Pollutant source analysis method and system for establishing coupling of air quality mode and receptor model

Abstract

The invention belongs to the technical field of air quality management, and discloses a pollutant source analysis method and system for establishing coupling of an air quality mode and a receptor model, and the method comprises the steps: carrying out three-dimensional regional air quality mode simulation on the basis of an air pollutant emission list and a meteorological field based on meteorological mode simulation; through verification of observation data, an adopted three-dimensional regional air quality mode system comprising a meteorological mode and an air quality mode has reliable simulation performance; a source analysis module embedded in a three-dimensional area air quality mode is used for carrying out source analysis on atmospheric pollutants. According to the method, respective advantages of a three-dimensional regional air quality mode and a receptor model in the aspect of source analysis are integrated, the accuracy of refined source analysis of atmospheric pollutants composed of multiple components such as PM2.5, PM10 and VOCs in the atmosphere of a single site or multiple sites is remarkably improved, and refined and scientific development of urban and regional atmospheric pollution prevention and control is promoted; therefore, the health of people is better protected.

Description

Technical Field

[0001] The present invention belongs to the technical field of air quality management, and in particular relates to a pollutant source analysis method and system for establishing a coupled air quality model and a receptor model. Background Art

[0002] Currently, there are many existing pollutant control systems. For example, the invention patent "Indoor Environmental Pollution Control System" (Patent No. 201710762642.9) discloses a method for monitoring, evaluating, and improving indoor environmental quality; and the utility model patent "Environmental Emergency Three-Level Control System" (Patent No. 201220604879.7) discloses a feasible solution for corporate environmental pollution. However, existing methods for atmospheric pollutant source attribution lack a single approach that integrates the respective advantages of three-dimensional regional air quality models and receptor models and significantly improves the accuracy of refined source attribution. Furthermore, currently, the main source attribution methods still rely on receptor models (such as PMF and CMB) or source attribution modules of regional air quality models. These two methods are used independently, failing to effectively integrate the respective advantages of the two methods to achieve more accurate source attribution. Therefore, there is an urgent need for a method that can effectively integrate the respective advantages of receptor models and regional air quality models, achieve complementary advantages, improve the refined source attribution of atmospheric pollutants, and enhance their accuracy, provide scientific guidance for atmospheric pollutant control, promote air quality improvement, and protect public health.

[0003] Through the above analysis, the problems and defects of the existing technologies are as follows: (1) The atmospheric pollutant source apportionment method based on regional air quality models lacks scientific constraints on the source contribution of observation data. 2) The source apportionment method based on receptor models lacks quantitative identification of the source contribution of different regions. The existing receptor model source apportionment method lacks consideration of the spatial distribution of atmospheric pollutant emissions, the spatial distribution of multiple meteorological elements, and the high-resolution temporal evolution of emissions and meteorology. 3) There is a lack of receptor model source apportionment methods that can cover the multiphase chemical reaction mechanisms of complex atmospheric physical and chemical processes.

[0004] Among existing source apportionment techniques, those based on three-dimensional regional air quality models (3DRAQs) suffer from uncertainty from three sources. First, the required inputs for the atmospheric pollutant emission inventory and simulated meteorological fields introduce certain errors. Second, the required inputs for atmospheric chemistry, due to their imperfect nature, also lead to errors. Finally, source apportionment methods based on 3DRAQs (whether ISAM, PSAT, OSAT, etc.) inherently rely on model construction, which is inherently subject to errors. These errors collectively affect the accuracy of the final source apportionment results, causing the sum of the contributions from various sources to deviate from the observed concentrations of atmospheric pollutants.

[0005] The source apportionment methods based on receptor models in existing source apportionment technologies all rely on the observation data of atmospheric pollutants at the receptor points and are implemented through statistical methods or mass conservation methods based on source spectrum data. However, these methods have the following three defects: (1) Lack of meteorological and emission information: These receptor model methods do not consider meteorological factors, resulting in a lack of regional or location information of contributing sources. Since they are not combined with the emission inventory of atmospheric pollutants, they cannot reflect the source emission intensity information of different regions or locations; (2) Incomplete source spectrum information: The receptor model source apportionment method relies on the source spectrum (source emission component spectrum) or uses source spectrum information to determine the source category of the contributing source, but the existing source spectrum is also incomplete, and the measurement methods of many components of its source emissions (source spectrum) are still under development. In the current source spectrum, the number of source categories is small, resulting in insufficient source classification; (3) Ignoring atmospheric physical and chemical reactions: These receptor model methods do not consider the atmospheric physical and chemical reactions that occur during the process from emission to atmospheric environment receptor points, which is inconsistent with the actual situation. Many components of many atmospheric pollutants (such as particulate matter and VOCs) are formed through secondary chemical reactions and phase change processes in the atmospheric environment. These factors lead to large errors in the existing source apportionment results based on receptor models, lack of refinement and accuracy, and are unable to meet the current needs of urban or regional air pollution prevention and control. Summary of the Invention

[0006] In response to the problems existing in the prior art, the present invention provides a pollutant source analysis method and system for establishing a coupled air quality model and a receptor model.

[0007] The present invention is implemented as follows: a pollutant source analysis method for establishing a coupled air quality model and a receptor model includes:

[0008] Step 1: Conduct a three-dimensional regional air quality model simulation based on the atmospheric pollutant emission inventory and the meteorological field simulated by the meteorological model;

[0009] Step 2: After verification by observation data, the three-dimensional regional air quality model system including the meteorological model and the air quality model has reliable simulation performance;

[0010] Step 3: Use the most advanced source apportionment method embedded in the 3D regional air quality model to simulate the source apportionment of atmospheric pollutants and obtain the source apportionment results;

[0011] Step 4: Using the time series concentration observation data of atmospheric pollutants and their components at the observation point, and based on the definition of relevant species based on the chemical mechanism used in the 3D regional air quality model simulation, compare them with the observed components, establish a corresponding relationship between the chemical mechanism species and the observed components, and further establish a conversion relationship between their concentration data, thereby obtaining the time series observation concentrations of the chemical mechanism species at the observation site;

[0012] Step 5: Couple the receptor model with the three-dimensional regional air quality model to construct an objective function and apply a nonlinear optimization algorithm to obtain the correction coefficient of each source's contribution to the receptor point pollutant under the constraints of the time series observation data.

[0013] Step 6: Obtain source apportionment results based on the coupling of air quality model and receptor model.

[0014] Furthermore, determine the single or multiple observation sites and target pollutants for source apportionment. The pollutants should be composed of multiple components, and time series concentration observation data for the pollutants and their components should be available. Analyze the time series concentration observations at the observation sites to understand the temporal evolution and distribution characteristics of the pollutant and its component concentration data, and clean the data.

[0015] Before conducting a three-dimensional regional air quality model simulation, it is necessary to set the simulation area and the grid resolution of each simulation area, and the innermost simulation area must cover the receptor points that require source analysis;

[0016] Conduct simulations of baseline emission scenarios using a three-dimensional regional air quality model system. The model used is the internationally renowned CMAQ (Community Multiscale Air Quality Model) chemical transport model. Prior to conducting simulations, the air quality model system must be configured, including the selection of parameterization schemes for various physical and chemical processes, the preparation of emission inventories for anthropogenic and natural sources, and other input data. The baseline scenario is then simulated based on the three-dimensional regional air quality model system. Observational data is then used to evaluate the model system's simulation performance. If performance is poor, error correction and parameter adjustments are performed to ensure the model system's reliability.

[0017] Adopting source apportionment methods embedded in the most advanced three-dimensional regional air quality models, including ISAM (The Integrated Source Apportionment Method) in the CMAQ model, PSAT (The Particulate Matter Source Apportionment Technology) and OSAT (The Ozone Source Apportionment Technology) in the CAMx (Comprehensive Air Quality Model with Extensions) model, as well as new source apportionment methods that will be improved or developed in the future;

[0018] Using the source apportionment method of the three-dimensional regional air quality model, we conduct source apportionment simulation for the target pollutants in the innermost simulation area and obtain the source apportionment results of the target pollutants for each grid in the entire simulation area.

[0019] The target pollutants selected include PM 2.5 , PM 10 , VOCs (Volatile organic compounds) are pollutants composed of multiple components;

[0020] The chemical mechanisms in the adopted three-dimensional regional air quality model include species related to the components of the target pollutants, also known as chemical mechanism species;

[0021] The chemical mechanism species in the air quality model may not be completely consistent with the components observed at the observation station. Therefore, it is necessary to construct a corresponding relationship between the two based on the definition of the species in the chemical mechanism and the specific components observed at the observation station, including the molar ratio relationship, and further establish a concentration conversion relationship between the two. Through this process, we can obtain (1) the observed concentration of the chemical mechanism species at the observation station, or (2) the simulated concentration of the observed component at the observation station. Only one of them is required. The purpose of obtaining (1) or (2) is to establish comparability between the observed concentration and the model simulated concentration, thereby laying the foundation for establishing a new source attribution method for coupling air quality models and receptor models, and ultimately achieving more accurate and refined source attribution.

[0022] Furthermore, the objective function (Formulas A, B, and C) for coupling the air quality model and the receptor model is constructed;

[0023]

[0024] here, is the uncertainty of the observed concentration of species i; is the source apportionment result obtained based on the three-dimensional regional air quality model, that is, the contribution concentration of source j to species i at the receptor point (grid); is a coefficient introduced, similar to the virtual source spectrum;

[0025] In order to reflect the influence of source contribution, avoid collinearity problems, and obtain the minimum x 2 , we define an R j , a correction factor applied to the contribution of source j to correct the source contribution simulated by the air quality model As shown in formula (B);

[0026]

[0027] Further for R j It is reasonable in practical sense and can explain the uncertainty of source contributions obtained from air quality model simulation. Here, additional constraints and a penalty term are added to solve an optimized R j , as shown in formula (C)

[0028]

[0029] here, represents the a priori uncertainty of all sources based on the air quality model for species i, and it can be assumed that the size of the a priori uncertainty is proportional to the observed concentration of species i, that is, it can be obtained by Obtain, where δ i is the normalized mean error (NME); the second term on the right side of formula (C) represents the uncertainty of the contribution of a single source due to emission errors; is the a priori uncertainty of the natural logarithm of the correction factor of source j; when performing the optimization calculation, we can initially set Where N is the number of species, J CTM Represents the number of contributing sources, so that the two terms on the right side of formula (C) have the same weight. Then, the nonlinear optimization problem is solved by a multi-iteration optimization algorithm to minimize the objective function and obtain the correction factor (Rj) of each source contribution. Then, the contribution concentration of each component (species) of the target pollutant at the receptor point (observation point) and the contribution concentration of each source (including different source categories, different areas, and different source categories in different specific areas) to the target pollutant at the receptor point (observation point) are obtained.

[0030] Furthermore, the optimization of the objective function (Formula C) includes minimizing the objective function by various optimization algorithms applicable to constrained nonlinear optimization problems.

[0031] Furthermore, including using the correction factor (Rj) contributed by each source and the source apportionment result of the air quality model, according to the following formula (D), the new source apportionment method based on the coupling of the air quality model and the receptor model is calculated to obtain the contribution concentration of each component (or species) of the target pollutant of each source (including different source categories, different regions, and different source categories in different specific regions) to the receptor point (observation point), and then the contribution concentration of each component (or species) of the same source to the receptor point is added to obtain the contribution concentration of the target pollutant of each source to the receptor;

[0032]

[0033] Furthermore, the source apportionment results obtained by the new source apportionment method based on the coupling of a three-dimensional regional air quality model and a receptor model were compared with those based only on the three-dimensional regional air quality model and only on the receptor model to evaluate the superiority of the new method.

[0034] As shown in formulas (E-J), the closer R and IOA (index of agreement) are to 1, the better the source apportionment results based on the new method are. The smaller the other statistical parameters are, the better the results are. Compared with the source apportionment results based only on the receptor model, the comparison can be made from the perspectives of the number of sources, degree of refinement, regional contribution, and uncertainty.

[0035]

[0036] Another object of the present invention is to provide a pollutant source analysis system for coupling an air quality model with a receptor model, comprising:

[0037] The simulation module is used to simulate the three-dimensional regional air quality model based on the atmospheric pollutant emission inventory and the meteorological field simulated by the meteorological model;

[0038] A verification module is used to verify that the three-dimensional regional air quality model system, including the meteorological model and the air quality model, has reliable simulation performance through observation data verification;

[0039] Source apportionment module, which uses the most advanced source apportionment methods embedded in the three-dimensional regional air quality model to simulate the source apportionment of atmospheric pollutants and obtain source apportionment results;

[0040] The observation module is used to use the time series concentration observation data of atmospheric pollutants and their components at the observation point, and based on the definition of relevant species of the chemical mechanism adopted in the three-dimensional regional air quality model simulation, compare it with the observed components, establish the corresponding relationship between the chemical mechanism species and the observed components, and further establish the conversion relationship between their concentration data, so as to obtain the time series observation concentration of the chemical mechanism species at the observation site;

[0041] The correction module is used to couple the receptor model with the three-dimensional regional air quality model, construct the objective function, and apply the nonlinear optimization algorithm to obtain the correction coefficient of each source's contribution to the receptor point pollutant under the constraints of the time series observation data.

[0042] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the pollutant source analysis method for establishing the coupling of the air quality model and the receptor model.

[0043] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to execute the steps of the pollutant source analysis method for establishing a coupling between an air quality model and a receptor model.

[0044] Another object of the present invention is to provide an information data processing terminal, which is used to implement the pollutant source analysis system that establishes the coupling of the air quality model and the receptor model.

[0045] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0046] First, the present invention provides a method for atmospheric pollutant source attribution based on the coupling of a three-dimensional regional air quality model and a receptor model. The present invention aims to provide a method for atmospheric pollutant source attribution that leverages the complementary strengths of a three-dimensional regional air quality model and a receptor model, using constraints based on time-series observational data of atmospheric pollutants and their components. This method enables more reliable source attribution of atmospheric pollutants by source category and by region, providing an important scientific basis for efficient, scientific, and refined decision-making on actions to improve urban or regional air quality.

[0047] The present invention provides a method for establishing an atmospheric pollutant source apportionment method based on the coupling of a three-dimensional regional air quality model and a receptor model. The method includes simulating a three-dimensional regional air quality model based on an atmospheric pollutant emission inventory and a meteorological field simulated by a meteorological model; utilizing a source apportionment module embedded in the three-dimensional regional air quality model to perform source apportionment on atmospheric pollutants; the simulation performance of the three-dimensional regional air quality model system used is verified by observation data to ensure its reliability; time series observation data of atmospheric pollutants and components at receptor points are used; according to the species definition in the chemical mechanism of the air quality model, the observed concentration of the mechanism species at the observation site is converted, and an objective function for the coupling of the three-dimensional regional air quality model and the receptor model is constructed, and then an optimization algorithm for a constrained nonlinear optimization problem is used to minimize the objective function, thereby obtaining a correction factor for each source contribution and a source apportionment result of the new method for source apportionment based on the coupling of the regional air quality model and the receptor model; and evaluating the degree of optimization of the source apportionment result of the new method for source apportionment of atmospheric pollutants based on the coupling of the three-dimensional regional air quality model and the receptor model. The present invention integrates the respective advantages of the three-dimensional regional air quality model and the receptor model in source apportionment, and significantly improves the atmospheric PM ions at single or multiple sites. 2.5 、PM 10 The accuracy of refined source analysis of multi-component atmospheric pollutants such as VOCs will promote the refined and scientific development of urban and regional air pollution prevention and control, thereby better protecting people's health.

[0048] Second, the technical solution of the present invention solves the following technical problems in the prior art in industrial applications and achieves significant technical progress:

[0049] 1) Limitations of existing air quality models and pollution source analysis methods

[0050] Existing air quality models typically rely solely on emissions inventories and simple meteorological simulations, failing to precisely distinguish the contributions of different pollution sources to specific monitoring points, resulting in inaccurate source identification. Furthermore, traditional receptor models rely solely on observational data, neglecting the dynamic processes of pollutant transport and chemical reactions in the atmosphere. These shortcomings often result in existing technologies lacking sufficient scientific basis for decision-making in air pollution prevention and control, making it difficult to accurately guide emission reduction measures.

[0051] By coupling a three-dimensional regional air quality model with a receptor model and utilizing embedded source attribution methods (such as ISAM in CMAQ), this approach overcomes the existing problem of insufficient analysis of pollution source contributions and significantly improves the accuracy of source attribution. This coupled model not only considers the transport, diffusion, and chemical reaction processes of pollutants in the atmosphere, but also integrates observational data from the receptor model to produce more accurate source attribution results.

[0052] 2) Existing source apportionment methods cannot fully utilize observational data

[0053] Existing source apportionment methods are typically based on model simulations, which fail to fully utilize the actual pollutant concentrations and temporal evolution characteristics contained in observational data, resulting in low source apportionment accuracy. Furthermore, because the chemical mechanisms used in air quality models do not fully correspond to the specific components at the observation sites, deviations between model simulations and actual observations are often present.

[0054] This method establishes a conversion relationship between chemical mechanism species and observed components, aligning model-simulated chemical mechanism species concentrations with actual observational data. This allows time series data from observation sites to be used to correct model simulation results. By introducing a nonlinear optimization algorithm and dynamically adjusting the results based on observed data, the accuracy of source apportionment is further improved. Compared to existing technologies, this method can better utilize actual observational data, overcoming the limitations of traditional models that rely solely on simulation results.

[0055] This invention significantly improves the accuracy of pollution source attribution by combining a three-dimensional regional air quality model with a receptor model and incorporating state-of-the-art source apportionment methods. This advancement is of great significance in pollution prevention and environmental management, providing a scientific basis for targeted emission reduction in cities and regions, significantly improving the efficiency and effectiveness of pollution control.

[0056] The coupled model proposed in this paper is not only applicable to source apportionment at a single site but can also be extended to pollutant source apportionment across large regions, supporting air quality assessments of varying scales. Furthermore, by combining observational data with a nonlinear optimization algorithm, the proposed model exhibits enhanced adaptability to diverse environmental and meteorological conditions, dynamically adjusting analysis results based on actual observations, thus enhancing the model's versatility and practicality.

[0057] Because this method can more accurately analyze the contributions of pollution sources, it can provide environmental decision-makers with highly accurate information on pollution source emissions, helping to formulate more targeted emission reduction policies and optimize pollution control strategies. This technological advancement can better balance the needs of environmental governance and economic development, especially in complex urban environments.

[0058] The present invention not only solves the accuracy problem in the prior art, but also achieves significant technical progress in the operability of pollution control and the scientific nature of environmental management.

[0059] Third, the present invention solves the technical problems of industrial application:

[0060] 1. Limitations of independent application of air quality models and receptor models

[0061] Current technical issues: Currently, three-dimensional regional air quality models and receptor models are typically applied independently. While air quality models offer advantages in global and spatiotemporal predictions, they are significantly affected by emission inventories and model parameters, and are prone to introducing systematic errors. Receptor models excel at quantitatively analyzing source contributions based on observational data, but lack comprehensive consideration of regional transport and chemical reactions. This independent application approach struggles to simultaneously achieve accurate simulations of regional atmospheric pollution and effective utilization of observational data, resulting in inaccurate and unreliable source attribution results.

[0062] Technical bottleneck: The source attribution results provided by air quality models are susceptible to the uncertainty of emission inventories, while the accuracy of receptor models decreases when there is insufficient observation data or limited observation coverage, making it difficult to fully reveal the sources and transmission characteristics of atmospheric pollutants.

[0063] 2. The accuracy of the source apportionment results of existing source apportionment methods is insufficient

[0064] Existing technical issues: Existing methods based solely on air quality models lack mechanisms to constrain and correct model source attribution results using observational data, leading to significant uncertainty in the results. Furthermore, traditional receptor model source attribution techniques lack systematic support for analyzing the dynamic changes in regional pollutant contributions and the contributions of multiple sources, hindering the scientific and targeted nature of pollution control measures.

[0065] Technical bottleneck: How to reuse the respective advantages of the receptor models of air quality models and overcome their respective weaknesses to form a more accurate and dynamic source contribution analysis method is the core technical difficulty in the current field of atmospheric pollution source analysis.

[0066] The key technical problems solved by the technical solution of the present invention are:

[0067] 1. Solve the coupling problem between air quality model and receptor model

[0068] Existing technical problems: There are essential differences between air quality models and receptor models in modeling methods and input data, making it difficult to directly couple them for application.

[0069] Technological Breakthrough: This invention establishes a unified conversion framework between chemical species simulated by air quality models and time-series concentration data from observation sites by establishing a correspondence between chemical mechanism species and observed components, as well as a conversion relationship between concentration data. This allows for data integration between the two models. Using a nonlinear optimization algorithm, an objective function constrained by time-series observation data is constructed, coupling the air quality model with the receptor model to refine the model results and improve the accuracy of pollutant source apportionment.

[0070] 2. Improve the accuracy and dynamism of source analysis results

[0071] Existing technical issues: Traditional air quality models lack effective utilization of observational data, and their analysis results are easily affected by uncertainty in emission inventories. This invention, through a coupled receptor model, leverages the constraints of observational data to dynamically modify the source contribution results of air quality models, significantly improving the accuracy and reliability of source analysis.

[0072] Technological breakthrough: Through the constraints of time series observation data, the pollutant contribution of each source to the receptor point is dynamically corrected, so that the analysis results can more accurately reflect the actual changing characteristics of the pollutant source.

[0073] 3. Multi-model collaboration improves the scientific nature of pollution prevention and control decisions

[0074] Existing technical issues: Source apportionment results from a single model are insufficient to meet the precise governance requirements under complex pollution situations. This invention combines the global simulation capabilities of a three-dimensional regional air quality model with the local observation and analysis capabilities of a receptor model to form a multi-model collaborative pollutant source apportionment method, significantly enhancing the targeted and scientific nature of pollution prevention and control measures.

[0075] Technological breakthrough: The analytical results based on the coupled model can not only accurately identify the contribution of pollution sources, but also dynamically adjust the emission inventory and model parameters, providing strong data support for regional pollution prevention and control and policy formulation.

[0076] Significant technical advancements of the present invention:

[0077] 1. Innovative application of two types of model coupling and data constraints

[0078] The coupling of the three-dimensional regional air quality model and the receptor model was achieved, and the advantages of the two models were complemented to form a new source apportionment method.

[0079] The innovative introduction of time series observation data as a constraint condition dynamically corrects the model results, significantly reducing the impact of the uncertainty of emission inventories and model parameters on the analysis results.

[0080] 2. High precision and reliability of source apportionment results

[0081] By introducing a nonlinear optimization algorithm, constructing an objective function based on observation data constraints, and dynamically optimizing the correction coefficient contributed by each pollution source, the accuracy of the analytical results is ensured.

[0082] The source apportionment results not only reflect the spatial distribution of pollutant sources, but also dynamically reveal the temporal change characteristics of pollution sources, providing comprehensive support for refined governance.

[0083] 3. Scientific decision-making support for air pollution prevention and control

[0084] The analysis results can provide more accurate pollution source contribution data and support the precise design and implementation of regional pollution prevention and control measures.

[0085] By dynamically revising and optimizing the emission inventory, governments and enterprises can be assisted in formulating more scientific and targeted pollution reduction strategies, reducing governance costs and improving governance effectiveness.

[0086] 4. Extensive industrial application value

[0087] It is suitable for tracing pollutant sources during severe atmospheric pollution processes, especially when multiple pollutants overlap and regional transmission is complex, and significantly improves the reliability of analysis results.

[0088] It is suitable for application by urban, regional and national air quality management departments, and provides a scientific basis for optimizing the construction layout of monitoring stations and comprehensive prevention and control of air pollution.

[0089] Through the above innovations and technological advancements, the present invention has important application value and driving role in the field of pollutant source analysis technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 This is a flow chart of a pollutant source analysis method for establishing an air quality model and a receptor model coupling provided by an embodiment of the present invention.

[0091] Figure 2 This is a structural block diagram of a pollutant source analysis system for establishing an air quality model and a receptor model coupling provided by an embodiment of the present invention.

[0092] Figure 3 This is a detailed flow chart of a method for establishing an atmospheric pollutant source analysis method based on the coupling of a three-dimensional regional air quality model and a receptor model, provided by an embodiment of the present invention.

[0093] Figure 4 Schematic diagram of the simulation area setting of the air quality model system provided by an embodiment of the present invention.

[0094] Figure 5 It is the cumulative distribution function of the VOCs contribution source correction factor of the Dianshan Lake station provided by the embodiment of the present invention.

[0095] Figure 6 The time series of TVOCs concentrations before and after correction (based on the air quality model only) and after correction (based on the new coupling method) for a typical month provided by the embodiment of the present invention are the same as the observed concentrations. DETAILED DESCRIPTION

[0096] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0097] The difficulty in solving existing technical problems and defects mainly lies in:

[0098] The variability of meteorological conditions and the complexity of atmospheric pollutants in the transmission process. Atmospheric pollutants undergo a series of complex atmospheric physical and chemical processes during transmission, and these processes are jointly affected by multiple meteorological factors (such as wind speed, wind direction, light radiation, precipitation, humidity, air pressure, etc.). In fact, accurately analyzing the sources of atmospheric pollutants in a city or receptor point is an extremely complex scientific problem. Although scientists have developed source attribution algorithms based on air quality models and source attribution methods based on receptor models, they still find it difficult to provide accurate and refined source attribution results. With the government's growing demand for refined control of atmospheric pollutants, especially in determining the key areas for the prevention and control of atmospheric pollutants and the scientific guidance methods for the intensity of prevention and control, the requirements for source attribution results have become higher. If there is a lack of reliable and refined source attribution results, it will be difficult to effectively meet the government's prevention and control needs, and further promoting the improvement of dynamic air quality will also face challenges.

[0099] The significance of solving the existing technical problems and defects is:

[0100] While existing source attribution methods based on three-dimensional regional air quality models and receptor models each have their own strengths and weaknesses, the shortcomings of three-dimensional regional air quality models can be offset by the advantages of receptor models, and vice versa. Therefore, coupling three-dimensional regional air quality models with receptor models, achieving complementary advantages, will help overcome the problems and shortcomings of existing technologies and holds great promise. This coupling holds important practical significance for promoting the development of precise and scientific air pollution prevention and control.

[0101] like Figure 1 As shown, an embodiment of the present invention provides a method for establishing a pollutant source analysis method by coupling an air quality model and a receptor model, comprising the following steps:

[0102] S101: Conduct three-dimensional regional air quality model simulation based on the atmospheric pollutant emission inventory and the simulated meteorological field based on the meteorological model;

[0103] S102: The three-dimensional regional air quality model system, including the meteorological model and the air quality model, has reliable simulation performance as verified by observation data;

[0104] S103, using the most advanced source apportionment method embedded in the three-dimensional regional air quality model to simulate the source apportionment of atmospheric pollutants and obtain source apportionment results;

[0105] S104: Using the time series concentration observation data of atmospheric pollutants and their components at the observation point, and based on the definition of relevant species based on the chemical mechanism used in the three-dimensional regional air quality model simulation, comparing them with the observed components, establishing a corresponding relationship between the chemical mechanism species and the observed components, and further establishing a conversion relationship between their concentration data, thereby obtaining the time series observation concentration of the chemical mechanism species at the observation site;

[0106] S105, couple the receptor model with the three-dimensional regional air quality model, construct an objective function, and apply a nonlinear optimization algorithm to obtain the correction coefficient of each source's contribution to the receptor point pollutant under the constraints of the time series observation data;

[0107] S106, obtaining source apportionment results based on the coupling of air quality model and receptor model.

[0108] The embodiments of the present invention provide for determining a single or multiple observation sites for source apportionment, as well as target pollutants; the pollutants should be composed of multiple components, and time series concentration observation data of the pollutants and their components should be available; data analysis is performed on the time series concentration observations of the observation sites to understand the temporal evolution and distribution characteristics of the concentration data of the pollutants and their components, and the data is cleaned;

[0109] Before conducting a three-dimensional regional air quality model simulation, it is necessary to set the simulation area and the grid resolution of each simulation area, and the innermost simulation area must cover the receptor points that require source analysis;

[0110] Conduct simulations of baseline emission scenarios using a three-dimensional regional air quality model system. The model used is one of the internationally renowned chemical transport models, CMAQ. Prior to conducting simulations, the air quality model system must be configured, including the selection of parameterization schemes for various physical and chemical processes, the preparation of emission inventories for anthropogenic and natural sources, and other input data. The baseline scenario is then simulated based on the three-dimensional regional air quality model system. Observational data is then used to evaluate the model system's simulation performance. If performance is poor, error correction and parameter adjustments are performed to ensure the model system's reliability.

[0111] Adopt source apportionment methods embedded in the most advanced three-dimensional regional air quality models, including ISAM in the CMAQ model, PSAT and OSAT in the CAMx model, as well as new source apportionment methods that will be improved or developed in the future;

[0112] Using the source apportionment method of the three-dimensional regional air quality model, we conduct source apportionment simulation for the target pollutants in the innermost simulation area and obtain the source apportionment results of the target pollutants for each grid in the entire simulation area.

[0113] The target pollutants selected include PM 2.5 , PM 10 , VOCs are pollutants composed of multiple components;

[0114] The chemical mechanisms in the adopted three-dimensional regional air quality model include species related to the components of the target pollutants, also known as chemical mechanism species;

[0115] The chemical mechanism species in the air quality model may not be completely consistent with the components observed at the observation station. Therefore, it is necessary to construct a corresponding relationship between the two based on the definition of the species in the chemical mechanism and the specific components observed at the observation station, including the molar ratio relationship, and further establish a concentration conversion relationship between the two. Through this process, we can obtain (1) the observed concentration of the chemical mechanism species at the observation station, or (2) the simulated concentration of the observed component at the observation station. Only one of them is required. The purpose of obtaining (1) or (2) is to establish comparability between the observed concentration and the model simulated concentration, thereby laying the foundation for establishing a new source attribution method for coupling air quality models and receptor models, and ultimately achieving more accurate and refined source attribution.

[0116] The objective function (Formulas A, B, and C) for constructing the coupling of the air quality model and the receptor model provided in the embodiment of the present invention;

[0117]

[0118] here, is the uncertainty of the observed concentration of species i; is the source apportionment result obtained based on the three-dimensional regional air quality model, that is, the contribution concentration of source j to species i at the receptor point (grid); is a coefficient introduced, similar to the virtual source spectrum;

[0119] In order to reflect the influence of source contribution, avoid collinearity problems, and obtain the minimum x 2 , we define an R j , a correction factor applied to the contribution of source j to correct the source contribution simulated by the air quality model As shown in formula (B);

[0120]

[0121] Further for R jIt is reasonable in practical sense and can explain the uncertainty of source contributions obtained from air quality model simulation. Here, additional constraints and a penalty term are added to solve an optimized R j , as shown in formula (C)

[0122]

[0123] here, represents the a priori uncertainty of all sources based on the air quality model for species i, and it can be assumed that the size of the a priori uncertainty is proportional to the observed concentration of species i, that is, it can be obtained by Obtain, where δ i is the normalized mean error (NME); the second term on the right side of formula (C) represents the uncertainty of the contribution of a single source due to emission errors; is the a priori uncertainty of the natural logarithm of the correction factor of source j; when performing the optimization calculation, we can initially set Where N is the number of species, J CT M represents the number of contributing sources, so that the two terms on the right side of formula (C) have the same weight. Then, the nonlinear optimization problem is solved by a multi-iteration optimization algorithm to minimize the objective function and obtain the correction factor (Rj) of each source contribution. Then, the contribution concentration of each component (species) of each source (including different source categories, different regions, and different source categories in different specific regions) to the target pollutant at the receptor point (observation point) after correction is obtained, as well as its contribution concentration to the target pollutant at the receptor point (observation point).

[0124] The optimization of the objective function (Formula C) provided by the embodiment of the present invention includes minimizing the objective function through various optimization algorithms applicable to constrained nonlinear optimization problems.

[0125] The embodiment of the present invention provides a new source apportionment method based on the coupling of the air quality model and the receptor model, including using the correction factor (Rj) contributed by each source and the source apportionment result of the air quality model, and calculating the contribution concentration of each component (or species) of the target pollutant of each source (including different source categories, different regions, and different source categories in different specific regions) to the receptor point (observation point) according to the following formula (D). Then, the contribution concentration of each component (or species) of the same source to the receptor point is added to obtain the contribution concentration of the target pollutant of each source to the receptor;

[0126]

[0127] The source apportionment results obtained by the new source apportionment method based on the coupling of a three-dimensional regional air quality model and a receptor model provided in an embodiment of the present invention were compared with the source apportionment results based only on the three-dimensional regional air quality model and only on the receptor model to evaluate the superiority of the new method.

[0128] As shown in formulas (E-J), the closer R and IOA are to 1, the better the source apportionment results based on the new method are, and the smaller the other statistical parameters are, the better the results are. Compared with the source apportionment results based only on the receptor model, the comparison can be made from the perspectives of the number of sources, degree of refinement, regional contribution, uncertainty, etc.

[0129]

[0130] Module structure is as follows Figure 2 As shown, an embodiment of the present invention provides a pollutant source analysis system for establishing a coupled air quality model and a receptor model, including:

[0131] The simulation module is used to simulate the three-dimensional regional air quality model based on the atmospheric pollutant emission inventory and the meteorological field simulated by the meteorological model;

[0132] A verification module is used to verify that the three-dimensional regional air quality model system, including the meteorological model and the air quality model, has reliable simulation performance through observation data verification;

[0133] Source apportionment module, which uses the most advanced source apportionment methods embedded in the three-dimensional regional air quality model to simulate the source apportionment of atmospheric pollutants and obtain source apportionment results;

[0134] The observation module is used to use the time series concentration observation data of atmospheric pollutants and their components at the observation point, and based on the definition of relevant species of the chemical mechanism adopted in the three-dimensional regional air quality model simulation, compare it with the observed components, establish the corresponding relationship between the chemical mechanism species and the observed components, and further establish the conversion relationship between their concentration data, so as to obtain the time series observation concentration of the chemical mechanism species at the observation site;

[0135] The correction module is used to couple the receptor model with the three-dimensional regional air quality model, construct the objective function, and apply the nonlinear optimization algorithm to obtain the correction coefficient of each source's contribution to the receptor point pollutant under the constraints of the time series observation data.

[0136] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the pollutant source analysis method for establishing the coupling of the air quality model and the receptor model.

[0137] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to execute the steps of the pollutant source analysis method for establishing a coupling between an air quality model and a receptor model.

[0138] Another object of the present invention is to provide an information data processing terminal, which is used to implement the pollutant source analysis system that establishes the coupling of the air quality model and the receptor model.

[0139] The present invention is specifically implemented:

[0140] The process of function implementation is as follows Figure 3 As shown, the present invention provides a method for establishing an atmospheric pollutant source apportionment based on the coupling of a three-dimensional regional air quality model and a receptor model, comprising:

[0141] 1) Based on the types of atmospheric pollutants and their components observed, select one or more observation sites and determine the pollutants for source apportionment. The pollutants must be composed of multiple components, and the concentrations of the pollutants and their components must be observed in time series.

[0142] 2) Analyze the observed concentrations of the time series at the observation sites to understand the temporal evolution and distribution characteristics of the concentrations of pollutants and their components, clean the data, and save it to the database;

[0143] 3) Set up the simulation area grid system required by the air quality model system. The simulation area includes three nested simulation areas. The outermost (first layer) simulation area covers the entire Chinese region, the second layer simulation area is a part of the area inside the first layer simulation area, and the third layer simulation area is a part of the area inside the second layer simulation area. The grid resolutions (side lengths of square grids in the horizontal direction) of the first, second, and third layer simulation areas are 9:3:1, respectively. The simulation data field of the first layer area provides boundary conditions for the second layer simulation area. Similarly, the boundary conditions of the third layer simulation area come from the simulation results of the second layer area. The third layer simulation area covers the study area, and the observation sites are also in the third layer simulation area;

[0144] 4) Select a research period and then use a meteorological model, such as WRF (Weather Research and Forecasting Model), to simulate the meteorological field of the three simulation areas.

[0145] 5) Prepare emission inventories of anthropogenic and natural sources for the three-layer simulation area. The anthropogenic emission inventory for the first and second-layer simulation areas adopts the MEIC (Multi-resolution Emission Inventory for China) inventory developed by Tsinghua University, while the third-layer simulation area adopts a more refined anthropogenic emission inventory. In addition, the anthropogenic emission inventory of the third layer needs to retain sub-inventory files of fine source categories for use in source parsing based on the air quality model. Each sub-inventory file represents a fine source class, and the number of sub-inventory files corresponds to the number of anthropogenic source categories that need to be parsed during source parsing. The natural source emission inventories for the three layers are generated based on the WRF simulated meteorological fields of the corresponding regions through the MEGAN (Model of Emissions of Gases and Aerosols from Nature) model. The emission inventory file formats of all categories in all regions must meet the format requirements of the air quality model;

[0146] 6) Perform a baseline scenario simulation based on a three-dimensional regional air quality model. This simulation is performed on a computer cluster running the Linux operating system. The three-dimensional regional air quality model used can be an internationally renowned chemical transport model such as CMAQ or CAMx. Using meteorological field files and emission inventory files, air quality simulation is performed for each layer of the simulated area. After the baseline scenario simulation is completed, the simulation performance of the model system is evaluated.

[0147] 7) Prepare the regional marker file required for source apportionment based on the air quality model. Source apportionment simulation is only performed on the third-layer simulation area, so the regional marker file is also only for the third-layer simulation area. Each grid in the regional file has a corresponding regional code. Different regional codes represent different regions. Usually, multiple grids correspond to one region, but it is also possible to have only one grid per region.

[0148] 8) Perform source attribution simulations based on the 3D regional air quality model. Using the 3D regional air quality model, you can use the ISAM tool in CMAQ or the PSAT and OSAT tools in CAMx to perform source attribution simulations for pollutants in the third-layer simulation area. The simulations ultimately yield the atmospheric pollutant source attribution results for the grid where the receptor (observation) point is located, including source category contributions, regional contributions, and contributions from different source categories within a region.

[0149] 9) Based on the names of the pollutant components observed at the observation site, the definitions of species in the chemical mechanism of the air quality model (depending on the chemical mechanism used in the air quality model), and the similarity of their chemical properties or functional groups, establish the correspondence between the observed components and the chemical mechanism species and the concentration conversion relationship, and obtain the observed concentrations of the chemical mechanism species at the observation site.

[0150] 10) By coupling the receptor model with the three-dimensional regional air quality model, the corrected source apportionment results are obtained. The simulation deviation of each species of the pollutant is decomposed into the sum of the deviations of the contribution concentrations of the corresponding species of the pollutant by each source. The purpose of the correction is to minimize the difference between the sum of the corrected source contributions and the observed concentration. A penalty term is introduced in the mathematical calculation, and the objective function is optimized (i.e., minimizing the difference) through a nonlinear optimization algorithm, and finally the correction coefficient of each source contribution is calculated. The original source contribution concentration is multiplied by the corresponding source contribution correction coefficient to obtain the corrected source contribution concentration of the receptor point pollutant. This result is the source apportionment result of the atmospheric pollutant source apportionment method based on the coupling of the three-dimensional regional air quality model and the receptor model.

[0151] 11) Evaluate the revised source atordination results, including comparing the time series simulated concentrations of individual pollutants before and after correction with the observed concentrations to evaluate the improvement effect, and also comparing the time series simulated concentrations of pollutants before and after correction with the observed concentrations to evaluate the degree of improvement of the source atordination results after coupling the three-dimensional regional air quality model with the receptor model compared with the source atordination results based only on the three-dimensional regional air quality model before coupling.

[0152] The present invention provides a new method for establishing atmospheric pollutant source apportionment based on the coupling of a three-dimensional regional air quality model and a receptor model. Ordinary technicians in the industry can also use other steps to implement it. Figure 3 What is provided is merely one specific embodiment of the present invention.

[0153] The observation sites in step 1) are selected based on the needs of the city or regional government. Pollutants composed of multiple components include PM 10 , PM 2.5 and VOCs, etc. The observation site is the receptor point.

[0154] The data quality of the observation site after data cleaning in step 2) is guaranteed to be reliable.

[0155] The horizontal grid resolutions for the first, second, and third layers of the simulation area in step 3) are typically 36 km × 36 km, 12 km × 12 km, and 4 km × 4 km, respectively. These resolutions can be adjusted as needed, as long as the horizontal grid side-to-side ratio requirement of 9:3:1 is met for each layer of the simulation area and the air quality model system used is verified and reliable.

[0156] The three-layer simulation area used in the meteorological model in step 4) is slightly larger than the surrounding areas of the corresponding layer simulation area of the air quality model, and the simulation performance of the meteorological model needs to be evaluated and its reliability.

[0157] The emission inventory prepared in step 5) needs to be verified using multiple data sources and multiple methods, and any errors or unreasonable deviations must be corrected.

[0158] In step 6), the simulation results for the outer simulation region provide the boundary conditions for the inner simulation region. The baseline scenario simulation results are evaluated using observed and simulated concentrations of multiple pollutants to verify the reliability of the model system. If the evaluation results are unsatisfactory, the parameters and parameterization schemes in the emissions inventory, meteorological model simulation, and air quality model are reviewed, and corrections and parameter tuning are performed until the air quality model simulation performance is reliable.

[0159] The number of regions in the region marker file prepared in step 7) represents the number of contributing regions for source apportionment in the 3D regional air quality model. The number of regions and the extent of each region are determined based on specific requirements.

[0160] The air quality model-based source apportionment methods (or modules) used in step 8) include ISAM in the CMAQ model, PSAT or OSAT in the CAMx model, and source apportionment methods and modules in other three-dimensional regional air quality models.

[0161] Step 9) The conversion rules for converting observed components and their concentrations into air quality model chemical mechanism species and their concentrations depend on the chemical mechanism selected in the air quality model.

[0162] Step 10) assumes that the observed data are of reliable quality. The difference between the observed concentration of a particular pollutant and the simulated concentration of that pollutant based on the air quality model is considered the simulation bias. Since the simulated concentration of a pollutant is equal to the sum of the concentrations contributed by each source, the simulation bias is also assumed to be the sum of the deviations from the concentrations contributed by each source.

[0163] The source contributions in step 10) include source category contributions, regional contributions, and source category contributions in a specific region. That is, the specific analysis of which source contributions are made is artificially defined according to needs.

[0164] The nonlinear optimization algorithm used in step 10) includes all reasonable algorithms that can achieve the purpose of this solution.

[0165] In step 10), the correction coefficients for the contributions of multiple components of a pollutant from the same source (source class, region, source class within a region, etc.) can be different. This includes assuming that the same correction coefficient is used for all components of the receptor site pollutant associated with it, and also assuming that the same source uses different correction coefficients for different components of the receptor site pollutant associated with it.

[0166] The source apportionment results obtained in step 10) are based on the source apportionment method after coupling the receptor model with the three-dimensional regional air quality model. They have a higher degree of refinement and reliability, and provide important scientific and technological support for the precise and scientific prevention and control of urban or regional air pollution, which helps to improve the air quality of cities or regions.

[0167] Step 11) involves comparing the source apportionment results from the coupled two models with those based solely on the receptor model to assess the extent to which the coupled approach improves the results. However, because source apportionment based solely on the receptor model cannot identify regional contributions or contributions from more refined source categories, and cannot account for the influence of multiple meteorological factors or the multiphase chemical mechanisms that reflect complex physical and chemical processes in the atmosphere, the results of source apportionment based solely on the receptor model can be significantly biased.

[0168] Example

[0169] Establishment of a scheme system for source apportionment of atmospheric VOCs in Shanghai Dianshan Lake for typical months in 2018 based on the coupling of a three-dimensional regional air quality model and a receptor model

[0170] (1) Selection of research area and preparation of data

[0171] The three-layer simulation area settings of the three-dimensional regional air quality model system (including meteorological model and air quality model) are as follows: Figure 4 As shown, the first-layer simulation area covers China; the second-layer covers the Beijing-Tianjin-Hebei region, the Yangtze River Delta, and the Pearl River Delta; and the third-layer covers the Yangtze River Delta. The horizontal grid resolutions for the three simulation layers are 36 km × 36 km, 12 km × 12 km, and 4 km × 4 km, respectively, all using the Lambert map projection. The horizontal grid sizes for the three simulation layers of the WRF meteorological model are 185 × 148, 147 × 240, and 204 × 228, respectively. The grid system of the three-dimensional regional air quality model is the same as that of WRF, except that each region is slightly narrowed inward by several columns. Vertically, the model is divided into 23 layers. The WRF model input data include meteorological reanalysis data (FNL) from the National Centers for Environmental Prediction (NCEP), land use data, and surface and upper-air observations for assimilation. The WRF model parameterization schemes include the Yonsei University (YSU) scheme, the Noah land surface scheme, the Purdue-Lin microphysics scheme, the Goddard shortwave radiation scheme, and the RRTM longwave radiation scheme. The three-dimensional regional air quality model CMAQv5.3 was used, with AERO7i and SAPRC07TIC selected for the aerosol and gas-phase mechanisms, respectively. Simulations were performed with a 14-day spin-up period to eliminate the effects of initial conditions.

[0172] The emission inventory data required for the air quality model are divided into anthropogenic and natural source emission inventories. The anthropogenic emission inventories for the first and second layer simulation areas are derived from the MEIC inventory established by Tsinghua University, while the anthropogenic emission inventory data for the third layer simulation area are derived from the Yangtze River Delta region emission inventory with a higher spatial resolution established by our team. The natural source emission inventory data are generated by using meteorological data and land use data simulated by WRF and simulated by the MEGAN model. In addition, the data required by the MEGAN model also include data such as the leaf area index (LAI) from MODIS (Moderate Resolution Imaging Spectroradiometer).

[0173] Meteorological observation data and pollutant concentration data used for the air quality model system evaluation were obtained from data published by the China Meteorological Administration and the China National Environmental Monitoring Center, respectively. The VOC and its component concentration data used for source apportionment at Dianshan Lake Station were obtained from hourly-resolution online observations of 55 ozone precursor VOC species by the Shanghai Environmental Monitoring Center at Dianshan Lake Station. These included 27 C2-C6 species and 28 C7-C12 species.

[0174] (2) Methods used in this implementation case

[0175] 1) WRF weather model

[0176] The WRF model is a numerical model used for weather forecasting and atmospheric simulation. This implementation case first uses the WPS (WRF Preprocessing System) tool to convert the format of the FNL meteorological data in the three-layer simulation area, interpolate horizontally and vertically onto the model grid, and generate initial and boundary condition data. Then, the preprocessed data is converted into the initial field and boundary conditions of the WRF model through the real.exe program and smoothed and adjusted. Next, the parameterization scheme is set, and the wrf.exe main program is used to numerically simulate the evolution of the atmospheric state in the three-layer simulation area over time. After the wrf.exe main program simulation is completed, the WRF output results are post-processed using post-processing tools, such as converting the data into a visual or analytical format for further study or verification. After the meteorological field simulation is completed, its simulation performance is evaluated using meteorological observation data. If the performance is poor, the WRF simulation is further optimized by adjusting the parameters or parameterization scheme.

[0177] 2) Basic simulation of the CMAQ three-dimensional regional air quality model

[0178] The CMAQ model is a three-dimensional regional air quality model based on a "One Atmosphere" framework, encompassing all necessary atmospheric chemical and physical processes. First, emission inventory data for the three simulation regions are processed using the SMOKE model into the CMAQ model format. WRF simulated meteorological field files are processed into the CMAQ model format using the MCIP (Meteorology-Chemistry Interface Processor) module. Initial and boundary condition files required for air quality simulations are generated using programs such as ICON (Initial Conditions) and BCON (Boundary Conditions). Then, the CCTM module, the main CMAQ program, numerically simulates atmospheric chemical and physical processes in each simulation region. The simulation results for the outermost simulation region provide the boundary conditions for the innermost simulation region. Finally, after the CCTM simulation is completed for the innermost simulation region, CMAQ post-processing tools are used to analyze the output results, including concentration profiles, time series, and statistics. Following the baseline CMAQ simulation, the model's performance is comprehensively evaluated using observed pollutant concentration data to ensure its reliability.

[0179] 3) ISAM source apportionment method based on CMAQ model

[0180] The ISAM source apportionment method, based on the CMAQ model, is an advanced tool for identifying the sources of atmospheric pollutants. ISAM tracks pollutants by source category and region, and quantifies the contribution of each source (including region) to pollutant concentrations in any grid within the simulation area through chemical reactions, phase changes, physical transport, and diffusion. ISAM simulations are performed only on the innermost simulation area.

[0181] First, prepare the sub-emission inventory and regional marking files for the source categories of the third-level simulation area. The corresponding source categories of the sub-emission inventory include process, natural source, mobile source, residential source, power plant, industrial boiler and kiln, solvent use and oil and gas storage and transportation, waste treatment source, straw burning, agricultural source, a total of 10 sub-category emission inventory files. Each sub-emission inventory file corresponds to a source category. The regional marking file of this implementation case sets up the Yangtze River Delta region and the third-level (innermost) simulation area excluding the Yangtze River Delta region. There are a total of 11 sources for ISAM source analysis in this implementation case, namely: process, natural source, mobile source, residential source, power plant, industrial boiler and kiln, solvent use and oil and gas storage and transportation, waste treatment source, straw burning, agricultural source, and long-distance transmission in the Yangtze River Delta region. The long-distance transmission contribution is the contribution of all areas outside the Yangtze River Delta region.

[0182] Then, building on the reliable simulation performance of the CMAQ base model, using the sub-emission inventory and regional tag files, the ISAM simulation within the CMAQ model was performed on the third-level simulation area. The contribution concentrations and contribution rates of 11 sources of VOCs and their components were obtained for any grid within the third-level simulation area. The source apportionment results were then visualized and analyzed for their time series and spatial distribution.

[0183] 4) Due to the wide variety of VOCs and their complex reactions, air quality models employ simplified gas-phase chemical mechanisms to reduce computational costs. Some species are not defined as single compounds, but rather represent a class of chemically similar substances, known as lumped species. In other words, chemically similar components are combined into a single lumped species. For example, PAR species in the carbon bond chemical mechanism refer to components containing carbon-carbon single bonds; ALKn species in the SAPRC chemical mechanism refer to a class of alkanes and other non-aromatic hydrocarbons with kOH within a certain range. Due to instrumental limitations, not all components within these lumped species have observed concentration data. Therefore, to ensure comparability between observed concentrations and model-simulated concentrations, it is necessary to derive concentration conversion rules based on the definition of lumped species in the chemical mechanism and the names of VOC components with observed data. The gas-phase chemical mechanism used by the CMAQ air quality model in this implementation is SAPRC07. Some species in this chemical mechanism are explicit (defined compounds), while others are lumped. According to the SAPRC07 Chemical Mechanism Handbook (https: / / intra.engr.ucr.edu / ~Carter / SAPRC), the observed concentration of the lumped species in the chemical mechanism can be calculated by using the molar ratio of the components with observed data in the lumped species in the chemical mechanism and the observed concentration of these components with observed data. The specific conversion formula is as follows:

[0184]

[0185] Among them is the observed concentration of the chemical mechanism lumped species calculated by the above formula, which is converted based on the observed concentration of the observed VOCs component. It is the observed concentration of the observed part contained in the chemical mechanism lumped species, that is, the sum of the observed concentrations of the observed components contained in the chemical mechanism lumped species; it represents the sum of the molar proportions of the observed components contained in the chemical mechanism lumped species in the lumped species. From the definition of the chemical mechanism and the list of VOCs components with observed data in this implementation case, it can be seen that the sum of the molar proportions of the observed components in the chemical mechanism lumped species is fixed. The names and definitions of the lumped species and explicit species related to the gas phase chemical mechanism and VOCs in the air quality model are shown in Table 1, and the merged relationship of Lumped species is shown in Table 2.

[0186] Table 1 Chemical mechanism species names and definitions of VOCs in air quality models (SAPRC07)

[0187]

[0188]

[0189] Table 2 Lumped species components and their molar proportions (i.e., weighting factors) in the chemical mechanism

[0190]

[0191]

[0192] 5) Source apportionment method using the coupling of CMAQ model and receptor model

[0193] Through the simulation of the ISAM source apportionment module of the CMAQ model, the contribution concentrations of each species of VOCs at the receptor point (chemical mechanism species) from the 11 sources were obtained, and the sum of the contributions of the 11 sources to a certain species is the total simulated concentration of the species at the receptor point (chemical mechanism species), and the sum of the simulated concentrations of each species at the receptor point is the simulated concentration of the total VOCs species at the receptor point (the sum of the concentrations of VOCs species in the chemical mechanism). At the same time, based on the receptor model method, the observed concentration of VOCs species i at the receptor point is also the sum of the contribution concentrations of species i from each source to the receptor point. Therefore, the air quality model is coupled with the receptor model to form the following formula:

[0194]

[0195] Where i is the chemical mechanism species, including lumped species and explicit species; j represents the source, and 11 sources are set in this implementation case; represents the observed concentration of species i at the receptor site. If i is an explicit species, then is the directly observed concentration of species i at the receptor site. If i is a lumped species, then The observed concentration of lumped species i at the receptor site is obtained by converting it using formula (1); represents the contribution concentration of source j to species i at the receptor site obtained through source apportionment of the air quality model; It represents a residual term introduced to make the left and right sides of formula (2) equal.

[0196] The simulated concentration of each species at the receptor site is the sum of the simulated contribution concentrations of that species at the receptor site from the 11 sources. Due to the large number of VOC species, the least squares method is introduced here to minimize the sum of squared errors between the simulated and observed multi-species concentrations at the receptor site, thereby obtaining the optimal source apportionment results. This constructs the core equation for coupling the three-dimensional regional air quality model with the receptor model, as shown in the following formula (3):

[0197]

[0198] here, is the uncertainty of the observed concentration of species i; is a coefficient introduced, similar to the virtual source spectrum.

[0199] However, formula (3) loses the influence of the source contribution size, and there are also collinearity and uniqueness problems. For example, different sources use similar source spectra, which still affects the solution of the equations. Therefore, in order to obtain the minimum x 2 , we define an R j , a correction factor applied to the contribution of source j to correct the source contribution simulated by the air quality model:

[0200]

[0201] here, Refers to the source contribution concentration before correction, that is, the concentration obtained through source analysis simulation of air quality model The correction of source contribution concentration can be solved by a method similar to the traditional receptor model CMB method, that is, solving x 2 Minimized R j , as shown in the following formula (5):

[0202]

[0203] However, without further constraints, Rj may be physically unrealistic and fail to account for the uncertainty in source contributions from air quality model simulations.j It is reasonable in practical sense and can explain the uncertainty of source contributions obtained from air quality model simulation. Here, additional constraints and a penalty term are added to solve an optimized R j :

[0204]

[0205] here, represents the a priori uncertainty of all sources based on the air quality model for species i, and it can be assumed that the size of the a priori uncertainty is proportional to the observed concentration of species i, that is, it can be obtained by Obtain, where δ i is the normalized mean error (NME). The second term on the right side of equation (6) represents the uncertainty of the contribution of a single source due to emission errors. is the a priori uncertainty of the natural logarithm of the correction factor for source j. When performing the optimization calculation, we can initially set Where N is the number of species, J CTM Represents the number of VOCs contributing sources, so that the two terms on the right side of Equation (6) have equal weights. The nonlinear optimization problem is then solved through a multi-iteration optimization algorithm to minimize the objective function. The preset ranges of variation for each VOCs emission source correction factor (Rj) are shown in Table 3. The optimization algorithm used in this implementation case employs Sequential Quadratic Programming (SQP).

[0206] Table 3 Preset ranges and a priori uncertainties of the correction factors for each VOCs emission source

[0207]

[0208] (3) Implementation process of this implementation case

[0209] The implementation process of this implementation case is as follows Figure 3 shown.

[0210] First, set the three-layer simulation area and grid resolution of the three-dimensional regional air quality model (including meteorological model and air quality model), such as Figure 4The parameterization schemes for the meteorological model WRF and the air quality model CMAQ were selected. CMAQv5.3 was used as the three-dimensional regional air quality model, and SAPRC07TIC was selected as the gas phase mechanism. Then, WRF meteorological simulations were performed for typical months in 2018 using FNL data and land use data. The emission inventory data required for the CMAQ model were divided into sub-emission inventories for anthropogenic sources and each source category required for source apportionment. Using the WRF simulation results, the natural source emission inventories for the three simulation areas were simulated using the MEGANv3.1 model. At the same time, the simulation performance of the WRF model and the CMAQ model was evaluated using observational data, and the reliability of the model system was ensured through parameter optimization.

[0211] Then, if Figure 5 As shown in the figure, the hourly concentration observation data of VOCs and their components at the Dianshan Lake Station in Shanghai are obtained and cleaned to ensure the quality and reliability of the cleaned data. Prepare the regional tag file required for source analysis based on the air quality model. The three-dimensional regional air quality model can be used for the ISAM tool in CMAQ to simulate the source analysis of VOCs in the long-term region. Through simulation, the VOCs source analysis results of the grid where the Dianshan Lake Station is located are finally obtained, including the source category contribution, regional contribution and the contribution of different source categories in a certain area. Then, according to the observed VOCs component names at the Dianshan Lake Station and the VOCs related species names and definitions of the SPRAC chemical mechanism in the CMAQ model, the chemical mechanism species observation concentrations at the Dianshan Lake Station are converted.

[0212] Finally, by coupling the CMB receptor model with the CMAQ model, the deviation between the simulated and observed concentrations of VOC chemical mechanism species at the Dianshan Lake station was attributed to the sum of the deviations in the concentrations of each source contributing to that chemical mechanism species. Least squares optimization was introduced to construct an objective function that minimized the sum of squared errors. Sequential quadratic programming (SQP) was used to optimize this objective function (i.e., minimize the difference), ultimately calculating the correction coefficients for each source contribution. The source contribution concentrations simulated by the ISAM module based on the CMAQ model were multiplied by the corresponding source contribution correction coefficients to obtain the corrected source contribution concentrations to the receptor species. The sum of the corrected source contribution concentrations of each VOC mechanism species to the receptor is the total contribution concentration of VOCs (the sum of the mechanism species) to the Dianshan Lake station. This result is the source apportionment result of the atmospheric pollutant source apportionment method based on the coupling of the 3D regional air quality model and the receptor model. Finally, the degree of improvement in the source apportionment results after coupling the 3D regional air quality model with the receptor model was evaluated compared to the source apportionment results based solely on the 3D regional air quality model before coupling. The source apportionment results of the atmospheric pollutant source apportionment method based on the coupling of three-dimensional regional air quality models and receptor models can provide important scientific and technological support for action decisions to continuously improve the air quality of cities or regions.

[0213] (IV) Implementation results of this implementation case

[0214] This implementation case establishes a source parsing method based on the coupling of the air quality model and the receptor model, and solves Rj through optimization calculations, that is, the contribution of each type of source has a correction factor. Generally speaking, the uncertainty of regulated or fixed emission sources is lower, such as power plants, industrial boilers and furnaces, waste treatment sources, etc., and the uncertainty of random emission sources such as industrial sources, solvent use sources and biomass combustion is higher. Because Rj is an hourly change rate, it can reflect the hourly correction amplitude of the source contribution. Rj less than 1 means that the source contribution after correction is less than the source contribution before correction (indicating that the air quality model simulation before correction is too high). On the contrary, Rj greater than 1 means that the source contribution has increased compared to before correction. The average value of Rj for 11 VOCs contribution sources over 4 months ranged from 0.83 to 1.11, as shown in Table 4.

[0215] Table 4 Source contribution correction factors for ozone pollution at the Dianshan Lake site in typical months (May, June, July, and September) in 2018

[0216]

[0217] Typically, emission sources with higher uncertainty have a wider range of Rj and a relatively large standard deviation; whereas emission sources with lower uncertainty have Rj values that are generally concentrated near 1. This leads to differences in the Rj cumulative distribution functions for different emission sources. The Rj values for process emissions sources in the four months basically cover the entire range of variation (0.1 to 5), resulting in the largest standard deviation, ranging from 0.78 to 0.97. In addition, with the exception of September, the average Rj values for other months are all greater than 1, indicating that the contribution of process emissions sources is often underestimated.

[0218] The standard deviations of Rj for solvent use and motor vehicles range from 0.25 to 0.35, with variations from 0.9 to 1.91 and 0.1 to 1.51, respectively. In air quality models, the time distribution of these two emission sources is fixed or only includes simple weekend effects, while their emissions are highly stochastic, with daily variations in vehicle numbers and solvent use.

[0219] The average Rj values for natural sources are all less than 1, indicating that their contribution was generally high before correction. Their standard deviations are large, ranging from 0.22 to 0.28. This is primarily due to the fact that the predominant VOC species in natural sources is the highly reactive isoprene, which leads to large model simulation errors. The Rj values for ultra-long-range transport vary widely, ranging from 0.1 to 1.67, with an average deviation greater than 0.15. This is because ultra-long-range transport is significantly influenced not only by meteorological conditions but also by uncertainty in emission sources.

[0220] Other emission sources, such as oil and gas storage and transportation and residential sources, have relatively regular emission schedules, resulting in smaller standard deviations, with average deviations ranging from 0.08 to 0.20. Emissions from industrial boilers and kilns, power plants, and waste treatment sources are even more regular, with the smallest average deviations, ranging from 0.01 to 0.07, and a smaller range of Rj, ranging from 0.93 to 1.0.

[0221] It is important to note that in addition to the impact of emission source uncertainty on Rj, the contribution of the emission source itself is also an important factor. For example, emission sources with large average deviations, such as industrial processes and motor vehicles, contribute much more to VOCs than emission sources with smaller average deviations, such as power plants and industrial boilers.

[0222] Figure 6Time series plots of TVOC concentrations compared to observed concentrations for typical months in this implementation case are shown, before and after correction (source apportionment results based solely on the air quality model) and after correction (based on the new source apportionment method coupled with the air quality model and receptor model). The figure shows that compared to TVOC concentrations based solely on source apportionment using the air quality model, TVOC concentrations based on the new source apportionment method coupled with the air quality model and receptor model are significantly closer to the observed concentrations. This further demonstrates the significant superiority of the new source apportionment method based on the coupling of the air quality model and receptor model in source apportionment.

[0223] By constructing a source apportionment method that couples the air quality model CMAQ with the CMB receptor model, and based on this method, further corrections are made on the basis of the source contributions simulated by the CMAQ model, obtaining more reliable contributions of various sources under the constraints of VOCs component observation data, and further improving the accuracy of refined source apportionment of VOCs in the atmospheric environment, providing a scientific basis for action decisions on urban and regional air pollution prevention and control and continuous improvement of air quality, and promoting the scientific and precise development of urban and regional air pollution prevention and control and air quality management.

[0224] 1) Urban Air Quality Monitoring and Pollution Control Decision Support System

[0225] This method can be applied to urban air quality monitoring and pollution control systems. By coupling the three-dimensional regional air quality model with the receptor model, it is possible to accurately analyze the contribution of different urban pollution sources (such as industrial emissions, traffic exhaust, construction dust, etc.) to pollutants at a certain monitoring point. In the process of urban pollution control, decision makers can use the results of this source analysis to accurately determine which sources are the main pollution contributors, thereby formulating more targeted emission reduction measures. For example, in industrial areas or traffic-intensive areas, the main sources of PM2.5 or VOCs can be refined through model results, thereby optimizing the direction of policy implementation and improving governance effects.

[0226] 2) Compilation and optimization of regional air pollution source emission inventories

[0227] This method can also be used to compile and optimize the management of regional atmospheric pollution source emission inventories. For example, in the process of regional development planning, by coupling air quality models with receptor models, it is possible to simulate changes in pollution source contributions under different development scenarios and assess the impact of industrial zone expansion, road construction, etc. on air quality in advance. Through the results of source analysis, managers can optimize the layout and adjustment of atmospheric pollution sources in key areas to reduce the potential risk of air quality deterioration in the future. In the planning of industrial parks, especially in areas involving high-pollution industries such as chemicals and energy, the use of this source analysis method can effectively identify and optimize major emission sources and reduce negative impacts on the environment.

[0228] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware portion can be implemented using dedicated logic; the software portion can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art will appreciate that the above-mentioned devices and methods can be implemented using computer-executable instructions and / or contained in processor control code, for example, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0229] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with this technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered by the scope of protection of the present invention.

Claims

1. A pollutant source analysis method for coupling an air quality model with a receptor model, characterized in that: The following steps are involved: Step 1: Conduct a three-dimensional regional air quality model simulation based on the atmospheric pollutant emission inventory and the meteorological field simulated by the meteorological model; Step 2: After verification by observation data, the three-dimensional regional air quality model system including the meteorological model and the air quality model has reliable simulation performance; Step 3: Use the most advanced source apportionment method embedded in the 3D regional air quality model to simulate the source apportionment of atmospheric pollutants and obtain the source apportionment results; Step 4: Using the time series concentration observation data of atmospheric pollutants and their components at the observation point, and based on the definition of relevant species based on the chemical mechanism used in the 3D regional air quality model simulation, compare them with the observed components, establish a corresponding relationship between the chemical mechanism species and the observed components, and further establish a conversion relationship between their concentration data, thereby obtaining the time series observation concentrations of the chemical mechanism species at the observation site; Step 5: Couple the receptor model with the three-dimensional regional air quality model to construct an objective function and apply a nonlinear optimization algorithm to obtain the correction coefficient of each source's contribution to the receptor point pollutant under the constraints of the time series observation data. Step 6: Obtain source apportionment results based on the coupling of air quality model and receptor model.

2. The pollutant source analysis method for establishing a coupled air quality model and receptor model according to claim 1, characterized in that: Identify the single or multiple observation sites for source apportionment and the target pollutant. The pollutant should be composed of multiple components, and time series concentration observation data for the pollutant and its components should be available. Analyze the time series concentrations at the observation sites to understand the temporal evolution and distribution characteristics of the concentrations of pollutants and their components, and clean the data; Before conducting a three-dimensional regional air quality model simulation, it is necessary to set the simulation area and the grid resolution of each simulation area, and the innermost simulation area must cover the receptor points that require source analysis; Conduct simulations of baseline emission scenarios using a three-dimensional regional air quality model system. The model used is one of the internationally renowned chemical transport models, CMAQ. Prior to conducting simulations, the air quality model system must be configured, including the selection of parameterization schemes for various physical and chemical processes, the preparation of emission inventories for anthropogenic and natural sources, and other input data. The baseline scenario is then simulated based on the three-dimensional regional air quality model system. Observational data is then used to evaluate the model system's simulation performance. If performance is poor, error correction and parameter adjustments are performed to ensure the model system's reliability. Adopt source apportionment methods embedded in the most advanced three-dimensional regional air quality models, including ISAM in the CMAQ model, PSAT and OSA in the CAMx model, as well as new source apportionment methods that will be improved or developed in the future; Using the source apportionment method of the three-dimensional regional air quality model, we conduct source apportionment simulation for the target pollutants in the innermost simulation area and obtain the source apportionment results of the target pollutants for each grid in the entire simulation area. The target pollutants selected include PM 2.5 , PM 10 , VOCs are pollutants composed of multiple components; The chemical mechanisms in the adopted three-dimensional regional air quality model include species related to the components of the target pollutants, also known as chemical mechanism species; The chemical mechanism species in the air quality model may not be completely consistent with the components observed at the observation station. Therefore, it is necessary to construct a corresponding relationship between the two based on the definition of the species in the chemical mechanism and the specific components observed at the observation station, including the molar ratio relationship, and further establish a concentration conversion relationship between the two. Through this process, we can obtain (1) the observed concentration of the chemical mechanism species at the observation station, or (2) the simulated concentration of the observed component at the observation station. Only one of them is required. The purpose of obtaining (1) or (2) is to establish comparability between the observed concentration and the model simulated concentration, thereby laying the foundation for establishing a new source attribution method for coupling air quality models and receptor models, and ultimately achieving more accurate and refined source attribution.

3. The pollutant source analysis method for establishing a coupled air quality model and receptor model according to claim 1, characterized in that: Construct the objective function (Formulas A, B, and C) for coupling the air quality model and the receptor model; here, is the uncertainty of the observed concentration of species i; is the source apportionment result obtained based on the three-dimensional regional air quality model, that is, the contribution concentration of source j to species i at the receptor point (grid); is a coefficient introduced, similar to the virtual source spectrum; In order to reflect the influence of source contribution, avoid collinearity problems, and obtain the minimum x 2 , we define an R j , a correction factor applied to the contribution of source j to correct the source contribution simulated by the air quality model As shown in formula (B); Further for R j It is reasonable in practical sense and can explain the uncertainty of source contributions obtained from air quality model simulation. Here, additional constraints and a penalty term are added to solve an optimized R j , as shown in formula (C) here, represents the a priori uncertainty of all sources based on the air quality model for species i, and it can be assumed that the size of the a priori uncertainty is proportional to the observed concentration of species i, that is, it can be obtained by Obtain, where δ i is the normalized mean error (NME); the second term on the right side of formula (C) represents the uncertainty of the contribution of a single source due to emission errors; is the a priori uncertainty of the natural logarithm of the correction factor of source j; when performing the optimization calculation, we can initially set Where N is the number of species, J CTM Represents the number of contributing sources, so that the two terms on the right side of formula (C) have the same weight. Then, the nonlinear optimization problem is solved by a multi-iteration optimization algorithm to minimize the objective function and obtain the correction factor (Rj) of each source contribution. Then, the contribution concentration of each component (species) of the target pollutant at the receptor point (observation point) and the contribution concentration of each source (including different source categories, different areas, and different source categories in different specific areas) to the target pollutant at the receptor point (observation point) are obtained.

4. The pollutant source analysis method for establishing a coupled air quality model and receptor model according to claim 3, characterized in that: The optimization of the objective function (Formula C) involves minimizing the objective function through various optimization algorithms that can be used for constrained nonlinear optimization problems.

5. The pollutant source analysis method for establishing a coupled air quality model and receptor model according to claim 3, characterized in that: This includes using the correction factor (Rj) contributed by each source and the source apportionment results of the air quality model to calculate the contribution concentration of each component (or species) of the target pollutant at the receptor point (observation point) of each source (including different source categories, different regions, and different source categories in different specific regions) based on the new source apportionment method coupled with the air quality model and the receptor model according to the following formula (D), and then summing the contribution concentrations of each component (or species) of the same source to the receptor point to obtain the contribution concentration of the target pollutant of each source to the receptor; 6. The pollutant source analysis method for establishing a coupled air quality model and receptor model according to claim 1, characterized in that: The source apportionment results obtained by the new source apportionment method based on the coupling of a three-dimensional regional air quality model and a receptor model were compared with those based only on the three-dimensional regional air quality model and only on the receptor model to evaluate the superiority of the new method. As shown in formulas (E-J), the closer R and IOA are to 1, the better the source apportionment results based on the new method are, and the smaller the other statistical parameters are, the better the results are. Compared with the source apportionment results based only on the receptor model, the comparison can be made from the perspectives of the number of sources, degree of refinement, regional contribution, uncertainty, etc.

7. A pollutant source attribution system for coupling an air quality model with a receptor model, which implements the pollutant source attribution method for coupling an air quality model with a receptor model as described in any one of claims 1 to 6, characterized in that: The pollutant source analysis system for establishing the coupling of the air quality model and the receptor model includes: The simulation module is used to simulate the three-dimensional regional air quality model based on the atmospheric pollutant emission inventory and the meteorological field simulated by the meteorological model; A verification module is used to verify that the three-dimensional regional air quality model system, including the meteorological model and the air quality model, has reliable simulation performance through observation data verification; Source apportionment module, which uses the most advanced source apportionment methods embedded in the three-dimensional regional air quality model to simulate the source apportionment of atmospheric pollutants and obtain source apportionment results; The observation module is used to use the time series concentration observation data of atmospheric pollutants and their components at the observation point, and based on the definition of relevant species of the chemical mechanism adopted in the three-dimensional regional air quality model simulation, compare it with the observed components, establish the corresponding relationship between the chemical mechanism species and the observed components, and further establish the conversion relationship between their concentration data, so as to obtain the time series observation concentration of the chemical mechanism species at the observation site; The correction module is used to couple the receptor model with the three-dimensional regional air quality model, construct the objective function, and apply the nonlinear optimization algorithm to obtain the correction coefficient of each source's contribution to the receptor point pollutant under the constraints of the time series observation data.

8. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of establishing a pollutant source analysis method for coupling an air quality model and a receptor model as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of establishing a pollutant source analysis method for coupling an air quality model and a receptor model as described in any one of claims 1 to 6.

10. An information data processing terminal, characterized in that: The information data processing terminal is used to implement the pollutant source analysis system that couples the air quality model and the receptor model as described in claim 7.

Images