Watershed water pollution migration and transformation simulation method and system based on MT-DHM model

Abstract

The invention relates to the technical field of migration simulation, in particular to a drainage basin water pollution migration and transformation simulation method and system based on an MT-DHM model. The method comprises the following steps: carrying out data acquisition on a watershed multi-source space, carrying out data preprocessing, and generating a hydrological response unit standard data set; performing vertical hierarchical calculation based on the hydrological response unit standard data set to obtain land runoff vertical analysis data; carrying out land pollutant migration simulation on the land runoff vertical analysis data, and carrying out real-time correction with an iteration step length of 1 h to obtain a dissolved state / adsorption state load matrix; therefore, by constructing a full-process multi-source heterogeneous data fusion and pollution migration simulation mechanism, the problems that a traditional model is insufficient in data uniformity, low in parameter calibration precision and weak in spatial distribution simulation capability are solved, and the precision, stability and management and control operability of drainage basin pollution load calculation are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of migration simulation, and particularly relates to a watershed water pollution migration and transformation simulation method and system based on an MT-DHM model. BACKGROUND

[0002] Traditional watershed pollution migration models mostly rely on single physical processes or statistical methods, which are difficult to comprehensively capture the migration and transformation mechanism of pollutants in complex hydrological environments, resulting in large deviations in simulation results. Secondly, existing models are generally difficult to effectively fuse multi-source heterogeneous data, especially in terms of spatial heterogeneity and temporal dynamics, and cannot accurately reflect the spatio-temporal variation characteristics of different sub-regions and pollution sources in the watershed. In addition, traditional models often use fixed parameters or static calibration methods, which lack dynamic and multi-objective optimization capabilities, making it difficult to adapt to the nonlinear changes of watershed environmental conditions and pollution processes, resulting in limited prediction accuracy. Furthermore, current simulation methods generally lack detailed differentiation and coupling analysis for the different migration mechanisms of dissolved and adsorbed states of pollutants, which cannot fully reflect the conversion and accumulation processes of pollutants in different media, affecting the mechanism expression ability of the model. Finally, traditional systems lack detailed processing of the load transfer mechanism between multi-branch tree nodes in the multi-level simulation of the dilution, dispersion and attenuation processes of pollutants, and are difficult to accurately reflect the spatial propagation law of pollution load and key influence areas. SUMMARY

[0003] Therefore, it is necessary to provide a watershed water pollution migration and transformation simulation method and system based on an MT-DHM model to solve at least one of the above technical problems.

[0004] To achieve the above-mentioned purpose, a watershed water pollution migration and transformation simulation method based on an MT-DHM model, the method comprising the following steps:

[0005] Step S1: data collection is performed on the multi-source space of the watershed, and data preprocessing is performed to generate a standard data set of hydrological response units;

[0006] Step S2: vertical layering calculation is performed based on the standard data set of hydrological response units to obtain land runoff vertical analysis data; land pollution migration simulation is performed on the land runoff vertical analysis data, and real-time correction with an iteration step of 1h is performed to obtain a dissolved / adsorbed load matrix;

[0007] Step S3: push flow migration calculation is performed on the dissolved / adsorbed load matrix, and dispersion and dilution level joint simulation is performed to obtain watershed dilution model level data; automatic load transfer of multi-branch tree nodes is performed on the watershed dilution model level data to obtain an inter-node attenuation load matrix, wherein the inter-node attenuation load matrix includes an attenuation data sufficient area and an attenuation data deficient area;

[0008] Step S4: Mechanism modeling is performed on the attenuation data sufficient area to obtain pollutant load mechanism result data; statistical modeling is performed on the attenuation data deficient area to obtain pollutant load statistical result data; Bayes fusion is performed on the pollutant load mechanism result data and the pollutant load statistical result data to obtain a mixed load output matrix;

[0009] Step S5: Multi-objective rating is performed on the mixed load output matrix to obtain a pollution contribution rate distribution map; accuracy verification is performed on the pollution contribution rate distribution map to obtain basin water pollution simulation verification data; and a Dianchi basin sediment / nutrient salt qualified report is constructed based on the basin water pollution simulation verification data.

[0010] The beneficial effects of the present application are that: through multi-source spatial data acquisition, DEM terrain data, land use data, soil parameters, rainfall and runoff measured data and other types of heterogeneous information are covered, and through unified projection, time alignment, rasterization and attribute reconstruction and other data cleaning means, a standard data set with consistent structure of hydrological response unit is formed, laying a high-quality data foundation for subsequent analysis. Secondly, through vertical layering calculation, continuous hydrological profile data from vegetation interception, soil infiltration to underground runoff are constructed, combined with measured parameters, pollutant migration simulation and real-time correction are realized, dynamic load calculation with 1 hour as a step is realized, and dissolved / adsorbed pollutant load matrix is obtained. On this basis, through push flow migration and multi-path dispersion dilution model, the transport and attenuation process of pollutants in the spatial dimension is modeled as a multi-branch tree structure, effectively expressing the load transfer and partition attenuation characteristics between upstream and downstream nodes. The modeling of attenuation data is divided into mechanism model and statistical model, and the structured and unstructured area load data are jointly processed through Bayes fusion technology, which significantly improves the robustness and data adaptability of the simulation results. In the model output stage, combined with the historical runoff, sediment, nitrogen and phosphorus measured data, through the Nash efficiency coefficient, multi-objective rating algorithm and other multi-criteria model evaluation methods, the simulation results are spatially distributed and error optimized, so that the pollution contribution rate distribution has high credibility, and then the pollution risk early warning and standard verification evaluation are completed. Finally, the compliance classification output of sediment and nutrient salt indicators at the basin scale is realized, which provides direct data support and decision basis for environmental management and pollution control. Therefore, through the construction of the whole process multi-source heterogeneous data fusion and pollution migration simulation mechanism, the present application solves the problems of insufficient data uniformity, low parameter rating accuracy and weak spatial distribution simulation capability of traditional models, and improves the accuracy, stability and controllability of basin pollution load calculation.

[0011] Preferably, step S1 comprises the following steps:

[0012] Step S11: obtaining a hydrological data set from a hydrological check station;

[0013] Step S12: Extract DEM data from the hydrological data set according to a resolution of 30 m, and perform basin boundary analysis to generate a sub-basin division map;

[0014] Step S13: Extract river network data from the hydrological data set according to river slope, and perform topological relationship analysis to obtain a multi-branch tree node coding table;

[0015] Step S14: Divide the land use / soil type data according to a 1 km grid for the hydrological data set, and perform spatial overlay analysis to obtain an HRU unit attribute matrix;

[0016] Step S15: Perform Thiessen polygon spatial interpolation on the hydrological data set according to an hourly level to generate a sub-basin area rainfall sequence;

[0017] Step S16: Construct soil adsorption parameters for the hydrological data set, and perform laboratory calibration to obtain a watershed adsorption parameter table;

[0018] Step S17: Perform data cleaning and format unification on the sub-basin division map, multi-branch tree node coding table, HRU unit attribute matrix, sub-basin area rainfall sequence, and watershed adsorption parameter table to obtain a standard hydrological response unit data set.

[0019] The application realizes the standardized expression of the basin scale hydrological response unit at the data level by the systematic data hierarchical construction method, and provides high-quality structured input support for subsequent pollution load migration modeling and dynamic simulation. First, the method takes hydrological check sites as the basic data source to obtain time series hydrological observation data, ensuring the spatial representativeness and time continuity of the input data. In the DEM data extraction link, the terrain undulation characteristics are identified and the water system flow direction is analyzed through the 30m resolution high-precision digital elevation model data, and then the refined sub-basin boundary is extracted to realize the spatial unit partition of the basin. Subsequently, the river network main system is constructed according to the river slope and spatial position, and the multi-branch tree node coding table for pollution path simulation is generated through topological relationship analysis, which effectively defines the pollution migration direction and upstream and downstream level relationship. After the land use and soil type information is overlaid and analyzed after being rasterized at 1km, the HRU (Hydrological Response Unit) unit attribute matrix is constructed, so that the hydrological response characteristics of different land-soil combinations can be reflected in the data structure. The generation of the areal rainfall sequence is based on the hour-level Thiessen polygon interpolation, which ensures the complete expression of the spatio-temporal dynamic process of rainfall; at the same time, the soil adsorption parameters are calibrated and inverted through laboratory samples, and are converted into a structured watershed adsorption parameter table. Finally, the above various data are unified, cleaned and integrated in terms of format, spatial coding, time granularity and other dimensions to form a standard data set of hydrological response unit with complete structure, clear semantics and consistent granularity. The data set not only solves the problem of unified expression of multi-source heterogeneous hydrological factors, but also improves the input accuracy and spatial coverage integrity of pollution transport modeling, and enhances the applicability and explanatory power of pollution process analysis models in complex terrain and multiple types of land cover conditions.

[0020] Preferably, the vertical layering calculation based on the hydrological response unit standard data set in step S2 comprises:

[0021] The hydrological response unit standard data set is subjected to rainfall data extraction to obtain watershed rainfall data, and the hydrological response unit standard data set is subjected to runoff process analysis to obtain land watershed runoff data;

[0022] The watershed rainfall data and the land watershed runoff data are subjected to vertical layering calculation to obtain land runoff vertical analysis data, wherein the vertical layering calculation comprises: based on the leaf area index LAI of the vegetation interception layer being greater than 2.0, the interception capacity of the vegetation interception layer of the watershed rainfall data and the land watershed runoff data is calculated to generate the ground rainfall data; based on the saturation rate of the upper soil layer being greater than 40%, the infiltration analysis of the vegetation interception layer of the watershed rainfall data and the land watershed runoff data is performed to obtain the land watershed infiltration water quantity component data;

[0023] The land runoff vertical analysis data is obtained by vertically integrating the data of the landfall rainfall and the water infiltration amount component of the land basin.

[0024] The application extracts the rainfall data and the surface runoff data consistent in time and space from the standard data set of the hydrological response unit based on the unified structure, and develops vertical structure modeling according to the rainfall data and the surface runoff data. In the vertical profile modeling process, the leaf area index (LAI) is first introduced as a discriminant factor of the vegetation interception process. When the LAI value is greater than 2.0, the quantitative estimation of the interception capacity is triggered, and the actual landfall rainfall is calculated, so as to peel off the weakening effect of the vegetation layer on the rainfall input. On this basis, according to the upper soil saturation, when the saturation rate is higher than 40%, the infiltration analysis mechanism is introduced, the infiltration water amount under the combined action of rainfall and runoff is estimated, and the surface runoff and the infiltration component in the vertical runoff are finely divided. By vertically integrating the landfall rainfall and the infiltration component data, the land runoff vertical analysis data set with process continuity and complete physical meaning is finally formed. The data set not only covers the initial action of rainfall, the interception behavior of the hydrological surface and the response process of the soil layer, but also establishes a multivariate data chain from rainfall driving to infiltration conversion, realizes the analysis and quantization of the hydrological process in the vertical direction, and especially provides fine input conditions and profile data support for the modeling of the pollutant migration path, groundwater recharge and soil moisture change. Through this nested multi-source data fusion mode, the ability to describe the vertical process in the complex land surface system is effectively improved, and the numerical expression accuracy of the interaction mechanism between the pollutants and the hydrological process is enhanced.

[0025] Preferably, the land runoff vertical analysis data in step S2 is simulated for land pollutant migration, and real-time correction with an iteration step of 1h is performed, including:

[0026] Obtaining a soil permeability coefficient;

[0027] Based on the soil permeability coefficient, the leaching rate calculation of the dissolved nitrogen or nitrate nitrogen of the land runoff vertical analysis data is performed to generate a dissolved migration flux;

[0028] According to the soil interception force, the adsorbed phosphorus stripping amount of the land runoff vertical analysis data is calculated, and an adsorbed migration amount is obtained.

[0029] The dissolved migration flux and the adsorbed migration amount are combined at the adsorption level, and real-time correction with an iteration step of 1h is performed to obtain a dissolved / adsorbed load matrix.

[0030] The application parameterizes the hydraulic conduction characteristics of different soil layers in the vertical analysis data of land runoff by collecting and inputting soil permeability coefficient data with spatial distribution attributes; based on the parameters, the migration process of dissolved nitrogen or nitrate nitrogen in vertical water flow is quantitatively calculated by using the leaching rate estimation method, and the dissolved migration flux expressed in the form of time series is formed. On the other hand, for the migration mechanism of adsorbed phosphorus, a stripping mechanism criterion based on soil shear force and slope condition is proposed. When the slope is greater than 5°, the stripping function is activated, and the order of phosphorus migration with particles is estimated by the shear threshold analysis method, and the adsorbed migration data is generated. Thereafter, the two types of migration data are combined and processed in the adsorption level to construct the spatial-morphology coupled expression of pollution load. In the dynamic simulation link, the iteration step is set to 1 hour, the initial migration results are corrected in real time, and the difference between the historical load monitoring value and the simulation output is gradually adjusted to improve the time adaptability and response accuracy of the load estimation. The final "dissolved / adsorbed load matrix" has the characteristics of clear level, balanced granularity and complete time series in data structure, which provides a high-precision and traceable data basis for subsequent pollutant push flow calculation, load attenuation modeling and watershed dilution analysis, and further promotes the transformation of non-point source pollution modeling system from empirical estimation to mechanism-driven and data-response parallel mode.

[0031] Preferably, step S3 comprises the following steps:

[0032] Step S31: performing push flow migration calculation on the dissolved / adsorbed load matrix, and calculating the flow velocity based on the roughness of the Manning formula being 0.05 to obtain the inflow push flow flux data;

[0033] Step S32: performing joint simulation of dispersion dilution based on the inflow push flow flux data to obtain watershed dilution model level data;

[0034] Step S33: performing automatic load transfer of the multi-way tree nodes on the watershed dilution model level data to obtain the inter-node attenuation load matrix, wherein the inter-node attenuation load matrix includes an attenuation data sufficient area and an attenuation data deficient area.

[0035] The application obtains high timeliness of inflow push flow flux data by using the dissolved / adsorbed load matrix as initial input, combining the boundary condition of setting the roughness of the Manning formula to 0.05, and dynamically calculating the parameterized modeling of the flow velocity of the river section. The data not only reflects the transport capacity of the pollutants in the channel, but also provides boundary driving variables for subsequent dilution simulation. In step S32, around the physical diffusion, turbulent action and dispersion characteristics of the pollutants in the flow process, physical parameters such as molecular diffusion coefficient, eddy viscosity correction factor and river curvature are introduced respectively to construct a dispersion dilution hierarchical model, thereby forming a "watershed dilution model hierarchical data" with controllable spatial scale and clear action mechanism. The hierarchical data contains independent flux subsets of each type of dilution process and integrated comprehensive flux expressions. In step S33, the multi-branch tree node system of the watershed river network topology structure is automatically identified and loaded, and the automatic transfer and attenuation estimation of the pollution load is performed based on the water power continuity between the push flow direction and the nodes. In this process, according to the historical monitoring density and the model data integrity, the "attenuation data sufficient area" and the "attenuation data deficient area" are automatically divided. The former can enter the mechanism modeling path, and the latter enters the statistical modeling path. This node attenuation structure builds a high-precision support system for the spatio-temporal estimation of pollution distribution. The overall process not only ensures that the pollution migration link from inflow to channel to node has physical rationality and data integrity, but also provides traceable and decomposable data support system for subsequent multi-type model fusion, regional pollution hierarchical diagnosis and watershed scale pollution regulation.

[0036] Preferably, step S32 comprises the following steps:

[0037] Step S321: molecular diffusion Fick's law calculation is performed on the inflow push flow flux data with a diffusion coefficient of 10 ―5 m 2 / s to obtain molecular diffusion dilution data;

[0038] Step S322: eddy viscosity correction is performed on the inflow push flow flux data with a Reynolds number greater than 4000 to obtain turbulent diffusion dilution data;

[0039] Step S323: shear flow dispersion analysis is performed on the inflow push flow flux data with a river curvature greater than 1.2 to obtain diffusion action dilution data;

[0040] Step S324: data integration is performed on the molecular diffusion dilution data, turbulent diffusion dilution data and diffusion action dilution data to obtain the watershed dilution model hierarchical data.

[0041] The application generates micro-diffusion dilution data based on molecular motion by performing molecular diffusion calculation on the inflow plug flow flux data with a set diffusion coefficient, which accurately describes the diffusion expansion characteristics of pollutants at the molecular level. Step S322 corrects the turbulent diffusion effect in the flow field by a correction mechanism of eddy viscosity coefficient for high Reynolds number (Re>4000) flow conditions, generates turbulent diffusion dilution data reflecting the mixing effect of turbulent flow enhancement, and effectively compensates for the limitations of molecular diffusion under intense flow conditions. Step S323 uses the river channel curvature threshold (greater than 1.2) as a criterion to carry out shear flow dispersion analysis, quantitatively assesses the influence of river meandering on fluid shear force, and then forms dilution data reflecting diffusion, which fully considers the local influence of complex hydraulic topography on pollutant distribution. Step S324 integrates the above three types of dilution data by a multi-source data fusion algorithm to generate unified watershed dilution model level data, which takes into account the physical nature and time-space dynamics of different diffusion processes, and supports subsequent watershed water quality evolution simulation and pollution risk assessment. The overall technology realizes multi-dimensional and multi-scale analysis of the pollutant diffusion process in the watershed, enhances the response capability of the model to the complexity of the actual water environment and the credibility of the prediction.

[0042] Preferably, step S4 comprises the following steps:

[0043] Step S41: Mechanism modeling is performed on the sufficient attenuation data area to obtain pollutant load mechanism result data;

[0044] Step S42: For the area with insufficient attenuation data, a multiple regression model is constructed by selecting forest land ratio greater than 70%, and forest land load estimation is performed to obtain pollutant load statistical result data;

[0045] Step S43: Bayesian fusion is performed on the pollutant load mechanism result data and the pollutant load statistical result data to obtain a hybrid load output matrix.

[0046] The application realizes physical mechanism analysis of dynamic change of pollutant load by parameterizing description of key biochemical processes such as nitrification and denitrification, and generates pollutant load mechanism result data with high spatio-temporal resolution and physical interpretation. Step S42 selects an area with forest coverage of more than 70% as a sample, and applies a multiple regression analysis method to construct a statistical prediction model, so as to realize quantitative estimation of forest pollution load in combination with land use, meteorological and existing load observation data, and obtain pollutant load statistical result data driven by empirical data. Step S43 realizes optimal fusion output of pollution load information by probability weighted integration of data output by the mechanism model and statistical model result through a Bayesian fusion algorithm, fully considers the reliability and uncertainty of each model, and generates a hybrid load output matrix. The matrix has physical rationality of the mechanism model and also has experience compensation ability of the statistical model, so as to make up for the limitations of a single modeling method under different data conditions. The overall process effectively enhances the spatial continuity and temporal dynamics of pollution load estimation, and provides high-quality input data basis for subsequent pollutant transport simulation and risk assessment in a basin.

[0047] Preferably, step S41 comprises the following steps:

[0048] Step S411: nitrification analysis is performed on the temperature greater than 15 DEG C based on the attenuation data sufficient area, and the ammonia nitrogen conversion amount is obtained;

[0049] Step S412: denitrification analysis is performed on the DO less than 2 mg / L based on the attenuation data sufficient area, and the denitrification amount data is obtained by using a first-order kinetics;

[0050] Step S413: the ammonia nitrogen conversion amount and the denitrification amount data are integrated according to the monitoring point density greater than 1 / 100 km 2 Mechanism modeling is performed to obtain pollutant load mechanism result data.

[0051] The application performs detailed calculation on the nitrification process of ammonia nitrogen based on field monitoring data and experimental calibration parameters under the environmental condition of temperature greater than 15 DEG C, generates time series data reflecting the conversion amount of ammonia nitrogen to nitrate nitrogen in combination with the temperature-dependent nitrification rate model, and the data accurately describes the kinetic characteristics of the nitrification link in the nitrogen cycle. Step S412 focuses on the anoxic environment with dissolved oxygen (DO) concentration less than 2 mg / L, and quantitatively simulates the denitrification process by using a first-order kinetics model, deduces the dynamic change of denitrification amount based on field DO monitoring data and denitrification rate parameters, and forms a spatio-temporal estimation of nitrogen loss. Step S413 integrates the above nitrification conversion amount and denitrification amount data according to the monitoring point spatial density (greater than 1 / 100 km2 )spatial interpolation and mechanism modeling, using spatial statistical methods to realize the continuous spatial distribution estimation of pollutant load, and generate high-precision pollutant load mechanism result data. The result data reflects the dynamic evolution of key nitrogen transformation processes in time and embodies the influence of environmental heterogeneity on pollution load in space, improving the physical interpretation and prediction ability of the pollutant load model, and providing a reliable data basis for watershed pollution control and management.

[0052] Preferably, step S5 comprises the following steps:

[0053] Step S51: multi-objective calibration is performed on the mixed load output matrix to obtain multi-objective calibration data; and a heat map rendering is performed on the multi-objective calibration data to obtain a pollution contribution rate distribution map;

[0054] Step S52: accuracy verification is performed on the pollution contribution rate distribution map to obtain watershed water pollution simulation verification data;

[0055] Step S53: the watershed water pollution simulation verification data is used to construct a Dianchi watershed sediment / nutrient salt qualification report.

[0056] The present application, through the multi-objective calibration process based on historical monitoring data and simulation data, adopts a multi-index evaluation system to jointly correct the spatial distribution and time evolution of pollutant load, and obtains multi-objective calibration data that can reflect the contribution characteristics of various pollutants in the watershed. After heat map rendering processing, the spatial heterogeneity of pollution contribution rate is intuitively presented, forming a pollution contribution rate distribution map, which provides quantitative basis for watershed pollution source identification and management strategy. In step S52, the pollution contribution rate distribution map is verified for accuracy with actual monitoring data, and statistical indicators such as correlation coefficient and root mean square error are used to evaluate the accuracy and reliability of the model simulation, to generate watershed water pollution simulation verification data, thereby ensuring the consistency of the model output and the actual water quality. In step S53, the simulation data after verification is systematically arranged, and according to the national and local water quality standards, a Dianchi watershed sediment and nutrient salt qualification report is constructed to reflect the water quality status and treatment effect of the watershed. The report not only provides scientific decision support for environmental management departments, but also provides data support for the formulation of subsequent pollutant load control and water quality improvement measures. The overall process forms a complete closed loop from data collection, model calibration to verification and report construction, effectively improving the accuracy and application value of watershed water pollution simulation, and promoting the scientific management and sustainable development of the watershed environment.

[0057] In the present specification, a watershed water pollution migration and transformation simulation system based on the MT-DHM model is provided for executing the above-mentioned watershed water pollution migration and transformation simulation method based on the MT-DHM model, and the watershed water pollution migration and transformation simulation system based on the MT-DHM model comprises:

[0058] The basin data acquisition and standardization module is used for data acquisition on the multi-source space of the basin, data preprocessing, and generation of a standard data set of the hydrological response unit;

[0059] The vertical runoff and migration correction module is used for vertical layering calculation based on the standard data set of the hydrological response unit, obtaining of land runoff vertical analysis data, land pollution migration simulation on the land runoff vertical analysis data, and real-time correction with an iteration step of 1h, and obtaining of a dissolved / adsorbed load matrix.

[0060] The water system migration and dilution level simulation module is used for push flow migration calculation on the dissolved / adsorbed load matrix, and dispersion and dilution level joint simulation, obtaining of basin dilution model level data, automatic load transfer of the multi-branch tree nodes of the basin dilution model level data, and obtaining of a node attenuation load matrix, wherein the node attenuation load matrix includes an attenuation data sufficient area and an attenuation data deficient area.

[0061] The mechanism / statistical modeling and Bayesian fusion module is used for mechanism modeling on the attenuation data sufficient area, obtaining of pollutant load mechanism result data, statistical modeling on the attenuation data deficient area, obtaining of pollutant load statistical result data, Bayesian fusion of the pollutant load mechanism result data and the pollutant load statistical result data, and obtaining of a mixed load output matrix.

[0062] The contribution rate calibration and qualified report generation module is used for multi-target calibration on the mixed load output matrix, obtaining of a pollution contribution rate distribution graph, precision verification on the pollution contribution rate distribution graph, obtaining of basin water pollution simulation verification data, and construction of a Dianchi Basin sediment / nutrient salt qualified report on the basin water pollution simulation verification data. BRIEF DESCRIPTION OF DRAWINGS

[0063] Figure 1 It is a step flow diagram of a basin water pollution migration and transformation simulation method based on the MT-DHM model.

[0064] Figure 2 It is a detailed implementation step flow diagram of step S3. Figure 1

[0065] The implementation of the object of the present application, functional features and advantages will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0066] ​The technical method of the patent of the application will be described clearly and completely in combination with the drawings. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the application.

[0067] In addition, the drawings are only schematic illustrations of the application and are not necessarily drawn to scale. Identical reference signs in the drawings represent identical or similar parts, so repeated descriptions thereof will be omitted. Some of the block diagrams shown in the drawings are functional entities, which do not necessarily have to correspond to physically or logically independent entities. The functional entities can be implemented in the form of software, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0068] It should be understood that although the terms "first", "second" and the like can be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, without departing from the scope of the exemplary embodiments, a first element can be called a second element, and similarly a second element can be called a first element. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0069] To achieve the above-mentioned purpose, please refer to Figures 1 to 2 A watershed water pollution migration and transformation simulation method based on an MT-DHM model, the method comprising the following steps:

[0070] Step S1: data collection is performed on a watershed multi-source space, and data preprocessing is performed to generate a hydrological response unit standard data set;

[0071] Step S2: vertical layering calculation is performed based on the hydrological response unit standard data set to obtain land runoff vertical analysis data; land pollution migration simulation is performed on the land runoff vertical analysis data, and real-time correction with an iteration step of 1 h is performed to obtain a dissolved / adsorbed load matrix;

[0072] Step S3: push flow migration calculation is performed on the dissolved / adsorbed load matrix, and dispersion dilution level joint simulation is performed to obtain watershed dilution model level data; automatic load transfer of multi-branch tree nodes is performed on the watershed dilution model level data to obtain an inter-node attenuation load matrix, wherein the inter-node attenuation load matrix includes an attenuation data sufficient area and an attenuation data deficient area;

[0073] Step S4: mechanism modeling is performed on the attenuation data sufficient area to obtain pollutant load mechanism result data, statistical modeling is performed on the attenuation data deficient area to obtain pollutant load statistical result data, Bayes fusion is performed on the pollutant load mechanism result data and the pollutant load statistical result data to obtain a mixed load output matrix;

[0074] Step S5: multi-objective rating is performed on the mixed load output matrix to obtain a pollution contribution rate distribution diagram; accuracy verification is performed on the pollution contribution rate distribution diagram to obtain basin water pollution simulation verification data; and a Dianchi basin sediment / nutrient salt qualification report is constructed based on the basin water pollution simulation verification data.

[0075] In the embodiment of the present application, as shown in the reference Figure 1 The method comprises the following steps:

[0076] Step S1: data collection is performed on a basin multi-source space, and data preprocessing is performed to generate a hydrological response unit standard data set;

[0077] In the embodiment of the present application, in terms of data collection, representative remote sensing images (such as Sentinel-2 or Landsat series) are preferentially selected as the source of basic ground information, and multiple types of heterogeneous data such as rainfall data of weather stations, flow data of hydrological stations, soil property database, land use present situation map, DEM elevation data and administrative division vector boundary information are synchronously collected. These data usually have problems such as inconsistent spatial scales, non-uniform time resolution, different observation accuracies, etc., so preliminary regularization needs to be carried out through resampling, interpolation, coordinate system unification, spatial clipping, etc. Then, based on the terrain analysis and flow-direction-flow-rate deduction method, the DEM data is divided into sub-basins and river network extraction operation is carried out, combined with the land use classification layer and soil type data for zoning processing, to construct the basic hydrological response unit (HRU) associated with the geomorphic unit, hydrological cycle and pollutant transport process. In order to ensure the availability of data in the subsequent vertical simulation and pollution transport process, the data points with missing or abnormal observation in the collected data are repaired and corrected by combining the historical time series trend regression and neighborhood spatial interpolation. At the same time, in order to ensure the consistency and operability of the data set, all input factors (such as slope, surface roughness, soil permeability, vegetation index, surface runoff coefficient, etc.) are normalized according to the preset standardization rules, and the processing results are mapped to the hydrological response unit grid structure, finally forming a data basic framework that is consistent in spatial attributes, unified in time dimension and suitable for downstream simulation analysis. The data set constructed in this step is not only the key basis for supporting the subsequent vertical runoff layering calculation and pollution migration modeling, but also an important precondition for ensuring the calculation accuracy and regional representativeness of the entire simulation system.

[0078] Step S2: based on the hydrological response unit standard data set, vertical layering calculation is carried out to obtain land runoff vertical analysis data; land pollution migration simulation is carried out on the land runoff vertical analysis data, and real-time correction with an iteration step of 1h is carried out to obtain a dissolved / adsorbed load matrix;

[0079] In the hydrological response unit grid in the embodiment of the application, in combination with DEM derived slope data, land use type and soil profile parameters (such as soil layer thickness, organic matter content, volumetric water content, saturated hydraulic conductivity and the like), vertical layering calculation is performed according to the surface-soil-underground three-layer structure. The hydraulic parameters between layers are extracted from the national soil database and the land use classification table through table lookup method and spatial interpolation technology, and are mapped to each response unit grid. After the vertical profile is established, the downward infiltration of surface precipitation, soil water storage and deep seepage process are dynamically simulated by the modified Green-Ampt model or the multi-layer Richards equation, and the water flux data between layers per time step are obtained. Subsequently, on the basis of the established water flux, the pollutant input source data are introduced, including surface pollutant deposition flux, fertilizer amount statistics, livestock and poultry breeding emission coefficient and domestic non-point source pollution intensity data, and are processed according to the form characteristics of the pollutants and are divided into two categories of dissolved and adsorbed states. In the migration simulation of the dissolved state pollutants, the Advection-Dispersion transport model is adopted in combination with real-time soil moisture content data to calculate the convection and diffusion behavior of the pollutants in the vertical profile; in the adsorbed state part, the Freundlich isothermal adsorption function is introduced to represent the solid-phase adsorption proportion of the soil particles to the pollutants, and the distribution relationship between the pollutants in the liquid phase and the solid phase is dynamically adjusted. The whole simulation process is iteratively executed at a 1-hour step, and is corrected in real time by using observation data or historical simulation output data to ensure the stability and rationality of the migration model under actual boundary conditions, and finally the dissolved / adsorbed state load matrix reflecting the multi-form transformation and enrichment process of the pollutants in the vertical system of the land is output. The matrix serves as the intermediate state result of the pollutant migration, and provides high-resolution and high-dynamic input data support for subsequent water system migration and dilution simulation.

[0080] Step S3: push flow migration calculation is performed on the dissolved / adsorbed state load matrix, and dispersion and dilution level joint simulation is performed to obtain the watershed dilution model level data; the watershed dilution model level data are subjected to automatic load transfer of the multi-way tree nodes to obtain the inter-node attenuation load matrix, wherein the inter-node attenuation load matrix includes an attenuation data sufficient area and an attenuation data deficient area;

[0081] In the embodiment of the present application, the spatial grid structure of the hydrological response unit is combined with DEM (Digital Elevation Model) to construct the flow-flow graph, and the confluence path and hydraulic connection relationship of each tributary unit are determined by combining the river network structure of the basin and the measured or simulated daily runoff data. Then the load matrix output in the previous step is mapped to the corresponding hydrological unit, and the calculation of the migration of pollutants with water flow is carried out through the river unit length load transmission rate and the hydrodynamic parameters (such as cross-section flow velocity, water depth, hydraulic slope). The advection model of the pollutant adopts one-dimensional or quasi-two-dimensional water quality transport equation (such as Advection-Dispersion Equation, ADE), and the load change is evolved step by step at the hour level time step. In order to simulate the dilution effect of the load in the transmission process, further introduce the process variables such as the inflow of the tributary, the rain source recharge and the evaporation loss, and through the construction of the dispersion dilution hierarchical structure based on the node intersection rule, the pollutant concentration of different levels of confluence unit is updated layer by layer, so as to form the multi-level pollution dilution simulation results in the basin. On this basis, in order to realize the automatic propagation and attenuation analysis of the pollution load in the complex basin network, the nodes of the river network are organized by using the multi-way tree structure, each node represents a river section or a confluence point, and the pollution load transmission between nodes is calculated according to the upstream load total, flow velocity factor, node retention time and local source input. In the construction of the attenuation relationship between nodes, the empirical attenuation model (such as exponential attenuation or power function attenuation model) is used to fit the attenuation law of the pollutant between nodes, and the data sufficient area and the data deficient area are divided, and the complete pollution load attenuation matrix between nodes is generated, which provides a clear structure and dynamic adjustable input basis for the subsequent mechanism modeling and statistical modeling. The matrix records the transmission amount, attenuation rate and time delay characteristics of the pollutant between nodes in the form of hour-level time series, which is the core carrier of subsequent multi-source modeling and fusion calculation.

[0082] Step S4: mechanism modeling is performed on the attenuation data sufficient area to obtain pollutant load mechanism result data; statistical modeling is performed on the attenuation data deficient area to obtain pollutant load statistical result data; Bayesian fusion is performed on the pollutant load mechanism result data and the pollutant load statistical result data to obtain a mixed load output matrix;

[0083] In the embodiments of the present application, mechanism modeling and statistical modeling methods are used to construct pollutant load evolution models, and the results of the two are integrated through Bayesian fusion to construct a mixed load output matrix from the data level. In the attenuation data sufficient area, by collecting historical observation data, measured water quality and quantity data and process simulation results, multi-dimensional variables such as time series pollutant concentration, flow rate, water temperature and water retention time are extracted, and based on hydrodynamics and water quality transmission theory, a one-dimensional reaction-transport model or a retention area coupled model is used to establish the transport and attenuation relationship of pollutants. Such mechanism model often uses first-order kinetic reaction equation or multi-component solute conversion mechanism to describe the physical and chemical processes such as degradation, transformation and sedimentation of pollutants in water body, and simultaneously couples the topological characteristics of river section, flow path and node in spatial structure. Model parameters such as reaction rate coefficient, sedimentation rate and partition volume ratio are determined by measured data inversion or sensitivity analysis, and dynamic simulation of pollutant attenuation is performed at hourly scale to output high spatio-temporal resolution of pollutant load response data. In the attenuation data deficient area, due to the lack of sufficient process data, it is necessary to turn to statistical modeling path, and by introducing historical data of spatial adjacent nodes, regional land use structure, hydrological response unit type and upstream load input data, a pollutant load estimation model based on linear regression, principal component regression or random forest regression is constructed. In the data processing process, normalization transformation, feature importance screening and cross-validation strategy are used to improve the robustness and generalization ability of the model, and finally the estimated value of the pollutant load of the corresponding node and its uncertainty range are output. In order to realize the collaborative fusion of mechanism model and statistical model results, the Bayesian fusion mechanism is introduced, the prior probability distribution of pollutant load is taken as the basis, the conditional likelihood function of mechanism model and the confidence prediction interval of statistical model are combined, the posterior probability distribution is constructed, and the mixed load output matrix of pollutants is obtained in the form of maximum posterior estimation. The matrix records the pollutant load and its confidence interval of each node at a given time scale, which provides quantitative data input basis for subsequent pollutant contribution rate calibration and simulation accuracy verification.

[0084] Step S5: multi-objective calibration is performed on the mixed load output matrix to obtain a pollutant contribution rate distribution map; the pollutant contribution rate distribution map is subjected to accuracy verification to obtain watershed water pollution simulation verification data; and the watershed water pollution simulation verification data is used to construct a Dianchi Lake watershed sediment / nutrient salt qualification report.

[0085] In the embodiment of the present application, a mixed load output matrix is taken as input, which contains information of pollutant concentration, form proportion and migration state of each node in the basin at different time scales. Combined with the measured water quality data (such as total phosphorus, total nitrogen, COD, TSS, etc.) and the corresponding hydrological data (such as flow, flow rate, water level, etc.), a multi-objective calibration function is constructed with the minimum deviation of pollutant concentration, the matching degree of pollution process response time and the consistency of spatial load distribution as the target. In order to balance the accuracy and stability of the calibration, a multi-objective genetic algorithm (MOGA) or a particle swarm optimization algorithm (PSO) is used for global parameter optimization, and at the same time, the sensitivity analysis of the key parameters (such as source strength coefficient, attenuation coefficient, settling rate, dilution factor) in the model structure is carried out, and the constraint boundary is set in the calibration process to avoid overfitting of the model. After the calibration is completed, the attribution source of each response unit or sub-basin node pollutant is analyzed, and the spatial distribution proportion of the pollution source is calculated according to the time and space load contribution relationship, and finally the pollution contribution rate distribution map is output. The map is based on the spatial interpolation (such as Kriging or IDW) method to express the continuous data of the discrete node contribution rate, and superimposes the administrative division, land use and water system layers to enhance the geographical recognition. In the accuracy verification link, the measured water quality data of different time periods from the calibration is selected as an independent verification set, and the fitting degree between the model output and the observation data is evaluated by calculating the Nash efficiency coefficient (NSE), the root mean square error (RMSE) and the relative error (RE). After the verification result meets the accuracy requirement, the pollution contribution rate distribution map is taken as the core reference, combined with the water quality standard of Dianchi Basin, the total amount control index, the historical pollution evolution trend and the key point monitoring data, the sediment and nutrient salt qualified report is constructed, which quantitatively lists the standard reaching situation, pollution source structure, potential over-standard risk and suggestion control strategy of each sub-region, forming a pollution simulation output result with quantitative support and spatial interpretation.

[0086] Preferably, step S1 comprises the following steps:

[0087] Step S11: obtaining a hydrological data set from a hydrological check station;

[0088] Step S12: extracting DEM data from the hydrological data set according to a resolution of 30m, and performing basin boundary analysis to generate a sub-basin division map;

[0089] Step S13: extracting river network data from the hydrological data set according to river slope, and performing topological relationship analysis to obtain a multi-branch tree node coding table;

[0090] Step S14: dividing land use / soil type data according to a 1km grid from the hydrological data set, and performing spatial overlay analysis to obtain an HRU unit attribute matrix;

[0091] Step S15: Perform Thiessen polygon spatial interpolation on the hydrological data set at the hour level to generate a sub-basin area rainfall sequence;

[0092] Step S16: Construct soil adsorption parameters for the hydrological data set and perform laboratory calibration to obtain a watershed adsorption parameter table;

[0093] Step S17: Perform data cleaning and format unification on the sub-basin division map, multi-fork tree node coding table, HRU unit attribute matrix, sub-basin area rainfall sequence, and watershed adsorption parameter table to obtain a standard hydrological response unit data set.

[0094] In the embodiment of the present application, the hydrological data set at the hourly or daily scale is obtained through the hydrological inspection station deployed in the river channel, tributaries and reservoir inlet and outlet, including time series data such as rainfall, evaporation, runoff, flow rate, water level and the like, and matching the corresponding spatial positioning information. In step S12, the digital elevation model (DEM) with a spatial resolution of 30 m is called, the flow direction extraction algorithm (such as D8 algorithm) and flow accumulation analysis are used to determine the flow direction and catchment area, the basin boundary is identified in combination with the confluence threshold setting, and the generation of the sub-basin division map is completed. In step S13, the river section slope derived from the DEM is used as an index to extract the river network system, and through the establishment of a directed graph structure, the numbering and topological relationship coding of the multi-branch tree nodes are realized based on the upstream-downstream hydraulic connection relationship, and a multi-branch tree node coding table containing node number, parent-child node pair, river section length, slope and the like is generated. In step S14, the land use data (such as remote sensing classification layer) and soil type data (such as national soil database) are spatially resampled and superimposed, and the HRU (Hydrological Response Unit) unit is constructed according to the land use-soil type combination logic, and the HRU unit attribute matrix is formed, in which each row corresponds to a grid unit, and contains soil type, land use type, slope grade, surface permeability and the like. In step S15, the point rainfall data is spatially interpolated, the Voronoi diagram is constructed, the station rainfall data is mapped to the corresponding polygon control area, and the hourly surface rainfall time series of each sub-basin is generated to reflect the spatial distribution continuity of rainfall. In step S16, the adsorption experiment is carried out based on the soil samples collected in the field, the Freundlich or Langmuir isothermal adsorption model is constructed, and the unit mass adsorption capacity, adsorption constant and the like are extracted by combining the experimental fitting results to form a structured watershed adsorption parameter table. Finally, in step S17, the sub-basin division map, multi-branch tree coding table, HRU attribute matrix, sub-basin rainfall sequence and adsorption parameter table are spatially registered in the unified coordinate system, the data cleaning technology is used to process the missing values and abnormal values, the field structure is standardized, the coding mode is unified, and the unified data framework is output as the hydrological response unit standard data set, which is the core data basis for supporting the modeling of watershed pollutant migration and multi-scale hydrological simulation analysis.

[0095] Preferably, the vertical layering calculation based on the hydrological response unit standard data set in step S2 includes:

[0096] The rainfall data of the hydrological response unit standard data set is extracted to obtain the watershed rainfall data, and the runoff process analysis of the hydrological response unit standard data set is performed to obtain the land watershed runoff data.

[0097] The land runoff vertical analysis data is obtained by vertically layering calculation of the basin rainfall data and the land basin runoff data, wherein the vertically layering calculation comprises: performing interception capacity calculation on the vegetation interception layer of the basin rainfall data and the land basin runoff data based on the leaf area index (LAI) of the vegetation interception layer being greater than 2.0 to generate the land rainfall data; performing infiltration analysis on the vegetation interception layer of the basin rainfall data and the land basin runoff data based on the saturation rate of the upper soil layer being greater than 40% to obtain the land basin infiltration water quantity component data;

[0098] The land runoff vertical analysis data is obtained by vertically layering calculation of the basin rainfall data and the land basin runoff data, wherein the vertically layering calculation comprises: performing interception capacity calculation on the vegetation interception layer of the basin rainfall data and the land basin runoff data based on the leaf area index (LAI) of the vegetation interception layer being greater than 2.0 to generate the land rainfall data; performing infiltration analysis on the vegetation interception layer of the basin rainfall data and the land basin runoff data based on the saturation rate of the upper soil layer being greater than 40% to obtain the land basin infiltration water quantity component data;

[0099] In the embodiment of the present application, the spatial boundary of the hydrological response unit is used as a mask to extract the sub-basin surface rainfall sequence corresponding to the hydrological response unit, and "basin rainfall data" with time index is generated; at the same time, a runoff generation model (such as SCS-CN method, Green-Ampt or simplified TOPMODEL) driven by DEM, land use, slope and HRU parameters is called in the same response unit to simulate the surface runoff process, and "land basin runoff data" is generated. Subsequently, the vertically layering calculation stage is entered, and the system constructs a vertical profile with a three-layer structure of "vegetation interception layer-soil surface layer-deep soil layer". In the vegetation interception modeling, the leaf area index (Leaf Area Index, LAI) obtained by remote sensing is called according to the vegetation type of each response unit, and the units with LAI greater than 2.0 are selected for simulating significant interception process. The interception capacity per unit time is calculated according to the empirical relationship between LAI and rainfall intensity by table lookup or empirical formula (such as Gash model or Rutter model), and then the corresponding interception amount is deducted from the original rainfall data to obtain "land rainfall data", that is, the net rainfall reaching the soil surface. In the infiltration process modeling, the saturated water content, initial water content and permeability coefficient data of different soil layers in the soil property database are combined, and the soil units with a saturation rate greater than 40% are selected, and the Green-Ampt or Richard equation is used for simplified infiltration estimation to simulate the infiltration behavior of "land rainfall" to the surface soil per unit time, and "infiltration water quantity component data" is output. Finally, the above-mentioned "land rainfall data" and "land basin infiltration water quantity component data" are aligned in space and time index, and the vertical path integration is performed to construct the complete "land runoff vertical analysis data". The data takes the response unit as the basic structure, records the distribution of net rainfall per unit area, infiltrable amount and actual infiltration amount at each time step, and provides a vertical input framework and boundary constraint condition for subsequent land pollution migration process modeling.

[0100] Preferably, the land runoff vertical analysis data is simulated in step S2, and real-time correction with an iteration step of 1h includes:

[0101] obtaining a soil permeability coefficient;

[0102] based on the soil permeability coefficient, calculating a leaching rate of dissolved nitrogen or nitrate nitrogen on land runoff vertical analysis data to generate a dissolved migration flux;

[0103] according to the soil intercepting force, calculating the stripping amount of the adsorbed state phosphorus on the land runoff vertical analysis data with a slope greater than 5° to obtain an adsorbed migration amount;

[0104] merging the dissolved migration flux and the adsorbed migration amount according to the adsorption level, and performing real-time correction with an iteration step of 1h to obtain a dissolved / adsorbed load matrix.

[0105] In the embodiment of the application, the migration behavior of nitrogen and phosphorus pollutants in the land runoff process is simulated and dynamically integrated by a quantitative parameter driven vertical water quality migration model. First, the saturated hydraulic conductivity (Ksat) and effective porosity of different types of soil in the soil data set are statistically analyzed to obtain the soil permeability coefficient data corresponding to the spatial unit, and the data are mapped according to the hydrological response unit to form a spatially matched permeability parameter grid. In the simulation of dissolved nitrogen or nitrate nitrogen migration, the previously generated land runoff vertical analysis data are called, the unit time infiltration water amount and the soil permeability coefficient are combined, and a first-order momentum conservation and mass conservation coupled model is used to estimate the leaching rate of nitrogen carried downward by water, and the core calculation is based on the coupling relationship between the hydraulic migration flux and the nitrogen concentration gradient, and a "dissolved migration flux" data matrix is generated with an hourly time step. For the migration process of adsorbed pollutants, spatial units with a slope greater than 5° are selected, the slope distribution is extracted from the existing HRU unit attribute matrix, an empirical slope erosion model (such as Revised Universal Soil Loss Equation, RUSLE) is constructed based on the relationship between the adsorbed phosphorus content of soil particles and the stripping critical shear force, and the soil erosion modulus and the erosion response coefficient are introduced for adjustment to estimate the stripping amount of adsorbed phosphorus per unit area, and a "adsorbed migration amount" data set is generated. In the form fusion stage, the "dissolved migration flux" and the "adsorbed migration amount" are aligned and weighted according to the spatial unit to construct a three-dimensional "adsorption level merging matrix", which retains the hourly granularity in the time dimension and records the dynamic transfer value of each form of pollutant in a specific unit. To ensure the consistency between the simulation results and the actual monitoring data, a real-time correction mechanism with an hourly scale is further introduced, Kalman filtering or least squares dynamic regression algorithm based on bias function is used to update the predicted value of the merged matrix hour by hour, and finally a "dissolved / adsorbed load matrix" is output, which fully expresses the water dynamic-chemical dynamic coupled migration law of pollutants in the vertical path of soil, and provides fine-grained input data for subsequent push flow evolution and dilution modeling.

[0106] As an example of the present application, reference is made to Figure 2 in the present example, the step S3 comprises:

[0107] Step S31: push flow migration calculation is performed on the dissolved / adsorbed load matrix, and the flow velocity is calculated based on the roughness of the Manning formula being 0.05, to obtain inflow push flow flux data;

[0108] Step S32: dispersion dilution level joint simulation is performed based on the inflow push flow flux data, to obtain the basin dilution model level data;

[0109] Step S33: automatic load transfer of the multi-branch tree node is performed on the basin dilution model level data, to obtain the inter-node attenuation load matrix, wherein the inter-node attenuation load matrix includes an attenuation data sufficient area and an attenuation data deficient area.

[0110] In the embodiment of the present application, in the push flow migration calculation stage, firstly, the dissolved / adsorbed load matrix is mapped according to the river section spatial unit, the pollutant load of each HRU unit or sub-basin is distributed to the corresponding river section entrance position, and the spatial position index of the pollutant input point is established. In terms of flow velocity calculation, the Manning formula wherein n is set to 0.05, the hydraulic radius R and the river section slope S are obtained from the DEM and the river section parameter extraction, the water flow velocity per unit time of each river section is calculated through the formula, and the volume flow rate is estimated in combination with the cross section width and depth, so as to calculate the “inflow push flow flux data” per unit time. In the dispersion dilution level joint simulation, the inflow push flow flux data is taken as the input, combined with the river flow along the flow change, the river water temperature, the background water quality concentration and other parameters, the multi-scale diffusion-convection equation is adopted to establish the pollutant dilution simulation model, the model considers the flow velocity difference and the concentration gradient change at the confluence point of the main river and the tributary, and introduces the diffusion coefficient and the distribution coefficient and other parameters, to generate the “basin dilution model level data” at different spatial scales, which retains the river section code, the pollutant concentration time evolution curve and the dilution attenuation factor and other fields in structure. Subsequently, under the multi-branch tree network node level, the system traces the path of the outlet load data of each river section, constructs an automatic load transfer mechanism, that is, the output of the upstream node is taken as the input of the downstream node, and the “inter-node attenuation load matrix” is formed through the accumulation and decrement mechanism. In this process, for the data complete area (that is, supported by the measured or historical verification data), the mechanism model based on the physical parameters is adopted to calculate the load attenuation value; and in the data deficient area, it is marked as an incomplete attenuation area, and a statistical interpolation or missing compensation strategy is introduced to retain the load transmission continuity between the nodes, so as to ensure that the “inter-node attenuation load matrix” meets the modeling requirements of subsequent pollutant load allocation and contribution rate analysis in structure and data consistency.

[0111] Preferably, step S32 comprises the following steps:

[0112] Step S321: molecular diffusion Fick law calculation on the inflow push flow flux data with a diffusion coefficient of 10 ―5 m 2 / s, to obtain molecular diffusion dilution data;

[0113] Step S322: eddy viscosity correction on the inflow push flow flux data with a Reynolds number greater than 4000, to obtain turbulent diffusion dilution data;

[0114] Step S323: shear flow dispersion analysis on the inflow push flow flux data with a river channel curvature greater than 1.2, to obtain dispersion dilution data;

[0115] Step S324: data integration of the molecular diffusion dilution data, the turbulent diffusion dilution data and the dispersion dilution data, to obtain the watershed dilution model level data.

[0116] In the embodiment of the application, the concentration gradient field is first constructed based on the inflow push flow flux data, and Fick's first law The flux of the pollutant concentration changing with space at the micro level of the fluid is calculated, where D represents the molecular diffusion coefficient (unit: m 2 / s), which is obtained from an empirical table or an experimental database according to the type of the pollutant and the water temperature condition, the calculation result generates “molecular diffusion dilution data”, and records the micro concentration diffusion intensity at each river section position. In step S322, the Reynolds number of each river section is calculated, the completely turbulent flow region with Re greater than 4000 is identified, and the eddy viscosity coefficient is introduced for correction based on the turbulent viscosity principle, an improved Elder formula or a similar eddy diffusion model D t =αuh is adopted, where u is the flow velocity, h is the water depth, and α is the turbulent diffusion empirical coefficient, the macroscopic dilution behavior under the action of turbulent flow is estimated, and “turbulent diffusion dilution data” is generated. Step S323 focuses on the change of the geometric characteristics of the river channel, according to the river section curvature index extracted in the geographic information system (GIS), the region with a curvature greater than 1.2 is screened out, a shear flow field profile is established, the shear effect between different longitudinal water layers due to the flow velocity gradient is calculated in the two-dimensional flow field, and then the longitudinal dispersion coefficient Where H is water depth, K is a correction factor for bending, simulating the dispersion migration process of pollutants caused by shear flow, and outputting "diffusion dilution data". The above three types of dilution data are attribute fused with river section number as index, and the linear superposition, empirical weighted average or Bayesian confidence fusion method is adopted to construct the "basin dilution model layer data" with unified structure, which takes the time series dilution concentration curve of each river section as the core and contains the comprehensive response of three dimensions of molecular diffusion, turbulent diffusion and shear dispersion: supporting the continuous evolution modeling and dynamic load transfer calculation of basin pollution load.

[0117] Preferably, step S4 comprises the following steps:

[0118] Step S41: mechanism modeling is performed on the attenuation data sufficient area to obtain pollutant load mechanism result data;

[0119] Step S42: a multiple regression model is constructed for the area with forest coverage greater than 70% in the attenuation data insufficient area, and forest load estimation is performed to obtain pollutant load statistical result data;

[0120] Step S43: Bayesian fusion is performed on the pollutant load mechanism result data and the pollutant load statistical result data to obtain a mixed load output matrix.

[0121] In the embodiment of the application, first, the attenuation data sufficient area, i.e., the multi-period node area with continuous measured concentration, flow rate and historical load record, is identified, and on this basis, a first-order attenuation model based on reaction kinetics is introduced or a river-level pollutant migration and transformation equation group (such as the Advection-Dispersion-Reaction model) is used to simulate the physical mechanism of the pollutant concentration evolution process in combination with hydrological driving data and diffusion parameters (such as hydraulic radius, flow rate, residence time), and output "pollutant load mechanism result data", which presents a pollutant load prediction matrix under space-time distribution, has fields such as river section / node code, time index and mass flux. In step S42, for the attenuation data insufficient area, the forest coverage in the land use data is extracted, the area with forest coverage greater than 70% is screened, and a multiple linear regression model or a ridge regression model is constructed in combination with soil type, slope, annual average precipitation and runoff depth and other variables. The expression is generally Y=β0+∑β i X i +, wherein Y represents the estimated pollution load, X i is a land use related independent variable, β iis a regression coefficient, and is a residual term. The coefficients are fitted by calibration data of a similar historical basin, and the current target area is estimated and inferred to output the "pollutant load statistical result data". In step S43, the Bayesian fusion method is used to combine the two types of results, and the mechanism model output is set as the prior distribution P(theta), and the statistical model result is the observation likelihood P(D|theta), and the posterior distribution after fusion is P(theta|D) proportional to P(D|theta)P(theta). In specific implementation, Markov Chain Monte Carlo (MCMC) sampling can be used to solve the parameters of the posterior distribution, and the confidence interval estimation of the model output is performed, and finally the "mixed load output matrix" with unified structure and stable numerical value is generated, which is used to depict the comprehensive output result of multi-source pollution load under different regional and different model application conditions, and provides quantitative input for subsequent pollution contribution rate decomposition.

[0122] Preferably, step S41 comprises the following steps:

[0123] Step S411: based on the attenuation data sufficient area, the nitrification analysis is carried out on the temperature greater than 15 DEG C, and the ammonia nitrogen conversion amount is obtained;

[0124] Step S412: based on the attenuation data sufficient area, the denitrification analysis is carried out on the DO less than 2 mg / L, and the denitrification amount data is obtained by using the first order kinetics processing;

[0125] Step S413: the ammonia nitrogen conversion amount and the denitrification amount data are combined according to the monitoring point density greater than 1 / 100 km 2 Mechanism modeling is carried out to obtain the pollutant load mechanism result data.

[0126] In the embodiment of the application, the temperature attribute layer screening is carried out on the attenuation data sufficient area, and the sample unit with water temperature higher than 15 DEG C is extracted, and the temperature threshold is the minimum physiological activity temperature limit of ammonia oxidizing bacteria. According to the empirical relationship between water temperature and nitrification rate, the temperature corrected first order reaction rate expression is adopted Wherein k 20 is the reaction rate constant at 20 DEG C, theta is the temperature sensitivity coefficient (usually the value range is 1.03-1.07), T is the measured water temperature, is the initial concentration of ammonia nitrogen, and the "ammonia nitrogen conversion amount" is calculated accordingly. In step S412, the water unit with DO concentration lower than 2 mg / L is selected, the potential occurrence area of denitrification process is identified, and the first order kinetics model is used to estimate the denitrification reaction, wherein k dn is the denitrification reaction rate constant, is the nitrate nitrogen concentration, and k dnThe region calibration or empirical correction is performed, and the C / N ratio is introduced as a process correction factor for precision checking, and finally a "denitrification amount data" matrix is generated. Step S413 integrates the nitrogen form conversion data generated in steps S411 and S412 in space, and combines the GIS platform to superimpose the monitoring point distribution information, and selects the spatial region with a monitoring point density greater than 1 / 100 km2 as the effective region of mechanism modeling. In this region, the pollutant conversion equation set is used to construct a water quality evolution model, the nitrification of ammonia nitrogen and the denitrification of nitrate are taken as coupled processes, a mechanism model including nitrogen mass balance, temperature-DO double factor driven is constructed, and "pollutant load mechanism result data" is generated. The data is indexed by river section number, and the structure includes time series nitrogen mass flux, conversion ratio, process rate parameter and attenuation factor fields, which has the integrity and consistency of supporting pollution source analysis and spatio-temporal coupling simulation.

[0127] Preferably, step S5 comprises the following steps:

[0128] Step S51: multi-objective rating is performed on the mixed load output matrix to obtain multi-objective rating data; and a heat map rendering is performed on the multi-objective rating data to obtain a pollution contribution rate distribution map;

[0129] Step S52: accuracy verification is performed on the pollution contribution rate distribution map to obtain watershed water pollution simulation verification data;

[0130] Step S53: the watershed water pollution simulation verification data is used to construct a Dianchi Basin sediment / nutrient salt qualification report.

[0131] In the embodiment of the present application, the "mixed load output matrix" is taken as the rating object, and a parameter optimization problem containing multiple objective functions is constructed. The multi-objective rating process usually contains typical water quality / quantity fitting indicators such as Nash-Sutcliffe efficiency coefficient (NSE), root mean square error (RMSE), and bias rate (PBIAS), and introduces multi-source measured concentration, flow rate, and load observation values as reference data for matching. Non-dominated sorting genetic algorithm (NSGA-II) or Pareto front optimization strategy is used to iteratively rate each key influencing parameter in the output matrix after Bayesian fusion until the multiple objective functions all reach the convergence standard. The obtained "multi-objective rating data" contains the parameter set of each sub-basin or HRU unit and the matching quality of the actual observation data. In the visualization stage, the pollution source contribution value after rating is normalized and color scale mapped according to the spatial coding, the spatial difference of pollution load is intuitively expressed in the form of color gradient by using the heat map rendering algorithm, and a "pollution contribution rate distribution map" is generated, in which each pixel point of the image maps a pollution source intensity value. In step S52, the system verifies the accuracy of the simulation results reflected by the heat map and the observation data of the actual monitoring points, mainly using error statistical test methods such as mean square error (MSE), mean absolute error (MAE), and R 2 correlation coefficient, and generates "basin water pollution simulation verification data" based on the point-to-area error comparison between the spatial interpolation results and the observation point data. In step S53, the simulation verification results are qualitatively judged and quantitatively evaluated by taking the sediment and nutrient salt concentration limits of the Dianchi Basin in the country or region as the qualified threshold standard, combining the contribution rate of each sub-basin and the probability interval of exceeding the limit value, and forming a "Dianchi Basin sediment / nutrient salt qualified report" containing elements such as pollutant type, spatial location, concentration level, and compliance status. The report is output in a structured format and has functions such as global evaluation, key area identification, and limit value comparison analysis, which provides a direct basis for basin water environment management and control.

[0132] Especially important is that step S51 includes:

[0133] obtaining historical measured data of runoff-sediment-nitrogen and phosphorus;

[0134] calculating the runoff volume of the historical measured data of runoff-sediment-nitrogen and phosphorus by Nash efficiency coefficient, and analyzing the relative error of sediment concentration to obtain target function error analysis data;

[0135] synchronously optimizing the hydrological and water quality parameters by using the target function error analysis data and the mixed load output matrix to obtain multi-objective rating data; and rendering a heat map of the multi-objective rating data to obtain a pollution contribution rate distribution map.

[0136] In the embodiment of the present application, the hydrological and water quality combined observation data set containing different time periods, different monitoring points, runoff, sediment concentration, and nitrogen and phosphorus concentration is obtained, a standard format historical measured data matrix is established, and the data fields include time index, spatial position, measured value, and corresponding environmental attributes. In the runoff calibration aspect, the Nash efficiency coefficient (Nash-Sutcliffe Efficiency, NSE) is used as the goodness-of-fit evaluation index, and its expression is: Where Q o bs,i and Q sim,i are the observed value and the simulated value, respectively, is the mean of the observed value, and the reconstruction ability of the simulated runoff to the observed data is evaluated by the index. In the error analysis of sediment concentration, the relative error (Relative Error, RE) or the mean absolute percentage error (MAPE) is used to calculate the deviation degree between the observed value and the simulated value, the tolerance type error evaluation standard is established for the high volatility sediment load data, and the "objective function error analysis data" is constituted. Then, the system performs hydrological and water quality parameter joint optimization on the above error data and the generated "mixed load output matrix", adjusts the key parameters such as confluence time, infiltration coefficient, soil adsorption rate, and nitrogen and phosphorus deposition rate through multi-objective optimization methods (such as weighted least squares fitting, genetic algorithm, or MCMC sampling), and forms "multi-objective calibration data" with observation support constraints. Finally, in the data visualization stage, the spatial encoding is used to render the pollution contribution value of each cell, the standard color scale mapping model (such as RGB or HSV gradient conversion) is used, and the pollution source contribution of each sub-basin is expressed by color intensity through spatial projection in the two-dimensional map to generate a "pollution contribution rate distribution map", realizing the intuitive quantitative expression of the spatial distribution of pollutant load. The process starts from the measured data, gradually realizes the accuracy alignment of the simulation output and the actual situation, and provides a data basis for subsequent simulation verification and management evaluation.

[0137] Especially important is that step S52 comprises:

[0138] The pollution contribution rate distribution map is classified according to the Nash coefficient greater than 0.75 to obtain the watershed pollution classification data;

[0139] The watershed pollution classification data is divided into red warning according to the contribution degree greater than 30%, and the rest is green warning to obtain the pollution risk control data;

[0140] The pollution risk control data is graded and coded to obtain the watershed water pollution simulation verification data.

[0141] In the embodiment of the present application, the simulated pollutant concentration values corresponding to each spatial unit in the pollution contribution rate distribution map are compared with the measured concentration values point by point, the Nash efficiency coefficient (Nash-Sutcliffe Efficiency, NSE) is used as the model fitting precision discrimination criterion, the area with NSE greater than 0.75 is extracted as the high fitting degree unit, and the "basin pollution classification data" is obtained. The classification data is stored according to the spatial code and contains the sub-basin number, pollution type, simulated and observed concentration sequence and NSE value field. On this basis, the system extracts the pollution contribution rate of each unit in the pollution classification data. If the corresponding contribution value is higher than 30%, it is marked as "red warning", representing the significant pollution risk area; the rest of the area is marked as "green warning", constituting the "pollution risk control data". In order to enhance the traceability and structural integrity of the data, the system introduces a hierarchical coding mechanism, generates a coding string according to the pollution source type (such as agricultural source, domestic source, non-point source, etc.), warning level (red / green) and sub-basin number combination, for example, "R-AG-0508" represents red warning, agricultural source, sub-basin 0508 unit, and finally generates "basin water pollution simulation verification data". This data set uses spatial index and pollution level nested combination in structure, which is convenient for subsequent model backtracking, policy intervention and zoning governance strategy formulation, and at the same time provides standardized verification samples for pollution spatio-temporal evolution trend analysis. The whole process emphasizes the coupling mechanism of parameter screening driven by measured data, contribution quantitative classification and structured output, ensuring the consistency of simulation accuracy and risk identification.

[0142] In the present specification, a basin water pollution migration and transformation simulation system based on the MT-DHM model is provided: the basin water pollution migration and transformation simulation method based on the MT-DHM model is executed, and the basin water pollution migration and transformation simulation system based on the MT-DHM model comprises:

[0143] A basin data acquisition and standardization module is used for data acquisition on a multi-source space of a basin, and data preprocessing is performed to generate a hydrological response unit standard data set;

[0144] A vertical runoff and migration correction module is used for vertical layering calculation based on the hydrological response unit standard data set to obtain land runoff vertical analysis data; land pollution migration simulation is performed on the land runoff vertical analysis data, and real-time correction with an iteration step of 1h is performed to obtain a dissolved / adsorbed load matrix;

[0145] The water system migration and dilution level simulation module is used for push flow migration calculation of the dissolved / adsorbed load matrix, and combined simulation of dispersion and dilution levels to obtain basin dilution model level data; the basin dilution model level data is subjected to automatic load transfer of the multi-branch tree node to obtain an inter-node attenuation load matrix, wherein the inter-node attenuation load matrix includes an attenuation data sufficient area and an attenuation data deficient area;

[0146] The mechanism / statistical modeling and Bayesian fusion module is used for mechanism modeling of the attenuation data sufficient area to obtain pollutant load mechanism result data; statistical modeling of the attenuation data deficient area to obtain pollutant load statistical result data; Bayesian fusion of the pollutant load mechanism result data and the pollutant load statistical result data to obtain a mixed load output matrix;

[0147] The contribution rate calibration and qualified report generation module is used for multi-objective calibration of the mixed load output matrix to obtain a pollution contribution rate distribution map; precision verification of the pollution contribution rate distribution map to obtain basin water pollution simulation verification data; and construction of a Dianchi basin sediment / nutrient salt qualified report based on the basin water pollution simulation verification data.

[0148] The beneficial effects of the present application are that: through multi-source spatial data acquisition, DEM terrain data, land use data, soil parameters, rainfall and runoff measured data and other types of heterogeneous information are covered, and through unified projection, time alignment, rasterization and attribute reconstruction and other data cleaning means, a standard data set with consistent structure of hydrological response unit is formed, laying a high-quality data foundation for subsequent analysis. Secondly, through vertical layering calculation, continuous hydrological profile data from vegetation interception, soil infiltration to underground runoff are constructed, combined with measured parameters, pollutant migration simulation and real-time correction are realized, dynamic load calculation with 1 hour as a step is realized, and dissolved / adsorbed pollution load matrix is obtained. On this basis, through push flow migration and multi-path dispersion and dilution model, the transport and attenuation process of pollutants in the spatial dimension is modeled as a multi-branch tree structure, effectively expressing the load transfer and partition attenuation characteristics between upstream and downstream nodes. The modeling of attenuation data is divided into mechanism model and statistical model, and through Bayesian fusion technology, the structured and unstructured regional load data are jointly processed, which significantly improves the robustness and data adaptability of the simulation results. In the model output stage, combined with the historical runoff, sediment, nitrogen and phosphorus measured data, through the Nash efficiency coefficient, multi-objective calibration algorithm and other multi-criteria model evaluation methods, the simulation results are subjected to spatial distribution mapping and error optimization, so that the pollution contribution rate distribution has high credibility, and then the pollution risk early warning and standard verification evaluation are completed. Finally, the compliance classification output of sediment and nutrient salt indicators on the basin scale is realized, which provides direct data support and decision basis for environmental management and pollution control.

[0149] Therefore, the embodiments should be regarded, at any point, as being exemplary and not limiting, the scope of the application being defined by the appended claims and not by the above description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

[0150] The foregoing is considered as illustrative only of the principles of the application. Numerous modifications and changes will readily occur to those skilled in the art, and the generic principles defined herein can be applied to other embodiments without departing from the spirit or scope of the application. Therefore, the application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for simulating the migration and transformation of water pollution in a watershed based on the MT-DHM model, characterized in that: The following steps are involved: Step S1: Collect data from multiple source spaces in the basin and perform data preprocessing to generate a standard data set of hydrological response units; Step S2: Perform vertical layered calculation based on the hydrological response unit standard data set to obtain vertical analysis data of land runoff; The land runoff vertical analysis data were used to simulate the transport of pollutants over land, and real-time correction was performed with an iteration step of 1 hour to obtain the dissolved / adsorbed load matrix. Step S3: Perform push flow migration calculation on the dissolved / adsorbed load matrix and conduct a joint simulation of the dispersed dilution level to obtain the watershed dilution model level data; Automatic load transfer of the watershed dilution model level data to the multi-branch tree nodes is performed to obtain the inter-node attenuation load matrix, where the inter-node attenuation load matrix includes the attenuation data sufficient area and the attenuation data scarce area; Step S4: performing mechanism modeling on the area with sufficient attenuation data to obtain pollutant load mechanism result data; Conduct statistical modeling for areas lacking attenuation data to obtain pollutant load statistics; The pollutant load mechanism result data and the pollutant load statistical result data are fused by Bayesian method to obtain the mixed load output matrix; Step S5: Perform multi-objective calibration on the mixed load output matrix to obtain a pollution contribution rate distribution map; perform accuracy verification on the pollution contribution rate distribution map to obtain basin water pollution simulation verification data; and use the basin water pollution simulation verification data to construct a qualified report on sediment / nutrient salts for the Dianchi Lake basin.

2. The method for simulating water pollution migration and transformation in a watershed based on the MT-DHM model according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: Obtaining a hydrological dataset from a hydrological inspection station; Step S12: extract DEM data from the hydrological dataset at a resolution of 30m, perform watershed boundary analysis, and generate a sub-watershed division map; Step S13: extracting river network data from the hydrological dataset according to the river section slope, and performing topological relationship analysis to obtain a multi-branch tree node coding table; Step S14: Divide the hydrological dataset into land use / soil type data according to a 1km grid, and perform spatial overlay analysis to obtain the HRU unit attribute matrix; Step S15: Performing Thiessen polygon spatial interpolation on the hydrological dataset at the hourly level to generate a sub-basin area rainfall series; Step S16: construct soil adsorption parameters for the hydrological data set and perform laboratory calibration to obtain a watershed adsorption parameter table; Step S17: clean the sub-basin division map, multi-branch tree node coding table, HRU unit attribute matrix, sub-basin surface rainfall series and basin adsorption parameter table, and format and unify them to obtain the hydrological response unit standard data set.

3. The method for simulating water pollution migration and transformation in a watershed based on the MT-DHM model according to claim 1, characterized in that: The vertical layered calculation based on the hydrological response unit standard data set in step S2 includes: Extract rainfall data from the hydrological response unit standard dataset to obtain basin rainfall data; perform runoff process analysis on the hydrological response unit standard dataset to obtain land basin runoff data; The basin rainfall data and land basin runoff data are vertically layered and calculated to obtain land runoff vertical analysis data, wherein the vertical layering calculation includes: based on the leaf area index LAI of the vegetation interception layer being greater than 2.0, the interception capacity of the basin rainfall data and land basin runoff data vegetation interception layer is calculated to generate the falling rainfall data; based on the saturation rate of the upper soil layer being greater than 40%, the infiltration analysis of the basin rainfall data and land basin runoff data vegetation interception layer is performed to obtain the land basin infiltration water component data; The vertical data of land runoff analysis are obtained by integrating the rainfall data and the water infiltration component data of the land basin.

4. The method for simulating water pollution migration and transformation in a watershed based on the MT-DHM model according to claim 1, characterized in that: In step S2, the land runoff vertical analysis data is simulated for land pollutant migration, and a real-time correction is performed with an iterative step length of 1 hour, including: Get soil permeability coefficient; The leaching rate of dissolved nitrogen or nitrate nitrogen is calculated based on the soil permeability coefficient of the vertical analysis data of land runoff to generate the dissolved nitrogen migration flux; The vertical analysis data of land runoff were used to calculate the stripping amount of adsorbed phosphorus according to the slope of soil shear force greater than 5° to obtain the adsorbed migration amount; The dissolved state migration flux and the adsorbed state migration flux were merged at the adsorption level, and a real-time correction was performed with an iterative step length of 1 h to obtain the dissolved state / adsorbed state loading matrix.

5. The method for simulating water pollution migration and transformation in a watershed based on the MT-DHM model according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: performing a flow migration calculation on the dissolved / adsorbed load matrix, and calculating the flow velocity based on the roughness of the Manning formula of 0.05 to obtain the inflow flow flux data; Step S32: performing a dispersed dilution level joint simulation based on the inflow flow rate data to obtain watershed dilution model level data; Step S33: Automatically transfer the load of the multi-branch tree nodes to the watershed dilution model hierarchical data to obtain an inter-node attenuation load matrix, wherein the inter-node attenuation load matrix includes an attenuation data sufficient area and an attenuation data scarce area.

6. The method for simulating water pollution migration and transformation in a watershed based on the MT-DHM model according to claim 5, characterized in that: Step S32 includes the following steps: Step S321: Diffusion coefficient of the inflow flux data is 10 ―5 m 2 / s molecular diffusion Fick law calculation, to obtain molecular diffusion dilution data; Step S322: performing eddy viscosity coefficient correction on the inflow flow rate data when the Reynolds number is greater than 4000 to obtain turbulent diffusion dilution data; Step S323: performing shear flow discrete analysis on the inflow flow data with a channel curvature greater than 1.2 to obtain diffusion dilution data; Step S324: Integrate the molecular diffusion dilution data, turbulent diffusion dilution data, and diffusion dilution data to obtain watershed dilution model level data.

7. The method for simulating water pollution migration and transformation in a watershed based on the MT-DHM model according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: performing mechanism modeling on the area with sufficient attenuation data to obtain pollutant load mechanism result data; Step S42: For areas lacking attenuation data, a multivariate regression model is constructed where the forest land accounts for more than 70%, and forest land load is estimated to obtain pollutant load statistical result data; Step S43: Perform Bayesian fusion on the pollutant load mechanism result data and the pollutant load statistical result data to obtain a mixed load output matrix.

8. The method for simulating water pollution migration and transformation in a watershed based on the MT-DHM model according to claim 7 is characterized in that: Step S41 includes the following steps: Step S411: performing nitrification analysis on the temperature greater than 15°C based on the sufficient attenuation data area to obtain the ammonia nitrogen conversion amount; Step S412: performing denitrification analysis on DO less than 2 mg / L based on the sufficient decay data area, and using first-order kinetic processing to obtain denitrification data; Step S413: The ammonia nitrogen conversion and denitrification data are collected according to the monitoring point density greater than 1 / 100km. 2 Conduct mechanism modeling and obtain pollutant load mechanism result data.

9. The method for simulating water pollution migration and transformation in a watershed based on the MT-DHM model according to claim 1, characterized in that: Step S5 includes the following steps: Step S51: performing multi-objective calibration on the mixed load output matrix to obtain multi-objective calibration data; performing heat map rendering on the multi-objective calibration data to obtain a pollution contribution rate distribution map; Step S52: Verify the accuracy of the pollution contribution rate distribution map to obtain watershed water pollution simulation verification data; Step S53: Using the water pollution simulation verification data of the Dianchi Lake Basin to construct a qualified report on sediment / nutrients.

10. A watershed water pollution migration and transformation simulation system based on the MT-DHM model, characterized by: The method for simulating water pollution migration and transformation in a watershed based on the MT-DHM model according to claim 1 is executed. The watershed water pollution migration and transformation simulation system based on the MT-DHM model comprises: The watershed data collection and standardization module is used to collect data from multiple sources in the watershed, perform data preprocessing, and generate a standard data set for hydrological response units; The vertical runoff and migration correction module is used to perform vertical stratification calculations based on the hydrological response unit standard data set to obtain vertical analysis data of land runoff; simulate the migration of terrestrial pollutants on the vertical analysis data of land runoff, and perform real-time correction with an iterative step length of 1 hour to obtain the dissolved / adsorbed load matrix; The water system migration and dilution level simulation module is used to perform flow migration calculations on the dissolved / adsorbed load matrix and conduct a joint simulation of the dispersed dilution level to obtain the watershed dilution model level data. The watershed dilution model level data is automatically transferred to the multi-branch tree nodes to obtain the inter-node attenuation load matrix, where the inter-node attenuation load matrix includes areas with sufficient attenuation data and areas with insufficient attenuation data. The mechanism / statistical modeling and Bayesian fusion module is used to perform mechanism modeling on areas with sufficient attenuation data to obtain pollutant load mechanism result data; perform statistical modeling on areas with insufficient attenuation data to obtain pollutant load statistical result data; and perform Bayesian fusion on the pollutant load mechanism result data and the pollutant load statistical result data to obtain a mixed load output matrix. The contribution rate calibration and qualified report generation module is used to perform multi-objective calibration on the mixed load output matrix to obtain a pollution contribution rate distribution map; the pollution contribution rate distribution map is used for accuracy verification to obtain basin water pollution simulation verification data; the basin water pollution simulation verification data is used to construct a qualified report on sediment / nutrients in the Dianchi Lake basin.

Images