radioactive contamination) is collected to obtain detailed information such as the location, type, release volume, and incident time of the pollution source. The downscaling meteorological model in this proposal is based on a physics-driven meteorological model (such as WRF). It uses meteorological data as input and generates a high-resolution output of the regional meteorological field using numerical methods. To improve model accuracy, the downscaling process can be further optimized by incorporating an ensemble Kalman filter (EnKF). Specifically, the downscaling process can be optimized by incorporating high-resolution meteorological data and pollutant dispersion data as observations to construct a meteorological model for each ensemble member, each representing a different forecast state. Each ensemble member generates a meteorological forecast based on a pre-defined physics-driven model (such as WRF). Combined with actual observations, the state estimate of each member is corrected using a Kalman gain, thereby reducing both model and observation errors. During this process, the ensemble Kalman filter dynamically adjusts model parameters based on patterns in historical data. Through repeated updates, the model output is optimized, ultimately achieving accurate predictions of pollutant dispersion behavior. This method, by handling nonlinearity, uncertainty, and high-dimensional data, further improves the accuracy of meteorological models under varying meteorological conditions and enhances the reliability of pollutant dispersion predictions. These methods can adjust the predictions of physical models by identifying patterns in historical data. The collected meteorological data is then cleaned and preprocessed, including filling missing values, removing outliers, and standardizing. Furthermore, consideration must be given to how to format pollutant-related data so that it can be combined with meteorological data for modeling. Key features influencing meteorological changes and pollution dispersion, such as meteorological factors (wind speed, temperature, humidity), topographical factors, pollutant release amount, and diffusion radius, are selected as model input features. Historical meteorological data and pollution incident data are then divided into training and validation sets. The training set is used to train the model, and the validation set is used to evaluate the model's generalization ability. During training, the model learns the complex relationship between meteorological data and pollutant dispersion, optimizing model parameters and enabling the model to predict pollutant dispersion behavior under future meteorological conditions. Then, standard error evaluation metrics (such as root mean square error, mean absolute error, etc.) are used to evaluate the prediction accuracy of the model on the validation set. Based on the evaluation results, the model is then optimized, which may include adjusting model hyperparameters, adding or reducing input features, optimizing the downscaling algorithm, etc., to improve the model's predictive ability.

[0174] S103, using the FLEXPART-WRF model to determine the simulated migration trajectory of pollutants based on the refined three-dimensional meteorological field data, the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution accident, and the amount of pollution accident released, as well as the simulated particle concentration and simulated dry and wet deposition of the adaptive WRF simulation grid.

[0175] The FLEXPART-WRF model can be a numerical simulation tool that combines FLEXPART (backward trajectory tracking model) and WRF (Weather Research and Forecasting Model) to simulate the diffusion and migration process of pollutants in the atmosphere. Specifically, FLEXPART is a backward trajectory model based on physical and numerical methods, which is widely used in air quality research, pollution source tracking, radiation transport and other fields. It predicts the migration and impact range of pollutants by simulating the diffusion, transport, sedimentation and other processes of air quality particles (such as pollutant molecules). WRF (Weather Research and Forecasting Model) is a numerical meteorological model used for weather forecasting, which can provide high-resolution simulation of meteorological variables such as atmospheric flow, temperature, humidity, and wind speed. It provides atmospheric data support for FLEXPART.

[0176] The FLEXPART-WRF model combines these two models, utilizing the high-resolution meteorological data provided by WRF to simulate pollutant migration. The FLEXPART-WRF model provides detailed simulations of pollutant diffusion, transport, particle concentration, wet and dry deposition in the atmosphere.

[0177] Simulating migration trajectories refers to simulating the paths and movements of pollutants (or their representative particles) in the atmosphere. Through backward trajectory tracking or forward particle tracking, it is possible to calculate how pollutants move and diffuse with airflow after release. In the FLEXPART-WRF model, pollutant migration trajectories are typically achieved by simulating the movement of particles in the air. Particles move under the influence of wind fields and atmospheric conditions, and over time, pollutants diffuse to different areas along airflow.

[0178] Simulated particle concentration can refer to the number or concentration of pollutant particles per unit volume within a given spatial area. By simulating the diffusion and transport of pollutants in the air, particle concentrations at different locations and times can be estimated. In the FLEXPART-WRF model, simulated particle concentrations are usually calculated based on the particle diffusion model under meteorological conditions and the transmission characteristics of the wind field. The concentration of particles in the air is affected by meteorological factors (such as wind speed, temperature, humidity, etc.), and the concentration of particles changes with changes in these meteorological factors.

[0179] Dry deposition and wet deposition are two mechanisms that describe the process by which pollutants from the atmosphere settle to the Earth's surface:

[0180] Dry deposition refers to the process by which pollutants, whether particulate or gaseous, settle directly onto the ground or a surface. This process is typically influenced by factors such as wind speed, particle size, and aerosol properties. During dry deposition, pollutants fall directly from the air to the ground without being washed by water or precipitation.

[0181] Wet deposition refers to the process by which pollutants are removed from the atmosphere by precipitation (such as rain, snow, and fog). Wet deposition can "clean" pollutants from the air through precipitation, carrying them from the atmosphere to the ground or water bodies. Wet deposition is influenced by factors such as precipitation intensity, precipitation amount, and the interaction between aerosols and precipitation.

[0182] The simulation process of the simulated migration trajectory of pollutants is:

[0183] Pollution source location: The FLEXPART-WRF model determines the location of the pollution source based on the latitude and longitude information of the pollution release point, which is used as the release source of the pollutant.

[0184] Initial Pollutant State: Based on the amount of pollution released by the accident and the type of pollution (point source, linear source, or non-point source), the model defines the initial concentration, distribution range, and intensity of the pollutants. For point source pollution, the pollutants diffuse evenly from the source to the surrounding area, with the distribution of pollutants being radial.

[0185] For line source pollution, pollutants expand along the line where the pollution source is located, and the distribution of pollutants expands linearly.

[0186] For non-point source pollution, pollutants spread along the non-point source area, and the distribution of pollutants diffuses in a planar manner.

[0187] Pollutant release time window: Based on the time of the pollution incident, the model starts the release of pollutants at that moment and stops the release of pollutants based on the end time of the pollution incident. The intensity of the pollution source (i.e., the release amount) and the change in the time step determine the continuous release amount of pollutants.

[0188] Simulating pollutant migration: The model uses refined three-dimensional meteorological data (such as wind speed, direction, temperature, and humidity) to simulate the diffusion of pollutants. These meteorological factors influence the speed, direction, and range of pollutant diffusion. Wind speed and direction play a dominant role in the migration of pollutants, influencing their direction and speed.

[0189] Temperature and humidity have indirect effects on the diffusion and deposition of pollutants, affecting the rise, diffusion and precipitation processes of pollutants.

[0190] The impact of time: The model simulates the release and diffusion of pollutants based on the time between the onset and end of a pollution incident. Following a pollution incident, the model adjusts the pollutant's migration trajectory based on meteorological data and the amount of pollution released at each time step. Within the time window between the onset and end of the pollution incident, the model simulates the diffusion of pollutants. As time passes, driven by meteorological conditions, the pollutant's diffusion range gradually increases and its concentration gradually decreases.

[0191] After the pollution accident ends, the model stops releasing new pollutants and continues to simulate the diffusion and deposition of remaining pollutants until the pollutants are completely deposited or diffused into the atmosphere.

[0192] Simulated particle concentration:

[0193] The model calculates pollutant concentrations at different points in time and space based on the pollutant's migration trajectory. Over time, pollutant concentrations gradually decrease until the release of pollutants ceases, and are affected by factors such as deposition and diffusion.

[0194] When simulating particle concentration, the model takes into account the amount of pollutant released. When the initial concentration of pollutants is high, the diffusion range of pollutants is relatively large and the concentration decays more slowly.

[0195] Simulating wet and dry deposition:

[0196] While pollutants are spreading, the FLEXPART-WRF model combines refined three-dimensional meteorological field data (such as precipitation, temperature, wind speed, etc.) to simulate the dry and wet deposition process.

[0197] The onset and end time of a pollution incident determine the period of pollutant deposition. The intensity of deposition is affected by the particle characteristics of the pollutants, precipitation, and meteorological conditions.

[0198] The differences between the FLEXPART-WRF model in this solution and the existing FLEXPART-WRF model are as follows:

[0199] The FLEXPART-WRF model of this solution: Compared with the existing FLEXPART-WRF model, the model of this solution integrates more types of input data, including the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the pollution type and the amount of pollution accident release. These data will serve as key parameters for the operation of the model.

[0200] The existing FLEXPART-WRF model typically relies on fixed meteorological datasets (such as standard meteorological field data output by WRF) to simulate particle trajectories, but lacks the spatiotemporal parameters of specific pollution incidents. Therefore, this solution uses dynamic monitoring of pollution incidents and simulation input to more accurately simulate the migration and diffusion of pollutants.

[0201] This solution: By fine-tuning the initial conditions for pollutant release based on pollution types (such as point source, line source, and non-point source pollution) and the amount of pollution accident release, the model can adjust the pollutant diffusion model according to different types of pollution sources (for example, radial diffusion of point sources, linear expansion of line sources, etc.).

[0202] Existing models usually focus on general pollution sources (such as the release or diffusion of pollutants in the atmosphere) and do not consider the dynamic changes of specific pollution accident types or release amounts.

[0203] The training process of the FLEXPART-WRF model is:

[0204] Collect and organize multi-dimensional input data such as the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude of the pollution release point, the type of pollution, and the amount of pollution accident release. Collect meteorological data, including wind speed, wind direction, temperature, humidity, air pressure, etc. These data will be downscaled through the WRF model to provide high-resolution meteorological field data. The meteorological data is then downscaled to ensure sufficient resolution within the study area so that the diffusion of pollutants can be accurately simulated. The pollution accident data is standardized to ensure that the input format of time, space and concentration distribution meets the requirements of the FLEXPART-WRF model. Based on the FLEXPART-WRF model, the behavioral characteristics of particle diffusion are adjusted according to the type of pollution (such as point source, line source, and surface source pollution). By refining the three-dimensional meteorological field data, the physical parameters such as wind field, temperature field, and humidity field in the model are updated to reflect the diffusion trend of pollutants. The intensity and time period of pollutant release are determined by combining the amount of pollution released and the time window of the pollution accident (from the time of occurrence to the time of end). Then, through historical pollution accident data, the migration trajectory, diffusion range, and deposition pattern of pollutants are trained, so that the model can dynamically adjust the diffusion path of pollutants according to meteorological conditions. During the training process, the changes in particle concentration of pollutants under different time steps and meteorological conditions are accurately simulated, and the final concentration of pollutants is adjusted by simulating precipitation (wet deposition) and aerosol deposition (dry deposition). Through iterative optimization, the model gradually adjusts and optimizes the particle distribution, concentration, and deposition characteristics to improve the accuracy and reliability of the model in actual accidents. Finally, the simulation results are compared with the actual pollutant diffusion data to evaluate the performance of the model under different meteorological conditions. Based on multiple simulation results, the parameters in the model, such as pollutant diffusion rate, deposition rate, etc., are adjusted to ensure that the model can give accurate pollution diffusion forecasts based on different types of pollution accidents.

[0205] S104: Determine a simulated meteorological forecast based on the refined three-dimensional meteorological field data, and determine a simulated pollutant diffusion forecast based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition.

[0206] A simulated weather forecast is a prediction of future weather conditions based on existing refined three-dimensional meteorological data (such as wind speed, temperature, humidity, and air pressure) combined with a meteorological model (such as WRF). The purpose of a simulated weather forecast is to provide information on future weather changes and background data for pollutant dispersion simulations.

[0207] The simulated pollutant diffusion forecast can be used to predict the diffusion range and concentration changes of pollutants in the future based on refined three-dimensional meteorological field data, pollutant migration trajectory, particle concentration, dry and wet deposition and other factors.

[0208] The time range for the simulated weather forecast can be determined (e.g., 1-hour, 24-hour, or 48-hour forecasts). Based on the meteorological data, an appropriate spatial resolution can be set to ensure sufficient meteorological data support at different locations in the study area. The refined three-dimensional meteorological field data obtained can then be used to analyze the changing trends of meteorological variables over the next few hours or days. Interpolation or other time series analysis methods can be used to predict meteorological variables (such as temperature, humidity, wind speed, and wind direction) at different time points. Numerical methods such as linear interpolation or more complex time series prediction algorithms can be used. During the simulated weather forecast process, localized corrections can be made to account for specific meteorological variations in the local area, such as the impact of complex terrain on wind speed and direction, and local climate effects. If the latest meteorological observation data is available, real-time corrections or adjustments can be made during the simulation to improve forecast accuracy. Finally, the simulated weather forecast results are output in a grid format, generating meteorological data (wind speed, wind direction, temperature, etc.) for a certain period in the future. This data can be used in subsequent pollutant dispersion simulations.

[0209] Based on existing simulated migration trajectories, the future diffusion paths of pollutants can be determined. Migration trajectories are often influenced by meteorological factors such as wind speed, wind direction, temperature, and humidity. Combining future weather forecast data (such as wind speed and direction), models can be used to simulate the future migration paths of pollutants and predict how they will spread to more distant areas. Trends in pollutant distribution over time and space are calculated, and potential areas of pollutant diffusion are mapped. Based on simulated particle concentration data, changes in pollutant concentrations at different points in time and locations are calculated. The distribution of pollutant concentrations often varies significantly with meteorological factors such as wind speed and temperature. Using simulated meteorological data and migration trajectories, particle concentrations can be predicted at future points in time. For example, pollutant concentrations may be higher in areas with lower wind speeds, forming concentration hotspots; whereas, pollutants may diffuse rapidly and have lower concentrations in areas with higher wind speeds. Based on precipitation data from weather forecasts, wet deposition of pollutants can be predicted. During rainfall, pollutant particles are carried to the ground by rain, reducing the concentration of pollutants in the air. In the absence of precipitation, pollutants slowly settle due to factors such as gravity. The model should consider the deposition rate of airborne particulate matter in stationary or slow-moving airflow and calculate the residence time of pollutants in the air. Combining the results of wet and dry deposition, the accumulation of pollutants on the ground and their impact on the environment are predicted. Then, based on migration trajectories, particle concentrations, and deposition data, the model can predict how pollutant concentrations vary over time and space. The diffusion paths and concentration distribution maps of pollutants are drawn to show the areas where pollutants will affect the environment over the coming period. This can include pollutant concentration contours, maps of pollutant migration paths, and concentration change maps under different deposition influences. Finally, a forecast of pollutant diffusion over the coming period is output, including pollutant migration paths, concentration distribution, and possible deposition areas. The output data is usually presented in the form of maps or spatial charts, which can show the most likely paths for pollutant diffusion, concentration hotspots, and affected areas.

[0210] In an embodiment of the present application, the time of occurrence of a pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution accident, and the amount of pollution released are obtained; the pollution type is determined based on the time of occurrence of the pollution accident, the time of end of the pollution accident, the type of pollution accident, and the amount of pollution released; and an adaptive WRF simulation grid is generated based on the longitude and latitude information of the pollution release point and a preset adaptive WRF simulation grid generation method; a study area is determined based on the longitude and latitude information of the pollution release point and the adaptive WRF simulation grid, and a downscaling meteorological diagnostic model is used to perform a three-dimensional meteorological field simulation on the study area to obtain refined three-dimensional meteorological field data; a FLEXPART-WRF model is used to determine the simulated migration trajectory of pollutants based on the refined three-dimensional meteorological field data, the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution, and the amount of pollution released, as well as the simulated particle concentration and simulated dry and wet deposition of the adaptive WRF simulation grid; a simulated meteorological forecast is determined based on the refined three-dimensional meteorological field data, and a simulated pollutant diffusion forecast is determined based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition. The FLEXPART-WRF-based adaptive pollution dispersion forecasting method, with its adaptive grid generation, precise meteorological field simulation, and real-time pollutant dispersion forecasting, provides precise pollutant dispersion information. Efficient automated simulation and sophisticated pollution source analysis provide quantitative and practical decision-making support tools for environmental management.

[0211] Based on the above technical solution, optionally, after determining the simulated meteorological forecast based on the refined three-dimensional meteorological field data, and determining the simulated pollutant diffusion forecast based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition, the method further includes:

[0212] Acquiring actual meteorological observation data, actual meteorological remote sensing data, and actual meteorological ground monitoring data, determining actual meteorological field data based on the actual meteorological observation data, actual meteorological remote sensing data, and actual meteorological ground monitoring data, and determining an actual meteorological forecast based on the actual meteorological field data;

[0213] If there is a difference between the actual weather forecast and the simulated weather forecast, the downscaled weather diagnostic model is adjusted according to the actual weather forecast.

[0214] In this scenario, actual meteorological observation data can come from real-time measurements from weather stations or satellites. This data includes various atmospheric meteorological variables, such as temperature, humidity, wind speed, wind direction, atmospheric pressure, and precipitation. Observational data can be obtained from fixed ground stations, aerial detection equipment (such as weather balloons and aircraft), and offshore platforms.

[0215] Actual meteorological remote sensing data can come from satellites, drones, or other remote sensing equipment. Remote sensing technology can provide large-scale, high-temporal and high-resolution meteorological information, such as cloud cover, surface temperature, atmospheric temperature, and humidity. Remote sensing data is often used for global-scale meteorological monitoring.

[0216] Actual ground-based meteorological data can be obtained from ground-based monitoring networks, including real-time data from weather stations, weather detectors, weather radar, and other equipment. Ground-based monitoring data is crucial for monitoring local meteorological changes and for the refinement and timeliness of weather forecasts. Examples include temperature, humidity, wind speed, and air pressure data provided by ground-based weather stations, as well as weather data collected by Automatic Weather Monitoring (AWS).

[0217] Actual meteorological data can be a combination of actual meteorological observations, remote sensing data, and ground-based monitoring data. It reflects meteorological conditions, such as temperature, humidity, wind speed, wind direction, and atmospheric pressure, distributed across a specific region or the globe, at various temporal and spatial scales. Actual meteorological data is obtained by integrating, calibrating, and simulating meteorological processes using all meteorological observations.

[0218] Actual weather forecasts can be short-term or long-term weather predictions generated based on actual meteorological observations and meteorological field data. Numerical weather prediction models (such as GFS and WRF) are used to predict weather changes over a period of time based on existing observation data.

[0219] Meteorological data can be collected from various sources, including weather stations, satellites, remote sensing equipment, and weather radar. Data sources can include real-time meteorological observation data provided by the National Meteorological Administration, satellite remote sensing data (such as MODIS and GOES satellite data), and monitoring data from ground-based weather stations. Using the actual meteorological data collected above, data assimilation techniques (such as interpolation and Kalman filtering) are used to integrate the actual meteorological observation data into a spatially and temporally continuous meteorological field data set. This process involves spatiotemporal interpolation, calibration, and synthesis of meteorological data from different sources. Specifically, spatial interpolation can be performed on various meteorological variables (such as temperature, humidity, and wind speed) to fill in incomplete observation data and generate a high-resolution meteorological field. Data assimilation techniques are used to fuse satellite remote sensing data with ground-based observation data to generate comprehensive meteorological field data. Based on the actual meteorological field data, numerical weather prediction models (such as WRF and GFS) are then used to forecast future meteorological conditions. Weather forecasts can range from a few hours to several days, covering meteorological elements such as temperature, precipitation, and wind speed. Specifically, actual meteorological data can be used as initial conditions to input into a numerical weather prediction model. The forecast model is then run to predict meteorological changes (such as temperature, humidity, and wind direction) for a certain period of time. If there are discrepancies between the actual forecast and the simulated forecast, the downscaled meteorological diagnostic model needs to be adjusted. Downscaling meteorological diagnostic models are often used to improve the spatial resolution and accuracy of weather forecasts, particularly in areas with significant local meteorological variability. Specifically, the differences between the actual forecast and the simulated forecast can be compared to identify model errors. Adjustments to the downscaling model parameters may include weightings of meteorological variables and spatiotemporal interpolation algorithms. The model is then validated and retrained using new observational data and actual meteorological data to optimize its predictive performance. By analyzing the forecast discrepancies, the model can be further adjusted through reverse analysis and iterative training. Based on the comparison between the actual and simulated forecasts, internal model parameters (such as weightings of meteorological variables and the model time step) are adjusted. Error correction and model calibration methods are used to update the model, thereby improving the model's ability to predict actual weather.

[0220] In this plan, by comparing actual meteorological data with simulated meteorological forecasts, real-time feedback and adjustments are carried out to optimize the downscaled meteorological diagnostic model, improve the accuracy and reliability of meteorological forecasts, enhance model adaptability, and improve decision support and emergency response capabilities.

[0221] Based on the above technical solution, optionally, after determining the simulated meteorological forecast based on the refined three-dimensional meteorological field data, and determining the simulated pollutant diffusion forecast based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition, the method further includes:

[0222] Obtaining actual pollutant ground monitoring data, actual pollutant remote sensing data, and actual pollutant ground deposition monitoring data, and determining pollutant concentration distribution, pollutant deposition information, and pollutant migration trajectory based on the actual pollutant ground monitoring data, actual pollutant remote sensing data, and actual pollutant ground deposition monitoring data;

[0223] Determining actual pollutant diffusion forecasts based on the pollutant concentration distribution, pollutant deposition information, and pollutant migration trajectories;

[0224] If there is a difference between the actual pollutant diffusion forecast and the simulated pollutant diffusion forecast, the FLEXPART-WRF model is adjusted according to the actual pollutant diffusion forecast.

[0225] In this solution, actual ground-based pollutant monitoring data can be pollutant concentration data measured by ground-based sensors (such as air quality monitoring stations and mobile monitoring equipment) at multiple geographical locations and time points. This monitoring data typically includes major air pollutants such as PM2.5, PM10, NO2, SO2, CO, and O3.

[0226] Actual pollutant remote sensing data can be obtained through remote sensing technologies such as satellites or drones. These data typically involve monitoring pollutants over large areas, such as the spatial distribution of ground-level pollutant concentrations, aerosols, and gas concentrations such as NO2 and CO2. These data can be obtained using sensors such as remote sensing satellites like MODIS and TROPOMI.

[0227] Actual pollutant ground deposition monitoring data can be information about pollutant deposition obtained through ground-based monitoring equipment, LiDAR (Light Detection and Ranging), or other remote sensing methods (such as Synthetic Aperture Radar (SAR)). It generally involves the deposition process of pollutants in the atmosphere, such as aerosol deposition and accumulation on the ground.

[0228] Pollutant concentration distribution can be described as the concentration of pollutants at different locations in the atmosphere at a specific moment or time period. Using ground-based monitoring data, remote sensing data, and deposition data, we can determine the spatial distribution of pollutants, including the concentration levels and changing trends in different regions.

[0229] Pollutant deposition information describes the process by which pollutants settle from the air to the ground, including the particle deposition velocity, deposition amount, and geographic distribution. The deposition process is affected by meteorological factors (such as precipitation, wind speed, and temperature) as well as pollutant characteristics (such as particle size and density).

[0230] A pollutant migration trajectory refers to the path that pollutants follow through the atmosphere, typically driven by meteorological conditions such as wind, temperature, and humidity. This trajectory involves the spread of pollutants from their release source to the affected area and can be linear, radial, or other complex patterns, depending on the type of pollution source and meteorological factors.

[0231] Actual pollutant dispersion forecasts predict the diffusion, migration, and deposition of pollutants in the atmosphere based on real-time monitoring and remote sensing data, generating forecasts of future pollutant dispersion trends. The goal of these forecasts is to provide decision-makers with real-time information on pollutant dispersion so they can take appropriate countermeasures. These forecasts include pollutant concentrations, migration trajectories, and deposition rates.

[0232] Pollutant concentration data can be collected through ground-based monitoring stations and sensors. Common monitoring stations include fixed air quality monitoring stations and mobile monitoring stations. This data covers real-time concentrations of pollutants such as PM2.5, PM10, CO, NO2, SO2, and O3. These data are typically collected regularly (e.g., every hour or every 10 minutes) and record pollutant concentrations at different geographic locations and time points. Pollutant data can also be obtained through satellite remote sensing or drone remote sensing technology. These data typically measure pollutant concentrations over large areas and are particularly suitable for spatial distribution and large-scale monitoring. Remote sensing data can include aerosol optical depth, NO2 concentration, CO2 concentration, and other data provided by satellites such as MODIS, TROPOMI, and Sentinel. Pollutant deposition data, obtained through LiDAR or other remote sensing methods (such as synthetic aperture radar (SAR),) reflects the process of pollutants settling from the atmosphere to the ground. Specifically, they include deposition rates, pollutant distribution on the ground, and concentrations after deposition. Spatial distribution maps of pollutant concentrations at different points in time and locations can then be constructed using both actual ground-based monitoring data and remote sensing data. This data can be spatially analyzed using GIS (Geographic Information System) software to generate heat maps or distribution maps of pollutant concentrations, helping to identify pollution hotspots. By monitoring the rate and area of pollutant deposition and combining it with deposition monitoring data, the process of pollutant deposition from the atmosphere to the ground over a specific time period can be analyzed to obtain deposition information. By monitoring the rate of pollutant deposition (such as deposition velocity) and the observed deposition area, the process of pollutant deposition from the atmosphere to the ground can be determined. Based on actual meteorological conditions (such as wind speed and direction) and observational data, diffusion models (such as FLEXPART-WRF) or other tracking algorithms are used to track the migration path of pollutants from the source to the affected area. These models can track the direction, speed, and range of pollutant diffusion at different temporal and spatial scales. Based on pollutant concentration distribution, deposition information, and migration trajectory, a pollutant diffusion forecast is calculated. This forecast is essentially a data-based estimate of the likely future diffusion trend of the pollutant. This process is often combined with meteorological forecasts to predict changes in pollutant concentrations in the air and help analyze the areas likely to be affected by the pollutants. If there is a discrepancy between the actual pollutant dispersion forecast (generated by monitoring and remote sensing data) and the simulated pollutant dispersion forecast (generated by the FLEXPART-WRF model), the FLEXPART-WRF model needs to be adjusted. The source of the error is then analyzed based on the discrepancy between the simulation results and the actual data. These errors may come from the following: Errors in meteorological data: Differences in meteorological data such as wind speed, wind direction, temperature, and humidity may lead to deviations in the pollutant diffusion path. Inaccurate pollution source characteristics: Inaccurate assumptions about parameters such as pollutant release amount, release time, and release location may affect the simulation of pollutant dispersion.Inappropriate diffusion model parameters: Parameters such as the diffusion coefficient, sedimentation rate, and sedimentation model may need adjustment. Local meteorological phenomena, such as the heat island effect and localized wind patterns, may not be fully accounted for in the simulation. Based on the results of the error analysis, adjust the FLEXPART-WRF model parameters. Adjustments may include:

[0233] a. Adjust meteorological input data

[0234] Introducing more accurate meteorological data: If the meteorological data in the model does not match the actual situation, you can consider introducing high-resolution meteorological data. For example, use high-precision observation data or data assimilation technology to combine actual meteorological observation data with the results of the numerical weather forecast model to improve the accuracy of meteorological input.

[0235] Adjust meteorological variables such as wind speed and wind direction: If the wind speed and wind direction in the simulation deviate significantly from the actual data, the weights of these parameters can be adjusted or a local wind field model can be introduced for correction.

[0236] b. Optimize diffusion parameters

[0237] Diffusion coefficient: The diffusion coefficient in the FLEXPART model controls how quickly pollutants diffuse through the atmosphere. Adjust the diffusion coefficient by comparing the simulated and actual pollutant diffusion ranges.

[0238] Deposition Rate: Adjust the dry and wet deposition rates of pollutants to better match the actual monitored deposition conditions.

[0239] Pollutant dissipation and removal: In the model, the dissipation rate and removal rate of pollutants may need to be adjusted according to actual conditions, especially when taking into account geographical characteristics and meteorological phenomena.

[0240] c. Adjust pollution source settings

[0241] Pollutant release: If the pollutant release assumptions in the model are inaccurate, the release can be adjusted based on actual monitoring data.

[0242] Pollution release time: Adjust the start and end time of pollutant release to make the simulation consistent with the time window of the actual pollution accident.

[0243] Pollution source location and type: Ensure that the location and type (point source, line source, area source) of the pollutant release point match the actual event. If there is a discrepancy, the pollution source configuration needs to be adjusted.

[0244] Model parameters can be optimized through inversion methods, including:

[0245] Minimize error: By comparing simulated data with actual monitoring data, use numerical optimization algorithms (such as least squares method, genetic algorithm, Monte Carlo method, etc.) to adjust the parameters in the model so that the simulation results are as close as possible to the actual observation data.

[0246] Inversion of meteorological data and pollution source characteristics: If there are multiple unknown factors, the optimal values of these factors can be inverted through optimization algorithms, thereby improving the accuracy of simulation results.

[0247] In this scheme, by adjusting the FLEXPART-WRF model according to actual pollutant ground monitoring data, remote sensing data and land subsidence monitoring data, the accuracy of pollutant dispersion forecast can be significantly improved, and the adaptability and reliability of the model can be enhanced.

[0248] Example 2

[0249] Figure 5 This is a flow chart of the adaptive pollution diffusion prediction method based on FLEXPART-WRF provided in Example 2 of this application, such as Figure 5 As shown, the specific method includes the following steps:

[0250] S501, obtain the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution accident and the amount of pollution released by the pollution accident, determine the pollution type based on the time of occurrence of the pollution accident, the time of end of the pollution accident, the type of pollution accident and the amount of pollution released by the pollution accident, and generate an adaptive WRF simulation grid based on the longitude and latitude information of the pollution release point and a preset adaptive WRF simulation grid generation method.

[0251] S502: Determine a study area based on the latitude and longitude information of the pollution release point and the adaptive WRF simulation grid, and use a downscaling meteorological diagnostic model to perform a three-dimensional meteorological field simulation on the study area to obtain refined three-dimensional meteorological field data.

[0252] S503: Using the FLEXPART-WRF model, the simulated migration trajectory of pollutants is determined based on the refined three-dimensional meteorological field data, the time when the pollution accident occurred, the time when the pollution accident ended, the longitude and latitude information of the pollution release point, the pollution type, and the amount of pollution accident released, as well as the simulated particle concentration and simulated dry and wet deposition of the adaptive WRF simulation grid.

[0253] S504: Determine a simulated meteorological forecast based on the refined three-dimensional meteorological field data, and determine a simulated pollutant diffusion forecast based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition.

[0254] S505: Determine a pollution warning method according to the pollution type, and transmit the pollution warning method to a control center for the control center to issue a warning according to the pollution warning method.

[0255] In this embodiment, the pollution warning method may refer to the means and channels used to transmit pollution accident alarms or reminders. Different warning methods may be selected according to different pollution types, accident severity, and pollution spread range.

[0256] A control center can refer to a specialized agency or organization that is responsible for coordinating, monitoring, handling, and responding to pollution incidents.

[0257] Based on the type of pollution (point source, line source, or surface source), the system will assess the scope of impact and propagation of the pollution incident, and thus select the appropriate pollution warning method. For example, point source pollution may be concentrated in a local area, so warnings can be issued to nearby residents or staff via text messages, phone calls, or local broadcasts; line source pollution (such as traffic pollution) may affect a linear area, and it is suitable to issue warning information through email, social media platforms, or dynamically updated maps; surface source pollution may involve a large area, and it is suitable to be widely disseminated through public broadcasts, government emergency platforms, or regional alarm systems. After the system selects the appropriate warning method based on the pollution type, it transmits the warning information to the control center.

[0258] In this plan, appropriate warning methods are adopted according to different pollution types (point source, line source, and surface source), which helps to ensure that warning information is conveyed in a timely and effective manner.

[0259] Example 3

[0260] Figure 6 This is a structural diagram of an adaptive pollution diffusion prediction system based on FLEXPART-WRF provided in Example 3 of the present application. Figure 6 As shown, the system is used to implement the adaptive pollution diffusion prediction method based on FLEXPART-WRF provided in Examples 1 and 2. The system specifically includes the following:

[0261] The data acquisition module 601 is used to obtain the pollution accident occurrence time, pollution accident end time, pollution release point latitude and longitude information, pollution accident type, and pollution accident release amount, determine the pollution type based on the pollution accident occurrence time, pollution accident end time, pollution accident type, and pollution accident release amount, and generate an adaptive WRF simulation grid based on the pollution release point latitude and longitude information and a preset adaptive WRF simulation grid generation method;

[0262] A refined three-dimensional meteorological field data generation module 602 is configured to determine a study area based on the latitude and longitude information of the pollution release point and the adaptive WRF simulation grid, and perform a three-dimensional meteorological field simulation on the study area using a downscaling meteorological diagnostic model to obtain refined three-dimensional meteorological field data;

[0263] The simulation module 603 is used to determine the simulated migration trajectory of pollutants based on the refined three-dimensional meteorological field data, the time when the pollution accident occurred, the time when the pollution accident ended, the longitude and latitude information of the pollution release point, the pollution type, and the amount of pollution released by the pollution accident using the FLEXPART-WRF model, and to determine the simulated particle concentration and simulated dry and wet deposition of the adaptive WRF simulation grid;

[0264] The forecast generation module 604 is used to determine a simulated meteorological forecast based on the refined three-dimensional meteorological field data, and to determine a simulated pollutant diffusion forecast based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition.

[0265] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0266] Example 4

[0267] like Figure 7 As shown, an embodiment of the present application also provides an electronic device 700, including a processor 701, a memory 702, and a program or instruction stored in the memory 702 and executable on the processor 701. When the program or instruction is executed by the processor 701, the various processes of the above-mentioned disk redirection device embodiment are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0268] It should be noted that the electronic devices in the embodiments of the present application include the mobile electronic devices and non-mobile electronic devices mentioned above.

[0269] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. An adaptive pollution diffusion prediction method based on FLEXPART-WRF, characterized by: The method comprises: Obtaining the occurrence time of the pollution accident, the end time of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution accident, and the amount of pollution released by the pollution accident; determining the pollution type according to the occurrence time of the pollution accident, the end time of the pollution accident, the type of pollution accident, and the amount of pollution released by the pollution accident; and generating an adaptive WRF simulation grid according to the longitude and latitude information of the pollution release point and a preset adaptive WRF simulation grid generation method; Determine the study area based on the latitude and longitude information of the pollution release point and the adaptive WRF simulation grid, and use the downscaling meteorological diagnostic model to perform a three-dimensional meteorological field simulation on the study area to obtain refined three-dimensional meteorological field data; Using the FLEXPART-WRF model, the simulated migration trajectory of pollutants is determined based on the refined three-dimensional meteorological field data, the time when the pollution accident occurred, the time when the pollution accident ended, the longitude and latitude information of the pollution release point, the pollution type, and the amount of pollution accident released, as well as the simulated particle concentration and simulated dry and wet deposition of the adaptive WRF simulation grid. A simulated meteorological forecast is determined based on the refined three-dimensional meteorological field data, and a simulated pollutant diffusion forecast is determined based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition.

2. The method according to claim 1, characterized in that in, After determining a simulated meteorological forecast based on the refined three-dimensional meteorological field data, and determining a simulated pollutant diffusion forecast based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition, the method further includes: Acquiring actual meteorological observation data, actual meteorological remote sensing data, and actual meteorological ground monitoring data, determining actual meteorological field data based on the actual meteorological observation data, actual meteorological remote sensing data, and actual meteorological ground monitoring data, and determining an actual meteorological forecast based on the actual meteorological field data; If there is a difference between the actual weather forecast and the simulated weather forecast, the downscaled weather diagnostic model is adjusted according to the actual weather forecast.

3. The method according to claim 1, characterized in that in, After determining a simulated meteorological forecast based on the refined three-dimensional meteorological field data, and determining a simulated pollutant diffusion forecast based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition, the method further includes: Obtaining actual pollutant ground monitoring data, actual pollutant remote sensing data, and actual pollutant ground deposition monitoring data, and determining pollutant concentration distribution, pollutant deposition information, and pollutant migration trajectory based on the actual pollutant ground monitoring data, actual pollutant remote sensing data, and actual pollutant ground deposition monitoring data; Determining actual pollutant diffusion forecasts based on the pollutant concentration distribution, pollutant deposition information, and pollutant migration trajectories; If there is a difference between the actual pollutant diffusion forecast and the simulated pollutant diffusion forecast, the FLEXPART-WRF model is adjusted according to the actual pollutant diffusion forecast.

4. The method according to claim 1, wherein in, Preset adaptive WRF simulation grid generation methods include: Obtain the duration of the accident and the current wind speed, and calculate the diffusion radius of the pollution accident based on the pollution accident release amount, the duration of the accident, the current wind speed, and a preset diffusion radius calculation formula; Determine the pollutant area based on the pollution accident diffusion radius and the longitude and latitude information of the pollution release point, determine the central area of the pollutant area based on a preset minimum radius threshold, and determine whether there is an area outside the maximum radius in the pollutant area based on a preset maximum radius threshold and the pollution accident diffusion radius; If there is an area outside the maximum radius, a preset highest resolution grid is used in the central area, and a preset lowest resolution grid is used in the area outside the maximum radius; Setting other areas within the pollutant area as dynamic transition areas, calculating the dynamic resolution of each grid in the dynamic transition area according to a preset dynamic resolution calculation formula, and setting dynamic resolution grids in the dynamic transition area according to the dynamic resolution; Determining an adaptive WRF simulation grid according to the preset highest resolution grid, the preset lowest resolution grid, and the dynamic resolution grid; Accordingly, after determining whether there is an area outside the maximum radius according to the preset maximum radius threshold and the pollution accident diffusion radius, the method further includes: If there is no area outside the maximum radius, the preset highest resolution grid is used in the central area; Setting other areas within the pollutant area as dynamic transition areas, calculating the dynamic resolution of each grid in the dynamic transition area according to a preset dynamic resolution calculation formula, and setting dynamic resolution grids in the dynamic transition area according to the dynamic resolution; An adaptive WRF simulation grid is determined according to the preset highest resolution grid and the dynamic resolution grid.

5. The method according to claim 4, characterized in that in, The default diffusion radius calculation formula is: Among them, R(t) is the diffusion radius of the pollution accident; Q is the release amount of the pollution accident; C is the preset pollution accident diffusion constant; U(t) is the current wind speed; t is the duration of the accident.

6. The method according to claim 4, characterized in that in, The preset dynamic resolution calculation formula is: Where Δx(r) is the dynamic resolution; Δx min is the preset minimum resolution; Δx max is the preset maximum resolution; r is the distance from each grid point to the pollution release point; R(t) is the diffusion radius of the pollution accident.

7. The method according to claim 4, characterized in that in, The preset first processing method also includes: Update the accident occurrence time and current wind speed according to a preset update step size, and update the pollution accident diffusion radius according to the pollution accident release amount, accident occurrence time, current wind speed and a preset diffusion radius calculation formula; Update the pollutant area according to the pollution accident diffusion radius and the longitude and latitude information of the pollution release point, determine the center area of the pollutant area according to a preset minimum radius threshold, and determine whether there is an area outside the maximum radius in the pollutant area according to a preset maximum radius threshold and the pollution accident diffusion radius; If there is an area outside the maximum radius, a preset highest resolution grid is used in the central area, and a preset lowest resolution grid is used in the area outside the maximum radius; Setting other areas within the pollutant area as dynamic transition areas, calculating the dynamic resolution of each grid in the dynamic transition area according to a preset dynamic resolution calculation formula, and setting dynamic resolution grids in the dynamic transition area according to the dynamic resolution; Determining an adaptive WRF simulation grid according to the preset highest resolution grid, the preset lowest resolution grid, and the dynamic resolution grid; Accordingly, after determining whether there is an area outside the maximum radius according to the preset maximum radius threshold and the pollution accident diffusion radius, the method further includes: If there is no area outside the maximum radius, the preset highest resolution grid is used in the central area; Setting other areas within the pollutant area as dynamic transition areas, calculating the dynamic resolution of each grid in the dynamic transition area according to a preset dynamic resolution calculation formula, and setting dynamic resolution grids in the dynamic transition area according to the dynamic resolution; An adaptive WRF simulation grid is determined according to the preset highest resolution grid and the dynamic resolution grid.

8. The method according to claim 1, characterized in that in, After determining a pollution warning method according to the pollution type and transmitting the pollution warning method to a control center, the method further includes: A pollution warning method is determined according to the pollution type, and the pollution warning method is transmitted to a control center for the control center to issue a warning according to the pollution warning method.

9. An adaptive pollution diffusion prediction system based on FLEXPART-WRF, used to execute the method according to any one of claims 1 to 8, characterized in that: The system comprises: A data acquisition module is used to obtain the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the type of pollution accident, and the amount of pollution released by the pollution accident; determine the pollution type based on the time of occurrence of the pollution accident, the time of end of the pollution accident, the type of pollution accident, and the amount of pollution released by the pollution accident; and generate an adaptive WRF simulation grid based on the longitude and latitude information of the pollution release point and a preset adaptive WRF simulation grid generation method; A refined three-dimensional meteorological field data generation module is used to determine the study area based on the latitude and longitude information of the pollution release point and the adaptive WRF simulation grid, and to perform a three-dimensional meteorological field simulation on the study area using a downscaling meteorological diagnostic model to obtain refined three-dimensional meteorological field data; A simulation module is used to determine the simulated migration trajectory of pollutants based on the refined three-dimensional meteorological field data, the time of occurrence of the pollution accident, the time of end of the pollution accident, the longitude and latitude information of the pollution release point, the pollution type, and the amount of pollution accident release using the FLEXPART-WRF model, as well as to determine the simulated particle concentration of the adaptive WRF simulation grid and simulated dry and wet deposition; A forecast generation module is used to determine a simulated meteorological forecast based on the refined three-dimensional meteorological field data, and to determine a simulated pollutant diffusion forecast based on the simulated migration trajectory, simulated particle concentration, and simulated dry and wet deposition.

10. An electronic device, characterized in that: The method comprises a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the adaptive pollution diffusion prediction method based on FLEXPART-WRF as described in any one of claims 1 to 8.

Images