A river nitrogen pollution source apportionment method based on process simulation and isotope tracing

Abstract

The application discloses a river nitrogen pollution source analysis method based on process simulation and isotope tracing, and particularly relates to the technical field of river sewage treatment, and comprises hydrological simulation, nitrogen isotope analysis, nitrogen source analysis, an analysis model, data processing and analysis. The application simulates the source and flux of nitrate nitrogen in the river system by using a hydrological model. When long-term hydrogeological chemical data of the river has been fully measured, standard verification and simulation methods are used. For the river without measurement data, the discharge and nitrate nitrogen concentration data of the field samples are used to help verify the hydrological model, so that the accuracy of simulation is improved. In turn, the hydrological model simulates the river runoff and nitrate nitrogen flux to support the calculation of the isotope source flux. Meanwhile, the hydrological model can also be used to verify the isotope-based results. The simulation results of the hydrological model and the isotope model are systematically integrated to realize reliable nitrate nitrogen space-time source tracing.

Description

Technical Field

[0001] This invention relates to the field of river wastewater treatment technology, and more specifically, to a method for analyzing river nitrogen pollution sources based on process simulation and isotope tracing. Background Technology

[0002] Nitrate nitrogen, as the most significant component of river nitrogen pollution, is highly mobile and widely distributed, making it a focal point for tracing river nitrogen pollution sources. Isotope mixing and hydrological models are effective and practical tools for quantifying the sources of nitrate nitrogen in rivers; however, both methods have significant limitations. Isotope-based nitrate nitrogen pollution source tracing methods primarily rely on the isotopic composition of different nitrogen pollution sources for nitrate nitrogen source identification. Their accuracy depends on the spatiotemporal resolution of isotopic and geochemical monitoring, as well as well-constrained isotopic inclusions. However, due to the overlapping characteristics and wide isotopic range of inclusions, accurately constraining the direction of inclusions... The challenges involved in identifying nitrate nitrogen sources are significant, leading to considerable uncertainty. When two isotopic inclusions are similar, it is difficult for models to accurately quantify their contributions. Hydrological models, such as SWAT, HSPF, and SPARROW models, can simulate the generation, migration, and transformation of nitrogen within a watershed based on parameters such as land use type, climate conditions, agricultural management practices, and soil type. The simulation results of hydrological models are significant in terms of temporal and spatial distribution, but they are also subject to uncertainty due to factors such as a lack of long-term data, subjective parameter settings, and omitted parameters (such as biological nitrate nitrogen removal).

[0003] Given that there is currently no single method, including isotope mixing models and hydrological models, that can accurately quantify the sources of nitrate nitrogen in rivers, it is impossible to achieve robust and accurate nitrate nitrogen source tracing in river systems, thus failing to achieve a methodological breakthrough in river nitrogen pollution research. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a method for apportioning river nitrogen pollution sources based on process simulation and isotope tracing. This method utilizes a hydrological model to simulate the sources and fluxes of nitrate nitrogen in a river system. When long-term hydrogeochemical data of the river has been adequately measured, standard verification and simulation methods are employed. For rivers without measurement data, field sample discharge and nitrate nitrogen concentration data are used to help verify the hydrological model, thereby improving the accuracy of the simulation. Conversely, the hydrological model simulates river runoff and nitrate nitrogen flux to support the calculation of isotope source fluxes. Simultaneously, the hydrological model can also be used to verify isotope-based results. The simulation results of the hydrological model and the isotope model are systematically integrated to achieve reliable spatiotemporal source tracing of nitrate nitrogen. By integrating the advantages of the isotope model and the hydrological model, the problems mentioned in the background art are solved.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for apportioning river nitrogen pollution sources based on process simulation and isotope tracing, comprising the following steps:

[0006] Hydrological simulation: Establish a river hydrological model to simulate the changes in river flow and nitrogen concentration over time. The changes in water flow are predicted by precipitation and evaporation.

[0007] Nitrogen isotope analysis: By measuring the isotope ratio of nitrogen in water, the characteristics of different pollution sources are identified and their contribution ratio is calculated;

[0008] Nitrogen source apportionment: Combining hydrological simulation and isotope analysis results, a multi-source apportionment model was used to calculate the contribution ratio of each pollution source to nitrogen concentration;

[0009] Analytical model: The contribution ratio of each pollution source is optimized through an optimization algorithm to accurately analyze the contribution of different nitrogen pollution sources;

[0010] Data processing and analysis: Combining hydrological data such as flow, precipitation, and isotope data, the data is preprocessed, cleaned, and denoised to generate accurate nitrogen pollution source analysis results.

[0011] In a preferred embodiment, the water flow calculation formula for hydrological simulation is based on the dynamic changes of hydrological processes, taking into account the effects of precipitation and evaporation, and is obtained through the water conservation formula:

[0012]

[0013] Where Q(t) represents the water flow rate at time t; Q0 represents the initial water flow rate; P(τ) represents the precipitation at time τ; E(τ) represents the evaporation at time τ; and t is time. In addition, the hydrological model includes nitrogen concentration calculation, assuming that the nitrogen concentration exhibits a time-varying relationship with water flow rate and pollution source input, expressed by the following formula:

[0014]

[0015] in, f represents the nitrogen concentration of the i-th pollution source; i Q represents the contribution ratio of the i-th pollution source; i This represents the water flow of each source. Through watershed data collection, the data required for establishing the hydrological model are collected, mainly including land use data, soil data, daily meteorological data, hydrological and water quality station data, sampling point data, crop planting information, etc., to establish the model database. The data are then imported into the model and parametrically processed. Through model calibration and verification, the collected field monitoring data, including hydrological and water quality data, are used to measure and verify the model, and accuracy analysis is performed until the simulated data matches the measured data.

[0016] In a preferred embodiment, nitrogen isotope analysis identifies pollution sources and calculates their contribution ratio by measuring the isotopic ratio of nitrogen in water, expressed by the following formula:

[0017]

[0018] Where, δ 15 N represents the nitrogen isotope ratio; R sample Indicates the isotopic ratio of nitrogen-15 to nitrogen-14 in the water sample; R standard This represents the nitrogen isotope ratio in the standard reference sample; based on the nitrogen isotope ratio δ of different pollution sources. 15 N. Establish a characteristic database of pollution sources, and use this database to perform matching analysis with the measured sample isotope ratios to determine the contribution ratio of each pollution source. The specific isotope source apportionment model is optimized to solve the following problems, and its expression formula is as follows:

[0019]

[0020] in, f represents the contribution ratio of each pollution source. i That is, to find a set of parameters that minimizes the objective function. δ 15 N measured (t) represents the actual isotopic ratio of nitrate nitrogen in the river collected at time t, providing observational data and laying the foundation for pollution source apportionment. f represents the nitrate nitrogen isotope ratio of the i-th pollution source at time t, distinguishing the contribution of each pollution source to nitrogen in the river; i Let i represent the contribution ratio of the i-th pollution source, and calculate the contribution of each pollution source to the total nitrogen pollution of the river; The objective function represents the nitrogen isotope value estimated through simulation, calculated using the model based on the combined isotopic contribution of all pollution sources at time t. The first part of the formula is expressed as:

[0021]

[0022] The formula calculates the squared difference between the observed values ​​and the model-calculated values, and the contribution ratio f of the pollution sources is optimized by minimizing this difference. i To make the model output values ​​closer to the actual observed values, we incorporate a regularization term, the formula of which is:

[0023]

[0024] Where λ is the regularization parameter, which controls the strength of the regularization term; This represents a summation expression; i indicates that 1 is the starting index for the summation; n indicates the ending index for the summation. Nitrogen pollution in rivers typically originates from multiple sources, such as agricultural runoff, domestic sewage, and industrial emissions. Isotope data provides a labeling method based on chemical composition to distinguish these sources. Relying solely on hydrological models is insufficient for source analysis. By introducing isotope labeling, the specific contribution of each source to nitrate nitrogen pollution can be estimated more accurately. In practical applications, this formula is part of a minimization problem used to optimize the contribution ratio of each source based on observational data and source characteristic data. By introducing a regularization term, the robustness of the optimization process can be ensured, and the model can be prevented from being overly sensitive to sources. The contributions of different sources vary with time and location. The time parameter t in the formula allows for independent estimation at each time point, which helps to establish a more dynamic and comprehensive source distribution model.

[0025] In a preferred embodiment, nitrogen source apportionment uses a multi-source apportionment model to weighted sum the contribution ratios of multiple pollution sources, expressed by the following formula:

[0026]

[0027] in, f is the nitrogen concentration of the i-th pollution source; i Let f be the contribution ratio of the pollution source; n is the number of pollution sources. The contribution ratio f of each pollution source is calculated using an optimization algorithm based on the sum of squared errors and the likelihood estimation method. i This optimizes the nitrogen source apportionment model, which is expressed as follows:

[0028]

[0029] Among them, C measured (t) represents the measured nitrogen concentration; f i λ represents the contribution ratio of the i-th source; λ is the regularization coefficient.

[0030] In a preferred embodiment, the analytical model optimizes the contribution ratio of each pollution source using an optimization algorithm. A nonlinear optimization algorithm based on gradient descent is employed, with iterative updates using a time step. The formula is as follows:

[0031]

[0032] in, The η represents the contribution ratio of the i-th pollution source in the t-th iteration; η represents the learning rate; and L represents the loss function. The loss function is represented by f. iThe partial derivatives; after setting various parameters such as simulation step size and simulation time, the model is run to simulate the hydrological cycle and nitrogen migration and transformation process in the watershed. The output results of the model are statistically analyzed to obtain the contribution ratio and flux of nitrate nitrogen from different sources.

[0033] In a preferred embodiment, the data processing steps include multi-stage preprocessing, denoising, and standardization of hydrological data, flow rate, precipitation, and nitrogen isotope data. The data preprocessing is performed using the following algorithm, expressed as follows:

[0034] D cleaned =D raw -Median(D raw )

[0035] Among them, D raw D represents the original dataset; cleaned This represents the cleaned dataset, which removes outliers and noise to ensure the quality of the model's input data.

[0036] In a preferred embodiment, the optimization algorithm optimizes the contribution ratio of each pollution source through an objective function, the objective function of which is defined by the following formula:

[0037]

[0038] Where, δ 15 N measured (t) represents the measured isotope ratio; f represents the nitrogen isotope ratio of the i-th source; i To represent the contribution ratio of pollution sources, sampling points were set up, and samples were taken from different pollution sources in the river and watershed according to seasonal variations. The isotopic ratio of nitrate nitrogen was analyzed using denitrification methods, and the concentration of nitrate nitrogen in the river was measured using ion chromatography to support the calibration and validation of the hydrological model. Isotopic calibration was performed on areas with significant denitrification, and data required for mixed model analysis were prepared. The MCMC method was used to analyze the isotopic composition of nitrate nitrogen and estimate the contribution ratio of different pollution sources in different seasons. The contribution ratio of each pollution source was estimated using the MCMC method based on the isotopic composition of nitrate nitrogen, expressed by the following formula:

[0039]

[0040] Where, δ 15 N measured (t) represents the measured nitrogen isotope ratio; f represents the nitrogen isotope ratio of the i-th pollution source; iThe value represents the contribution ratio of the pollution source; λ is the regularization coefficient. Based on the flow monitoring data at the sampling points, the nitrogen flux of each pollution source is calculated. If there is no flow monitoring data at the sampling points, it is estimated using the runoff and total nitrate nitrogen flux simulated by the hydrological model. Based on the flow rate and nitrogen concentration, the nitrate nitrogen flux T from different sources is calculated. i Its formula is:

[0041]

[0042] Where Q(t) is the flow rate, Let T be the nitrogen concentration of the i-th pollution source; i Let be the nitrogen flux from the i-th source; for sampling points without flow monitoring, the total nitrate nitrogen flux is estimated based on the runoff and nitrate nitrogen flux simulated by the hydrological model, expressed by the following formula:

[0043]

[0044] Among them, Q simulated (t) represents the simulated flow rate. The changes in flow rate and nitrogen concentration are calculated through dataset integration and numerical methods. An integral method is used for simulation calculations. The nitrogen isotope ratios of different pollution sources are calculated using the MCMC algorithm to obtain the optimal solution, thereby achieving accurate analysis of pollution sources. Seasonal river sampling is conducted by setting sampling points, and watershed pollution source samples are collected. Isotope analysis is performed using denitrification. Ion chromatography is used to analyze the nitrate nitrogen concentration in the river to determine the nitrate nitrogen removal status and provide support for the calibration and validation of the hydrological model. If significant denitrification is achieved, isotope calibration is performed to prepare for the MCMC mixture model. The MCMC mixture model is used to estimate the contribution ratio of potential nitrate nitrogen sources in each sampling point throughout the four seasons using the isotopic composition of nitrate nitrogen. The nitrate nitrogen flux from different sources is calculated. For sampling points with flow monitoring, flux calculations for different sources are performed based on monitoring data. If there are no monitoring stations, calculations are performed based on the runoff and total nitrate nitrogen flux simulated by the hydrological model.

[0045] The technical effects and advantages of this invention are as follows:

[0046] 1. The source of nitrate nitrogen is estimated using the isotopic MCMC model. At the same time, the source and flux of nitrate nitrogen in the river system are simulated using the hydrological model. When the long-term hydrogeochemical data of the river have been fully measured, standard validation and simulation methods are used. For rivers without measurement data, the discharge and nitrate nitrogen concentration data of field samples are used to help validate the hydrological model, thereby improving the accuracy of the simulation. Conversely, the hydrological model simulates river runoff and nitrate nitrogen flux to support the calculation of isotopic source flux. At the same time, the hydrological model can also be used to validate isotope-based results, especially when isotopic inclusions are not well constrained or overlapped. Finally, the simulation results of the hydrological model and the isotopic model will be systematically integrated to achieve reliable spatiotemporal source tracing of nitrate nitrogen.

[0047] 2. First, collect watershed data to establish the hydrological model, including land use data, soil data, daily meteorological data, hydrological and water quality station data, sampling point data, and crop planting information. Second, establish the model database by importing the data into the model and performing parameterization. Third, calibrate and validate the model based on the collected field monitoring data (including hydrological and water quality data), determine and validate the model, and perform accuracy analysis until the simulated data matches the measured data. This step is the core of hydrological simulation and determines the accuracy of the results. Fourth, run the model. After setting parameters such as simulation step size and simulation time, run the model to simulate the hydrological cycle and nitrogen migration and transformation process within the watershed. Statistically analyze the model's output results to obtain the contribution ratio and flux of nitrate nitrogen from different sources.

[0048] 3. First, set up sampling points, conduct seasonal river sampling, and collect samples from watershed pollution sources. Second, use denitrification methods for isotope analysis and ion chromatography to analyze the nitrate nitrogen concentration in the river, determine the nitrate nitrogen removal status, and provide support for the calibration and validation of the hydrological model. Third, if significant denitrification is found, perform isotope calibration to prepare for the MCMC mixture model. Fourth, use the MCMC mixture model to estimate the contribution ratio of potential nitrate nitrogen sources in each season at each sampling point using the isotopic composition of nitrate nitrogen. Fifth, calculate the nitrate nitrogen flux from different sources. For sampling points with flow monitoring, calculate the flux from different sources based on the monitoring data. If there are no monitoring stations, calculate based on the runoff and total nitrate nitrogen flux simulated by the hydrological model. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the process of the present invention.

[0050] Figure 2 This is a schematic diagram of hydrological simulation and isotopic source analysis of river nitrogen pollutants. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Refer to the instruction manual appendix Figure 1 An embodiment of the present invention provides a method for apportioning river nitrogen pollution sources based on process simulation and isotope tracing, comprising the following steps:

[0053] Hydrological simulation: Establish a river hydrological model to simulate the changes in river flow and nitrogen concentration over time. The changes in water flow are predicted by precipitation and evaporation.

[0054] Nitrogen isotope analysis: By measuring the isotope ratio of nitrogen in water, the characteristics of different pollution sources are identified and their contribution ratio is calculated;

[0055] Nitrogen source apportionment: Combining hydrological simulation and isotope analysis results, a multi-source apportionment model was used to calculate the contribution ratio of each pollution source to nitrogen concentration;

[0056] Analytical model: The contribution ratio of each pollution source is optimized through an optimization algorithm to accurately analyze the contribution of different nitrogen pollution sources;

[0057] Data processing and analysis: Combining hydrological data such as flow, precipitation, and isotope data, the data is preprocessed, cleaned, and denoised to generate accurate nitrogen pollution source analysis results.

[0058] The formula for calculating water flow in hydrological simulation is based on the dynamic changes of hydrological processes, taking into account the effects of precipitation and evaporation, and is obtained through the water conservation formula:

[0059]

[0060] Where Q(t) represents the water flow rate at time t; Q0 represents the initial water flow rate; P(τ) represents the precipitation at time τ; E(τ) represents the evaporation at time τ; and t is time. In addition, the hydrological model includes nitrogen concentration calculation, assuming that the nitrogen concentration exhibits a time-varying relationship with water flow rate and pollution source input, expressed by the following formula:

[0061]

[0062] in, f represents the nitrogen concentration of the i-th pollution source; i Q represents the contribution ratio of the i-th pollution source; iThis represents the water flow of each source. Through watershed data collection, the data required for establishing the hydrological model are collected, mainly including land use data, soil data, daily meteorological data, hydrological and water quality station data, sampling point data, crop planting information, etc., to establish the model database. The data are then imported into the model and parametrically processed. Through model calibration and verification, the collected field monitoring data, including hydrological and water quality data, are used to measure and verify the model, and accuracy analysis is performed until the simulated data matches the measured data.

[0063] Nitrogen isotope analysis identifies pollution sources and calculates their contribution ratio by measuring the isotopic ratio of nitrogen in water. The formula is as follows:

[0064]

[0065] Where, δ 15 N represents the nitrogen isotope ratio; R sample Indicates the isotopic ratio of nitrogen-15 to nitrogen-14 in the water sample; R standard This represents the nitrogen isotope ratio in the standard reference sample; based on the nitrogen isotope ratio δ of different pollution sources. 15 N. Establish a characteristic database of pollution sources, and use this database to perform matching analysis with the measured sample isotope ratios to determine the contribution ratio of each pollution source. The specific isotope source apportionment model is optimized to solve the following problems, and its expression formula is as follows:

[0066]

[0067] in, f represents the contribution ratio of each pollution source. i That is, to find a set of parameters that minimizes the objective function. δ 15 N measured (t) represents the actual isotopic ratio of nitrate nitrogen in the river collected at time t, providing observational data and laying the foundation for pollution source apportionment. f represents the nitrate nitrogen isotope ratio of the i-th pollution source at time t, distinguishing the contribution of each pollution source to nitrogen in the river; i Let i represent the contribution ratio of the i-th pollution source, and calculate the contribution of each pollution source to the total nitrogen pollution of the river; The objective function represents the nitrogen isotope value estimated through simulation, calculated using the model based on the combined isotopic contribution of all pollution sources at time t. The first part of the formula is expressed as:

[0068]

[0069] The formula calculates the squared difference between the observed values ​​and the model-calculated values, and the contribution ratio f of the pollution sources is optimized by minimizing this difference.i To make the model output values ​​closer to the actual observed values, the regularization term is used, and its expression is as follows:

[0070]

[0071] Where λ is the regularization parameter, which controls the strength of the regularization term; The expression represents a summation; i indicates that 1 is the starting index for the summation; n indicates the ending index for the summation. Nitrogen pollution in rivers typically originates from multiple sources, such as agricultural runoff, domestic sewage, and industrial emissions. Isotope data provides a labeling method based on chemical composition to distinguish these sources. Relying solely on hydrological models is insufficient for source analysis. By introducing isotope labeling, the specific contribution of each source to nitrate nitrogen pollution can be estimated more accurately. In practical applications, this formula is part of a minimization problem, used to optimize the contribution ratio of each source based on observational data and source characteristic data. By introducing a regularization term, the robustness of the optimization process can be ensured, and the model can be prevented from being overly sensitive to sources. The contributions of different sources vary with time and location. The time parameter t in the formula allows for independent estimation at each time point, which helps to establish a more dynamic and comprehensive source distribution model.

[0072] Nitrogen source apportionment uses a multi-source apportionment model to weighted sum the contribution ratios of multiple pollution sources. The formula for this is:

[0073]

[0074] in, f is the nitrogen concentration of the i-th pollution source; i Let f be the contribution ratio of the pollution source; n is the number of pollution sources. The contribution ratio f of each pollution source is calculated using an optimization algorithm based on the sum of squared errors and the likelihood estimation method. i This optimizes the nitrogen source apportionment model, which is expressed as follows:

[0075]

[0076] Among them, C measured (t) represents the measured nitrogen concentration; f i λ represents the contribution ratio of the i-th source; λ is the regularization coefficient.

[0077] The analytical model optimizes the contribution ratio of each pollution source using an optimization algorithm. A nonlinear optimization algorithm based on gradient descent is employed, with iterative updates using a time step. The formula is as follows:

[0078]

[0079] in, The η represents the contribution ratio of the i-th pollution source in the t-th iteration; η represents the learning rate; and L represents the loss function. The loss function is represented by f. i The partial derivatives; after setting various parameters such as simulation step size and simulation time, the model is run to simulate the hydrological cycle and nitrogen migration and transformation process in the watershed. The output results of the model are statistically analyzed to obtain the contribution ratio and flux of nitrate nitrogen from different sources.

[0080] The data processing steps include multi-stage preprocessing, denoising, and standardization of hydrological data, flow rate, precipitation, and nitrogen isotope data. Data preprocessing is performed using the following algorithm, expressed as follows:

[0081] D cleaned =D raw -Median(D raw )

[0082] Among them, D raw D represents the original dataset; cleaned This represents the cleaned dataset, which removes outliers and noise to ensure the quality of the model's input data.

[0083] The optimization algorithm optimizes the contribution ratio of each pollution source through an objective function, which is defined by the following formula:

[0084]

[0085] Where, δ 15 N measured (t) represents the measured isotope ratio; f represents the nitrogen isotope ratio of the i-th source; i To represent the contribution ratio of pollution sources, sampling points were set up, and samples were taken from different pollution sources in the river and watershed according to seasonal variations. The isotopic ratio of nitrate nitrogen was analyzed using denitrification methods, and the concentration of nitrate nitrogen in the river was measured using ion chromatography to support the calibration and validation of the hydrological model. Isotopic calibration was performed on areas with significant denitrification, and data required for mixed model analysis were prepared. The MCMC method was used to analyze the isotopic composition of nitrate nitrogen and estimate the contribution ratio of different pollution sources in different seasons. The contribution ratio of each pollution source was estimated using the MCMC method based on the isotopic composition of nitrate nitrogen, expressed by the following formula:

[0086]

[0087] Where, δ 15 N measured (t) represents the measured nitrogen isotope ratio; f represents the nitrogen isotope ratio of the i-th pollution source;i The value represents the contribution ratio of the pollution source; λ is the regularization coefficient. Based on the flow monitoring data at the sampling points, the nitrogen flux of each pollution source is calculated. If there is no flow monitoring data at the sampling points, it is estimated using the runoff and total nitrate nitrogen flux simulated by the hydrological model. Based on the flow rate and nitrogen concentration, the nitrate nitrogen flux T from different sources is calculated. i Its formula is:

[0088]

[0089] Where Q(t) is the flow rate, Let T be the nitrogen concentration of the i-th pollution source; i Let be the nitrogen flux from the i-th source; for sampling points without flow monitoring, the total nitrate nitrogen flux is estimated based on the runoff and nitrate nitrogen flux simulated by the hydrological model, expressed by the following formula:

[0090]

[0091] Among them, Q simulated (t) represents the simulated flow rate. The changes in flow rate and nitrogen concentration are calculated through dataset integration and numerical methods. An integral method is used for simulation calculations. The nitrogen isotope ratios of different pollution sources are calculated using the MCMC algorithm to obtain the optimal solution, thereby achieving accurate analysis of pollution sources. Seasonal river sampling is conducted by setting sampling points, and watershed pollution source samples are collected. Isotope analysis is performed using denitrification methods, and ion chromatography is used to analyze the nitrate nitrogen concentration in the river to determine the nitrate nitrogen removal status and provide support for the calibration and validation of the hydrological model. If significant denitrification is achieved, isotope calibration is performed to prepare for the MCMC mixture model. The MCMC mixture model is used to estimate the contribution ratio of potential nitrate nitrogen sources in each season at each sampling point using the isotopic composition of nitrate nitrogen, and to calculate the nitrate nitrogen flux from different sources. For sampling points with flow monitoring, the analysis is based on the monitoring data. Flux calculations are performed from different sources. If no monitoring stations are available, the calculations are based on the runoff and total nitrate nitrogen flux simulated by the hydrological model. Field geochemical monitoring data of nitrate nitrogen from the isotope mixing method are also applied to the calibration or validation of the hydrological model, providing effective support for the hydrological model to simulate nitrate nitrogen sources. This solves the problem that hydrological models cannot be carried out due to a lack of long-term monitoring data or the lack of necessary indicators in the monitoring data. The isotope mixing model relies on field sampling data from the watershed, which can more realistically reflect the actual situation of the watershed. This technique compares the field monitoring and source allocation results from isotope studies with the simulation results of the hydrological model, which helps to verify the accuracy of the hydrological model. This method solves the problem that the model may deviate from the actual situation due to subjective parameter settings and simplification of macroscopic data, but verification is difficult, thereby improving the credibility of the model.

[0092] Assessing and even quantifying the removal of nitrate nitrogen in a watershed using isotopes can provide an important supplement to hydrological model simulations, thus addressing the uncertainty arising from the incomplete consideration of processes such as biological nitrate removal in hydrological models. To reflect nitrate nitrogen loads at different watershed extents, isotope mixing models require sampling at multiple sampling points. In the absence of monitoring stations, the instantaneous outflow at the sampling point is typically used to represent the average flow for the entire month. This method may lead to errors in flux results due to the variability of flow. This technique utilizes hydrological models, enabling simulations even without hydrological measurement data. In cases of insufficient data, this technique provides a reference for river discharge and nitrogen flux data for isotope mixing, and verifies the representativeness of field samples with the support of hydrological model simulation results. This addresses the problem that isotope mixing is difficult to apply in long-term monitoring due to its high economic costs in terms of manpower and resources. The accuracy of isotope mixing models depends on the spatiotemporal resolution of isotope and geochemical monitoring and well-constrained isotopic inclusions. However, when two isotopic inclusions are similar, it is difficult for isotope methods to accurately allocate their contributions. Furthermore, spontaneous biotransformation of samples during sampling can affect the results. This technique utilizes hydrological model simulation results to distinguish nitrate nitrogen fluxes from different land use types. Therefore, it can provide a reliable nitrate nitrogen source for isotope mixing even when the nitrate nitrogen isotopic composition is very similar. This solves the uncertainty that may arise in quantifying the source due to the overlapping characteristics of inclusions, the wide isotopic range, and the potential for bio-driven nitrogen transformation to alter the initial isotopic values ​​of rivers.

[0093] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for apportioning river nitrogen pollution sources based on process simulation and isotope tracing, characterized in that, Includes the following steps: Hydrological simulation: Establish a river hydrological model to simulate the changes in river flow and nitrogen concentration over time. The changes in water flow are predicted by precipitation and evaporation. Nitrogen isotope analysis: By measuring the isotope ratio of nitrogen in water, the characteristics of different pollution sources are identified and their contribution ratio is calculated; Nitrogen source apportionment: Combining hydrological simulation and isotope analysis results, a multi-source apportionment model was used to calculate the contribution ratio of each pollution source to nitrogen concentration; Analytical model: The contribution ratio of each pollution source is optimized through an optimization algorithm to analyze the contribution of different nitrogen pollution sources; Data processing and analysis: Combining hydrological data such as flow, precipitation, and isotope data, the data is preprocessed, cleaned, and denoised to generate accurate nitrogen pollution source analysis results; The formula for calculating water flow in hydrological simulation is based on the dynamic changes of hydrological processes, taking into account the effects of precipitation and evaporation, and is obtained through the water conservation formula: Where Q(t) represents the water flow rate at time t; Q0 represents the initial water flow rate; P(τ) represents the precipitation at time τ; E(τ) represents the evaporation at time τ; and t is time. In addition, the hydrological model includes nitrogen concentration calculation, assuming that the nitrogen concentration exhibits a time-varying relationship with water flow rate and pollution source input, expressed by the following formula: in, f represents the nitrogen concentration of the i-th pollution source; i Q represents the contribution ratio of the i-th pollution source; i This represents the water flow rate of each source; n is the number of pollution sources.

2. The method for apportioning river nitrogen pollution sources based on process simulation and isotope tracing according to claim 1, characterized in that: Nitrogen isotope analysis identifies pollution sources and calculates their contribution ratio by measuring the isotopic ratio of nitrogen in water. The formula is as follows: Where, δ 15 N represents the nitrogen isotope ratio; R sample Indicates the isotopic ratio of nitrogen-15 to nitrogen-14 in the water sample; R standard This represents the nitrogen isotope ratio in the standard reference sample; based on the nitrogen isotope ratio δ of different pollution sources. 15 N. Establish a characteristic database of pollution sources, and use this database to perform matching analysis with the measured sample isotope ratios to determine the contribution ratio of each pollution source. The specific isotope source apportionment model is optimized to solve the following problems, and its expression formula is as follows: in, This represents the contribution ratio f of each pollution source in the optimization process. i ;δ 15 N measured (t) represents the actual isotopic ratio of nitrate nitrogen in the river collected at time t, providing observational data and a basis for pollution source analysis; f represents the nitrate nitrogen isotope ratio of the i-th pollution source at time t, distinguishing the contribution of each pollution source to nitrogen in the river; i Let i represent the contribution ratio of the i-th pollution source, and calculate the contribution of each pollution source to the total nitrogen pollution of the river; The objective function represents the nitrogen isotope value estimated through simulation, calculated using the model based on the combined isotopic contribution of all pollution sources at time t. The first part of the formula is expressed as: The formula calculates the squared difference between the observed and model-calculated values, and uses this difference to optimize the pollution source contribution ratio f. i To make the model output values ​​closer to the actual observed values, the regularization term is used, and its expression is as follows: Where λ is the regularization parameter, which controls the strength of the regularization term; This represents the summation expression; i indicates that 1 is the starting index for the summation; n indicates the ending index for the summation.

3. The method for apportioning river nitrogen pollution sources based on process simulation and isotope tracing according to claim 2, characterized in that: Nitrogen source apportionment uses a multi-source apportionment model to weighted sum the contribution ratios of multiple pollution sources. The formula for this is: Where n is the number of pollution sources, the contribution ratio f of each pollution source is calculated using an optimization algorithm based on the sum of squared errors and the likelihood estimation method. i This optimizes the nitrogen source apportionment model, which is expressed as follows: Among them, C measured (t) represents the measured nitrogen concentration.

4. The method for apportioning river nitrogen pollution sources based on process simulation and isotope tracing as described in claim 3, characterized in that: The analytical model optimizes the contribution ratio of each pollution source using an optimization algorithm. A nonlinear optimization algorithm based on gradient descent is employed, with iterative updates using a time step. The formula is as follows: in, The η represents the contribution ratio of the i-th pollution source in the t-th iteration; η represents the learning rate; and L represents the loss function. The loss function is represented by f i The partial derivatives of .

5. A method for apportioning river nitrogen pollution sources based on process simulation and isotope tracing according to claim 4, characterized in that: Data processing and analysis include multi-stage preprocessing, denoising, and standardization of hydrological data, flow, precipitation, and nitrogen isotope data. Data preprocessing is performed using the following algorithm, expressed as follows: D cleaned =D raw -Median(D raw ) Among them, D raw D represents the original dataset; cleaned This represents the cleaned dataset, which removes outliers and noise to ensure the quality of the model's input data.

6. A method for apportioning river nitrogen pollution sources based on process simulation and isotope tracing according to claim 5, characterized in that: The optimization algorithm optimizes the contribution ratio of each pollution source through an objective function, which is defined by the following formula: This process involved setting up sampling points and sampling different pollution sources within the river and watershed according to seasonal variations. Denitrification was used to analyze the isotopic ratio of nitrate nitrogen, and ion chromatography was used to determine the nitrate nitrogen concentration in the river to support the calibration and validation of the hydrological model. Isotopic calibration was performed on areas with significant denitrification, and data required for mixed model analysis were prepared. The MCMC method was used to analyze the isotopic composition of nitrate nitrogen and estimate the contribution ratio of different pollution sources in different seasons. The MCMC method uses the isotopic composition of nitrate nitrogen to estimate the contribution ratio of each pollution source, expressed by the following formula: Based on flow monitoring data from sampling points, nitrogen flux from each pollution source is calculated. If flow monitoring data is unavailable at sampling points, it is estimated using runoff and total nitrate nitrogen flux simulated by a hydrological model. Based on flow rate and nitrogen concentration, the nitrate nitrogen flux T from different sources is calculated. i Its formula is: Where T i Let be the nitrogen flux from the i-th source. For sampling points without flow monitoring, the flux is estimated based on the runoff and total nitrate nitrogen flux simulated by the hydrological model. The formula is as follows: Among them, Q simulated (t) represents the simulated flow rate, and the changes in flow rate and nitrogen concentration are calculated through dataset integration and numerical methods.

Images