A method and system for tracing the source of contamination of soil organic matter

Abstract

This invention provides a method and system for tracing the source of soil organic matter pollution, relating to the field of data processing technology. The method includes: collecting soil and end-member samples from the site to be tested and simultaneously performing sampling and measurement on the soil and end-member samples to construct a three-dimensional feature vector of the samples and an end-member feature matrix; calculating the end-member discrimination index of the end-member feature matrix using the Euclidean distance algorithm, checking and correcting the end-member feature vectors in the end-member feature matrix to obtain the target end-member feature matrix; calculating a dynamic weight compensation factor based on the target end-member carbon-nitrogen mass ratio in the target end-member feature matrix; establishing an isotope mass balance equation based on the sample three-dimensional feature vector and the target end-member feature matrix, and embedding the dynamic weight compensation factor into the isotope mass balance equation to construct a weighted correction equation set; substituting the weighted correction equation set into a Bayesian linear mixture model to calculate the average contribution rate of each pollution source end-member to the accumulation of soil organic matter, thus completing the pollution source tracing.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and system for tracing the source of pollution of soil organic matter. Background Technology

[0002] Soil organic pollutants are widely present in industrial parks, steel smelting areas, areas surrounding major transportation routes, and historical industrial sites. Their pollution sources are typically characterized by multiple overlapping sources, complex migration and transformation processes, and strong spatial heterogeneity. To support pollution risk assessment, remediation responsibility determination, and the selection of optimal remediation measures, it is urgent to conduct research on the identification of soil organic pollutant sources and quantitative analysis of their contribution rates, in order to achieve a scientific determination of the types of pollution sources and their proportions.

[0003] Currently, soil organic pollution source tracing technologies mainly include physical classification, biomarker fingerprinting, spectroscopic characterization, and stable isotope tracing. Among these, stable isotope tracing measures the stable isotope ratios of elements such as carbon and nitrogen in soil samples, and combines this with end-member mixing models or mass conservation models to quantitatively calculate the contribution ratios of different potential pollution sources (such as coal combustion, oil, traffic emissions, and natural background sources). This method has advantages such as small sample requirements and the ability to achieve quantitative analysis, and has become an important technical approach for tracing pollution sources in multi-source complex pollution sites.

[0004] However, existing stable isotope endmember analysis methods still have significant limitations. Traditional molecular diagnostic ratio methods are prone to diagnostic ratio drift during organic matter migration, weathering, and microbial degradation, leading to unstable source apportionment results. Furthermore, whole-sample stable isotope methods typically assume that each pollution source endmember has a relatively balanced elemental load when performing endmember mixing calculations, lacking a concentration-dependent correction mechanism for differences in endmember carbon and nitrogen content. This can easily result in isotope signals being masked or over-captured, leading to a systematic underestimation or overestimation of the contribution rate of some pollution sources, thus reducing the accuracy of pollution source identification. Summary of the Invention

[0005] To address the challenges of traditional molecular diagnostic ratio methods, which are prone to diagnostic ratio drift during organic matter migration, weathering, and microbial degradation, leading to unstable source apportionment results, and the issues arising from the whole-sample stable isotope method's assumption of relatively balanced elemental loads for each pollution source endmember during endmember mixing calculations, lacking a concentration-dependent correction mechanism for differences in endmember carbon and nitrogen content, this invention provides a pollution source tracing method and system suitable for soil organic matter.

[0006] The technical solutions provided by the embodiments of the present invention are as follows: First aspect: This invention provides a method for tracing the source of soil organic matter pollution, comprising: S1: Collect soil samples from the site to be tested; S2: Simultaneously measure the soil samples from the site to be tested, and construct a three-dimensional feature vector of the samples; S3: Collect end-member samples of potential pollution sources in the test site and perform simultaneous measurements to construct an end-member feature matrix; S4: Calculate the endmember discrimination index of the endmember feature matrix using the Euclidean distance algorithm, and check and correct the endmember feature vectors in the endmember feature matrix according to the endmember discrimination index to obtain the target endmember feature matrix. S5: Based on the carbon-nitrogen mass ratio of the target endmember in the target endmember feature matrix, calculate the dynamic weight compensation factor including the carbon mass weight operator and the nitrogen mass weight operator; S6: Based on the three-dimensional feature vector of the sample and the feature matrix of the target endmember, establish an isotope mass balance equation, and embed the dynamic weight compensation factor into the isotope mass balance equation to construct a weighted correction equation set. S7: Substitute the weighted correction equations into the Bayesian linear mixture model to calculate the average contribution rate of each pollution source end-member to the accumulation of soil organic matter. S8: Based on the average contribution rate of each of the above, trace the source of soil organic matter pollution at the site to be tested.

[0007] The second aspect: This invention provides a pollution source tracing system for soil organic matter, comprising: processor; A memory storing computer-readable instructions, which, when executed by the processor, implement the pollution source tracing method for soil organic matter as described in the first aspect.

[0008] Third aspect: The present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the pollution source tracing method for soil organic matter as described in the first aspect.

[0009] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: In this embodiment of the invention, soil samples from the site to be tested are collected and simultaneously sampled and measured with end-member samples to construct a three-dimensional feature vector and an end-member feature matrix. Then, the end-member discrimination index is calculated using Euclidean distance, and the end-member feature vector is checked and corrected to obtain the target end-member feature matrix. Furthermore, based on the carbon-nitrogen mass ratio of the target end-member, a concentration-dependent dynamic weight compensation factor is introduced and an isotope mass balance equation is embedded to construct a weighted correction equation set. The equation set is then substituted into a Bayesian linear mixture model to calculate the mean contribution rate of each pollution source end-member and complete the pollution source tracing. This enhances the adaptability and result stability of the source apportionment process to factors such as organic matter migration, weathering, and microbial degradation, improves the concentration-dependent correction ability of end-member mixture calculation to differences in carbon and nitrogen content, reduces the systematic bias of contribution rate caused by isotope signal masking or over-capture, and thus improves the accuracy and reliability of pollution source contribution rate estimation and pollution source identification. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic flowchart of a pollution source tracing method for soil organic matter provided in an embodiment of the present invention.

[0012] Figure 2 This is a schematic diagram of a pollution source tracing system for soil organic matter provided in an embodiment of the present invention. Detailed Implementation

[0013] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0014] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0015] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0016] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0017] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0018] Reference manual attached Figure 1 The diagram illustrates a flow chart of a pollution source tracing method for soil organic matter provided by an embodiment of the present invention.

[0019] This invention provides a method for tracing the source of pollution of soil organic matter. This method can be implemented using a pollution source tracing device suitable for soil organic matter, which can be a terminal or a server. The processing flow of this method for tracing the source of pollution of soil organic matter may include the following steps:

[0020] S1: Collect soil samples from the site to be tested.

[0021] S2: Simultaneous sampling and measurement of soil samples from the test site to construct a three-dimensional feature vector of the sample.

[0022] Among them, simultaneous injection and determination refers to loading the pretreated soil powder sample into the injection container of the elemental analyzer or other detection equipment according to the preset mass, and obtaining multiple target parameters simultaneously in the same set of analysis processes through a single injection and the same determination procedure.

[0023] In one possible implementation, S2 specifically includes: S201: Remove impurities, dry, grind and sieve the soil sample from the test site to obtain a soil powder sample with uniform particle size.

[0024] S202: Weigh the soil powder sample according to the preset mass range, and put the weighed soil powder sample into the sample injection container of the elemental analyzer to form the sample injection unit to be tested.

[0025] The sample unit to be tested is a standardized sample unit that can be directly tested after the soil powder sample obtained through pretreatment is weighed according to a preset mass range, loaded into the sample container specified by the elemental analyzer (or isotope mass spectrometry system), and then packaged and numbered.

[0026] It should be noted that those skilled in the art can set the size of the preset quality range according to actual needs, and this invention does not limit this.

[0027] S203: Perform a single injection into the sample unit to be tested, and simultaneously obtain the carbon stable isotope value and nitrogen stable isotope value of the whole soil sample in the same measurement process, and simultaneously determine the total organic carbon content and total nitrogen content of the sample unit to be tested.

[0028] S204: Calculate the carbon-nitrogen mass ratio of the soil sample from the test site based on the total organic carbon content and total nitrogen content.

[0029] Specifically, the carbon-nitrogen mass ratio is the ratio of total organic carbon content to total nitrogen content.

[0030] S205: By combining the stable carbon isotope values, stable nitrogen isotope values, and the carbon-nitrogen mass ratio of the soil sample from the testing site, a three-dimensional feature vector of the sample is constructed. in, This represents the three-dimensional feature vector of the soil sample from the site to be tested. This indicates the carbon stable isotope value (δ¹⁸) of the soil sample from the testing site. 13 C), This indicates the nitrogen stable isotope value (δ¹⁸) of the soil sample from the testing site. 15 N), This indicates the carbon-to-nitrogen mass ratio of the soil sample from the testing site. The symbols represent vectors.

[0031] In this embodiment of the invention, by simultaneously measuring soil samples from the test site during injection, the stable carbon isotope values, stable nitrogen isotope values, total organic carbon content, and total nitrogen content of the whole soil sample are obtained simultaneously in a single injection and the same measurement process. Furthermore, the carbon-nitrogen mass ratio is calculated to construct a three-dimensional feature vector for the sample. This method enables the coordinated acquisition of multiple parameters under unified experimental conditions and consistent error sources, significantly improving the comparability and stability of sample fingerprint data and avoiding systematic bias and cumulative errors caused by multiple injections.

[0032] S3: Collect end-member samples of potential pollution sources in the test site and perform simultaneous measurements to construct an end-member feature matrix.

[0033] In one possible implementation, S3 specifically includes: S301: Based on the production process, historical pollution records, functional zoning, and on-site survey results of the site to be tested, identify the types of potential pollution sources that contribute to soil organic matter pollution within the site to be tested, and construct an end-member list.

[0034] It should be noted that the end-source list includes coal-related sources, oil-related sources, transportation emission sources, and natural soil background sources.

[0035] S302: Based on the end-member list, set up end-member sampling points at typical locations corresponding to each potential pollution source and collect multiple end-member samples.

[0036] Optionally, the determination of typical locations includes: First, delineating high-risk areas of potential pollution sources by combining the production process flow, material storage and transportation routes, distribution of emission outlets and leakage points, and historical pollution records of the site to be tested. Second, inferring the main diffusion channels and enrichment areas of pollutants based on environmental conditions such as site functional zoning, topography, surface runoff direction, prevailing wind direction, and downwind deposition zone. Finally, verifying the surrounding areas of key facilities (such as boiler / chimney outlets, oil tank areas and pipeline interfaces, loading and unloading areas, road intersections, and parking areas) and their downstream / downwind deposition zones through on-site reconnaissance, and prioritizing representative locations with minimal external interference and repeatable sampling conditions as typical end-member sampling locations corresponding to each potential pollution source.

[0037] Specifically, coal-related source end-unit samples include ash, coal dust, or combustion residue; oil-related source end-unit samples include sludge, lubricating oil residue, or oil deposits; natural soil background source end-unit samples include deep, uncontaminated soil; and traffic emission source end-unit samples include road dust, tire wear particles, or traffic emission deposits.

[0038] The number of sampling points and the number of samples collected in each area are shown in the table below: S303: Using the same measurement method as the soil samples at the test site, determine the end-member carbon stable isotope value, end-member nitrogen stable isotope value, and end-member carbon-nitrogen mass ratio of each end-member sample.

[0039] S304: Combine the endmember carbon stable isotope value, endmember nitrogen stable isotope value, and endmember carbon-nitrogen mass ratio of each endmember term to construct multiple endmember feature matrix vectors: in, Indicates the first i The endmember feature vector of each endmember term. The symbols represent vectors. Indicates the first i Carbon stable isotope values ​​of each end-member term Indicates the firsti The nitrogen stable isotope values ​​of each endmember term, Indicates the first i The carbon-nitrogen mass ratio of each end-member term.

[0040] S305: Summarize the endmember feature matrix vectors of each endmember sample according to the endmember sample number, and construct the endmember feature matrix: in, Represents the endmember feature matrix, with superscripts T Represents the transpose of a matrix. Indicates the first i The endmember feature vector corresponding to each endmember sample. i =1,2,… m , m This indicates the total number of end-member samples.

[0041] In this embodiment of the invention, potential pollution source types are identified and an end-member list is constructed based on the production process, historical pollution records, functional zoning, and on-site reconnaissance of the site to be tested. End-member sampling points are then set up at typical locations corresponding to each pollution source to collect end-member samples. The carbon stable isotope value, nitrogen stable isotope value, and carbon-nitrogen mass ratio of each end-member are obtained by using the same sampling and synchronous measurement method as the soil sample. This constructs an end-member feature matrix, which can ensure that the end-member fingerprint data source is reliable, highly representative, and has the same measurement conditions and error structure as the sample to be tested. This effectively reduces the source tracing deviation caused by improper end-member selection, external interference, and measurement differences.

[0042] It should be noted that, in order to address the problems in existing technologies where end-member sample feature vectors overlap in the isotope-element ratio feature space due to differences in sampling location, fluctuations in operating conditions, and external interference, leading to multiple solutions, false solutions, and unstable contribution rate estimation during the solution of end-member hybrid models, this invention innovatively introduces an end-member identifiability evaluation and adaptive correction mechanism based on Euclidean distance before pollution source tracing analysis. By standardizing the distance calculation of the end-member feature matrix and constructing an end-member discrimination index, the invention achieves quantitative judgment of the end-member discrimination ability, significantly improving the accuracy and engineering applicability of pollution source tracing in complex multi-source sites.

[0043] S4: Calculate the endmember discriminant index of the endmember feature matrix using the Euclidean distance algorithm, and check and correct the endmember eigenvectors in the endmember feature matrix based on the endmember discriminant index to obtain the target endmember feature matrix.

[0044] The Euclidean distance algorithm is a distance calculation method used to measure the degree of difference between two samples (or vectors) in a multidimensional feature space. Its basic idea is to square and sum the differences of each feature dimension and then take the square root to obtain the straight-line distance between the two points in the feature space.

[0045] The endmember discrimination index is an evaluation metric used to quantitatively characterize the degree of distinguishability among endmembers in the endmember feature matrix. The larger the index, the more significant the differences between endmembers, the more stable the solution of the mixture model, and the less likely it is to produce multiple solutions or false solutions. Conversely, the smaller the index, the higher the degree of overlap of endmember features, which can easily lead to unstable contribution rate estimation. It is necessary to merge or resample overlapping endmembers to improve the reliability of source tracing analysis.

[0046] In one possible implementation, S4 specifically includes: S401: Standardize the endmember eigenvectors in the endmember feature matrix to obtain the standard endmember feature matrix.

[0047] It should be noted that Z-score standardization was chosen for the standardization process. This is used to ensure comparability in distance calculations. Z-score standardization is existing technology and will not be elaborated upon here.

[0048] S402: Calculate the Euclidean distance between any two endmember eigenvectors in the standard endmember eigenma matrix to form a distance matrix.

[0049] S403: Obtain the minimum distance between each endmember feature vector and the remaining endmember feature vectors in the distance matrix, and calculate the arithmetic mean of all minimum distances as the endmember discrimination index.

[0050] It should be noted that there must be sufficient differences between endmembers; otherwise, the model may have multiple solutions, false solutions, or unstable contribution rates.

[0051] S404: Determine whether the endmember discrimination index of the endmember feature matrix is ​​greater than or equal to the preset endmember discrimination index. If yes, determine that the discrimination between the endmember feature vectors in the endmember feature matrix meets the source tracing parsing requirements, and determine the endmember feature matrix as the target endmember feature matrix, then proceed to S5. Otherwise, determine that the discrimination between the endmember feature vectors in the endmember feature matrix does not meet the source tracing parsing requirements, and proceed to the next step.

[0052] It should be noted that those skilled in the art can set the size of the preset end-member discrimination index according to actual needs, and this invention does not limit this.

[0053] S405: Identify candidate overlapping endmembers in the distance matrix and generate multiple candidate endmember groups.

[0054] Specifically, based on the minimum distances corresponding to the feature vectors of each endmember, the endmembers with minimum distances less than a preset overlap threshold are identified as candidate overlapping endmembers, and the candidate overlapping endmembers are divided into multiple candidate endmember groups according to the distance clustering rules.

[0055] It should be noted that those skilled in the art can set the size of the preset overlap threshold according to actual needs, and this invention does not limit this.

[0056] S406: Merge the endmember feature vectors within each candidate endmember group according to the average value of the endmember feature vectors within the group, update the endmember feature matrix, and return to S401 to re-execute.

[0057] It should be noted that the feature vectors of each endmember in the candidate endmember group are relatively close in the isotope and element ratio feature space, and belong to the same potential pollution source type under different sampling locations or different operating conditions. The average value of the endmember feature vectors in the group is used to construct a hybrid endmember, which can be used as the representative feature value of the endmember group, thereby eliminating redundant endmembers and improving the identifiability of the endmember feature matrix and the stability of source tracing analysis.

[0058] Optionally, if the discriminative power between the endmember feature vectors in the endmember feature matrix fails to meet the source tracing parsing requirements multiple times, manual adjustment should be requested.

[0059] In this embodiment of the invention, the degree of difference between endmembers is objectively assessed by calculating the endmember discrimination index. When the discrimination index is insufficient, candidate overlapping endmembers are automatically identified and merged. This effectively eliminates redundant endmembers of the same pollution source type generated at different sampling locations or under different operating conditions, avoiding problems such as multiple solutions, false solutions, or fluctuations in contribution rate estimation in the hybrid model caused by highly similar endmember features. Simultaneously, Z-score standardization ensures that δ¹³C and δ¹³C are within acceptable limits. 15 By having N and C / N participate in distance calculation at a unified scale, the endmember screening and correction process becomes more stable and repeatable, ultimately resulting in a target endmember feature matrix with higher discriminative power. This provides reliable prior input for subsequent dynamic weight compensation and Bayesian mixture model inference, significantly improving the accuracy of quantitative analysis of pollution source contribution rates.

[0060] S5: Based on the carbon-nitrogen mass ratio of the target endmembers in the target endmember feature matrix, calculate the dynamic weight compensation factor, which includes the carbon mass weight operator and the nitrogen mass weight operator.

[0061] In one possible implementation, S5 specifically includes: S501: Extract the endmember carbon-nitrogen mass ratio corresponding to each endmember item from the target endmember feature matrix to form a set of endmember carbon-nitrogen mass ratios.

[0062] It should be noted that the end-member carbon-nitrogen mass ratio is the end-member term. iThe carbon and nitrogen mass loading characteristics are used to characterize the degree of asymmetry in the carbon and nitrogen contributions of end-members in the soil organic pool.

[0063] S502: Perform validity verification on the endmember carbon-nitrogen mass ratios in the endmember carbon-nitrogen mass ratio set.

[0064] Optionally, the validity verification specifically involves: when there is a missing or abnormal endmember carbon-nitrogen mass ratio, the corresponding endmember item is removed or re-measured to ensure the stability of the weight compensation factor calculation.

[0065] Optionally, anomalies include, but are not limited to, endmember total nitrogen content approaching zero, resulting in an illegal or anomalous amplification of the endmember carbon-nitrogen mass ratio.

[0066] It should be noted that, in response to the severe asymmetry of carbon and nitrogen loads in the soil organic pool caused by different pollution sources (such as petroleum sources contributing a large amount of carbon but almost no nitrogen), this invention introduces a concentration-dependent correction factor. By using the end-member carbon-nitrogen mass ratio (C / N) as the basis for weight adjustment, carbon mass weight operators and nitrogen mass weight operators are constructed respectively, and then embedded into the isotope mass balance equation to form a weighted correction equation set. This achieves the reinforcement of isotope signals from "high carbon and low nitrogen" sources and the suppression of background interference from "high nitrogen and low carbon" sources.

[0067] S503: Based on the carbon-nitrogen mass ratio of each endmember after validity verification, calculate the carbon mass weight operator and nitrogen mass weight operator for each endmember term: in, Indicates the first i Carbon mass weighting operator corresponding to each end-member term Indicates the first i The carbon-nitrogen mass ratio of each end-member term, i =1,2,… m , m This indicates the total number of endmember samples. Indicates the first i The nitrogen mass weighting operator corresponding to each endmember term.

[0068] In this embodiment of the invention, by calculating the carbon mass weighting operator and the nitrogen mass weighting operator through the endmember carbon-nitrogen mass ratio, the asymmetric differences in carbon and nitrogen loads of different endmembers in the soil organic pool can be explicitly introduced into the model in the form of weights. This allows the "high C / N" endmember to obtain a higher effective weight in carbon isotope balance and be reasonably weakened in nitrogen isotope balance, while the "low C / N" endmember maintains an effective characterization in nitrogen isotope balance and avoids excessive interference to carbon channels. This achieves adaptive compensation and suppression of endmember contribution signals, reduces the problem of underestimation or overestimation of contribution rate caused by traditional equal-weight analysis, and improves the stability and accuracy of multi-source mixed source tracing calculation.

[0069] S504: Summarize the carbon mass weight operator and nitrogen mass weight operator corresponding to each end-member term, and calculate the dynamic weight compensation factor.

[0070] Specifically, the dynamic weight compensation factor is a set that includes the carbon mass weight operator and the nitrogen mass weight operator of the endmember terms.

[0071] It should be noted that the comparative test of actual samples from 37 typical regions in this invention verified the effectiveness of the element ratio weight correction algorithm in multi-source analysis. The test results compared the average proportion of each endmember in the traditional equal-weight analysis method and the weight correction method of this invention. The specific data are shown in the table below:

[0072] Analysis of the calculation results shows that natural soil sources are significantly overestimated (27.02%), usually due to the proximity of their isotopic characteristic values ​​to the mixed sample points. This algorithm, through weighting factors, identifies the mismatch between the high organic matter content in the sample points and the extremely low nutrient abundance of the natural soil, thus correcting its proportion by 21.28%, effectively eliminating the interference of environmental background noise on pollution source tracing. Traffic emission sources are almost ignored in traditional methods (only 3.16%), which is seriously inconsistent with the actual urban environmental conditions. This invention utilizes factors to capture the high C / N ratio mass balance characteristics of traffic and petroleum sources, increasing the contribution rate of traffic sources to 28.87% and petroleum sources to 10.01%. This correction shifts the analysis results from "background simulation" to "emission-driven," more realistically reflecting the contribution intensity of anthropogenic sources to environmental PAHs. Coal combustion, as the main source of regional pollution, has its proportion corrected from 63.51% to 49.07%. This 14.44% reduction does not negate its dominant position, but rather corrects the "signal suppression" phenomenon in the traditional method, where strong coal isotope signals mask other subtle sources, through weight redistribution. This makes the proportion of each pollution source more consistent with the actual quality proportion of the local energy consumption structure.

[0073] In this embodiment of the invention, the dynamic weight compensation factor can automatically adjust the analysis weights based on the chemical abundance characteristics of the source end-members when dealing with complex organic pollution systems. By compensating for key anthropogenic sources such as petroleum and transportation, and by selectively filtering natural background sources, the scientific rigor and evidentiary value of environmental source tracing are significantly improved.

[0074] S6: Based on the sample's three-dimensional feature vector and the target endmember feature matrix, establish the isotope mass balance equation, and embed the dynamic weight compensation factor into the isotope mass balance equation to construct a weighted correction equation set.

[0075] Among them, the isotope mass balance equation is a basic mathematical model used to describe the "source-sink" relationship of a mixed system in the source tracing of stable isotopes. Its core idea is based on the principle of mass conservation, which holds that the isotope characteristic values ​​of the sample to be tested can be obtained by weighted mixing of multiple pollution source end-members according to their contribution ratio.

[0076] In one possible implementation, the isotope mass balance equation includes a carbon isotope mass balance equation and a nitrogen isotope mass balance equation. S6 specifically includes:

[0077] S601: Based on the three-dimensional feature vector of the sample and the feature matrix of the target endmember, establish the isotope mass balance equation.

[0078] S602: The carbon mass weighting operator and the nitrogen mass weighting operator are respectively embedded into the carbon isotope mass balance equation and the nitrogen isotope mass balance equation to construct a weighted correction equation set: in, f i Indicates the first i The percentage of each term contributing to the total organic load of the soil at the test site is expressed as the mass contribution of each term.

[0079] It should be noted that this is the "number". i "The basic contribution ratio of each terminus to the total soil organic matter", while "and" represent "the effective contribution weight in the carbon / nitrogen isotope balance".

[0080] In this embodiment of the invention, a dynamic weighting compensation factor is introduced based on the traditional isotope mass balance equation. The carbon-nitrogen mass loading difference (C / N) of the end-members is transformed into carbon mass weighting operators and nitrogen mass weighting operators, which are then embedded into the carbon and nitrogen isotope mass balance equations to form a weighted correction equation set. This allows the contribution of each end-member in the carbon and nitrogen isotope balance to no longer be treated equally, but rather to be adaptively adjusted according to the degree of asymmetry in the end-member element loading. This mechanism can effectively avoid the problem of "high carbon, low nitrogen" end-members (such as petroleum sources and transportation sources) being underestimated in traditional models due to weak nitrogen signals, while suppressing the phenomenon of falsely high contributions from "high nitrogen, low carbon" end-members (such as natural soil background) due to strong nitrogen signals. This makes the isotope mixing analysis results more consistent with the laws of mass conservation and physical contribution, improving the stability, accuracy, and engineering applicability of contribution rate solutions in multi-source complex pollution scenarios.

[0081] S7: Substitute the weighted correction equations into the Bayesian linear mixture model to calculate the average contribution rate of each pollution source end-member to the accumulation of soil organic matter.

[0082] Among them, the Bayesian linear mixture model is a mixture analysis method based on the Bayesian statistical framework, which is used to make quantitative inferences on multi-source mixture systems under the condition of observation error and uncertainty. Its core idea is to represent the observed index of the sample to be tested as a linear combination of the end-member indexes of each pollution source weighted by the contribution rate, and to introduce the contribution rate as the parameter to be estimated into the prior distribution. By constructing a likelihood function to characterize the error relationship between the observed data and the model prediction, the posterior probability distribution of the contribution rate is obtained by using the Bayesian formula.

[0083] In one possible implementation, S7 specifically includes: S701: Construct a contribution rate parameter vector by taking the contribution rate of each endmember item corresponding to the target endmember feature matrix as the parameter to be estimated.

[0084] S702: Based on the weighted correction equation set, construct the carbon isotope prediction function and nitrogen isotope prediction function for the soil samples of the site to be tested: in, This indicates the contribution rate parameter vector. f Under the given conditions, the predicted carbon stable isotope values ​​of the soil samples from the test site are calculated based on the weighted correction equations. This indicates the contribution rate parameter vector. f Under the given conditions, the predicted values ​​of nitrogen stable isotopes in the soil samples of the test site are calculated based on the weighted correction equation set.

[0085] S703: Correspond the observed isotope values ​​in the sample's three-dimensional feature vector to the carbon isotope prediction function and the nitrogen isotope prediction function, respectively, and establish the observation equation: in, This represents the carbon isotope observation error term. This represents the nitrogen isotope observation error term.

[0086] It should be noted that the error term follows a normal distribution.

[0087] S704: Based on the observation equation, construct the likelihood function of the observation data of the soil sample at the test site under the condition of contribution rate parameter vector.

[0088] Optionally, the likelihood function includes the carbon isotope likelihood function and the nitrogen isotope likelihood function, and the overall likelihood function is the product of the two.

[0089] S705: Set the prior distribution of the contribution rate parameter vector and construct the posterior distribution based on Bayes' theorem.

[0090] Among them, Bayes' theorem is the core foundation of Bayesian statistical inference. It is used to update the probability of unknown parameters or events after obtaining observation data. Its basic idea is to combine prior knowledge and observational evidence to obtain posterior conclusions.

[0091] Optionally, the posterior distribution is as follows: in, This represents the contribution rate parameter vector under the condition that the carbon and nitrogen stable isotope values ​​of the soil samples at the test site are observed. f The posterior probability distribution, Indicates the direct proportion sign. Represents the contribution rate parameter vector f The prior probability distribution.

[0092] S706: Based on the likelihood function, variational Bayesian inference is used to approximate the posterior distribution and construct an approximate posterior distribution.

[0093] Variational Bayesian Inference (VB) is a probabilistic inference method used to approximate the posterior distribution of a Bayesian model. It is mainly used to solve the problem that the posterior distribution in complex models is difficult to calculate analytically or that the computation cost of using MCMC sampling is too high.

[0094] Optionally, to avoid the high computational complexity caused by directly solving the posterior distribution, a variational Bayesian inference method is used to construct an approximate posterior distribution, and maximizing the lower bound of evidence (ELBO) is used as the optimization objective. in, This represents the lower bound of evidence, characterizing how closely the approximate posterior distribution approximates the true posterior distribution; a larger value indicates a better approximation. Represents the contribution rate parameter vector f The approximate posterior distribution, This indicates that the expression within the parentheses is subjected to the expected value operation under the condition of approximate posterior distribution. log ( ) represents a logarithmic function.

[0095] Among them, maximizing the evidence lower bound (ELBO) is the core optimization objective used in variational Bayesian inference to approximate the posterior distribution. Its role is to approximate the true posterior distribution with a computable approximate distribution when the true posterior distribution is difficult to calculate directly, and to make the approximate posterior distribution as close as possible to the true posterior by maximizing ELBO.

[0096] It should be noted that ELBO maximization can achieve efficient approximate inference of the posterior distribution.

[0097] S707: Initialize the approximate posterior distribution parameters of the contribution rate parameter vector and perform iterative updates until the convergence condition is met, then stop the iteration.

[0098] In one possible implementation, the convergence condition specifically includes the ELBO increment between two adjacent iterations being less than a preset increment or the number of iterations reaching a preset maximum number of iterations.

[0099] It should be noted that those skilled in the art can set the preset increment and preset maximum number of iterations according to actual needs, and this invention does not limit them.

[0100] Optionally, the iterative update process is as follows: First, initialize the approximate posterior distribution parameters. Then, in each iteration, generate contribution rate estimates based on the current approximate posterior distribution and substitute them into the carbon isotope prediction function and nitrogen isotope prediction function to obtain predicted values. Next, calculate the log-likelihood value of the current iteration based on the likelihood function, and calculate the evidence lower bound ELBO accordingly. Finally, update the approximate posterior distribution parameters with the goal of maximizing ELBO, until the ELBO increment between two adjacent iterations is less than a preset threshold or the number of iterations reaches the upper limit, at which point the iteration stops.

[0101] S708: After the variational Bayesian inference converges, extract the posterior mean of the contribution rate of each endmember term from the approximate posterior distribution, and use it as the mean contribution rate.

[0102] Optionally, the average contribution rate is as follows: in, Indicates the first i The average contribution rate corresponding to each terminus.

[0103] It should be noted that by introducing a dynamic weight compensation factor, the weighted correction equation set is used as the prediction function and observation equation of the Bayesian linear mixture model, thereby constructing an improved Bayesian mixture analytical model suitable for endmember carbon and nitrogen load asymmetry scenarios, and improving the stability and accuracy of contribution rate inference.

[0104] In this embodiment of the invention, a weighted correction equation set is introduced into a Bayesian linear mixture model. While considering the asymmetry of endmember carbon and nitrogen loadings, observation errors and model uncertainties are explicitly incorporated into the contribution rate solution process in probabilistic form. A likelihood function is formed by constructing carbon and nitrogen isotope prediction functions and observation equations, and the posterior probability distribution is obtained by combining the prior distribution of the contribution rate parameter vector. This avoids the problems of multiple solutions, false solutions, and result fluctuations that easily occur in traditional deterministic solutions when endmember overlap, noise is high, or equations are underdetermined. Furthermore, variational Bayesian inference is employed to maximize the evidence lower bound (ELBO) to achieve an efficient approximate solution for the posterior distribution. While ensuring computational efficiency, the posterior mean of the contribution rate of each endmember is output, enabling the source tracing results to not only have a quantitative contribution ratio but also statistical robustness and interpretability. This significantly improves the stability, accuracy, and engineering applicability of source analysis for soil organic pollution in complex multi-source sites.

[0105] S8: Based on the average contribution rate of each factor, trace the source of pollution of soil organic matter at the test site.

[0106] Specifically, the average contribution rates of each end-factor were normalized to obtain the contribution ratios of coal combustion, oil, traffic emissions, and natural soil background sources to soil organic matter accumulation. Secondly, the pollution sources were ranked according to their average contribution rates to identify dominant and secondary pollution sources. The rationality of the source tracing results was then verified in conjunction with the functional zoning, production processes, and pollution diffusion characteristics of the site under test. Finally, the source tracing results were output as a distribution map or statistical table of the contribution rates of each pollution source, illustrating the main sources of soil organic matter pollution at the site and their relative contribution levels. This provides a basis for subsequent pollution risk assessment, remediation measure selection, and governance priority formulation.

[0107] In this embodiment of the invention, the results are output in the form of a distribution map or statistical table, which facilitates pollution risk assessment and remediation decision-making, enables targeted formulation of remediation measures and priority allocation of resources, thereby improving the engineering application value of site pollution investigation and remediation management.

[0108] Reference manual attached Figure 2 The diagram shows a structural schematic of a pollution source tracing system for soil organic matter provided by the present invention.

[0109] The present invention also provides a pollution source tracing system 20 suitable for soil organic matter, applied to the above-mentioned pollution source tracing method for soil organic matter, comprising: Processor 201.

[0110] The memory 202 stores computer-readable instructions, which, when executed by the processor 201, implement the pollution source tracing method for soil organic matter as described in the method embodiment.

[0111] The pollution source tracing system 20 for soil organic matter provided by the present invention can execute the pollution source tracing method for soil organic matter described above and achieve the same or similar technical effects. To avoid duplication, the present invention will not elaborate further.

[0112] It should be understood that the processor in the embodiments of the present invention can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0113] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0114] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0115] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0116] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0117] It should be understood that, in various embodiments of the present invention, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0118] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0119] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0120] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0121] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0122] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0123] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a 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 this 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.

[0124] This invention provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the pollution source tracing method for soil organic matter as described in the method embodiments.

[0125] The present invention provides a computer-readable storage medium that can implement the steps and effects of the above-described method embodiments applicable to the source tracing method of soil organic matter pollution. To avoid repetition, the present invention will not repeat them.

[0126] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

[0127] The following points need to be explained: (1) The accompanying drawings of the embodiments of the present invention only involve the structures involved in the embodiments of the present invention. Other structures can refer to the general design.

[0128] (2) For clarity, the thickness of layers or regions is enlarged or reduced in the drawings used to describe embodiments of the invention, i.e., these drawings are not drawn to scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being “above” or “below” another element, the element may be “directly” located “above” or “below” the other element or there may be intermediate elements.

[0129] (3) Where there is no conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other to obtain new embodiments.

[0130] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. The scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for tracing the source of soil organic matter pollution, characterized in that, include: S1: Collect soil samples from the site to be tested; S2: Simultaneously measure the soil samples from the site to be tested, and construct a three-dimensional feature vector of the samples; S3: Collect end-member samples of potential pollution sources in the test site and perform simultaneous measurements to construct an end-member feature matrix; S4: Calculate the endmember discrimination index of the endmember feature matrix using the Euclidean distance algorithm, and check and correct the endmember feature vectors in the endmember feature matrix according to the endmember discrimination index to obtain the target endmember feature matrix. S5: Based on the carbon-nitrogen mass ratio of the target endmember in the target endmember feature matrix, calculate the dynamic weight compensation factor including the carbon mass weight operator and the nitrogen mass weight operator; S6: Based on the three-dimensional feature vector of the sample and the feature matrix of the target endmember, establish an isotope mass balance equation, and embed the dynamic weight compensation factor into the isotope mass balance equation to construct a weighted correction equation set. S7: Substitute the weighted correction equations into the Bayesian linear mixture model to calculate the average contribution rate of each pollution source end-member to the accumulation of soil organic matter. S8: Based on the average contribution rate of each of the above, trace the source of soil organic matter pollution at the site to be tested.

2. The method for tracing the source of soil organic matter pollution according to claim 1, characterized in that, S2 specifically includes: S201: The soil sample from the test site is subjected to impurity removal, drying, grinding and sieving to obtain a soil powder sample with uniform particle size. S202: Weigh the soil powder sample according to the preset mass range, and load the weighed soil powder sample into the sample injection container of the elemental analyzer to form the sample injection unit to be tested. S203: Perform a single injection on the sample unit to be tested, and simultaneously obtain the carbon stable isotope value and nitrogen stable isotope value of the whole soil sample in the same measurement process, and simultaneously measure the total organic carbon content and total nitrogen content of the sample unit to be tested. S204: Based on the total organic carbon content and total nitrogen content, calculate the carbon-nitrogen mass ratio of the soil sample from the site to be tested; S205: Combine the carbon stable isotope value, the nitrogen stable isotope value, and the carbon-nitrogen mass ratio of the soil sample from the site to be tested to construct the three-dimensional feature vector of the sample.

3. The method for tracing the source of soil organic matter pollution according to claim 1, characterized in that, S3 specifically includes: S301: Based on the production process, historical pollution records, functional zoning, and on-site survey results of the site to be tested, identify the types of potential pollution sources that contribute to soil organic matter pollution in the site to be tested, and construct an end-member list. S302: According to the end-member list, set up end-member sampling points at typical locations corresponding to each potential pollution source and collect multiple end-member samples. S303: Determine the end-member carbon stable isotope value, end-member nitrogen stable isotope value, and end-member carbon-nitrogen mass ratio of each end-member sample using the same determination method as the soil samples at the test site. S304: Combine the end-member carbon stable isotope value, the end-member nitrogen stable isotope value, and the end-member carbon-nitrogen mass ratio of each end-member term to construct multiple end-member feature matrix vectors. S305: Summarize the endmember feature matrix vectors of each endmember sample according to the endmember sample number, and construct the endmember feature matrix.

4. The method for tracing the source of soil organic matter pollution according to claim 1, characterized in that, S4 specifically includes: S401: Standardize the endmember feature vectors in the endmember feature matrix to obtain a standard endmember feature matrix; S402: Calculate the Euclidean distance between any two endmember feature vectors in the standard endmember feature matrix to form a distance matrix; S403: Obtain the minimum distance between each of the endmember feature vectors and the remaining endmember feature vectors in the distance matrix, and calculate the arithmetic mean of all the minimum distances as the endmember discrimination index; S404: Determine whether the endmember discrimination index of the endmember feature matrix is ​​greater than or equal to the preset endmember discrimination index; if yes, determine that the discrimination between each endmember feature vector in the endmember feature matrix meets the source tracing parsing requirements, determine the endmember feature matrix as the target endmember feature matrix, and proceed to S5; otherwise, determine that the discrimination between each endmember feature vector in the endmember feature matrix does not meet the source tracing parsing requirements, and proceed to the next step. S405: Identify candidate overlapping endmembers in the distance matrix and generate multiple candidate endmember groups; S406: Merge the endmember feature vectors within each candidate endmember group according to the average value of the endmember feature vectors within the group, update the endmember feature matrix, and return to S401 to re-execute.

5. The method for tracing the source of soil organic matter pollution according to claim 1, characterized in that, S5 specifically includes: S501: Extract the end-member carbon-nitrogen mass ratio corresponding to each end-member item from the target end-member feature matrix to form an end-member carbon-nitrogen mass ratio set. S502: Verify the validity of the endmember carbon-nitrogen mass ratios in the set of endmember carbon-nitrogen mass ratios; S503: Based on the carbon-nitrogen mass ratio of each endmember after validity verification, calculate the carbon mass weight operator and nitrogen mass weight operator for each of the endmember items; S504: Summarize the carbon mass weight operator and nitrogen mass weight operator corresponding to each of the aforementioned end-member terms, and calculate the dynamic weight compensation factor.

6. The method for tracing the source of soil organic matter pollution according to claim 1, characterized in that, The isotope mass balance equations include a carbon isotope mass balance equation and a nitrogen isotope mass balance equation; S6 specifically includes: S601: Based on the three-dimensional feature vector of the sample and the feature matrix of the target endmember, establish the isotope mass balance equation; S602: The carbon mass weighting operator and the nitrogen mass weighting operator are respectively embedded into the carbon isotope mass balance equation and the nitrogen isotope mass balance equation to construct the weighted correction equation set.

7. The method for tracing the source of soil organic matter pollution according to claim 1, characterized in that, Specifically, S7 includes: S701: Use the contribution rate of each endmember item corresponding to the target endmember feature matrix as the parameter to be estimated, and construct a contribution rate parameter vector. S702: Based on the weighted correction equation set, construct the carbon isotope prediction function and nitrogen isotope prediction function for the soil sample of the site to be tested; S703: The observed isotope values ​​in the three-dimensional feature vector of the sample are respectively matched with the carbon isotope prediction function and the nitrogen isotope prediction function to establish the observation equation; S704: Based on the observation equation, construct the likelihood function of the observation data of the soil sample of the site to be tested under the condition of the contribution rate parameter vector; S705: Set the prior distribution of the contribution rate parameter vector, and construct the posterior distribution based on Bayes' theorem; S706: Based on the likelihood function, variational Bayesian inference is used to approximate the posterior distribution and construct an approximate posterior distribution; S707: Initialize the approximate posterior distribution parameters of the contribution rate parameter vector and perform iterative updates until the convergence condition is met, then stop the iteration; S708: After the variational Bayesian inference converges, the posterior mean of the contribution rate of each endmember term is extracted from the approximate posterior distribution and used as the mean of the contribution rate.

8. The method for tracing the source of soil organic matter pollution according to claim 1, characterized in that, The convergence conditions specifically include the ELBO increment between two adjacent iterations being less than a preset increment or the number of iterations reaching a preset maximum number of iterations.

9. A pollution source tracing system suitable for soil organic matter, characterized in that, include: processor; A memory storing computer-readable instructions, which, when executed by the processor, implement the pollution source tracing method for soil organic matter as described in any one of claims 1 to 8.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the pollution source tracing method for soil organic matter as described in any one of claims 1 to 8.

Images