Surface water source tracing method fusing water quality concentration and chemical composition characteristics

Abstract

This invention relates to the field of quantitative source tracing, and discloses a quantitative method for surface water source tracing that integrates water quality concentration and chemical composition characteristics, providing a more accurate technical means for surface water source tracing. The method includes screening chemical characteristic indicators, constructing a pollution source chemical characteristic matrix after water sample analysis, calculating the degree of variation of each indicator and generating a dynamic weight vector, inputting this weight vector and the recipient water sample vector into a non-negative constraint optimization model to obtain a preliminary contribution ratio. Through multiple iterations with introduced random perturbations, the method performs statistical analysis and outputs quantitative source tracing results. The results are also used for compliance assessment to identify suspected pollution sources. Based on the source tracing results, a key monitoring list is generated, monitoring plans are optimized, and the source tracing results are updated. This invention integrates multi-dimensional characteristics, considers uncertainties, and forms a closed-loop management system, enabling precise source tracing of surface water pollution sources and providing a basis for environmental management.

Description

Technical Field

[0001] This invention relates to the field of source tracing and quantitative analysis, and more particularly to a method for source tracing and quantitative analysis of surface water that integrates water quality concentration and chemical composition characteristics. Background Technology

[0002] With rapid industrialization and urbanization, surface water pollution has become increasingly severe, posing a serious threat to the ecological environment and human health. Accurately identifying surface water pollution sources and their contribution ratios is a crucial prerequisite for developing effective pollution control strategies and improving water quality. Traditional surface water source tracing methods mainly rely on single water concentration analysis or simple chemical composition comparison; however, these methods face numerous challenges in practical applications.

[0003] Pollutants in surface water come from a wide range of sources and are complex. Pollutants emitted from different sources have a certain degree of similarity and overlap in chemical composition and concentration. It is difficult to accurately distinguish and quantify the contribution of each pollution source by relying on a single water quality concentration index. Multiple industrial pollution sources may discharge wastewater containing similar heavy metal ions. It is difficult to accurately determine the specific contribution of each pollution source by detecting heavy metal concentration alone.

[0004] While chemical composition characteristics can provide more information about pollution sources, different chemical indicators exhibit varying migration and transformation processes in the natural environment, resulting in different levels of stability and variability. Existing methods for tracing pollution sources using chemical composition characteristics often neglect the dynamic changes of these indicators and fail to fully consider the relative importance of each indicator in reflecting the characteristics of the pollution source, thus affecting the accuracy and reliability of the tracing results.

[0005] In actual sampling and analysis, measurement data inevitably contain certain errors and fluctuations due to various reasons such as instrument errors and environmental interference. Current technologies fail to fully consider the impact of these uncertainties on the source tracing results, making the obtained pollution source contribution ratios lack sufficient credibility and persuasiveness, and thus failing to meet the actual needs of environmental management and decision-making.

[0006] Therefore, we propose a quantitative method for tracing the source of surface water that integrates water quality concentration and chemical composition characteristics to address the above-mentioned problems. Summary of the Invention

[0007] This invention provides a quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics, which can provide a more accurate technical means for tracing the source of surface water.

[0008] The first aspect of this invention provides a quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics. This method includes: based on chemical analysis of surface water samples from potential pollution sources and target recipients, converting the measured concentrations of various chemical characteristic indicators into relative concentrations to construct a pollution source chemical characteristic matrix; based on the pollution source chemical characteristic matrix, calculating the degree of variation of each chemical characteristic indicator among different pollution sources, and generating a dynamic weight vector according to this degree of variation; inputting the pollution source chemical characteristic matrix, the standardized chemical characteristic vector of the recipient water sample, and the dynamic weight vector into a preset non-negative constraint optimization model to obtain a preliminary contribution ratio of each pollution source, which minimizes the overall deviation between the weighted chemical characteristics of each pollution source and the chemical characteristics of the recipient; introducing random perturbations within a preset range into the pollution source chemical characteristic matrix and the standardized chemical characteristic vector of the recipient water sample, and performing multiple iterative solutions to statistically analyze the preliminary contribution ratio of each pollution source and output quantitative source tracing results.

[0009] Optionally, in a first implementation of the first aspect of the present invention, the method includes: screening a set of chemical characteristic indicators including ion concentration ratios, stable isotope ratios, and proportions of characteristic organic components to form a set of characteristic indicators for source tracing analysis; collecting water samples from each potential pollution source and target receptor according to the set of characteristic indicators, and performing chemical analysis to obtain a table of original concentration data for each water sample under each indicator; processing the original concentration data table to generate a chemical analysis dataset; and for the chemical analysis dataset, dividing the concentration of each indicator for each pollution source water sample by the sum of the concentrations of all pollution sources at that indicator, converting the absolute concentration into a relative concentration characterizing its relative abundance at that indicator, collecting the relative concentration data of all pollution sources, and constructing a pollution source chemical characteristic matrix.

[0010] Optionally, in a second implementation of the first aspect of the present invention, the method includes: based on the pollution source chemical feature matrix, calculating the degree of dispersion of each chemical feature index among the pollution sources to obtain a set of values ​​including the degree of variation of each index; dividing each value in the set of values ​​of the degree of variation of each index by the sum of all values ​​in the set to convert the degree of variation of each index into a relative importance coefficient; and collecting the relative importance coefficients corresponding to all chemical feature indices, arranging them in index order to generate a dynamic weight vector.

[0011] Optionally, in the third implementation of the first aspect of the present invention, before the step of collecting the relative importance coefficients corresponding to all chemical characteristic indicators, arranging them in order of indicators, and generating a dynamic weight vector, the method further includes: querying and obtaining the stability level information of each chemical characteristic indicator during the migration process in the natural environment based on a preset indicator stability knowledge base, and generating a set of stability coefficients for each indicator; correcting the corresponding values ​​in the numerical set of the degree of variation of each indicator according to the values ​​of each indicator in the stability coefficient set, and generating a set of comprehensive degree of variation numerical values; and processing the set of comprehensive degree of variation numerical values ​​to convert them into comprehensive importance coefficients of each indicator for generating a dynamic weight vector.

[0012] Optionally, in a fourth implementation of the first aspect of the present invention, the method includes: constructing an objective function form based on the pollution source chemical feature matrix and the dynamic weight vector; setting constraints, including requiring that the contribution ratio of all pollution sources be non-negative and that the sum of all contribution ratios be one, to form a constraint parameter set; solving the objective function form under the constraint parameter set to obtain an initial contribution ratio vector; and processing the initial contribution ratio vector to generate a set of contribution ratios for each pollution source.

[0013] Optionally, in the fifth implementation of the first aspect of the present invention, in order to solve for the contribution ratios x1, x2, and x3 of the three pollution sources, a non-negative constraint optimization model is constructed, and an objective function is set. For the weighted sum of squared deviations:

[0014] ,

[0015] Among them, w j R represents the dynamic weight of the j-th indicator. j S is the observed value of the i-th index of the receptor water sample. ij Let be the characteristic value of the i-th pollution source on the j-th indicator. The contribution ratio of pollution sources is subject to the following constraints: and .

[0016] Optionally, in the sixth implementation of the first aspect of the present invention, after generating the preliminary set of contribution ratios of each pollution source that meets the proportional summation requirement, the method further includes: substituting the preliminary set of contribution ratios of each pollution source back into the chemical mass balance equation set, calculating the absolute difference between the predicted value and the measured value corresponding to each chemical characteristic indicator, and generating a mass balance residual vector; based on a preset allowable error range, performing compliance judgment on each residual value in the mass balance residual vector, and generating a set of balance state flags for each indicator; according to the set of balance state flags, if the balance state of all indicators is marked as compliant, then the preliminary set of contribution ratios of each pollution source is marked as a reliable solution; if there are indicators marked as non-compliant, then based on the weight coefficient in the dynamic weight vector corresponding to the non-compliant indicator, identifying and generating a list of suspicious pollution sources, which is used to prompt for review or supplementary sampling analysis of the characteristic data of the pollution sources in the list.

[0017] Optionally, in the seventh implementation of the first aspect of the present invention, the method includes: according to a preset analytical error range, superimposing independent random perturbation values ​​on the elements in the chemical feature matrix of the pollution source and the elements in the standardized chemical feature vector of the receptor water sample to generate multiple sets of simulated perturbation source matrices and receptor vectors; for each set of simulated perturbation source matrices and receptor vectors, re-executing the dynamic weight generation and optimization solution process to obtain the corresponding pollution source contribution ratio solution; by repeating this process multiple times, forming a pollution source contribution ratio solution sequence including multiple solution samples; performing statistical analysis on the proportion value corresponding to each pollution source in the pollution source contribution ratio solution sequence, calculating its statistical mean as the final point estimate of the pollution source contribution rate, and calculating its numerical interval at a specified confidence level as the confidence interval; and collecting the final point estimates and confidence intervals of all pollution sources to form a quantitative source tracing result report.

[0018] Optionally, in the eighth implementation of the first aspect of the present invention, the method further includes: based on the quantitative source tracing result report, identifying the main pollution sources whose contribution rate to the target receptor surface water exceeds a preset threshold, and generating a list of key monitored pollution sources; according to the list of key monitored pollution sources, adjusting or optimizing the sampling point layout scheme and monitoring frequency of the surface water environmental monitoring network in the target area to form an enhanced monitoring scheme; according to the enhanced monitoring scheme, implementing a new round of key pollution source and receptor water sample collection and analysis to obtain an updated chemical analysis dataset; using the updated chemical analysis dataset, updating the pollution source chemical feature matrix and receptor chemical feature vector, and re-executing the subsequent weight calculation, model solution and uncertainty analysis steps to output the updated quantitative source tracing result.

[0019] Optionally, in the ninth implementation of the first aspect of the present invention, after implementing a new round of key pollution source and receptor water sample collection and analysis according to the enhanced monitoring scheme to obtain an updated chemical analysis dataset, the method further includes: calculating the change in contribution rate of each major pollution source based on the updated chemical analysis dataset and the list of key monitored pollution sources, and generating a pollution contribution change trend analysis table; matching recommended treatment measures from a preset pollution treatment technology library to pollution sources with significantly increased contribution rates according to the pollution contribution change trend analysis table, forming a pollution source treatment priority and measure recommendation table; guiding environmental management departments to implement engineering or management intervention measures for pollution sources according to the pollution source treatment priority and measure recommendation table, and recording them to form a pollution control measure implementation log; collecting and analyzing the chemical characteristic data of the intervened pollution source and receptor water samples again at a preset period after the intervention measures are implemented, and generating a measure post-effect evaluation dataset; quantifying the actual effect of the treatment measures on reducing the contribution of specific pollution sources by comparing the measure post-effect evaluation dataset with the updated chemical analysis dataset, and outputting a treatment effect evaluation report.

[0020] The mechanism of this invention is as follows: It provides an iterative optimization tracing framework that systematically integrates the distinguishability of chemical fingerprints, the physical constraints of quality balance, and the quantification of uncertainty.

[0021] Beneficial effects: By integrating water quality concentration and chemical composition characteristics, multiple types of chemical characteristic indicators, including ion concentration ratio, stable isotope ratio and characteristic organic component ratio, are screened to characterize pollution source features from multiple dimensions. This effectively solves the limitations of single-indicator source tracing, greatly improves the accuracy of source tracing results, and can more accurately identify the contribution of each pollution source to the target receiver surface water.

[0022] The variation degree of each indicator is calculated based on the chemical feature matrix of the pollution source, and the variation degree is further corrected by considering the information of the indicator stability knowledge base. A dynamic weight vector is generated, which can reasonably allocate the indicator weight according to different situations, making the source tracing model more in line with the actual situation and enhancing the model's adaptability and flexibility to different pollution scenarios.

[0023] The chemical feature matrix of pollution sources, the standardized chemical feature vector of the receiver water sample, and the dynamic weight vector are input into a preset non-negative constraint optimization model. The initial contribution ratio is solved with the goal of minimizing the overall deviation between the weighted chemical features of each pollution source and the chemical features of the receiver. This not only ensures the non-negativity of the contribution ratio and the rationality of the ratio summation, but also makes the result more consistent with the principle of chemical mass balance.

[0024] By introducing random perturbations within a preset range, the chemical characteristic matrix of pollution sources and the standardized chemical characteristic vector of receptor water samples are solved iteratively multiple times. The resulting solution sequence of pollution source contribution ratios is statistically analyzed, and the statistical mean is calculated as the final point estimate and the confidence interval at a specified confidence level. This fully considers the uncertainty of the data, making the source tracing results more credible and persuasive, and providing more reliable decision-making basis for environmental management departments. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of an embodiment of the quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics in this invention.

[0026] Figure 2 This is a schematic diagram of another embodiment of the quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics in this invention. Detailed Implementation

[0027] This invention provides a quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics, offering a more accurate technical means for surface water source tracing. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0028] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics in this invention includes:

[0029] 101. Based on the chemical analysis of surface water samples from various potential pollution sources and target receptors, the measured concentrations of various chemical characteristic indicators are converted into relative concentrations to construct a chemical characteristic matrix of pollution sources.

[0030] It is understood that the executing entity of this invention can be a surface water source tracing and quantitative device that integrates water quality concentration and chemical composition characteristics, or it can be a terminal or a server; the specific implementation is not limited here. This embodiment of the invention will be described using a server as an example.

[0031] It should be noted that a river (receiving water body) in a certain area has exceeded the standards for nitrogen and phosphorus nutrients, and its source needs to be traced. Preliminary investigation has identified three main potential sources of pollution: A (domestic sewage discharge outlet), B (agricultural runoff collection point), and C (food processing plant drainage outlet).

[0032] Water samples were collected from pollution sources A, B, and C, as well as at the target receptor river section. Each sample underwent chemical analysis, determining the concentrations of a pre-selected set of chemical characteristic indicators. In this example, the selected indicators included: conventional nutrients: ammonia nitrogen (NH3). - N), nitrate nitrogen (NO3) - -N), total phosphorus (TP); ionic components: chloride ions (Cl-N), total phosphorus (TP); - ), sulfate ions (SO4) 2- Characteristic trace organic pollutants (PPCPs): caffeine, sulfamethoxazole (SMX);

[0033] The raw concentration data obtained after the measurement (unit: μg / L, PPCPs unit is ng / L) are shown in Table 1 below as an example:

[0034]

[0035] Concentration data standardization (converting to relative concentration) eliminates the influence of differences in dimensions and orders of magnitude among various indicators. The absolute concentration of each indicator in each pollution source is converted into the relative proportion (i.e., relative concentration) of the total concentration of all selected indicators within that pollution source. The calculation method is: for a specific indicator in a given pollution source, its relative concentration = (measured concentration of that indicator) / (sum of concentrations of all selected indicators in that pollution source).

[0036] Ammonia nitrogen (NH3) from pollution source A (domestic sewage) - Taking NH3 as an example: The sum of the concentrations of all indicators from source A = 25000 + 800 + 3500 + 40000 + 60000 + 0.45 + 0.12 ≈ 127300.57 (Note: PPCPs units are converted from ng / L to μg / L, i.e., 450 ng / L = 0.45 μg / L, 120 ng / L = 0.12 μg / L); - The relative concentration of N = 25000 / 127300.57≈0.196; the relative concentration of each indicator for each pollution source is calculated accordingly.

[0037] A pollution source chemical characteristic matrix is ​​constructed by organizing the relative concentration values ​​of all indicators for each pollution source obtained from the above calculation into a matrix form, namely the pollution source chemical characteristic matrix. The rows of this matrix correspond to different pollution sources, and the columns correspond to different chemical characteristic indicators.

[0038] The matrix constructed based on the example data is roughly as shown in Table 2 below (the values ​​are for illustrative purposes only):

[0039]

[0040] This matrix clearly reflects the chemical composition "fingerprint" characteristics of different pollution sources. Domestic sewage (A) shows a higher Cl... - SO4 2- The relative abundance of characteristic PPCPs (caffeine); agricultural runoff (B) was characterized by a high proportion of NO3. - -N is a significant feature; the relative concentration distribution of various indicators in food factory wastewater (C) falls between the two.

[0041] 102. Based on the chemical feature matrix of pollution sources, calculate the degree of variation of each chemical feature index among different pollution sources, and generate a dynamic weight vector according to the degree of variation, in which the index with stronger ability to distinguish pollution sources is given a higher weight.

[0042] It should be noted that, building upon the previously constructed chemical characteristic matrix of pollution sources, the rows of this matrix represent three pollution sources: A (domestic sewage), B (agricultural runoff), and C (food processing plant wastewater), while the columns represent the relative concentrations of each chemical characteristic indicator.

[0043] The core of this step is to calculate the degree of variability of each chemical characteristic indicator, assessing the magnitude of the difference between different pollution sources. The greater the difference in an indicator, the stronger its ability to distinguish between different pollution sources. Standard deviation or coefficient of variation is typically used to measure this degree of variability. Based on the relative concentration matrix obtained in the previous step, we calculate the coefficient of variation (the ratio of standard deviation to mean, eliminating the influence of dimensions) for each indicator among the three pollution sources A, B, and C. The calculation results are shown in Table 3 below.

[0044]

[0045] To generate a dynamic weight vector based on the degree of variation, the degree of variation of each indicator needs to be converted into weights. The basic principle is: the larger the coefficient of variation, the higher the weight, because it contributes more to distinguishing pollution sources. Baseline weight allocation: The coefficients of variation are normalized so that their sum is 1. Caffeine has the largest coefficient of variation (2.50), indicating that its concentration difference among the three pollution sources is very significant, making it a strong "fingerprint" indicator; while sulfate ions (SO4)... 2- The coefficient of variation of ) is relatively small (0.15), indicating that its content is relatively stable in each source and its ability to distinguish is weak.

[0046] Generate a weight vector: After calculation and adjustment, a weight vector is generated, where each weight value corresponds to a chemical characteristic index. This weight vector is "dynamic," meaning that if the monitored pollution source changes or new indicators are added, the weights need to be recalculated.

[0047] Based on the example data, the generated dynamic weight vectors are shown in Table 4 below (values ​​are for illustrative purposes only):

[0048]

[0049] 103. Input the chemical feature matrix of the pollution source, the standardized chemical feature vector of the receiver water sample, and the dynamic weight vector into a preset non-negative constraint optimization model, and solve to obtain a set of preliminary contribution ratios of each pollution source. This set of ratios minimizes the overall deviation between the weighted chemical features of each pollution source and the chemical features of the receiver.

[0050] It should be noted that this example continues to address the issue of excessive nitrogen and phosphorus nutrients in a certain river, and quantitatively traces three potential pollution sources (A: domestic sewage, B: agricultural runoff, and C: wastewater from a food processing plant). Before starting this step, the following three key input data were obtained:

[0051] Pollution Source Chemical Characteristic Matrix: This matrix contains the relative concentrations of multiple chemical characteristic indicators for each pollution source, used to characterize its "chemical fingerprint." A simplified version of the matrix is ​​shown in Table 5 below (values ​​are for example only):

[0052]

[0053] Standardized chemical characteristic vector of recipient water samples: The concentrations of chemical indicators in the target recipient river water samples have also been standardized to relative concentrations, forming a vector: [0.120, 0.280, 0.090, 0.220, 0.380, 0.0000015, 0.00000080] (in the same order as the indicators above). Dynamic weight vector: Calculated based on the degree of variation of each indicator among pollution sources, used to emphasize indicators with strong distinguishing ability: [0.11, 0.18, 0.05, 0.13, 0.02, 0.35, 0.16] (weight order is consistent with the indicator order, and the sum is 1).

[0054] The core of this step is to use a non-negativity-constrained optimization model to solve for the contribution ratio of a set of pollution sources. Model objective: To find a set of contribution ratios (denoted as source A proportion p). A B source ratio p B C source ratio p CThis approach minimizes the overall deviation between the pollution source "chemical fingerprint" synthesized by weighting these proportions and the receiver water sample's "chemical fingerprint." When calculating the deviation, the difference in each indicator is multiplied by its dynamic weight; the greater the weight of an indicator, the greater its impact on the overall deviation.

[0055] Constraints: The contribution ratio must satisfy two basic physical constraints: Non-negativity constraint: The contribution ratio of each pollution source must be greater than or equal to zero, i.e., p A ≥0,p B ≥0,p C ≥0. The sum must be 1: the sum of the contribution percentages of all pollution sources must equal 1 (i.e., 100%). This reflects the principle of conservation of mass. Solution process: The optimization algorithm in the server will automatically perform iterative calculations, continuously adjusting p. A ,p B ,p C Given the constraints mentioned above, we seek the ratio combination that minimizes the overall weighted deviation.

[0056] After optimization model calculations, the following preliminary contribution ratios were obtained: Pollution source A (domestic sewage) contribution ratio: 0.35 (35%); Pollution source B (agricultural runoff) contribution ratio: 0.52 (52%); Pollution source C (food processing plant) contribution ratio: 0.13 (13%). This indicates that, under the current model and input data, agricultural runoff (B) is preliminarily identified as the main source of nitrogen and phosphorus nutrient exceedances at the recipient river section, followed by domestic sewage (A). The contribution of the food processing plant (C) is relatively small.

[0057] 104. By introducing random perturbations within a preset range into the chemical characteristic matrix of pollution sources and the standardized chemical characteristic vector of the receptor water sample, and performing multiple iterative solutions, the preliminary contribution ratio of each pollution source is statistically analyzed, and quantitative source tracing results including the estimated contribution rate of each pollution source and its confidence interval are output.

[0058] It should be noted that, building upon the preliminary contribution ratios already obtained, the core of this step lies in assessing the uncertainty and stability of these preliminary results, thereby providing more reliable quantitative conclusions.

[0059] Introducing a pre-defined range of random perturbation acknowledges the unavoidable errors in actual monitoring data (instrument measurement errors, spatiotemporal variations in water samples, etc.). To assess the impact of these uncertainties on the source tracing results, the server introduces a pre-defined range of random perturbation for each relative concentration value in the pollution source chemical characteristic matrix and the receptor water sample standardized chemical characteristic vector. Perturbation method: The pre-defined perturbation range is set to ±5%. This means that each concentration value will independently undergo a small random increase or decrease; for example, if the original relative concentration of an indicator is 0.15, after perturbation it may become 0.152 (increase) or 0.143 (decrease). Purpose: By introducing this controllable randomness, to simulate the small fluctuations that may occur in actual measurements and observe whether the initial contribution ratio is sensitive to these fluctuations.

[0060] The above perturbation and solution process is not performed only once, but repeated hundreds or thousands of times (1000 iterations). Process: In each iteration, the server independently and randomly perturbs the original feature matrix and receptor vector, generating a "new" set of simulated data. This perturbated data, along with the dynamic weight vector determined in the previous steps, is then input again into the non-negative constraint optimization model for solution. The contribution ratios of pollution sources calculated in this iteration are recorded. Results: After 1000 iterations, we will obtain 1000 sets of slightly different contribution ratio results: the contribution ratio of pollution source A (domestic sewage) may be 36% in the first iteration, 34% in the second, ..., and 35% in the 1000th iteration. The contribution ratios of pollution source B (agricultural runoff) and pollution source C (food processing plant) also fluctuate slightly accordingly.

[0061] Statistical Analysis and Results Output: The server performs statistical analysis on these 1000 sets of results, providing an estimate and confidence interval for the contribution rate of each pollution source. Contribution Rate Estimate: Typically, the median or arithmetic mean of all iteration results is taken as the best estimate of the pollution source's contribution rate. The median contribution rate of pollution source A over 1000 iterations is likely 35.2%. Confidence Interval: A specific confidence interval (95% confidence interval) is calculated for the contribution rate of each pollution source. This indicates that 95% of the iteration results fall within this interval, quantifying the range of uncertainty in the contribution rate estimate. Based on exemplary data, the quantitative source tracing results are shown in Table 6 below:

[0062]

[0063] This result not only provides the best estimate of the contribution ratio of each pollution source, but more importantly, it reveals the accuracy of this estimate through confidence intervals. Agricultural runoff (B) is the main contributing source, and its confidence interval is relatively narrow, indicating that this conclusion is relatively stable and reliable. Although the contribution of food processing plants (C) is small, its confidence interval is relatively wide, suggesting that the uncertainty of this estimate is slightly greater.

[0064] Please see Figure 2 Another embodiment of the quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics in this invention includes:

[0065] 201. Based on the chemical analysis of surface water samples from various potential pollution sources and target receptors, the concentrations of various chemical characteristic indicators were converted into relative concentrations to construct a chemical characteristic matrix of pollution sources.

[0066] Specifically, a set of chemical characteristic indicators, including ion concentration ratios, stable isotope ratios, and proportions of characteristic organic components, is selected to form a set of characteristic indicators for source tracing analysis. Based on the set of characteristic indicators, water samples from each potential pollution source and target receptor are collected and chemically analyzed to obtain a table of original concentration data for each water sample under each indicator. Outliers and missing values ​​in the original concentration data table are processed to generate a chemical analysis dataset. For the chemical analysis dataset, the concentration of each indicator for each pollution source water sample is divided by the sum of the concentrations of all pollution sources at that indicator, converting the absolute concentration into a relative concentration characterizing its relative abundance at that indicator. The relative concentration data of all pollution sources are collected to construct a chemical characteristic matrix of pollution sources.

[0067] It should be noted that, taking the source tracing of a polluted river section as an example, three potential pollution sources were identified: textile dyeing wastewater (source 1), agricultural non-point source runoff (source 2), and urban domestic sewage (source 3).

[0068] A set of characteristic indicators was selected, including: chloride / sulfate ion ratio (indicator A), δ¹⁵N nitrogen stable isotope ratio (indicator B), and caffeine component ratio (indicator C). Chemical analysis was performed on water samples collected from the three pollution sources to obtain raw concentration data tables. At this stage, if indicator B data for source 2 was found to be missing, interpolation was performed using the mean of historical monitoring data from the same area to complete the data, generating a complete chemical analysis dataset.

[0069] The raw concentration data, after processing, are set as follows: Indicator A (ion ratio): Source 1 value is 10.0, Source 2 value is 40.0, Source 3 value is 50.0 (sum of three is 100.0); Indicator B (isotope): Source 1 value is 5.0, Source 2 value is 10.0, Source 3 value is 35.0 (sum of three is 50.0); Indicator C (organic matter): Source 1 value is 0.1, Source 2 value is 0.1, Source 3 value is 0.8 (sum of three is 1.0).

[0070] The relative concentration conversion and matrix construction, based on algorithmic logic, divide the value of each pollution source for each indicator by the sum of the values ​​of all pollution sources under that indicator, converting the absolute value into relative abundance:

[0071] For the calculation of index A: Relative concentration of source 1 = 10.0 divided by 100.0 = 0.10; Relative concentration of source 2 = 40.0 divided by 100.0 = 0.40; Relative concentration of source 3 = 50.0 divided by 100.0 = 0.50;

[0072] For the calculation of index B: relative concentration of source 1 = 5.0 divided by 50.0 = 0.10; relative concentration of source 2 = 10.0 divided by 50.0 = 0.20; relative concentration of source 3 = 35.0 divided by 50.0 = 0.70;

[0073] For the calculation of index C: relative concentration of source 1 = 0.10; relative concentration of source 2 = 0.10; relative concentration of source 3 = 0.80;

[0074] The above data were used to construct a chemical characteristic matrix of pollution sources (rows represent pollution sources, and columns represent the relative concentrations of indicators A, B, and C): Source 1 vector: [0.10, 0.10, 0.10]; Source 2 vector: [0.40, 0.20, 0.10]; Source 3 vector: [0.50, 0.70, 0.80]; This matrix eliminates the differences in dimensions and orders of magnitude between different indicators, thus fulfilling the construction requirements of step 201.

[0075] 202. Based on the chemical feature matrix of pollution sources, calculate the degree of variation of each chemical feature index among different pollution sources, and generate a dynamic weight vector according to the degree of variation, in which the index with stronger ability to distinguish pollution sources is given a higher weight.

[0076] Specifically, based on the chemical characteristic matrix of pollution sources, the dispersion measure of each chemical characteristic index among the pollution sources is calculated to obtain a set of values ​​including the degree of variation of each index. Each value in the set of values ​​of the degree of variation of each index is divided by the sum of all values ​​in the set to complete the normalization process, thereby converting the degree of variation of each index into a relative importance coefficient between 0 and 1. The relative importance coefficients corresponding to all chemical characteristic indices are collected, arranged in order of index, and a dynamic weight vector is generated for subsequent optimization model calculation.

[0077] Furthermore, before aggregating the relative importance coefficients of all chemical characteristic indicators, arranging them in order of indicators, and generating a dynamic weight vector, the process includes: querying and obtaining the stability level information of each chemical characteristic indicator during migration in the natural environment based on a pre-set indicator stability knowledge base, and generating a set of stability coefficients for each indicator; correcting the corresponding values ​​in the numerical set of the degree of variation of each indicator according to the values ​​of each indicator in the stability coefficient set, and generating a comprehensive set of numerical values ​​of the degree of variation that integrates distinguishability and stability; and normalizing the comprehensive set of numerical values ​​of the degree of variation to convert it into the comprehensive importance coefficients of each indicator, which are used to generate the dynamic weight vector.

[0078] It should be noted that we have obtained a chemical characteristic matrix of pollution sources consisting of textile dyeing and printing wastewater (source 1), agricultural non-point source runoff (source 2), and urban domestic sewage (source 3). This step aims to calculate the weights of three characteristic indicators: the chloride / sulfate ion ratio (index A), the δ¹⁵N nitrogen stable isotope ratio (index B), and the caffeine component ratio (index C).

[0079] The degree of variability (discrimination ability assessment) is calculated based on the relative concentration data obtained in the previous step. The numerical dispersion of each indicator among the three pollution sources is calculated (using standard deviation as the measure): Indicator A (ion ratio): Source 1 is 0.10, Source 2 is 0.40, Source 3 is 0.50. The three values ​​are relatively evenly distributed, and the calculated dispersion measure is 0.17. Indicator B (isotopes): Source 1 is 0.10, Source 2 is 0.20, Source 3 is 0.70. Source 3 is significantly higher, with good discrimination ability, and the dispersion measure is 0.26. Indicator C (organic matter): Source 1 is 0.10, Source 2 is 0.10, Source 3 is 0.80. Source 3 is extremely high, and the others are extremely low, exhibiting the strongest discrimination ability, and the dispersion measure is 0.32. The resulting original set of variability values ​​is {0.17, 0.26, 0.32}. To prevent misleading results from indicators that, while highly discriminative, are unstable in the environment, a stability coefficient correction is introduced. The system queries the "Indicator Stability Knowledge Base" to obtain correction coefficients: Indicator A is an inorganic ion, extremely stable, with a stability coefficient of 0.95. Indicator B is an isotope, exhibiting slight fractionation during migration, with a stability coefficient of 0.80. Indicator C is an organic compound, easily degraded in natural water bodies, with relatively poor stability, and a stability coefficient of 0.50.

[0080] The overall variability is generated and normalized using stability coefficients to correct the original variability (the two are multiplied): Correction value for index A: 0.17 multiplied by 0.95 equals 0.1615. Correction value for index B: 0.26 multiplied by 0.80 equals 0.2080. Correction value for index C: 0.32 multiplied by 0.50 equals 0.1600.

[0081] Note that although indicator C has the strongest original discriminative power, its score decreases after correction due to its instability. Therefore, the set of corrected values ​​{0.1615, 0.2080, 0.1600} is normalized. The sum of the three is 0.5295. The final weight of indicator A = 0.1615 / 0.5295 ≈ 0.305; the final weight of indicator B = 0.2080 / 0.5295 ≈ 0.393; the final weight of indicator C = 0.1600 / 0.5295 ≈ 0.302; a dynamic weight vector [0.305, 0.393, 0.302] is generated, arranged in the order of indicators A, B, and C.

[0082] 203. Input the pollution source chemical feature matrix, the standardized chemical feature vector of the receiver water sample, and the dynamic weight vector into a preset non-negative constraint optimization model to solve for a set of preliminary contribution ratios of each pollution source. This set of ratios minimizes the overall deviation between the weighted chemical features of each pollution source and the chemical features of the receiver.

[0083] Specifically, based on the chemical characteristic matrix of pollution sources and the dynamic weight vector, an objective function is constructed with the contribution ratio of each pollution source as the variable and the goal of minimizing the weighted sum of squared deviations. Constraints are set, including the requirement that the contribution ratio of all pollution sources be non-negative and that the sum of all contribution ratios is one, forming a set of constraint parameters for the non-negative constraint optimization model. A quadratic programming algorithm is used to solve the objective function under the constraint parameter set, calculating an initial contribution ratio vector that satisfies the constraints and represents the initial contribution share of each source. The initial contribution ratio vector is normalized to ensure that the sum of its elements is strictly 1, generating a preliminary set of contribution ratios for each pollution source that satisfies the requirement of proportional summation.

[0084] Furthermore, after generating a preliminary set of contribution ratios for each pollution source that meets the proportional summation requirement, the process further includes: substituting the preliminary set of contribution ratios for each pollution source back into the chemical mass balance equation set, calculating the absolute difference between the predicted and measured values ​​for each chemical characteristic indicator, and generating a mass balance residual vector; based on a preset allowable error range associated with the analytical accuracy of each chemical characteristic indicator, performing compliance judgment on each residual value in the mass balance residual vector, and generating a set of balance state indicators for each indicator; according to the set of balance state indicators, if the balance state of all indicators is marked as compliant, then the preliminary set of contribution ratios for each pollution source is marked as a reliable solution; if any indicator is marked as non-compliant, then based on the weight coefficients in the dynamic weight vector corresponding to the non-compliant indicator, identifying and generating a list of suspected pollution sources, which is used to prompt for review or supplementary sampling analysis of the characteristic data of the pollution sources in the list.

[0085] It should be noted that the known chemical characteristic matrix of pollution sources (source 1: textile wastewater, source 2: agricultural runoff, source 3: domestic sewage) and the dynamic weight vector generated in step 202 are... .

[0086] This step requires combining the recipient water sample data to calculate the contribution ratio of each pollution source and verify the reliability of the results.

[0087] Input data preparation and model building: Water samples were collected from the polluted river section, and their standardized values ​​were measured for indicators A (ion ratio), B (isotope), and C (organic matter) to obtain the receptor vector. To determine the contribution ratios x1, x2, and x3 of the three pollution sources, a non-negativity-constrained optimization model is constructed. The objective function is defined. The weighted sum of squared deviations is expressed mathematically as follows:

[0088] ,

[0089] Among them, w j R represents the dynamic weight of the j-th indicator. j S is the observed value of the i-th index of the receptor water sample. ij Let be the characteristic value of the i-th pollution source on the j-th indicator. The constraints are set as follows: and .

[0090] The quadratic programming solution substitutes the source matrix data (source 1: [0.1, 0.1, 0.1]; source 2: [0.4, 0.2, 0.1]; source 3: [0.5, 0.7, 0.8]) and the receptor vector into the above formula. The quadratic programming algorithm is used iteratively to minimize this value. The calculated initial contribution ratio vectors are: source 1 = 0.10 (10%), source 2 = 0.30 (30%), and source 3 = 0.60 (60%). After normalization, the sum of the three is strictly 1, confirming it as the initial contribution ratio set.

[0091] Quality balance backsubstitution and residual compliance judgment: Substitute the above proportions back into the quality balance equation, calculate the predicted value and compare it with the measured receptor vector to generate the residual vector: Indicator A: Predicted value Measured value: 0.43, residual: 0. Indicator B: Predicted value The measured value was 0.49, and the residual was 0. Indicator C: Predicted value. The measured value was 0.55, and the absolute residual was 0.03.

[0092] The result judgment and list generation are based on the preset tolerance error range (0.05 for indicators A and B, and 0.02 for indicator C due to its higher analysis difficulty). The judgment is as follows: if the residuals of indicators A and B are both less than 0.05, they are marked as "compliant". If the residual of indicator C is 0.03, which is greater than the tolerance error of 0.02, it is marked as "non-compliant".

[0093] Due to the presence of non-compliant indicators, the system classifies the solution as "pending". Based on the high-weight coefficients and source feature matrix corresponding to indicator C, source 3 (domestic sewage) has the highest relative abundance (0.80) for indicator C, making it the main potential factor causing the prediction bias of this indicator. Therefore, a list of suspected pollution sources is generated, with source 3 listed as the primary target for review, indicating the need for secondary sampling analysis of the organic component characteristic data of source 3 to correct the model bias.

[0094] 204. By introducing random perturbations within a preset range into the chemical characteristic matrix of pollution sources and the standardized chemical characteristic vector of the receptor water sample, and performing multiple iterative solutions, the preliminary contribution ratio of each pollution source is statistically analyzed, and quantitative source tracing results including the estimated contribution rate of each pollution source and its confidence interval are output.

[0095] Specifically, based on a preset analytical error range, independent random perturbation values ​​are superimposed on the elements in the chemical feature matrix of the pollution source and the elements in the standardized chemical feature vector of the receptor water sample to generate multiple sets of simulated perturbation source matrices and receptor vectors. For each set of simulated perturbation source matrices and receptor vectors, the dynamic weight generation and optimization solution process is re-executed to obtain a corresponding pollution source contribution ratio solution. By repeating this process multiple times, a pollution source contribution ratio solution sequence including multiple solution samples is formed. Statistical analysis is performed on the proportion value corresponding to each pollution source in the pollution source contribution ratio solution sequence, and its statistical mean is calculated as the final point estimate of the pollution source contribution rate, and its numerical interval at a specified confidence level is calculated as the confidence interval. The final point estimates and confidence intervals of all pollution sources are collected to form a quantitative source tracing result report including uncertainty assessment.

[0096] It should be noted that although preliminary calculations show that textile dyeing wastewater (source 1), agricultural non-point source runoff (source 2), and urban domestic sewage (source 3) contribute 10%, 30%, and 60% respectively, Monte Carlo simulation is needed to assess the reliability of the results, considering the unavoidable errors in the sampling and chemical analysis process.

[0097] The relative standard deviation of the laboratory analysis was set at 5%. Using this as a benchmark, the system generated random noise following a normal distribution and superimposed it onto each data point of the original "pollution source chemical characteristic matrix" and "receptor water sample vector". In the first simulation, the relative concentration of source 3 in the key indicator C was randomly perturbed from 0.80 to 0.78, and the value of the receptor water sample in indicator C was perturbed from 0.55 to 0.56. This new data was re-input into the optimization model, yielding the first set of simulated solutions: source 1 contributed 11.2%, source 2 contributed 28.5%, and source 3 contributed 60.3%. In the second simulation, the value of source 2 in indicator A shifted positively, and the value of the receptor sample shifted negatively. After resolving, the second set of simulated solutions was obtained: source 1 contributed 9.5%, source 2 contributed 31.2%, and source 3 contributed 59.3%.

[0098] The above process was repeated, performing a total of 2000 independent simulations. The system recorded the contribution ratio of the three pollution sources obtained in each calculation, forming a solution sequence containing 2000 sets of data. These 2000 results demonstrate the possible range of fluctuations in the results, taking into account analytical errors.

[0099] Statistical analysis was performed on the solution sequence. The arithmetic mean was calculated as the corrected final point estimate. Next, the standard deviation was calculated to measure the magnitude of fluctuation, and a 95% confidence interval was determined based on the 2.5% and 97.5% quantiles (i.e., there is a 95% confidence that the true value falls within this range).

[0100] The generated quantitative traceability results report is shown in Table 7 below:

[0101]

[0102] 205. Based on the quantitative source tracing results report, identify the main pollution sources whose contribution rate to the target receptor surface water exceeds the preset threshold, and generate a list of key monitored pollution sources; based on the list of key monitored pollution sources, adjust or optimize the sampling point layout scheme and monitoring frequency of the surface water environmental monitoring network in the target area to form a targeted enhanced monitoring scheme; according to the enhanced monitoring scheme, implement a new round of key pollution source and receptor water sample collection and analysis to obtain an updated chemical analysis dataset; using the updated chemical analysis dataset, update the pollution source chemical feature matrix and receptor chemical feature vector, and re-execute the subsequent weight calculation, model solution and uncertainty analysis steps to output the updated quantitative source tracing results for evaluating the effectiveness of pollution control measures.

[0103] Furthermore, according to the enhanced monitoring plan, after implementing a new round of key pollution source and receptor water sample collection and analysis to obtain an updated chemical analysis dataset, the process also includes: calculating the change in contribution rate of each major pollution source based on the updated chemical analysis dataset and the list of key monitored pollution sources, and generating a pollution contribution change trend analysis table; matching recommended treatment measures from a pre-set pollution treatment technology library to pollution sources with significantly increased contribution rates based on the pollution contribution change trend analysis table, forming a targeted pollution source treatment priority and measure recommendation table; guiding environmental management departments to implement engineering or management intervention measures for specific pollution sources based on the pollution source treatment priority and measure recommendation table, and recording the implementation of pollution control measures in a pollution control measure implementation log; collecting and analyzing the chemical characteristic data of the water samples of the intervened pollution sources and receptors again at a pre-set period after the implementation of intervention measures, generating a measure post-effect evaluation dataset; and quantifying the actual effect of treatment measures on reducing the contribution of specific pollution sources by comparing the measure post-effect evaluation dataset with the updated chemical analysis dataset, and outputting a treatment effect evaluation report.

[0104] It should be noted that the quantitative source tracing report in step 204 shows that the contribution rate of urban domestic sewage (source 3) is as high as 59.9%, and the lower limit of the confidence interval exceeds 50%.

[0105] The system identified Source 3 as a "major pollution source" (exceeding the preset 40% threshold) and automatically generated a list of key monitored pollution sources. Based on this list, the environmental protection department dynamically adjusted the monitoring network: increasing the sampling frequency of Source 3's emission outlet from "once a month" to "once a week," and adding two temporary, encrypted sampling points 500 meters downstream of its discharge outlet.

[0106] After a month of intensive monitoring, updated chemical analysis datasets were obtained. System calculations revealed that the contribution rate of Source 3 fluctuated consistently between 58% and 62% over the past month, showing a persistently high trend. Based on the "Pollution Contribution Change Trend Analysis Table," and considering Source 3's high eigenvalue in index C (organic matter), the system automatically retrieved and matched "MBR membrane bioreactor upgrading and retrofitting" as a recommended measure from the "Pollution Control Technology Library," generating a "Level 1 Priority" treatment recommendation table.

[0107] Based on recommendations, the environmental management department urged the urban wastewater treatment plant to implement upgrading and renovation projects, and recorded the implementation in the log. Three months after the project was completed and put into operation, water samples from source 3 and the recipient were collected again for analysis (post-treatment effect evaluation dataset). The new data was input into the model for recalculation, revealing a significant reduction in the concentration of characteristic organic matter in source 3. By comparing the contribution rate data before and after the treatment, the system quantified the emission reduction effect and output the following 8-point treatment effect evaluation report:

[0108]

[0109] The control measures successfully reduced Source 3 to a non-dominant pollution source. Due to the significant decrease in Source 3's contribution, the relative contribution of Source 2 (agricultural non-point source pollution) passively increased. Based on this, the system recommends that the next round of control efforts shift the focus to agricultural non-point source pollution control.

[0110] The present invention also provides a surface water source tracing and quantitative device that integrates water quality concentration and chemical composition characteristics. The surface water source tracing and quantitative device that integrates water quality concentration and chemical composition characteristics includes a memory and a processor. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor performs the steps of the surface water source tracing and quantitative method that integrates water quality concentration and chemical composition characteristics in the above embodiments.

[0111] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the surface water source tracing and quantitative method that integrates water quality concentration and chemical composition characteristics.

[0112] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0113] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0114] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics, characterized in that, include: Based on the chemical analysis of surface water samples from various potential pollution sources and target receptors, the concentrations of various chemical characteristic indicators were converted into relative concentrations to construct a chemical characteristic matrix of pollution sources. Based on the chemical characteristic matrix of pollution sources, the degree of variation of each chemical characteristic index among different pollution sources is calculated, and a dynamic weight vector is generated according to the degree of variation. The pollution source chemical feature matrix, the standardized chemical feature vector of the receiver water sample, and the dynamic weight vector are input into a preset non-negative constraint optimization model to obtain the preliminary contribution ratio of each pollution source. This ratio minimizes the overall deviation between the weighted chemical features of each pollution source and the chemical features of the receiver, including: Based on the pollution source chemical feature matrix and the dynamic weight vector, the objective function is constructed. Set constraints, including requiring that the contribution ratio of all pollution sources be non-negative and that the sum of all contribution ratios be one, to form a set of constraint parameters; Solving the objective function under the constraint parameter set yields the initial contribution ratio vector; The initial contribution ratio vector is processed to generate a set of contribution ratios for each pollution source; After generating a preliminary set of contribution proportions from each pollution source that meets the proportional summation requirement, the following steps are also included: Substitute the initial set of contribution ratios of each pollution source back into the chemical mass balance equation set, calculate the absolute difference between the predicted and measured values ​​of each chemical characteristic index, and generate the mass balance residual vector. Based on the preset tolerance range, compliance judgment is performed on each residual value in the quality balance residual vector to generate a set of balance state flags for each indicator. According to the set of equilibrium state indicators, if the equilibrium state of all indicators is marked as compliant, the preliminary set of contribution ratios of each pollution source is marked as a reliable solution; if there are indicators marked as non-compliant, a list of suspected pollution sources is identified and generated based on the weight coefficients in the dynamic weight vector corresponding to the non-compliant indicators, which is used to prompt the review or supplementary sampling analysis of the characteristic data of the pollution sources in the list. By introducing random perturbations within a preset range into the chemical characteristic matrix of the pollution sources and the standardized chemical characteristic vector of the receptor water sample, and performing multiple iterative solutions, the preliminary contribution ratio of each pollution source is statistically analyzed, and quantitative source tracing results are output.

2. The quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics according to claim 1, characterized in that, include: A set of chemical characteristic indicators, including ion concentration ratios, stable isotope ratios, and proportions of characteristic organic components, was selected to form a set of characteristic indicators for source tracing analysis. Based on the set of characteristic indicators, water samples from each potential pollution source and target receptor were collected and chemically analyzed to obtain the original concentration data table of each water sample under each indicator. The original concentration data table is processed to generate a chemical analysis dataset; For the chemical analysis dataset, the concentration of each index of each pollutant source water sample is divided by the sum of the concentrations of all pollutants at that index, and the absolute concentration is converted into a relative concentration that characterizes its relative abundance at that index. The relative concentration data of all pollutants are collected to construct a chemical characteristic matrix of the pollutants.

3. The quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics according to claim 2, characterized in that, include: Based on the pollution source chemical characteristic matrix, the degree of dispersion of each chemical characteristic index among the pollution sources is calculated to obtain a set of values ​​including the degree of variation of each index. Divide each value in the set of values ​​for the degree of variation of each indicator by the sum of all values ​​in the set to convert the degree of variation of each indicator into a relative importance coefficient. The relative importance coefficients corresponding to all chemical characteristic indicators are collected, arranged in order of indicators, and a dynamic weight vector is generated.

4. The quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics according to claim 3, characterized in that, Before the process of aggregating the relative importance coefficients corresponding to all chemical characteristic indicators, arranging them in order of indicator, and generating a dynamic weight vector, the following steps are also included: Based on a pre-defined knowledge base of indicator stability, the stability level information of each chemical characteristic indicator during the migration process in the natural environment is queried and obtained, and a set of stability coefficients for each indicator is generated. Based on the values ​​of each indicator in the stability coefficient set, the corresponding values ​​in the numerical set of the degree of variation of each indicator are corrected to generate a comprehensive numerical set of degree of variation. The set of comprehensive variation values ​​is processed and transformed into comprehensive importance coefficients for each indicator, which are then used to generate a dynamic weight vector.

5. The quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics according to claim 1, characterized in that, To determine the contribution ratios x1, x2, and x3 of the three pollution sources, a non-negativity-constrained optimization model is constructed, and an objective function is defined. For the weighted sum of squared deviations: Among them, w j R represents the dynamic weight of the j-th indicator. j S is the observed value of the i-th index of the receptor water sample. ij Let be the characteristic value of the i-th pollution source on the j-th indicator. The contribution ratio of pollution sources is subject to the following constraints: and .

6. The quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics according to claim 1, characterized in that, include: Based on the preset analytical error range, independent random perturbation values ​​are superimposed on the elements in the pollution source chemical feature matrix and the elements in the standardized chemical feature vector of the receptor water sample to generate multiple sets of simulated perturbed source matrices and receptor vectors. For each set of simulated perturbation source matrix and receptor vector, the dynamic weight generation and optimization solution process is re-executed to obtain the corresponding pollution source contribution ratio solution. By repeating this process multiple times, a pollution source contribution ratio solution sequence including multiple solution samples is formed. Statistical analysis is performed on the proportion value corresponding to each pollution source in the solution sequence of pollution source contribution ratios, and its statistical mean is calculated as the final point estimate of the contribution rate of the pollution source. The numerical interval at the specified confidence level is calculated as the confidence interval. The final point estimates and confidence intervals of all pollution sources are compiled to form a quantitative source tracing results report.

7. The quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics according to claim 1, characterized in that, Also includes: Based on the quantitative source tracing results report, identify the main pollution sources whose contribution rate to the surface water of the target receptor exceeds a preset threshold, and generate a list of key monitored pollution sources. Based on the list of key pollution sources, the sampling point layout and monitoring frequency of the surface water environment monitoring network in the target area are adjusted or optimized to form an enhanced monitoring plan; According to the enhanced monitoring plan, a new round of water sample collection and analysis of key pollution sources and receptors will be carried out to obtain updated chemical analysis datasets. Using the updated chemical analysis dataset, the source chemical feature matrix and receptor chemical feature vector are updated, and subsequent weight calculation, model solving and uncertainty analysis steps are re-executed to output the updated quantitative source tracing results.

8. The quantitative method for tracing the source of surface water by integrating water quality concentration and chemical composition characteristics according to claim 7, characterized in that, After implementing a new round of water sample collection and analysis of key pollution sources and receptors according to the enhanced monitoring plan to obtain updated chemical analysis datasets, the process also includes: Based on the updated chemical analysis dataset and the list of key monitored pollution sources, the change in the contribution rate of each major pollution source is calculated, and a pollution contribution change trend analysis table is generated. Based on the pollution contribution change trend analysis table, recommended treatment measures from the preset pollution treatment technology library are matched to pollution sources with significantly increased contribution rates, forming a pollution source treatment priority and measure recommendation table; Based on the aforementioned priority and recommended measures table for pollution source control, guide environmental management departments to implement engineering or management intervention measures for pollution sources and record them in a pollution control measures implementation log; At a predetermined period after the implementation of the intervention measures, chemical characteristic data of water samples from the intervened pollution sources and recipients are collected and analyzed again to generate a post-implementation effect assessment dataset. By comparing the post-implementation effect assessment dataset with the updated chemical analysis dataset, the actual effect of the treatment measures on reducing the contribution of specific pollution sources is quantified, and a treatment effect assessment report is output.

Images