Assessment method of agricultural non-point source pollution in the intersection area based on coupling of CFD and SWAT models

Abstract

The present invention discloses an assessment method for agricultural non-point source pollution in a confluence area based on the coupling of CFD and SWAT models: S1, assess the type of agricultural non-point source pollution in the confluence area, construct a SWAT model, calibrate model parameters, and calculate the temporal and spatial distribution of typical agricultural non-point source pollution in the target confluence area basin; S2, assess the inlet flow and pollutant load of the main and tributary streams in the target confluence area according to the simulation results; S3, establish a two-dimensional hydrodynamic water quality model, perform numerical simulation, and obtain the distribution characteristics and transportation conditions of pollutants in the target confluence area; S4, construct an integrated model to assess agricultural non-point source pollution. The assessment method for agricultural non-point source pollution in a confluence area based on the coupling of CFD and SWAT models provided by the present invention can efficiently study the distribution and transportation laws of pollutants in the confluence area, and clarify the source and destination of pollutants in the confluence area, which solves the problem that traditional research methods consume a lot of manpower and material resources and reduce the cost of water environment management in the basin.

Description

Technical Field

[0001] The invention belongs to the field of river water environment management, and in particular relates to an assessment method for agricultural non-point source pollution in a confluence area based on coupling of CFD and SWAT models. Background Art

[0002] The forms of water pollution mainly include point source pollution and non-point source pollution. Point source pollution refers to water pollution caused by pollutants discharged from fixed discharge points, while non-point source pollution refers to water pollution caused by pollutants discharged from non-specific areas entering receiving water bodies through surface runoff, farmland drainage and underground seepage.

[0003] The phenomenon of water confluence is widely present in the natural runoff process. The confluence of main and tributary rivers carrying pollutants of different concentrations causes a sudden change in the spatial and temporal distribution of pollutants at the confluence due to the effects of top support and mixing, which directly affects the migration and transformation of pollutants in the river network. If the distribution and transportation of pollutants in the river confluence area are clear, the pollution of the downstream and even the entire basin can be judged. However, there are few hydrological and water quality stations set up at the confluence, and it is difficult to directly obtain the hydrological and water quality data of the confluence area. At present, the research on pollutants in the confluence area mainly adopts a combination of field measurement and numerical simulation. Monitoring sections and test points are arranged in the confluence area, and the pollutants and sections measured at each measuring point are obtained by manual collection or unmanned boats, and then the water quality of the confluence area is simulated. The results obtained by this technical method are relatively accurate, but it will consume a lot of time and energy. If the basin hydrological and water quality model is coupled with the hydrodynamic and water quality model of the confluence area, it may be possible to improve the efficiency of the research while maintaining the accuracy of the research. Summary of the invention

[0004] The purpose of the present invention is to provide an assessment method for agricultural non-point source pollution in the intersection area based on the coupling of CFD and SWAT models, which can accurately and efficiently study the distribution and transport laws of pollutants in the intersection area, and clarify the sources and destinations of pollutants in the intersection area, thus solving the problem that traditional research methods consume a lot of manpower and material resources.

[0005] The technical solution adopted by the present invention is: an assessment method for agricultural non-point source pollution in the intersection area based on the coupling of CFD and SWAT models, which is specifically implemented according to the following steps:

[0006] Step 1: Evaluate the typical agricultural non-point source pollution types in the target intersection area, construct a SWAT model, calibrate the SWAT model parameters, and calculate the temporal and spatial distribution of typical agricultural non-point source pollution in the target intersection area;

[0007] Step 2: Based on the simulation results of step 1, the natural breakpoint classification method is used to evaluate the inlet flow and pollutant load of the main and tributary streams in the target confluence area in different hydrological periods;

[0008] Step 3: Establish a two-dimensional hydrodynamic water quality model of the target intersection area, compile UDF, and use the finite volume method to perform numerical simulation to obtain the distribution characteristics and transport conditions of pollutants in the target intersection area;

[0009] Step 4: Use step 2 as an interface to connect step 1 and step 3, build an integrated model, and evaluate agricultural non-point source pollution based on the model.

[0010] The present invention is also characterized in that:

[0011] In step 1, field surveys and online questionnaires were used to obtain information such as farmland farming patterns and fertilizer application amounts in the target confluence area, and then the typical agricultural non-point source pollution types in the target confluence area were evaluated.

[0012] The basic data required to construct the SWAT model in step 1 include the digital elevation model of the watershed, the spatial distribution of land use types, the spatial distribution of soil types, the physical and chemical property data of soil, meteorological data, pollution source data, and hydrological and water quality data.

[0013] The digital elevation model of the watershed was obtained through hydrological analysis using ArcGIS, which was used to generate river networks and divide sub-watersheds. The spatial distribution of land use types and soil types was obtained through ArcGIS masking and reclassification. The physical and chemical property data of soil were obtained by searching the HWSD database, calculating with SPAW software and formulas, and then written into the SWAT database. Meteorological data included precipitation, temperature, wind speed, relative humidity and solar radiation. Daily measured meteorological data of at least 20 years at the meteorological stations in the study area were used, and the parameters of the weather generator were calculated using the Swat Weather program. The parameters were written into the SWAT database and an index table was created. Hydrological and water quality data required the use of monthly measured hydrological and water quality data of 5 years or more from the hydrological stations and water quality stations in the study area. Pollution source data were agricultural management measures in the study area, including crop cultivation patterns, types of crops planted, time for planting and harvesting crops, and types and time of fertilization.

[0014] Step 1: Construct the SWAT model. The specific steps are as follows:

[0015] Step 1.1: Automatically divide the sub-basins and their outlets based on the digital elevation model of the watershed and the river network map, and then manually add an outlet for the main stream and tributary at a location 5-10 km upstream of the confluence to obtain the boundary conditions for the pollutant transport study in the target confluence area. At the same time, manually add outlets at the spatial locations of the hydrological station and the water quality station for subsequent model verification;

[0016] Step 1.2, overlay the spatial database in the same projection coordinate system and divide it into hydrological response units according to the attribute database. The hydrological response unit is a combination of hydrological characteristics such as the same flow, the same flow velocity, and the same sediment content;

[0017] Step 1.3: Write the time series data of the spatial meteorological stations in the study area, set agricultural management measures, and complete the model construction;

[0018] Step 1.4: Set a warm-up period of two years or more and run the SWAT model.

[0019] The SWAT model parameter calibration items in step 1 include runoff and pollutant load, which are specifically implemented as follows:

[0020] The model parameters were selected based on the actual conditions of the target confluence area, and the SUFI2 algorithm in the SWAT-CUP2012 software was used for calibration. The correlation coefficient R between the model simulation value and the measured value was used. 2 and Nash coefficient NSE as evaluation criteria, in the simulation based on the monthly scale, when R 2 ≥0.6, NSE ≥0.5, the simulation results are considered feasible, and the correlation coefficient R 2 The calculation formula is as follows:

[0021]

[0022] In formula 1: Q i and P i are the ith measured value and simulated value, respectively. and are the measured average and simulated average of the simulation time period, respectively;

[0023] The calculation formula of Nash coefficient NSE is as follows:

[0024]

[0025] In formula 2: Q i and P i are the ith measured value and simulated value, respectively. and are the measured average and simulated average of the simulation time period, respectively;

[0026] After the parameter calibration is completed, the established SWAT model is substituted and run again to obtain the temporal and spatial distribution of typical agricultural non-point source pollution in the target confluence area.

[0027] Step 2 can be implemented by calling the jenkspy package in Python.

[0028] Step 3 The specific steps are as follows:

[0029] Step 3.1, obtaining a mathematical model of a two-dimensional natural river channel through satellite image extraction of the target confluence area;

[0030] Step 3.2, use ICEM software to divide the grid, select unstructured grid in ICEM software, and select All Tri as the grid type. The river channel contour of natural rivers is complex and irregular. Using unstructured grid and All Tri can get better results. After dividing the grid, check the grid quality and control the grid quality to be greater than 0.3. If the grid quality is unqualified, further optimize the grid until it is qualified.

[0031] Step 3.3, FLUENT software is used for calculation. The turbulence model in the software uses the RNG k-ε model. The water quality model is realized by compiling UDF using Visual Studio software and then loading it into FLUENT software. The conditions set during calculation are ammonia nitrogen conditions. In FLUENT software, the model boundary conditions need to be set. The inlet boundary conditions of the intersection area include mainstream flow, mainstream pollutant concentration, tributary flow, and tributary pollutant concentration, which are obtained through step 2. The wall boundary condition sets the river roughness to 0.03-0.05m, and the outlet boundary condition is set to atmospheric pressure.

[0032] Step 3.4: Set the convergence residual to 10 in FLUENT software. -5 , run FLUENT software, perform calculations, and obtain the pollutant distribution characteristics and transport conditions in the target intersection area.

[0033] Step 4 is as follows: Use step 2 as the interface to connect step 1 and step 3, and build an integrated model by integrating the spatiotemporal distribution of typical agricultural non-point source pollution in the confluence area and the distribution characteristics and transport conditions of pollutants in the confluence area.

[0034] The beneficial effects of the present invention are as follows: the present invention is based on the intersection area agricultural non-point source pollution assessment method coupled with CFD and SWAT models, takes the SWAT model simulation results verified by long-term series measured hydrological and water quality data as an interface, combines the SWAT model with CFD, and constructs an integrated model, thereby efficiently and accurately simulating the pollutant distribution and transport rules in the intersection area, so as to judge the incoming flow pollution downstream and even the entire river basin. Compared with the traditional method of obtaining flow value and pollutant value by field measurement, the present invention can effectively save manpower and material resources and improve the efficiency of river basin water environment management. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of the method for assessing agricultural non-point source pollution in the intersection area based on the coupling of CFD and SWAT models of the present invention;

[0036] Figure 2 It is a schematic diagram of the sub-basin division and hydrological and water quality stations of the SWAT model in the study area in the embodiment;

[0037] Figure 3Schematic diagram of the fitting results of the simulated value and the measured value of the runoff and ammonia nitrogen load at each station in the embodiment;

[0038] Figure 4 is a simulation result diagram of ammonia nitrogen load in the study area in the embodiment;

[0039] Figure 5 is the monthly value of ammonia nitrogen load at the main and branch stream inlets from 2005 to 2019 in the embodiment;

[0040] Figure 6 is the monthly value of the inlet flow of the main and tributary rivers from 2005 to 2019 in the embodiment;

[0041] Figure 7 Schematic diagram of mesh division of the model in the embodiment;

[0042] Figure 8 is a velocity field distribution diagram of the intersection area of ​​the calculation results in the embodiment;

[0043] Fig. 9 1 is a distribution diagram of ammonia nitrogen concentration in the intersection area of ​​Jingwei according to the calculation results in the embodiment. DETAILED DESCRIPTION

[0044] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0045] The present invention is based on the intersection area agricultural non-point source pollution assessment method coupled with CFD and SWAT models, such as Figure 1 As shown, the specific implementation steps are as follows:

[0046] Step 1: Select the target intersection area, obtain information such as farmland farming patterns and fertilizer application through field surveys and online questionnaires, and then evaluate its typical agricultural non-point source pollution types and build a SWAT model. The basic data required for building the SWAT model include the digital elevation model of the watershed, the spatial distribution of land use types, the spatial distribution of soil types, soil physical and chemical property data, meteorological data, pollution source data, and hydrological and water quality data. The digital elevation model of the watershed is obtained by hydrological analysis through ArcGIS, which is used to generate river networks and divide sub-watersheds; the spatial distribution of land use types and soil types is obtained by ArcGIS masking and reclassification; the soil physical and chemical property data is obtained by searching the HWSD database, SPAW software calculation and formula calculation, and written into the SWAT database; the meteorological data includes precipitation, temperature, wind speed, relative humidity and solar radiation. It is necessary to select at least 20 years of daily measured meteorological data from meteorological stations in the study area and use Swat The Weather program calculates the parameters of the weather generator, writes the parameters into the SWAT database and creates an index table; the hydrological and water quality data must be selected from the monthly measured hydrological and water quality data of the hydrological stations and water quality stations in the study area for 5 years or more; the pollution source data are the agricultural management measures in the study area, including crop farming patterns, types of crops planted, time for planting and harvesting crops, types and time of fertilization, etc.

[0047] The specific steps to build the SWAT model are as follows:

[0048] Step 1.1: Automatically divide the sub-basins and their outlets based on the digital elevation model of the watershed and the river network map, and then manually add an outlet for the main stream and tributary at a location 5-10 km upstream of the confluence to obtain the boundary conditions for the pollutant transport study in the target confluence area. At the same time, manually add outlets at the spatial locations of the hydrological station and the water quality station for subsequent model verification;

[0049] Step 1.2, overlay the spatial database in the same projection coordinate system, and divide it into hydrological response units according to the attribute database, wherein the hydrological response unit is a combination of hydrological characteristics such as the same flow, the same flow velocity, and the same sediment content;

[0050] Step 1.3: Write the time series data of the spatial meteorological stations in the study area, set agricultural management measures, and complete the model construction;

[0051] Step 1.4: Set a warm-up period of two years or more and run the SWAT model.

[0052] The SWAT model parameters were calibrated to calculate the temporal and spatial distribution of typical agricultural non-point source pollution in the target confluence area. The SWAT model parameter calibration items included runoff and pollutant load. The specific implementation was as follows: the model parameters were selected based on the actual situation of the target confluence area, and the SUFI2 algorithm in the SWAT-CUP2012 software was used for calibration. The correlation coefficient R between the model simulation value and the measured value was used. 2 and Nash coefficient NSE as evaluation criteria, in the simulation based on the monthly scale, when R 2 ≥0.6, NSE ≥0.5, the simulation results are considered feasible, and the correlation coefficient R 2 The calculation formula is as follows:

[0053]

[0054] In formula 1: Q i and P i are the ith measured value and simulated value, respectively, and are the measured average and simulated average of the simulation time period, respectively;

[0055] The calculation formula of Nash coefficient NSE is as follows:

[0056]

[0057] In formula 2: Q i and P i are the ith measured value and simulated value, respectively. is the measured average value during the simulation period;

[0058] After the parameter calibration is completed, the established SWAT model is substituted and run again to obtain the temporal and spatial distribution of typical agricultural non-point source pollution in the target confluence area.

[0059] Step 2: Based on the simulation results of step 1 (spatial and temporal distribution data of typical agricultural non-point source pollution in the target confluence area), the monthly flow and pollutant values ​​of the long-term series at the inlet of the main and tributary rivers in the confluence area are obtained. The natural breakpoint classification method is used to evaluate the flow and pollutant load at the inlet of the main and tributary rivers in the target confluence area in different hydrological periods, i.e., the flood season, the normal water season and the dry season. This can be achieved by calling the jenkspy package in Python.

[0060] Step 3: Establish a two-dimensional hydrodynamic water quality model of the target intersection area, compile UDF, and use the finite volume method to perform numerical simulation to obtain the distribution characteristics and transport conditions of pollutants in the target intersection area;

[0061] The specific steps are as follows:

[0062] Step 3.1, obtaining a mathematical model of a two-dimensional natural river channel through satellite image extraction of the target confluence area;

[0063] Step 3.2, use ICEM software to divide the grid, select unstructured grid in ICEM software, and select All Tri as the grid type. The river channel contour of natural rivers is complex and irregular. Using unstructured grid and All Tri can get better results. After dividing the grid, check the grid quality and control the grid quality to be greater than 0.3. If the grid quality is unqualified, further optimize the grid until it is qualified.

[0064] Step 3.3, FLUENT software is used for calculation. The turbulence model in the software uses the RNG k-ε model. The water quality model is realized by compiling UDF using Visual Studio software and then loading it into FLUENT software. The conditions are set as ammonia nitrogen conditions during calculation. The transport equation of ammonia nitrogen is:

[0065]

[0066] Where C is the saturated ammonia nitrogen concentration, mg / L; D is the molecular diffusion coefficient, m 2 / s;S c is the Schmidt number, which is 1; S is the source term, mg / (L·s), P is pressure, P = 1000-0.019549(t-4.0) 1.68 , t is temperature, ℃; K is the comprehensive degradation coefficient of ammonia nitrogen, K = K 20 θ t-20 , K 20 =0.08d -1 , θ=1.083; C is the saturated ammonia nitrogen concentration, mg / L.

[0067] In FLUENT software, the model boundary conditions need to be set. The boundary conditions at the inlet of the intersection area include the mainstream flow, mainstream pollutant concentration, tributary flow, and tributary pollutant concentration, which are obtained through step 2. The wall boundary condition sets the river channel roughness to 0.03-0.05m, and the outlet boundary condition is set to atmospheric pressure.

[0068] Step 3.4: Set the convergence residual to 10 in FLUENT software. -5 , run FLUENT software, perform calculations, and obtain the pollutant distribution characteristics and transport conditions in the target intersection area.

[0069] Step 4, use step 2 as the interface to connect step 1 and step 3 to construct an integrated model, as follows: Use step 2 as the interface to connect step 1 and step 3, comprehensively consider the spatiotemporal distribution of typical agricultural non-point source pollution in the confluence area and the distribution characteristics and transport conditions of pollutants in the confluence area, construct an "integrated" model of the sources and destinations of agricultural non-point source pollution in the confluence area - overall and local - time and space, and evaluate the agricultural non-point source pollution situation based on the model.

[0070] Example

[0071] The Wei River is the largest tributary of the Yellow River, and the Jing River is the largest tributary of the Wei River. The confluence of the two is located in the Guanzhong Plain, one of the main grain-producing areas in the Yellow River Basin. The Yellow River Basin is an important economic zone in my country. As its main grain-producing area, the Guanzhong Plain applies a large amount of fertilizers in agricultural production to achieve high yields, resulting in serious agricultural non-point source pollution in the region. After years of governance, the water quality there has improved significantly, but it is still not optimistic. Through field research and online questionnaires, the typical agricultural non-point source pollution in the basin is ammonia nitrogen. Therefore, this embodiment selects the intersection of the Jing River and the Wei River as the target object and ammonia nitrogen as the target agricultural non-point source pollutant to fully understand the agricultural non-point source pollution in the Jingwei intersection area and provide certain theoretical references and data support for its water quality improvement. The specific operation steps are as follows:

[0072] Step 1. Construct the SWAT model of the Jingwei confluence area: the digital elevation model is obtained by hydrological analysis through ArcGIS; the spatial distribution of land use and soil types is obtained by ArcGIS masking and reclassification; the soil physical and chemical property data are obtained by searching the HWSD database, SPAW software calculation and formula calculation, and written into the SWAT database; meteorological data include precipitation, temperature, wind speed, relative humidity and solar radiation. The long-term series of daily measured meteorological data from 25 meteorological stations in the Jingwei confluence area from 2000 to 2019 are selected, and the parameters of the weather generator are calculated using the Swat Weather program and written into the SWAT database; the hydrological data are the measured monthly runoff data of 7 hydrological stations in the confluence area, and the water quality data are the monthly monitoring data of ammonia nitrogen concentration of 1 water quality station in the confluence area; the pollution source data are estimated through field surveys and online questionnaire results;

[0073] The digital elevation model was added to ArcGIS software, and the SWAT module was called to automatically divide the sub-basins and their outlets. An outlet was manually added for the main stream and tributary at a location 5 km upstream of the confluence to obtain the boundary conditions for the study of pollutant transport in the confluence area. The sub-basin division results are shown in Figure 2. Figure 2As shown, in addition, the locations of the hydrological stations and water quality stations should also be manually added as exports for subsequent model verification. Then the spatial database is overlaid in the "WGS_1984_UTM_Zone_49N" projection coordinate system, and divided into combinations with the same hydrological characteristics such as flow, velocity, and sediment content based on factors such as slope, land use, and soil characteristics, namely, hydrological response units. Then write the time series data of the spatial meteorological station, set agricultural management measures, and complete the model construction. Finally, set the warm-up period to 5 years and run the SWAT model.

[0074] The runoff values ​​of seven hydrological stations (Wushan, Linjiacun, Xianyang, Jiaqiao, Jingchuan, Jingcun, and Zhangjiashan) and the ammonia nitrogen load value of one water quality station (Vineyard (Niubei)) were selected to calibrate the model. The locations of the hydrological and water quality stations are shown in Figure 2. Figure 2 As shown, each time a station is calibrated, the upstream station is calibrated first. After the parameters of the upstream station are fixed, the downstream station is calibrated. After the runoff calibration is completed, the parameters are brought back to the SWAT model, and after re-running, the ammonia nitrogen is calibrated again until the R of each station is 2 If both R and NSE meet the evaluation criteria, the model calibration is completed. In this example, the minimum NSE of the runoff simulation rate period at each station is 0.52 and the maximum is 0.82; the minimum NSE of the validation period is 0.57 and the maximum is 0.88; the R 2 The minimum is 0.60 and the maximum is 0.83; the validation period R 2 The minimum is 0.65 and the maximum is 0.96, and the results meet the simulation requirements.

[0075] The above data show that the constructed SWAT model of the Jingwei confluence area can better reflect the hydrological characteristics of the basin. The NSE of the ammonia nitrogen simulation results in the regular and validation periods are 0.51 and 0.53, respectively, and the R 2 are 0.60 and 0.93 respectively. The results meet the simulation requirements, indicating that the constructed SWAT model for the confluence of Jing and Wei can better reflect the water quality characteristics of the basin. The comparison results between the simulated values ​​and measured values ​​of model runoff and ammonia nitrogen at each station in this example are as follows Figure 3 As shown in the figure, the fluctuation trends of the simulated values ​​and the measured values ​​of the model are generally consistent, and the peak fitting results are good. The overall results show that the SWAT model can be well applied to the simulation study of ammonia nitrogen pollution in this study area.

[0076] In the SWAT model, the calibrated parameters were used to replace the original parameters of the model, and the SWAT model was run again to obtain the simulation results of agricultural non-point source pollution in the Jingwei intersection area. Figure 4 In the simulation results, find the continuous monthly values ​​of the output of the sub-basin and its upstream sub-basin in the confluence area of ​​the main and tributary rivers, and calculate the flow value at the entrance of the main and tributary rivers (such as Figure 5 ) and pollutant values ​​(as shown Figure 6 shown).

[0077] Step 2: Based on the simulation results of step 1, the monthly flow and pollutant values ​​of the main and tributary inlets in the confluence area from 2005 to 2019 were obtained. The natural breakpoint classification method was used to obtain the main stream inlet flow rates of 46.49, 165.37, and 341.17 m in the dry season, normal season, and flood season, respectively. 3 / s, and the flow rates at the tributary inlets are 15.41, 36.63, and 74.45 m 3 / s, the ammonia nitrogen concentrations at the main stream inlet were 0.48, 2.19 and 4.70 mg / L respectively, and the ammonia nitrogen concentrations at the tributary inlet were 0.06, 0.23 and 0.45 mg / L respectively.

[0078] Step 3: Extract the closed surface area (5 km upstream and 10 km downstream of the confluence) from the satellite image of the Jing-Wei confluence as the computational fluid domain and establish a mathematical model. Then use ICEM software to perform meshing, using unstructured grids. Taking the flat water period as an example, a total of 1,542,743 grids are divided, and the grid quality is 0.33-1. The meshing results are shown in Figure 2. Figure 7 shown.

[0079] The flow rate and ammonia nitrogen concentration obtained in step 3 are used as the model boundary conditions. The finite volume method is used for numerical simulation. The FLUENT software is used for calculation. The solver type is set to pressure basis, absolute velocity, steady-state calculation, gravity is not considered, and the turbulence model is selected as the RNG k-ε model. The ammonia nitrogen transport model is constructed, and the ammonia nitrogen source term UDF is loaded. The boundary conditions set the Jinghe inlet as velocity inlet 1 and the Weihe inlet as velocity inlet 2. The velocity directions are perpendicular to the inlet section, the outlet is free outflow, the riverbank contour line is the wall boundary, and the flow near the wall is simulated by the standard wall function. According to the riverbed conditions, the roughness is set to 0.03m, and the solution method is Coupled. The least squares method is used to calculate the gradient at the center of the control body. The pressure interpolation format is Linear, and the discrete formats of the remaining parameters are all second-order upwind formats. The relaxation factor remains the default, and the monitoring convergence residual is set to 1×10 -5 After the calculation is completed, the distribution and transport rules of pollutants in the confluence area of ​​Jing and Wei rivers are obtained. Taking the normal water period as an example, Figure 8 The velocity field distribution in the intersection area of ​​Jing and Wei is shown in Fig. 9 Shown is the ammonia nitrogen concentration distribution field in the river confluence area.

[0080] Step 4: Based on the simulation results of steps 1 and 3, an integrated model is constructed with step 2 as the interface.

[0081] Through the above method, the present invention is applied to the confluence of the Jinghe River and the Weihe River, and the flow and pollutant values ​​at the entrances of the Jinghe River and the Weihe River are obtained by using the SWAT model of the Jingwei confluence area verified based on long-sequence measured hydrological and water quality data, and then the pollutant distribution and transport law of the Jingwei confluence area are calculated by CFD. The present invention solves the drawbacks of traditional research methods that consume a lot of manpower and material resources, and can accurately and efficiently study the distribution and transport law of pollutants in the confluence area, reduce the cost of water environment management in the basin, and effectively improve the efficiency of water environment management of relevant departments.

Claims

1. The agricultural non-point source pollution assessment method in the intersection area based on the coupling of CFD and SWAT models is characterized by: Follow the steps below to implement it: Step 1: Evaluate the typical agricultural non-point source pollution types in the target intersection area, construct a SWAT model, calibrate the SWAT model parameters, and calculate the temporal and spatial distribution of typical agricultural non-point source pollution in the target intersection area; Step 2: Based on the simulation results of step 1, the natural breakpoint classification method is used to evaluate the inlet flow and pollutant load of the main and tributary streams in the target confluence area in different hydrological periods; Step 3: Establish a two-dimensional hydrodynamic water quality model of the target intersection area, compile UDF, and use the finite volume method to perform numerical simulation to obtain the distribution characteristics and transport conditions of pollutants in the target intersection area; Step 4: Use step 2 as an interface to connect step 1 and step 3, build an integrated model, and evaluate the agricultural non-point source pollution according to the model; The specific steps of step 1 to construct the SWAT model are as follows: Step 1.1: Automatically divide the sub-basins and their outlets based on the digital elevation model of the watershed and the river network map, and then manually add an outlet for the main stream and tributary at a location 5-10 km upstream of the confluence to obtain the boundary conditions for the pollutant transport study in the target confluence area. At the same time, manually add outlets at the spatial locations of the hydrological station and the water quality station for subsequent model verification; Step 1.2, overlay the spatial database in the same projection coordinate system, and divide it into hydrological response units according to the attribute database, wherein the hydrological response unit is a combination of hydrological characteristics such as the same flow, the same flow velocity, and the same sediment content; Step 1.3: Write the time series data of the spatial meteorological stations in the study area, set agricultural management measures, and complete the model construction; Step 1.4: Set a warm-up period of two years or more and run the SWAT model; The SWAT model parameter calibration items in step 1 include runoff and pollutant load, which are specifically implemented as follows: The model parameters were selected based on the actual conditions of the target confluence area, and the SUFI2 algorithm in the SWAT-CUP2012 software was used for calibration. The correlation coefficient R² and the Nash coefficient NSE between the model simulation value and the measured value were used as evaluation criteria. In the simulation based on the monthly scale, when R 2 When ≥0.6 and NSE ≥0.5, the simulation results are considered feasible. After the parameter calibration is completed, they are substituted into the established SWAT model and run again to obtain the temporal and spatial distribution of typical agricultural non-point source pollution in the target confluence area. The specific steps of step 3 are as follows: Step 3.1, obtaining a mathematical model of a two-dimensional natural river channel by extracting satellite images of the target confluence area; Step 3.2, use ICEM software to divide the grid, select unstructured grid in ICEM software, and select AllTri as the grid type. The river channel contour of natural rivers is complex and irregular. Using unstructured grid and All Tri can get better results. After dividing the grid, check the grid quality and control the grid quality to be greater than 0.

3. If the grid quality is unqualified, further optimize the grid until it is qualified. Step 3.3, FLUENT software is used for calculation. The turbulence model in the software uses the RNG k-ε model. The water quality model is realized by compiling UDF using Visual Studio software and then loading it into FLUENT software. The conditions are set as ammonia nitrogen conditions during calculation. The transport equation of ammonia nitrogen is: （3） （4） In the formula, C is the saturated ammonia nitrogen concentration, mg / L; D is the molecular diffusion coefficient, m 2 / s; is the Schmidt number, which is 1; is the source term, , ; P For pressure, , t is temperature, °C; K is the comprehensive degradation coefficient of ammonia nitrogen, , K 20 = 0.08d -1 , θ = 1.083; C is the saturated ammonia nitrogen concentration, mg / L; In FLUENT software, the model boundary conditions need to be set. The boundary conditions at the inlet of the intersection area include the mainstream flow, mainstream pollutant concentration, tributary flow, and tributary pollutant concentration, which are obtained through step 2. The wall boundary condition sets the river channel roughness to 0.03~0.05m, and the outlet boundary condition is set to atmospheric pressure; Step 3.4, set the convergence residual to 10-5 in the FLUENT software, run the FLUENT software, perform calculations, and obtain the pollutant distribution characteristics and transport conditions in the target intersection area; The specific details of step 4 are as follows: use step 2 as an interface to connect step 1 and step 3, and comprehensively consider the spatiotemporal distribution of typical agricultural non-point source pollution in the confluence area and the distribution characteristics and transport conditions of pollutants in the confluence area to construct an integrated model.

2. The method for assessing agricultural non-point source pollution in the intersection area based on coupling of CFD and SWAT models according to claim 1 is characterized in that: In step 1, the information such as the farming patterns and fertilizer application amount of the target confluence area watershed is obtained through field surveys and online questionnaires, and then the typical agricultural non-point source pollution types in the target confluence area are evaluated.

3. The method for assessing agricultural non-point source pollution in the intersection area based on coupling of CFD and SWAT models according to claim 1 is characterized in that: The basic data required for constructing the SWAT model in step 1 include a watershed digital elevation model, spatial distribution of land use types, spatial distribution of soil types, soil physical and chemical property data, meteorological data, pollution source data, and hydrological and water quality data.

4. The method for assessing agricultural non-point source pollution in the intersection area based on coupling of CFD and SWAT models according to claim 3 is characterized in that: The digital elevation model of the watershed is obtained by hydrological analysis through ArcGIS, which is used to generate river networks and divide sub-watersheds; the spatial distribution of land use types and soil types is obtained by ArcGIS masking and reclassification; the soil physical and chemical property data are obtained by searching the HWSD database, SPAW software calculations and formula calculations, and are written into the SWAT database; the meteorological data include precipitation, temperature, wind speed, relative humidity and solar radiation, and it is necessary to use at least 20 years of daily measured meteorological data from meteorological stations in the study area, use the Swat Weather program to calculate the parameters of the weather generator, write each parameter into the SWAT database and create an index table; the hydrological and water quality data must use 5 years or more of monthly measured hydrological and water quality data from hydrological stations and water quality stations in the study area; the pollution source data are agricultural management measures in the study area, including crop farming patterns, types of crops planted, time for planting and harvesting crops, types and time of fertilization, etc.

5. The method for assessing agricultural non-point source pollution in the intersection area based on coupling of CFD and SWAT models according to claim 1 is characterized in that: Step 2 can be implemented by calling the jenkspy package in Python.

Images