Underground water environment monitoring system and method

Abstract

The invention provides an underground water environment monitoring system and method, and the method comprises the steps: carrying out the three-dimensional fluorescence spectrum full-wave band scanning of a known type of pollution source reference sample in a target monitoring region, and obtaining a three-dimensional fluorescence spectrum data set; performing characteristic analysis on the three-dimensional fluorescence spectrum data set to obtain fluorescence characteristic components, and determining a characteristic excitation wavelength corresponding to each fluorescence characteristic component; the classification contribution degree of each characteristic excitation wavelength is evaluated, and a characteristic excitation wavelength combination during source tracing of the pollution source is obtained through screening; performing fixed wavelength scanning on the to-be-detected underground water sample in the target monitoring area according to the characteristic excitation wavelength combination, and determining the initial type of the pollution source; and determining the relative contribution degree of the type of the pollution source of the to-be-detected underground water sample, simulating the migration trend of the pollution plume according to the underground water migration and diffusion model, and generating the position of the pollution source of the to-be-detected underground water sample. By adopting the scheme of the application, the pollution source of the underground water can be quickly identified and positioned at low cost.

Description

Technical Field

[0001] This application relates to environmental monitoring, and more specifically, to groundwater environmental monitoring systems and methods. Background Technology

[0002] Groundwater pollution source tracing is a crucial aspect of environmental monitoring and remediation. With accelerated industrialization and urbanization, groundwater pollution is becoming increasingly complex, with diverse pollution sources and concealed migration paths. Traditional monitoring methods struggle to quickly and accurately identify pollution sources and their spatial distribution. Therefore, there is an urgent need to establish an efficient, precise, and practical groundwater pollution source tracing technology system to enable rapid identification, dynamic tracking, and accountability of pollution sources.

[0003] Current technologies for tracing groundwater pollution sources primarily rely on methods such as chemical analysis, geophysical exploration, and spectral comparison, including source apportionment through pollutant fingerprinting, geophysical imaging, or full-band spectral scanning. However, full-band scanning is time-consuming and complex to process; single fingerprinting techniques are susceptible to environmental interference and struggle to distinguish between multiple mixed pollution sources; and traditional physical exploration methods cannot directly correlate pollution source types with spectral characteristics, resulting in limited accuracy in tracing. Therefore, how to rapidly and cost-effectively identify and locate groundwater pollution sources has become a significant challenge for the industry. Summary of the Invention

[0004] This application provides a groundwater environmental monitoring system and method that can quickly and cost-effectively identify and locate groundwater pollution sources.

[0005] Firstly, this application provides a method for tracing the source of groundwater pollution, used by a groundwater environmental monitoring system to trace the source of water pollution, comprising the following steps: Three-dimensional fluorescence spectroscopy full-band scanning was performed on reference samples of known types of pollution sources in the target monitoring area to obtain a three-dimensional fluorescence spectroscopy dataset; Parallel factor analysis was used to analyze the fluorescent components in the three-dimensional fluorescence spectral dataset to obtain multiple independent fluorescent characteristic components, and then the characteristic excitation wavelength corresponding to each fluorescent characteristic component was determined. The contribution of each characteristic excitation wavelength to the classification of groundwater pollution sources in the target monitoring area is evaluated, and then the characteristic excitation wavelength combination for tracing the pollution sources of groundwater in the target monitoring area is selected. Based on the aforementioned characteristic excitation wavelength combination, the groundwater sample to be tested in the target monitoring area is scanned at a fixed wavelength to determine the preliminary type of pollution source; The relative contribution of the pollution source type of the groundwater sample to be tested is determined by using a preset pollution source fluorescent fingerprint database and the preliminary type. Then, the dynamic migration trend of the pollution plume is simulated based on the groundwater migration and diffusion model to generate the pollution source location of the groundwater sample to be tested.

[0006] In some embodiments, a three-dimensional fluorescence spectral full-band scan is performed on reference samples of known types of pollution sources in the target monitoring area to obtain a three-dimensional fluorescence spectral dataset, specifically including: Collect reference samples of known types of pollution sources within the target monitoring area; The pollution source reference samples are classified, numbered, and stored according to the preset pollution source categories; The reference sample of the pollution source was scanned in three dimensions across the entire wavelength range using a fluorescence spectrometer to obtain the original excitation-emission matrix data. The original excitation-emission matrix data is preprocessed and spectral normalized to obtain a three-dimensional fluorescence spectral dataset.

[0007] In some embodiments, parallel factor analysis is used to analyze the fluorescence components in the three-dimensional fluorescence spectral dataset to obtain multiple independent fluorescence feature components, specifically including: The three-dimensional fluorescence spectrum dataset is reconstructed along the sample dimensions; Set the component range of the reconstructed 3D fluorescence spectroscopy dataset; Within the range of the group fractions, parallel factor modeling and kernel consistency analysis are performed sequentially on the reconstructed three-dimensional data array to obtain the optimal fluorescence feature group fractions that represent the data and are mutually independent. Parallel factor analysis was performed on the reconstructed three-dimensional data array using the optimal fluorescence feature component fractions to extract multiple independent fluorescence feature components.

[0008] In some embodiments, determining the characteristic excitation wavelength corresponding to each fluorescent characteristic component specifically includes: Determine the excitation spectral loading vector for each fluorescence feature group; Determine the characteristic excitation wavelength corresponding to each excitation spectral loading vector at the maximum loading peak.

[0009] In some embodiments, evaluating the contribution of each characteristic excitation wavelength to the classification of groundwater pollution sources in the target monitoring area specifically includes: Extract the emission spectral intensity data of the pollution source reference sample at various characteristic excitation wavelengths; Based on the mapping relationship between spectral characteristics and pollution source categories, all emission spectral intensity data are evaluated and classified to obtain spectral discrimination criteria for different pollution source categories; Based on the aforementioned spectral discrimination criteria, the classification contribution of each characteristic excitation wavelength in distinguishing groundwater pollution sources in the target monitoring area is determined.

[0010] In some embodiments, the characteristic excitation wavelength combination for tracing the pollution sources of groundwater in the target monitoring area specifically includes: Determine the Pearson correlation coefficient of the emission spectral fragments corresponding to any two characteristic excitation wavelengths on the three-dimensional fluorescence spectral dataset; Based on all Pearson correlation coefficients, each characteristic excitation wavelength is divided into multiple information redundancy groups; By filtering and eliminating the characteristic excitation wavelengths in all information redundancy groups based on all classification contributions, the characteristic excitation wavelength combination for tracing the pollution sources of groundwater in the target monitoring area is obtained.

[0011] In some embodiments, performing fixed-wavelength scanning on groundwater samples from the target monitoring area based on the characteristic excitation wavelength combination to determine the preliminary type of pollution source specifically includes: Collect groundwater samples from the target monitoring area; The groundwater sample to be tested is scanned at a fixed wavelength under the characteristic excitation wavelength combination to obtain the characteristic fluorescence spectrum data of the groundwater sample to be tested. The preliminary type of pollution source is determined based on the characteristic fluorescence spectral data.

[0012] In some embodiments, determining the relative contribution of the pollution source type of the groundwater sample to be tested using a preset pollution source fluorescent fingerprint database and the preliminary type specifically includes: Retrieve the standard spectral vector corresponding to the preliminary type from the pollution source fluorescent fingerprint database; Determine multiple weighted matching degrees between the standard spectral vector and the characteristic fluorescence spectral data of the groundwater sample to be tested; The relative contribution of the pollution source type of the groundwater sample to be tested is determined based on all weighted matching degrees.

[0013] In some embodiments, the dynamic migration trend of the pollution plume is simulated based on a groundwater migration and diffusion model to generate the pollution source location of the groundwater sample to be tested, specifically including: Determine the groundwater migration and diffusion model for the target monitoring area; Using the currently detected contaminated groundwater sampling points as constraints, a backtracking simulation is performed in the groundwater migration and diffusion model to obtain the backtracking trajectory of the pollution plume and the distribution area of ​​potential pollution sources. By combining the fluorescent fingerprint database of pollution sources with the characteristics of the water sample to be tested, candidate pollution sources are screened and verified to obtain the location of the pollution source of the groundwater sample to be tested.

[0014] Secondly, this application provides a groundwater environment monitoring system, which includes a water pollution source tracing unit, the water pollution source tracing unit comprising: The acquisition module is used to perform three-dimensional fluorescence spectrum full-band scanning on reference samples of known types of pollution sources in the target monitoring area to obtain a three-dimensional fluorescence spectrum dataset; The processing module is used to perform feature analysis on the fluorescence components in the three-dimensional fluorescence spectrum dataset using parallel factor analysis to obtain multiple independent fluorescence feature components, and then determine the characteristic excitation wavelength corresponding to each fluorescence feature component. The processing module is also used to evaluate the contribution of each characteristic excitation wavelength to the classification of groundwater pollution sources in the target monitoring area, and then screen out the characteristic excitation wavelength combination for tracing the pollution sources of groundwater in the target monitoring area. The processing module is also used to perform fixed-wavelength scanning on the groundwater sample to be tested in the target monitoring area according to the characteristic excitation wavelength combination, so as to determine the preliminary type of pollution source; The execution module is used to determine the relative contribution of the pollution source type of the groundwater sample to be tested by using a preset pollution source fluorescent fingerprint database and the preliminary type, and then simulate the dynamic migration trend of the pollution plume according to the groundwater migration and diffusion model to generate the pollution source location of the groundwater sample to be tested.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: The groundwater environmental monitoring system and method provided in this application firstly perform a three-dimensional fluorescence spectrum full-band scan on reference samples of known types of pollution sources in the target monitoring area to obtain a three-dimensional fluorescence spectrum dataset. Parallel factor analysis is then used to analyze the characteristics of the fluorescence components in the three-dimensional fluorescence spectrum dataset, resulting in multiple independent fluorescence characteristic components. The characteristic excitation wavelength corresponding to each fluorescence characteristic component is then determined. The contribution of each characteristic excitation wavelength to the classification of groundwater pollution sources in the target monitoring area is evaluated, and a combination of characteristic excitation wavelengths for tracing the pollution sources of groundwater in the target monitoring area is then selected. Based on the combination of characteristic excitation wavelengths, a fixed-wavelength scan is performed on the groundwater sample to be tested in the target monitoring area to determine the preliminary type of pollution source. The relative contribution of the pollution source type of the groundwater sample to be tested is determined using a preset pollution source fluorescence fingerprint database and the preliminary type. Then, the dynamic migration trend of the pollution plume is simulated based on a groundwater migration and diffusion model to generate the pollution source location of the groundwater sample to be tested.

[0016] Therefore, in the groundwater pollution source tracing method of this application, firstly, a three-dimensional fluorescence spectrum full-band scan is performed on reference samples of known types of pollution sources in the target monitoring area to obtain a three-dimensional fluorescence spectrum dataset; parallel factor analysis is used to analyze the characteristics of the fluorescence components in the three-dimensional fluorescence spectrum dataset to obtain multiple independent fluorescence characteristic components, and then the characteristic excitation wavelength corresponding to each fluorescence characteristic component is determined; wherein, the characteristic excitation wavelength refers to the specific excitation wavelength value corresponding to the maximum load peak on the excitation spectral load vector of an independent fluorescence characteristic component, which is used to characterize the key site where the corresponding fluorescence component is most effectively excited, so as to improve the most significant spectral fingerprint information for subsequent differentiation of different pollution source types; secondly, the contribution of each characteristic excitation wavelength to the classification of groundwater pollution sources in the target monitoring area is evaluated, and then selected to obtain This method employs a combination of characteristic excitation wavelengths for tracing pollution sources in groundwater within a target monitoring area. By scanning the groundwater sample using only this combination of wavelengths, it avoids the traditional method of continuous full-band scanning. Only the spectral slices most relevant to identifying the target pollution source are acquired, allowing for the rapid assembly of a complete characteristic fluorescence spectrum data matrix of the groundwater sample, resulting in an order-of-magnitude improvement in detection efficiency. Finally, a fixed-wavelength scan is performed on the groundwater sample in the target monitoring area based on the aforementioned characteristic excitation wavelength combination to determine the preliminary type of pollution source. The relative contribution of the pollution source type to the groundwater sample is determined using a pre-set pollution source fluorescence fingerprint database and the preliminary type. Then, the dynamic migration trend of the pollution plume is simulated based on a groundwater migration and diffusion model to generate the pollution source location of the groundwater sample. This method enables rapid and low-cost identification and location of groundwater pollution sources. Attached Figure Description

[0017] Figure 1 This is an exemplary flowchart of a groundwater pollution tracing method according to some embodiments of this application.

[0018] Figure 2 This is an exemplary flowchart illustrating the determination of fluorescent characteristic components according to some embodiments of this application.

[0019] Figure 3 This is an exemplary flowchart illustrating the determination of a preliminary type of pollution source according to some embodiments of this application.

[0020] Figure 4 This is a schematic diagram of the structure of a water pollution tracing unit according to some embodiments of this application.

[0021] Figure 5 This is a schematic diagram of the structure of a computer device for implementing a groundwater pollution source tracing method according to some embodiments of this application. Detailed Implementation

[0022] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0023] refer to Figure 1 The figure is an exemplary flowchart of a groundwater pollution tracing method according to some embodiments of this application. The groundwater pollution tracing method mainly includes the following steps: In step 101, a three-dimensional fluorescence spectrum full-band scan is performed on reference samples of known types of pollution sources in the target monitoring area to obtain a three-dimensional fluorescence spectrum dataset.

[0024] In some embodiments, the following steps can be used to perform a three-dimensional fluorescence spectral full-band scan on reference samples of known types of pollution sources in the target monitoring area to obtain a three-dimensional fluorescence spectral dataset: Collect reference samples of known types of pollution sources within the target monitoring area; The pollution source reference samples are classified, numbered, and stored according to the preset pollution source categories; The reference sample of the pollution source was scanned in three dimensions across the entire wavelength range using a fluorescence spectrometer to obtain the original excitation-emission matrix data. The original excitation-emission matrix data is preprocessed and spectral normalized to obtain a three-dimensional fluorescence spectral dataset.

[0025] In practice, the collection of reference samples of known types of pollution sources within the target monitoring area can be achieved in the following way: First, based on environmental survey data within the target monitoring area, raw water samples from all potential pollution sources within the target area need to be collected to obtain reference samples of known types of pollution sources within the target monitoring area. Potential pollution sources can be, for example, designated chemical plant discharge outlets, municipal sewage pipe network manholes, landfill leachate collection ponds, etc. The aforementioned pollution source reference samples refer to a collection of different and clearly defined pollution source samples collected from the direct discharge outlets or representative locations of various potential pollution sources within the target monitoring area. Second, the coordinates, time, and surrounding environment of the sampling points are recorded simultaneously during collection. The classification, numbering, and storage of the pollution source reference samples according to preset pollution source categories can be achieved in the following way: In the laboratory, in accordance with the "Technical Regulations for the Preservation and Management of Water Quality Samples" (HJ... (493-2009) Preprocessing of pollution source reference samples, based on the industry attributes of the pollution source reference samples, such as chemical fiber dyeing and finishing, "metal surface treatment", pre-coded and stored in information; other methods may be used in other embodiments, which are not limited here.

[0026] In practice, the original excitation-emission matrix data is obtained by performing a three-dimensional full-band fluorescence spectral scan on the pollution source reference sample using a fluorescence spectrometer. This can be achieved as follows: The pollution source reference sample is brought to room temperature and shaken well. Then, under dark conditions, a blank scan is performed using ultrapure water to subtract the instrument background and the Raman scattering background of the water. Next, a steady-state fluorescence spectrometer is used to scan the pollution source reference sample, which is sequentially placed in a standard quartz cuvette. The instrument software automatically records and outputs a two-dimensional data table with excitation and emission wavelengths as the x and y axes and fluorescence intensity as the matrix values; this is the original excitation-emission matrix. The original excitation-emission matrix data is then subjected to standardization preprocessing and... Spectral normalization to obtain a three-dimensional fluorescence spectral dataset can be achieved as follows: First, scattering correction is performed on the original excitation-emission matrix data to remove optical interference. Second, the excitation source intensity correction factor curve and detector quantum efficiency response curve provided with the fluorescence spectrometer at the factory are used to perform point-by-point calculation and correction on each data point in the original excitation-emission matrix data to eliminate instrument system errors. Finally, a characteristic fluorescence region that exists and is stable for all reference samples is selected, and the spectral data of all samples are scaled to a relative intensity scale based on common fluorescence characteristics to finally obtain the three-dimensional fluorescence spectral dataset. Other methods can also be used in other embodiments, which are not limited here.

[0027] It should be noted that the three-dimensional fluorescence spectroscopy dataset in this application refers to a collection of matrix data obtained by measuring with excitation wavelength, emission wavelength, and fluorescence intensity as three dimensions. It includes three-dimensional fluorescence spectral data of multiple known types of pollution source samples within the target monitoring area after preprocessing, and an equally spaced excitation wavelength sequence and an emission wavelength sequence determined by scanning parameters, which serve as coordinate axes of the data in the excitation and emission dimensions, respectively. The three-dimensional fluorescence spectroscopy dataset is used to characterize the fluorescence characteristics of pollutants in the pollution source reference samples, which facilitates subsequent pollution source fingerprinting and tracing.

[0028] In step 102, parallel factor analysis is used to analyze the fluorescent components in the three-dimensional fluorescence spectral dataset to obtain multiple independent fluorescent characteristic components, and then the characteristic excitation wavelength corresponding to each fluorescent characteristic component is determined.

[0029] In some embodiments, reference Figure 2 As shown, this figure is an exemplary flowchart for determining fluorescent characteristic components in some embodiments of this application. In this embodiment, parallel factor analysis is used to analyze the fluorescent components in the three-dimensional fluorescence spectral dataset to obtain multiple independent fluorescent characteristic components. This can be achieved through the following steps: First, in step 1021, the three-dimensional fluorescence spectroscopy dataset is reconstructed along the sample dimension; Secondly, in step 1022, the component range of the reconstructed three-dimensional fluorescence spectrum dataset is set; Furthermore, in step 1023, parallel factor modeling and kernel consistency analysis are performed sequentially on the reconstructed three-dimensional data array within the range of the number of groups, thereby obtaining the optimal fluorescence feature groups that represent the data and are mutually independent. Finally, in step 1024, parallel factor analysis is performed on the reconstructed three-dimensional data array using the optimal fluorescence feature component fractions to extract multiple independent fluorescence feature components.

[0030] In specific implementation, the data reconstruction of the three-dimensional fluorescence spectrum dataset along the sample dimension can be achieved in the following way: each three-dimensional fluorescence spectrum data in the three-dimensional fluorescence spectrum dataset is regarded as a data slice. Using a scientific computing library, the two-dimensional data slices of all pollution source samples in the pollution source reference sample are stacked in the sample order to construct a three-dimensional data array. The array has three distinct dimensions: the first dimension represents the excitation wavelength, the second dimension represents the emission wavelength, and the third dimension represents different pollution source samples. The reconstructed three-dimensional fluorescence spectrum dataset is the standard input format for parallel factor analysis. Other methods can also be used in other embodiments, which are not limited here.

[0031] It should be noted that the core of the data reconstruction in this application is to ensure that the start and end wavelengths and intervals of the excitation and emission wavelength axes of all pollution source samples are completely consistent, so as to perform subsequent mathematical decomposition on the same numerical grid.

[0032] In specific implementation, the number range of components in the reconstructed three-dimensional fluorescence spectrum dataset can be set as follows: the number range of components in the reconstructed three-dimensional fluorescence spectrum dataset is set to 2 to 7. The preset reasonable number range of components is set based on the common quantity experience of fluorescent components of dissolved organic matter in environmental water samples and the principle of model interpretability, and is usually 2 to 7. Other methods can also be used in other embodiments, which are not limited here.

[0033] In specific implementation, parallel factor modeling and kernel consistency analysis are sequentially performed on the reconstructed three-dimensional data array within the specified number of components to obtain the optimal fluorescence feature components that represent the data independently. This can be achieved in the following way: using a kernel consistency analysis diagnostic tool, such as writing or calling a parallel factor analysis function in the chemometrics toolbox; secondly, iterative modeling is performed within the number of components: for each set number of components N, the algorithm performs alternating least squares optimization, decomposing the three-dimensional fluorescence spectral dataset into N trilinear components and outputting the corresponding excitation spectral load, emission spectral load, and sample relative concentration load; then, a high-dimensional array based on model residuals is constructed using the parallel factor analysis function to calculate the kernel consistency value of the model for each N value, and a line graph of the kernel consistency value changing with the number of components N is plotted, thereby selecting the N value before the kernel consistency value first falls below a preset threshold as the optimal fluorescence feature components of the reconstructed three-dimensional fluorescence spectral dataset; other methods can also be used in other embodiments, which are not limited here.

[0034] It should be noted that the optimal fluorescence characteristic component number in this application refers to the number of core components that can explain the spectral variation of all pollution source reference samples in the target monitoring area to the greatest extent with the simplest trilinear structure in parallel factor analysis; it is used to provide a quantitative model complexity standard for the entire source tracing method; the preset threshold is set to 95%; parallel factor analysis is based on the basic assumption of trilinearity, that is, any sample spectrum actually measured can be represented as a linear superposition of a finite number of pure component spectra. However, if the set component number exceeds its inherent true chemical component number, the model will forcibly decompose noise or minor variation into additional pseudo-components, resulting in model distortion. Therefore, kernel consistency analysis is used to judge that the closer the kernel consistency value is to 100%, the more the model conforms to the trilinear assumption and the more realistic the components are.

[0035] In specific implementation, parallel factor analysis is performed on the reconstructed three-dimensional data array using the optimal fluorescence feature component score to extract multiple independent fluorescence feature component scores. This can be achieved as follows: once the optimal fluorescence feature component score N is determined, the standard parallel factor analysis algorithm model in the chemometrics toolbox is called, using the reconstructed three-dimensional data array and the optimal fluorescence feature component score N as core input parameters. The three-dimensional data array has dimensions I×J×K, where I is the number of excitation wavelengths, J is the number of emission wavelengths, and K is the number of samples. This algorithm uses an alternating least squares optimization framework for solution, internally iteratively updating three load matrices until convergence: where the excitation wavelength... Each column of the spectral loading matrix represents the excitation spectral loading vector of a fluorescent feature component; each column of the emission spectral loading matrix represents the emission spectral loading vector of a component; each row of the sample score matrix represents the relative concentration distribution of a sample on each component; the termination condition of the algorithm can be set to the change of the sum of squared residuals of the model in two adjacent iterations being less than a preset threshold, such as 1e-6; the model finally outputs N sets of determined vectors: including N excitation spectral loading vectors, N emission spectral loading vectors, and the score of each sample on these N components; thus, the obtained N sets of vectors are all regarded as mutually independent fluorescent feature components; other methods can also be used in other embodiments, which are not limited here.

[0036] It should be noted that the multiple independent fluorescent characteristic components in this application refer to a set of mathematical vectors with stable spectral morphology, which are resolved from the three-dimensional fluorescence spectral data of all pollution source reference samples in the target monitoring area through parallel factor analysis. These vectors are used to characterize the fluorescence signals generated by pollutants in the samples under the excitation of a specific wavelength light source, and are excitation optical fingerprints that can stably distinguish different pollution source types. Each excitation spectral loading vector directly reflects the relative efficiency of the corresponding fluorescent component being excited at different excitation wavelengths, and its profile features (such as peak position and peak shape) represent the excitation spectral characteristics of the component.

[0037] In some embodiments, determining the characteristic excitation wavelength corresponding to each fluorescent characteristic component can be achieved by the following steps: Determine the excitation spectral loading vector of each fluorescent characteristic component; Determine the characteristic excitation wavelength corresponding to each excitation spectral loading vector at the maximum loading peak.

[0038] In specific implementation, the excitation spectral loading vector of each fluorescent characteristic component can be determined in the following way: the parallel factor analysis algorithm model includes an excitation loading matrix, an emission loading matrix, and a sample concentration matrix. Each column of the excitation loading matrix corresponds to the excitation spectral loading vector of a fluorescent characteristic component. These vectors can be obtained directly by calling the corresponding attributes of the model results to obtain the excitation spectral loading vector of each fluorescent characteristic group. Other methods can also be used in other embodiments, which are not limited here.

[0039] In specific implementation, the characteristic excitation wavelength corresponding to the maximum load peak value of each excitation spectral load vector can be determined in the following way: perform peak finding analysis on each excitation spectral load vector: traverse all numerical points in the vector, find its maximum load peak value, and record the specific wavelength value corresponding to the peak value on the excitation wavelength axis. Use this wavelength as the characteristic excitation wavelength of the excitation spectral load vector corresponding to the fluorescent characteristic component. Other methods can also be used in other embodiments, which are not limited here.

[0040] It should be noted that the characteristic excitation wavelength in this application refers to the specific excitation wavelength value corresponding to the maximum load peak on the excitation spectral load vector of an independent fluorescent characteristic component. It is used to characterize the key site where the corresponding fluorescent component is most effectively excited, so as to improve the most significant spectral fingerprint information for subsequent differentiation of different pollution source types.

[0041] In step 103, the contribution of each characteristic excitation wavelength to the classification of groundwater pollution sources in the target monitoring area is evaluated, and then the combination of characteristic excitation wavelengths for tracing the pollution sources of groundwater in the target monitoring area is selected.

[0042] In some embodiments, the assessment of the contribution of each characteristic excitation wavelength to the classification of groundwater pollution sources in the target monitoring area can be achieved by the following steps: Extract the emission spectral intensity data of the pollution source reference sample at various characteristic excitation wavelengths; Based on the mapping relationship between spectral characteristics and pollution source categories, all emission spectral intensity data are evaluated and classified to obtain spectral discrimination criteria for different pollution source categories; Based on the aforementioned spectral discrimination criteria, the classification contribution of each characteristic excitation wavelength in distinguishing groundwater pollution sources in the target monitoring area is determined.

[0043] In specific implementation, the emission spectral intensity data of the pollution source reference sample at each characteristic excitation wavelength can be extracted in the following way: when the fluorescence spectrometer scans the sample, an array of equally spaced wavelength values ​​is generated based on the set excitation wavelength scanning range and step size, and this array is used as the full-spectrum excitation wavelength sequence; secondly, for each characteristic excitation wavelength, the index of the characteristic excitation wavelength in the full-spectrum excitation wavelength sequence is obtained, and all samples can be sliced. Therefore, for any sample, its three-dimensional data is extracted along the entire emission wavelength dimension when the excitation wavelength dimension is the index in the full-spectrum excitation wavelength sequence, thereby obtaining an emission spectral curve. By traversing all characteristic excitation wavelengths and all pollution source samples, a two-dimensional feature matrix is ​​finally constructed. Each row of this matrix represents the splicing of the emission spectral information of a pollution source sample at all characteristic wavelengths, thereby using this matrix as the emission spectral intensity data of the pollution source reference sample at each characteristic excitation wavelength; other methods can also be used in other embodiments, which are not limited here.

[0044] In specific implementation, all emission spectral intensity data are evaluated and classified according to the mapping relationship between spectral features and pollution source categories. The spectral discrimination criteria for distinguishing different pollution source categories can be achieved in the following way: A support vector machine classifier is used to divide the emission spectral intensity data and the corresponding pollution source sample category label vectors into training and test sets, thereby initializing an SVC (Support Vector Classification) model. A radial basis function is selected as the kernel function to map the data to a high-dimensional space. The penalty coefficient and kernel function coefficient of the model are searched in a preset log space parameter grid through grid search cross-validation, with the classification accuracy on the validation set as the optimization objective, to find the optimal combination of penalty coefficient and kernel function coefficient parameters. Then, the model is retrained on the complete training set using the optimal parameter combination. The trained SVM model is the spectral discrimination criteria for distinguishing different pollution source categories. Other methods can also be used in other embodiments, which are not limited here.

[0045] It should be noted that the spectral discrimination criterion refers to the basis for judging the spectral information of pollutants, which is instantiated as a training model in this application. In the original emission spectrum data, the sample points of different pollution source categories may be nonlinearly mixed in a high-dimensional feature space. The core function of the radial basis function kernel is to map the data to a higher-dimensional feature space by calculating the similarity distance between sample points. In this new space, the originally complex nonlinear classification problem is transformed into a linear problem of finding a linear hyperplane to separate sample points of different categories. This captures and solidifies the essential difference pattern between the emission spectrum characteristics of different pollution source categories, providing a reliable judgment standard for the subsequent accurate classification of new unknown water samples.

[0046] In specific implementation, the classification contribution of each characteristic excitation wavelength in distinguishing groundwater pollution sources in the target monitoring area based on the spectral discrimination criteria can be achieved in the following way: First, based on the emission spectral data of all characteristic excitation wavelengths and known pollution source labels, a random forest classification model is trained, and its baseline classification accuracy on the validation set is recorded. Then, all spectral data columns corresponding to each characteristic excitation wavelength are randomly permuted in turn to destroy their association with the labels, and the model performance is re-evaluated. At the same time, the decrease in model accuracy before and after the characteristic permutation is calculated. The decrease value is the contribution of the characteristic excitation wavelength to the classification. Finally, the decrease values ​​calculated for all characteristic excitation wavelengths are normalized to obtain the contribution of each characteristic excitation wavelength in distinguishing groundwater pollution sources in the target monitoring area. Other methods can also be used in other embodiments, which are not limited here.

[0047] It should be noted that the classification contribution in this application is a parameter value that characterizes the spectral information carried by the feature excitation wavelength and its importance and usefulness in distinguishing different pollution source types. A random forest classification model is trained using spectral data under all feature wavelengths, and its baseline recognition accuracy is recorded. Then, for a single feature wavelength, its corresponding spectral data in all samples is randomly shuffled, thereby severing the inherent connection between the spectral feature and the actual pollution source category. Subsequently, the model performance is re-evaluated using this disrupted data. The magnitude of the decrease in model accuracy directly proves the uniqueness and importance of the information carried by the feature wavelength, i.e., the greater the decrease, the greater the contribution.

[0048] In some embodiments, the characteristic excitation wavelength combination for tracing the pollution sources of groundwater in the target monitoring area can be obtained by the following steps: Determine the Pearson correlation coefficient of the emission spectral fragments corresponding to any two characteristic excitation wavelengths on the three-dimensional fluorescence spectral dataset; Based on all Pearson correlation coefficients, each characteristic excitation wavelength is divided into multiple information redundancy groups; By filtering and eliminating the characteristic excitation wavelengths in all information redundancy groups based on all classification contributions, the characteristic excitation wavelength combination for tracing the pollution sources of groundwater in the target monitoring area is obtained.

[0049] In specific implementation, dividing each characteristic excitation wavelength into multiple information redundancy groups based on all Pearson correlation coefficients can be achieved in the following way: First, calculate the Pearson correlation coefficients between all pairs of characteristic excitation wavelengths to form a symmetrical correlation matrix. Then, using a hierarchical clustering algorithm, each characteristic excitation wavelength is initially treated as an independent cluster. The dissimilarity between any two clusters is calculated, and then the two clusters with the smallest dissimilarity (i.e., the highest correlation) are iteratively merged until the dissimilarity between all clusters is greater than a preset redundancy threshold. In this application, the preset redundancy threshold is set to 0.9. Thus, by analyzing the clustering dendrogram, characteristic excitation wavelengths with pairwise Pearson correlation coefficients not less than 0.9 are grouped into multiple internally similar information redundancy groups. The formula for calculating dissimilarity is: Dissimilarity = |1 - Correlation Coefficient|. An information redundancy group refers to a set of characteristic excitation wavelengths whose pollution source identification information is largely repeated. Other methods can also be used in other embodiments, which are not limited here.

[0050] In specific implementation, the characteristic excitation wavelengths in all information redundancy groups are screened and eliminated based on their classification contribution, thus obtaining the characteristic excitation wavelength combination for tracing the pollution source of groundwater in the target monitoring area. This can be achieved in the following way: within each information redundancy group, the classification contribution of each characteristic excitation wavelength is compared, and only the characteristic excitation wavelength with the highest classification contribution is retained as the representative of that group; subsequently, the characteristic excitation wavelengths represented by all information redundancy groups are combined to form the characteristic excitation wavelength combination for tracing the pollution source of groundwater in the target monitoring area; other methods can also be used in other embodiments, which are not limited here. It should be noted that the characteristic excitation wavelength combination in this application refers to the combination of excitation wavelengths with the highest specific contribution when distinguishing all potential pollution source types in the target monitoring area, which is used to reflect the chemical composition characteristics of pollutants in the area.

[0051] In step 104, the groundwater sample to be tested in the target monitoring area is scanned at a fixed wavelength according to the characteristic excitation wavelength combination to determine the preliminary type of pollution source; In some embodiments, reference Figure 3 As shown, this figure is an exemplary flowchart for determining the preliminary type of a pollution source in some embodiments of this application. In this embodiment, the determination of the preliminary type of the pollution source by scanning the groundwater sample to be tested in the target monitoring area with a fixed wavelength according to the characteristic excitation wavelength combination can be achieved by the following steps: First, in step 1041, groundwater samples to be tested are collected from the target monitoring area; Secondly, in step 1042, the groundwater sample to be tested is subjected to a fixed wavelength scan under the characteristic excitation wavelength combination to obtain the characteristic fluorescence spectrum data of the groundwater sample to be tested. Finally, in step 1043, the preliminary type of pollution source is determined based on the characteristic fluorescence spectral data.

[0052] In practice, the collection of groundwater samples from the target monitoring area can be achieved in the following manner: according to the hydrogeological survey and monitoring plan, at the groundwater observation point in the target monitoring area, a disposable sterile sampler is used to collect water samples in accordance with the "Technical Specification for Groundwater Environmental Monitoring" (HJ 164-2020) to obtain the groundwater samples to be tested; at the same time, the coordinates of the sampling point, water temperature, pH, conductivity and other parameters are recorded; in other embodiments, other methods can also be used, which are not limited here.

[0053] In specific implementation, the characteristic fluorescence spectral data of the groundwater sample to be tested can be obtained by performing a fixed wavelength scan under the characteristic excitation wavelength combination on the groundwater sample to be tested in the following way: the pre-treated groundwater sample to be tested is placed in the sample chamber of the fluorescence spectrometer, and in the instrument control software, only the monochromator is set and driven to switch to each specific wavelength value in the characteristic excitation wavelength combination in sequence. Under each wavelength, the complete emission spectrum is scanned and recorded, thereby obtaining the characteristic fluorescence spectral data of the groundwater sample to be tested; wherein, the characteristic fluorescence spectral data refers to the structured fluorescence response dataset obtained by exciting the groundwater sample to be tested under each characteristic excitation wavelength; other methods can also be used in other embodiments, which are not limited here.

[0054] It should be noted that by scanning the groundwater sample under test using only the combination of characteristic excitation wavelengths, the traditional full-band continuous scanning is avoided. Only the spectral slices most relevant to the identification of the target pollution source are obtained, so that a complete characteristic fluorescence spectrum data matrix of the groundwater sample under test can be assembled in a very short time, achieving an order-of-magnitude improvement in detection efficiency.

[0055] In specific implementation, determining the preliminary type of pollution source based on the characteristic fluorescence spectral data can be achieved in the following way: the characteristic fluorescence spectral data is expanded into a one-dimensional feature vector and directly input into a pre-trained support vector machine classification model. This classification model is trained based on data from local pollution source reference samples of known types at the same characteristic wavelength. The model uses internal kernel and decision functions to calculate the similarity in high-dimensional space between the spectral characteristics of the groundwater sample to be tested and the templates of the spectral fingerprints of each type of pollution source, thereby outputting a probabilistic classification result, for example: chemical wastewater: 85%, domestic sewage: 15%. The category with the highest probability is then taken as the preliminary type of pollution source for the groundwater sample to be tested. Other methods can also be used in other embodiments, which are not limited here.

[0056] In step 105, the relative contribution of the pollution source type of the groundwater sample to be tested is determined by the preset pollution source fluorescent fingerprint database and the preliminary type. Then, the dynamic migration trend of the pollution plume is simulated according to the groundwater migration and diffusion model to generate the pollution source location of the groundwater sample to be tested.

[0057] In some embodiments, determining the relative contribution of the pollution source type of the groundwater sample to be tested by using a preset pollution source fluorescent fingerprint database and the preliminary type can be achieved through the following steps: Retrieve the standard spectral vector corresponding to the preliminary type from the pollution source fluorescent fingerprint database; Determine multiple weighted matching degrees between the standard spectral vector and the characteristic fluorescence spectral data of the groundwater sample to be tested; The relative contribution of the pollution source type of the groundwater sample to be tested is determined based on all weighted matching degrees.

[0058] In specific implementation, determining the weighted matching degrees of the standard spectral vector and the characteristic fluorescence spectral data of the groundwater sample to be tested can be achieved in the following way: the characteristic fluorescence spectral data of the sample to be tested is used as the dependent variable vector, and the standard spectral vectors of all selected pollution source types in the pollution source fluorescence fingerprint database are used as the independent variable matrix. Then, a multiple linear regression model is used to calculate the weighted matching degree for fitting, and then the least squares method is used to estimate the regression coefficient vector. Each regression coefficient vector obtained represents the weight of the corresponding standard spectral vector in reconstructing the spectrum to be tested. All the obtained weights are used as the weighted matching degree of the characteristic fluorescence spectral data of the groundwater sample to be tested. The higher the weighted matching degree, the more significant the fingerprint characteristics of the pollution source in the spectrum to be tested. Other methods can also be used in other embodiments, which are not limited here.

[0059] In specific implementation, the relative contribution of the pollution source type of the groundwater sample to be tested can be determined based on all weighted matching degrees in the following way: the weighted matching degrees of each pollution source obtained by the above multiple linear regression are normalized to obtain a set of quantitative relative contributions; wherein, the relative contribution is a set of values ​​representing the proportion of each pollution source in the overall fluorescence signal of the groundwater sample to be tested; other methods can also be used in other embodiments, which are not limited here.

[0060] In some embodiments, the location of the pollution source in the groundwater sample to be tested can be generated by simulating the dynamic migration trend of the pollution plume based on a groundwater migration and diffusion model using the following steps: Determine the groundwater migration and diffusion model for the target monitoring area; Using the currently detected contaminated groundwater sampling points as constraints, a backtracking simulation is performed in the groundwater migration and diffusion model to obtain the backtracking trajectory of the pollution plume and the distribution area of ​​potential pollution sources. By combining the fluorescent fingerprint database of pollution sources with the characteristics of the water sample to be tested, candidate pollution sources are screened and verified to obtain the location of the pollution source of the groundwater sample to be tested.

[0061] In specific implementation, the current groundwater sampling point where pollution is detected is used as a constraint. A backtracking simulation is performed in the groundwater migration and diffusion model to obtain the pollution plume backtracking trajectory and the distribution area of ​​potential pollution sources. This can be achieved in the following way: Based on the constructed numerical model of groundwater flow and solute transport, the location of the currently detected pollution sampling point, the pollutant concentration, and the detection time are used as the model's inversion constraints. A virtual pollution source located upstream of the model area and conforming to hydrogeological conditions is set as a candidate release point. The inverse particle tracking algorithm or the adjoint state method is used to inversely solve for the most probable pollution release history and transport path under given constraints. Through multiple inversion calculations and uncertainty analysis, one or more pollution plume backtracking trajectories with the highest probability are finally output, and the distribution area of ​​one or more potential pollution sources upstream of the most probable pollutants is delineated. Other methods can also be used in other embodiments, which are not limited here.

[0062] In specific implementation, the location of the pollution source in the groundwater sample is determined by combining the pollution source fluorescence fingerprint database with the characteristics of the water sample to be tested for screening and verification of candidate pollution sources. This can be achieved in the following way: First, the electronic maps of the pollution source distribution area and the target area are spatially overlaid for analysis to initially screen out the candidate pollution sources located in the area whose industry type matches the preliminary identification and quantitative analysis results of the fluorescence spectrum. Then, the standard spectra of these candidate sources in the pollution source fluorescence fingerprint database are retrieved, and a refined secondary similarity calculation is performed on each one with the characteristics of the water sample to be tested. The candidate sources are then comprehensively scored based on information such as hydraulic connectivity and spatial distance with the pollution sampling point. The geographical coordinates of the candidate source with the highest comprehensive score and the best consistency of the evidence chain are determined as the location of the pollution source in the groundwater sample to be tested. Other methods can also be used in other embodiments, which are not limited here.

[0063] Furthermore, in another aspect of this application, in some embodiments, this application provides a groundwater environment monitoring system, which includes a water pollution source tracing unit, referencing... Figure 4 The figure is a schematic diagram of the structure of a water pollution tracing unit according to some embodiments of this application. The water pollution tracing unit includes: an acquisition module 401, a processing module 402, and an execution module 403, which are described below: The acquisition module 401 in this application is mainly used to perform three-dimensional fluorescence spectrum full-band scanning on reference samples of known types of pollution sources in the target monitoring area to obtain a three-dimensional fluorescence spectrum dataset. Processing module 402, in this application, is used to perform feature analysis on the fluorescence components in the three-dimensional fluorescence spectrum dataset using parallel factor analysis to obtain multiple independent fluorescence feature components, and then determine the characteristic excitation wavelength corresponding to each fluorescence feature component. It should be noted that the processing module 402 in this application is also used to evaluate the contribution of each characteristic excitation wavelength to the classification of groundwater pollution sources in the target monitoring area, and then to screen out the characteristic excitation wavelength combination for tracing the pollution sources of groundwater in the target monitoring area. Additionally, it should be noted that the processing module 402 in this application is also used to perform fixed-wavelength scanning on the groundwater sample to be tested in the target monitoring area according to the characteristic excitation wavelength combination, so as to determine the preliminary type of pollution source. The execution module 403 in this application is mainly used to determine the relative contribution of the pollution source type of the groundwater sample to be tested by using a preset pollution source fluorescent fingerprint database and the preliminary type, and then to simulate the dynamic migration trend of the pollution plume based on the groundwater migration and diffusion model to generate the pollution source location of the groundwater sample to be tested.

[0064] In addition, this application also provides a computer device, the computer device including a memory and a processor, the memory storing code, and the processor being configured to acquire the code and execute the above-described groundwater pollution tracing method.

[0065] In some embodiments, reference Figure 5 The figure is a schematic diagram of the structure of a computer device for implementing a groundwater pollution source tracing method according to some embodiments of this application. The groundwater pollution source tracing method in the above embodiments can... Figure 5 The computer device shown is used to implement this, and the computer device includes at least one processor 501, a communication bus 502, a memory 503, and at least one communication interface 504.

[0066] Processor 501 can be a general-purpose central processing unit (CPU) or an application-specific integrated circuit (ASIC).

[0067] The communication bus 502 can be used to transmit information between the aforementioned components.

[0068] Memory 503 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. Memory 503 may exist independently and be connected to processor 501 via communication bus 502. Memory 503 may also be integrated with processor 501.

[0069] The memory 503 stores program code for executing the scheme of this application, and its execution is controlled by the processor 501. The processor 501 executes the program code stored in the memory 503. The program code may include one or more software modules. The method used in the above embodiments can be implemented by the processor 501 and one or more software modules in the program code in the memory 503.

[0070] Communication interface 504 uses any transceiver-like device to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0071] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0072] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0073] In addition, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described groundwater pollution tracing method.

[0074] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0075] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A groundwater pollution tracing method for tracing water pollution by a groundwater environmental monitoring system, characterized by, The method comprises the following steps: The three-dimensional fluorescence spectrum data set of the reference sample of the known type of pollution source in the target monitoring area is obtained by performing full-band scanning on the three-dimensional fluorescence spectrum of the reference sample; The characteristic analysis of the fluorescence components in the three-dimensional fluorescence spectrum data set is performed by using parallel factor analysis method, and a plurality of mutually independent fluorescence characteristic components are obtained, and then the characteristic excitation wavelength corresponding to each fluorescence characteristic component is determined; The classification contribution degree of each characteristic excitation wavelength to the groundwater pollution source in the target monitoring area is evaluated, and then the characteristic excitation wavelength combination for the pollution source tracing of the groundwater in the target monitoring area is screened out; The preliminary type of the pollution source is determined by performing fixed wavelength scanning on the to-be-tested groundwater sample in the target monitoring area according to the characteristic excitation wavelength combination; The relative contribution degree of the pollution source type of the to-be-tested groundwater sample is determined through the preset pollution source fluorescence fingerprint database and the preliminary type, and then the dynamic migration trend of the pollution plume is simulated according to the groundwater migration and diffusion model, and the pollution source position of the to-be-tested groundwater sample is generated.

2. The method of claim 1, wherein, The three-dimensional fluorescence spectrum data set obtained by performing full-band scanning on the three-dimensional fluorescence spectrum of the reference sample of the known type of pollution source in the target monitoring area specifically comprises: The reference sample of the known type of pollution source in the target monitoring area is collected; The reference sample of the known type of pollution source in the target monitoring area is classified, numbered and stored according to the preset pollution source category; The original excitation emission matrix data are obtained by performing three-dimensional fluorescence spectrum full-band scanning on the reference sample of the known type of pollution source in the target monitoring area using a fluorescence spectrometer; The three-dimensional fluorescence spectrum data set is obtained by performing standardization preprocessing and spectral normalization on the original excitation emission matrix data.

3. The method of claim 1, wherein, The characteristic analysis of the fluorescence components in the three-dimensional fluorescence spectrum data set is performed by using parallel factor analysis method, and a plurality of mutually independent fluorescence characteristic components are obtained, and then the characteristic excitation wavelength corresponding to each fluorescence characteristic component is determined; The three-dimensional fluorescence spectrum data set is reconstructed along the sample dimension; The component number range of the reconstructed three-dimensional fluorescence spectrum data set is set; The optimal fluorescence characteristic component number is obtained by sequentially performing parallel factor modeling and kernel consistency analysis on the reconstructed three-dimensional data array within the component number range, and then the characteristic data and mutually independent optimal fluorescence characteristic components are obtained; The plurality of mutually independent fluorescence characteristic components are extracted by performing parallel factor analysis on the reconstructed three-dimensional data array through the optimal fluorescence characteristic component number.

4. The method of claim 1, wherein, The characteristic excitation wavelength corresponding to each fluorescence characteristic component is determined, which specifically comprises: The excitation spectrum load vector of each fluorescence characteristic group is determined; The characteristic excitation wavelength corresponding to each excitation spectrum load vector at the maximum load peak value is determined.

5. The method of claim 1, wherein, The classification contribution degree of each characteristic excitation wavelength to the groundwater pollution source in the target monitoring area is evaluated, which specifically comprises: The emission spectrum intensity data of the reference sample of the pollution source at each characteristic excitation wavelength are extracted; The emission spectrum intensity data of all the pollution source reference samples are evaluated and classified according to the mapping relationship between the spectral characteristics and the pollution source categories, and the spectral discrimination basis for distinguishing different pollution source categories is obtained; The classification contribution degree of each characteristic excitation wavelength in the process of distinguishing the groundwater pollution source in the target monitoring area is determined based on the spectral discrimination basis.

6. The method of claim 1, wherein, The characteristic excitation wavelength combination for tracing the pollution source of the underground water in the target monitoring area comprises: determining the Pearson correlation coefficients of the emission spectrum segments corresponding to any two characteristic excitation wavelengths on the three-dimensional fluorescence spectrum dataset; dividing the characteristic excitation wavelengths into multiple information redundancy groups according to all the Pearson correlation coefficients; screening and removing the characteristic excitation wavelengths in all the information redundancy groups through all the classification contribution degrees, and then obtaining the characteristic excitation wavelength combination for tracing the pollution source of the underground water in the target monitoring area.

7. The method of claim 1, wherein, The fixed wavelength scanning of the to-be-tested underground water sample in the target monitoring area according to the characteristic excitation wavelength combination to determine the preliminary type of the pollution source comprises: collecting the to-be-tested underground water sample in the target monitoring area; performing the fixed wavelength scanning of the to-be-tested underground water sample under the characteristic excitation wavelength combination to obtain the characteristic fluorescence spectrum data of the to-be-tested underground water sample; determining the preliminary type of the pollution source according to the characteristic fluorescence spectrum data.

8. The method of claim 1, wherein, The relative contribution degree of the pollution source type of the to-be-tested underground water sample is determined through the preset pollution source fluorescence fingerprint database and the preliminary type, which comprises: calling the standard spectrum vector corresponding to the preliminary type from the pollution source fluorescence fingerprint database; determining multiple weighted matching degrees of the standard spectrum vector and the characteristic fluorescence spectrum data of the to-be-tested underground water sample; determining the relative contribution degree of the pollution source type of the to-be-tested underground water sample according to all the weighted matching degrees.

9. The method of claim 1, wherein, The pollution source position of the to-be-tested underground water sample is generated by simulating the dynamic migration trend of the pollution plume according to the underground water migration and diffusion model, which comprises: determining the underground water migration and diffusion model of the target monitoring area; performing backtracking simulation in the underground water migration and diffusion model by taking the current detected contaminated underground water sampling point as a constraint condition to obtain the backtracking trajectory of the pollution plume and the potential pollution source distribution area; combining the pollution source fluorescence fingerprint database and the characteristics of the to-be-tested water sample to screen and verify the candidate pollution source, and obtaining the pollution source position of the to-be-tested underground water sample.

10. A groundwater environment monitoring system comprising a water pollution tracing unit, characterized by, The water pollution tracing unit comprises: an acquisition module configured to perform three-dimensional fluorescence spectrum full-band scanning on a reference sample of a known type of pollution source in a target monitoring area to obtain a three-dimensional fluorescence spectrum dataset; a processing module configured to perform characteristic analysis on fluorescence components in the three-dimensional fluorescence spectrum dataset by using a parallel factor analysis method to obtain multiple independent fluorescence characteristic components, and then determine characteristic excitation wavelengths corresponding to each fluorescence characteristic component; the processing module is further configured to evaluate the classification contribution degree of each characteristic excitation wavelength to the pollution source of the underground water in the target monitoring area, and then screen a characteristic excitation wavelength combination for tracing the pollution source of the underground water in the target monitoring area; the processing module is further configured to perform fixed wavelength scanning on a to-be-tested underground water sample in the target monitoring area according to the characteristic excitation wavelength combination to determine the preliminary type of the pollution source. The execution module is used for determining the relative contribution degree of the pollution source type of the to-be-tested groundwater sample through the preset pollution source fluorescence fingerprint database and the preliminary type, and then simulating the dynamic migration trend of the pollution plume according to a groundwater migration and diffusion model to generate the pollution source position of the to-be-tested groundwater sample.

Images