Plain small watershed nitrogen and phosphorus pollutant collaborative source analysis method

Abstract

The invention discloses a plain small watershed nitrogen and phosphorus pollutant collaborative source analysis method, which belongs to the technical field of environmental pollution traceability, and comprises the following steps: 1, collecting multi-source basic data in a target plain small watershed, carrying out unified coordinate system, format and precision preprocessing, and constructing a watershed basic database; and 2, constructing a response relation model of various pollution sources and potential pollution loads on a GIS platform based on a drainage basin basic database, qualitatively identifying main pollution sources through spatial visualization analysis, and delineating key risk areas. According to the method, macroscopic spatial analysis, microchemical tracing and mathematical model calculation are organically fused, high-precision spatial data, hydrochemical and isotope analysis, multi-model quantitative analysis and uncertainty quantification are integrated, and high-precision and high-reliability cooperative source analysis of nitrogen and phosphorus pollutants is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of environmental pollution source tracing, and particularly relates to a nitrogen and phosphorus pollutant source analysis method for a small plain watershed. BACKGROUND

[0002] In small watersheds in plain areas, nitrogen and phosphorus are the main pollutants leading to water body eutrophication. With the increase of agricultural fertilization intensity, the expansion of livestock and poultry breeding scale, and the continuous discharge of domestic sewage and industrial wastewater, a large amount of nitrogen and phosphorus enters the surface water body, making the risk of water body algae outbreak and ecosystem degradation increasing. The sources of nitrogen and phosphorus pollution are complex, including not only non-point source pollution but also point source pollution, which brings great challenges to water environment governance. Existing pollution source analysis research mainly relies on single model method. For example, positive matrix factorization (PMF) can reveal the contribution mode of pollution sources through statistical decomposition, absolute principal component score and multiple linear regression (APCS-MLR) can be combined with factor score for quantitative analysis, and isotope tracing can identify pollution sources through stable isotope characteristics. These methods can provide certain source tracing information under certain conditions, but each has certain limitations.

[0003] In practical application, a single model often shows insufficient applicability. Different models have strong dependence on the type, quantity and quality of input data, and when the monitoring data is insufficient or the environmental conditions are complex, the uncertainty of the analysis result is likely to increase significantly. At the same time, different methods are based on different theoretical frameworks, and the results obtained are often different, making it difficult to form a unified and comprehensive conclusion, and thus there is a certain one-sidedness in explaining the pollution sources. In addition, the sources of nitrogen and phosphorus pollution often involve agricultural non-point source, domestic sewage and industrial point source and other types. Traditional single analysis method has insufficient discrimination between these pollution sources, and it is often difficult to accurately identify the contribution proportion of each type of pollution source. This deficiency not only limits the reliability of the research conclusion, but also weakens the application value of the result in the precise governance and policy making of water environment.

[0004] In summary, the existing technology still has significant deficiencies in nitrogen and phosphorus pollution source tracing. It is urgent to propose a comprehensive method that can integrate multi-source data and various analysis methods and can quantitatively characterize uncertainty, so as to improve the accuracy and stability of nitrogen and phosphorus pollution source analysis, and thus provide a more reliable scientific basis for the eutrophication governance of small plain watershed.

[0005] Chinese patent document with publication number CN 110390494 A and publication date of October 29, 2019 discloses a method for tracing "three nitrogen" in groundwater of agricultural area domestic waste landfill, which comprises the following steps: The first stage is field investigation and data collection, mainly to understand the environmental characteristics of the region; The second stage is point distribution and sampling monitoring, scientific point distribution is carried out based on the field investigation result, and sampling monitoring is carried out on the groundwater, soil and typical pollution sources in the region respectively; The third stage is data arrangement and analysis, according to the complexity of the 'three nitrogen' pollution of the regional groundwater and the tracing requirement, the pollution source of the 'three nitrogen' of the groundwater is analyzed by using water quality space-time analysis, water quality chemical analysis and nitrate nitrogen oxygen stable double isotope tracing method in turn.

[0006] Based on this, the present application designs a nitrogen and phosphorus pollutant collaborative source analysis method for small plain watersheds to solve the above problems. SUMMARY

[0007] In view of the above-mentioned shortcomings of the prior art, the present application provides a nitrogen and phosphorus pollutant collaborative source analysis method for small plain watersheds.

[0008] To achieve the above object, the present application is realized by the following technical scheme: A nitrogen and phosphorus pollutant collaborative source analysis method for small plain watersheds, comprising the following steps: Step one: collect multi-source basic data in the target small plain watershed, and carry out unified coordinate system, format and accuracy pretreatment to construct a watershed basic database; Step two: build response relationship models of various pollution sources and potential pollution load based on the watershed basic database on the GIS platform, and qualitatively identify the main pollution sources and delineate the key risk areas by spatial visualization analysis; Step three: in the delineated key risk area, water sample collection points are arranged according to the water system trend and pollution source distribution requirements, surface water and groundwater samples are collected, and the water samples are analyzed to obtain analysis data; Step four: determine the analysis model based on the sampling data conditions, watershed characteristics and project constraints, and the pollution source contribution rate is determined by the analysis model through the analysis data; Step five: the nitrogen and phosphorus pollution source type and spatial position are obtained based on the key risk area and the pollution source contribution rate, the source affecting the concentration of nitrogen and phosphorus pollutants is determined through the nitrogen and phosphorus pollution source type and spatial position, and the control strategy is formulated according to the source affecting the concentration of nitrogen and phosphorus pollutants.

[0009] Further, the multi-source basic data includes (1) high-resolution land use / land cover data, digital elevation model, soil type distribution map, etc.; (2) high-precision population spatial distribution data, agricultural planting structure and fertilizer application amount statistics, location and pollution discharge amount of large-scale livestock and poultry breeding farms, location and pollution discharge permit information of industrial enterprise point source discharge outlets, and urban sewage treatment plant information, etc.; (3) watershed river system map, flow and water level data of hydrological monitoring stations, and historical water quality monitoring data.

[0010] Further, the spatial visualization analysis qualitatively identifies the main pollution sources and delineates the key risk areas as follows: Step 1: Superimpose the agricultural land use data such as farmland and garden land with the unit area fertilization intensity data to identify the areas where the fertilization intensity is higher than the agricultural non-point source pollution risk threshold; Step 2: Superimpose the population density distribution with the sewage treatment facility coverage rate to identify the areas where the domestic sewage is directly discharged or the treatment capacity does not meet the standard; Step 3: Locate the industrial discharge outlets and identify the industrial discharge outlets with discharge amount greater than the set value; Step 4: The areas where the fertilization intensity is higher than the agricultural non-point source pollution risk threshold, the areas where the domestic sewage is directly discharged or the treatment capacity does not meet the standard, and the industrial discharge outlets with discharge amount greater than the set value are delineated as the key risk areas.

[0011] Further, the analysis of water samples specifically includes: (1) Determining routine water quality indicators, including but not limited to total nitrogen, ammonia nitrogen, nitrate nitrogen, total phosphorus, chemical oxygen demand, pH, and conductivity.

[0012] (2) Focusing on measuring isotope indicators related to nitrogen pollution sources, such as δ 15 N-NO3 – , δ 18 O-NO3 – , which are used to assist in identifying the source of nitrogen pollution.

[0013] Further, the data conditions include sample quantity, data completeness, and existing historical database.

[0014] Further, the specific steps of step four are as follows: Step 41: Input the data conditions, watershed characteristics, and project constraints into the AHP model that has been constructed, and score them according to 1-5 points; Step 42: Calculate the maximum weight of data conditions, watershed characteristics, and project constraints, the score of data conditions, watershed characteristics, and project constraints, and the weighted total score of each scheme through the maximum weight and score; Step 43: Determine the analysis model by the highest weighted score; Step 44: Substitute the analysis data obtained in step three into the analysis model to calculate the pollution source contribution rate.

[0015] Further, the AHP model is as follows: Model 1: PMF; Model 2: APCS-MLR; Model 3: PMF + APCS-MLR; Model 4: PMF + isotope tracing; Model 5: APCS-MLR+isotope tracing; Model 6: PMF+APCS-MLR+isotope tracing; Model 7: PMF+Monte Carlo; Model 8: APCS-MLR+Monte Carlo; Model 9: PMF+APCS-MLR+Monte Carlo; Model 10: PMF+isotope tracing+Monte Carlo; Model 11: APCS-MLR+isotope tracing+Monte Carlo; Model 12: PMF+APCS-MLR+isotope tracing+Monte Carlo.

[0016] Further, when PMF is used in the analysis model of step 43, the specific operation of step 44 is as follows: Step 44: input the analysis data obtained in step three into the positive matrix factorization model to calculate the contribution rate of each type of pollution source to the nitrogen and phosphorus concentrations in the water body, i.e., the pollution source contribution rate; Alternatively, when APCS-MLR is used in the analysis model of step 43, the specific operation of step 44 is as follows: Step 44: input the analysis data obtained in step three into the absolute principal component score-multivariate linear regression model to calculate the contribution rate of each type of pollution source to the nitrogen and phosphorus concentrations in the water body, i.e., the pollution source contribution rate; Alternatively, when isotope tracing is used in the analysis model of step 43, the specific operation of step 44 is as follows: Step 44: based on the δ 15 N and δ 18 O isotope data, the Bayesian mixture model is used to quantitatively estimate the contribution proportion of different nitrogen sources, i.e., the pollution source contribution rate; Alternatively, when Monte Carlo is used in the analysis model of step 43, the specific operation of step 44 is as follows: Step 44: random disturbance is introduced to the input parameters and running process of PMF, APCS-MLR and the Bayesian mixture model by using the Monte Carlo simulation method, multiple simulation calculations are performed, and the uncertainty range of each model output result, i.e., the pollution source contribution rate, is quantified.

[0017] In order to better achieve the purpose of the present application, the present application also provides Compared with the prior art, the present application has the beneficial effects that: 1. macroscopic spatial analysis, microscopic chemical tracing and mathematical model calculation are organically integrated, high-precision spatial data, water chemistry and isotope analysis, multi-model quantitative analysis and uncertainty quantification are integrated, a method for high-precision and high-reliability collaborative source analysis of nitrogen and phosphorus pollutants is realized, the limitations of single technology are avoided, and the best source analysis combination can be selected according to the actual situation; 2. By multi-model parallel analysis and Monte Carlo simulation uncertainty quantification, the uncertainty of a single model is effectively reduced, making the final pollution source contribution rate estimation result more stable and reliable; 3. The hydrological and pollution characteristics of the plain small watershed are fully considered, and the applicability and analysis accuracy are improved. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0019] Figure 1 A flowchart of a plain small watershed nitrogen and phosphorus pollutant co-source analysis method of the present application; Figure 2 A PMF+APCS-MLR source contribution rate diagram of the present application; Figure 3 A PMF model source contribution rate diagram of the present application; Figure 4 A Bayesian mixture model source contribution rate diagram of the present application. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical scheme and advantages of the embodiments of the present application more clear, the technical scheme in the embodiments of the present application will be clearly and completely described below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0021] Embodiment one: in some embodiments, referring to the drawings of the specification Figures 1-4 A plain small watershed nitrogen and phosphorus pollutant co-source analysis method, comprising the following steps: Step one: collect multi-source basic data in the target plain small watershed, and perform unified coordinate system, format and accuracy preprocessing to construct a watershed basic database; Before unified coordinate system, format and accuracy preprocessing, the missing values of the multi-source basic data are filled by linear interpolation or mean substitution method, and then the filled data are verified for spatial consistency, temporal consistency and logical consistency; The statistical indicators of data from different sources may conflict (e.g., "agricultural planting structure" is counted by administrative units, while "land use data" is counted by grid units), and the method does not mention how to achieve "administrative unit-grid unit" data matching, which may cause spatial misplacement in step two "layer overlay analysis" (e.g., farmland and fertilizer intensity overlay), affecting the accuracy of risk area identification. Spatial consistency check: For spatial and statistical data, use the "spatial interpolation + unit matching" method - such as converting the fertilizer amount of administrative units to grid unit fertilizer intensity through "farmland area proportion weighted interpolation", and then overlaying with the farmland grid of the land use map to ensure spatial unit uniformity.

[0022] Temporal consistency check: For statistical and monitoring data, use the "time scale matching" method - such as decomposing annual fertilizer amount into "quarterly fertilizer amount" (combined with planting cycles such as spring planting and autumn harvest season fertilizer peak), and then associating with water quality monitoring data of the same period to ensure temporal consistency.

[0023] Logical consistency check: For the same type of data, use the "cross-validation" method - such as industrial discharge outlet data and downstream water quality data, calculate the theoretical concentration through the "discharge amount-concentration conversion formula" (concentration = discharge amount / (flow rate × time)), and compare it with the actual monitoring concentration. If the deviation exceeds the threshold (such as ±15%), recheck the data to ensure logical consistency.

[0024] Unified coordinate system: Convert the coordinate systems of all multi-source basic data with spatial information (such as land use map, DEM, discharge outlet location, hydrological station, etc.) to the same type of coordinate system, such as the national geodetic coordinate system (such as CGCS2000).

[0025] In ArcGIS software, use "projection transformation" or "define projection" tool for accurate conversion to ensure that the spatial references of all layers are completely consistent. Unified format: Convert the collected multi-source basic data formats (such as Shapefile, GeoTIFF, Excel table, text file, etc.) to a unified, compatible and easy-to-operate data format on the ArcGIS platform.

[0026] Unified precision: Align the boundaries of all multi-source basic data to ensure consistent spatial resolution and geographic registration accuracy.

[0027] Constructing the basin basic database: Import and save the pre-processed data on the ArcGIS platform; Step two: Based on the basin basic database, build response relationship models of various pollution sources and potential pollution loads on the GIS platform, and visually analyze and qualitatively identify the main pollution sources and delineate the key risk areas; The response relationship model uses GIS spatial overlay and analysis tools to correlate the spatial distribution of pollution sources with their emission intensity or probability of occurrence.

[0028] The association is achieved by constructing an expert knowledge-driven judgment rule base: for example, the intersection of "arable land or orchard land use type and high fertilization intensity area" is judged as a high-risk area for agricultural non-point source pollution; the intersection of "high population density and insufficient sewage treatment capacity area" is judged as a key area for domestic pollution; and the intersection of "industrial discharge outlets and high discharge permit level" is judged as a key industrial point source impact area. In addition, in the GIS platform, key risk areas can be delineated and output by overlaying and analyzing multi-source spatial data such as land use, fertilization intensity, population distribution, sewage treatment facilities, and industrial discharge outlets.

[0029] Step 3: Within the designated key risk areas, water sampling points are set up according to the water system direction and pollution source distribution requirements to collect surface water and groundwater samples. The water samples are then analyzed to obtain analytical data. Step 4: Determine the analysis model based on the sampling data conditions, watershed characteristics, and project constraints. The analysis model determines the contribution rate of pollution sources by analyzing the data. Step 5: Based on key risk areas and pollution source contribution rates, obtain the types and spatial locations of nitrogen and phosphorus pollution sources, identify the sources affecting nitrogen and phosphorus pollutant concentrations, and formulate control strategies based on the sources affecting nitrogen and phosphorus pollutant concentrations.

[0030] For example, Figure 3 The factor contributing 23.9% to NO2 – Cl – and Na + The higher proportion of sources explained by NO2 indicates that this factor is the main source of these components, as previous studies have revealed. – Generally originating from sewage discharge, agricultural fertilizers, etc., Cl – and Na + Generally originating from the dissolution of rock salt in nature or from domestic sewage discharge, Figure 3 The study area contains no rock salt minerals and has dense urban areas. NO2 – Cl – and Na + High concentrations of NO2 are mainly concentrated in these areas. – Cl – and Na + This can be interpreted as wastewater discharge. Spatial location can be interpolated on a GIS platform based on the contribution of this factor to each sampling point to obtain a more accurate location of the pollution source.

[0031] In nitrogen and oxygen isotope source tracing, the pollution source has been identified, and the spatial location of the pollution source is the area involved in which it makes a high contribution to certain sampling points, for example...Figure 4 The high contribution rate of sewage discharge to the pollution source indicates that the source of sewage discharge is nearby, and attention should be paid to these areas.

[0032] The control strategy is developed according to the source affecting the concentration of nitrogen and phosphorus pollutants; For example: the main source of nitrogen and phosphorus pollutants is sewage discharge, so the control strategy is mainly to limit sewage discharge, such as increasing the punishment for unreasonable sewage discharge; publicizing the harm of sewage discharge; upgrading wastewater treatment facilities, more strictly implementing emission standards, and implementing decentralized treatment systems in densely populated or suburban areas.

[0033] If the main source is agricultural fertilizer, the control strategy is to optimize the amount and time of fertilizer application, use slow-release fertilizer, and use vegetation buffer strips to intercept the loss of nutrients.

[0034] The multi-source basic data includes: (1) high-resolution land use / land cover data, digital elevation model, soil type distribution map, etc. (2) high-precision population spatial distribution data, agricultural planting structure and fertilizer application amount statistics, location and pollution discharge of large-scale livestock and poultry farms, location and pollution discharge permit information of industrial enterprises, and information of urban sewage treatment plants; (3) river system map, flow and water level data of hydrological monitoring stations, and historical water quality monitoring data (such as TN, TP, NO3–-N, etc.).

[0035] The spatial visualization analysis qualitatively identifies the main pollution sources and delineates the key risk areas as follows: Step 1: Overlay the agricultural land use data such as farmland and garden land with the unit area fertilizer intensity data to identify areas with fertilizer intensity higher than the agricultural non-point source pollution risk threshold; Step 2: Overlay the population density distribution with the sewage treatment facility coverage rate to identify areas where domestic sewage is directly discharged or the treatment capacity does not meet the standard; Step 3: Locate industrial pollution outlets and identify industrial pollution outlets with discharge greater than the set value; Step 4: The areas with fertilizer intensity higher than the agricultural non-point source pollution risk threshold, the areas where domestic sewage is directly discharged or the treatment capacity does not meet the standard, and the industrial pollution outlets with discharge greater than the set value are delineated as key risk areas.

[0036] The analysis of water samples includes: (1) Determination of routine water quality indicators, including but not limited to total nitrogen, ammonia nitrogen, nitrate nitrogen, total phosphorus, chemical oxygen demand, pH, conductivity, etc.

[0037] (2) Key determination of isotope indicators related to nitrogen pollution sources, such as δ 15N-NO3 – δ 18 O-NO3 – , for the auxiliary identification of nitrogen pollution sources (such as chemical fertilizers, domestic sewage, livestock manure).

[0038] Data conditions include sample size, data completeness, and existing historical database; Sample size is the key to selecting statistical models. Small samples (≤20) are difficult to meet the stable requirements of models such as PMF on sample size, and are more suitable for APCS-MLR or need to rely on specific isotope data for constraint; medium samples (20-50) provide basic operating conditions for most models; large samples (>50) can fully exert the advantages of models such as PMF in identifying subtle sources.

[0039] Data completeness: sample size must be complete; If there are a large number of missing or abnormal data, models that are not sensitive to data quality problems or have built-in robust processing mechanisms (such as PMF can integrate data uncertainty) should be preferred, rather than models that are demanding on data quality.

[0040] Existing historical database is long-term water quality data and local pollution source isotope "fingerprint" library: Basin characteristics include pollution source complexity and hydrogeochemical process activity; Pollution source complexity is used to measure the degree of interweaving of pollution sources, to ensure that the selected method has sufficient resolution to distinguish all important pollution sources.

[0041] Pollution source complexity: For basins with single or clear source, simple AHP models can be used; for basins with complex sources (such as agricultural, domestic, and industrial sources coexisting), AHP models with "blind source analysis" ability, such as PMF, which can identify more factors, are needed.

[0042] Hydrogeochemical process activity (such as denitrification intensity) is a key adaptive factor for small plain basins with slow flow and easy biochemical processes.

[0043] Strong denitrification and other processes can significantly change nitrogen isotope ratios (fractionation effect), making it impossible to use fixed "fingerprints" for source identification. Hydrogeochemical process activity directly determines whether it is necessary to use isotope tracing (especially δ 15 N-δ 18 O double marker method) to identify and correct this impact and avoid drawing incorrect conclusions Project constraints include budget cost, tolerance for result uncertainty, and project time requirements; Budget cost is a constraint factor.

[0044] Isotopic analysis is costly, while conventional water chemistry analysis is relatively cheap. This budget cost plays a role of “veto” or “strong recommendation” in the scheme selection, ensuring that the recommended technical path is economically feasible within the project budget.

[0045] The tolerance of result uncertainty distinguishes whether the research result is used for preliminary screening or precise decision-making.

[0046] For high-risk management decisions (such as delineating ecological compensation areas), a very low result uncertainty is required, and multi-model integration and Monte Carlo simulation must be used to quantify and reduce uncertainty; for trend analysis, the requirement can be appropriately relaxed. This factor ensures that the accuracy of the analysis result matches the strictness of the management requirement.

[0047] Project time requirements affect technical selection.

[0048] Isotopic testing usually has a long cycle, while conventional water chemistry model calculation is faster. This factor ensures that, under a tight timeline, a solution that can deliver preliminary results quickly can be prioritized, or when time is available, a more detailed but time-consuming technical combination can be selected.

[0049] Step four has the following specific steps: Step 41: Input data conditions, basin characteristics, and project constraints into the AHP model that has been built, and score them according to 1-5 points. The scoring standard is based on the dependency and utilization ability of the AHP model on the influencing factors; for example, taking the sample size as an influencing factor, assuming there are 15 data, APCS-MLR is based on traditional multivariate statistics, and has a relatively high tolerance for small samples. However, PMF requires a relatively large number of samples, so a small sample of 15 data is not suitable for implementing PMF. When scoring specifically, APCS-MLR can be given a score of 5 points under small sample conditions, while PMF can be given a score of 1 point; similarly, a combination of Model 3 can be given a score of 3 points. For example, 35 samples can be considered as a medium sample size, so APCS-MLR can be scored as 5 points, PMF can be scored as 3 points, and the combination of the two can be scored as 4 points. If there are 80 samples, they can be considered as a large sample, so APCS-MLR can be scored as 5 points, PMF can be scored as 5 points, and the combination of the two can be scored as 5 points.

[0050] Step 42: Calculate the maximum weight of data conditions, basin characteristics, and project constraints, the score of data conditions, basin characteristics, and project constraints, and the weighted total score of each scheme by multiplying the maximum weight and the score. The specific operation of step 42 is as follows: Step 421: Calculate the criterion layer judgment matrix, the criterion layer data condition judgment matrix, the criterion layer basin feature judgment matrix and the criterion layer project constraint judgment matrix; Step 422: Calculate the maximum weight by the criterion layer judgment matrix, the criterion layer data condition judgment matrix, the criterion layer basin feature judgment matrix and the criterion layer project constraint judgment matrix; Step 423: Calculate the weighted total score of each scheme by the maximum weight and the score; Each criterion layer judgment matrix is as follows:

[0051] The criterion layer data condition judgment matrix is as follows:

[0052] The criterion layer basin feature judgment matrix is as follows:

[0053] The criterion layer project constraint judgment matrix is as follows:

[0054] The maximum weight result based on AHP is as follows:

[0055] Step 43: Determine the analysis model by the weighted maximum score; Step 44: Put the analysis data obtained in step three into the analysis model to calculate the pollution source contribution rate.

[0056] The AHP model is as follows: Model 1: PMF; Model 2: APCS-MLR; Model 3: PMF+APCS-MLR; Model 4: PMF+isotope tracing; Model 5: APCS-MLR+isotope tracing; Model 6: PMF+APCS-MLR+isotope tracing.

[0057] Model 7: PMF+Monte Carlo.

[0058] Model 8: APCS-MLR+Monte Carlo.

[0059] Model 9: PMF+APCS-MLR+Monte Carlo.

[0060] Model 10: PMF+isotope tracing+Monte Carlo.

[0061] Model 11: APCS-MLR+isotope tracing+Monte Carlo.

[0062] Model 12: PMF+APCS-MLR+Isotope Tracing+Monte Carlo.

[0063] If the analysis model in step 43 contains a PMF, the specific operation in step 44 is as follows: Step 44: Input the analytical data obtained in Step 3 into the positive definite matrix factor decomposition model to calculate the contribution rate of various pollution sources to the nitrogen and phosphorus concentrations in the water body, i.e., the pollution source contribution rate. Alternatively, if the analysis model in step 43 includes APCS-MLR, the specific steps in step 44 are as follows: Step 44: Input the analysis data obtained in Step 3 into the absolute principal component score-multiple linear regression model to calculate the contribution rate of various pollution sources to the nitrogen and phosphorus concentration in the water body, i.e., the pollution source contribution rate. Alternatively, if the analysis model in step 43 includes isotope tracing, step 44 will proceed as follows: Step 44: Based on δ 15 N and δ 18 Using O isotope data and a Bayesian mixture model, the contribution ratio of different nitrogen sources, i.e., the pollution source contribution rate, is quantitatively estimated. Alternatively, if the analysis model in step 43 includes Monte Carlo methods, then step 44 will proceed as follows: Step 44: Using the Monte Carlo simulation method, random perturbations are introduced into the input parameters and operation process of the PMF, APCS-MLR and Bayesian mixture models. Multiple simulation calculations are performed to quantify the uncertainty range of the output results of each model, i.e. the pollution source contribution rate.

[0064] This invention generates a pollution source list containing the types, spatial locations, and contribution ratios of nitrogen and phosphorus pollution sources by identifying key risk areas and their contribution rates. The sources influencing nitrogen and phosphorus pollutant concentrations are then determined based on this list. Since this invention primarily traces the sources of nitrogen and phosphorus, when isotope tracing is included, the contribution rate is mainly based on isotope tracing results, while receptor model results primarily reflect the sources of other components. If the method does not include isotope tracing, the results of a single receptor model or a comprehensive discussion of both are used. This provides a scientific basis for watershed managers to formulate differentiated and precise pollution control strategies. For example, ecological agriculture can be promoted in agricultural non-point source pollution areas with high contribution rates; and the construction of sewage pipe networks can be accelerated in areas with significant domestic pollution.

[0065] Specific Case 1: For example Figures 2-3 As shown: The sample size is 48 (medium), there are few missing data, there are few previous studies, the watershed pollution sources are relatively complex, the denitrification conditions are mature, the budget is limited (no isotopes), and the uncertainty of the results is not considered. The weighted total score statistics are as follows:

[0066] From the above table, it can be seen that the weighted total score of model 3: PMF+APCS-MLR is the highest, and therefore, the analysis model is model 3. Mixed sources and agricultural non-point sources are the main contributors of NO3 – The control strategy for agricultural non-point sources is to promote ecological ditches and soil testing and formula fertilization in the middle and lower reaches of agricultural areas.

[0067] Specific case 2: as shown in the following table: Figure 4 The sample size is large, the data is missing, there are related nitrogen pollution source researches, the pollution sources are complex, there is nitrification, the budget is sufficient, there is no requirement for result uncertainty, and the project time is sufficient. The weighted total score statistics table is as follows:

[0068] The PMF model is run using the water chemical data of 29 samples, and the contribution rates of 4 main pollution sources factors (representing natural sources, agricultural sources, domestic sewage sources and manure sources respectively) are preliminarily analyzed. Using the δ 15 N and δ 18 O data of 10 samples, the Bayesian mixed model is run to quantify the contribution proportion of soil nitrogen, chemical fertilizer, domestic sewage and livestock and poultry manure to nitrate.

[0069] Domestic sewage is the main contributor of nitrogen pollutants (62.3%). In addition, there are additional inputs from soil nitrogen (13.4%), agricultural fertilizer (8.7%) and manure (12.9%). Therefore, the control strategy gives priority to reducing untreated or insufficiently treated sewage into the water system. The specific operation upgrades the wastewater treatment facilities, strictly implements the discharge standard, and implements the decentralized treatment system in the densely populated or suburban areas.

[0070] The above examples are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for part of the technical features; and these modifications or replacements will not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.​

Claims

1. A method for analyzing the synergistic source of nitrogen and phosphorus pollutants in a small watershed in a plain, characterized in that, The method comprises the following steps: Step 1: Collecting multi-source basic data in the target plain watershed, and performing unified coordinate system, format and precision preprocessing to construct a watershed basic database; Step 2: Constructing response relationship models of various pollution sources and potential pollution load based on the watershed basic database on a GIS platform, and qualitatively identifying main pollution sources and delineating key risk areas through spatial visualization analysis; Step 3: In the delineated key risk area, water sample collection points are arranged according to the water system trend and pollution source distribution, surface water and groundwater samples are collected, and the water samples are analyzed to obtain analysis data; Step 4: Determining an analysis model based on sampling data conditions, watershed characteristics and project constraints, and determining the pollution source contribution rate through the analysis data of the analysis model; Step 5: Obtaining nitrogen and phosphorus pollution source types and spatial positions based on the key risk area and the pollution source contribution rate, determining the source affecting the concentration of nitrogen and phosphorus pollutants through the nitrogen and phosphorus pollution source types and spatial positions, and formulating a control strategy according to the source affecting the concentration of nitrogen and phosphorus pollutants.

2. The method according to claim 1, wherein, The multi-source basic data comprises (1) high-resolution land use / land cover data, digital elevation model, soil type distribution map and the like; (2) high-precision population spatial distribution data, agricultural planting structure and fertilization amount statistics, location and pollution discharge amount of large-scale livestock and poultry breeding farms, location and pollution discharge permit information of industrial enterprise point source discharge outlets, and information of urban sewage treatment plants and the like; (3) watershed water system map, flow and water level data of hydrological monitoring stations, and historical water quality monitoring data.

3. The method according to claim 2, wherein, The specific operation of qualitatively identifying main pollution sources and delineating key risk areas through spatial visualization analysis is as follows: Step 1: Superimposing and analyzing the agricultural land distribution data such as farmland and garden land with the unit area fertilization intensity data to identify the areas with fertilization intensity higher than the agricultural non-point source pollution risk threshold value; Step 2: Superimposing the population density distribution with the sewage treatment facility coverage rate to identify the areas with direct discharge of domestic sewage or treatment capacity not meeting the standard; Step 3: Positioning industrial discharge outlets to identify the industrial discharge outlets with discharge amount greater than a set value; Step 4: The areas with fertilization intensity higher than the agricultural non-point source pollution risk threshold value, the areas with direct discharge of domestic sewage or treatment capacity not meeting the standard, and the industrial discharge outlets with discharge amount greater than a set value are delineated as key risk areas.

4. The method according to claim 3, wherein, The analysis of the water samples specifically comprises: (1) Determining conventional water quality indexes, including but not limited to total nitrogen, ammonia nitrogen, nitrate nitrogen, total phosphorus, chemical oxygen demand, pH, conductivity and the like; (2) Key determination of isotope indicators related to nitrogen pollution sources, such as δ 15 N-NO3 – , δ 18 O-NO3 – , for auxiliary identification of nitrogen pollution sources.

5. The method according to claim 4, wherein, The data conditions comprise sample quantity, data integrity and existing historical database.

6. The method according to claim 5, wherein, The specific steps of Step 4 are as follows: Step 41: Inputting the data conditions, watershed characteristics and project constraints into the constructed AHP model to score according to 1-5 points; Step 42: Calculating the maximum weight of the data conditions, watershed characteristics and project constraints, the score of the data conditions, watershed characteristics and project constraints, and the weighted total score of each scheme through the maximum weight and the score; Step 43: Determining the analysis model through the highest weighted score; Step 44: Substituting the analysis data obtained in Step 3 into the analysis model to calculate the pollution source contribution rate.

7. The method according to claim 6, wherein, The AHP model is specifically as follows: Model 1: PMF; Model 2: APCS-MLR; Model 3: PMF+APCS-MLR; Model 4: PMF+isotope tracing; Model 5: APCS-MLR+isotope tracing; Model 6: PMF+APCS-MLR+isotope tracing; Model 7: PMF+Monte Carlo; Model 8: APCS-MLR+Monte Carlo; Model 9: PMF+APCS-MLR+Monte Carlo; Model 10: PMF+isotope tracing+Monte Carlo; Model 11: APCS-MLR+isotope tracing+Monte Carlo; Model 12: PMF+APCS-MLR+isotope tracing+Monte Carlo. 8.The method according to claim 7, wherein, When PMF is in the analysis model of step 43, step 44 is specifically operated as follows: Step 44: input the analysis data obtained in step three into the positive definite matrix factorization model, and calculate the contribution rate of each type of pollution source to the nitrogen and phosphorus concentrations in the water body, that is, the pollution source contribution rate; Or, when APCS-MLR is in the analysis model of step 43, step 44 is specifically operated as follows: Step 44: input the analysis data obtained in step three into the absolute principal component score-multivariate linear regression model, and calculate the contribution rate of each type of pollution source to the nitrogen and phosphorus concentrations in the water body, that is, the pollution source contribution rate; Or, when isotope tracing is in the analysis model of step 43, step 44 is specifically operated as follows: Step 44: Based on delta 15 N and delta 18 O isotope data, the contribution ratio of different nitrogen sources, i.e. the pollution source contribution rate, is quantitatively estimated by using a Bayesian mixture model. Or, when Monte Carlo is in the analysis model of step 43, step 44 is specifically operated as follows: Step 44: using Monte Carlo simulation method, random disturbance is introduced to the input parameters and running process of PMF, APCS-MLR and Bayesian mixed model, multiple simulation calculations are carried out, and the uncertainty range of each model output result is quantified, that is, the pollution source contribution rate.

Images