Fine particulate matter distribution prediction system based on WRF-CHEM

Abstract

The invention discloses a fine particulate matter distribution prediction system based on a WRF-CHEM, and belongs to the technical field of environment monitoring and air quality prediction, and the system comprises a data input unit which is configured to obtain meteorological data and pollution source emission list data; at least one processor; and a memory storing an atmospheric chemistry transfer model program executable by the at least one processor. The WRF-CHEM-based fine particulate matter distribution prediction system can improve the accurate prediction capability of the concentration of fine particulate matters in a target area by integrating meteorological data, pollution source emission list data and a heterogeneous chemical reaction mechanism. The dynamic adjustment effect of humidity on the heterogeneous chemical reaction rate can effectively simulate the generation process of particulate matters under different humidity conditions, so that the prediction accuracy and reliability are improved. Through dynamic adjustment of the humidity dependence function, the system can simulate and adjust the influence of the humidity on the heterogeneous chemical reaction rate, which provides a basis for accurately predicting the influence of the humidity change on the atmospheric chemical reaction.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring and air quality prediction technology, and more specifically, to a fine particulate matter distribution prediction system based on WRF-CHEM. Background Technology

[0002] Fine particulate matter (PM2.5) refers to particulate matter with a diameter of 2.5 micrometers or less, primarily originating from pollution sources such as transportation, industry, and agriculture. Because it can penetrate deep into the lungs and enter the bloodstream, PM2.5 poses a significant threat to health; long-term exposure can lead to respiratory and cardiovascular diseases. As the impact of PM2.5 concentrations on human health becomes increasingly severe, PM2.5 concentration prediction and pollution source apportionment have become key focuses of environmental monitoring and air quality management.

[0003] Traditional forecasting methods rely on empirical models and statistical analysis, but their accuracy is poor under complex meteorological conditions and the influence of multi-source pollution. With the development of atmospheric science and computing technology, numerical model-based fine particulate matter (FPM) forecasting methods have become mainstream, especially those combined with atmospheric chemical transport models, which can more accurately simulate the distribution and changes of FPM in the atmosphere.

[0004] The WRF-CHEM model is a widely used numerical simulation tool that combines meteorological and atmospheric chemical processes to simulate aerosol generation, transport, deposition, and chemical reactions. By integrating meteorological data with chemical reaction mechanisms, the WRF-CHEM model can more accurately predict fine particulate matter concentrations and provide valuable forecasts in dynamically changing environments.

[0005] Humidity plays a crucial role in the physical and chemical properties of aerosols. It affects aerosol particle size, reaction rates, and diffusion processes, especially under high humidity conditions where aerosols may undergo hydration, leading to changes in particulate matter reaction rates. Accurately simulating the impact of humidity on chemical reaction rates is key to improving the accuracy of fine particulate matter prediction.

[0006] Pollution source apportionment is a crucial aspect of air quality management. Traditional methods rely on monitoring data, while source-receptor model-based pollution source apportionment methods can quantitatively analyze the contribution of pollution sources to fine particulate matter concentrations using atmospheric chemical transport models. This approach can clearly define the spatial and temporal distribution of pollution sources and their impact on air quality.

[0007] The aerosol microphysics module simulates processes such as aerosol nucleation, condensation, and collision-coalescence, thereby optimizing atmospheric chemical transport models and improving their ability to describe the aerosol lifecycle. The size, quantity, and chemical composition of aerosols are all influenced by these microphysical processes; therefore, accurate simulation of these processes is crucial for improving prediction accuracy.

[0008] Against this backdrop, the WRF-CHEM model, combining humidity-dependent functions, aerosol microphysics modules, and source-acceptor models, can comprehensively analyze and predict the distribution of fine particulate matter (PM2.5). This model can provide pollution source apportionment and dynamic humidity regulation, helping to improve the prediction accuracy of PM2.5 concentrations, thus providing strong support for air pollution control and policy formulation. Summary of the Invention

[0009] 1. Technical problems to be solved

[0010] To address the problems existing in the prior art, the purpose of this invention is to provide a fine particulate matter distribution prediction system based on WRF-CHEM, which can provide pollution source analysis and dynamic humidity adjustment, helping to improve the prediction accuracy of fine particulate matter concentration, thereby providing strong support for air pollution control and policy formulation.

[0011] 2. Technical Solution

[0012] To solve the above problems, the present invention adopts the following technical solution.

[0013] The WRF-CHEM-based fine particulate matter distribution prediction system includes a data input unit configured to acquire meteorological data and pollution source emission inventory data.

[0014] At least one processor; and

[0015] The memory stores an atmospheric chemical transport model program that can be executed by the at least one processor;

[0016] The atmospheric chemical transport model program is executed as follows:

[0017] The heterogeneous chemical reaction mechanism on the aerosol surface is embedded into the core chemical reaction equation of the model;

[0018] Based on the real-time relative humidity within the model simulation grid, the reaction rate of the heterogeneous chemical reaction is dynamically adjusted through a preset humidity dependence function; and after the simulation is completed, the sensitivity analysis module is invoked to quantitatively assess the contribution rate of each regional pollution source to the fine particulate matter concentration in the target area based on the source-receptor model or the adjoint model.

[0019] Furthermore, the atmospheric chemical transport model program is also configured as follows:

[0020] An integrated aerosol microphysics module is used to simulate the nucleation, aggregation, and collision processes of particulate matter. A non-uniform vertical stratification structure is adopted, and the density distribution of the stratification is dynamically determined based on atmospheric boundary layer height and topographic elevation data.

[0021] Furthermore, the atmospheric chemical transport model program communicates with the emission inventory processing module through a standardized data interface, which supports NETCDF or HDF5 data formats. The emission inventory processing module is configured to perform three-dimensional spatial interpolation on the input emission source inventory data at each simulation time step, and map the emission source species to the species corresponding to the chemical mechanisms in the atmospheric chemical transport model according to a preset mapping table.

[0022] Furthermore, the atmospheric chemical transport model program has a regional source resolution engine embedded within it;

[0023] The regional source resolution engine uses tracer labeling to identify SO2 and NO from different geographical regions. x Each NH3 emission source is assigned a unique virtual digital tag.

[0024] The heterogeneous chemical reaction mechanism simultaneously processes the transfer and conversion of the virtual digital tags during chemical reaction calculations to maintain the mass conservation of the tags before and after the reaction.

[0025] Furthermore, the atmospheric chemical transport model program is coupled with a boundary layer parameterization scheme;

[0026] The boundary layer parameterization scheme is based on the Monin-Obukhov similarity theory and uses turbulent kinetic energy and potential temperature gradient to calculate the vertical diffusion coefficient of aerosols.

[0027] Furthermore, when a temperature inversion event is detected, the preset critical relative humidity threshold in the heterogeneous chemical reaction mechanism is dynamically lowered, and the magnitude of the reduction is negatively correlated with the intensity of the temperature inversion layer.

[0028] Furthermore, the WRF-CHEM-based method for predicting the distribution of fine particulate matter includes the following steps:

[0029] 1) Obtain meteorological data and pollution source emission inventory data through the data input unit;

[0030] 2) Execute the atmospheric chemical transport model program in memory via at least one processor to perform the following operations:

[0031] ① The heterogeneous chemical reaction mechanism on the aerosol surface is embedded into the core chemical reaction equation of the model;

[0032] ②Based on the real-time relative humidity within the model simulation grid, the reaction rate of the heterogeneous chemical reaction is dynamically adjusted through a preset humidity dependence function;

[0033] ③ After the simulation is completed, call the sensitivity analysis module to quantitatively assess the contribution rate of each regional pollution source to the fine particulate matter concentration in the target area based on the source-receptor model or the adjoint model.

[0034] 3. Beneficial effects

[0035] Compared with the prior art, the advantages of this invention are:

[0036] 1) The WRF-CHEM-based fine particulate matter (FPM) distribution prediction system can improve the accuracy of FPM concentration prediction in target areas by integrating meteorological data, pollution source emission inventory data, and heterogeneous chemical reaction mechanisms. The dynamic regulation effect of humidity on heterogeneous chemical reaction rates can effectively simulate the particulate matter formation process under different humidity conditions, thereby improving the accuracy and reliability of predictions.

[0037] 2) By dynamically adjusting the humidity-dependent function, the system can simulate and adjust the effect of humidity on heterogeneous chemical reaction rates, providing a basis for accurately predicting the impact of humidity changes on atmospheric chemical reactions. Experimental results show that the important role of humidity in aerosol surface reactions further emphasizes the value of the humidity-dependent function in improving the prediction of chemical reaction rates.

[0038] 3) By integrating an aerosol microphysics module, the system can simulate the nucleation, aggregation, and collision processes of particulate matter, further optimizing the atmospheric chemical transport model's description of the aerosol life cycle. The WRF-CHEM-based model can simulate the impact of pollution sources on fine particulate matter concentrations under different meteorological conditions, thus providing real-time data support for air pollution prevention and control. Detailed Implementation

[0039] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0040] Example 1:

[0041] The WRF-CHEM-based fine particulate matter distribution prediction system includes a data input unit configured to acquire meteorological data and pollution source emission inventory data.

[0042] At least one processor; and

[0043] The memory stores an atmospheric chemical transport model program that can be executed by at least one processor;

[0044] The atmospheric chemical transport model program was executed as follows:

[0045] The heterogeneous chemical reaction mechanism on the aerosol surface is embedded into the core chemical reaction equation of the model;

[0046] Based on the real-time relative humidity within the model simulation grid, the reaction rate of the heterogeneous chemical reaction is dynamically adjusted through a preset humidity dependence function; and after the simulation is completed, the sensitivity analysis module is invoked to quantitatively assess the contribution rate of each regional pollution source to the fine particulate matter concentration in the target area based on the source-receptor model or the adjoint model.

[0047] Assessments are typically performed using source-receptor models or concomitant models. Source-receptor models calculate the contribution of each source to pollution levels at a specific location by tracking emissions from different sources and combining this with meteorological and atmospheric chemical models. Specific steps include:

[0048] Source analysis engine: for each pollution source (such as SO2, NO) x Each NH3 is assigned a unique virtual tag.

[0049] Sensitivity analysis: Based on the model results, the contributions from different sources are identified. Sensitivity analysis calculates the impact of changes in each source on the pollution concentration changes in the target area.

[0050] The adjoint model, similar to the source-receptor model, reflects the influence of the source on the concentration distribution, but focuses on calculating the influence of the source in reverse, helping to determine the transport path of pollutants from the receptor to the source.

[0051] The atmospheric chemical transport model program is also configured as follows:

[0052] An integrated aerosol microphysics module simulates the nucleation, aggregation, and collision processes of particulate matter; and a non-uniform vertical stratification structure is adopted, with the density distribution of the strata dynamically determined based on atmospheric boundary layer height and topographic elevation data.

[0053] The atmospheric chemical transport model program communicates with the emission inventory processing module through a standardized data interface that supports NETCDF or HDF5 data formats. The emission inventory processing module is configured to perform three-dimensional spatial interpolation on the input emission source inventory data at each simulation time step, and map the emission source species to the species corresponding to the chemical mechanisms in the atmospheric chemical transport model according to a preset mapping table.

[0054] The atmospheric chemical transport model program has a regional source resolution engine embedded within it;

[0055] The regional source resolution engine uses tracer labeling to identify SO2 and NO from different geographic regions. x Each NH3 emission source is assigned a unique virtual digital tag.

[0056] The heterogeneous chemical reaction mechanism simultaneously processes the transfer and transformation of the virtual digital tags during chemical reaction calculations to maintain the mass conservation of the tags before and after the reaction.

[0057] The atmospheric chemical transport model program is coupled with a boundary layer parameterization scheme.

[0058] The boundary layer parameterization scheme for atmospheric chemical transport models is based on the Monin-Obukhov similarity theory, which describes the similarity between atmospheric stability and turbulence. It typically involves the following steps:

[0059] Turbulent kinetic energy: Calculates the turbulent kinetic energy caused by factors such as wind speed and temperature.

[0060] Potential temperature gradient: Estimating the intensity of turbulence using the vertical temperature gradient.

[0061] Vertical diffusion coefficient: Based on these turbulence and potential temperature gradient data, the vertical diffusion coefficient of aerosols is calculated, which is the intensity of particulate matter diffusion in the vertical direction. When the model detects a temperature inversion event (a temperature inversion refers to an increase in atmospheric temperature with increasing altitude), the humidity threshold is dynamically lowered. This is because temperature inversions lead to airflow instability, affecting the vertical diffusion of pollutants.

[0062] The boundary layer parameterization scheme is based on the Monin-Obukhov similarity theory and uses turbulent kinetic energy and potential temperature gradient to calculate the vertical diffusion coefficient of aerosols.

[0063] Based on the Monin-Obukhov similarity theory, the vertical diffusion coefficient (Kz) of aerosols is calculated from the turbulent kinetic energy (TKE) and the potential temperature gradient (θ), and the usual calculation formula is as follows:

[0064]

[0065] Where C k It is a constant, usually taken as 0.4. TKE is turbulent kinetic energy, representing the intensity of turbulence, and potential temperature gradient represents the vertical rate of temperature change.

[0066] Furthermore, when a temperature inversion event is detected, the preset critical relative humidity threshold in the heterogeneous chemical reaction mechanism is dynamically lowered, with the magnitude of the reduction being negatively correlated with the intensity of the temperature inversion layer.

[0067] The method for predicting the distribution of fine particulate matter based on WRF-CHEM includes the following steps:

[0068] 1) Obtain meteorological data and pollution source emission inventory data through the data input unit;

[0069] 2) Execute the atmospheric chemical transport model program in memory via at least one processor to perform the following operations:

[0070] ① The heterogeneous chemical reaction mechanism on the aerosol surface is embedded into the core chemical reaction equation of the model;

[0071] Heterogeneous chemical reactions in aerosols typically refer to chemical reactions occurring on the surfaces of gases and aerosol particles. For example, SO2 and NO... x It can react with the particle surface to generate secondary particulate matter. The model requires:

[0072] These heterogeneous reactions are incorporated into the reaction rate equations as part of the chemical reaction mechanism.

[0073] The reaction rate can be adjusted using dynamic chemical reaction parameters (e.g., the reactivity of aerosol surfaces, the humidity dependence of aerosols).

[0074] ②Based on the real-time relative humidity within the model simulation grid, the reaction rate of the heterogeneous chemical reaction is dynamically adjusted through a preset humidity dependence function;

[0075] Humidity plays a crucial role in aerosol surface reactions. Typically, humidity-dependent functions are used to describe how humidity affects the reaction rate. The following methods can be used:

[0076] The reaction rate can be adjusted using humidity-dependent functions, such as exponential or linear models.

[0077] Rate adjusted =Rate0·f(RH)

[0078] Where Rate0 is the baseline reaction rate, f(RH) is the humidity function, and RH is the relative humidity.

[0079] When the humidity in the model changes, the reaction rate will be dynamically adjusted, especially since the intensity of heterogeneous chemical reactions will be significantly different under different humidity conditions.

[0080] ③ After the simulation is completed, call the sensitivity analysis module to quantitatively assess the contribution rate of each regional pollution source to the fine particulate matter concentration in the target area based on the source-receptor model or the adjoint model.

[0081] Experiment Example 1: Dynamic Regulation of Aerosol Heterogeneous Chemical Reaction Rate

[0082] The purpose of this experiment is to verify the role of humidity-dependent functions in heterogeneous chemical reactions, especially how to dynamically adjust the reaction rate through changes in humidity.

[0083] Experimental methods:

[0084] 1. Experimental Design: Select typical aerosol types (such as SO2 and NO) x The reaction with the particle surface was simulated experimentally under different relative humidity (RH) conditions.

[0085] 2. Simulation conditions: Set up the WRF-CHEM model, simulate different humidity conditions (e.g., 40%, 60%, 80% RH), and adjust the humidity dependence function.

[0086] 3. Humidity-dependent function: The following humidity-dependent function is used to adjust the reaction rate:

[0087] Rate adjusted =Rate0·f(RH)

[0088] Where Rate0 is the baseline reaction rate, f(RH) is the humidity function, and RH is the relative humidity.

[0089] 4. Experimental steps:

[0090] 1) Data input unit: Acquire meteorological data and pollution source emission inventory data.

[0091] 2) Model execution: The reaction mechanism is executed through the WRF-CHEM model, embedding the heterogeneous reaction mechanism of aerosols into the core reaction equation of the model, and adjusting the reaction rate according to the real-time RH.

[0092] 4) Humidity Adjustment: Adjust the reaction rate according to the humidity dependence function, and simulate SO2 and NO under different humidity conditions. x The reaction process with aerosol particles.

[0093] 5) Analysis results: Based on the simulation results, compare the changes in reaction rate under different humidity conditions and evaluate the effect of humidity on the reaction rate.

[0094] 5. Experimental Results:

[0095] Under low humidity (40% RH) conditions, SO2 and NO x The heterogeneous reaction rate is relatively low, resulting in fewer secondary particulate matter formations. Under high humidity (80% RH) conditions, the reaction rate increases significantly, leading to an increase in the amount of secondary particulate matter formed.

[0096] The results show that humidity has a significant impact on the heterogeneous chemical reaction rate of aerosols, and the humidity-dependent function can effectively regulate the reaction rate.

[0097] Humidity dependence function: f(RH)=a·RH+b is the fitting parameter, and a and b are obtained through regression analysis.

[0098] 6. Parameter confidence:

[0099] Simulated humidity range: 40%-80% RH

[0100] Initial reaction rate: Rate0 is based on literature values ​​and experimental measurements, and is typically set to 10. -4mol / m 2 / s.

[0101] Experimental Example 2: Source-receptor model for assessing the contribution rate of regional pollution sources

[0102] This experiment uses a source-receptor model to quantitatively assess the contribution of different pollution sources to the concentration of fine particulate matter in the target area.

[0103] Experimental methods:

[0104] 1. Experimental Design: Select typical pollution sources (such as SO2, NO) x The study simulated emissions of NH3 and other gases in different geographical regions (such as urban, industrial, and rural areas).

[0105] 2. Simulation conditions: The concentration of fine particulate matter at different time steps was simulated using the WRF-CHEM model, and the contribution of pollution sources in each region was analyzed using the source-receptor model.

[0106] 3. Sensitivity Analysis: Based on the model results, use the sensitivity analysis module to assess the impact of changes in pollution sources on the concentration in the target area.

[0107] 4. Experimental steps:

[0108] 1) Data input unit: Acquire meteorological data, emission source data and geographic data.

[0109] 2) Source resolution engine: uses tracer labeling to identify SO2 and NO. x Each source of NH3 emissions is assigned a unique virtual digital label.

[0110] 3) Source-receptor model calculation: Run the source-receptor model to calculate the contribution rate of each regional pollution source to the fine particulate matter concentration in the target area.

[0111] 4) Sensitivity analysis: After the simulation is completed, sensitivity analysis is performed using the adjoint model or source-receptor model to quantitatively assess the contribution of each pollution source to the concentration of fine particulate matter.

[0112] 5. Experimental Results:

[0113] SO2 and NO in urban areas x The emission sources contributed the most to the concentration of fine particulate matter in the target area, reaching 40%, while the NH3 emission sources in industrial areas contributed less, accounting for only 5%. Overall, the pollution sources in urban and industrial areas contributed far more than those in rural areas.

[0114] The calculation formula for the source receptor model is:

[0115]

[0116] Where Ctarget is the fine particulate matter concentration in the target area, fi is the source sensitivity coefficient, and Si is the source emission intensity.

[0117] 6. Parameter confidence:

[0118] Source emission data: Based on regional emission inventories, using the latest pollution source data.

[0119] Model accuracy: Standard error analysis is used to evaluate model error and ensure the reliability of the results.

[0120] Experiment Example 3: Calculation and Influence of Vertical Diffusion Coefficient

[0121] This experiment uses the Monin-Obukhov similarity theory to calculate the vertical diffusion coefficient of aerosols and assess the effects of turbulence and potential temperature gradient on the diffusion coefficient.

[0122] Experimental methods:

[0123] 1. Experimental design: Select typical meteorological conditions (wind speed, temperature, humidity) and calculate the vertical diffusion coefficient at different times and altitudes.

[0124] 2. Simulation conditions: WRF-CHEM was used to simulate turbulent kinetic energy (TKE) and potential temperature gradient (θ) at different altitudes and geographical locations to calculate the vertical diffusion coefficient of aerosols.

[0125] 3. Experimental steps:

[0126] 1) Data input unit: Acquire meteorological data and boundary layer height data.

[0127] 2) Calculation of turbulent kinetic energy and potential temperature gradient: TKE and potential temperature gradient are calculated based on the Monin-Obukhov similarity theory.

[0128] 3) Calculation of vertical diffusion coefficient: The vertical diffusion coefficient of aerosols is calculated using the formula.

[0129] 4) Simulation and evaluation: Simulate aerosol diffusion in different geographical regions and at different times, and evaluate the effects of turbulence and potential temperature gradient on the vertical diffusion coefficient.

[0130] 4. Experimental Results:

[0131] The stronger the turbulent kinetic energy, the greater the vertical diffusion coefficient, and the faster the vertical diffusion rate of aerosols.

[0132] Under inverted temperature conditions, the vertical diffusion coefficient decreases, and the diffusion of aerosols is inhibited.

[0133]

[0134] Where Ck = 0.4. TKE is the turbulent kinetic energy, and θ is the potential temperature gradient.

[0135] Parameter confidence:

[0136] Turbulent kinetic energy: based on WRF simulation data, with values ​​ranging from 0.5m. 2 / s 2 up to 2.0m 2 / s 2 ;

[0137] Potential temperature gradient: Based on model output data, the value ranges from 0.01 K / m to 0.1 K / m.

[0138] The above description is merely a preferred embodiment of the present invention; however, the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and its improved concepts, should be covered within the scope of protection of the present invention.

Claims

1. A fine particulate matter distribution prediction system based on WRF-CHEM, characterized in that: include: The data input unit is configured to acquire meteorological data and pollution source emission inventory data; At least one processor; as well as The memory stores an atmospheric chemical transport model program that can be executed by the at least one processor; The atmospheric chemical transport model program is executed as follows: The heterogeneous chemical reaction mechanism on the aerosol surface is embedded into the core chemical reaction equation of the model; Based on the real-time relative humidity within the model simulation grid, the reaction rate of the heterogeneous chemical reaction is dynamically adjusted through a preset humidity dependence function; and after the simulation is completed, the sensitivity analysis module is invoked to quantitatively assess the contribution rate of each regional pollution source to the fine particulate matter concentration in the target area based on the source-receptor model or the adjoint model.

2. The fine particulate matter distribution prediction system based on WRF-CHEM according to claim 1, characterized in that: The atmospheric chemical transport model program is also configured to: An integrated aerosol microphysics module is used to simulate the nucleation, aggregation, and collision processes of particulate matter. A non-uniform vertical stratification structure is adopted, and the density distribution of the stratification is dynamically determined based on atmospheric boundary layer height and topographic elevation data.

3. The fine particulate matter distribution prediction system based on WRF-CHEM according to claim 1, characterized in that: The atmospheric chemical transport model program communicates with the emission inventory processing module through a standardized data interface, which supports NETCDF or HDF5 data formats. The emission inventory processing module is configured to perform three-dimensional spatial interpolation on the input emission source inventory data at each simulation time step, and map the emission source species to the species corresponding to the chemical mechanisms in the atmospheric chemical transport model according to a preset mapping table.

4. The fine particulate matter distribution prediction system based on WRF-CHEM according to claim 1, characterized in that: The atmospheric chemical transport model program has a regional source resolution engine embedded within it. The regional source resolution engine uses tracer labeling to identify SO2 and NO from different geographical regions. x Each NH3 emission source is assigned a unique virtual digital tag. The heterogeneous chemical reaction mechanism simultaneously processes the transfer and conversion of the virtual digital tags during chemical reaction calculations to maintain the mass conservation of the tags before and after the reaction.

5. The fine particulate matter distribution prediction system based on WRF-CHEM according to claim 1, characterized in that: The atmospheric chemical transport model program is coupled with a boundary layer parameterization scheme. The boundary layer parameterization scheme is based on the Monin-Obukhov similarity theory and uses turbulent kinetic energy and potential temperature gradient to calculate the vertical diffusion coefficient of aerosols. Furthermore, when a temperature inversion event is detected, the preset critical relative humidity threshold in the heterogeneous chemical reaction mechanism is dynamically lowered, and the magnitude of the reduction is negatively correlated with the intensity of the temperature inversion layer.

6. A method for predicting the distribution of fine particulate matter based on WRF-CHEM according to claims 1 to 5, characterized in that... Includes the following steps: 1) Obtain meteorological data and pollution source emission inventory data through the data input unit; 2) Execute the atmospheric chemical transport model program in memory via at least one processor to perform the following operations: ① The heterogeneous chemical reaction mechanism on the aerosol surface is embedded into the core chemical reaction equation of the model; ②Based on the real-time relative humidity within the model simulation grid, the reaction rate of the heterogeneous chemical reaction is dynamically adjusted through a preset humidity dependence function; ③ After the simulation is completed, call the sensitivity analysis module to quantitatively assess the contribution rate of each regional pollution source to the fine particulate matter concentration in the target area based on the source-receptor model or the adjoint model.

Images