Nitrogen and oxygen isotope tracing coupled correction accounting method for agricultural non-point source pollution

Abstract

The application discloses a nitrogen-oxygen isotope tracing coupled agricultural non-point source pollution correction and accounting method, relates to the technical field of agricultural non-point source pollution monitoring, and comprises the following steps: first, dividing an agricultural catchment unit, arranging equipment to build a multi-point sampling network, collecting runoff parallel water samples and flow time series data of each sampling point; detecting the concentration of water sample physicochemical pollutants and the nitrogen-oxygen isotope ratio, calling a local pollution source isotope threshold database, calculating the two-dimensional Euclidean distance between the measured isotope and the interval center value of the pollution source, and converting the relative contribution proportion of each pollution source; combining the concentration and flow to calculate the total pollution flux of single-point single rainfall, distributing to obtain the absolute pollution load of various pollution sources, removing the industrial, domestic point source and soil, atmospheric background interference load to obtain the real data of agricultural non-point source; through the subdivision of the load of the water inlet and outlet points, deducing and analyzing to obtain the absolute pollution load of each subdivision type and the overall real agricultural non-point source of the catchment unit, and generating a pollutant physical mass classification load account book.

Description

Technical Field

[0001] This invention relates to the field of agricultural non-point source pollution monitoring technology, and in particular to an agricultural non-point source load correction and accounting method based on nitrogen and oxygen isotope source tracing coupling. Background Technology

[0002] Accurate calculation of agricultural non-point source pollution load is the foundation for water environment management and emission reduction assessments in farmland and mixed agricultural areas. In recent years, automatic sampling technology has developed rapidly. Existing schemes mostly employ single-point sampling or simple multi-point synchronous sampling modes, relying on precipitation and flow velocity to trigger adaptive water sample collection, enabling the acquisition of watershed runoff samples and calculation of pollution fluxes. However, in actual agricultural catchment areas, the sources of nitrogen in runoff water bodies are extremely complex.

[0003] Existing conventional sampling methods can only obtain total nitrogen and total phosphorus concentration data. Relying on flow-weighted load calculations, they cannot distinguish the specific contributions of the aforementioned multiple sources. The final agricultural non-point source pollution load calculation will include industrial point sources, domestic sewage, soil-borne sources, and atmospheric deposition, resulting in severely inflated and distorted results. Summary of the Invention

[0004] To address the aforementioned problems, this invention provides a method for correcting and calculating agricultural non-point source loads based on nitrogen and oxygen isotope source tracing coupling. This method introduces nitrogen and oxygen dual isotopes (δ¹⁰ and δ¹⁰). 5 N、δ¹ 8 The O) Feature discrimination model, built on a multi-point linked adaptive sampling platform, constructs a complete local isotope database containing eight types of pollution sources, including three major agricultural non-point source types, point source interference types, and natural background types. By comparing the measured isotope ratios of water samples with the database, the contribution ratio of each source is broken down. Based on this, the isotope source apportionment results are directly coupled and weighted with the flow time series data from multi-point linked synchronous sampling. From the total pollution load calculated based on the measured values ​​of flow × concentration, the absolute load of industrial point sources, domestic point sources, soil endogenous sources, and atmospheric deposition is removed one by one. The absolute contribution load of planting, aquaculture, and livestock farming is further subdivided. Finally, the accurate classification, correction, and accounting of the real agricultural non-point source load of the catchment unit are achieved, and the classified load ledger data that can directly support the allocation of total emission control and emission reduction tasks is output.

[0005] The agricultural non-point source load correction and accounting method based on nitrogen and oxygen isotope source tracing coupling provided by this invention includes the following steps: S1: Deploy sampling equipment in different agricultural catchment units to continuously collect parallel runoff water samples and flow time series data at each sampling point; S2: Perform physicochemical and isotopic tests on the parallel runoff water samples to determine the pollutant concentration data and nitrogen-oxygen isotope ratio data in the water samples; S3: Call the local threshold database of pollution isotopes and calculate the relative contribution ratio of the nitrogen and oxygen isotope ratio data to the main pollution sources in the threshold database; S4: Perform flux calculation on the pollutant concentration data and the flow time series data to obtain the total pollution flux of a single rainfall event at a single location; and based on the relative contribution ratio of the main pollution sources, perform pollution source allocation analysis on the total pollution flux of a single rainfall event at a single location to obtain the absolute pollution load corresponding to each type of pollution source at a single location. S5: Based on the total pollution flux of a single rainfall event at a single location, the absolute pollution loads corresponding to industrial point sources, domestic point sources, soil endogenous sources, and atmospheric deposition are removed to obtain the real pollution data for agricultural non-point source pollution. S6: Based on the pollution load corresponding to the subdivided types of the outflow and inflow points of the agricultural water catchment unit, the pollution load generation at the inflow and outflow points is extrapolated and analyzed to obtain the absolute pollution load of the subdivided types. After summarizing, the total absolute pollution load of the agricultural non-point source pollution of the agricultural water catchment unit is obtained, and a classified load ledger containing the physical quantity of pollutants is generated.

[0006] In summary, the present invention has at least the following beneficial effects: 1. This invention establishes a complete technical chain from on-site monitoring to operational results output, forming a complete engineering closed loop of "sampling-source tracing-correction-segmentation-output". This invention constructs a multi-point linked adaptive sampling network and a localized isotope threshold database, simultaneously conducting physicochemical index and nitrogen and oxygen isotope detection of runoff samples. Through isotope source analysis, it decomposes the contribution ratio of eight types of pollution sources and calculates the absolute pollution load of each source by coupling measured flow-concentration data. It quantitatively separates the interference loads from industrial point sources, domestic sewage, soil endogenous sources, and atmospheric deposition, and screens out the true agricultural non-point source pollution load within the catchment unit. This overcomes the defects of traditional flow-weighted accounting, which cannot distinguish pollution sources and misclassifies non-agricultural interference sources as agricultural non-point sources, resulting in inflated loads and data distortion. This invention deeply couples automatic field hydrological and water quality monitoring, stable isotope laboratory analysis, and numerical analytical calculation. It does not rely on complex commercial hydrological and water quality models and can be directly applied to complex agricultural catchment units. It outputs classified load ledgers to support ecological and environmental business such as agricultural non-point source emission reduction responsibility accounting, water environment capacity allocation, and governance effectiveness assessment. It has strong engineering practicality.

[0007] 2. This invention adopts the design concept of a "local pollution source isotope threshold database." The database only uniformly defines the pollution source classification framework, which is fixed as eight types of pollution sources: three major agricultural non-point sources (crop farming non-point sources, aquaculture non-point sources, and livestock farming non-point sources), three types of point source interference (industrial point sources, urban domestic sewage point sources, and rural domestic scattered discharge point sources), and two types of natural background interference (endogenous soil natural erosion and atmospheric dry and wet deposition). The δ¹ values ​​corresponding to each type of pollution source are... 5 N、δ¹ 8 The specific threshold ranges for O dual isotopes are not globally fixed. In practical applications, implementers combine the geographical and climatic conditions, industrial structure, planting and fertilization patterns, aquaculture production characteristics, historical sample testing data, or on-site supplementary sampling and testing results of the target study area to complete localized calibration. Due to significant differences in soil parent material, fertilizer types, aquaculture patterns, and atmospheric deposition conditions in different regions, isotopic fingerprint characteristics will experience regional shifts. This invention abandons the practice of directly copying general fixed isotopic ranges from the literature, and adapts to the actual pollution fingerprint characteristics of different regions through localized calibration, effectively reducing source apportionment bias caused by applying fixed parameters across regions. This design makes the entire accounting method unrestricted by region, and can be adapted to various agricultural areas such as the southern plains and river network, the northern dryland farming areas, and hilly and mountainous mixed crop-livestock catchment units. It has good portability and wide applicability, overcoming the shortcomings of traditional isotopic analysis methods that are constrained by regional conditions.

[0008] 3. The data output by this invention, after correction and calculation, breaks through the limitations of traditional agricultural non-point source load calculation, which can only obtain a general total load. The output results are rich in dimensions, including not only the total absolute pollution load of agricultural non-point sources after correction of the catchment unit, but also the absolute nitrogen and phosphorus pollution loads of the three major agricultural industries: crop farming, aquaculture, and livestock farming. Simultaneously, it can also output the absolute flux and flux percentage data of various interference sources such as industrial point sources, domestic point sources, soil-borne sources, and atmospheric deposition. All results use the actual mass of pollutants as the unit of measurement, which differs from simple water concentration indicators and dimensionless contribution rates, and can objectively reflect the actual pollution levels of different industries within the catchment unit. Based on this calculation result, environmental management departments can accurately locate the main polluting industries in the region, identify the pollution contribution of crop farming, aquaculture, and livestock farming, clarify the contribution boundaries between anthropogenic agricultural emissions and point source and natural background interference, and avoid misallocation of emission reduction responsibilities due to inflated loads. Based on this, the pollution reduction tasks of each industry and each control unit can be scientifically and rationally decomposed, and targeted, differentiated agricultural non-point source pollution control measures can be formulated. At the same time, it can provide quantitative and traceable data support for watershed water quality target management, sewage control, and evaluation of the effectiveness of agricultural non-point source pollution control projects, and strongly support the refined and differentiated environmental management decisions of the watershed water environment. Detailed Implementation

[0009] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0010] This invention provides a method for agricultural non-point source load correction and calculation based on nitrogen and oxygen isotope source tracing coupling, comprising the following steps: Step 1: Construction of multi-point linkage sampling network and adaptive synchronous sampling.

[0011] The purpose of this step is to establish a field monitoring system with comprehensive spatial coverage and a highly unified temporal reference, and to acquire parallel water samples from multiple cross-sections of the catchment unit, hourly rainfall, and raw monitoring data of flow time series. This provides time-synchronized and spatially matched measured samples and hydrological basic data for subsequent water sample physicochemical and isotope detection, pollution flux integration calculation, isotope source apportionment, and load allocation of various pollution sources. This avoids problems that are easily encountered in traditional single-point sampling, such as confusion of pollution source sources, asynchronous rainfall sampling processes at different locations, and mismatch between water sample and flow time series.

[0012] First, a field survey and analysis of the target agricultural catchment unit is required to clarify the catchment boundaries, water flow direction, regional industrial spatial distribution, and the distribution of various pollution sources. Three key monitoring locations must be identified: the catchment unit inflow section, the concentrated output section of agricultural non-point source pollution, and representative monitoring points for various disturbance sources. The catchment unit inflow section is used to characterize the inflow of water from outside the unit, providing boundary conditions for subsequent calculations of the output load from primary agricultural non-point sources within the catchment unit by subtracting the input load from the output load.

[0013] The main cross-sections for concentrated agricultural non-point source activity include the total outlet of drainage from planting areas, the concentrated drainage outlet of aquaculture, and the total drainage outlet of livestock and poultry farming. These sections are used to capture runoff samples generated by the three core agricultural non-point sources: planting, aquaculture, and livestock and poultry farming. Representative monitoring points for various interference sources cover the monitoring locations corresponding to industrial wastewater, urban domestic sewage discharge, rural scattered discharge, soil background, and atmospheric deposition. These points correspond to point source interference and natural background interference sources among the eight types of pollution sources, ensuring that interference source samples and agricultural runoff samples have the conditions for comparative analysis. Spatially, this approach takes into account external inputs, internal agricultural emissions, and various interference sources within the unit, reducing analytical bias caused by the lack of comparative samples in subsequent isotope source tracing.

[0014] After completing the site planning, automatic runoff sampling equipment, flow monitoring equipment, and rainfall collection equipment need to be deployed at all selected monitoring sites to form a multi-point coordinated sampling network. The sampling start-stop logic of all sites within the network is fully linked and synchronized. A unified time clock is used for interference source sites, agricultural non-point source outlet sites, and inflow / outflow cross-section sites to ensure that water samples collected from different sites belong to the same rainfall-runoff event. Achieving a unified time reference across multiple sites at the hardware level overcomes the shortcomings of inconsistent sampling times at different sites in traditional decentralized sampling models, ensuring the time comparability of subsequent isotope source apportionment results and pollution flux calculation results.

[0015] Once the equipment network is complete, each monitoring point will enter a continuous monitoring state, continuously collecting hourly rainfall data and cross-sectional runoff velocity data. Real-time cross-sectional flow will be obtained through velocity conversion, and baseflow flow during non-rainfall periods will be recorded as a baseline for judging runoff changes. All monitoring data will be cached and stored locally on the equipment. The continuously accumulated rainfall and flow time-series baselines serve as the basis for automatically triggering subsequent rainfall-runoff events, thus distinguishing between baseflow status and rainfall-generated runoff status. This avoids collecting baseflow samples under no-rainfall conditions, which could interfere with the calculation of area source load under rainfall-runoff conditions.

[0016] The system continuously performs conditional judgments on the monitoring signals. When it detects that the hourly rainfall is not zero and the cross-sectional runoff flow is higher than the baseflow flow during the non-rainfall period, it determines that a rainfall-runoff event has occurred in the area. The system then issues a command to the entire sampling network, and all points simultaneously start water sampling, collecting parallel runoff water samples—that is, collecting multiple parallel water samples from the same point at the same time. The sampling equipment follows the complete runoff process, continuously collecting a series of water samples at preset time intervals, and simultaneously recording the cross-sectional flow at the corresponding time for each set of water samples, establishing a one-to-one pairing relationship between water sample and flow data time. The dual-condition triggering mechanism can filter out non-point source runoff driven by rainfall and eliminate interference from baseflow without rainfall; the synchronous start-up mode of all points ensures that all water samples come from the same rainfall event; parallel water samples can reduce experimental errors in subsequent physicochemical index and isotope detection; and the pairing results of water sample and flow time also provide necessary input parameters for the subsequent trapezoidal integral calculation of total pollution flux.

[0017] When hourly rainfall reaches zero and cross-sectional runoff returns to the baseflow level during the non-rainfall period, the rainfall-runoff event is considered to have completely ended, and water sampling at all points within the network is simultaneously terminated. After the event, all raw data must be archived. Parallel runoff samples collected from multiple points should be refrigerated and transferred to the laboratory for subsequent physicochemical and isotope ratio testing. Simultaneously, a complete time-series dataset of flow rates from all points throughout the entire rainfall event should be exported, retaining complete timestamps. This processing method ensures that the collected samples fully cover the entire process of rainfall-runoff rise, peak flow, and recession, preventing the interception of pollution output at the tail end of the runoff. The archived time-series flow data is matched with the water sample data, establishing a data link from on-site runoff observation to laboratory sample testing. This stage ultimately outputs parallel water samples from multiple runoff points, time-series flow data at each point with timestamps, and hourly rainfall monitoring sequences. It also forms a reusable multi-point linkage sampling and monitoring network. The collected water samples are directly used in the water sample testing and isotope ratio determination stages. The flow time-series data will also be called up multiple times in subsequent pollution flux calculation and catchment unit full-area load correction calculation steps.

[0018] Step 2: Water sample testing and isotope ratio determination.

[0019] This step uses parallel runoff water samples collected by a multi-point linked sampling network as the processing object. Following the physical water samples obtained from the previous on-site sampling, it simultaneously acquires two core pieces of information—the concentrations of conventional physicochemical pollutants and the nitrogen-oxygen isotope ratio—through laboratory analysis. The physicochemical concentration data will be used for subsequent quantitative calculations of pollution flux, while the nitrogen-oxygen isotope ratio data is a fundamental characteristic parameter for source apportionment of pollution. Both types of data must originate from the same set of water samples to ensure that subsequent flux calculations and source apportionment results are coupled and matched, avoiding calculation biases caused by sample mismatches. After the water samples arrive at the laboratory, pretreatment work is first required on the parallel runoff water samples transported from the site. The water samples are preserved, filtered, and aliquoted according to water quality testing standards, separating physicochemical index detection subsamples from isotope detection subsamples. Repeated experimental groups are set up for each parallel water sample to verify the stability of the testing process, reduce random errors introduced during transportation and pretreatment, and ensure that the test results accurately reflect the actual pollution state of the water body at the time of rainfall runoff occurrence.

[0020] After water sample pretreatment, routine physicochemical pollutant testing is conducted, primarily measuring total nitrogen and total phosphorus. Simultaneously, ammonia nitrogen, chemical oxygen demand, and other pollutants can be tested according to operational needs. Physicochemical testing strictly adheres to relevant water environment monitoring standards and methods to obtain pollutant concentration values ​​for each group of water samples. Each test result is linked to the corresponding sampling point and sampling timestamp. The core function of this physicochemical concentration data is to complement the flow time-series data obtained in previous steps, serving as input for the trapezoidal integral flux calculation formula. This formula calculates the total pollution flux at a single point under a single rainfall event, enabling a quantitative solution for the actual pollution load. Without accurate synchronous concentration data, subsequent calculations of the absolute pollution load from various pollution sources become impossible, yielding only dimensionless source contribution ratios, which cannot meet the operational requirements for physical load records in the refined management of agricultural non-point source pollution.

[0021] After completing the physicochemical index testing, nitrogen and oxygen isotope analysis needs to be conducted using pretreated water samples from the same batch. The δ¹⁸O values ​​for nitrate nitrogen will be determined using an isotope ratio mass spectrometer. 15 N and δ 18 The O isotope ratio was used to obtain the measured dual isotope characteristic values ​​for each water sample. Isotope detection also utilized parallel water samples collected on-site to set up detection replicates, controlling the testing error of the mass spectrometer and ensuring accurate measured δ¹⁸O values. 15 N, δ 18 The reliability of the O results. This set of isotope ratios characterizes the source fingerprint features of nitrogen in water. Different pollution sources have their own unique nitrogen and oxygen isotope distribution ranges, which is the core basis for this invention to distinguish eight types of pollution sources. Subsequently, this set of measured isotope coordinates will be compared with the interval center values ​​of various pollution sources in the local pollution source isotope threshold database to calculate the two-dimensional Euclidean distance, and then the relative contribution ratio of various pollution sources will be separated.

[0022] Step 3: Constructing a local pollution source isotope threshold database.

[0023] This step builds upon the preliminary field survey data of the catchment unit and the historical sample testing data of the region. The core value of this database is to provide a reference benchmark for subsequent measured isotopic characteristics of water samples. If a local isotopic threshold range suitable for the study area is lacking, calculations will be affected by differences in geographical conditions, industrial models, and farming and aquaculture habits in different regions, resulting in distorted analysis of the contribution ratio of pollution sources. This, in turn, will affect the subsequent allocation of absolute pollution load and the effectiveness of interference source removal.

[0024] The database construction does not directly apply numerical values ​​from general literature, but is based on the actual situation of the target agricultural catchment unit. First, it completes the classification and sorting of regional pollution sources, fully identifying all major nitrogen pollution sources within the catchment area. All pollution sources are divided into three major groups: agricultural non-point source pollution, point source pollution interference, and natural background interference. These are further subdivided into eight categories: crop non-point source pollution, aquaculture non-point source pollution, livestock and poultry farming non-point source pollution, industrial point source pollution, urban domestic sewage point source pollution, rural domestic scattered discharge point source pollution, soil natural erosion internal sources, and atmospheric dry and wet deposition. This fully covers all source types that this invention needs to distinguish and identify. This classification operation clarifies the list of objects that the database needs to label, ensuring that subsequent source apportionment can simultaneously distinguish different industries within agriculture, as well as point sources and natural background interference, providing a classification basis for subsequently separating non-agricultural non-point source loads from the total load.

[0025] After establishing the eight-category pollution source classification framework, it is necessary to collect basic data corresponding to the target catchment units, including regional industrial survey results, historical monitoring sample test results of various pollution source emission outlets, and, when conditions permit, targeted supplementary sampling and testing can be carried out to obtain the delta values ​​of various pollution sources. 15 N, δ 18 O Measured sample data. For each type of pollution source, the isotope detection results of multiple sets of samples were summarized, and the corresponding δ¹⁸O values ​​for that pollution source were statistically obtained. 15 N and δ 18 The O-isotope threshold range refers to the upper and lower limits of the isotopic fingerprint distribution of this type of pollution source. It's important to note that isotopic threshold ranges have a significant regional attribute. Different fertilizer application patterns, aquaculture species, soil parent material, and atmospheric environmental conditions in different regions can all cause shifts in isotopic characteristics. Therefore, fixed parameters from other regions cannot be directly applied. All threshold ranges are calibrated based on samples from the local region. This improves the database's adaptability to local pollution sources and reduces analytical errors caused by applying parameters across regions. For each determined threshold range, the interval's center value is further calculated. This center coordinate represents a typical representative point of the isotopic characteristics of each type of pollution source. In subsequent steps, the measured isotopic points of water samples are determined by calculating the two-dimensional Euclidean distance between these center coordinates to complete the source similarity discrimination.

[0026] After completing the calibration calculation of the dual isotope threshold ranges and corresponding center values ​​for eight types of pollution sources, all data needs to be structured and stored to form a local pollution source isotope threshold database that can be called by the program. Each data record in the database is bound to the pollution source category name, its major category, and δ. 15 N threshold upper and lower limits, δ 18The upper and lower limits of the O threshold and the center coordinates of the isotope intervals ensure that all parameters can be quickly retrieved according to the pollution source number during subsequent calculations. The database is not static after its establishment. When the industrial structure of the study area is adjusted, the planting and breeding patterns are significantly changed, or more new pollution source sample data are accumulated, the isotope threshold intervals of each pollution source can be iteratively updated to continuously optimize the localization of the database.

[0027] After the database is built, it outputs a callable local pollution source isotope threshold database to the downstream isotope source analysis and contribution ratio decomposition steps. In subsequent calculations, the program will directly retrieve the center coordinates of the isotope intervals of the eight types of pollution sources in the database, and the measured δ¹⁸O values ​​obtained from water sample testing. 15 N, δ 18 The O-ratio is used to perform distance calculations and convert the relative contribution percentages of various pollution sources. The establishment of the entire database enables the fingerprint characteristics of regional pollution sources to be solidified into calculable digital parameters, and transforms the qualitative isotope identification characteristics of pollution sources into coordinate parameters that can participate in mathematical calculations. This provides a discrimination standard for realizing a complete closed-loop accounting system of "isotope source tracing - flux coupling - interference elimination - load subdivision," which is the key foundation for this invention to distinguish itself from traditional isotope analysis schemes and achieve complete separation of eight types of sources.

[0028] Step 4: Isotope source analysis and contribution percentage breakdown.

[0029] The input for this step comes from the measured nitrogen and oxygen isotope ratio data of water samples output from the previous step, as well as the established local pollution source isotope threshold database. The relative contribution percentages of various pollution sources output in this step are dimensionless intermediate calculation parameters and are not directly used as the final operational results. They need to be combined with the total pollution flux of the rainfall event to obtain the absolute pollution load of each pollution source. If this analysis step is skipped and only the physicochemical concentration and flow data are used, only the total pollution flux of the catchment unit can be obtained. It is impossible to distinguish the contribution of each of the eight different sources, and therefore it is impossible to eliminate non-agricultural disturbance loads such as industrial, domestic, soil, and atmospheric pollution, making it difficult to achieve the corrected calculation of the actual agricultural non-point source load.

[0030] First, based on the nitrogen and oxygen isotope detection results obtained from the water sample laboratory, the measured values ​​of each parallel runoff water sample are used as the x and y coordinates of a Cartesian coordinate system to construct the measured isotope coordinate points of the water samples: ; This coordinate system transforms the nitrogen and oxygen isotope fingerprint of the water sample into calculable geometric points on a plane, converting isotopic characteristic differences into spatial distance differences, thus establishing a mathematical foundation for subsequent similarity quantification calculations. Simultaneously, it calls upon a local pollution source isotope threshold database, extracting the pre-calculated isotope interval center values ​​for eight types of pollution sources within the database to form the isotope interval center coordinates corresponding to the k-th pollution source: ; In the formula, These correspond to eight categories of pollution sources: agricultural non-point source pollution, aquaculture non-point source pollution, livestock and poultry farming non-point source pollution, industrial point source pollution, urban domestic sewage point source pollution, rural domestic scattered discharge point source pollution, soil natural erosion internal source pollution, and atmospheric dry and wet deposition pollution. By retrieving the center coordinates of each type of pollution source, it is equivalent to marking the typical characteristic points of the eight pollution sources on the same two-dimensional isotopic plane, establishing all reference objects for comparative calculations between water sample measurement points and characteristic points of each pollution source.

[0031] In one implementation, a two-dimensional Euclidean distance-reciprocal distance normalization method is used to calculate the relative contribution of major pollution sources. This method has simple calculation logic, high computational efficiency, and does not require complex model parameter debugging, making it suitable for grassroots ecological and environmental departments to conduct rapid business accounting of agricultural non-point source load.

[0032] Specifically, the two-dimensional Euclidean distances from the measured isotopic coordinates of the water sample to the center coordinates of the isotopic intervals of the eight types of pollution sources are calculated using the following formula: ; In the formula, The measured isotope coordinates of the water sample are as follows: The two-dimensional Euclidean distance between the center coordinates of pollution sources is used. The magnitude of the distance represents the similarity between the isotopic fingerprint of the water sample and the characteristics of the corresponding pollution source. The smaller the distance value, the closer the nitrogen isotopic composition of the water sample is to the isotopic characteristics of this type of pollution source, and the higher the probability that the pollution source has contributed to the pollution of the water sample; conversely, the larger the distance, the lower the probability that the pollution source has contributed. Quantifying the similarity through Euclidean distance solves the problem that isotopic characteristics can only be qualitatively identified and it is difficult to directly assign the contribution ratio.

[0033] Obtain the distances corresponding to all eight types of pollution sources Next, the relative contribution percentage of each pollution source was calculated using the inverse distance normalization method. First, the inverse distance to each type of pollution source was calculated; the smaller the distance, the larger the inverse distance, thus amplifying the weight of nearby pollution sources. Then, the inverse distances of all eight types of pollution sources were normalized to obtain the relative contribution percentage of the k-th type of pollution source. The calculation formula is: ; After conversion using this formula, the relative contribution percentages of the eight types of pollution sources corresponding to each water sample meet the requirements. That is, the sum of the contribution percentages of all sources is 100%. The purpose of the inverse distance normalization process is to transform the geometric distance into a normalized contribution weight, realizing the mathematical transformation from isotopic spatial similarity to the relative contribution percentage of pollution sources, and obtaining the dimensionless contribution percentage of each of the eight sources in each water sample.

[0034] In one implementation, when calculating the relative contribution of major pollution sources, the MixSIAR isotope mixing model is used to quantify the contribution rate of different pollution sources, breaking down the pollution contribution percentage of each source in each water sample. This method is suitable for scientific research assessment scenarios with sufficient sample base and high requirements for result accuracy.

[0035] Specifically, the existing Bayesian mixture model is used to perform source apportionment calculations. The probability distributions of various pollution source isotopes obtained from the local pollution source isotope threshold database are used as inputs to the model endmembers. Prior parameters of isotope fractionation can also be introduced based on the regional nitrogen migration and transformation characteristics. The model outputs the probability distribution and confidence interval of the relative contribution of each pollution source, thus completing the uncertainty quantification of the source apportionment results.

[0036] It should be noted that the Bayesian mixture model is a well-known existing tool in the field for isotope source analysis. It uses Markov-Monte Carlo (MCMC) sampling operations to solve the posterior probability of the contribution ratio of pollution sources based on the endmember isotope probability distribution, and then obtains the probability distribution and confidence interval of the contribution ratio. The specific calculation implementation can be completed using existing open source programs such as MixSIAR. This invention does not modify the internal calculation logic of the Bayesian model, but only adapts it as an optional analysis module to be integrated into this calculation process.

[0037] After calculating the proportions, the calculated relative contribution proportions of each pollution source need to be linked and archived with the corresponding sampling points, rainfall event numbers, and water sample numbers to ensure that each set of proportion results can be traced back to the corresponding field water samples and test records. This set of proportion data can only characterize the relative composition of various sources and cannot directly reflect the actual physical mass of pollutants emitted; therefore, it cannot be used alone as a result to evaluate the pollution emission level of the region. The calculation results will then flow down to the multi-source interference pollution flux elimination and agricultural non-point source subdivision steps, multiplying with the total pollution flux to transform the dimensionless contribution proportions into the absolute pollution load of various pollution sources. This completes the coupling of isotope source tracing results and hydrological flux data, providing core allocation coefficients for eliminating point sources and natural background interference, and breaking down the loads of crop farming, aquaculture, and livestock farming.

[0038] Step 5: Multi-source interference pollution flux elimination and agricultural non-point source subdivision.

[0039] This step builds upon the point flow time-series data, water sample pollutant concentration data, and the relative contribution ratios of each pollution source obtained from isotope source apportionment from the previous steps. It couples the dimensionless source contribution ratios with the measured hydrological-concentration data, converting the relative ratios into absolute pollution load quantities for each type of pollution source. Then, it isolates the interference loads from non-agricultural non-agricultural sources such as industrial point sources, domestic point sources, soil-borne sources, and atmospheric deposition from the total pollution load, separating the agricultural non-agricultural non-agricultural loads for crop farming, aquaculture, and livestock farming. Traditional accounting methods directly attribute all pollution fluxes obtained from cross-sectional monitoring to agricultural non-agricultural sources, resulting in a large amount of pollution from non-agricultural sources and inflated, distorted accounting results. This step addresses this issue by correcting for this problem. The output absolute loads of various pollution sources for single-point rainfall events will provide the basic calculation data for subsequent load extrapolation and accounting at the overall scale of the catchment unit.

[0040] First, retrieve the pollutant concentration sequences corresponding to all water samples from this sampling point under a single rainfall event. and the matched cross-sectional flow time series The trapezoidal integral method was used to calculate the total pollution flux of a single rainfall event at a single location. The calculation formula is as follows: ; In the formula, This represents the total pollution flux at this location during a single rainfall event. For the first Pollutant concentration data for the group of water samples; For the first The cross-sectional flow rate of the water samples at the corresponding time points; The time interval between two adjacent sets of samples; This represents the number of sampling groups for a single rainfall event. The trapezoidal integral method can adapt to the rapid fluctuations in concentration and flow during rainfall runoff. It integrates and sums the entire process of rising water, peak flow, and receding water. Compared with a simple flow-weighted average, it can better reflect the actual changes in rainfall runoff and more accurately restore the total cross-sectional pollution flux generated by a rainfall event. The total flux obtained includes the sum of contributions from all eight types of pollution sources.

[0041] After obtaining the total pollution flux of a single rainfall event at a single location, combine it with the relative contribution percentages of various pollution sources calculated in the previous step. To calculate the pollution load allocation at a given location, the total pollution flux is multiplied by the contribution percentage of each corresponding pollution source to obtain the absolute pollution load of the k-th type of pollution source. : ; This calculation transforms the dimensionless contribution percentages obtained from isotope analysis into physical pollutant masses with actual physical meaning, measured in kilograms or tons. It distinguishes between industrial point sources, urban domestic sewage point sources, rural domestic scattered discharge point sources, internal sources of natural soil erosion, atmospheric dry and wet deposition, and non-point sources from agriculture, aquaculture, and livestock farming, each with its corresponding absolute load. The core function of this step is to establish a conversion relationship between source apportionment ratios and actual pollution loads. It no longer merely provides the percentage of each source, but quantifies the specific amount of nitrogen and phosphorus pollutants emitted by each type of pollution source, providing a directly subtractable quantitative value for subsequent deduction of interference terms.

[0042] Based on obtaining the absolute pollution loads of the eight types of pollution sources, a process of eliminating multi-source interference loads is carried out. The absolute pollution loads corresponding to industrial point sources, urban domestic sewage point sources, rural domestic scattered discharge point sources, natural soil erosion sources, and atmospheric dry and wet deposition are removed from the total flux. The remaining portion represents the actual pollution data for agricultural non-point source pollution at that location. It is important to clarify that the objects removed here include both anthropogenic point source pollution interference and nitrogen and phosphorus inputs from natural backgrounds such as soil erosion and atmospheric deposition. These sources are not considered non-point source pollution from agricultural production activities; if they are not removed, they will be incorrectly included in the agricultural non-point source load, leading to an overestimation of agricultural pollution contribution. The remaining actual agricultural non-point source data after elimination can be further broken down into three sub-types: crop farming non-point sources, aquaculture non-point sources, and livestock and poultry farming non-point sources. The independent load values ​​for each of these three agricultural sources are recorded, achieving sub-quantification of different industrial sources within agricultural non-point sources.

[0043] After the calculations are completed, the total pollution flux of a single rainfall event at a single point, the absolute loads of each of the eight pollution sources, the total agricultural non-point source load after removing interference, and the sub-loads of crop farming, aquaculture, and livestock farming need to be associated and archived with the sampling point number and rainfall event identifier. This ensures that each set of calculation results can be traced back to the original water sample, flow record, and isotope analysis parameters. The output of this step is still a calculation result at the single-section scale, representing the various loads observed at the monitoring section location. It cannot be directly equated to the pollution load originally generated within the entire catchment unit, because it also includes the external pollution input brought into this unit by the upstream inflow section. Therefore, this part of the calculation result will be completely transferred to the agricultural non-point source classification load correction and accounting step of the catchment unit. By extrapolating the load difference between the inflow and outflow sections, the influence of upstream water is eliminated, and finally the absolute pollution load of each sub-type of agricultural non-point source actually produced within the entire catchment unit is solved.

[0044] Step 6: Calculation of agricultural non-point source pollution classification load for water catchment units.

[0045] This step builds upon the previous step's elimination of multi-source interference pollution flux and subdivision of agricultural non-point sources, which yielded absolute pollution loads for each monitoring point. It also reuses the time-series data of influent and effluent flow rates collected by the multi-point linkage sampling network. The core objective is to eliminate interference from external pollution carried by upstream influent into the catchment unit and transform the observed cross-sectional load into the actual agricultural non-point source pollution load generated within the catchment unit.

[0046] The loads obtained from the previous steps are only flux values ​​observed at the monitoring sections. The agricultural non-point source loads monitored at the outlet sections are mixed with external pollution brought into the unit by the upstream inflow. If the observed loads at the outlet sections are directly used as the regional pollution generation, pollution that does not belong to this catchment unit will be included in the statistics, resulting in an overestimation of the agricultural non-point source load in this region. Therefore, it is necessary to use the load difference between the inflow and outflow sections to extrapolate, remove the influence of external inputs, and obtain the true pollution generation level of the unit's internal sources, so as to provide quantitative physical data at the whole-area scale for environmental management and emission reduction task allocation.

[0047] First, the absolute pollution loads of various agricultural non-point sources at the outflow and inflow sections of the catchment unit, after isotope tracing and interference removal, were obtained and denoted as follows: and The subscript 'm' represents the three subcategories of agricultural non-point sources: crop farming, aquaculture, and livestock farming. The native absolute pollution load of each agricultural non-point source within the catchment unit is calculated by subtracting the corresponding subcategories of pollution loads from the influent point from the pollution loads of each subcategories at the effluent point. The computational relationship is expressed as: ; The core logic of this difference calculation is that a certain type of agricultural pollution load monitored at the outlet section comes partly from upstream inflows carried into this unit, and partly from the pollution actually generated and output by various agricultural production activities within the catchment unit. By subtracting the loads of the same type, the contribution of upstream transit pollution can be eliminated, restoring the sub-item pollution load generated only by production activities within this catchment unit, thus avoiding interference from external water inflows on the local pollution generation accounting results.

[0048] The unit-specific absolute pollution loads of agricultural non-point source pollution, aquaculture non-point source pollution, and livestock and poultry farming non-point source pollution were calculated separately. Then, the loads of the three types of agricultural non-point sources are summed to obtain the total absolute pollution load of agricultural non-point sources for the entire agricultural catchment unit, which satisfies: ; This total load represents the total physical amount of nitrogen and phosphorus pollutants produced by all agricultural activities within the target catchment unit, measured in kilograms or tons. It differs from the simple increase in concentration at river sections and from the dimensionless contribution percentage output by isotope analysis; it is a physical quality indicator that can be directly used for operational management. By calculating and then summing the results for each item, we can obtain the total load for the entire region while retaining the independent load information for each agricultural industry. This facilitates the identification of the pollution contribution of different agricultural types and supports the development of differentiated governance plans, rather than simply obtaining a general regional total value.

[0049] After completing the numerical calculations of the individual and total loads, a categorized load ledger containing physical mass information of nitrogen and phosphorus pollutants needs to be constructed. This ledger comprehensively records the total absolute pollution load from agricultural non-point sources within the entire catchment unit. It also separately records the absolute physical quantities of pollution from crop farming, aquaculture, and livestock farming. Furthermore, it can simultaneously include flux data from various disturbance sources within the catchment unit, such as industrial point sources, residential point sources, natural soil erosion, and atmospheric dry and wet deposition. Each load result in the ledger is associated with the corresponding rainfall event, sampling batch, original detection and calculation parameters, ensuring the traceability of all calculation results. This allows for reverse tracing back to on-site sampling records, water sample physicochemical and isotope detection reports, database parameters, and each step of the calculation process.

[0050] After completing the data processing, the corrected agricultural non-point source load data can be simultaneously uploaded to the business terminal platform for data visualization. This allows for the presentation of the original cross-sectional observation load, the corrected load after removing various interference sources, the detailed contribution loads of the three major agricultural non-point sources, and the flux proportions of various interference sources. This step enables managers to intuitively compare the differences between the uncorrected original results and the corrected actual results, clearly distinguishing the contributions of local agricultural endogenous output and external transit pollution, point sources, and natural background interference. At this point, the entire technical process of "multi-point sampling - water sample testing - isotope database construction - source apportionment proportion breakdown - interference load removal - catchment unit correction and accounting" is fully closed-loop. The final output of the categorized load ledger can directly serve practical environmental management scenarios such as watershed water environment capacity accounting, total agricultural non-point source pollution control, decomposition of emission reduction tasks for various industries, and evaluation of regional water environment governance effectiveness.

[0051] Example: Application of composite agricultural catchment units in plain river network areas.

[0052] This embodiment uses a typical river network area in a plain in southern my country as an example to illustrate the method of the present invention in detail. The catchment unit covers an area of ​​approximately 25 km². 2The region is characterized by a mix of crop farming (rice and vegetables), aquaculture (fish ponds and shrimp ponds), and livestock farming (pig farms). It also includes one small food processing enterprise, one urban sewage treatment plant discharge outlet, and three rural residential areas.

[0053] A total of 18 sampling points were set up at the following locations within the catchment unit: one inlet section, two outlets for drainage from the planting area, three aquaculture outlets, two livestock and poultry farm outlets, one industrial enterprise drainage monitoring point, one urban domestic sewage discharge monitoring point, three monitoring points for interference near residential areas, two atmospheric deposition monitoring points, and three soil background monitoring points. Sampling equipment at each point collected hourly rainfall and cross-sectional runoff velocity data in real time. When the hourly rainfall was not zero and the cross-sectional runoff flow rate was higher than the base flow rate during the non-rainfall period, the sampling at each point was automatically triggered to start and stop simultaneously, obtaining parallel runoff samples from multiple points and synchronously recording the time-series runoff flow data for each point during the corresponding time period.

[0054] Total nitrogen, total phosphorus, ammonia nitrogen, and chemical oxygen demand (COD) concentrations were measured in all synchronously collected water samples. Simultaneously, stable isotope ratio mass spectrometry (SMR) was used to detect the δ¹⁸O₂ concentration in the water samples. 15 N and δ 18 O nitrogen-oxygen isotope ratio.

[0055] Based on the regional background survey and historical sample testing data of this catchment unit, the δ¹⁸O values ​​of eight types of pollution sources were pre-identified. 15 N and δ 18 The O dual isotope characteristic threshold ranges are stored in the database. The measured δ¹ values ​​of water samples from each location are... 5 N and δ¹ 8 The O isotope ratio was compared with the threshold database for discrimination. A two-dimensional Euclidean distance-reciprocal distance normalization model was used to quantify the contribution rate of different pollution sources and to break down the pollution contribution percentage (%) of each source in each water sample.

[0056] Based on synchronously recorded flow time-series data and measured concentrations of total nitrogen, total phosphorus, ammonia nitrogen, and chemical oxygen demand, the absolute load of each pollutant at the inlet and outlet sections of the catchment unit is calculated using the trapezoidal integral formula. In this embodiment, the absolute load of total nitrogen at the outlet of the catchment unit under a certain rainfall event is 45 kg. The absolute load of each pollution source is calculated by multiplying the relative contribution percentage of each source obtained in step five by the absolute load of total nitrogen at the outlet; among them, industrial point sources contribute approximately 18% with an absolute load of approximately 8 kg; urban residential point sources contribute approximately 22% with an absolute load of approximately 10 kg; rural scattered discharge point sources contribute approximately 4% with an absolute load of approximately 1.8 kg; soil endogenous sources contribute approximately 14% with an absolute load of approximately 6.3 kg; and atmospheric deposition contributes approximately 6% with an absolute load of approximately 2.7 kg. The absolute loads of each of the above-mentioned interference sources are stripped from the 45 kg total nitrogen absolute load (totaling approximately 28.8 kg), resulting in a true absolute load of agricultural non-point source pollution of approximately 16.2 kg. Further breakdown reveals that the non-point source contribution from crop farming is approximately 6.5 kg, from aquaculture approximately 5.2 kg, and from livestock and poultry farming approximately 4.5 kg.

[0057] The calculated absolute loads of each sub-type of agricultural non-point source pollution are combined with the loads of the corresponding sub-types at the influent section for differential correction. External pollution input from upstream water is deducted to obtain the original absolute pollution loads of each sub-type of agricultural non-point source pollution within the catchment unit. The total absolute pollution load of the entire catchment unit is summarized as 16.2 kg. The absolute loads contributed by each sub-type are: crop farming 6.5 kg, aquaculture 5.2 kg, and livestock farming 4.5 kg. The absolute fluxes of various disturbance sources are: industrial point sources 8 kg, urban residential point sources 10 kg, rural scattered discharge point sources 1.8 kg, soil endogenous sources 6.3 kg, and atmospheric deposition 2.7 kg. All data are simultaneously uploaded to the terminal platform in absolute load (kg) format, forming a categorized load ledger that can directly support total emission control and emission reduction task allocation. The original observed loads, corrected agricultural non-point source sub-loads, and the flux proportions of various disturbance sources are visualized at each location.

[0058] The above are merely preferred embodiments of the invention and are not intended to limit the invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the invention should be included within the protection scope of the invention.

Claims

1. A method for correcting and calculating agricultural non-point source loads based on nitrogen and oxygen isotope source tracing coupling, characterized in that, Includes the following steps: S1: Deploy sampling equipment in different agricultural catchment units to continuously collect parallel runoff water samples and flow time series data at each sampling point; S2: Perform physicochemical and isotopic tests on the parallel runoff water samples to determine the pollutant concentration data and nitrogen-oxygen isotope ratio data in the water samples; S3: Call the local threshold database of pollution isotopes and calculate the relative contribution ratio of the nitrogen and oxygen isotope ratio data to the main pollution sources in the threshold database; S4: Perform flux calculation on the pollutant concentration data and the flow time series data to obtain the total pollution flux of a single rainfall event at a single location; and based on the relative contribution ratio of the main pollution sources, perform pollution source allocation analysis on the total pollution flux of a single rainfall event at a single location to obtain the absolute pollution load corresponding to each type of pollution source at a single location. S5: Based on the total pollution flux of a single rainfall event at a single location, the absolute pollution loads corresponding to industrial point sources, domestic point sources, soil endogenous sources, and atmospheric deposition are removed to obtain the real pollution data for agricultural non-point source pollution. S6: Based on the pollution load corresponding to the subdivided types of the outflow and inflow points of the agricultural water catchment unit, the pollution load generation at the inflow and outflow points is extrapolated and analyzed to obtain the absolute pollution load of the subdivided types. After summarizing, the total absolute pollution load of the agricultural non-point source pollution of the agricultural water catchment unit is obtained, and a classified load ledger containing the physical quantity of pollutants is generated.

2. The method for agricultural non-point source load correction and accounting based on nitrogen and oxygen isotope source tracing coupling according to claim 1, characterized in that, The specific steps for deploying sampling equipment within different agricultural catchment units and continuously collecting parallel runoff water samples and flow time-series data at each sampling point are as follows: Agricultural catchment units are pre-divided, and sampling equipment is deployed at the inlet section of the catchment unit, the concentrated output section of agricultural non-point source activity, and representative detection points of various interference sources to build a multi-point sampling network. The sampling equipment collects hourly rainfall and cross-sectional runoff velocity data in real time. When the hourly rainfall is not zero and the cross-sectional runoff flow rate is higher than the flow rate during the non-rainfall period, it triggers the synchronous start of water sampling at all points. After the rainfall and runoff process ends, water sampling at each location is terminated simultaneously to obtain parallel water samples from multiple locations. At the same time, the flow time series data of each location during the sampling process are recorded.

3. The method for agricultural non-point source load correction and accounting based on nitrogen and oxygen isotope source tracing coupling according to claim 1, characterized in that, The steps for constructing a threshold database of local contamination isotopes are as follows: Pollution types within agricultural catchment units are summarized and categorized into agricultural non-point source pollution, point source pollution disturbance, and natural background disturbance. Agricultural non-point source pollution includes non-point sources from crop farming, aquaculture, and livestock farming; point source pollution disturbance includes industrial point sources, urban domestic sewage point sources, and rural domestic scattered discharge point sources; and natural background disturbance includes internal sources of natural soil erosion and atmospheric dry and wet deposition. Based on historical sample detection data, δ¹ corresponding to eight types of pollution sources were respectively calibrated. 5 N、δ¹ 8 The O dual isotope threshold range stores the calibrated dual isotope threshold ranges of various pollution sources, generating a local pollution source isotope threshold database.

4. The method for agricultural non-point source load correction and accounting based on nitrogen and oxygen isotope source tracing coupling according to claim 3, characterized in that, The specific steps for calculating the relative contribution percentage of the nitrogen-oxygen isotope ratio data to the main pollution sources in the threshold database by calling the local pollution isotope threshold database are as follows: δ¹ measured from parallel water samples of runoff 5 N、δ¹ 8 The nitrogen and oxygen isotope ratio data were used as the x and y coordinates of a Cartesian coordinate system to construct the measured isotope coordinate points of the water sample. : ； Retrieve the dual isotope threshold intervals of eight types of pollution sources from the local pollution source isotope threshold database, calculate the center value of the dual isotope threshold interval for each type of pollution source, and form the first... The center coordinates of the isotopic intervals corresponding to the type of pollution source : ； In the formula, These correspond to eight categories of pollution sources, in turn. Calculate the two-dimensional Euclidean distance from the measured isotopic coordinates of the water sample to the center coordinates of the isotopic intervals of the eight types of pollution sources. The distance calculation formula is as follows: ； In the formula, The measured isotope coordinates of the water sample are as follows: Two-dimensional Euclidean distance between the center coordinates of the pollution source.

5. The method for agricultural non-point source load correction and accounting based on nitrogen and oxygen isotope source tracing coupling according to claim 4, characterized in that, The specific steps for converting two-dimensional Euclidean distance into the relative contribution percentage of major pollution sources are as follows: First calculate the reciprocal of the two-dimensional Euclidean distance, then calculate the first... The relative contribution of different types of pollution sources : ； The relative contribution percentage of each type of pollution source for each water sample was calculated. and satisfy .

6. The method for agricultural non-point source load correction and accounting based on nitrogen and oxygen isotope source tracing coupling according to claim 3, characterized in that, The specific steps for calculating the relative contribution percentage of the nitrogen-oxygen isotope ratio data to the main pollution sources in the threshold database by calling the local pollution isotope threshold database are as follows: Using a Bayesian mixture model, the probability distribution of isotopes of each pollution source in the local pollution isotope threshold database is used as the input of the Bayesian mixture model, and the isotope fractionation offset is used as the prior parameter to calculate and output the probability distribution and confidence interval of the relative contribution of each major pollution source.

7. The method for agricultural non-point source load correction and accounting based on nitrogen and oxygen isotope source tracing coupling according to claim 5, characterized in that, The specific steps for calculating the total pollution flux of a single rainfall event at a single location by combining the pollutant concentration data and the flow time series data are as follows: By retrieving the flow time-series data and pollutant concentration data for each location, and substituting them into the flux calculation formula in the form of a trapezoidal integral, the total pollution flux of a single rainfall event at a single location can be calculated. : ； In the formula, This represents the total pollution flux at this location during a single rainfall event. For the first Pollutant concentration data for the group of water samples; For the first The cross-sectional flow rate of the water samples at the corresponding time points; The time interval between two adjacent sets of samples; This represents the number of sampling groups for a single rainfall event.

8. The method for agricultural non-point source load correction and accounting based on nitrogen and oxygen isotope source tracing coupling according to claim 1, characterized in that, Based on the pollution load corresponding to the subdivided types of the outflow and inflow points of agricultural water catchment units, the pollution load generation at the inflow and outflow points is extrapolated and analyzed to obtain the absolute pollution load of each subdivided type. These are then summarized to obtain the overall true absolute agricultural non-point source pollution load of the agricultural water catchment unit. The specific steps for generating a categorized load ledger containing the physical quantities of pollutants are as follows: By subtracting the corresponding sub-type pollution load brought in from the inlet point from the pollution load of each sub-type at the outlet point, the absolute pollution load of each sub-type of non-point source from crop farming, aquaculture, and livestock farming within the catchment unit is obtained. The absolute pollution loads of crop farming non-point sources, aquaculture non-point sources, and livestock and poultry farming non-point sources are summarized to obtain the overall true total absolute pollution load of agricultural non-point sources for the agricultural catchment unit. Based on the absolute pollution load of each sub-type, a classified load ledger containing physical mass information of nitrogen and phosphorus pollutants is constructed. The ledger records the physical absolute load of each pollution source.

Images