Refining source analysis method and system for heavy metals settled in air in soil

Abstract

The invention relates to the technical field of particulate matter source analysis, and discloses a refined source analysis method and system.The method comprises the steps that a plurality of pollution sources and dust falling particulate matter sampling points are determined according to soil pollution distribution of a target area and an atmospheric pollution source survey result, by collecting single particle mass spectrum data of pollution sources, sampling points and receptors, constructing a mass spectrum library and combining an ART-2a algorithm and a deep learning model, the heavy metal pollution contribution proportion is accurately analyzed and traced to specific discharge ports of enterprises, and the problem that specific pollution enterprises and working sections cannot be positioned in the prior art is solved. According to the method, a pollution source and an environment receptor are sampled and analyzed by utilizing a single-particle mass spectrometry technology aiming at atmospheric dust fall particles in soil, and the heavy metal in the atmospheric dust fall particles in the soil is finely traced to specific enterprises and working sections by analyzing the heavy metal, so that a fine technical support is provided for soil heavy metal pollution management and control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of particulate matter source analysis, and particularly relates to a method and system for fine source analysis of atmospheric deposition heavy metals in soil. BACKGROUND

[0002] At present, the soil in China is mainly exceeded by inorganic pollutants, and the organic pollutants are exceeded secondly. Among the inorganic pollutants, the heavy metals are the most important exceeded category. Atmospheric input is still an important input source of regional scale soil heavy metals. However, the atmospheric deposition heavy metal tracing methods mainly include source list method, diffusion model method and receptor model method. Compared with the conventional analysis method, the on-line single particle aerosol mass spectrometer (Single Partical Aerosol Mass Spectrometer, SPAMS) can realize in-situ real-time sampling and analysis of single aerosol particles without pretreatment through a single device. The on-line single particle aerosol mass spectrometer can simultaneously obtain the particle size, mass spectrum characteristics, isotope characteristics and other information of a single particle, reflect the most original micro characteristics of the particle, and has the characteristics of rich chemical component information and high time resolution. In addition to single element components, the on-line single particle aerosol mass spectrometer can identify element-related ion clusters, and has more tracer components than conventional techniques. The massive single particle mass spectrum data obtained by the on-line single particle aerosol mass spectrometer can be used for fine clustering through machine learning, and the fine characteristics of different source particles can be deeply mined, so as to realize fine differentiation of different source particles.

[0003] Based on the above characteristics, the SPAMS technology has been widely used in the overall source research of fine particles in the atmosphere. In recent years, some research teams have constructed source spectrum characteristics, and analyzed environmental air mass spectrum data through correlation analysis and pre-trained machine learning models, so as to further realize fine source analysis of atmospheric particles. However, the correlation with soil dust is lacking, and enterprise-level tracing has not been realized. At present, the particle sources can only be divided into specific industries, and it is still impossible to locate specific pollution enterprises and sections. In addition, the collection, analysis and tracing methods of atmospheric and soil samples are quite different. In addition, Xie Ruige et al. analyzed the differences in lead isotope peak area ratios of different industries by using the single particle aerosol mass spectrometer, and determined the main sources of soil lead pollution (cement industry, ceramic industry, non-road mobile source, etc.). However, the analysis result of the method can only be fine to the specific source industry of the overall soil heavy metals, and cannot be further analyzed to the specific enterprise or section. SUMMARY

[0004] Therefore, the present application provides a method and system for fine source analysis of atmospheric deposition heavy metals in soil, so as to solve the problems of being unable to finely divide the atmospheric deposition contribution and analyze to the specific enterprise or section.

[0005] In a first aspect, the present invention provides a method for refined source apportionment of atmospheric deposition heavy metals in soil, the method comprising:

[0006] Based on the distribution of soil pollution and the results of the investigation of air pollution sources in the target area, multiple heavy metal pollution sources and multiple soil dust particulate matter sampling points were identified.

[0007] Obtain receptor mass spectrometry data of soil dust particles at each sampling point and source mass spectrometry data of each heavy metal pollution source;

[0008] Based on receptor mass spectrometry and source mass spectrometry data, a refined source apportionment of soil dust particles at each sampling point was performed to determine the contribution ratio of each heavy metal pollution source to the heavy metal pollution in soil dust particles at each sampling point.

[0009] The present invention provides a method for refined source apportionment of heavy metals in atmospheric deposition in soil. This method uses single-particle mass spectrometry to sample and analyze pollution sources and environmental receptors for atmospheric dust particles in soil. Through the analysis of heavy metals, it enables refined source tracing of heavy metals in atmospheric deposition particles in soil to specific enterprises and processes, thereby providing refined guidance for pollution control.

[0010] In one optional implementation, receptor mass spectrometry data of soil dust particles at each sampling point are acquired, including:

[0011] If the sampling point cannot meet the detection conditions of the single-particle aerosol mass spectrometer, then gas samples and / or liquid samples and / or solid samples are obtained at the sampling point.

[0012] Mass spectrometry was performed on gaseous and / or liquid and / or solid samples to obtain receptor mass spectrometry data of soil dust particles at the sampling point.

[0013] In one optional implementation, acquiring receptor mass spectrometry data of soil dust particles at each sampling point further includes:

[0014] If the sampling point meets the detection conditions of the single-particle aerosol mass spectrometer, then place the single-particle aerosol mass spectrometer at the sampling point to obtain particulate matter in the atmospheric environment.

[0015] Mass spectrometry was used to detect particulate matter in the atmospheric environment to obtain receptor mass spectrometry data of soil dust particles at the sampling point.

[0016] The present invention provides a refined source apportionment method for atmospheric deposition heavy metals in soil. Different sampling and detection methods are selected according to the location characteristics of the sampling point, making sampling and detection more flexible and adaptable to different scenarios. The mass spectrometer is placed directly at the sampling point for sampling and detection, avoiding the influence of other factors during sample transportation. The detection accuracy and sensitivity are higher. By collecting particulate matter samples in three forms—gas, liquid, and solid—it meets the sampling needs of different environments.

[0017] In one optional implementation, source mass spectrometry data of each heavy metal pollution source are acquired, including:

[0018] Vacuum bottles were used to collect particulate matter from various heavy metal pollution sources.

[0019] Particles from each pollution source were passed through a single-particle aerosol mass spectrometer for detection, and the mass spectrometric characteristics of each pollution source particle were obtained as source mass spectrometry data for each heavy metal pollution source.

[0020] The present invention provides a refined source apportionment method for heavy metals in atmospheric deposition in soil. By collecting pollutant particles from various pollution sources and performing detection, the source mass spectrometry data of heavy metal pollution sources are obtained, which improves the authenticity of heavy metal detection data in pollution sources. Targeted sampling and detection from various heavy metal pollution sources is beneficial for refined analysis of heavy metal pollution sources, more accurately identifying pollution sources and carrying out targeted treatment.

[0021] In one optional implementation, the heavy metal pollution source includes: multiple pollution outlets of at least one enterprise emitting heavy metals in the target area; and refined source apportionment of soil dust particles at each sampling point based on receptor mass spectrometry data and source mass spectrometry data, including:

[0022] A mass spectrometry library was constructed using source mass spectrometry data. The mass spectrometry library includes multiple source spectra, each of which represents the industry to which the heavy metal pollution source in the target area belongs.

[0023] The receptor mass spectrometry data and the source spectra in the mass spectrometry library are calculated by dot product based on the ART-2a algorithm, and the industry to which the heavy metal pollution source corresponding to the receptor mass spectrometry data belongs is determined based on the principle of maximum similarity.

[0024] By classifying receptor mass spectrometry data from the same industry using a source apportionment model, the contribution percentage of heavy metal emissions from enterprises corresponding to each heavy metal in the receptor mass spectrometry data can be obtained.

[0025] The present invention provides a refined source apportionment method for atmospheric deposition heavy metals in soil. It uses the similarity method to analyze receptor mass spectrometry data to determine the industry to which the pollution source belongs. Even with limited data, it can still ensure the accuracy of the analysis, providing a reliable basis for scientifically assessing soil pollution in different regions and for refined management.

[0026] In one optional implementation, receptor mass spectrometry data from the same industry are classified using a source apportionment model to obtain the contribution percentage of each heavy metal emission outlet of the enterprise corresponding to each heavy metal in the receptor mass spectrometry data, including:

[0027] Based on the source mass spectrometry data of particles from each pollution source, the corresponding heavy metal types are analyzed, and based on the heavy metal types and corresponding heavy metal pollution sources, a pre-set deep learning model is trained to obtain the source resolution model.

[0028] Based on the source apportionment model, the pollution discharge outlets corresponding to the source mass spectrometry data of each heavy metal pollution source are determined;

[0029] The receptor mass spectrometry data is input into the source resolution model to obtain the heavy metal type corresponding to the receptor mass spectrometry data;

[0030] Based on the heavy metal categories in the receptor mass spectrometry data and the heavy metal categories at each pollution discharge outlet, the sources of heavy metals in the soil dust particles at the sampling points were determined.

[0031] The present invention provides a refined source apportionment method for atmospheric deposition heavy metals in soil. It utilizes deep learning algorithms to extract deep features from large amounts of complex data, accurately identify features related to pollution sources, and fit receptor mass spectrometry data and source mass spectrometry data based on these features to achieve high-precision pollution source analysis. The model parameters can be adjusted according to the specific situation of the pollution source, and it has good adaptability and generalization ability.

[0032] Secondly, this invention provides a refined source apportionment system for atmospheric deposition heavy metals in soil, the system comprising:

[0033] The pollution source and sampling point determination module is used to determine multiple heavy metal pollution sources and multiple soil dust particulate matter sampling points based on the soil pollution distribution and air pollution source survey results of the target area.

[0034] The mass spectrometry data analysis module is used to acquire receptor mass spectrometry data of soil dust particles at each sampling point and source mass spectrometry data of each heavy metal pollution source.

[0035] The source apportionment module is used to perform refined source apportionment of soil dust particles at each sampling point based on receptor mass spectrometry data and source mass spectrometry data, and to determine the contribution ratio of each heavy metal pollution source to the heavy metal pollution in the soil dust particles at each sampling point.

[0036] Thirdly, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the method described in the first aspect or any corresponding embodiment thereof.

[0037] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the method described in the first aspect or any corresponding embodiment thereof.

[0038] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to perform the method described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0039] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0040] Figure 1 This is a flowchart illustrating the method for refined source apportionment of atmospheric deposition heavy metals in soil according to an embodiment of the present invention.

[0041] Figure 2 This is a flowchart illustrating another method for refined source apportionment of atmospheric deposition heavy metals in soil according to an embodiment of the present invention.

[0042] Figure 3 This is a schematic diagram of the vacuum bottle sampling device in the refined source apportionment method for atmospheric deposition heavy metals in soil according to an embodiment of the present invention;

[0043] Figure 4 This is a schematic diagram of the structure used in the refined source apportionment method for atmospheric deposition heavy metals in soil according to an embodiment of the present invention to detect gas samples, liquid samples and solid samples respectively using SPAMS;

[0044] Figure 5 This is a schematic diagram of model training based on a deep learning algorithm in the refined source apportionment method for atmospheric deposition heavy metals in soil according to an embodiment of the present invention.

[0045] Figure 6 This is a specific embodiment of the refined source apportionment method for atmospheric deposition heavy metals in soil according to an embodiment of the present invention, showing the mass spectra of heavy metals from each enterprise in all detected particles;

[0046] Figure 7 This is a mass spectrum of four discharge outlets of a smelter in a specific embodiment of the method for refined source apportionment of atmospheric deposition heavy metals in soil according to an embodiment of the present invention.

[0047] Figure 8 This is a mass spectrum of two discharge outlets of a steel plant in a specific embodiment of the refined source apportionment method for atmospheric deposition heavy metals in soil according to an embodiment of the present invention.

[0048] Figure 9This is a schematic diagram of the source spectrum characteristics of Pb, Zn, and Cu in different enterprises in a specific embodiment of the refined source apportionment method for atmospheric deposition heavy metals in soil according to an embodiment of the present invention.

[0049] Figure 10 This is a structural block diagram of a refined source apportionment system for atmospheric deposition heavy metals in soil according to an embodiment of the present invention;

[0050] Figure 11 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Agricultural land soil environmental management is an important prerequisite for ensuring food security and sustainable agricultural development. Strengthening the prevention and control of agricultural land soil pollution is related to the quality and safety of national agricultural products and the health of the people, and has very important practical significance.

[0053] Methods for tracing the sources of heavy metals in atmospheric deposition mainly include source inventory methods, diffusion model methods, and receptor model methods. Source inventory methods estimate emissions from different sources by investigating and statistically analyzing emission factors and activity levels, and then identify major emission sources contributing to receptors based on these emissions. While this method yields simple and clear results, it suffers from problems such as high uncertainty in emission factors, lack of data on human pollution activity levels, and difficulty in accurately calculating emissions from various sources. Diffusion model methods focus on the possible sources of a pollutant within a specific area. Based on the emission volume and mode of pollutants from different sources, as well as the natural and social environmental characteristics of the region, mathematical models calculate the impact of different sources on soil pollution in that area. However, due to the significant uncertainties in the required source emission inventory and various environmental parameters, and the complexity and difficulty in accurately describing the transport and accumulation processes of heavy metals, the relationship between pollution sources and receptors is difficult to establish, greatly limiting the application of this method. Receptor modeling methods do not require consideration of pollution emission inventories and pollutant transport and accumulation processes. They directly target receptor samples for measurement, analysis, and source apportionment, making them the primary and commonly used source apportionment methods. These methods primarily include chemical mass balance (CMB) models, positive definite matrix factorization (PMF) methods, absolute factor analysis-multiple linear regression, and isotope tracing. However, existing receptor models often use data based on overall pollutant concentrations obtained through traditional monitoring methods. Limited by the granularity of receptor and source spectrum data, they generally suffer from insufficient granularity in source apportionment results, making it difficult to pinpoint specific polluting enterprises or processes. Consequently, they fail to provide further effective support for environmental management, necessitating the development of new methods for refined source apportionment of heavy metals.

[0054] Existing traditional heavy metal source apportionment techniques, such as source inventory methods and diffusion models, suffer from significant uncertainties in emission factors, a lack of data on human pollution activity levels, and difficulties in accurately calculating emissions from various sources. These factors lead to substantial uncertainty in source apportionment. Furthermore, the most commonly used receptor model is limited by insufficient refinement of source spectral features, hindering the achievement of refined heavy metal source apportionment. Single-particle aerosol mass spectrometry (SPA) methods can acquire refined characteristics of individual particles, possessing the potential for refined source apportionment. However, current technologies only address the source apportionment of specific particulate matter in the atmosphere or soil for large-scale applications, failing to address the refined source apportionment of heavy metals from atmospheric dustfall in soil.

[0055] To address the aforementioned issues, this invention provides a method and system for refined source apportionment of heavy metals from atmospheric deposition in soil. By sampling and analyzing pollution sources and environmental receptors, a refined source apportionment model for heavy metals is constructed, thereby achieving the effect of tracing the source of heavy metals from atmospheric deposition in soil to specific enterprises and processes.

[0056] According to an embodiment of the present invention, a method for refined source apportionment of atmospheric deposition heavy metals in soil is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0057] This embodiment provides a refined source apportionment method for atmospheric deposition heavy metals in soil, which can be used in the aforementioned computer system. Figure 1 This is a flowchart of a method for refined source apportionment of atmospheric deposition heavy metals in soil according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps:

[0058] Step S101: Based on the soil pollution distribution and air pollution source investigation results of the target area, identify multiple heavy metal pollution sources and multiple soil dust particle sampling points.

[0059] Specifically, by monitoring the local soil pollution distribution and air pollution source investigation reports of the target city, key industries and typical polluting enterprises are screened out. The basic principle of screening is to refer to the pollutant emissions of various industries in the local area, with a focus on emitting enterprises and emission outlets with large emissions of target heavy metals. The source spectrum sampling range must cover all suspected sources. In this embodiment, the heavy metal pollution source is the specific emission outlet of a specific enterprise. Sampling points for atmospheric deposition and surface soil are scientifically set up for farmland. Soil dust particle sampling points can be selected according to the location of pollution sources and key farmland locations, such as a certain farmland, a village street, or the roof of a neighborhood committee building, etc. This is just an example and is not a limitation.

[0060] Step S102: Obtain receptor mass spectrometry data of soil dust particles at each sampling point and source mass spectrometry data of each heavy metal pollution source.

[0061] Specifically, in this embodiment, the heavy metal pollution sources are refined to specific discharge outlets or processes of specific enterprises. The exhaust gas discharged from the outlet can be directly collected at the specific discharge outlet location. The exhaust gas contains particulate matter from the pollution source, and mass spectrometry detection of the particulate matter from the pollution source is performed to obtain the source mass spectrometry data of each heavy metal pollution source.

[0062] At each sampling point, water samples can be collected using sedimentation tanks, long-term accumulated dust samples can be collected, and particulate matter samples from the atmospheric environment can be collected to obtain receptor samples of soil dust particles. For the detection of receptor samples, an appropriate mass spectrometry detection method can be selected according to the state of the receptor sample to obtain receptor mass spectrometry data. This is only an example and is not a limitation.

[0063] Step S103: Based on receptor mass spectrometry data and source mass spectrometry data, perform refined source apportionment of soil dust particles at each sampling point to determine the contribution ratio of each heavy metal pollution source to heavy metal pollution in soil dust particles at each sampling point.

[0064] Specifically, heavy metal particles were extracted from both receptor mass spectrometry and source mass spectrometry data. Metal-containing particles in the contaminated soil samples were extracted using isotope peaks or oxide peaks of various heavy metals. The main heavy metals and their screening conditions are shown in Table 1.

[0065] Table 1

[0066]

[0067]

[0068] Based on the analysis of heavy metal particles in source mass spectrometry data, the heavy metal analysis results in each heavy metal pollution source can be determined. Based on the analysis of heavy metal particles in acceptor mass spectrometry data, the heavy metal analysis results in soil dust particles at each sampling point can be determined. Combining the heavy metal analysis results, similarity methods or deep learning algorithms can be used for analysis to determine the contribution percentage of each heavy metal pollution source to the heavy metal pollution in soil dust particles at each sampling point. For example, in the acceptor mass spectrometry data of soil dust particles at a certain sampling point, the sources of copper metal content include: 40% from the first discharge outlet of the first company, 20% from the second discharge outlet of the second company, 25% from the third discharge outlet of the second company, and 15% from the fourth discharge outlet of the third company. This is just an example and is not a limitation.

[0069] The method for refined source apportionment of heavy metals in atmospheric deposition in soil provided in this embodiment uses single-particle mass spectrometry to sample and analyze pollution sources and environmental receptors for atmospheric dust particles in soil. Through the analysis of heavy metals, the method can achieve refined source tracing of heavy metals in atmospheric deposition particles in soil to specific enterprises and processes, thereby providing refined guidance for pollution control.

[0070] This embodiment provides a refined source apportionment method for atmospheric deposition heavy metals in soil, which can be used on the aforementioned mobile terminals, such as mobile phones and tablets. Figure 2 This is a flowchart of a method for refined source apportionment of atmospheric deposition heavy metals in soil according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps:

[0071] Step S201: Based on the soil pollution distribution and air pollution source survey results of the target area, identify multiple heavy metal pollution sources and multiple soil dust particle sampling points. For details, please refer to [link to relevant documentation]. Figure 1 Step S101 of the illustrated embodiment will not be described again here.

[0072] Step S202: Obtain receptor mass spectrometry data of soil dust particles at each sampling point and source mass spectrometry data of each heavy metal pollution source.

[0073] Specifically, in step S202 above, obtaining receptor mass spectrometry data of soil dust particles at each sampling point includes:

[0074] Step S2021: If the sampling point cannot meet the detection conditions of the single-particle aerosol mass spectrometer, then obtain gas samples and / or liquid samples and / or solid samples at the sampling point.

[0075] Specifically, single-particle aerosol mass spectrometers have a certain volume and generally require power to operate. However, some sampling points, due to space, location, or other reasons, cannot accommodate a single-particle aerosol mass spectrometer or cannot be powered, thus failing to meet the detection requirements. In such cases, it is necessary to collect soil dust particles from the sampling point separately. The collected samples can be one or more of aerosol, liquid, or solid samples, depending on the characteristics of the sampling point or actual needs. For aerosol samples, vacuum bottles can be used for sampling, such as... Figure 3 The diagram shows the structure of the vacuum bottle sampling device. Before sampling, the clean 3L vacuum bottle is pre-evacuated to 0.1 MPa. Upon arrival at the sampling site, the piston is opened to begin sampling. When the pressure gauge reading reaches zero (equal to atmospheric pressure), the piston is closed, and the aerosol sample collection is complete. For liquid samples, atmospheric deposition sampling points are set up by monitoring the local soil pollution distribution in the target city and selecting locations with heavy soil heavy metal pollution or other points of interest. Water samples are collected from the deposition tank, impurities are removed, and after mixing, at least 50 ml of sample is taken. For solid samples, depending on the sampling location, including soil dust, dust from pollution sources, and dust from enterprise dust removal facilities, we selected farmland contaminated with metal and divided it into three areas according to the degree of contamination. We collected three soil dust samples from each area, with 500g of each sample collected. Sampling points were set up in different functional areas. At the sampling points, dust accumulated over a long period of time on dust-carrying platforms such as windowsills, shop windows, and shelves at different heights in buildings, warehouses, and shops was brushed into plastic bags, with each bag weighing about 500g. When bagging, we carefully removed large clods of soil, branches, leaves, grass, paint powder, wall paint, and other debris.

[0076] It should be noted that, in order to minimize the loss of particulate matter after collection, each sample must be sent to the laboratory for testing within a specified time: less than 4 hours for pollution source samples, less than 72 hours for sedimentation tank samples, and less than 1 week for soil dust and pollution source dustfall.

[0077] Step S2022: Mass spectrometry is performed on gaseous and / or liquid and / or solid samples to obtain receptor mass spectrometry data of soil dust particles at the sampling point.

[0078] Specifically, before sample injection, check that the instrument's injection pressure and mass drift range are within the allowable range. If they exceed the instrument's quality control range, perform instrument calibration to ensure the accuracy of mass spectrometry detection. Different mass spectrometry detection methods are used for different sample forms, such as... Figure 4 The diagram shows the structure for detecting gas, liquid, and solid samples using SPAMS. For gas samples, they are directly connected to the SPAMS instrument via a conductive silicone tube without any pretreatment. If the gas sample has high humidity, a drying tube is connected during injection. After the sample gas passes through the drying tube, it enters the SPAMS mass spectrometry system for detection, thereby obtaining the particulate mass spectral characteristics of the gas sample.

[0079] For liquid samples, an appropriate amount of liquid sample is placed in an aerosol generator. High-purity nitrogen gas with a stable flow is input by a pump and sprayed onto the surface of the liquid to be tested through a nozzle to form an aerosol. After the aerosol sample is dried by a drying tube, it enters a single-particle aerosol mass spectrometer for detection to obtain the particulate mass spectral characteristics of the liquid sample.

[0080] For solid samples, after drying and sieving, they are placed in a simple resuspension device. The collected dust samples are resuspended by nitrogen or clean air to remove particulate matter, and then input into a single-particle aerosol mass spectrometer for detection to obtain the particulate mass spectral characteristics of the solid samples.

[0081] Particulate matter spectral characteristics from gaseous and / or liquid and / or solid samples are used as receptor mass spectrometry data. If a sample has more than one morphology, the particulate matter spectral characteristics from multiple samples can be comprehensively analyzed to obtain the final receptor mass spectrometry data. Source apportionment analysis is performed on the receptor mass spectrometry information of different morphologies around the sampling point. For example, meteorological data can be combined, with the source apportionment results of aerosol samples as the main component, combined with the sampling location and source apportionment results of solid and liquid samples, to comprehensively analyze the source of heavy metal pollution in the soil. Aerosol samples mainly reflect the real-time changes of the source during the sampling period, while dust samples mainly reflect the long-term cumulative effect.

[0082] In some optional implementations, acquiring receptor mass spectrometry data of soil dust particles at each sampling point further includes:

[0083] If the sampling point meets the detection conditions of the single-particle aerosol mass spectrometer, then place the single-particle aerosol mass spectrometer at the sampling point to obtain particulate matter in the atmospheric environment.

[0084] Specifically, if the sampling point meets the detection conditions of a single-particle aerosol mass spectrometer, the SPAMS or SPAMS monitoring vehicle is placed at the target location to collect particulate matter in the atmospheric environment directly as the detection sample for the sampling point.

[0085] Mass spectrometry was used to detect particulate matter in the atmospheric environment to obtain receptor mass spectrometry data of soil dust particles at the sampling point.

[0086] Specifically, the basic principle of SPAMS analysis and detection is as follows: Aerosol particles pass through a 0.1 mm diameter sample inlet and an aerodynamic lens, and are focused into a straight-moving particle beam that enters the particle size detection zone. After entering the particle size detection zone, the flight time of the particles is recorded by two parallel laser beams (532 nm) spaced 6 cm apart and two PMTs, thereby calculating the aerodynamic diameter of the particles. The particle velocity can also be calculated in real time, indicating the time it takes for the particles to reach the ionization laser beam of the mass spectrometer. After entering the ionization zone, the particles are struck into positive and negative ion fragments by a triggered ionization laser (266 nm Nd:YAG ultraviolet pulsed laser). These ion fragments fly towards the poles of the mass spectrometer under the influence of an electric field, and are ultimately detected and converted into information about the chemical composition of the particles.

[0087] The refined source apportionment method for atmospheric deposition heavy metals in soil provided in this embodiment selects different sampling and detection methods according to the location characteristics of the sampling point, making sampling and detection more flexible and adaptable to different scenarios. The mass spectrometer is placed directly at the sampling point for sampling and detection, avoiding the influence of other factors during sample transportation, resulting in higher detection accuracy and sensitivity. By collecting particulate matter samples in three forms—gas, liquid, and solid—it meets the sampling needs of different environments.

[0088] Specifically, obtaining source mass spectrometry data for each heavy metal pollution source in step S202 above includes:

[0089] Step S2023: Collect pollutant particles from each heavy metal pollution source using a vacuum bottle.

[0090] Specifically, the vacuum bottle sampling method is selected for sampling exhaust gas from pollution sources. The vacuum bottle sampling device is as follows: Figure 3 As shown, before sampling, a clean 3L vacuum bottle is pre-evacuated to 0.1 MPa. Upon arrival at the sampling site (exhaust outlet), the vacuum bottle inlet is connected to the sampling gun. Once the sampling gun is inserted into the center of the exhaust chimney, the piston is opened for sampling. Sampling is completed when the pressure gauge reading reaches zero (equal to atmospheric pressure), and then the piston is closed. Vacuum bottle sampling is fast but not easily carried in large quantities. To ensure sample reliability, the following principles should be followed during sampling: select a location after a combustion boiler or process waste gas dust removal device; avoid locations with bends and changes in pipe diameter, and try to sample in straight flues; avoid sampling locations with high moisture content after wet desulfurization or water spray treatment of the waste gas; and try not to sample too close to the fan. Generally, three parallel samples should be collected from each sampling point to ensure data reliability and that the amount of data obtained meets the requirements for analysis and statistics. After sampling each pollution source, the sampling equipment should be thoroughly cleaned to avoid contaminating the next source sample.

[0091] Step S2024: Pass the particles from each pollution source into a single-particle aerosol mass spectrometer for detection to obtain the mass spectrometry characteristic information of each pollution source particle, which serves as the source mass spectrometry data of each heavy metal pollution source.

[0092] Specifically, after the sample gas containing pollutant particles is collected, it is directly connected to the SPAMS instrument for analysis using a conductive silicone tube without any treatment. Figure 4 As shown, if the sample gas has high humidity, a drying tube is connected during sample injection. After the sample gas passes through the drying tube, it enters the SPAMS mass spectrometry injection system for detection, obtaining the particulate mass spectrometry characteristics of the waste gas sample, which serves as the source mass spectrometry data for the corresponding heavy metal pollution source.

[0093] The method for refined source apportionment of atmospheric deposition heavy metals in soil provided in this embodiment obtains source mass spectrometry data of heavy metal pollution sources by collecting pollution source particles from various pollution sources and performing detection. This improves the authenticity of heavy metal detection data in pollution sources. Targeted sampling and detection from various heavy metal pollution sources is beneficial for refined analysis of heavy metal pollution sources, more accurately identifying pollution sources and carrying out targeted treatment.

[0094] Step S203: Based on receptor mass spectrometry data and source mass spectrometry data, perform refined source apportionment of soil dust particles at each sampling point to determine the contribution ratio of each heavy metal pollution source to heavy metal pollution in soil dust particles at each sampling point.

[0095] Specifically, heavy metal pollution sources include: multiple pollution outlets of at least one enterprise emitting heavy metals in the target area; and in step S203 above, based on receptor mass spectrometry data and source mass spectrometry data, a refined source apportionment of soil dust particles at each sampling point is performed, including:

[0096] Step S2031: Construct a mass spectrometry library using source mass spectrometry data. The mass spectrometry library includes multiple source spectra, each of which represents the industry to which the heavy metal pollution source in the target area belongs.

[0097] Specifically, a mass spectrometry library is constructed using source mass spectrometry data. The elements in the mass spectrometry library are source spectra. Each source spectrum represents the industry to which the heavy metal pollution source in the target area belongs, such as industrial sources, vehicle exhaust sources, etc. This is just an example, but not a limitation.

[0098] Step S2032: Perform dot product calculation between the receptor mass spectrometry data and the source spectra in the mass spectrometry library based on the ART-2a algorithm, and determine the industry to which the heavy metal pollution source corresponding to the receptor mass spectrometry data belongs based on the principle of maximum similarity.

[0099] Specifically, the ART-2a algorithm, as an unsupervised learning pattern recognition algorithm, can automatically classify data without prior labels. This algorithm achieves adaptive learning and memorization of data patterns by constructing a three-layer network structure comprising an input layer, a comparison layer, and a recognition layer. When matching receptor mass spectrometry data with source spectra in a mass spectrometry library, the ART-2a algorithm is used to perform dot product calculations. The dot product operation measures the similarity between two vectors (i.e., the receptor mass spectrometry data vector and the source spectrum vector); a larger dot product value indicates a higher similarity between the two vectors in the feature space.

[0100] The specific calculation process is as follows: The preprocessed receptor mass spectrometry data is used as the input vector, and a dot product is sequentially performed with each source spectrum vector in the mass spectrometry library. During the calculation, the ART-2a algorithm dynamically adjusts the weight parameters within the network based on the characteristics of the input data, ensuring that similar data patterns are grouped into the same category. Each dot product calculation yields a similarity score, reflecting the degree of matching between the receptor mass spectrometry data and the corresponding source spectrum.

[0101] The industry to which the heavy metal pollution source corresponding to the receptor mass spectrometry data belongs is determined based on the principle of maximum similarity. After calculating the dot product with all source spectra in the mass spectrometry library, all similarity values ​​are compared, and the industry corresponding to the source spectrum with the highest similarity is selected as the most likely industry to which the heavy metal pollution source of the receptor mass spectrometry data belongs. For example, if the dot product calculation of the receptor mass spectrometry data and the source spectrum of the mining and smelting industry in the mass spectrometry library yields the highest similarity value, then the heavy metal pollution source corresponding to that receptor mass spectrometry data is determined to most likely originate from the mining and smelting industry.

[0102] Step S2033: Classify the receptor mass spectrometry data of the same industry using the source apportionment model to obtain the contribution ratio of each heavy metal in the receptor mass spectrometry data to the heavy metal discharge outlets of the corresponding enterprises.

[0103] Specifically, in the target area, pollution sources in the same industry may include multiple discharge outlets of multiple enterprises. Therefore, it is necessary to further refine the classification of receptor mass spectrometry data of the same industry using source apportionment models to determine the contribution ratio of each heavy metal in the receptor mass spectrometry data to the heavy metal discharge outlets of the corresponding enterprises. For example, the receptor mass spectrometry data of soil dust particles at a certain sampling point, after source apportionment model analysis, shows that the sources of copper metal content include: the first discharge outlet of the first enterprise (40%), the second discharge outlet of the second enterprise (20%), the third discharge outlet of the second enterprise (25%), and the fourth discharge outlet of the third enterprise (15%). This is just an example, but it is not a limitation.

[0104] The method for refined source apportionment of atmospheric deposition heavy metals in soil provided in this embodiment uses the similarity method to analyze receptor mass spectrometry data to determine the industry to which the pollution source belongs. Even with limited data, the accuracy of the analysis can still be guaranteed, providing a reliable basis for scientifically assessing soil pollution in different regions and for refined management.

[0105] In some optional implementations, step S2033 above includes:

[0106] Step a1: Analyze the corresponding heavy metal types based on the source mass spectrometry data of particles from each pollution source, and train a preset deep learning model based on the heavy metal types and corresponding heavy metal pollution sources to obtain the source apportionment model.

[0107] Specifically, the source mass spectrometry data of particles from each pollution source are vectorized so that each mass spectrometry feature information forms a 500-dimensional vector.

[0108] Model training is performed using deep learning algorithms, such as Figure 5 As shown, the process is as follows:

[0109] Input → Calculate loss value → Adjust parameters → Calculate loss value again → Adjust parameters again → Repeat this process until the parameters no longer change or the loss value no longer decreases → End. Figure 5 In this model, x is the input, w is the weight matrix, b is the bias, f(z) is the activation function, f is the output value, y is the expected output value (the initial given label used to distinguish different source classes, for example, dividing exhaust particles into two categories: gasoline vehicles are labeled (10), and diesel vehicles are labeled (0 1)), and C is the loss function value. After iterative correction of w and b until the loss function value no longer decreases (or w and b no longer change), the model training ends. Output all parameter sets of the model, including w, b, and the parameters required in the model calculation process. Repeat the above source spectrum training process, with n iterations, and finally output n sets of parameters such as w and b.

[0110] Optimal Parameter Selection: Import standardized historical atmospheric particle vector data. For each standardized atmospheric particle vector, perform an x→z→f operation using the n sets of parameters output above (each set of parameters is calculated separately), outputting a k-dimensional vector, where k is the source class dimension. Each dimension constantly represents a source class. Find the dimension with the largest value in the k-dimensional vector. If this value is greater than a preset threshold, the atmospheric particle is classified into the source class represented by that dimension. After all atmospheric particle vectors have been calculated, a classification result can be obtained by counting, namely the number of gasoline vehicle particles and the number of diesel vehicle particles. Finally, after calculation using n sets of parameters, n classification results are obtained. Calculate the average proportion of each source class in the n results. Using the average as a benchmark, the bounding range is equal to the average multiplied by plus or minus x%. When x is the smallest and the proportion of all source classes falls within the bounding range, the parameter set corresponding to the result is the optimal parameter set.

[0111] A refined source resolution model is generated based on the optimal parameter set.

[0112] Step a2: Based on the source apportionment model, determine the pollution discharge outlets corresponding to the source mass spectrometry data of each heavy metal pollution source.

[0113] Specifically, based on the source apportionment model, the pollution discharge outlets corresponding to the source mass spectrometry data of each heavy metal pollution source can be determined.

[0114] Step a3: Input the receptor mass spectrometry data into the source resolution model to obtain the heavy metal type corresponding to the receptor mass spectrometry data.

[0115] Specifically, the receptor mass spectrometry data to determine the industry to which the pollution source belongs are vectorized, so that each receptor mass spectrometry data is composed of a 500-dimensional vector. The vector is then input into the source resolution model to perform x→z→f operations, and finally the classification result is obtained.

[0116] Step a4: Based on the heavy metal categories in the receptor mass spectrometry data and the heavy metal categories at each pollution discharge outlet, determine the source of heavy metals in the soil dust particles at the sampling point.

[0117] Specifically, particulate matter is categorized into industrial sources, vehicle exhaust, and other sources based on the similarity of characteristic ions in the actual sample spectrum or the source spectrum characteristics of heavy metal particles in the source spectrum library. Further, deep learning algorithms are used to further classify industrial sources into steel plant sources, power plant sources, etc., or into more refined source categories such as steel plant discharge outlet #1, steel plant discharge outlet #2, etc.

[0118] It should be noted that the mass spectrometry data of known heavy metal pollution sources can be randomly divided into two parts. One part is used as the source spectrum input to the source apportionment model, and the other half is used as the receptor input (equivalent to a blind sample) to the source apportionment model. Using the constructed source apportionment model, based on the input half of the source spectrum data, the other half of the "receptor" dataset is classified into sources. By comparing the proportion of source particles that are accurately classified in the "receptor" dataset, the classification accuracy of the source apportionment model can be verified.

[0119] The refined source apportionment method for atmospheric deposition heavy metals in soil provided in this embodiment uses deep learning algorithms to extract deep features from a large amount of complex data, accurately identify features related to pollution sources, and fit receptor mass spectrometry data and source mass spectrometry data based on these features to achieve high-precision pollution source analysis. The model parameters can be adjusted according to the specific situation of the pollution source, and it has good adaptability and generalization ability.

[0120] In one specific embodiment, for four consecutive days, emission characteristics of heavy metals from seven discharge outlets of three different types of enterprises with large discharge volumes were detected in a key soil pollution area in Guangdong Province. Additionally, dust samples from the rooftop of a village committee building, stairwell, and settling tanks near a high-value heavy metal concentration in the soil, as well as air samples from those areas, were collected over four consecutive days.

[0121] One sample was collected from each of the smelter, steel mill, and power plant. Based on the actual number of discharge outlets and operating conditions, four, two, and one discharge outlets were collected from the smelter, steel mill, and power plant, respectively. One dust sample was collected from a stairwell, one from a rooftop, and one from a sedimentation tank at a neighborhood committee office in a village near a soil heavy metal pollution site. Online monitoring of atmospheric aerosols was also conducted at the village committee office.

[0122] This embodiment takes heavy metals (Zn, Pb, Cu, Cd, Cr, As, Ni, Hg) as examples for research, and the screening conditions for each heavy metal are shown in Table 1.

[0123] The results of the analysis of heavy metal components in the samples are shown in Table 2. As can be seen from Table 2, heavy metals were detected in all samples, with significant differences between different types of samples. The proportion of heavy metal particles was significant, and Zn, Pb, and Cu were the main heavy metal types detected at the enterprise's discharge outlet.

[0124] Table 2

[0125]

[0126]

[0127] Based on the above results, analysis is conducted from different dimensions:

[0128] (1) Different Enterprises: As shown in Table 3, this represents the percentage of heavy metals in all detected particles from each enterprise.Figure 6 The image shows the mass spectra of heavy metals detected in all particles from various enterprises, indicating significant differences in source spectra among different types of enterprises. Therefore, single-particle data based on single-particle aerosol mass spectrometry can be used to distinguish between different types of enterprises. Smelters, steel mills, and power plants share the characteristic of containing heavy metals Zn, Pb, and Fe. The differences are that smelters contain heavy metals Li and Al; steel mills contain significant amounts of Cu and Pb; and power plants contain significant amounts of Cd and Al.

[0129] Table 3

[0130] Business Name Zn / Mass Pb / Mass Cu / Mass Cd / Mass Cr / Mass Ni / Mass As / Mass Smelter 31.9％ 34.6％ 6.2％ 2.0％ 1.7％ 0.1％ 0.0％ Steel Mill 18.6％ 15.8％ 0.9％ 13.4％ 2.2％ 0.0％ 0.0％ Power Plant 11.0％ 0.4％ 3.3％ 0.4％ 0.4％ 0.1％ 0.0％

[0131] (2) Different discharge outlets: As shown in Table 4, the metal emission characteristics of different discharge outlets in the smelter were compared. The results show that there are significant differences among the various discharge outlets of the smelter, such as... Figure 7 The image shows the mass spectra of four discharge outlets from the smelter. A common characteristic of the coexisting chemical components in the heavy metal-containing particles emitted from these four outlets is the presence of organic nitrogen and elemental carbon mass spectrum peaks. The differences are: outlets DA001 and DA018 show significant Ca and Al signals, while DA001 contains particles with a mass-to-charge ratio of 59. Discharge outlet DA024 emits heavy metal particles containing significant Pb and Cu signals. Discharge outlet DA031 emits heavy metal particles containing significant Zn and phosphate (PO3-) signals.

[0132] Table 4

[0133]

[0134]

[0135] like Figure 8 The image shows the mass spectra of two discharge outlets from a steel plant. Similar to the smelter, the heavy metal particles emitted from the two outlets also show significant differences. A common feature is the coexistence of Fe and EC characteristic peaks in the heavy metal particles. DA001 also contains the heavy metal Zn, and the negative spectrum shows S2-, AlO-, NaCl-, and silicates. DA003 also contains the heavy metals Cr, Cd, Cu, and Pb, and contains a component with a mass-to-charge ratio of 74.

[0136] (3) Differences between different heavy metal types at enterprises and discharge outlets: Further analysis was conducted on the differentiation of individual heavy metal categories between enterprises and discharge outlets for Pb, Zn, and Cu, for which there was relatively abundant valid data. For example... Figure 9The diagram shows the source spectral characteristics of Pb, Zn, and Cu from different companies. The mass spectra reveal the following: ① The average mass spectra of lead-containing particles from all three companies are accompanied by heavy metals calcium and iron. However, there are differences among the three companies; for example, the lead signal from the smelter is significantly higher than that from the steel mill and power plant. Lead-containing particles emitted by the smelter also contain zinc and copper, while those from the power plant contain cadmium. ② The average mass spectra of zinc-containing particles from all three companies contain heavy metals calcium and iron. Significant differences exist between the companies. The average mass spectra of lead-containing particles emitted by the smelter contain both lead and aluminum / lithium signals; zinc-containing particles emitted by the steel mill also contain silicates and phosphates; and zinc-containing particles emitted by the power plant contain aluminum, with a higher intensity than those from the smelter. ③ The average mass spectra of copper-containing particles from all three companies are accompanied by heavy metals calcium and iron. Significant differences exist between different enterprises. The average mass spectrum of copper-containing particles emitted by smelters also shows a clear signal of heavy metal lead, as well as aluminum, zinc, and copper. The phosphate signal from steel plants is more significant than that from the other two enterprises. Copper-containing particles emitted by power plants contain aluminum, and the intensity is higher than that from smelters.

[0137] Taking the similarity method as an example, similarity matching was used to analyze the sources of heavy metals. To verify the differences in source spectra and the accuracy of source apportionment results, 50% of the total heavy metal-containing particles were randomly selected as source spectra, and the other 50% were used as receptors from known sources for similarity matching to evaluate the accuracy of the matching. A similarity threshold of 0.7 was selected. The results showed that the accuracy of the heavy metal source apportionment results from enterprises was high, all greater than 80.0%. The accuracy of the heavy metal source apportionment results from most discharge outlets was high, greater than 75%. The reason why the accuracy of the steel A003 discharge outlet was much lower than that of other discharge outlets may be related to the small number of samples tested and poor representativeness. The accuracy of the source apportionment results for Pb, Zn, and Cu from enterprises was high, above 80.0%. The poor Pb source apportionment results from power plants were related to the small sample size and large uncertainty. Table 5 shows the accuracy verification results of total heavy metal particulate matter from different enterprises; Table 6 shows the accuracy verification results of total heavy metal particulate matter from different discharge outlets; and Table 7 shows the accuracy verification results of single heavy metal particulate matter from different enterprises. The evaluation results show that, based on single-particle mass spectrometry monitoring data, the similarity method can effectively distinguish heavy metals from different enterprises and different discharge outlets.

[0138] Table 5

[0139] Total Heavy Metals 50% Air Sample Match Value Accuracy % Smelter 1776 1654 93.1％ Steel Mill 201 181 89.6％ Power Plant 244 207 84.8％

[0140] Table 6

[0141] Total Heavy Metals 50% Air Sample Match Value Accuracy % Smelter DA001 Emission Port 23 20 87.0％ Smelter DA018 Emission Port 514 416 80.9％ Smelter DA024 Emission Port 872 816 93.6％ Smelter DA031 Emission Port 368 349 94.8％ Steel Mill DA001 Emission Port 196 179 91.3％ Steel Mill DA003 Emission Port 6 4 66.7％ Power Plant Unit 10 244 189 77.5％

[0142] Table 7

[0143]

[0144] The Adaptive Resonance Neural Network (ART-2a) algorithm was used to perform pairwise similarity analysis between the source spectra of heavy metal-containing particles and atmospheric heavy metal-containing particles. The specific calculation steps included:

[0145] Import source spectrum A (MASS1, MASS2, MASS3..., a matrix composed of all source samples of each type of pollution source) and source spectrum B (MASS, an air sample or a source sample of that type of pollution source);

[0146] The MASS of source spectrum A is subjected to Art-2a iterative classification to construct the source spectrum matrix;

[0147] The source spectrum B is classified using Art-2a iteration, and the similarity value between each class of PIDCell and the source spectrum A matrix is ​​calculated.

[0148] The average of the maximum similarity values ​​between them is used as the similarity value between them;

[0149] The threshold (usually 0.8) used in the actual matching of fine particles in the atmospheric environment using the similarity method is used as the evaluation threshold. It is believed that pollution source spectra with similarity greater than the threshold will interfere with the matching of pollution sources for fine particles in the atmospheric environment.

[0150] Calculation command:

[0151] [CelResultTable,CelCalTable,CelMASSTable]=get_similarity_with_interchoos e({MASS1,MASS2,MASS3},MASS,0.8,90)

[0152] Where 0.8 is the Art-2a classification threshold, and 90 is the number of particles extracted after Art-2a classification for subsequent calculations.

[0153] Table 8 shows the pollution source proportions matched for atmospheric heavy metals and various heavy metal particulate matter under different similarity threshold parameters: 0.5, 0.55, 0.6, 0.65, 0.70, 0.75, 0.8, 0.85, 0.9, and 0.95. The similarity matching process involves calculating the median proportion of each source based on the multi-threshold calculation results, further calculating the deviation between the source apportionment results and the median value under different similarity levels, and averaging the deviations of each source under each similarity level to obtain the average deviation under different similarity thresholds. The optimal similarity apportionment result is selected based on the principle of minimizing the deviation and the proportion of "other" sources. The similarity thresholds used for total heavy metals, Pb, Zn, Cu, and Cd in stair dust are 0.7, 0.7, 0.7, 0.55, and 0.7, respectively.

[0154] Table 8

[0155]

[0156] Calculations were performed to analyze the sources of heavy metals and particles containing Pb, Zn, Cu, or Cd in stairwell dust. The results showed that over 55.0% of the heavy metals Pb, Zn, Cu, and Cd in the receptor data originated from smelters; over 15.0% of the heavy metals Cu and Cd originated from power plants.

[0157] A deep learning neural network algorithm was used to perform a refined comparative analysis of the source spectra of heavy metal particulate matter from different discharge outlets and atmospheric heavy metal particulate matter. The specific calculation steps included:

[0158] Importing source spectrum to generate a refined model: 1) Importing source spectrum data; 2) Standardizing source spectrum data vectors so that each known industry source spectrum mass spectrometry information forms a 500-dimensional vector; 3) Training source spectrum data based on deep learning algorithms; 4) Selecting optimal parameters; 5) Generating a refined model based on the optimal parameter set.

[0159] Atmospheric heavy metal particulate matter data are fed into a trained model, resulting in a vector for each data point. This vector is then compared with the source spectrum labels to determine the closest possible source class: 1) Import atmospheric heavy metal particulate matter data; 2) Standardize the measured data vectors to form a 500-dimensional vector for each source spectrum mass spectrometry information; 3) Calculate the model using the optimal parameters to obtain the final classification result.

[0160] Table 9 shows the matching results of heavy metal particles from the four emission outlets of the smelter with atmospheric heavy metals and Pb- and Zn-containing particles (the data used in this study were the matching data of heavy metal- and Pb- and Zn-containing particles in stairwell dust with the source spectra of heavy metal-containing particles from the smelter at a similarity of 0.7, i.e., the optimal similarity analysis results in Table 8. The Cu- and Cd-containing particles from the smelter and the various heavy metal particles from the steel plant had too few corresponding source spectra to meet the algorithm's requirements).

[0161] The results showed that the heavy metals Pb and Zn in the smelter mainly came from emissions from DA031 and DA018 emission outlets.

[0162] Table 9

[0163]

[0164] This embodiment also provides a refined source apportionment system for atmospheric deposition heavy metals in soil. This system is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the system described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0165] This embodiment provides a refined source apportionment system for atmospheric deposition heavy metals in soil, such as... Figure 10 As shown, it includes:

[0166] The pollution source and sampling point determination module 1001 is used to determine multiple heavy metal pollution sources and multiple soil dust particulate matter sampling points based on the soil pollution distribution and air pollution source investigation results of the target area.

[0167] The mass spectrometry data analysis module 1002 is used to acquire receptor mass spectrometry data of soil dust particles at each sampling point and source mass spectrometry data of each heavy metal pollution source.

[0168] The source apportionment module 1003 is used to perform refined source apportionment of soil dust particles at each sampling point based on receptor mass spectrometry data and source mass spectrometry data, and to determine the contribution ratio of each heavy metal pollution source to the heavy metal pollution in the soil dust particles at each sampling point.

[0169] In some alternative implementations, the mass spectrometry data analysis module 1002 includes:

[0170] The sample acquisition unit is used to acquire gaseous and / or liquid and / or solid samples at the sampling point if the sampling point cannot meet the detection conditions of the single-particle aerosol mass spectrometer.

[0171] The sample detection unit is used to perform mass spectrometry detection on gaseous and / or liquid and / or solid samples to obtain receptor mass spectrometry data of soil dust particles at the sampling point.

[0172] The pollution source particle collection unit is used to collect pollution source particles from various heavy metal pollution sources using a vacuum bottle.

[0173] The pollution source particle detection unit is used to pass each pollution source particle into a single particle aerosol mass spectrometer for detection, and obtain the mass spectrometry characteristic information of each pollution source particle, which serves as the source mass spectrometry data of each heavy metal pollution source.

[0174] In some alternative implementations, the source resolution module 1003 includes:

[0175] The mass spectrometry library construction unit is used to construct a mass spectrometry library using source mass spectrometry data. The mass spectrometry library includes multiple source spectra, each of which represents the industry to which the heavy metal pollution source in the target area belongs.

[0176] The similarity calculation unit is used to perform dot product calculations between receptor mass spectrometry data and source spectra in the mass spectrometry library based on the ART-2a algorithm, and determine the industry to which the heavy metal pollution source corresponding to the receptor mass spectrometry data belongs based on the principle of maximum similarity.

[0177] The source apportionment unit is used to classify receptor mass spectrometry data of the same industry using a source apportionment model to obtain the contribution ratio of each heavy metal in the receptor mass spectrometry data to the heavy metal discharge outlets of the corresponding enterprises.

[0178] In some optional implementations, the source resolution unit includes:

[0179] The model training subunit is used to analyze the corresponding heavy metal types based on the source mass spectrometry data of particles from each pollution source, and to train a preset deep learning model based on the heavy metal types and corresponding heavy metal pollution sources to obtain the source resolution model.

[0180] The discharge outlet determination subunit is used to determine the pollution discharge outlets corresponding to the source mass spectrometry data of each heavy metal pollution source based on the source apportionment model.

[0181] The heavy metal type determination subunit is used to input receptor mass spectrometry data into the source analysis model to obtain the heavy metal type corresponding to the receptor mass spectrometry data.

[0182] The heavy metal source determination subunit is used to determine the source of heavy metals in soil dust particles at sampling points based on the heavy metal categories in receptor mass spectrometry data and the heavy metal categories at each pollution discharge outlet.

[0183] Further functional descriptions of the above modules and units are the same as those in the corresponding embodiments described above, and will not be repeated here.

[0184] The soil atmospheric deposition heavy metal fine source apportionment system in this embodiment is presented in the form of functional units. Here, a unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0185] This invention also provides a computer device having the above-described features. Figure 10 The system shown is a fine source apportionment system for atmospheric deposition heavy metals in soil.

[0186] Please see Figure 11 , Figure 11This is a schematic diagram of the structure of a computer device provided in an optional embodiment of the present invention, such as... Figure 11 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 11 Take a processor 10 as an example.

[0187] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0188] The memory 20 stores instructions executable by at least one processor 10 to cause the at least one processor 10 to perform the method shown in the above embodiments.

[0189] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0190] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0191] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0192] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0193] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0194] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for refined source apportionment of atmospheric deposition heavy metals in soil, characterized in that, The method includes: Based on the distribution of soil pollution and the results of the investigation of air pollution sources in the target area, multiple heavy metal pollution sources and multiple dust particulate matter sampling points were identified. Obtain receptor mass spectrometry data of soil dust particles at each sampling point and source mass spectrometry data of each heavy metal pollution source; Based on the receptor mass spectrometry data and the source mass spectrometry data, a refined source apportionment of soil dust particles at each sampling point was performed to determine the contribution ratio of each heavy metal pollution source to the heavy metal pollution in the soil dust particles at each sampling point.

2. The method according to claim 1, characterized in that, Obtain receptor mass spectrometry data of soil dust particles at each sampling point, including: If the sampling point cannot meet the detection conditions of the single-particle aerosol mass spectrometer, then gas samples and / or liquid samples and / or solid samples are obtained at the sampling point. Mass spectrometry was performed on the gaseous and / or liquid and / or solid samples to obtain receptor mass spectrometry data of soil dust particles at the sampling point.

3. The method according to claim 2, characterized in that, Obtaining receptor mass spectrometry data of soil dust particles at each sampling point also includes: If the sampling point meets the detection conditions of the single-particle aerosol mass spectrometer, then place the single-particle aerosol mass spectrometer at the sampling point to obtain particulate matter in the atmospheric environment. Mass spectrometry was used to detect particulate matter in the atmospheric environment to obtain receptor mass spectrometry data of soil dust particles at the sampling point.

4. The method according to claim 1, characterized in that, Obtain source mass spectrometry data for each heavy metal pollution source, including: Vacuum bottles were used to collect particulate matter from various heavy metal pollution sources. Particles from each pollution source were passed through a single-particle aerosol mass spectrometer for detection, and the mass spectrometric characteristics of each pollution source particle were obtained as source mass spectrometry data for each heavy metal pollution source.

5. The method according to claim 1, characterized in that, The heavy metal pollution sources include: multiple pollution outlets of at least one enterprise emitting heavy metals in the target area; based on the receptor mass spectrometry data and the source mass spectrometry data, a refined source apportionment of soil dust particles at each sampling point is performed, including: A mass spectrometry library was constructed using source mass spectrometry data. The mass spectrometry library includes multiple source spectra, each of which represents the industry to which the heavy metal pollution source in the target area belongs. The receptor mass spectrometry data and the source spectra in the mass spectrometry library are calculated by dot product based on the ART-2a algorithm, and the industry to which the heavy metal pollution source corresponding to the receptor mass spectrometry data belongs is determined based on the principle of maximum similarity. By classifying receptor mass spectrometry data from the same industry using a source apportionment model, the contribution percentage of heavy metal emissions from enterprises corresponding to each heavy metal in the receptor mass spectrometry data can be obtained.

6. The method according to claim 1, characterized in that, Receptor mass spectrometry data from the same industry were classified using a source apportionment model to obtain the contribution percentage of each heavy metal's emissions from the corresponding enterprise's heavy metal discharge outlets, including: Based on the source mass spectrometry data of particles from each pollution source, the corresponding heavy metal types are analyzed, and based on the heavy metal types and corresponding heavy metal pollution sources, a pre-set deep learning model is trained to obtain the source resolution model. Based on the source apportionment model, the pollution discharge outlets corresponding to the source mass spectrometry data of each heavy metal pollution source are determined; The receptor mass spectrometry data is input into the source resolution model to obtain the heavy metal type corresponding to the receptor mass spectrometry data; Based on the heavy metal categories in the receptor mass spectrometry data and the heavy metal categories at each pollution discharge outlet, the sources of heavy metals in the soil dust particles at the sampling points were determined.

7. A refined source apportionment system for atmospheric deposition heavy metals in soil, characterized in that, The system includes: The pollution source and sampling point determination module is used to determine multiple heavy metal pollution sources and multiple soil dust particulate matter sampling points based on the soil pollution distribution and air pollution source survey results of the target area. The mass spectrometry data analysis module is used to acquire receptor mass spectrometry data of soil dust particles at each sampling point and source mass spectrometry data of each heavy metal pollution source. The source apportionment module is used to perform refined source apportionment of soil dust particles at each sampling point based on the receptor mass spectrometry data and the source mass spectrometry data, and to determine the contribution ratio of each heavy metal pollution source to the heavy metal pollution in the soil dust particles at each sampling point.

8. A computer device, characterized in that, include: A memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, the processor executing the computer instructions to perform the method of any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a computer to perform the method of any one of claims 1 to 6.

10. A computer program product, characterized in that, Includes computer instructions for causing a computer to perform the method of any one of claims 1 to 6.

Images