Lake water heavy metal pollution risk assessment method based on multi-source analytical model

Abstract

The invention relates to a lake water body heavy metal pollution risk assessment method based on a multi-source analytical model, and the method comprises the following steps: S1, sample collection: collecting lake water body samples by using a water collector in a dry season and a wet season respectively, carrying out membrane treatment on the samples through a microporous filter membrane, and removing suspended matters in the samples to obtain a water sample to be detected; s2, sample detection: pretreating a water sample to be detected, determining the concentrations of Pb, Cu, Zn, Co, As, Cd, Cr, Mn and Ni in the water sample by using ICP-MS (Inductively Coupled Plasma Mass Spectrometry), and determining the concentration of Hg in the sample by using atomic fluorescence spectrometry; and S3, heavy metal spatial distribution characteristics: analyzing the heavy metal spatial distribution characteristics based on inverse distance weighted interpolation in Arc-Gis. According to the method, the double-stage sampling design of the dry season and the wet season is combined, the spatial and temporal distribution characteristics of the heavy metals under different hydrological conditions can be comprehensively reflected, the limitation of a single-index evaluation method is made up, and the method is high in adaptability and suitable for preliminary diagnosis of heavy metal risks of various water bodies.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of heavy metal pollution detection, and particularly relates to a lake water body heavy metal pollution risk assessment method based on a multi-source analysis model. BACKGROUND

[0002] With the rapid economic development and the acceleration of urbanization and industrialization, the heavy metal pollution problem in rivers, lakes, reservoirs and other types of water bodies is becoming increasingly serious. According to relevant research, the main pollutants include mercury (Hg), lead (Pb), cadmium (Cd), arsenic (As) and other highly toxic metal ions. Once these heavy metals enter the water environment, they not only have significant persistence and migration, but also exhibit strong biological accumulation effects, can be long-term enriched in plants and animals, and can damage their normal physiological functions, and even pose a threat to human health through the food chain. For example, Cd can cause kidney dysfunction and osteoporosis, and Pb can cause irreversible damage to the red blood cell system and the central nervous system. Because of its difficulty in natural degradation, heavy metals have become one of the key pollution factors that threaten the safety of the ecological system and the living environment of human beings.

[0003] At present, the research on conventional pollutants such as nutrients and suspended solids in lake water bodies is relatively in-depth, but there are still obvious deficiencies in the systematic monitoring and pollution source identification of heavy metal pollution in water bodies. On the one hand, the existing monitoring is mostly based on point data, which is not strong in spatial representation, and it is difficult to fully reflect the distribution characteristics of heavy metal pollution in the lake area. On the other hand, the sources of heavy metals are complex, including industrial wastewater, domestic sewage, and may also be related to geological background or agricultural runoff. The existing research methods are limited in pollution source identification, lack of comprehensiveness and systematicness, and it is difficult to provide accurate basis for regional water environment management and pollution prevention and control.

[0004] In the aspect of pollution risk assessment, the single-factor index method and the Nemerow comprehensive pollution index method are commonly used, which have the advantages of simplicity and practicality, but have significant technical limitations: the single-factor method often overestimates the pollution risk and is difficult to reflect the overall pollution level, while the Nemerow index is dominated by the largest single indicator and easily hides the superposition effect of other pollutants, which is insufficient in comprehensiveness. At the same time, these evaluation models mostly lack geographical spatial expression ability and cannot depict the continuous distribution pattern and spatial heterogeneity of heavy metal pollution at the lake scale. In the aspect of pollution source analysis, the traditional correlation analysis and principal component analysis method is commonly used in practice, but its results depend on manual experience for principal component interpretation, lack of quantitative support, in addition, PCA does not consider measurement error and data uncertainty, which is easy to cause source category confusion or contribution rate deviation. In recent years, the PMF model has certain quantitative resolution ability, but it has high requirements for the quality and uncertainty of input data, and the model stability is limited, which is not suitable for lake type regions lacking continuous monitoring basis. At present, there is no mature heavy metal pollution system analysis method system. SUMMARY

[0005] In view of the above problems, the present application provides a lake water heavy metal pollution risk assessment method based on a multi-source analysis model, which not only combines the double-period sampling design of dry season and wet season to comprehensively reflect the spatial and temporal distribution characteristics of heavy metals under different hydrological conditions, but also makes up for the limitations of single indicator evaluation method, has strong adaptability, and is suitable for preliminary diagnosis of heavy metal risk in various water bodies.

[0006] In order to achieve the above objects and other related objects, the technical scheme provided by the present application is as follows: A lake water heavy metal pollution risk assessment method based on a multi-source analysis model, the method comprising: S1. Sample collection: collecting lake water samples using a water sampler in dry season and wet season, and performing membrane treatment on the samples through a microporous filter to remove suspended solids, to obtain a to-be-tested water sample; S2. Sample detection: after pretreatment of the to-be-tested water sample, using ICP-MS to determine the concentrations of Pb, Cu, Zn, Co, As, Cd, Cr, Mn and Ni in the water sample, and using atomic fluorescence spectrometry to determine the concentration of Hg in the sample; S3. Heavy metal spatial distribution characteristics: analyzing the spatial distribution characteristics of heavy metals based on inverse distance weighted interpolation in Arc-Gis; S4. Heavy metal pollution degree evaluation: identifying the pollution degree of each heavy metal by the single-factor index method, and then analyzing the synergistic pollution effect of multiple heavy metals by the Nemerow comprehensive pollution index method to evaluate the pollution risk level; S5. Heavy metal source analysis: S51. Correlation analysis and principal component analysis: correlation analysis is performed using a Spearman correlation coefficient matrix to identify heavy metal clusters with homologous correlation, and principal component analysis is used to identify the dominant pollution source category characteristics; S52. Positive matrix factorization model: PMF is used to quantitatively analyze the contribution rate of pollution sources to heavy metals.

[0007] Further, in step S1, the sampling points of the water sampler collecting the lake water sample are uniformly distributed in space gradient, covering the near-shore area and the center area of the lake.

[0008] Further, in step S2, the pretreatment of the water sample to be measured includes: in the laboratory, using a disposable syringe, the water sample brought back to the laboratory is subjected to membrane treatment through a 0.4-0.5 μm microporous filter membrane, the purpose is to separate solid and liquid, and remove suspended matter, then 4-6 ml of the membrane-treated water sample is taken into a microwave digestion tube, and a CEM-Mars6 microwave digestion instrument is used to digest the water sample, the heating time is set to 9-11 min, the digestion temperature is 165-175℃, the digestion power is 900-1100W, and the holding time is 15-25 min, 2-4 ml of high-purity nitric acid is added before digestion, and after digestion is completed, the liquid in the digestion tube is transferred to an acid removal tube, and perchloric acid is added for acid removal, until the liquid is only a few drops and white smoke is generated, then the acid is considered to be removed, the temperature is set to 75-85℃, and the acid removal is performed for 23-25 h, after the acid removal is completed, the sample is subjected to membrane treatment again, and is diluted to 95-105 ml with 1-3% nitric acid, then is loaded into a sample tube, and the corresponding label paper is attached for measurement.

[0009] Further, in step S2, the determination of the heavy metal content includes: before measurement, the water sample to be measured needs to be diluted with 1-3% nitric acid solution, and is prepared into 0 μg / L, 10 μg / L, 50 μg / L, 100 μg / L, 200 μg / L, 300 μg / L, 400 μg / L, and 500 μg / L standard water samples, at the same time, In and Ge are used as internal standards, the standard water sample concentration is used as the abscissa, and the ratio of the sample signal to the internal standard signal is used as the ordinate to draw a standard curve, then the water sample solution is taken for machine measurement to obtain the concentration of heavy metals in the river water body, and when the heavy metals are measured, the blank sample and the parallel sample are used for quality control.

[0010] Further, in step S3, the calculation function for analyzing the spatial distribution characteristics of heavy metals based on the inverse distance weighted interpolation in Arc-Gis is, , wherein Z so is the estimated value of the heavy metal concentration of so at the undetected point, and Z siis the monitoring value of the i-th point around so, n is the number of monitoring points in the local area, W i is the weight coefficient from point si to so.

[0011] Further, in step S4, the calculation function of the single-factor index method for identifying the pollution degree of each heavy metal is, , wherein P i is the single-factor index of heavy metal i, C i is the detected concentration value of heavy metal i, C o is the standard limit value corresponding to heavy metal i; The classification standard of water pollution degree is: when the single-factor index P i of heavy metal i is less than 1, the water quality level is non-polluted, when P i is 1-2, the water quality level is low-polluted, when P i is 2-3, the water quality level is moderately polluted, and when P i is greater than 3, the water quality level is heavily polluted.

[0012] Further, in step S4, the calculation function of the Nemerow comprehensive pollution index method for analyzing the synergistic pollution effect of multiple heavy metals and evaluating the pollution risk level is, , wherein P is the comprehensive pollution evaluation index, P max is the maximum pollution index in the water quality parameter, P i is the single pollution index; The classification standard of the evaluation of pollution risk level is: when the comprehensive pollution evaluation index P is less than 1, the water quality level is non-polluted, when P is 1-2, the water quality level is moderately polluted, when P is 2-3, the water quality level is polluted, when P is 3-5, the water quality level is heavily polluted, and when P is greater than 5, the water quality level is severely polluted.

[0013] Further, in step S51, the correlation analysis is used to analyze the degree of correlation between various heavy metals, and the homologous heavy metals are identified, wherein the correlation coefficient r reflects the degree of correlation between variables in a numerical manner, generally r is between -1 and 1, if 0 < r < 1, it represents that there is a positive correlation between two variables, if -1 < r < 0, it represents that there is a negative correlation between two variables, |r| = 1 represents that the value of one variable completely depends on the other variable, r = 0 represents that there is no linear correlation between two variables, but it does not exclude the possibility of other nonlinear relationships, 0.8 ≤ |r| ≤ 1 is considered as high correlation, 0.5 ≤ |r| < 0.8 is considered as moderate correlation, 0.3 ≤ |r| < 0.5 is considered as low correlation, 0 ≤ |r| < 0.3 indicates that the degree of correlation between two variables is weak and can be considered as unrelated, and the calculation steps of r are as follows: 1) The original data is arranged in rows to form a matrix X; 2) The data is standardized to make the mean value zero and the variance 1; 3) The correlation coefficient matrix R is calculated; , wherein, r ij ( i,j = 1,2…p) is the correlation coefficient of the original variable x i and x j , r ij = r ji , and the calculation function is as follows: , wherein, x ki is the element of the kth row and the ith column in the matrix X, x kj is the element of the kth row and the jth column in the matrix X, is the average rank of the original variable x i , is the average rank of the original variable x j .

[0014] Further, in step S51, the principal component analysis to identify the dominant pollution source class characteristics includes: 1) Solve the characteristic equation |λE-R| = 0 to obtain the characteristic value and arrange it according to its size; 2) Solve the characteristic vector a i = [a i1 , a i2 ,⋯,a ip ], i = 1,2,…,p, and ; 3) Calculate the contribution rate and the cumulative contribution rate, , , wherein R is a correlation coefficient matrix, lambda is an eigenvalue matrix, E is an identity matrix, lambda i and lambda k are elements in the eigenvalue matrix lambda, p is a positive integer, a ij is an element in the eigenvector a i corresponding to the eigenvalue.

[0015] Further, in step S52, the function model for quantitatively analyzing the contribution rate of the pollution source to the heavy metal by using the PMF is, , wherein X ij is the concentration of the jth element in the ith sample, g ik is the contribution of the kth source to the ith sample, f kj is the concentration of the jth element in the kth source, e ij is the error value of the concentration of the jth element in the ith sample.

[0016] The present application has the following positive effects: 1. The present application adopts spatially uniform arrangement of sampling points, covers different functional areas and entry channels of the lake, and simultaneously combines double-period sampling design of dry season and wet season, so that the spatial and temporal distribution characteristics of heavy metals under different hydrological conditions can be comprehensively reflected. Sample analysis adopts high-sensitivity detection method combining ICP-MS and atomic fluorescence method, so as to guarantee the representativeness and accuracy of data.

[0017] 2. In the evaluation method, the present application comprehensively adopts single-factor pollution index and Nemerow comprehensive index, so that the risk of exceeding the standard of single heavy metal can be identified, and the degree of combined pollution of multiple pollutants can be evaluated, thereby better making up for the limitation of single index evaluation method, and being strong in adaptability and suitable for preliminary diagnosis of heavy metal risks of various water bodies.

[0018] 3. With the aid of GIS platform and inverse distance weighted (IDW) interpolation method, the present application converts point data into continuous surface distribution map, so as to realize the visual expression of the heavy metal pollution pattern of the lake area, and facilitate the identification of pollution high-value area, potential input channel and partition management unit.

[0019] 4. The present application can effectively distinguish industrial, agricultural and other composite sources by the combination of correlation analysis, principal component analysis and PMF model, so as to clearly define the contribution proportion of different sources to various heavy metals, and has strong operability while guaranteeing scientific explanation, and is suitable for pollution tracing analysis and responsibility division in environmental management.

[0020] 5. The application can be applied to the heavy metal pollution investigation and diagnosis of typical lakes, artificial reservoirs, aquaculture areas and wetlands and other types of water bodies, and is especially suitable for popularization and application in water environment management, protection of backup water source areas, ecological compensation evaluation and the like, and has good practical value and popularization prospect. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 FIG. 1 is a schematic diagram of different sampling point positions in the experimental example of the application; Figure 2 FIG. 2 is a water body heavy metal spatial distribution graph (A) in the experimental example of the application; Figure 3 FIG. 3 is a water body heavy metal spatial distribution graph (B) in the experimental example of the application; Figure 4 FIG. 4 is a schematic diagram of water body heavy metal single factor pollution index in the experimental example of the application; Figure 5 FIG. 5 is a schematic diagram of water body heavy metal Nemerow pollution index in the experimental example of the application; Figure 6 FIG. 6 is a water body heavy metal correlation matrix graph in the experimental example of the application; Figure 7 FIG. 7 is a water body heavy metal principal component analysis graph in the experimental example of the application; Figure 8 FIG. 8 is a water body heavy metal PMF source component spectrum in the experimental example of the application; Figure 9 FIG. 9 is the total contribution rate of various pollution sources to water body heavy metals in the experimental example of the application. DETAILED DESCRIPTION

[0022] Exemplary embodiments of the present disclosure are described below with reference to the accompanying drawings, which include various details of the embodiments of the present disclosure to assist in understanding, and should be considered as merely exemplary. Thus, those of ordinary skill in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and structures are omitted in the following description for clarity and conciseness.

[0023] Embodiment 1: A lake water body heavy metal pollution risk assessment method based on a multi-source analysis model, the method comprising: S1. Sample collection: lake water samples are collected using a water sampler in the dry season and the wet season, and the samples are subjected to membrane treatment using a microporous filter to remove suspended solids, to obtain test water samples; S2. Sample detection: after pretreatment of the test water samples, the concentrations of Pb, Cu, Zn, Co, As, Cd, Cr, Mn and Ni in the water samples are determined using ICP-MS, and the concentration of Hg in the samples is determined using atomic fluorescence spectrometry; S3. Heavy metal spatial distribution characteristics: based on inverse distance weighted interpolation in Arc-Gis to analyze the spatial distribution characteristics of heavy metals; S4. Heavy metal pollution degree evaluation: identify the pollution degree of each heavy metal by single factor index method, and analyze the synergistic pollution effect of multiple heavy metals by Nemerow comprehensive pollution index method to evaluate the pollution risk grade; S5. Heavy metal tracing analysis: S51. Correlation analysis and principal component analysis: correlation analysis is carried out by using Spearman correlation coefficient matrix to identify heavy metal clusters with homologous correlation, and principal component analysis is used to identify the characteristics of the main pollution source category; S52. Positive matrix factorization model: PMF is used to quantitatively analyze the contribution rate of pollution sources to heavy metals.

[0024] In this embodiment, in step S1, the sampling points of the water sampler collecting the lake water body sample are uniformly distributed in space gradient, covering the near-shore area and the center area of the lake.

[0025] In this embodiment, in step S2, the pretreatment of the water sample to be tested includes: in the laboratory, using a disposable syringe, the water sample taken back to the laboratory is subjected to membrane treatment through a 0.4-0.5 μm microporous filter film, the purpose is to separate solid and liquid, and remove suspended matter, then 4-6 ml of the water sample after membrane treatment is taken in a microwave digestion tube, and a CEM-Mars6 microwave digestion instrument is used to digest the water sample, the heating time is set to 9-11 min, the digestion temperature is 165-175℃, the digestion power is 900-1100W, and the holding time is 15-25 min, 2-4 ml of high-purity nitric acid is added before digestion, after digestion is completed, the liquid in the digestion tube is transferred to an acid removal tube, and perchloric acid is added for acid removal, until the liquid is only a few drops and white smoke is generated, then the acid is considered to be removed, the temperature is set to 75-85℃, and the acid removal is performed for 23-25 h, after the acid removal is completed, the sample is subjected to membrane treatment again, and is diluted to 95-105 ml with 1-3% nitric acid, then is loaded into a sample tube, and the corresponding label paper is attached for measurement.

[0026] In this embodiment, in step S2, the determination of the heavy metal content includes: before measurement, the water sample to be tested needs to be diluted with 1-3% nitric acid solution, and is prepared into standard water samples with concentrations of 0 μg / L, 10 μg / L, 50 μg / L, 100 μg / L, 200 μg / L, 300 μg / L, 400 μg / L and 500 μg / L, In and Ge are used as internal standards, a standard curve is drawn with the standard water sample concentration as the abscissa and the ratio of the sample signal to the internal standard signal as the ordinate, then the water sample solution is taken for machine measurement to obtain the concentration of heavy metals in the river water body, and blank samples and parallel samples are used for quality control when the heavy metals are measured.

[0027] In the present embodiment, in step S3, the calculation function of analyzing the spatial distribution characteristics of heavy metals based on the inverse distance weighted interpolation analysis in Arc-Gis is, , wherein Z so is the estimated value of the heavy metal concentration of so at the undetected point, Z si is the monitoring value of the i-th point around so, n is the number of monitoring points in the local area, W i is the weight coefficient pointing from point si to so.

[0028] In the present embodiment, in step S4, the calculation function of identifying the pollution degree of each heavy metal by the single factor index method is, , wherein P i is the single factor index of heavy metal i, C i is the detected concentration value of heavy metal i, and C o is the standard limit value corresponding to heavy metal i. The division standard of water pollution degree is: when the single factor index P i of heavy metal i is less than 1, the water quality level is non-polluted, when P i is 1-2, the water quality level is low pollution, when P i is 2-3, the water quality level is moderate pollution, and when P i is greater than 3, the water quality level is strong pollution.

[0029] In the present embodiment, in step S4, the calculation function of analyzing the synergistic pollution effect of multiple heavy metals by the Nemerow comprehensive pollution index method and evaluating the pollution risk level is, , wherein P is the comprehensive pollution evaluation index, P max is the maximum pollution index in the water quality parameter, and P i is the single pollution index. The division standard of the evaluation of pollution risk level is: when the comprehensive pollution evaluation index P is less than 1, the water quality level is non-polluted, when P is 1-2, the water quality level is moderate pollution, when P is 2-3, the water quality level is pollution, when P is 3-5, the water quality level is heavy pollution, and when P is greater than 5, the water quality level is serious pollution.

[0030] In the embodiment, in step S51, the correlation analysis is used to analyze the correlation between various heavy metals, and the homologous heavy metals are identified, wherein the correlation coefficient r reflects the correlation between variables in a numerical manner, generally, r is between -1 and 1, if 0 < r < 1, it represents that there is a positive correlation between two variables, if -1 < r < 0, it represents that there is a negative correlation between two variables, |r| = 1 represents that the value of one variable completely depends on the other variable, r = 0 represents that there is no linear correlation between two variables, but the possibility of other nonlinear relationships cannot be ruled out, 0.8 ≤ |r| ≤ 1 is considered as high correlation, 0.5 ≤ |r| < 0.8 is considered as moderate correlation, 0.3 ≤ |r| < 0.5 is considered as low correlation, 0 ≤ |r| < 0.3 indicates that the degree between two variables is weak and can be considered as unrelated, and the calculation steps of r are as follows: 1) The original data is arranged in a matrix X according to rows; 2) The data is standardized to make the mean value zero and the variance 1; 3) The correlation coefficient matrix R is calculated; , wherein, r ij ( i,j = 1,2…p) is the correlation coefficient of the original variable x i and x j , r ij = r ji , and the calculation function is as follows: , wherein, x ki is the element of the kth row and the ith column in the matrix X, x kj is the element of the kth row and the jth column in the matrix X, x i is the average rank of the original variable x j , and x i is the average rank of the original variable x i1 .

[0031] In the embodiment, in step S51, the principal component analysis to identify the dominant pollution source category characteristics includes: 1) The characteristic equation |λE-R| = 0 is solved to obtain the characteristic value and arrange it according to the size; 2) The characteristic vector a i2 = [a ip , a i ,⋯,a k ], i = 1,2,…,p corresponding to the characteristic value is calculated, and ; 3) The contribution rate and the cumulative contribution rate are calculated, , , wherein R is a correlation coefficient matrix, λ is an eigenvalue matrix, E is an identity matrix, λ i and λ k are elements in the eigenvalue matrix λ, p is a positive integer, a ij is an element in the eigenvector a i corresponding to the eigenvalue.

[0032] In this embodiment, in step S52, the function model for quantitatively analyzing the contribution rate of the pollution source to the heavy metal by using the PMF is , wherein X ij is the concentration of the jth element in the ith sample, g ik is the contribution of the kth source to the ith sample, f kj is the concentration of the jth element in the kth source, e ij is the error value of the concentration of the jth element in the ith sample.

[0033] Example 2: Based on the lake water heavy metal pollution risk assessment method based on the multi-source analysis model in Example 1, the present application is further described and explained as follows.

[0034] Taking Gehu Lake as an example, as Figure 1 or Figure 2 or Figure 3 , the present application provides a lake water heavy metal pollution risk assessment and tracing method based on a multi-source analysis model, comprising the following steps: Step 1: Selecting Gehu Lake in Jiangsu Province as the target water body, the sampling area covers the north, middle and south of the lake body and the main inlet channel. A total of 10 sampling points are set, numbered as S1 to S10, and the point distribution takes into account the spatial uniformity and typicality. Two rounds of sampling were carried out in October 2020 (dry season) and June 2021 (wet season) to understand the heavy metal pollution status in different hydrological periods.

[0035] Step 2: Using a polyethylene sampling bottle to collect water samples on site, and after being brought back to the laboratory, filtering through a 0.45 μm filter membrane, moving into a microwave digestion tube, adding 3 mL of high-purity nitric acid, and performing microwave digestion under the conditions of 10 min of warming time, 170 ℃ of digestion temperature, 20 min of holding time, and 1000 W of power. After digestion, the acid was chased, and the volume was adjusted to 100 mL before measurement.

[0036] The concentrations of Pb, Cu, Zn, Co, As, Cd, Cr, Mn and Ni in the water samples were determined by ICP-MS, and the concentration of Hg was determined by atomic fluorescence spectrometry. Before determination, the instrument was calibrated by configuring a multi-point standard curve, and the quality control was performed by setting a blank sample and a parallel sample.

[0037] In this example, the measured concentrations of all elements were lower than the limit values of the Class III standard in the Environmental Quality Standards for Surface Water (GB 3838-2002), indicating that the overall water quality of Lake Gi met the basic environmental function requirements of surface water during the investigation period. Further statistical analysis of the coefficient of variation (CV) of each element concentration was used to evaluate the uniformity of its spatial distribution and the degree of human influence. In this example, the coefficient of variation of Hg was 14%, indicating that its spatial distribution was relatively uniform and less affected by human disturbance. The coefficients of variation of Pb, Cu, Zn, Co, As, Cd, Cr, Mn and Ni were between 30% and 60%, which belonged to the medium variation level, indicating that there were significant differences in space and that they were greatly affected by human activities. The results reflect that there may be heavy metal inputs related to industrial emissions, urban sewage and agricultural runoff in some areas of Lake Gi. Among them, there are more electroplating, printing and dyeing, metal processing and other industrial enterprises, as well as concentrated residential areas and agricultural land in the northern and western regions of Lake Gi, indicating that industrial wastewater, domestic sewage and pesticide application may be the main exogenous input paths of the above elements. Overall, the heavy metal concentrations in the water body of Lake Gi are within a controllable range, but the spatial variability is obvious, and some elements are significantly affected by human activities, with typical regional input characteristics. This analysis provides a data basis and judgment basis for subsequent pollution risk assessment and pollution source analysis.

[0038] Step three: Based on the heavy metal concentrations of each sample point, the inverse distance weighted interpolation method (IDW) in Arc-Gis was used to analyze the spatial distribution characteristics of heavy metals, and the spatial distribution pattern of 10 heavy metals was simulated. The calculation formula is as follows: , where Z so is the estimated value of the heavy metal concentration of so at the undetected point, Z si is the monitoring value of the i-th point around so, n is the number of monitoring points in the local area, and W i is the weight coefficient of point si to so.

[0039] In this example, the concentrations of Hg, Cu, Co and Cd show a distribution trend of increasing from south to north, with high values mainly appearing in the northern region of Gehu Lake. The concentrations of As and Cr are relatively high in the northwest and southeast regions, and relatively low in the northeast and southwest regions, both of which have similar spatial distribution, suggesting potential homology between their sources. Mn shows a distribution pattern of low in the south and high in the north, low in the east and high in the west, with a clear concentration gradient, which may be related to the direction of water flow and changes in the sedimentary environment. The spatial pattern of Ni concentration is special, showing an increasing trend from the southwest and northeast to the east of the lake, with a concentration peak near S6. The low concentration areas of Pb are scattered, mainly near S1, S3, S4, S6 and S10, and the high concentration appears in S2, S7, S8 and other sites. The concentration of Zn is relatively high in the west and low in the east, with high concentration in S1, S4 and S9, and low concentration in S2 and S8. Overall, the concentration of heavy metals in the water body shows the characteristics of high in the north and low in the south, high in the west and low in the east. This distribution pattern is related to the geographical location of the industrial intensive area in the north of Gehu Lake and the pollution load of the west rivers (such as Huangli River, Beigan River and Mengjin River) entering the lake.

[0040] It is shown that the IDW interpolation model established by the present application can effectively reflect the spatial migration characteristics and regional input path of pollutants. Through the interpolation results, the spatial distribution map of the main heavy metals can be obtained to identify the high value area of pollution, the trend direction and the potential input channel.

[0041] Step four: as shown in Figure 4 or Figure 5 The present application uses the single factor pollution index method and the Nemerow comprehensive pollution index method to evaluate the pollution degree of 10 heavy metals (Hg, Pb, Cu, Zn, Co, As, Cd, Cr, Mn and Ni) in the water body of Gehu Lake.

[0042] (1) The single factor index method is used to evaluate the pollution degree of a single heavy metal, and the calculation formula is: , Wherein, P i is the single factor index of heavy metal i, C i is the concentration value of heavy metal i, and C o is the standard limit value corresponding to heavy metal i.

[0043] The single factor index pollution classification standard is as follows:

[0044] (2) The Nemerow pollution index method is used to comprehensively evaluate the overall pollution level of multiple heavy metals, and the calculation formula is: , P = Pmax max Pmax is the maximum pollution index of water quality parameters, P i Pmax is the maximum pollution index of water quality parameters, P

[0045] The classification of the comprehensive pollution index is as follows:

[0046] In this example, the single-factor pollution index of each heavy metal element is less than 1, and the water body is in a non-polluted state. The pollution index of each heavy metal from high to low is: Hg (0.294) > Pb (0.118) > Ni (0.100) > Zn (0.082) > Cr (0.061) > As (0.059) > Cd (0.055) > Cu (0.005). Among them, the index of Pb in the wet season is 0.178, which is significantly higher than that in the normal water period (0.058), showing that the pollution level increases in the wet season, while the other elements are opposite, indicating that most of the heavy metals except Pb are obviously affected by the dilution effect of rainwater in the wet season, while Pb increases in the wet season, which may be related to the input of Pb in the atmosphere and soil caused by the increase of surface runoff, showing strong characteristics of non-point source pollution. Combined with the calculation of the pollution index of each sampling point in the wet and dry seasons by the Nemerow index method, the results are all less than 1, which determines that the overall water quality of the water area is non-polluted. From the spatial dimension, the Nemerow average value of each point from high to low is: S1 (0.332) > S6 (0.326) > S5 (0.316) > S7 (0.307) > S2 (0.303) > S10 (0.297) > S4 (0.295) > S9 (0.293) > S3 (0.285) > S8 (0.280); From the time dimension, the dry season (0.328) > the wet season (0.279), the index fluctuation between different sampling points is small, indicating that the overall water quality of the lake is relatively uniform in space distribution, and there is no local high-risk pollution area. The combination of the above two methods can identify the key pollution factors and the overall pollution risk, and provide a basis for subsequent source analysis.

[0047] Step five: heavy metal source analysis (1) Correlation analysis: such as Figure 6To analyze the correlation between different heavy metal elements in the water body of Gehu Lake and preliminarily infer the potential source, the SPSS 22.0 statistical analysis software is used to perform correlation analysis on the sampling data. Before the correlation analysis, firstly, the normality test is performed on the data. Considering that the sample amount of each heavy metal is less than 50, the Shapiro-Wilk (S-W) test is mainly used. The test result shows that the concentration data of the 10 kinds of heavy metal elements do not pass the normality test (p<0.05), and present the non-normal distribution characteristics. Therefore, the Spearman rank correlation coefficient is selected to analyze the correlation between the elements, and the calculation steps of the correlation coefficient r are as follows: 1) the original data is arranged in a matrix X according to rows; 2) the data standardization is performed on X, so that the mean value becomes zero and the variance becomes 1; 3) the correlation coefficient matrix is solved; , wherein, r ij (i,j=1,2…p) is the correlation coefficient of the original variable x i and x j , r ij =r ji , and the calculation function is as follows: , wherein, x ki is the element of the kth row and the ith column in the matrix X, x kj is the element of the kth row and the jth column in the matrix X, is the average rank of the original variable x i , is the average rank of the original variable x j .

[0048] r accurately reflects the correlation between variables in the form of a numerical value, and generally r is in-1-1. If 0<r<1, it represents that there is a positive correlation between the two variables, if-1<r<0, it represents that there is a negative correlation between the two variables, |r|=1 represents that the value of one variable completely depends on the other variable, r=0 represents that there is no linear correlation between the two variables, but the possibility of other nonlinear relationships cannot be excluded, 0.8≤|r|≤1 is regarded as high correlation, 0.5≤|r|<0.8 is regarded as moderate correlation, 0.3≤|r|<0.5 is regarded as low correlation, and 0≤|r|<0.3 indicates that the degree between the two variables is weak and can be regarded as unrelated.

[0049] The correlation coefficients of Hg, Cu, Co and Cd are 0.67, 0.73, 0.75 respectively, which are significantly positively correlated, indicating that the four heavy metals have high cooperative variation characteristics in the water body, and it can be inferred that they have strong homology; the correlation coefficients of Pb, Zn, Co, Cr and As are 0.67, 0.77, 0.67, 0.84 respectively, among which Pb, Zn, Cr and As also have significant positive correlation between each other, and the four heavy metals also have high homology; the correlation coefficients of Zn, Pb, As, Cd, Cr and Mn are 0.67, 0.63, 0.49, 0.64, 0.77 respectively, which also show extensive significant positive correlation, indicating that Zn has great relevance with Pb, As, Cd, C and Mn, and the above heavy metals may be affected by the same factor.

[0050] (2) Principal component analysis (PCA): as shown in Figure 7 , in order to further identify the potential source structure of heavy metal pollution in the water body of Geyang Lake, the present application adopts SPSS software to perform principal component analysis on the concentration data of 10 kinds of heavy metal elements, and the calculation method is as follows: 1) Solve the characteristic equation | λE-R|=0, and arrange the characteristic values according to their size; 2) Solve the characteristic vector a i =[a i1 , a i2 ,⋯,a ip ], i=1,2,…,p, and ; 3) Calculate the contribution rate and cumulative contribution rate, , , wherein R is the correlation coefficient matrix, λ is the characteristic value matrix, E is the unit matrix, λ i and λ k are elements in the characteristic value matrix λ, p is a positive integer, and a ij is an element in the characteristic vector a i corresponding to the characteristic value.

[0051] Three principal components were finally extracted, and the total variance contribution rate was 93.28%, which indicated that the extracted principal components could effectively reflect the main information of the data. According to the load coefficient, the load degree can be divided into four levels: the load coefficient is greater than 0.75, which is in high load; the load coefficient is between 0.5 and 0.75, which is in medium load; the load coefficient is between 0.3 and 0.5, which is in low load; and the load coefficient is less than 0.3, which is in extremely low load. The principal component analysis results are combined with the industry characteristic pollutants, the spatial distribution of water body heavy metal concentration and the correlation analysis results to qualitatively analyze the source of heavy metals.

[0052] (3) as shown in Figure 8 Positive matrix factorization model (PMF): in order to accurately identify the source of heavy metal pollution in the lake, PMF is used to quantitatively analyze the contribution rate of pollution source to heavy metals, and the model specifically includes: , Wherein, X ij is the concentration of the jth element in the ith sample, g ik is the contribution of the kth source to the ith sample, f kj is the concentration of the jth element in the kth source, and e ij is the error value of the concentration of the jth element in the ith sample.

[0053] The original matrix is decomposed by multiple calculations, so that the objective function Q reaches the minimum value, , Wherein, μ ij is the uncertainty, and the calculation method of the uncertainty is shown in the following two formulas:

[0054] , Wherein: MDL is the method detection limit of the pollutant; P is the relative standard deviation of the concentration of the pollutant in percentage unit.

[0055] The present application utilizes EPA PMF5.0 software, calculates the signal-to-noise ratio (S / N) according to the concentrations of Hg, Pb, Cu, Zn, Co, As, Cd, Cr, Mn and Ni in the sediment and the corresponding uncertainty, and obtains that the signal-to-noise ratio of each element is greater than 1, which is defined as "Strong". After the model is run, the regression fitting effect of each heavy metal is good, the R² value is between 0.94 and 0.99, the fitting slope is close to 1, and the intercept is close to 0, indicating that the model fitting precision is high.

[0056] The model extracts 5 pollution source factors, and combined with industry-specific pollutants, the main pollution source industry can be determined. Among them, factor 1 contributes to Cr (66.1%) and Pb (36.1%), and factor 1 is identified as the printing and dyeing textile industry; factor 2 contributes to Hg (25.1%), Ni (31.6%) and Cu (24.8%), and factor 2 is identified as the chemical industry; factor 3 contributes to Mn (36.8%) and Cu (29.4%), and factor 3 is identified as the metal manufacturing industry; factor 4 contributes to Cu (40.0%), Co (37.6%), Cd (33.1%) and Hg (30.2%), and factor 4 is identified as the electroplating industry; factor 5 contributes to Zn (38.0%), As (31.9%) and Ni (34.6%). Factor 5 is identified as agriculture. The PMF model can realize the quantitative identification of the source of water body heavy metal pollution, the structure of each factor is clear, and the fitting effect is ideal, which provides data support and method verification for the pollution tracing strategy proposed in the application.

[0057] Step six: as shown in Figure 9 The total contribution rate of various pollution sources is analyzed According to the PMF model decomposition result, the source structure of the heavy metals in the water body of the lake and the relative contribution rate of each type of pollution source to the total concentration are obtained. The specific order is as follows: electroplating industry (28.31%) > printing and dyeing textile industry (22.37%) > agriculture (18.46%) > metal manufacturing industry (15.80%) > chemical industry (15.07%). The above results show that the heavy metal pollution sources of the lake water body are mainly industrial pollution, and agricultural activities are also an important contribution factor. The contribution proportion of each type of pollution source is clear, and the pollution cause structure is clear, which verifies the applicability and accuracy of the multi-source analysis method based on the PMF model in the heavy metal pollution tracing of the water body, and can provide a basis for subsequent pollution prevention and control and management strategies.

[0058] In summary, the application not only combines the double-period sampling design of dry season and wet season, which can comprehensively reflect the temporal and spatial distribution characteristics of heavy metals under different hydrological conditions, but also makes up for the limitations of single index evaluation method, has strong adaptability, and is suitable for preliminary diagnosis of heavy metal risk in various water bodies.

[0059] The above specific embodiments do not constitute a limitation on the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for assessing the risk of heavy metal pollution in lake water based on a multi-source analysis model, characterized in that, The method includes: S1. Sample collection: During the dry season and the wet season, water samples were collected from the lake using a water sampler. The samples were then filtered through a microporous membrane to remove suspended solids, and the water samples to be tested were obtained. S2. Sample detection: After pretreatment of the water sample to be tested, the concentrations of Pb, Cu, Zn, Co, As, Cd, Cr, Mn and Ni in the water sample were determined by ICP-MS, and the concentration of Hg in the sample was determined by atomic fluorescence spectrometry. S3. Spatial Distribution Characteristics of Heavy Metals: Analysis of Spatial Distribution Characteristics of Heavy Metals Based on Inverse Distance Weighted Interpolation in Arc-GIS; S4. Assessment of Heavy Metal Pollution Level: The pollution level of each heavy metal is identified by the single-factor index method, and the synergistic pollution effect of multiple heavy metals is analyzed by the Nemerow comprehensive pollution index method to assess the pollution risk level. S5. Heavy Metal Source Tracing Analysis: S51. Correlation analysis and principal component analysis: The Spearman correlation coefficient matrix is ​​used to perform correlation analysis to identify heavy metal clusters with common source associations, and principal component analysis is used to identify the characteristics of the dominant pollution source categories. S52. Positive Deterministic Matrix Factorization Model: Using PMF to quantitatively analyze the contribution rate of pollution sources to heavy metals.

2. The method for assessing the risk of heavy metal pollution in lake water based on a multi-source analysis model according to claim 1, characterized in that, In step S1, the sampling points of the water sampler that collect lake water samples are evenly distributed according to spatial gradient, covering the nearshore area and the central area of ​​the lake.

3. The method for assessing the risk of heavy metal pollution in lake water based on a multi-source analysis model according to claim 1, characterized in that, In step S2, the pretreatment of the water sample to be tested includes: using a disposable syringe in the laboratory, passing the water sample brought back to the laboratory through a 0.4-0.5 μm microporous membrane for membrane treatment to separate solids and remove suspended solids. Then, 4-6 ml of the membrane-treated water sample is placed in a microwave digestion tube and digested using a CEM-Mars6 microwave digester. The heating time is set to 9-11 min, the digestion temperature to 165-175℃, and the digestion power to 900 kW. -1100W, holding time 15-25min, add 2-4ml of analytical grade nitric acid before digestion. After digestion, transfer the liquid in the digestion tube to the acid removal tube, add perchloric acid to remove the acid. When only a few drops of liquid remain and white fumes are emitted, the acid is considered to have been completely removed. Set the temperature to 75-85℃ and remove the acid for 23-25h. After the acid removal is complete, pass the sample through a membrane again and dilute it to 95-105ml with 1-3% nitric acid. Then, put it into the sample tube, affix the corresponding label, and prepare for measurement.

4. The method for assessing the risk of heavy metal pollution in lake water based on a multi-source analysis model according to claim 1, characterized in that, In step S2, the determination of heavy metal content includes: before measurement, the water sample to be tested needs to be diluted with 1-3% nitric acid solution to prepare standard water samples with concentrations of 0 μg / L, 10 μg / L, 50 μg / L, 100 μg / L, 200 μg / L, 300 μg / L, 400 μg / L, and 500 μg / L respectively. At the same time, In and Ge are used as internal standards. A standard curve is plotted with the concentration of the standard water sample as the abscissa and the ratio of the sample signal to the internal standard signal as the ordinate. Then, the water sample solution is taken for instrumental measurement to obtain the concentration of heavy metals in the river water. When measuring heavy metals, blank samples and parallel samples are used for quality control.

5. The method for assessing the risk of heavy metal pollution in lake water based on a multi-source analysis model according to claim 1, characterized in that, In step S3, the calculation function for analyzing the spatial distribution characteristics of heavy metals based on inverse distance weighted interpolation in Arc-Gis is ， Among them, Z so Z is an estimated value of the heavy metal concentration of SO at the undetected point. si W is the monitoring value of the i-th point surrounding so, where n is the number of monitoring points in the local area. i It refers to the weight coefficient pointing from point si to point so.

6. The method for assessing the risk of heavy metal pollution in lake water based on a multi-source analysis model according to claim 1, characterized in that, In step S4, the calculation function for identifying the pollution degree of each heavy metal by the single factor index method is ， Among them, P i C is a single-factor index of heavy metal i. i C represents the concentration value detected for heavy metal i. o Here is the standard limit for heavy metal i; The criteria for classifying the degree of water pollution are: the single-factor index P of heavy metal i. i When P < 1, the water quality level is unpolluted. i When the value is 1-2, the water quality level is low pollution, P i When the concentration is 2-3, the water quality level is moderately polluted, P i When the value is greater than 3, the water quality level is considered severely polluted.

7. The method for assessing the risk of heavy metal pollution in lake water based on a multi-source analysis model according to claim 1, characterized in that, In step S4, the calculation function for analyzing the combined pollution effect of multiple heavy metals and evaluating the pollution risk level by the Nemerow comprehensive pollution index method is ， Where P is the comprehensive pollution assessment index, P max P is the highest pollution index among water quality parameters. i For a single pollution index; The classification standard for evaluating the pollution risk level is as follows: when the comprehensive pollution evaluation index P < 1, the water quality level is pollution-free; when P is 1 - 2, the water quality level is moderately polluted; when P is 2 - 3, the water quality level is polluted; when P is 3 - 5, the water quality level is heavily polluted; when P > 5, the water quality level is severely polluted.

8. The method for assessing the risk of heavy metal pollution in lake water based on a multi-source analysis model according to claim 1, characterized in that, In step S51, the correlation analysis is used to analyze the strength of the correlation between various heavy metals and identify homologous heavy metals. The correlation coefficient r reflects the strength of the correlation between variables in numerical form. Generally, r is between -1 and 1. If 0 < r < 1, it represents a positive correlation between two variables. If -1 < r < 0, it represents a negative correlation between two variables. |r| = 1 means that the value of one variable completely depends on the other variable. r = 0 means that there is no linear correlation between the two variables, but it does not exclude the possibility of other non-linear relationships. When 0.8 ≤ |r| ≤ 1, it is regarded as highly correlated. When 0.5 ≤ |r| < 0.8, it is regarded as moderately correlated. When 0.3 ≤ |r| < 0.5, it is regarded as lowly correlated. When 0 ≤ |r| < 0.3, it indicates that the degree between the two variables is weak and can be regarded as uncorrelated. The calculation steps of r are as follows: 1) Arrange the original data in rows to form a matrix X; 2) Standardize the data of X so that its mean becomes zero and the variance becomes 1; 3) Calculate the correlation coefficient matrix R; ， Where, r ij (i,j = 1,2…p) represents the original variable x. i With x j Correlation coefficient, r ij =r ji Its calculation function is as follows: ， Where, x ki Let x be the element in the k-th row and i-th column of matrix X. kj Let X be the element in the k-th row and j-th column. For the original variable x i Average rank, For the original variable x j The average rank.

9. The method for assessing the risk of heavy metal pollution in lake water based on a multi-source analysis model according to claim 8, characterized in that, In step S51, the principal component analysis for identifying the characteristics of the dominant pollution source category includes: 1) Solve the characteristic equation |λE - R| = 0 to find the eigenvalues and arrange them in descending order; 2) Find the eigenvector a corresponding to the eigenvalue. i =[a i1 , a i2 ,⋯,a ip ], i=1,2,…,p, and ； 3) Calculate the contribution rate and cumulative contribution rate, ， ， Where R is the correlation coefficient matrix, λ is the eigenvalue matrix, E is the identity matrix, and λ i and λ k Let p be an element in the eigenvalue matrix λ, where p is a positive integer and a ij The eigenvector a corresponding to the eigenvalue i The elements in.

10. The method for assessing the risk of heavy metal pollution in lake water based on a multi-source analysis model according to claim 1, characterized in that, In step S52, the function model for quantitatively analyzing the contribution rate of pollution sources to heavy metals by PMF is ， Among them, X ij Let g be the concentration of the j-th element in the i-th sample. ik f represents the contribution of the k-th source to the i-th sample. kj Let e ​​be the concentration of the j-th element in the k-th source. ij This represents the error value of the concentration of the j-th element in the i-th sample.

Images