Agricultural watershed pollution traceability method, system, storage medium, and electronic device

Abstract

The application provides an agricultural watershed pollution tracing method, system, storage medium and electronic equipment, which comprises the following steps: obtaining the to-be-tested sample data of different river sections in a target watershed; comparing the to-be-tested sample data with a constructed pollution source comparison database to determine the target pollution source in the to-be-tested sample data, wherein the pollution source comparison database comprises molecular matrix data and isotope ratio data of pollution sources in the target watershed; and calculating the relative contribution degree of the target pollution source to the to-be-tested sample of the river section according to the pollution source characteristic group to obtain a tracing ratio to generate a tracing report. The application can more accurately trace the pollution source, comprehensively considers the influence of hydrological factors on pollution source discharge, enhances the time and space adaptability and accuracy of the tracing result, and has higher resolution and accuracy and a wider application range.

Description

Technical Field

[0001] This invention belongs to the field of data processing and environmental protection technology, and in particular relates to a method, system, storage medium and electronic equipment for tracing pollution sources in agricultural watersheds. Background Technology

[0002] The sources of water pollution in rural watersheds are complex, including not only pollution from domestic sources but also a large amount of agricultural non-point source pollution. Agricultural non-point source pollution is characterized by multiple sources, discontinuity, and concealment, making it difficult to carry out corresponding source tracing work and accurately track the source of major pollutants. Therefore, targeted prevention and control of water pollution in rural watersheds faces enormous challenges.

[0003] Currently, among the commonly used technologies for tracing agricultural watershed pollution sources, nitrogen and oxygen isotope-based methods are widely used. However, nitrogen and oxygen isotope ratios in agricultural pollution sources may overlap or be similar, especially in adjacent areas or areas with similar agricultural activities. This makes it difficult to effectively distinguish the contributions of different sources by simply relying on isotope ratios. Furthermore, agricultural activities and natural environmental factors (such as precipitation and soil properties) have a significant impact on isotope ratios, resulting in strong fluctuations in the spatiotemporal changes of isotope ratios, which further increases the complexity of pollution source tracing.

[0004] In addition, two-dimensional river water quality models are often used to trace the source of pollution in agricultural watersheds. However, two-dimensional water quality models are usually based on certain assumptions and may not accurately reflect the complex hydrodynamics and water quality changes in reality. For example, the model assumes that the water body is uniformly mixed, but in reality, the water flow in rivers is often highly non-uniform, which affects the accuracy of the model results. Furthermore, the model requires a large amount of water quality monitoring data and hydrological data as input. The lack or inaccuracy of data may lead to deviations in the model output results, especially in rural areas where there are fewer monitoring stations and insufficient data collection, which affects the reliability of the model. Summary of the Invention

[0005] In view of the shortcomings of the prior art described above, the purpose of this invention is to provide a method, system, storage medium and electronic device for tracing pollution sources in agricultural watersheds, so as to solve the problem of tracing pollution sources in complex agricultural watershed environments in the prior art.

[0006] In a first aspect, the present invention provides a method for tracing the source of pollution in agricultural watersheds, the method comprising:

[0007] Acquire the sample data of different river sections within the target watershed;

[0008] The data of the sample to be tested is compared with an existing pollution source comparison database, which includes molecular matrix data and isotope ratio data of pollution sources within the target watershed.

[0009] Based on the test sample data and the specific molecular formulas in the molecular matrices of different pollution sources, a preliminary screening is performed to determine the target pollution sources in the test sample data.

[0010] The relative contribution of the target pollution source to the sample to be tested at the river section is calculated based on the pollution source feature group to obtain the source tracing ratio and generate a source tracing report. The pollution source feature group includes the molecular matrix and isotope ratio of the sample to be tested.

[0011] In one possible implementation of the first aspect, the preliminary screening based on the sample data to be tested, combined with specific molecular formulas in the molecular matrices of different pollution sources, to determine the target pollution source in the sample data to be tested specifically includes:

[0012] The molecular matrix and isotope ratio of the sample to be tested are obtained by deeply mining the Fourier transform ion cyclotron resonance mass spectrometry data in the sample data as the pollution source feature group, wherein the sample data to be tested is input by the user terminal;

[0013] Preliminary screening is performed by comparing the specific molecular formula with all molecular formulas in the molecular matrix of the sample to be tested to identify the target pollution source in the sample data that may cause downstream contamination.

[0014] In one possible implementation of the first aspect, constructing the pollution source comparison database specifically includes:

[0015] The user terminal input data is obtained, including land use type data, meteorological data, and pollution source sample data, wherein the pollution source samples include water samples or sediment samples.

[0016] Hydrological data for the target watershed is determined based on the land use type data and the meteorological data, wherein the hydrological data includes hydrological processes and water quantity data.

[0017] Based on the pollution source sample data, the molecular matrix and isotope ratio of different pollution source samples are determined, thereby constructing the pollution source comparison database.

[0018] In one possible implementation of the first aspect, Fourier transform ion cyclotron resonance mass spectrometry is used to analyze sample data from different pollution sources to obtain specific molecular formulas and molecular structure characteristic parameters as the molecular matrix of pollution source samples, and the isotope ratios of different pollution source samples are measured to construct the pollution source comparison database, wherein the molecular matrix of pollution source samples is specifically the molecular matrix of dissolved organic matter.

[0019] In one possible implementation of the first aspect, the method further includes comparing molecular matrices of different pollution source samples with a publicly available mass spectrometry database to determine the specific molecular formula corresponding to each pollution source sample.

[0020] In one possible implementation of the first aspect, the land use type data and the meteorological data are input into the SWAT model to obtain the hydrological data, wherein the hydrological process is used to determine the upstream pollution source and the direction of water flow, and the water volume data is used to calculate the ratio of pollution source discharge and river flow to obtain the allocation weight.

[0021] In one possible implementation of the first aspect, the relative contribution of the pollution source feature group in the target pollution source is calculated to obtain the source tracing ratio in order to generate a source tracing report, specifically including:

[0022] Based on the pollution source characteristic group data, the relative contribution of the target pollution source to the test sample of the river section is calculated by using the MixSIAR model in combination with the ratio weight to obtain the source tracing ratio result.

[0023] A visual source tracing report is generated based on the source tracing ratio results, wherein the source tracing report corresponds to the contribution ratio of different pollution sources in the current sample to be tested.

[0024] Secondly, the present invention provides an agricultural watershed pollution tracing system, the system comprising:

[0025] The acquisition module is used to acquire the test sample data of different river cross sections within the target watershed;

[0026] The comparison module is used to compare the data of the sample to be tested with the established pollution source comparison database, which includes molecular matrix data and isotope ratio data of pollution sources in the target watershed.

[0027] The initial screening module is used to perform initial screening based on the sample data to be tested and the specific molecular formula in the molecular matrix of the pollution source to determine the target pollution source in the sample data to be tested.

[0028] The generation module is used to calculate the relative contribution of the target pollution source to the sample to be tested in the river section based on the pollution source feature group to obtain the source tracing ratio and generate a source tracing report. The pollution source feature group includes the molecular matrix and isotope ratio of the sample to be tested.

[0029] Thirdly, the present invention provides an electronic device, the electronic device comprising: a processor and a memory;

[0030] The memory is used to store computer programs;

[0031] The processor is used to execute the computer program stored in the memory, so that the electronic device performs the above-described method for tracing the source of pollution in agricultural watersheds.

[0032] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by an electronic device, implements the above-described method for tracing the source of pollution in agricultural watersheds.

[0033] As described above, the agricultural watershed pollution tracing method, system, storage medium, and electronic device of the present invention have the following beneficial effects:

[0034] 1. It can more accurately trace the source of pollution, comprehensively consider the impact of hydrological factors on pollution source emissions in rural areas, and enhance the temporal and spatial adaptability and accuracy of the source tracing results;

[0035] 2. Based on time and space factors, the pollution sources in rural watersheds were divided in detail, and the pollution sources in different seasons and periods were distinguished. This breaks through the shortcomings of the traditional method, which treats agricultural pollution in rural areas as agricultural non-point source pollution, resulting in unclear source tracing results. This facilitates the subsequent refined management of pollution emissions in rural watersheds.

[0036] 3. Analysis by Fourier transform ion cyclotron resonance mass spectrometry and 15 N / 14 N、 18 O / 16 The coupling and combined use of two O isotope detection methods with the MixSIAR model enables the simultaneous tracing of pollution sources in rural areas from both organic and inorganic perspectives. It also provides more characteristic endmember features for the MixSIAR model, making the calculation results more valuable for practical reference.

[0037] 4. It can automatically update the pollution source database and simulate hydrological processes, realize real-time and efficient calculation of pollution source tracing in rural watersheds, and generate real-time reports on the analysis results of major pollution sources and hydrological change trends in the target rural area, providing strong data support for the real-time formulation of targeted pollution source prevention and control measures;

[0038] 5. It can be applied to the source tracing of pollution sources in various water body types (such as rivers and lakes), and can also be extended to the management and control of pollution sources in agriculture, industry and cities. It has a wide range of universal application prospects and provides technical support for the prevention and control of water pollution in designated areas. Attached Figure Description

[0039] Figure 1 The diagram shows a scenario application of the agricultural watershed pollution tracing method of the present invention in one embodiment;

[0040] Figure 2 The diagram shows a step-by-step illustration of the agricultural watershed pollution tracing method of the present invention in one embodiment.

[0041] Figure 3This is a flowchart illustrating land use analysis in one embodiment of the agricultural watershed pollution tracing method of the present invention.

[0042] Figure 4 The diagram shows a SWAT model simulation flowchart of an embodiment of the agricultural watershed pollution tracing method of the present invention.

[0043] Figure 5 The diagram shows a schematic flow of FT-ICR-MS data analysis in one embodiment of the agricultural watershed pollution tracing method of the present invention.

[0044] Figure 6 The diagram shows a flowchart of a machine learning model in one embodiment of the agricultural watershed pollution tracing method of the present invention.

[0045] Figure 7 This is a schematic diagram of the DOM molecular matrix of a paddy field in one embodiment of the agricultural watershed pollution tracing method of the present invention.

[0046] Figure 8 This is a schematic diagram of the DOM molecular matrix of aquaculture in an embodiment of the agricultural watershed pollution tracing method of the present invention.

[0047] Figure 9 The diagram shows a schematic diagram of the DOM molecular matrix of livestock and poultry farming in one embodiment of the agricultural watershed pollution tracing method of the present invention.

[0048] Figure 10 The diagram shows a schematic diagram of the molecular matrix (DOM) of domestic sewage in a residential area in one embodiment of the agricultural watershed pollution tracing method of the present invention.

[0049] Figure 11 The diagram shows a dryland DOM molecular matrix in one embodiment of the agricultural watershed pollution tracing method of the present invention.

[0050] Figure 12 The diagram shows a MixSIAR model flowchart in one embodiment of the agricultural watershed pollution tracing method of the present invention.

[0051] Figures 13A-13B The diagram shows the data results of pollutant tracing using the MixSIAR model in one embodiment of the agricultural watershed pollution tracing method of the present invention.

[0052] Figure 14 The diagram shown is a structural schematic of the agricultural watershed pollution tracing system of the present invention in one embodiment.

[0053] Figure 15 The diagram shown is a structural schematic of an embodiment of the electronic device of the present invention.

[0054] Component designation explanation

[0055] Steps S202~S208

[0056] 140 Agricultural Watershed Pollution Source Tracing System

[0057] 141 Acquisition Module

[0058] 142 Comparison Module

[0059] 143 Initial Screening Module

[0060] 144 Generation Module Detailed Implementation

[0061] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0062] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0063] Furthermore, the use of terms such as "first" and "second" in this application is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. If the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed in this application.

[0064] Currently, although existing source tracing technologies can provide clues for tracing pollution sources to some extent, due to the special and complex nature of pollution sources in rural areas, single technologies are often insufficient to address these challenges. To more accurately trace pollution sources, there is an urgent need for a comprehensive source tracing method applicable to complex rural watershed environments, possessing high accuracy and computational efficiency. Therefore, this application proposes an agricultural watershed pollution source tracing method. Specifically, addressing the current situation where the main pollution contributors to river sections in rural areas change constantly with time and space, making it difficult to implement effective targeted remediation measures, this method is based on organic molecule diffusion inversion tracing for agricultural watershed pollution source tracing. Figure 1 As shown in the diagram, this is a scenario application illustration. After determining the target watershed, staff identify different pollution sources through on-site testing. They then collect land use type data, meteorological data, and pollution source sample data corresponding to the target watershed for data analysis to construct a pollution source comparison database. Samples are then collected at different river sections within the target watershed for analysis. The collected samples include water samples and sediment samples. Sediment samples are collected for analysis when there are many sunny days in the current period. Further analysis of the molecular matrix and isotope ratios of the samples is performed. Combined with the unique molecular formulas of the pollution source samples in the pollution source comparison database, a preliminary screening is conducted to identify the target pollution sources in the current sample data. This may include all pollution sources or several. The relative contribution of the pollution source characteristic groups in the target pollution sources is then calculated to obtain the source tracing ratio, generating a source tracing report. Correspondingly, the source tracing report shows the distribution ratio of different pollution sources, where the pollution source characteristic groups correspond to the molecular matrix and isotope ratios of the samples.

[0065] Specifically, the technical solutions in the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0066] like Figure 2 As shown, in one embodiment of the invention, the agricultural watershed pollution tracing method of the present invention includes the following steps:

[0067] Step S202: Obtain the sample data of different river sections within the target watershed;

[0068] Step S204: Compare the data of the sample to be tested with the established pollution source comparison database;

[0069] Step S206: Based on the test sample data and the specific molecular formulas in the molecular matrices of different pollution sources, a preliminary screening is performed to determine the target pollution source in the test sample data.

[0070] Step S208: Calculate the relative contribution of the target pollution source to the sample to be tested in the river section based on the pollution source characteristic group to obtain the source tracing ratio and generate a source tracing report.

[0071] It should be noted that, in this embodiment, when tracing pollution sources in agricultural watersheds, pollution sources within the current target watershed have been identified in advance, such as livestock and poultry farming, shrimp ponds and fish ponds, and domestic sewage from residential areas. This allows for the comparison of the collected sample data with known pollution sources to obtain the distribution of different pollution sources. Specifically, in application, the user collects samples from different river sections within the target watershed, thereby obtaining a pollution source comparison database constructed from the sample data input by the user. The sample data is then subjected to FT-ICR-MS (Fourier Transform Ion Cyclotron Resonance Mass Spectrometry) and isotope analysis to obtain the molecular matrix and isotope ratio of the sample as the pollution source feature group.

[0072] Further, specifically, based on the sample data to be tested and combined with the special molecular formulas in the molecular matrices of different pollution sources in the constructed pollution source comparison database, a preliminary screening is performed to determine the target pollution sources in the sample data to be tested. Each pollution source in the target watershed corresponds to a unique special molecular formula. After determining the target pollution sources in the current sample data to be tested, the relative contribution of the pollution source feature groups in the target pollution source is calculated based on the MixSIAR (Mixing Model for Stable Isotope Analysis in R, Bayesian Framework) model to obtain the source tracing ratio and generate a source tracing report, thereby showing the distribution ratio of different pollution sources located in the current watershed in the current sample to be tested. The invention patent CN117368298A describes a method for tracing the source of overflow pollution in municipal pipe networks based on DOM molecular group discrimination tracing. However, in practical applications, the abundance values ​​obtained by FT-ICR-MS can only indicate the relative abundance within a group and cannot reflect the absolute amount of the molecular formula. Therefore, the method in invention patent CN117368298A has defects and cannot guarantee the accuracy of the data. In contrast, this invention first qualitatively identifies the type of pollution source by using the characteristics of special molecular formulas and isotopes. Then, it calculates the contribution by using the DOM molecular characteristic parameters (which do not have abundance issues) obtained by FT-ICR-MS and the isotope ratio. Under the premise of ensuring qualitative accuracy, it can achieve quantitative calculation and solves the existing defects in invention patent CN117368298A.

[0073] Furthermore, in one embodiment of the invention, the preliminary screening based on the sample data to be tested, combined with the specific molecular formulas in the molecular matrices of different pollution sources, to determine the target pollution source in the sample data to be tested specifically includes:

[0074] The molecular matrix and isotope ratio of the sample to be tested are obtained by deeply mining the Fourier transform ion cyclotron resonance mass spectrometry data in the sample data as the pollution source feature group, wherein the sample data to be tested is input by the user terminal;

[0075] Preliminary screening is performed by comparing the specific molecular formula with all molecular formulas in the molecular matrix of the sample to be tested to identify the target pollution source in the sample data that may cause downstream contamination.

[0076] It should be noted that, in this embodiment, the sample data input by the user terminal is obtained, and Fourier transform ion cyclotron resonance mass spectrometry and isotope analysis are performed to obtain the molecular matrix and isotope ratio of the sample as the pollution source characteristic group. The molecular matrix of the sample includes the molecular formula and molecular characteristic parameters, such as C 1～∝ H 1～∝ O 1～∝ N 0-3 S 0-1 P 0-1 Molecular formulas and molecular characteristic parameters, such as saturation index H / C, oxidation degree O / C, nitrogen-containing organic matter characteristic N / O, phosphorus-containing organic matter characteristic P / C, sulfur-containing organic matter characteristic S / C, degree of unsaturation DBE, and nominal carbon oxidation state NOSC, can be used to reduce computational load, save costs, and ensure data stability. This can be achieved by first screening the sample to be tested, comparing the specific molecular formulas of pollution sources with those of the sample to determine which pollution sources are present in the current sample. This helps identify the specific pollution sources and ultimately pinpoint the target pollution sources that pose a downstream contamination risk in the current sample data. Specifically, the special molecular formula in this invention refers to the unique molecular formula of each pollution source. The molecular matrix formed by combining molecular characteristics has a strong directional relationship with the corresponding pollution source. Although the academic journal that analyzes the sources and mechanisms of urban river pollution based on the molecular fingerprint characteristics of dissolved organic matter also describes the research on the pollution mechanism of river DOM based on FT-ICR-MS, the special molecular formula studied in that article refers to the common molecular formula found in dry fields and rainy days DOM. It is not the same concept as the special molecular formula corresponding to the specific pollution source in this application. Therefore, given that the author of the academic journal is the same as the inventor of this application, the above explanation is made.

[0077] Furthermore, in one embodiment of the invention, constructing the pollution source comparison database specifically includes:

[0078] The user terminal input data is obtained, including land use type data, meteorological data, and pollution source sample data, wherein the pollution source samples include water samples or sediment samples.

[0079] Hydrological data for the target watershed is determined based on the land use type data and the meteorological data, wherein the hydrological data includes hydrological processes and water quantity data.

[0080] Based on the pollution source sample data, the molecular matrix and isotope ratio of different pollution source samples are determined, thereby constructing the pollution source comparison database.

[0081] It should be noted that this embodiment specifically describes how to construct a pollution source comparison database. After determining the target watershed, staff will collect data from different pollution sources within the target watershed to obtain user-end input data for data processing. The input data specifically includes land use type data, meteorological data, and pollution source sample data. The land use type data is obtained by analyzing the main land use types of the target watershed using ArcGIS, specifically high-resolution land use classification results. Further, pollutant samples include water samples or sediment samples, specifically differentiated based on meteorological data. For example, sediment samples are collected in application scenarios with a long period of sunny days, otherwise water samples are collected. In one embodiment, such as... Figure 3The diagram shows a flowchart of the land use analysis process for a watershed using ArcGIS software. In the data preparation stage, remote sensing images (high-resolution remote sensing images) and basic geographic data (including DEM (Digital Elevation Model) elevation data, administrative boundaries, and water systems) are collected. In the data preprocessing stage, the remote sensing images are preprocessed (radiometric correction, atmospheric correction, image stitching and cropping), and watershed boundaries are established (using the DEM to divide the watershed and extract confluence areas). The land use classification stage includes training sample selection (manually selecting training points based on existing ground survey data or high-resolution map patches), supervised classification (using support vector machines, maximum likelihood methods, etc. for land use classification), and accuracy assessment (constructing a confusion matrix, calculating classification accuracy, and Kappa coefficient). The land use type statistics and spatial analysis stage includes land use type identification and coding (identifying dryland, paddy fields, residential areas, aquaculture, etc.) and calculating the land use area ratio (using zonal mapping). The statistics tool calculates the area and percentage of various types of land use. In the final output stage, a high-resolution land use classification map and table can be output. Based on the current land use situation in the watershed, the main land use types (i.e. potential sources of pollution) in the region are identified, including dry land (20%), paddy fields (10%), livestock and poultry farming areas (20%), aquaculture areas (20%), and residential areas (30%). Dry land is mainly broad bean fields and wheat fields, paddy fields are mainly lotus ponds and water hyacinth ponds, livestock and poultry farming is mainly poultry such as chickens and ducks, and aquaculture is mainly shrimp ponds and fish ponds.

[0082] Further, based on the land use type data and the meteorological data, hydrological data for the target watershed is determined. The obtained hydrological data includes hydrological processes and water volume data. Specifically, the land use type data and the meteorological data are input into a SWAT (Soil and Water Assessment Tool) model to obtain the hydrological data. The hydrological processes are used to determine upstream pollution sources and water flow direction, and the water volume data is used to calculate the ratio of pollution source emissions to river flow to obtain a weighting factor. Figure 4The diagram shows the SWAT model simulation flowchart. In the data preparation stage, it includes a digital elevation model (DEM) for delineating watershed boundaries and river networks, land use data (which may be sourced from remote sensing classification results), and meteorological data (rainfall, temperature, wind speed, and humidity). In the watershed modeling stage, it includes sub-watershed delineation (automatically delineating sub-watersheds and river networks using DEM data and flow direction analysis) and hydrological response units (HRUs). The system is divided into units, specifically generating HRUs based on land use, soil, and slope combinations to reflect watershed heterogeneity. The parameter setting and model configuration phase includes setting agricultural activity parameters (crop type, cultivation method, fertilization / irrigation system, etc.) and pollutant load coefficients (initial nitrogen and phosphorus output coefficients for various land use types). The model operation and calibration phase includes SWAT simulation (simulating runoff, pollutant molecular diffusion, soil erosion, etc.) and model calibration and validation (comparing with measured water quality and hydrological data and adjusting parameters). The results analysis and output phase includes pollution output load analysis (obtaining pollutant output from each HRU, sub-watershed, and land use type) and spatial distribution maps and contribution rate statistics (outputting pollutant molecular diffusion maps, etc., overlaid with land use maps for analysis). Specifically, in the example above, the meteorological data for the watershed includes annual precipitation of "1200 mm", annual evaporation of "800 mm", and land use information. A SWAT model is established with a watershed area of ​​"500 km²". 2 The model primarily covers agricultural and aquaculture areas, simulating hydrological processes within the basin during different seasons, with particular focus on water volume changes during agricultural drainage seasons. The SWAT model simulation results are as follows: Rainy season (June-September): water flow rate of 30 m³ / h. 3 / s, agricultural wastewater accounts for "60%", livestock wastewater accounts for "25%", and domestic sewage accounts for "15%"; dry season (October-May): water flow rate is "5m 3 / s, agricultural wastewater accounts for 45%, livestock wastewater accounts for 30%, and domestic sewage accounts for 25%.

[0083] Furthermore, based on the pollution source sample data, the molecular matrix and isotope ratio of different pollution source samples are determined, thereby constructing the pollution source comparison database. By analyzing the data of different pollution source samples through Fourier transform ion cyclotron resonance mass spectrometry, the specific molecular formula and molecular structure characteristic parameters are obtained as the molecular matrix of pollution source samples, and the isotope ratio of different pollution source samples is measured to construct the pollution source comparison database. Specifically, the molecular matrix of pollution source samples is the molecular matrix of dissolved organic matter.

[0084] It should be noted that, in this embodiment, the molecular matrix and isotope ratio of different pollution source samples are determined based on the input pollution source sample data. In application, based on land use analysis and SWAT model simulation results, this embodiment specifically describes eight pollution sources: broad bean fields, wheat fields, lotus ponds, water hyacinth ponds, livestock and poultry farming, shrimp ponds and fish ponds, and domestic sewage from residential areas. Staff collected five pollution source samples for each pollution source, totaling 40 pollution source samples, which included the main types of crops and aquaculture in the current season.

[0085] Furthermore, for the collected pollution source samples, after determining the extraction method with the highest overall DOM (Dissolved Organic Matter) extraction efficiency through comparative experiments, the staff performed solid-phase extraction on the collected samples under optimal extraction conditions. Solid-phase extraction eliminates salts and limits artifacts caused by differences in DOM concentration in the samples during FT-ICR-MS analysis. The volume of the PPL column was adjusted according to the amount of dissolved organic carbon, and the target DOC concentration of the eluent was 100 mg / L. The detailed solid-phase extraction method is as follows:

[0086] (1) After passing the sample through a 0.45μm filter membrane, adjust the pH to 2 with HCl for later use;

[0087] (2) The PPL column is pre-activated by rinsing it with 10 mL of HPLC-grade methanol at a rate of 2 mL / min.

[0088] (3) Rinse the PPL column with 10 mL of 0.01 M HCl at a rate of 2 mL / min;

[0089] (4) Pass the acidified sample with pH=2 through the PPL column at a rate of 5 mL / min;

[0090] (5) Rinse the PPL column with 10 mL of 0.01 M HCl at a rate of 2 mL / min.

[0091] (6) Dry the PPL column with ultrapure N2;

[0092] (7) Elute the DOM enriched in the PPL column with “5 mL” of HPLC-grade methanol at a rate of “2 mL / min”.

[0093] (8) Further, in this embodiment, the sample data to be tested is detected by FT-ICR-MS. The methanol sample after DOM extraction is detected by Bruker SolariX FT-ICR-MS. The ion source is an electrospray ionization source (ESI) in negative ion mode. The main detection parameters are: continuous injection, injection rate of 120 μL / h, capillary inlet voltage of 4.0 kV, ion accumulation time of 0.1 s, acquisition mass range of 100-1600 Da, number of sampling points of 4M 32-bit data, and time domain signal superposition of 300 times to improve the signal-to-noise ratio. The instrument is calibrated with 10 mmol / L sodium formate before the pollution source sample is detected. After the sample is detected, soluble organic matter (known molecular formula) is used for internal standard calibration. After calibration, the detection quality error is less than 1 ppm.

[0094] Furthermore, in one embodiment of the invention, the method further includes comparing the molecular matrices of different pollution source samples with a publicly available mass spectrometry database to determine the specific molecular formula corresponding to each pollution source sample.

[0095] Specifically, such as Figure 5 The diagram shows a schematic of the FT-ICR-MS data analysis workflow. Firstly, the mass spectrometry peak data is processed using an isotope classification and removal algorithm to obtain molecular ion peak data. Then, a multi-element molecular matching algorithm is used to determine the C atoms present in the sample. 1～∝ H 1～∝ O 1～∝ N 0-3 S 0-1 P 0-1 Molecular formula; further calculate molecular characteristic parameters such as H / C, O / C, N / O, P / C, S / C, AImod, DBE, NOSC, KMD, etc.; on the other hand, through statistical analysis, screen out the DOM molecular formulas that are commonly found in multiple samples of the same pollution source type as the common molecular formulas of the corresponding pollution sources; compare the common molecular formulas of different sources, screen out the DOM molecular formulas that exist only in a single pollution source as the specific DOM molecular formulas of the current pollution source, that is, the corresponding special molecular formulas. Among them, compare with public mass spectrometry databases, such as SDBS (Spectral Database for Organic Compounds), Shanghai Institute of Organic Chemistry Chemical Database, etc., to match and identify the specific DOM molecular formulas (special molecular formulas) corresponding to the "8" pollution sources, namely broad bean fields, wheat fields, lotus ponds, seaweed ponds, livestock and poultry farming, shrimp ponds and fish ponds, and domestic sewage from residential areas.

[0096] Furthermore, it should be noted that the above embodiments illustrate the determination of the molecular matrix of samples from different pollution sources, including molecular formulas and molecular characteristic parameters. In this embodiment, the determination of the isotope ratios of samples from different pollution sources is further explained. Specifically, the isotope ratios of samples from eight pollution sources, including broad bean fields, wheat fields, lotus ponds, water hyacinth ponds, livestock and poultry farms, shrimp ponds and fish ponds, and domestic sewage from residential areas, were measured. 15 N / 14 N、 18 O / 16 O isotope ratio, of which, broad bean field: 15 N / 14 N = 3.3” 18 O / 16 O = 1.6”; Wheat field: 15 N / 14 N = 3.5” 18 O / 16 O = 1.7”; Lotus root pond: 15 N / 14 N = 4.0” 18 O / 16 O = 1.8”; Water lily pond: 15 N / 14 N = 4.2” 18 O / 16 O = 1.9”; Livestock and poultry breeding area: 15 N / 14 N = 4.6” 18 O / 16 O = 1.9”; Shrimp pond: 15 N / 14 N = 3.8” 18 O / 16 O = 1.7”; Fishpond: 15 N / 14 N = 4.3” 18 O / 16 O = 1.8”; Domestic sewage in residential areas: 15 N / 14 N = 5.1” 18 O / 16 O = 2.2".

[0097] Furthermore, in practical applications, due to the sheer volume of data, machine learning models are employed for feature extraction of DOM molecular feature parameters and isotopic data. Molecular feature parameters are molecular structural characteristics, such as the oxygen-to-carbon ratio, degree of unsaturation, and the range of proportions of different elements. Isotopic data features correspond to the range of isotopic ratios. Specifically, for example... Figure 6The diagram shows the flowchart of the machine learning model. In the data preparation stage, it includes data acquisition (collecting samples from various pollution sources and receiving water samples, performing FT-ICR-MS and isotope detection to obtain metadata) and data preprocessing (completing molecular fingerprint feature (corresponding to molecular matrix) analysis and isotope analysis). In the feature selection stage, the boots package is used to perform feature selection, combining random forest, PAMR, and GBM algorithms. The bootstrap method selects the features that have the most significant impact on the model's classification ability. Specifically, the bootstrap method is used for variable screening. The bootstrap method evaluates the robustness of features through repeated sampling ("100" iterations, repeated "20" times). In each sampling, features are randomly sampled into a new training set. After model training, the importance of each feature is evaluated, and finally, the features that have the most significant impact on the model's classification ability are selected. After all iterations are completed, the frequency of selection for each feature is calculated to obtain the most discriminative feature list. To ensure the conciseness and effectiveness of the final model, a threshold is set to an "85%" selection frequency, meaning only those features selected in "85%" iterations will be included in the final model.

[0098] Further, in the model building and training phase, a PLS-DA model (Partial Least Squares Discriminant Analysis, a supervised classification model based on partial least squares regression) is constructed to reduce the dimensionality of high-dimensional data and find the optimal linear relationship between pollution source categories and features. PLS-DA is a supervised learning method whose advantage lies in its ability not only to reduce the dimensionality of high-dimensional data but also in finding the optimal linear relationship between features and labels by combining the output labels (pollution source categories). In addition, a variable importance score (VIP) is established. Scores: Each variable (DOM molecular fingerprint) in the model receives a variable importance score based on its contribution to the classification result. Variables with a VIP value higher than "0.8" are considered to be the most important for classification. Data standardization and model training: To eliminate the influence of data scale, all compound data were transformed and standardized by log10 before entering the model to ensure that the influence of different features is on the same scale. Specifically, 5-fold cross-validation is used to evaluate the generalization ability of the model. Cross-validation divides the dataset into "5" subsets. Each time, "4" subsets are used for training, and the remaining "1" subset is used for testing. The predictive performance of the model is calculated through repeated validation.

[0099] Further, in the model validation and evaluation phase, the model performance and results are evaluated. The "predictive ability" of the PLS-DA model is assessed by the Q2 value, and the "goodness of fit" is assessed by the R2 value. The explained variance of the PLS-DA model represents the degree to which the model can explain the variation in the data, and is expected to be greater than "50%". For model validation and robustness evaluation, to verify the robustness of the model results, the study conducts "2000" permutation tests. By randomly shuffling the sample labels, the model is retrained to obtain the Q2 value distribution. If the Q2 value of the original model is significantly higher than the Q2 value of the random labels, the statistical significance and robustness of the model can be verified (p<0.0005).

[0100] Furthermore, in the model application stage, including pollution source identification and feature attribution, unknown samples are input into the trained model to output their main pollution source types, and key molecular fingerprint feature groups (i.e., DOM molecular matrices) corresponding to different pollution sources are obtained. Specifically, DOM molecular matrices with significant differences among eight types of pollution sources are extracted, including broad bean fields, wheat fields, lotus ponds, seaweed ponds, livestock and poultry farming, shrimp ponds and fish ponds, and domestic sewage from residential areas.

[0101] Furthermore, specifically, since pollution sources actually correspond to different land use types, and each land use type corresponds to different characteristic groups, molecular formulas, molecular structure parameters, and isotope ratios are used for refined differentiation to determine different pollution sources. Traditional methods, which use a single condition to differentiate pollution sources, have low resolution and low specificity. In this embodiment, pollution sources are differentiated and determined by constructing different molecular formulas, different molecular characteristic parameters, and different isotope ratios. For the "8" pollution sources in the above embodiment, there are "8" DOM ​​molecular matrices. For example, as illustrated in the diagram... Figures 7-11 As shown, this is a schematic diagram of the DOM molecular matrix corresponding to different pollution sources. Specifically, Figure 7 This is a schematic diagram of the DOM molecular matrix of paddy fields. Figure 8 This is a schematic diagram of the DOM molecular matrix for aquaculture. Figure 9 This is a schematic diagram of the molecular matrix of DOM (Domain of Origin) for livestock and poultry farming. Figure 10 This is a schematic diagram of the DOM (Matrices of Determinants) molecular matrix for domestic sewage in residential areas. Figure 11 This is a schematic diagram of the molecular matrix of DOM in dryland.

[0102] Furthermore, in one embodiment of the invention, the relative contribution of pollution source characteristic groups in the target pollution source is calculated to obtain the source tracing ratio in order to generate a source tracing report, specifically including:

[0103] Based on the pollution source characteristic group data, the relative contribution of the target pollution source to the test sample of the river section is calculated by using the MixSIAR model in combination with the ratio weight to obtain the source tracing ratio result.

[0104] A visual source tracing report is generated based on the source tracing ratio results, wherein the source tracing report corresponds to the contribution ratio of different pollution sources in the current sample to be tested.

[0105] It should be noted that, in this embodiment, the relative contribution of the pollution source feature group is calculated based on the MixSIAR model combined with the weighted ratio to obtain the source tracing ratio result. Based on the source tracing ratio result, a visual source tracing report is generated. Accordingly, the source tracing report corresponds to the distribution ratio of different pollution sources in the current sample to be tested. For example, agricultural pollution source in broad bean field: "25%"; agricultural pollution source in wheat field: "15%"; pollution source in lotus pond: "10%"; pollution source in water hyacinth pond: "5%"; livestock and poultry breeding wastewater: "20%"; aquaculture wastewater in shrimp pond: "10%"; aquaculture wastewater in fish pond: "5%"; and domestic sewage in residential area: "10%".

[0106] Furthermore, specifically, such as Figure 12 The diagram shows the MixSIAR model flowchart. In this embodiment, when constructing the MixSIAR model, the input data includes the feature groups composed of the DOM molecular matrix and isotope ratios described in the previous embodiment. This means converting the feature group data of each pollution source into the format required by the MixSIAR model for input. Further, model parameters are set, including prior distributions and mixed source assumptions. Appropriate isotope data are selected as input variables, and existing pollution source feature data are used as a reference. The model is then run using the Markov Chain Monte Carlo (MCMC) method for training, calculating the relative contribution ratio of different pollution sources in the target river water sample. Finally, the output result is the relative contribution of pollution sources in the target river water sample calculated using the MixSIAR model. Calculating the relative contribution based on the MixSIAR model is a technical solution that can be chosen by those skilled in the art; this embodiment only uses it as an application, and the specific process is not described in detail. Figures 13A-13B The image shows a schematic diagram of the data results for pollutant source tracing using the MixSIAR model.

[0107] Please see Figure 14 In one embodiment, this embodiment provides an agricultural watershed pollution tracing system 140, the system comprising:

[0108] The acquisition module 141 is used to acquire the test sample data of different river sections within the target watershed;

[0109] The comparison module 142 is used to compare the sample data to be tested with the constructed pollution source comparison database, which includes molecular matrix data and isotope ratio data of pollution sources in the target watershed.

[0110] The initial screening module 143 is used to perform initial screening based on the test sample data and the specific molecular formula in the pollution source molecular matrix to determine the target pollution source in the test sample data.

[0111] The generation module 144 is used to calculate the relative contribution of the target pollution source to the sample to be tested in the river section based on the pollution source feature group to obtain the source tracing ratio and generate a source tracing report. The pollution source feature group includes the molecular matrix and isotope ratio of the sample to be tested.

[0112] Since the specific implementation of this embodiment corresponds to the aforementioned method embodiment, the same details will not be repeated here, and those skilled in the art should also understand this. Figure 14 The division of the modules in the embodiments is only a logical functional division. In actual implementation, they can be fully or partially integrated into one or more physical entities. These modules can be fully implemented in software through processing element calls, fully implemented in hardware, or some modules can be implemented in software through processing element calls and some modules can be implemented in hardware.

[0113] In the several embodiments provided by this invention, it should be understood that the disclosed apparatus or method can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules / units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or units may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of apparatuses or modules or units may be electrical, mechanical, or other forms.

[0114] The modules / units described as separate components may or may not be physically separate. The components shown as modules / units may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules / units can be selected to achieve the objectives of the embodiments of the present invention, depending on actual needs. For example, the functional modules / units in the various embodiments of the present invention may be integrated into one processing module, or each module / unit may exist physically separately, or two or more modules / units may be integrated into one module / unit.

[0115] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0116] This invention also provides an electronic device, such as... Figure 15 As shown, the electronic device includes a processor and a memory.

[0117] The memory is used to store computer programs.

[0118] The processor is configured to execute a computer program stored in the memory to cause the electronic device to perform any of the methods described above.

[0119] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, or methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules / units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or units may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of apparatuses or modules or units may be electrical, mechanical, or other forms.

[0120] The modules / units described as separate components may or may not be physically separate. The components shown as modules / units may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules / units can be selected to achieve the objectives of the embodiments of this application, depending on actual needs. For example, the functional modules / units in the various embodiments of this application may be integrated into one processing module, or each module / unit may exist physically separately, or two or more modules / units may be integrated into one module / unit.

[0121] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0122] This application also provides a computer-readable storage medium. Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing a processor. The program can be stored in a computer-readable storage medium, which is a non-transitory medium, such as random access memory, read-only memory, flash memory, hard disk, solid-state drive, magnetic tape, floppy disk, optical disk, and any combination thereof. The storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., digital video disc (DVD)), or a semiconductor medium (e.g., solid-state drive (SSD)).

[0123] This application embodiment may also provide a computer program product comprising one or more computer instructions. When the computer instructions are loaded and executed on a computing device, all or part of the processes or functions described in this application embodiment are generated. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means.

[0124] When the computer program product is executed by a computer, the computer performs the method described in the foregoing method embodiments. The computer program product can be a software installation package; when the foregoing method is required, the computer program product can be downloaded and executed on the computer.

[0125] The descriptions of the processes or structures corresponding to the above figures each have their own emphasis. For parts of a process or structure that are not described in detail, please refer to the relevant descriptions of other processes or structures.

[0126] The above embodiments are merely illustrative of the principles and effects of this application and are not intended to limit this application. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in this application should still be covered by the claims of this application.

Claims

1. A method for tracing the source of pollution in agricultural watersheds, characterized in that, include: Acquire the sample data of different river sections within the target watershed; The data of the sample to be tested is compared with an existing pollution source comparison database, which includes molecular matrix data and isotope ratio data of pollution sources within the target watershed. Based on the test sample data and the specific molecular formulas in the molecular matrices of different pollution sources, a preliminary screening is performed to determine the target pollution sources in the test sample data. The relative contribution of the target pollution source to the sample to be tested in the river section is calculated based on the pollution source feature group to obtain the source tracing ratio and generate a source tracing report. The pollution source feature group includes the molecular matrix and isotope ratio of the sample to be tested. The initial screening based on the sample data to be tested, combined with the specific molecular formulas in the molecular matrices of different pollution sources, to determine the target pollution sources in the sample data specifically includes: The molecular matrix and isotope ratio of the sample to be tested are obtained by deeply mining the Fourier transform ion cyclotron resonance mass spectrometry data in the sample data as the pollution source feature group, wherein the sample data to be tested is input by the user terminal; Based on the specific molecular formula, a preliminary screening is performed by comparing all molecular formulas in the molecular matrix of the sample to be tested to identify the target pollution source in the sample data that may cause downstream contamination. The construction of the pollution source comparison database specifically includes: The user terminal input data is obtained, including land use type data, meteorological data, and pollution source sample data, wherein the pollution source samples include water samples or sediment samples. Hydrological data for the target watershed is determined based on the land use type data and the meteorological data, wherein the hydrological data includes hydrological processes and water quantity data. Based on the pollution source sample data, the molecular matrix and isotope ratio of different pollution source samples are determined, thereby constructing the pollution source comparison database; By analyzing the data of samples from different pollution sources using Fourier transform ion cyclotron resonance mass spectrometry, specific molecular formulas and molecular structure characteristic parameters were obtained as the molecular matrix of pollution source samples. The isotope ratios of samples from different pollution sources were also measured to construct the pollution source comparison database. Specifically, the molecular matrix of pollution source samples is the molecular matrix of dissolved organic matter. Calculating the relative contribution of pollution source characteristic groups in the target pollution source to obtain the source tracing ratio and generating a source tracing report, specifically including: Based on the pollution source characteristic group data, the relative contribution of the target pollution source to the test samples of the river section is calculated by using the MixSIAR model combined with the weighting ratio to obtain the source tracing ratio result. A visual source tracing report is generated based on the source tracing ratio results, wherein the source tracing report corresponds to the contribution ratio of different pollution sources in the current sample to be tested.

2. The method for tracing the source of agricultural watershed pollution according to claim 1, characterized in that, The method also includes comparing the molecular matrices of different pollution source samples with publicly available mass spectrometry databases to determine the specific molecular formula corresponding to each pollution source sample.

3. The method for tracing the source of agricultural watershed pollution according to claim 1, characterized in that, The land use type data and the meteorological data are input into the SWAT model to obtain the hydrological data. The hydrological process is used to determine the upstream pollution source and the direction of water flow, and the water volume data is used to calculate the ratio of pollution source discharge and river flow to obtain the allocation weight.

4. An agricultural watershed pollution source tracing system, characterized in that, include: The acquisition module is used to acquire the test sample data of different river cross sections within the target watershed; The comparison module is used to compare the data of the sample to be tested with the established pollution source comparison database, which includes molecular matrix data and isotope ratio data of pollution sources in the target watershed. The construction of the pollution source comparison database specifically includes: acquiring input data from the user terminal, the input data including land use type data, meteorological data, and pollution source sample data, wherein the pollution source samples include water samples or sediment samples; determining the hydrological data of the target watershed based on the land use type data and the meteorological data, the hydrological data including hydrological process and water quantity data; determining the molecular matrix and isotope ratio of different pollution source samples based on the pollution source sample data, thereby constructing the pollution source comparison database; analyzing the different pollution source sample data by Fourier transform ion cyclotron resonance mass spectrometry to obtain specific molecular formulas and molecular structure characteristic parameters as the pollution source sample molecular matrix, and determining the isotope ratio of different pollution source samples to construct the pollution source comparison database, wherein the pollution source sample molecular matrix is ​​specifically the dissolved organic matter molecular matrix; The initial screening module is used to perform initial screening based on the sample data to be tested and the specific molecular formulas in the molecular matrix of the pollution sources to determine the target pollution sources in the sample data to be tested. Specifically, this initial screening based on the sample data to be tested and the specific molecular formulas in the molecular matrices of different pollution sources to determine the target pollution sources in the sample data to be tested includes: deeply mining the Fourier transform ion cyclotron resonance mass spectrometry data in the sample data to be tested to obtain the molecular matrix and isotope ratios of the sample as the pollution source feature group, wherein the sample data to be tested is input by the user; and comparing the specific molecular formulas with all molecular formulas in the molecular matrix of the sample to be tested to determine the target pollution sources in the sample data that have downstream pollution potential. The generation module is used to calculate the relative contribution of the target pollution source to the sample to be tested in the river section based on the pollution source feature group to obtain the source tracing ratio and generate a source tracing report. The pollution source feature group includes the molecular matrix and isotope ratio of the sample to be tested. Specifically, calculating the relative contribution of the pollution source feature group in the target pollution source to obtain the source tracing ratio and generating the source tracing report includes: calculating the relative contribution of the target pollution source to the sample to be tested in the river section based on the pollution source feature group data using the MixSIAR model combined with weighting to obtain the source tracing ratio result; and generating a visualized source tracing report based on the source tracing ratio result, wherein the source tracing report corresponds to the contribution percentage of different pollution sources in the current sample to be tested.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the agricultural watershed pollution tracing method according to any one of claims 1 to 3.

6. An electronic device, characterized in that, The electronic device includes a processor and a memory; wherein the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to enable the electronic device to perform the agricultural watershed pollution tracing method as described in any one of claims 1 to 3.

Images