Water body sediment heavy metal pollution diagnosis and risk zoning method

Abstract

The invention relates to the technical field of environmental science and engineering, and discloses a water body sediment heavy metal pollution diagnosis and risk zoning method, which comprises the following steps of: firstly, performing functional zoning and differentiated point distribution sampling based on hydrodynamic topographic features; secondly, carrying out multi-dimensional index detection on the sediment sample to obtain a multi-dimensional data set containing a chemical occurrence form; secondly, combining a positive definite matrix factorization model with bivariate local space autocorrelation analysis, quantitatively analyzing the contribution of the pollution source and verifying the space fate of the pollution source; then, constructing a two-dimensional risk assessment system based on the chemical occurrence form, and calculating a migration release risk index and a bio-availability risk index; and finally, according to source-sink positioning and risk assessment results, a differentiated partition control strategy is generated. According to the invention, through coupling chemical morphology and spatial analysis, the concealment risk that the total amount does not exceed the standard but the bioavailability is high can be accurately identified, and accurate positioning and migration path verification of the composite pollution source are realized.

Description

Technical Field

[0001] This invention relates to the field of environmental science and engineering technology, specifically to a method for diagnosing and risk zoning heavy metal pollution in aquatic sediments. Background Technology

[0002] As a sink and potential source of pollutants, the heavy metal pollution status of aquatic sediments is a key indicator for assessing the health of aquatic ecosystems. Currently, the assessment and management of heavy metal pollution in sediments typically relies on monitoring total heavy metal concentrations. The degree of pollution is determined by comparing the measured total concentration with environmental quality standards or geological background values. While this method is intuitive, it has revealed a series of technical limitations in practical applications.

[0003] First, focusing solely on the total amount of heavy metals often masks the true ecological risks of pollutants. The bioavailability and migration / transformation capacity of heavy metals are primarily determined by their chemical forms (such as weakly acid-extractable forms and reducible forms) rather than their total amount. Therefore, total amount assessments may misclassify some heavy metals with low total amounts but extremely high proportions of active forms as low-risk, thus ignoring their potential biotoxicity and secondary release risks.

[0004] Secondly, when dealing with large water bodies such as lakes or reservoirs with multiple input sources and complex hydrodynamics, traditional pollution assessment methods struggle to accurately trace pollution sources and clarify their spatial transport pathways. While pollution source apportionment can identify source types, it lacks effective coupling with spatial geographic information, failing to effectively distinguish between directly input local pollution sources and secondary enrichment sources formed in specific areas (such as deep water areas or backwater bays) after long-distance hydrodynamic transport. This results in vague pollution source location and a lack of spatial precision in remediation.

[0005] Finally, due to the lack of clear risk identification and source location, existing remediation strategies often tend to adopt uniform, large-scale engineering measures, such as dredging the entire reservoir area. This approach is not only extremely costly and inefficient, but also lacks specificity because it fails to fully consider the pollution causes and dominant risk factors in different areas, making it difficult to achieve the expected environmental benefits. Therefore, there is an urgent need for a comprehensive diagnostic method that integrates pollutant speciation, spatial distribution, and pollution source contributions to achieve accurate risk assessment and zoned management of heavy metal pollution in sediments. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a method for diagnosing and risk zoning of heavy metal pollution in aquatic sediments. This method aims to solve the problems of existing technologies that rely solely on total heavy metal assessments, which fail to identify actual ecological risks, are difficult to accurately trace complex pollution sources and their spatial migration paths, and lack targeted control strategies.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for diagnosing and risk zoning of heavy metal pollution in aquatic sediments, comprising the following steps: S10. Based on the basic geographic information data of the target water body, perform functional zoning based on hydrodynamic topographic features, divide the target water body into an external input driving zone and a deep-water sedimentation collection zone, and generate differentiated sampling point layout schemes for collecting surface sediment samples. S20. Perform multidimensional index detection on the surface sediment sample to obtain a multidimensional dataset; the multidimensional dataset includes physicochemical environmental factor data, total heavy metal data, and chemical occurrence speciation data obtained through chemical speciation classification extraction. S30. Using the source-sink spatial coupling analysis module, the multidimensional dataset is subjected to source-sink spatial coupling analysis. The contribution rate of pollution sources is quantitatively analyzed using the positive definite matrix factorization model, and the spatial trend of pollution sources is verified by combining bivariate local spatial autocorrelation analysis, thereby identifying the pollution source factor type and the spatial clustering pattern of heavy metal elements. S40. Based on the chemical occurrence data, construct a two-dimensional ecological risk quantitative assessment system, and calculate the migration and release risk index and the bioavailability risk index respectively, which are used to quantify the anthropogenic enrichment, release potential and biotoxicity of heavy metal elements. S50. Generate a zoned control strategy that coordinates morphology and space. Based on the spatial location information of pollution sources obtained by the source-sink spatial coupling analysis and the risk assessment results output by the dual-dimensional ecological risk quantification assessment system, identify the type of pollution target area using preset logical judgment rules and generate corresponding differentiated control strategies.

[0008] Preferably, the differentiated sampling point layout scheme specifically includes: implementing a dense sampling point strategy in the external input driving area, setting high-density sampling points along the shoreline and tributary inlets, with the sampling point spacing being smaller than the sampling point spacing in the deep-water sedimentation and collection area, to capture concentration abrupt change signals caused by point source emissions; and implementing a grid-based sampling point strategy in the deep-water sedimentation and collection area, using a uniform grid method with sampling points located at the grid center, to reflect the accumulation characteristics of pollutants in a homogeneous hydrodynamic environment.

[0009] Preferably, the chemical speciation data is obtained by classifying heavy metals by chemical speciation using an improved BCR sequential extraction method, specifically including: weakly acid-extractable state, including exchangeable and carbonate-bound heavy metals; reducible state, including iron-manganese oxide-bound heavy metals; oxidizable state, including organic and sulfide-bound heavy metals; and residual state, including heavy metals present in silicate mineral lattices.

[0010] Preferably, before quantitatively analyzing the pollution source contribution using a positive definite matrix factorization model during the source-sink spatial coupling analysis process, the method further includes constructing the input matrix required for the model. The input matrix includes a sample concentration matrix and an uncertainty matrix. The uncertainty value in the uncertainty matrix is ​​calculated using a piecewise function: when the measured heavy metal concentration is less than or equal to the method detection limit, the uncertainty value is calculated using a first calculation rule based on the method detection limit; when the measured heavy metal concentration is greater than the method detection limit, the uncertainty value is calculated using a second calculation rule based on the error fraction and the measured concentration.

[0011] Preferably, the quantitative analysis of pollution source contributions using a positive definite matrix factorization model specifically involves minimizing an objective function through iterative computation to reflect the model's fit. The minimized objective function is defined as the sum of squares of the ratios of residual values ​​to uncertainty. Based on the convergence of the minimized objective function, the contribution rate of each factor and the relative content of heavy metal elements in each factor are analyzed.

[0012] Preferably, the step of verifying the spatial fate of pollution sources by combining bivariate local spatial autocorrelation analysis specifically includes: calculating the global bivariate Moran index to measure the correlation strength of heavy metal speciation or environmental factors within the overall spatial range; based on confirming the global correlation, calculating the bivariate local spatial autocorrelation index to identify the clustering pattern of local areas; and performing spatial overlay analysis on the high-load element combinations analyzed by the positive definite matrix factor decomposition model and the High-High clustering regions. If the element combinations form High-High clusters in the deep-water sedimentation collection area, it is determined to be distant-source sedimentation pollution after hydrodynamic transport.

[0013] Preferably, the calculation of the migration release risk index specifically includes: classifying the weakly acid extractable state, the reducible state, and the oxidizable state as secondary phases, and classifying the residue state as primary phases; calculating the ratio of the sum of the contents of the secondary phases to the contents of the primary phases, and using the ratio as the migration release risk index; when the migration release risk index is greater than or equal to a preset threshold, it is determined that the heavy metal element has a high migration risk.

[0014] Preferably, the calculation of the bioavailability risk index specifically includes: identifying the weak acid extract as a bioavailability contributing component; calculating the percentage of the content of the weak acid extract in the total heavy metal content, and using the percentage as the bioavailability risk index; when the bioavailability risk index is greater than or equal to a preset percentage threshold, determining that the heavy metal element has a high biotoxicity risk.

[0015] Preferably, the identified pollution target area types include agricultural residue source target areas and traffic and hydrodynamic combined source target areas. The logical rule for identifying agricultural residue source target areas is as follows: the agricultural residue source target area is located in the exogenous input driving area, and the heavy metal forms in the sediment are predominantly oxidizable, or the source apportionment results show that the exogenous input driving area is a high-value distribution area of ​​the agricultural residue source factor. The logical rule for identifying traffic and hydrodynamic combined source target areas is as follows: the traffic and hydrodynamic combined source target area covers the connecting path extending from the exogenous input driving area to the deep-water sedimentation collection area, and bivariate local spatial autocorrelation analysis shows that the region exhibits high-high clustering characteristics of heavy metal elements, while the heavy metal forms in the sediment are predominantly weakly acid-extractable or reducible.

[0016] Preferably, the identification of the pollution target area type also includes the identification of the natural weathering source reference area; the judgment logic rule for this area is: the total amount of heavy metals is evenly distributed throughout the entire reservoir area, there is no obvious spatial cluster center, and in terms of chemical characteristics, the residual state occupies the dominant position in the total amount of heavy metals, while the migration and release risk index and the bioavailability risk index are both in the low risk range.

[0017] This invention provides a method for diagnosing and risk zoning of heavy metal pollution in aquatic sediments. It has the following beneficial effects: 1. This invention introduces data on the chemical occurrence forms of heavy metals and constructs a two-dimensional assessment system that includes migration and release risk index and bioavailability risk index. This overcomes the limitations of traditional assessments that rely solely on total heavy metal content. It can effectively identify hidden risks, such as those in Qingshitan Reservoir where the total amount of Mn does not significantly exceed the standard, but the Mn content is dominated by the weakly acidic extractable form, which has high biotoxicity. This provides a more accurate risk warning for environmental management.

[0018] 2. This invention combines the quantitative source analysis results of the positive definite matrix factorization model with bivariate local spatial autocorrelation analysis. Through this coupled analysis, it can not only quantitatively distinguish the contributions of different pollution sources such as agriculture and transportation, but also verify the migration path and final destination of pollutants after hydrodynamic transport from the external source input area by identifying the High-High spatial clustering pattern of specific heavy metal element combinations in the deep water settling zone, thus providing spatial statistical evidence for the precise location of pollution sources.

[0019] 3. Based on the spatial positioning of sources and sinks and the results of dual-dimensional risk assessment, this invention divides water bodies into different types of control areas, such as agricultural residue source target areas and traffic and hydrodynamic composite source target areas. It can formulate differentiated control strategies such as input blocking and emission prevention and reduction for the dominant pollution sources and their risk characteristics in different areas, which solves the problem of insufficient targeting of the traditional unified governance model. It provides technical support for the formation of a scientific paradigm of morphological and spatial coordinated control, and improves governance efficiency and economic benefits. Attached Figure Description

[0020] Figure 1 This is a system framework diagram of the present invention; Figure 2 This is a schematic diagram of the method flow of the present invention; Figure 3 This is a map showing the distribution of physicochemical properties in reservoir sediments; Figure 4 Map showing the distribution of heavy metal content in reservoir sediments; Figure 5 The percentage of heavy metal speciation in reservoir sediments; Figure 6 Distribution of Mn in reservoir sediments by various speciations; Figure 7A A bar chart showing the contribution rate of heavy metals in sediments of the study area based on the PMF model and correlation analysis; Figure 7B A correlation diagram showing the contribution rate of heavy metals in sediments of the study area based on the PMF model and correlation analysis; Figure 8 This is a local Moran I space clustering diagram for Pb-Mn and Pb-Zn bivariate clusters; Figure 9 The RSP and RAC risk indices for heavy metals in reservoir sediments.

[0021] Among them, 100 is the hydrodynamic topography zoning module; 200 is the multi-dimensional indicator detection module; 300 is the source-sink spatial coupling analysis module; 400 is the dual-dimensional ecological risk quantitative assessment module; and 500 is the zoning control strategy generation module. Detailed Implementation

[0022] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] See attached document Figure 1This invention provides a system for diagnosing and risk zoning of heavy metal pollution in water sediments, the system comprising: The hydrodynamic topography zoning module 100 is used to construct a physical spatial framework based on the topographic data and hydrodynamic field distribution characteristics of the target water body. This module divides the water body into an external input-driven zone and a deep-water sedimentation and collection zone, and generates differentiated sampling point layout schemes for different regions. The multi-dimensional index detection module 200 is used to receive the collected sediment samples and obtain data on total heavy metal content, chemical occurrence speciation, and physicochemical environmental factors. The source-sink spatial coupling analysis module 300 processes the detection data based on a positive definite matrix factor decomposition model and a bivariate local spatial autocorrelation analysis algorithm to achieve quantitative analysis and spatial location verification of pollution sources. The dual-dimensional ecological risk quantitative assessment module 400 is used to calculate the migration and release risk index and the bioavailability risk index based on chemical speciation data. The zoning control strategy generation module 500 outputs targeted zoning control strategies based on the analyzed spatial location and risk type of pollution sources.

[0024] See attached document Figure 2 This invention provides a method for diagnosing and risk zoning of heavy metal pollution in aquatic sediments based on morphological and spatial coupling. This method is implemented using the aforementioned system architecture and includes the following steps: S10, based on the topographic data and hydrodynamic field distribution characteristics of the target water body, the water body is divided into an external input driving zone and a deep-water sedimentation and collection zone, and differentiated sampling points are implemented to collect surface sediment samples. S20 involves multidimensional index detection of collected sediment samples to obtain total heavy metal data, chemical speciation data based on the improved BCR sequential extraction method, and physicochemical environmental factor data. S30 uses a positive definite matrix factor decomposition model to calculate the concentration and uncertainty of heavy metals in order to analyze the contribution rate of pollution source factors. It also combines bivariate local spatial autocorrelation analysis to identify the spatial clustering pattern of heavy metal elements, so as to realize the mutual verification of quantitative analysis and spatial location of pollution sources. S40, based on chemical occurrence data, constructs a two-dimensional evaluation system to calculate the migration and release risk index characterizing migration and release potential and the bioavailability risk index characterizing biotoxicity, respectively. S50 identifies pollution target area types based on the spatial location results of pollution source apportionment and the two-dimensional index of risk assessment, divides water bodies into agricultural residue source target areas, traffic hydrodynamic composite source target areas, and natural weathering source reference areas, and generates differentiated management and control strategies.

[0025] The above steps will be described in detail below with reference to specific embodiments and accompanying drawings.

[0026] In a specific implementation of step S10, the hydrodynamic topography zoning module 100 is configured to perform functional zoning and differentiated point placement operations based on hydrodynamic topography features. The hydrodynamic topography zoning module 100 first acquires basic geographic information data of the target water body, including underwater topography data and flow field distribution data. For underwater topography data, a multibeam echo sounder or existing large-scale isobaths can be used to construct a high-precision digital elevation model (DEM). For flow field distribution data, it can be obtained through on-site measurements using an acoustic Doppler current profiler (ADCP) or through simulation based on a hydrodynamic numerical model. Based on the acquired basic geographic information data, the hydrodynamic topography zoning module 100 divides the target water body into an externally driven input zone and a deep-water sedimentation collection zone with different environmental functional attributes, according to the differences in water depth gradient and flow velocity.

[0027] The external input-driven zone is defined as the shallow water area at the edge of the water body or where tributaries flow into the reservoir, while the deep-water sedimentation collection zone is defined as the still water area far from the shoreline and with greater depth. After completing the functional zoning, the hydrodynamic topography zoning module 100 generates differentiated sampling point layout schemes to adapt to the pollution distribution characteristics of different areas.

[0028] In the externally driven area, a dense sampling strategy is implemented. Due to the influence of point source emissions and non-point source runoff, the distribution of pollutants in this area exhibits high spatial heterogeneity. Therefore, high-density sampling points are set up along the shoreline and at tributary inlets. In the deep-water sedimentation catchment area, a grid-based sampling strategy is implemented. Because the hydrodynamic environment in this area is relatively homogeneous, the distribution of pollutants is mainly controlled by large-scale sedimentation processes, resulting in a relatively smooth spatial distribution.

[0029] Based on the generated sampling plan, surface sediment samples were collected during the dry season. The dry season was chosen because the water level is low and the hydrodynamic conditions are relatively stable, which is conducive to sediment collection and maintaining representativeness. After collection, impurities such as stones and plant debris were removed on-site. The samples were then placed in polyethylene self-sealing bags or clean glass bottles and immediately stored in a 4°C vehicle refrigerator away from light. After being transported back to the laboratory, they underwent freeze-drying, grinding, and sieving for subsequent analysis by the multidimensional index detection module 200.

[0030] After sample collection and pretreatment, step S20 uses a multidimensional index detection module 200 to perform subsequent physicochemical environmental factor determination, heavy metal total determination, and chemical speciation extraction on the collected sediment samples. The multidimensional index detection module 200 first analyzes the physicochemical properties of the sediment samples, specifically measuring pH and electrical conductivity (EC). During the measurement process, the air-dried and sieved sediment samples are mixed with decarbonylated distilled water at a solid-liquid ratio of 1:2.5. After shaking and settling, the pH value is measured using the glass electrode method. Simultaneously, a suspension is prepared at a solid-liquid ratio of 1:5, and the electrical conductivity (EC) value is measured using a conductivity meter.

[0031] The multi-dimensional index detection module 200 determines the total amount of heavy metal elements according to environmental standards. The ground and sieved sediment sample is weighed and placed in a polytetrafluoroethylene digestion vessel. The system underwent digestion treatment. Specifically, appropriate amounts of concentrated nitric acid and hydrogen peroxide were added, and the sediment matrix was destroyed under high temperature and high pressure conditions, allowing the analytes to be completely released into the solution. After acid removal, volume adjustment, and filtration, the digestate was analyzed using inductively coupled plasma optical emission spectrometry (ICP-OES) to determine the mass fractions of arsenic (As), lead (Pb), zinc (Zn), copper (Cu), manganese (Mn), and chromium (Cr). During the determination process, national standard soil component analysis reference materials, such as the GSS series, were introduced for simultaneous analysis to control data quality and ensure that the recovery rate met the analytical requirements.

[0032] To elucidate the bioavailability and migration / transformation mechanisms of heavy metals, the multidimensional index detection module 200 employs an improved BCR sequential extraction method for chemical speciation of heavy metals. This method operationally defines the occurrence forms of heavy metals as four binding states, and the extraction steps and environmental implications of each state are as follows: Extraction of weak acid extract state An acetic acid solution with a concentration of 0.11 mol / L was added to the weighed sediment sample, and the mixture was continuously shaken at room temperature for 16 hours. After centrifugation, the supernatant was collected. This form mainly contains exchangeable and carbonate-bound heavy metals, is highly sensitive to changes in environmental pH, represents the portion that can be directly utilized by organisms, and is a major contributor to short-term ecological risk.

[0033] Extracting reducible state At that time, extract into the weak acid state The extracted residue was treated with a 0.5 mol / L hydroxylamine hydrochloride solution, and the pH was adjusted to 1.5. The mixture was continuously shaken at room temperature for 16 hours, and the supernatant was collected after centrifugation. This form mainly contains iron-manganese oxide-bound heavy metals, which are sensitive to changes in the redox potential (Eh). Under anoxic or reducing conditions in the water, the iron-manganese oxides undergo reductive dissolution, releasing the heavy metals adsorbed on them. This is a key component posing a secondary release risk in deep-water sedimentation and accumulation zones.

[0034] Extracting oxidizable states At that time, towards the reducible state The residue after extraction was added in portions with 30% hydrogen peroxide solution and heated to near dryness in an 85°C water bath. Then, ammonium acetate solution with a concentration of 1.0 mol / L was added, and the mixture was shaken and extracted, followed by collection of the supernatant. This form mainly contains organic matter and sulfide-bound heavy metals, reflecting the impact of agricultural non-point source pesticide residues or organic pollution inputs. It is prone to mineralization and release under strong oxidizing conditions.

[0035] Determining the state of residue In this case, the final residue after extraction in the aforementioned steps is digested using the same digestion system as the total amount determination, or it is calculated by subtracting the sum of the first three states from the total amount. This state mainly exists in the silicate mineral lattice, is stable, and does not easily migrate or transform under natural environmental conditions, primarily indicating its natural geological origin.

[0036] Between each extraction step, the multidimensional index detection module 200 performs a cleaning procedure, washing the residue with deionized water and centrifuging to remove residual reagents and prevent cross-contamination between different forms. Through the above steps, the system obtains a multidimensional dataset containing physicochemical properties, total amount, and classification of the four forms, and transmits it to the source-sink space coupling analysis module 300 for further processing.

[0037] In step S30, the source-sink spatial coupling analysis module 300 uses a systematic analysis framework of data preprocessing, spatial analysis, source analysis, and risk assessment to perform in-depth mining of the multidimensional dataset. In this embodiment, spatial analysis and visualization are mainly completed using ArcGIS (e.g., version 10.8) and GeoDa (e.g., version 1.22) software platforms, while source analysis calculation uses the PMF model software released by the US EPA.

[0038] To ensure the comparability and spatial continuity of multi-source heterogeneous data in the source-sink spatial coupling analytical model, the source-sink spatial coupling analytical module 300 first executes the data standardization and interpolation simulation sub-step in step S30.

[0039] For the discrete sampling point data transmitted by the multidimensional index detection module 200, the module first performs a distribution characteristic test. The Shapiro-Wilk test is used to test the normality of all heavy metal contents and physicochemical indicators. For data that does not conform to a normal distribution, the Box-Cox transformation algorithm is used for data transformation. Subsequently, to eliminate the interference of differences in the units and concentrations of different heavy metal elements on the weights of subsequent coupling analysis, the multidimensional index detection module 200 performs Z-score standardization on all variables. The standardization calculation formula is as follows: ; In the formula, These are the standardized values; These are the original observations; This is the arithmetic mean of the variable; This represents the standard deviation of the variable.

[0040] Based on standardized data, a spatially continuous distribution layer of heavy metal content, speciation, and physicochemical indicators was constructed using ordinary kriging interpolation for the multidimensional index detection module 200. During the interpolation process, a spherical model was selected as the semi-variogram model, and the search radius was set to 1.5 kilometers based on the average spacing of the sampling points. Cross-validation was used to evaluate the interpolation accuracy, ensuring that the root mean square error (RMSE) was less than 15%, thus providing a high-precision spatial data foundation for subsequent source analysis and spatial verification.

[0041] After data preprocessing, the source-sink spatial coupling analytical module 300 constructs the input matrices required for the positive definite matrix factorization (PMF) model, including the sample concentration matrix and the uncertainty matrix. To improve the model's analytical accuracy for low-concentration data, a piecewise function is used to calculate the uncertainty of each data point based on the detection limit (MDL) of the detection method and the error fraction of the measurement process.

[0042] When the measured heavy metal concentration is less than or equal to the method detection limit, the first uncertainty calculation formula is used: ; In the formula, This represents the uncertainty value for the corresponding data point; This represents the method detection limit for the heavy metal element in the detection method. When the measured heavy metal concentration value is greater than the method detection limit, the second uncertainty is calculated using the following formula: ; In the formula, This represents the uncertainty value for the corresponding data point; This represents the error fraction, which is usually determined based on the relative standard deviation of laboratory quality control data. This represents the measured heavy metal concentration value for that data point; This represents the method detection limit for the heavy metal element in the detection method.

[0043] The source-sink spatial coupling analytical module 300 inputs the constructed concentration matrix and uncertainty matrix into the PMF model, and uses the weighted least squares method to decompose the heavy metal content matrix of the receptor sample into a factor score matrix, a factor component spectrum matrix, and a residual matrix. The matrix decomposition calculation formula is as follows: ; In the formula, For the first The first sample Content of heavy metals (unit: milligrams per kilogram); For the first The pollution source affects the first Contribution value of each sample (unit: milligrams per kilogram); For the first Individual filth The first of the dye sources The relative content (percentage) of each heavy metal; The number of pollution sources; Let be the residual matrix. Summation satisfies the given meaning.

[0044] To obtain the optimal solution of the model, the objective function is minimized through iterative computation. This reflects the model's fit. Objective function The calculation formula is as follows: ; In the formula, Number of samples; The number of different types of heavy metal elements; The residual value; For the first The first sample The uncertainty of the content of a certain heavy metal. The calculation is based on the relationship between concentration and method detection limit (MDL) in the aforementioned steps. (Module based on...) The convergence of values ​​reveals the contribution rate of each factor, identifying factor types such as natural sources, agricultural sources, and transportation sources.

[0045] To verify the spatial rationality of the PMF analysis results and identify "pollution communities," the source-sink spatial coupling analysis module 300 performs a two-scale spatial autocorrelation analysis. This module first calculates the global bivariate Moran index to quantitatively measure the correlation strength and direction between two variables (e.g., a heavy metal speciation and an environmental factor, or two heavy metal elements) in the overall spatial range. The formula for calculating the global bivariate Moran index is as follows: ; In the formula, The number of samples; Spatial weights; and Variables and variables At the sampling point , The standardized value; , Variables , The standardized mean of the is used for a significance test with 999 random permutations. This is considered as significant spatial autocorrelation. Based on the confirmed global correlation, the module further calculates a bivariate local spatial autocorrelation index to identify clustering patterns in local regions. The autocorrelation index is calculated using the following formula: ; In the formula, Sampling points The local Moran index, with the other parameters having the same meaning as before.

[0046] Based on the calculation results, the module classifies spatial clustering patterns into four categories: High-High clustering (HH, high values ​​surrounded by high values), Low-Low clustering (LL, low values ​​surrounded by low values), High-Low clustering (HL, high values ​​surrounded by low values), and Low-High clustering (LH, low values ​​surrounded by high values). The module performs spatial overlay analysis between the high-load element combinations resolved by the PMF and the High-High clustering regions: if the PMF analysis shows that elements A and B are homologous, and the LISA analysis shows that they form a significant HH cluster in the deep-water sedimentation collection area, then it is determined to be distant-source sedimentation pollution after hydrodynamic transport; if the HH cluster is located in the external input-driven area, it is determined to be near-source direct input. Through this step, the quantitative analysis and spatial location of the pollution source are cross-verified.

[0047] After the source-sink spatial coupling analysis module 300 completes the quantitative analysis and spatial location of pollution sources, the two-dimensional ecological risk quantification assessment module 400 utilizes the chemical speciation data provided by the multi-dimensional index detection module 200. Step S40 initiates the first-dimensional risk assessment procedure, namely, the construction and calculation of the migration and release risk index (RSP). This step aims to overcome the limitations of traditional assessment methods that rely solely on the total amount of heavy metals for risk characterization. By distinguishing the proportional relationship between secondary and primary phases, it accurately quantifies the anthropogenic enrichment of heavy metals in sediments and their potential release into water bodies.

[0048] The dual-dimensional ecological risk quantification assessment module 400 first classifies the four occurrence forms of heavy metals based on their geochemical properties. The weakly acid-extractable, reducible, and oxidizable forms are classified as secondary phases. These components typically originate from anthropogenic pollution inputs, are chemically unstable, and are prone to migration and transformation under changing environmental conditions. The residual form is classified as the primary phase. These components originate from geological weathering processes, are bound within stable mineral lattices, and are not easily released. Based on this classification, the dual-dimensional ecological risk quantification assessment module 400 calculates the ratio of secondary phases to primary phases; this ratio represents the migration and release risk index. The formula for calculating the migration release risk index is as follows: ; In the formula, The migration release risk index is used to characterize the potential migration capacity of heavy metals and the intensity of anthropogenic disturbances. This represents the content of weakly acidic extracts. Represents the content of reducible states; This represents the content of oxidizable states; This represents the content of the residual state. The unit for the content of each of the above forms is milligrams per kilogram.

[0049] After the calculation is completed, the dual-dimensional ecological risk quantification assessment module 400 analyzes the calculated results according to the preset grading standards. The value is used to determine the risk level. When the calculated value... When the value is less than 1, it is judged as a pollution-free state, indicating that heavy metals mainly exist in a stable residual form, with extremely low environmental risk; when A value greater than or equal to 1 and less than 2 is considered a state of mild pollution; when A value greater than or equal to 2 and less than 3 is considered a moderate pollution state; when When the value is greater than or equal to 3, it is judged as a state of severe pollution.

[0050] In this assessment system, the dual-dimensional ecological risk quantitative assessment module 400 is specifically designed for... A high migration risk warning is issued when the value is greater than or equal to 3. The values ​​indicate that most of the heavy metals in the sediments exist in unstable forms and have extremely high release potential.

[0051] The first dimension, migration and release risk index, was completed in the dual-dimensional ecological risk quantification assessment module 400. Following the initial calculation, the module immediately initiates the second-dimensional assessment process: the construction and calculation of the Bioavailability Risk Index (RAC). Unlike the migration-release risk index, which focuses on assessing the geochemical potential of heavy metals released into the aquatic environment, the bioavailability risk index aims to quantify the short-term ecotoxicity of heavy metals directly ingested and utilized by aquatic organisms, thereby improving the risk assessment system at the physiological and toxicological level.

[0052] The dual-dimensional ecological risk quantification assessment module 400, based on the heavy metal speciation theory, identifies the weakly acid-extractable form F1 as a contributing component to bioavailability. Since weakly acid-extractable F1 contains both exchangeable and carbonate-bound forms, this form of heavy metal has the weakest binding force to the surface of sediment particles, making it highly susceptible to entering pore water through ion exchange or weak acid dissolution processes, and subsequently being absorbed by benthic organisms or transferred via the food chain. Therefore, the dual-dimensional ecological risk quantification assessment module 400 defines the proportion of weakly acid-extractable F1 in the total heavy metal content as the risk assessment code, i.e., the bioavailability risk index. The formula for calculating the bioavailability risk index is as follows: ; In the formula, Represents the bioavailability risk index, used to characterize the direct biotoxic potential of heavy metals; , , and The physical meaning of these terms is consistent with the definition in the formula for calculating the migration and release risk index, representing the content of the weakly acid-extractable, reducible, oxidizable, and residual forms, respectively. The denominator is the sum of the contents of the four forms, representing the total amount of the heavy metal element.

[0053] The dual-dimensional ecological risk quantitative assessment module 400 is based on the calculated... Numerical values ​​are used to classify biotoxicity risks according to globally accepted environmental monitoring grading standards. When When the value is less than 1%, it is considered risk-free; when A value greater than or equal to 1% and less than 10% is considered low risk; when A value greater than or equal to 10% and less than 30% is considered medium risk; when A value greater than or equal to 30% and less than 50% is considered high-risk; when When the value is greater than or equal to 50%, it is judged as extremely high risk.

[0054] In this assessment system, the dual-dimensional ecological risk quantitative assessment module identifies 400 key areas. High-risk elements with a value greater than or equal to 30%. This is addressed by introducing a bioavailability risk index. The technical solution constructed in this invention can convert elements with high biotoxicity (high... By effectively distinguishing between risk characteristics and elements with high environmental mobility (high RSP characteristics), the risk type can be accurately identified, thus laying a data foundation for the subsequent development of differentiated control strategies for different risk mechanisms.

[0055] Based on the risk assessment results output by the dual-dimensional ecological risk quantification assessment module 400 and the pollution source spatial location information provided by the source-sink spatial coupling analysis module 300, the zoning control strategy generation module 500 executes the precise zoning control strategy generation step of morphological and spatial coordination in step S50.

[0056] The zoning control strategy generation module 500 first identifies agricultural residue source target areas, which are defined as input-blocking control areas. The logical rules for determining this target area are: spatially, the area must be located in the exogenous input-driven area defined in step S10; chemically, the heavy metals in the sediments of this area must be in an oxidizable state. The primary factor is agricultural residue, or source apportionment results show that the contribution rate of agricultural residue source factors to the region is higher than a preset threshold. For the identified agricultural residue source target areas, the module generates governance strategies centered on input blocking and activity regulation.

[0057] The zoning control strategy generation module 500 further identifies target areas with combined traffic and hydrodynamic sources, which are defined as emission prevention and reduction control zones. The logical rules for determining this target area are as follows: spatially, the area covers a connected path extending from the external input driving zone to the deep-water sedimentation collection zone, and bivariate local spatial autocorrelation analysis shows a significant high-high clustering characteristic of heavy metal elements in the area; in terms of risk characteristics, the heavy metals in the sediments of this area are primarily in a weakly acid-extractable state. or reducible state Dominated by this, corresponding to a higher risk index of migration and release. or bioavailability risk index For this target area, the zoning control strategy generation module 500 generates a comprehensive treatment strategy encompassing source reduction and process release prevention. At the source, it controls shipping fuel emissions and domestic sewage input; at the collection point, it focuses on implementing redox environmental regulation at the sediment-water interface. Because the reducible state... Highly sensitive to reducing environments, technical solutions include installing bottom-layer aeration devices or covering the sediment with oxidizing modifiers in deep water areas to maintain the redox potential (Eh) of the sediment surface at the oxidized state level, thereby maintaining the stability of iron and manganese oxides and preventing the secondary release of bound heavy metals under anoxic reducing conditions. Finally, the zoning control strategy generation module 500 identifies the natural weathering source reference area, which is defined as the background monitoring control area. The logical rule for determining this target area is: the total amount of heavy metals is evenly distributed throughout the entire reservoir area, with no obvious spatial clustering center; in terms of chemical characteristics, the residue state... It occupies an absolute dominant position in the total amount of heavy metals and has a migration and release risk index. With bioavailability risk index All areas are within the low-risk range. For this region, the zoned control strategy generation module 500 generates a background monitoring strategy, incorporating it as an environmental background reference point into the routine monitoring network, but exempting it from engineering remediation or treatment measures to avoid unnecessary waste of resources.

[0058] To further verify the effectiveness and advancement of the water sediment heavy metal pollution diagnosis and risk zoning system and method based on morphological and spatial coupling proposed in this invention, a typical large reservoir in Southwest China (Qingshitan Reservoir) was selected as the research object for verification, and the following embodiments are given.

[0059] The spatial distribution of As, Mn, Cr, Pb, Cu, and Zn in the sediments of Qingshitan Reservoir exhibits a typical west-to-east convergence pattern. The densely populated human-inhabited area along the western shore is the main source of external pollutants, while the deeper water area in the east becomes a pollutant deposition and enrichment zone due to weakened hydrodynamics. This macroscopic pattern, together with the spatial gradients of key environmental factors such as pH and electrical conductivity (EC) in the sediments, constitutes the environmental geochemical background for heavy metal migration and transformation. The statistical characteristics of their content are shown in Table 1.

[0060] Table 1. Statistics on heavy metal content in sediments of the study area Table 1 shows significant differences in the content of the six heavy metals, with Zn having the highest average content, followed by Mn. The coefficient of variation (CV) reveals the spatial heterogeneity of the elements. As's CV > 0.90 indicates a significant influence from local anthropogenic inputs and a highly uneven spatial distribution; while Cr and Mn have CVs of 0.43 and 0.41 respectively, showing relatively uniform distributions, presumably mainly controlled by natural soil formation processes or dispersed anthropogenic sources. Skewness and kurtosis analysis shows that Cu and As have skewness values ​​> 0 and high kurtosis values, exhibiting a right-skewed frequency distribution with sharp peaks and thick tails. This suggests the existence of a small number of extremely high concentrations in the samples, far exceeding the average, providing evidence for point source pollution.

[0061] Compared with the background values ​​of soils in northeastern Guangxi, the average contents of Zn, Pb, and As all exceeded the background values. Among them, Zn enrichment was the most obvious (2.7 times the background value), while Pb and As were 2.2 times and 1.7 times the background values, respectively. The average Cr content was slightly lower than the background value, and although Cu was higher than the background value, the excess was small. This indicates that the sediments in the study area have been generally affected by the enrichment of Zn, Pb, and As.

[0062] The results of key physicochemical parameters of the sediments in Qingshitan Reservoir are shown in Table 2. The pH value of the sediments ranged from 4.98 to 6.90, indicating a generally weakly acidic environment; the EC ranged from 22.90 to 415.00 μS / cm, showing significant regional differences.

[0063] Table 2. Physicochemical characteristics of reservoir sediments Reference Appendix Figure 3The sediment pH exhibits a spatial gradient increasing from west to east. Along the western coast, the pH ranges from 4.98 to 5.65, creating an acidic environment that promotes the dissolution and release of carbonate-bound heavy metals (such as F1-bound Mn and Zn). In the deeper eastern waters, the pH ranges from 6.46 to 6.87, tending towards neutral, which is more conducive to the adsorption and fixation of heavy metal ions, constituting a key environmental factor influencing the heavy metal dissolution-precipitation balance. Bivariate global Moran's index analysis (Table 3) shows a significant negative spatial correlation between pH and all detected heavy metals (p < 0.05), with Moran's indices ranging from -0.117 to -0.158. This spatial statistical analysis confirms the synchronous distribution of low pH areas and high heavy metal content, further revealing the promoting effect of acidic environments on heavy metal enrichment.

[0064] Table 3. Moran's Index of Heavy Metals and pH in Reservoir Sediments Reference Appendix Figure 3 The spatial distribution of EC (23.89-414.32 μS / cm) shows an increasing trend from west to east. The high-value area (>294.76 μS / cm) is stably distributed in the deep water area in the east, and it coincides with the area rich in heavy metals.

[0065] This spatial coupling relationship may be driven by multiple processes, with hydrodynamic sorting playing a key role. The reservoir's topography, characterized by shallower water in the west and deeper water in the east, results in weaker water flow dynamics in the deeper eastern area, which is more conducive to the sedimentation of fine particulate matter suspended with heavy metals. This leads to an increase in the total ion content in the sediments and pore water in this area. Early diagenetic processes may also be a significant influencing factor. The stable sedimentary environment in the eastern deep-water area may have fostered reducing conditions at the sediment-water interface, potentially triggering the dissolution of solid components such as iron and manganese oxides. This would release previously adsorbed heavy metals into the pore water, a process that could further contribute to the increase in pore water conductivity. Therefore, the high EC values ​​in the eastern area may not be directly caused by external inputs, but rather are more likely a comprehensive manifestation of a series of geochemical processes involving the migration, sedimentation, and early diagenesis of pollutants within the reservoir area. This result supports the inference that the eastern deep-water area is a major sink for pollutants.

[0066] Reference Appendix Figure 4Based on the reservoir's topography, hydrodynamic conditions, and the distribution of human activities, the study area was divided into three functional zones: the northern inlet area, the western coastal driving area, and the eastern deep-water collection area. Spatial interpolation analysis showed that all heavy metals exhibited two high-value zones, located at the reservoir inlet and the eastern deep-water collection area, respectively. This distribution pattern was mainly controlled by the synergistic effects of reservoir topography, hydrodynamic processes, and human activities. The formation of the high-value zone at the northern inlet was primarily related to external inputs and physical sedimentation: as the direct receiver of watershed materials, the sudden decrease in water flow velocity at the inlet led to primary sedimentation of suspended particulate matter from various sources, resulting in heavy metal enrichment. The formation of the high-value zone in the eastern deep-water area was closely related to the reservoir's "shallow in the west and deep in the east" topographical characteristics. This area, with its greater water depth and slower flow, became the final collection area for fine-grained sediments and bound heavy metals.

[0067] The specific spatial distribution characteristics of the elements are as follows: As: The content ranges from 2.52 to 184.22 mg / kg, with the highest concentration concentrated along the northern and southeastern coasts. This distribution highly coincides with the distribution in agricultural planting areas and aquaculture areas, indicating that fishery feed input and historical application of arsenic-containing pesticides are the main input pathways, exhibiting obvious point source pollution characteristics. Common heavy metals Zn, Pb, and Cu: All show a combination of non-point source input and migration / deposition characteristics. The Zn content ranges from 13.22 to 485.74 mg / kg, with a high-value zone along the western coast. This originates from non-point source inputs such as domestic sewage from coastal villages and drainage from fishponds. The high-concentration area extends into the central and eastern parts of the reservoir area, forming a broad medium-to-high concentration deposition zone, clearly reflecting the process of Zn being transported from the western source area to open waters and then deposited. The Pb content ranged from 24.84 to 191.19 mg / kg, with a spatial distribution similar to Zn but slightly weaker migration. The highest value area was closer to the western coast, suggesting that it rapidly settled after combining with larger particles. Meanwhile, a second-highest value area appeared in the eastern deep water area, indicating that some fine-grained Pb settled in a still water environment after long-distance transport. The Cu content ranged from 10.92 to 155.66 mg / kg, showing a point-source-area source coherent distribution. A high-value core existed in the west, which is speculated to be related to the peeling off of anti-corrosion coatings from fish farm cages and the loss of feed additives. Furthermore, a widespread medium-concentration plateau formed in the central reservoir area, reflecting the mixing and diffusion process of Cu after being exported from point sources.

[0068] The concentration of Cr, the dominant natural background element, ranged from 0.05 to 121.82 mg / kg, with a relatively uniform spatial distribution. The concentration in most areas of the reservoir ranged from 43.16 to 78.72 mg / kg, indicating that the primary source was weathering of natural parent material. However, small-scale high-value areas appeared in the western part of the reservoir, suggesting possible anthropogenic input from industrial or transportation sources. Mn concentrations ranged from 264.03 to 1645.89 mg·kg⁻¹, with a relatively uniform spatial distribution. However, the average concentration across the entire reservoir and the concentration in the deep-water areas of the central and eastern parts significantly exceeded the regional background value, with the highest value being 2.8 times the background value. The extreme values ​​varied considerably, exhibiting a complex characteristic of natural background pollution combined with anthropogenic pollution. The high concentration areas were mainly distributed in the central and eastern deep-water areas, reflecting the reduction and dissolution of manganese oxides in the sediments (diffusion upwards in the form of Mn²⁺) under the quiescent and anoxic environment at the reservoir bottom. Subsequently, particulate matter was reformed and deposited at the oxidation interface, leading to secondary enrichment in the deep-water areas. This reveals a complex geochemical cycle and spatial redistribution process of Mn in the reservoir.

[0069] Reference Appendix Figure 5 Based on an improved BCR sequential extraction method, the distribution characteristics and correlation patterns of six heavy metals (weakly acid-extractable form F1, reducible form F2, oxidizable form F3, and residual form F4) in the sediments of Qingshitan Reservoir were investigated. According to their speciation and environmental behavior, they can be classified into highly reactive elements (Mn), medium-risk elements (Pb, As), and low-risk elements (Cr, Cu, Zn). Significant differences exist in the speciation, environmental sensitivity, and potential risks of each element category.

[0070] The highly reactive element Mn exhibits the strongest chemical activity and bioavailability, with its F1 and F2 states accounting for nearly 60% combined. It is extremely sensitive to changes in environmental pH and redox conditions. The medium-risk elements Pb and As are mainly in stable forms, but they have unique morphological characteristics and potential release risks—Pb is dominated by the F2 state, while As shows obvious spatial differentiation, with the reactive form significantly enriched in the western anthropogenic activity area. The low-risk elements Cr, Cu, and Zn are absolutely dominated by the F4 state, each accounting for more than 60%.

[0071] Mn is the element with the highest environmental reactivity and potential risk among the six elements. Its speciation is mainly composed of F1 state (31.21%) and F2 state (27.56%), accounting for a total of 58.77%. The total Mn content is significantly positively correlated with the F1 state (r=0.722, p<0.01), indicating that Mn is mainly found in unstable forms such as carbonate-bound and iron-manganese oxide-bound forms, and is highly sensitive to environmental conditions. Under acidic conditions, F1 state Mn is easily released directly; in reducing environments, F2 state Mn can be activated by the dissolution of iron-manganese oxides.

[0072] Reference Appendix Figure 5 -Appendix Figure 6The spatial distribution of Mn-F2 states exhibits a unique pattern, with abundant accumulation in the near-shore oxidizing environment of the west and significant reduction in the deep-water reducing environment of the east.

[0073] The speciation of Pb exhibits significant specificity, with the F2 state accounting for as much as 51.43%, indicating that it is mainly bound to iron and manganese oxides. The total Pb content shows a very significant positive correlation with all four speciations, reflecting the complexity of its origin—it includes both a stable geological background (F4) and significant anthropogenic input of active components (F1, F2, F3). This dominance of the F2 state in its occurrence means that it faces a significant risk of secondary release when the sedimentary environment changes to reducing conditions.

[0074] As exhibits significant morphological differentiation and a clear spatial pattern. High values ​​of As-F3 are observed along the western coast, in scattered areas in the south, and at the northern reservoir head, closely coinciding with areas of intensive agricultural activity, indicating that anthropogenic input plays a dominant role in the distribution of its reactive form. As-F4 is enriched in the eastern deep-water area, which is related to the higher pH and fine particle sedimentation processes in this region. The predominance of reactive As in the western region suggests a potential release risk under specific environmental conditions.

[0075] Both Cr and Cu are predominantly in the F4 state, accounting for over 60%, and their total content shows a strong positive correlation with the residual state content, but no significant correlation with the active state. This indicates that these elements are mainly fixed in the primary mineral lattice, and their total amount is largely controlled by natural background values. They are not easily affected by changes in pH and redox conditions in exogenous environments, exhibiting extremely low mobility and bioavailability, and their ecological risks are controllable. Specifically, the distribution of Cr forms shows a pattern of active form in the west and stable form in the east. The Cr-F3 state is slightly enriched in the western human-inhabited areas, but the absolute content is low. The Cu-F2 state content of Cu is close to the detection limit, and the distribution of active forms is mainly related to the high organic matter content in the western region.

[0076] The Zn-F4 state accounted for 68.01%, but its total content showed a very significant positive correlation with all non-residual states (F1+F2+F3), indicating that Zn enrichment mainly originated from exogenously introduced anthropogenic active components. The non-residual states accounted for 31.9% in total, of which the F1 state, sensitive to environmental acidification, and the F2 state, releaseable under reducing conditions, accounted for 17.8%, giving them some migration potential and bioavailability, but their overall risk remained lower than that of Mn and Pb.

[0077] To investigate the potential sources and synergistic migration behavior of heavy metals in the sediments of Qingshitan Reservoir, Pearson correlation analysis was conducted on six elements. The results showed that there were significant differences in the correlations among the elements, revealing the complexity of their sources and geochemical behavior.

[0078] The analysis results show that Mn, Cu, Zn, and Pb all exhibit extremely significant positive correlations, with Zn and Pb having a correlation coefficient as high as 0.875 and Mn and Zn having a correlation coefficient of 0.727, strongly suggesting that they may be controlled by a common source or have similar geochemical behaviors and migration and transformation pathways in sedimentary environments. Cr is significantly positively correlated with all elements except Pb, indicating that its distribution has a certain degree of synergy with other elements, but its source is more complex, influenced by both natural background and some anthropogenic inputs. As has unique correlations with other elements, showing significant positive correlations with Cu, Mn, and Cr, but a weak correlation with Zn (r=0.266, p<0.05) and only a moderate correlation with Pb, indicating that although As has some overlapping sources or behaviors with other heavy metals, it may have a relatively independent dominant source.

[0079] In summary, the correlation analysis preliminarily reveals that heavy metals in the sediments of Qingshitan Reservoir are influenced by multiple sources. Among them, Mn, Cu, Zn, and Pb may originate from a strongly correlated combination of common sources, while As and Cr show some independent source characteristics, providing preliminary evidence for subsequent quantitative source apportionment.

[0080] Source apportionment analysis was performed on the concentrations and uncertainties of six metals in 72 sediment samples using EPA PMF 5.0 software. All elements had a signal-to-noise ratio (S / N) > 9, classifying them as "Strong". After experimenting with different numbers of factors and runs, 20 runs and 3 factors were ultimately determined, with most residuals within the range of -3 to 3, indicating good model fit. Uncertainty analysis of the model showed robustness and reliability: Bootstrap analysis showed a matching degree of over 90% for all factors, and DISP analysis confirmed that the fluctuation range of each factor's contribution rate was less than ±5%. Table 4 shows that, except for slightly lower R² values ​​for Mn and Cu, the correlation coefficients (R²) between the measured and predicted values ​​for the other elements were all > 0.80, and the predicted-to-measured ratio (P / O) was close to 1.00, verifying the reliability of the PMF model. The relatively low R² values ​​for Mn and Cu may be related to their active biogeochemical cycles and multi-source nature in the sedimentary environment.

[0081] Table 4. Fitting results of measured and simulated predicted values ​​of heavy metal content in reservoir sediments Reference Appendix Figure 7A -Appendix Figure 7B The PMF model identified three main pollution source factors, and the contribution rates of each factor to each heavy metal are as follows.

[0082] PM1: It has the highest contribution rate to Cr (83.50%), and also has obvious contributions to Cu (46.1%) and Zn (40.5%). The secondary contributing elements are Mn and Pb. Combining that Cr is significantly positively correlated with Cu, Zn, Mn, and As (P < 0.01), and the variation coefficients of Cr and Zn are relatively low (0.4 < CV < 0.6), it is speculated that this factor reflects natural sources (such as weathering of soil parent materials). In addition, the occurrence form of Cr mainly in residual state and the characteristics that the Cr content in most samples is lower than the background value further rule out the possibility dominated by human input. A large number of studies have shown that Cr, Cu, and Mn originate from soil parent materials. Therefore, PM1 is identified as the background source dominated by regional parent material weathering and natural geochemical processes, and its contribution can be used for environmental baseline deduction.

[0083] PM2: Its contribution rate to As reaches 79.3%, and the contributions to the rest of the elements are all < 11.6% (Cr 11.2%, Pb 8.7%, Cu 11.6%), indicating its high specificity for the source of As. The As-F3 state (oxidizable state) shows obvious high-value distribution near farmland and fish ponds on the western coast, further verifying from the morphological spatial pattern that PM2 mainly represents the source of agricultural activities. As in the agricultural system mainly comes from the residues of historical arsenic-containing pesticides (such as lead arsenate, calcium arsenate), livestock and poultry feed additives (roxarsone), and long-term application of arsenic-containing phosphate fertilizers. A large number of studies have confirmed that As is significantly correlated with agricultural activities, especially forming high-value aggregations in areas with intensive use of chemical fertilizers and pesticides. The surrounding area of the study area is a traditional planting area of rice and citrus, where arsenic-containing pesticides were widely used historically, and there is a lack of systematic soil remediation measures, resulting in the continuous accumulation of As in sediments. At the same time, cage fish farming was intensively carried out in the study area, and the bait feeding further increased the input of As. Therefore, PM2 is identified as the source of historical agricultural activities and pesticide residues, and is the core target for As pollution control.

[0084] PM3: Its contribution rate to Pb reaches 61.8%, and it has a secondary contribution of 56.95% to Zn, and basically no contribution to Cr (≤0.2%). The residents around the study area used small oil-consuming boats as means of transportation and tourism vehicles. In addition, after heavy metals such as Pb from other sources in the region are adsorbed on particulate matter, they can migrate to the reservoir along with the runoff formed by atmospheric rainfall, and under the regulation of hydrodynamic differentiation, settle in static water environments such as the west side and the deep water area in the east of the reservoir. The spatial distributions of both the Pb-F1 state and the Pb-F2 state show continuous high values in the west side and the deep water area in the east of the reservoir, highly coinciding with the ship channels and static water areas of the reservoir. And these two forms usually indicate human input sources. Among them, Zn mainly comes from ship anti-rust paint and galvanized steel plates, Pb comes from the exhaust gas emitted by the combustion of leaded gasoline in ships, Cu comes from domestic sewage and traffic pollution (ships, road traffic), and Mn may come from domestic sewage (detergents) and traffic pollution. Therefore, PM3 is identified as the traffic pollution - hydrodynamic composite anthropogenic source, which is the main anthropogenic pollution source faced by the reservoir.

[0085] To explore the intrinsic connections and synergistic patterns in the spatial distribution of different heavy metals, bivariate local spatial autocorrelation analysis (Bivariate LISA) was used to analyze the spatial correlations of six heavy metals. This method can effectively identify spatial co-enrichment or dispersion patterns among elements, providing spatial statistical evidence for judging source homology and migration process similarity. As shown in Table 6, there are significant differences in the spatial correlation strength among the elements, but all passed the significance test (p<0.001), revealing that the spatial distribution of heavy metals in the study area has obvious structural characteristics.

[0086] Table 6. Double Moran Index of Heavy Metals in Reservoir Sediments Reference Appendix Figure 8 The bivariate Moran indices for Zn-Mn, Zn-Pb, and Pb-Mn were 0.584, 0.567, and 0.516, respectively, showing extremely strong to very strong positive correlations. Furthermore, a large-scale, continuous "high-high (HH)" cluster was formed at the northern inlet of the reservoir and in the eastern deep-water area. This spatial model is highly consistent with the PMF source apportionment results, confirming that Zn, Pb, and Mn have a high degree of homology (mainly originating from a combined transportation-hydrodynamic source), and undergo similar migration-deposition paths after entering the reservoir, ultimately accumulating together in the same hydrodynamic and geochemical sink area, forming a clear "pollution community."

[0087] Spatial manifestation of elemental origin complexity: The bivariate Moran index of Cu with Zn, Pb, and Mn ranges from 0.253 to 0.476, showing a moderate to strong positive correlation, but the correlation strength is lower than that of Zn-Mn and Zn-Pb. Spatial verification of the PMF model results: Although Cu partly originates from a combined transportation-hydrodynamic source, there are independent local inputs related to fishery activities along the western coast (such as anti-corrosion coatings for net cages and feed additives), causing its spatial distribution to deviate somewhat from that of core elements such as Zn and Pb.

[0088] The Moran index for the bivariate Cr and Zn was 0.292, showing a moderate positive correlation. However, unlike other indicators, the "high-high" clustering of Cr-Zn was concentrated at the northern inlet of the reservoir, rather than in the deep water area in the eastern part of the reservoir. This phenomenon may be caused by hydrodynamic sorting.

[0089] Based on RSP and RAC, a quantitative assessment of the potential ecological risks of heavy metals in the sediments of Qingshitan Reservoir was conducted from two dimensions: geochemical activity (migration and release potential) and bioavailability (short-term ecological risk). The results are attached. Figure 9 As shown, there are significant differences in the risk levels of the six heavy metal elements.

[0090] The RSP assessment results showed that Pb had the most prominent risk of migration and release, with 72% of the sites having an RSP value >3 and an average RSP of 3.53, indicating a heavily polluted level. This suggests that Pb was highly activated in the sediment and was significantly affected by anthropogenic input. Mn had an average RSP value of 1.80, which is in the range of light to moderate pollution and has a moderate migration potential. Zn, As, Cu, and Cr all had average RSP values ​​<1, indicating an overall unpolluted level, with only a few sites showing moderate pollution due to local anthropogenic input.

[0091] The RAC assessment results showed that Mn had the highest bioavailability, with an average RAC value of 33% and a maximum of 57%, which is classified as high-risk and has significant short-term ecotoxicity potential. Pb and Zn had average RAC values ​​of 5% and 7%, respectively, which are at low-risk levels, with low bioavailability and low risk of environmental release. As, Cu, and Cr all had RAC values ​​of <1%, extremely low bioavailability, and negligible short-term ecological risks.

[0092] In sedimentary heavy metal environmental geochemical studies, the sum of the weakly acid-extractable (F1), reducible (F2), and oxidizable (F3) states is often defined as the non-residual state. This component represents the active portion of heavy metals with high bioavailability, strong migration capacity, and the greatest potential ecological risk in the natural environment. Based on the combined RSP and RAC assessment results, the potential ecological risk ranking of heavy metals in the sediments of Qingshitan Reservoir is: Pb > Mn > As > Zn > Cu > Cr.

Claims

1. A method for diagnosing and risk zoning heavy metal pollution in water sediments, characterized in that, Includes the following steps: S10. Based on the basic geographic information data of the target water body, perform functional zoning operation based on hydrodynamic topographic features to divide the target water body into an external input driving zone and a deep water sedimentation and collection zone, and generate a differentiated sampling point layout scheme for collecting surface sediment samples. S20. Perform multidimensional index detection on the surface sediment sample to obtain a multidimensional dataset; The multidimensional dataset includes physicochemical environmental factor data, full data of heavy metals, and chemical occurrence data obtained through chemical speciation extraction. S30. Using the source-sink spatial coupling analysis module, perform source-sink spatial coupling analysis on the multidimensional dataset, use the positive definite matrix factorization model to quantitatively analyze the contribution rate of pollution sources, and combine bivariate local spatial autocorrelation analysis to verify the spatial trend of pollution sources, thereby identifying the pollution source factor types and the spatial clustering pattern of heavy metal elements. S40. Based on the chemical occurrence data, construct a two-dimensional ecological risk quantitative assessment system, and calculate the migration and release risk index and the bioavailability risk index respectively to quantify the anthropogenic enrichment, release potential and biotoxicity of heavy metals. S50. Generation of zoned control strategies that combine execution form and spatial coordination: Based on the spatial positioning information of pollution sources obtained by the source-sink spatial coupling analysis and the risk assessment results output by the dual-dimensional ecological risk quantification assessment system, the types of pollution target areas are identified using preset logical judgment rules, and corresponding differentiated control strategies are generated.

2. The method for diagnosing and risk zoning heavy metal pollution in aquatic sediments according to claim 1, characterized in that, In step S10, the generation of differentiated sampling point layout scheme specifically includes: A dense sampling point deployment strategy is implemented in the external input driving area, with high-density sampling points set up along the shoreline and tributary inlets. The spacing between sampling points is smaller than that between sampling points in the deep-water sedimentation collection area, in order to capture concentration mutation signals caused by point source discharge. A grid-based sampling strategy was implemented in the deep-water sedimentation and collection area. A uniform grid method was used, with sampling points located at the center of the grid, to reflect the accumulation characteristics of pollutants in a homogeneous hydrodynamic environment.

3. The method for diagnosing and risk zoning heavy metal pollution in aquatic sediments according to claim 1, characterized in that, In step S20, the chemical speciation data obtained through chemical speciation fractionation is obtained by fractionating the heavy metals using a modified BCR sequential extraction method. The chemical speciation data includes: The weak acid extractable state includes exchangeable and carbonate-bound heavy metals; Reduceable state, including heavy metals bound to iron and manganese oxides; Oxidizable state, including organic matter and sulfide-bound heavy metals; The residue state contains heavy metals present in the silicate mineral lattice.

4. The method for diagnosing and risk zoning heavy metal pollution in aquatic sediments according to claim 1, characterized in that, In step S30, before quantitatively analyzing the pollution source contribution rate using the positive definite matrix factorization model, the following steps are also included: The input matrix required to construct the positive definite matrix factorization model includes a sample concentration matrix and an uncertainty matrix. The uncertainty values ​​in the uncertainty matrix are calculated using a piecewise function: When the measured heavy metal concentration is less than or equal to the method detection limit, the uncertainty value is solved using the first uncertainty calculation formula; when the measured heavy metal concentration is greater than the method detection limit, the uncertainty value is solved using the second uncertainty calculation formula.

5. The method for diagnosing and risk zoning heavy metal pollution in aquatic sediments according to claim 4, characterized in that, In step S30, the contribution of pollution sources is quantitatively analyzed using a positive definite matrix factorization model. Specifically, the objective function is minimized through iterative calculation to reflect the model's fit. The minimized objective function is defined as the sum of squares of the ratios of residual values ​​to uncertainty. Based on the convergence of the minimization objective function, the contribution rate of each factor and the relative content of heavy metals in each factor are analyzed.

6. The method for diagnosing and risk zoning heavy metal pollution in aquatic sediments according to claim 1, characterized in that, In step S30, the spatial fate of the pollution source is verified by combining bivariate local spatial autocorrelation analysis, specifically including: The global bivariate Moran index is calculated using the global bivariate Moran index calculation formula and is used to measure the correlation strength of heavy metal speciation or environmental factors in the overall spatial range. Based on the confirmation of global correlation, the autocorrelation index of the bivariate local space is calculated using the autocorrelation index calculation formula to identify the clustering patterns of local regions. Spatial overlay analysis is performed on the high-load element combination obtained from the positive definite matrix factorization model and the High-High clustering region. If the element combination forms a High-High cluster in the deep-water sedimentation collection area, it is determined to be a distant-source sedimentation pollution after hydrodynamic transport.

7. The method for diagnosing and risk zoning heavy metal pollution in aquatic sediments according to claim 3, characterized in that, In step S40, calculating the migration release risk index specifically includes: The weakly acid extractable state, the reducible state, and the oxidizable state are classified as secondary phases, and the residue state is classified as primary phase; Calculate the ratio of the sum of the contents of the secondary phases to the contents of the primary phases, and use the ratio as the migration and release risk index; When the migration release risk index is greater than or equal to a preset threshold, the heavy metal element is determined to have a high migration risk.

8. The method for diagnosing and risk zoning heavy metal pollution in aquatic sediments according to claim 3, characterized in that, In step S40, calculating the bioavailability risk index specifically includes: The weak acid extract state was identified as a bioavailability-contributing component; Calculate the percentage of the content of the weakly acidic extract in the total heavy metal content, and use the percentage as the bioavailability risk index; When the bioavailability risk index is greater than or equal to a preset ratio threshold, the heavy metal element is determined to have a high biotoxicity risk.

9. The method for diagnosing and risk zoning heavy metal pollution in aquatic sediments according to claim 1, characterized in that, In step S50, the identification of pollution target area types includes identifying agricultural residue source target areas and traffic and hydrodynamic combined source target areas; The logical rule for identifying agricultural residue source target areas is as follows: the agricultural residue source target area is located in the external input driving area, and the heavy metal form in the sediment is dominated by the oxidizable state, or the source apportionment results show that the external input driving area is a high-value distribution area of ​​the agricultural residue source factor; The logical rule for identifying the combined traffic and hydrodynamic source target area is as follows: the combined traffic and hydrodynamic source target area covers the connecting path extending from the external input driving area to the deep water sedimentation collection area, and bivariate local spatial autocorrelation analysis shows that the area has high-high clustering characteristics of heavy metal elements, while the heavy metal forms in the sediment are mainly in the weakly acid extractable or reducible form.

10. The method for diagnosing and risk zoning heavy metal pollution in aquatic sediments according to claim 1, characterized in that, In step S50, identifying the type of pollution target area also includes identifying a reference area for natural weathering sources; The logical rule for identifying the reference area of ​​natural weathering source is as follows: the total amount of heavy metals is evenly distributed throughout the entire reservoir area, with no obvious spatial clustering center, and in terms of chemical characteristics, the residual state dominates the total amount of heavy metals. At the same time, the migration and release risk index and the bioavailability risk index are both in the low-risk range.

Images