Water quality and quantity accurate simulation method and system based on swat model and storage medium

Abstract

The application relates to the technical field of data processing, and discloses a water quality and quantity accurate simulation method and system based on a SWAT model and a storage medium. The method comprises the following steps: analyzing precipitation and evapotranspiration data through double-parameter trend test analysis, and generating standard hydrological data by adopting a triple configuration algorithm; combining terrain information to divide sub-basins, and calculating hydrological response units; performing water balance operation and water source conservation evaluation; analyzing pollutant diffusion and solving diffusion equations; implementing operation state monitoring and designing an early warning system; comparing and correcting measured data, and constructing a hierarchical monitoring mechanism. The application realizes accurate management of the whole process from hydrological data collection to water quality evaluation to early warning monitoring, and significantly improves the accuracy and reliability of water quality and quantity monitoring.

Description

Technical Field

[0001] This application relates to the field of data processing, and in particular to a method, system and storage medium for accurate simulation of water quality and quantity based on the SWAT model. Background Technology

[0002] Current water quality and quantity monitoring methods primarily employ traditional approaches such as fixed-point sampling and analysis, and manual inspections, while also incorporating technologies like remote sensing and online monitoring to monitor water quality and quantity in water source areas. Water quality assessment mainly relies on physicochemical index analysis and biological monitoring to evaluate pollutant concentrations and eutrophication levels in water bodies. Water quantity regulation primarily utilizes hydrological station networks and precipitation monitoring networks to collect hydrological data, which is then analyzed and predicted using hydrological models. Commonly used hydrological models include the SWAT model and the MIKE model, which play a crucial role in simulating and predicting hydrological processes.

[0003] However, existing technologies have the following shortcomings: traditional monitoring methods often suffer from insufficient spatiotemporal resolution of monitoring data and incomplete coverage of monitoring points; water quality assessment methods struggle to accurately describe the spatial diffusion patterns and temporal evolution characteristics of pollutants in water bodies; in terms of water quantity regulation, there is a lack of precise assessment methods for water source conservation capacity, making it difficult to achieve coordinated optimization of water quality and quantity; although existing hydrological models can simulate hydrological processes, their simulation accuracy and prediction accuracy still need improvement when faced with complex water quality and quantity coupling problems; at the same time, existing monitoring and early warning systems lack dynamic correction and accuracy assessment mechanisms for monitoring data, making it difficult to guarantee the reliability of monitoring results. Summary of the Invention

[0004] This application provides a method, system, and storage medium for accurate simulation of water quality and quantity based on the SWAT model, which solves the data accuracy problem in the coupled optimization process of water quality and quantity in water source areas. By establishing a complete data processing system, it achieves precise management of the entire process from hydrological data acquisition and water quality assessment to early warning monitoring, significantly improving the accuracy and reliability of water quality and quantity monitoring.

[0005] Firstly, this application provides a method for accurate simulation of water quality and quantity based on the SWAT model, the method comprising:

[0006] The spatiotemporal characteristics of multi-source precipitation and evapotranspiration data are analyzed by dual-parameter trend test and slope evaluation to obtain initial hydrological data. The initial hydrological data are then processed by a triple configuration algorithm to generate a standard hydrological dataset.

[0007] Based on the standard hydrological dataset, regional hydrological division is carried out in combination with topographic elevation information to generate sub-basin division data. Based on the sub-basin division data, hydrological response units of sub-basins are calculated to construct a basic hydrological calculation dataset.

[0008] Water balance calculations are performed using the basic hydrological calculation dataset to extract water conservation characteristic data. The water conservation characteristic data is then evaluated in conjunction with land use change to generate water conservation assessment data.

[0009] Based on the water source conservation assessment data, pollutant diffusion analysis of inflow water is performed, water quality change prediction data is output, the water quality change prediction data is integrated and the pollutant diffusion equation is solved to establish water quality distribution data.

[0010] Based on the water quality distribution data, reservoir operation status monitoring is carried out to obtain pollution diffusion trend data; based on the pollution diffusion trend data, an early warning indicator system is designed and early warning threshold data is determined.

[0011] The warning threshold data is compared and corrected with the measured data to establish accuracy assessment data. A hierarchical monitoring mechanism is constructed based on the accuracy assessment data to output water quality and quantity monitoring results.

[0012] Secondly, this application provides a precise water quality and quantity simulation system based on the SWAT model, the precise water quality and quantity simulation system based on the SWAT model includes:

[0013] The acquisition module is used to analyze the spatiotemporal characteristics of multi-source precipitation and evapotranspiration data through dual-parameter trend testing and slope evaluation, acquire initial hydrological data, process the initial hydrological data using a triple configuration algorithm, and generate a standard hydrological dataset.

[0014] The partitioning module is used to perform regional hydrological partitioning based on the standard hydrological dataset and combined with topographic elevation information, generate sub-basin partitioning data, calculate the hydrological response units of the sub-basins based on the sub-basin partitioning data, and construct a basic hydrological calculation dataset.

[0015] The calculation module is used to perform water balance calculations using the basic hydrological calculation dataset, extract water conservation characteristic data, evaluate the water conservation characteristic data with land use change, and generate water conservation assessment data.

[0016] The analysis module is used to perform pollutant diffusion analysis of the inflow water body based on the water source conservation assessment data, output water quality change prediction data, integrate the water quality change prediction data and solve the pollutant diffusion equation to establish water quality distribution data.

[0017] The monitoring module is used to monitor the reservoir's operational status based on the water quality distribution data, obtain pollution diffusion trend data, design an early warning indicator system based on the pollution diffusion trend data, and determine early warning threshold data.

[0018] The calibration module is used to compare and calibrate the warning threshold data with the measured data, establish accuracy assessment data, construct a hierarchical monitoring mechanism based on the accuracy assessment data, and output water quality and quantity monitoring results.

[0019] A third aspect of this application provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the aforementioned method for accurate simulation of water quality and quantity based on the SWAT model.

[0020] The technical solution provided in this application analyzes the spatiotemporal characteristics of multi-source precipitation and evapotranspiration data through dual-parameter trend testing and slope assessment. A triple-configuration algorithm is used to process the initial hydrological data, ensuring its accuracy and reliability. Regional hydrological division is performed based on standard hydrological datasets and topographic elevation information, generating sub-basin division data and calculating hydrological response units, thus improving the spatial accuracy of hydrological analysis. Water balance calculations are performed using basic hydrological calculation datasets to extract water conservation characteristic data and evaluate it in conjunction with land use change data, achieving accurate assessment of water conservation capacity. Pollutant diffusion analysis of inflow water is conducted based on water conservation assessment data, and water quality distribution data is established by solving pollutant diffusion equations, improving the accuracy of water quality assessment. Reservoir operation status monitoring is implemented based on water quality distribution data, acquiring pollution diffusion trend data and designing an early warning indicator system, enhancing the real-time nature of water quality monitoring. Accuracy assessment data is established by comparing and correcting early warning threshold data with measured data, and a hierarchical monitoring mechanism is constructed, ensuring the reliability of monitoring results. Attached Figure Description

[0021] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of an embodiment of the water quality and quantity accurate simulation method based on the SWAT model in this application.

[0023] Figure 2 This is a schematic diagram of an embodiment of the water quality and quantity accurate simulation system based on the SWAT model in this application. Detailed Implementation

[0024] This application provides a method, system, and storage medium for accurate simulation of water quality and quantity based on a SWAT model. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0025] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the water quality and quantity accurate simulation method based on the SWAT model in this application includes:

[0026] Step S101: Analyze the spatiotemporal characteristics of multi-source precipitation and evapotranspiration data through dual-parameter trend test and slope evaluation to obtain initial hydrological data. Process the initial hydrological data using a triple configuration algorithm to generate a standard hydrological dataset.

[0027] Step S102: Based on the standard hydrological dataset and combined with topographic elevation information, regional hydrological division is carried out to generate sub-basin division data. Based on the sub-basin division data, the hydrological response units of the sub-basins are calculated to construct the basic hydrological calculation dataset.

[0028] Step S103: Perform water balance calculation using the basic hydrological calculation dataset, extract water conservation characteristic data, evaluate the water conservation characteristic data and land use change, and generate water conservation assessment data.

[0029] Step S104: Based on the water source conservation assessment data, conduct pollutant diffusion analysis of the water body entering the reservoir, output water quality change prediction data, integrate the water quality change prediction data and solve the pollutant diffusion equation to establish water quality distribution data.

[0030] Step S105: Monitor the reservoir's operational status based on water quality distribution data to obtain pollution diffusion trend data; design an early warning indicator system based on the pollution diffusion trend data and determine the early warning threshold data.

[0031] Step S106: Compare and correct the early warning threshold data with the measured data, establish accuracy assessment data, construct a hierarchical monitoring mechanism based on the accuracy assessment data, and output the water quality and quantity monitoring results.

[0032] It is understood that the executing entity of this application can be a precise water quality and quantity simulation system based on the SWAT model, or it can be a terminal or a server; the specific implementation is not limited here. This application's embodiment uses a server as an example for illustration.

[0033] Specifically, spatiotemporal characteristics of multi-source precipitation and evapotranspiration data were analyzed using a two-parameter trend test and slope assessment. The two-parameter trend test employed the Mann-Kendall method, which calculates the S-statistic for time-series data and then standardizes it to obtain the Z-statistic, thereby determining the trend of the data series. The slope assessment used the Sen slope method, estimating the overall rate of change by calculating the median slope between any two points in the time series. The CMFD, CHMPRE, and MSWEP precipitation datasets were time-aligned, and data fusion was achieved using an extended triple configuration method. This extended triple configuration method establishes data quality assessment indicators, assigns appropriate weights to different data sources, and performs a weighted average calculation. The same method was used for the ERA5Land, CR, and GLEAM evapotranspiration datasets. During data fusion, spatiotemporal consistency was first checked in each dataset to remove outliers, and then weight configuration was used to achieve effective integration of multi-source data.

[0034] The flow direction was extracted based on topographic elevation information to calculate cumulative discharge and determine the river network structure. The D8 algorithm was used for flow direction extraction, dividing the study area into a regular grid where the flow direction of each grid cell points to the lowest elevation among its eight neighboring cells. Cumulative discharge calculation was based on a recursive algorithm, accumulating catchment areas from upstream to downstream. Sub-basin division employed a critical catchment area threshold method, setting appropriate thresholds based on the topographic characteristics of different regions to complete the division of 109 sub-basins. The calculation of hydrological response units was based on factors such as land use type, soil properties, and slope, using overlay analysis to identify regional units with similar hydrological characteristics. Land use type data came from remote sensing image interpretation; soil properties included physicochemical properties such as texture, organic matter content, and permeability; and slope was calculated using a digital elevation model. These data were spatially overlaid using a geographic information system to generate a distribution map of hydrological response units.

[0035] Water balance calculations are based on the principle of water conservation, considering hydrological elements such as precipitation, runoff, evaporation, and infiltration to establish a water balance equation. During the extraction of water conservation characteristics, land use change data is incorporated to analyze the impact of different land use types on water conservation and calculate the conservation capacity index. The water balance equation considers three components: surface runoff, interflow, and groundwater recharge. The spatiotemporal distribution of each hydrological element is obtained by solving a continuity equation. A pollutant diffusion equation is established, considering convection diffusion and degradation processes. The pollutant concentration distribution is obtained through numerical solutions. The pollutant diffusion equation is solved using the finite difference method, discretizing the spatial and temporal dimensions, and obtaining a steady-state solution through iterative calculation. The parameters in the diffusion equation are calibrated using measured data, including key parameters such as the diffusion coefficient and degradation coefficient.

[0036] A monitoring network is established based on water quality distribution data to analyze pollutant diffusion trends through real-time data acquisition. The network layout considers water flow characteristics and pollutant diffusion patterns, with monitoring points set up at key cross-sections. The early warning indicator system is designed considering multiple dimensions, including water quality parameters, diffusion rates, and impact ranges, and employs a multi-level grading standard to determine early warning thresholds. Early warning indicators include routine water quality indicators and characteristic pollutant indicators; the early warning thresholds for each indicator are determined through statistical analysis. An accuracy assessment system is established by comparing predicted values ​​with measured data and calculating error statistics. Error statistics use multiple indicators, such as root mean square error and relative error, to comprehensively evaluate prediction accuracy. A tiered monitoring mechanism sets monitoring frequencies and response measures according to different accuracy levels to ensure the reliability of monitoring results. The monitoring mechanism includes two levels: routine monitoring and emergency monitoring, with monitoring frequencies dynamically adjusted based on water quality changes.

[0037] Taking the Guanshan River basin, a typical tributary flowing into the reservoir, as an example, precipitation data was processed. Significant trends were identified through a two-parameter trend test, and the rate of change was determined by slope assessment. In the sub-basin division process, the study area was divided into multiple sub-basins based on topographic features, and each sub-basin was further subdivided into several hydrological response units. Water balance calculations revealed quantitative relationships between various hydrological elements, and these relationships were used to infer water conservation capacity. Pollutant diffusion analysis, based on measured data, established a diffusion model to predict the migration and transformation processes of pollutants in the water body. Accuracy assessment of the monitoring data showed that this method can accurately reflect the dynamic changes in water quality and quantity, providing a scientific basis for water environment management.

[0038] For example, in the delineation of hydrological response units, the critical catchment area threshold is determined using measured hydrological data; in pollutant diffusion analysis, the diffusion coefficient and degradation coefficient are calibrated using experimental data. This parameter calibration method based on measured data ensures the reliability of the simulation results. Simultaneously, multi-source data fusion and multi-scale analysis improve simulation accuracy. During the data fusion phase, outliers are eliminated through consistency checks; during parameter calibration, multi-objective optimization methods are used to determine the optimal parameter combination; and during model validation, cross-validation is performed using independent datasets. These quality control measures ensure the standardization of the method implementation and the reliability of the results.

[0039] In this embodiment, the spatiotemporal characteristics of multi-source precipitation and evapotranspiration data are analyzed through dual-parameter trend testing and slope assessment. A triple-configuration algorithm is used to process the initial hydrological data, ensuring its accuracy and reliability. Regional hydrological division is performed based on standard hydrological datasets and topographic elevation information, generating sub-basin division data and calculating hydrological response units, thus improving the spatial accuracy of hydrological analysis. Water balance calculations are performed using basic hydrological calculation datasets to extract water conservation characteristic data and evaluate it in conjunction with land use change data, achieving accurate assessment of water conservation capacity. Pollutant diffusion analysis of inflow water is conducted based on water conservation assessment data, and water quality distribution data is established by solving pollutant diffusion equations, improving the accuracy of water quality assessment. Reservoir operation status monitoring is implemented based on water quality distribution data, acquiring pollution diffusion trend data and designing an early warning indicator system, enhancing the real-time nature of water quality monitoring. Accuracy assessment data is established by comparing and correcting early warning threshold data with measured data, and a hierarchical monitoring mechanism is constructed, ensuring the reliability of monitoring results.

[0040] In one specific embodiment, the process of performing step S101 may specifically include the following steps:

[0041] (1) Extract time weight coefficients from multi-source precipitation and evapotranspiration data, use weighted filtering to eliminate random disturbance components, and generate a time series matrix;

[0042] (2) Perform least squares operation on the time series matrix, calculate the mutation detection threshold, construct a trend test sequence based on the MK statistic, and output the significance level;

[0043] (3) Based on the significance level as the benchmark parameter, perform segmented calculation of the time series rate of change, use Sen slope to assess and determine the trend change, and generate spatiotemporal characteristic parameters through quantile correction;

[0044] (4) Based on the spatiotemporal characteristic parameters as constraints, perform Kriging spatial interpolation to calculate the spatial autocorrelation coefficient, and output the initial hydrological data through co-evolution analysis;

[0045] (5) Perform heteroscedasticity test matrix operation on the initial hydrological data, remove outliers by combining Mahalanobis distance discrimination method, and supplement the data by Bayesian estimation;

[0046] (6) Perform triple configuration weight calculation on the processed hydrological data, perform adaptive fusion iterative operation, and generate a standard hydrological dataset through normalization processing.

[0047] Specifically, time-weighted coefficients were extracted from multi-source precipitation and evapotranspiration data. The multi-source data included three precipitation datasets (CMFD, CHMPRE, and MSWEP) and three evapotranspiration datasets (ERA5Land, CR, and GLEAM). The time-weighted coefficient extraction process performed a temporal consistency check on each dataset. By calculating the rate of change of data between adjacent time points, points with abnormal fluctuations were assigned lower weights, while areas of stable change were assigned higher weights. Weighted filtering employed a sliding window method, with the window size set according to the data sampling frequency. A weighted average was calculated for the data within each time window to eliminate the influence of random disturbances. After these two steps, the original data was transformed into a matrix form with temporal attributes.

[0048] Least squares operations are performed on the time series matrix to identify abrupt changes in the data sequence. The least squares operation fits a trend line to the time series and calculates the sum of squared residuals between the actual and fitted values. A point where the residual exceeds a preset threshold is considered a mutation point. The mutation detection threshold is determined based on the statistical characteristics of historical data, using the MK statistic to construct a trend test sequence. During the MK statistic calculation, pairwise comparisons of time series data are performed, counting the number of increases or decreases, and then standardization is applied to obtain the Z-value, outputting the significance level. With the significance level as a baseline parameter, the time series rate of change is calculated piecewise. The time series is divided into different change stages based on the significance level, and the rate of change for each stage is calculated separately. The Sen slope assessment method determines the overall trend change by calculating the median slope between all possible points in the time series. Quantile correction eliminates the influence of extreme values ​​by calculating the rate of change at different quantiles to obtain spatiotemporal characteristic parameters. Based on the obtained spatiotemporal characteristic parameters, spatial interpolation is performed. Kriging spatial interpolation is an optimal linear unbiased estimation method based on regionalized variable theory. It calculates the spatial correlation between the interpolation point and known points by constructing a variogram model. The spatial autocorrelation coefficient is calculated using the Moran index method to quantify the relationships between spatial units. Co-evolutionary analysis, on the other hand, constructs a spatial correlation matrix to analyze the changing patterns of hydrological elements in different regions and outputs initial hydrological data.

[0049] Further processing of the initial hydrological data includes heteroscedasticity testing and outlier removal. Heteroscedasticity testing uses matrix operations to calculate the variance at different spatial locations, thus testing for spatial heterogeneity. Mahalanobis distance is a multidimensional distance calculation method that considers the correlation between variables; it identifies and removes outliers by setting a threshold. For data gaps after outlier removal, Bayesian estimation is used to fill the gaps, estimating the optimal solution for the missing values ​​based on the known distribution characteristics of the data.

[0050] The triple configuration algorithm is a multi-source data fusion method based on data quality, spatiotemporal distribution characteristics, and physical consistency. Weight calculation considers three aspects: data accuracy, spatiotemporal coverage, and physical rationality, and continuously optimizes the weight coefficients through adaptive iteration. Normalization transforms data from different sources into a unified scale, generating a standard hydrological dataset.

[0051] For example, consider the precipitation data processing for a specific study area. Precipitation data are obtained from three data sources: CMFD, CHMPRE, and MSWEP. Time-weighting coefficients are used to determine the different variation characteristics of the data during the flood season and non-flood season, assigning different weighting coefficients accordingly. Weighted filtering smooths out significant anomalous fluctuations. Least squares fitting and the MK test show a significant upward trend in the data on an interannual scale, and the specific rate of change is determined using Sen slope assessment. During spatial interpolation, the influence of topographic factors is considered, and the Kriging method is used to interpolate between stations. Heteroscedasticity testing reveals differences in precipitation variability between mountainous and plain areas, and outliers that do not conform to regional precipitation patterns are identified using Mahalanobis distance. A triple-configuration algorithm fuses the processed multi-source data to obtain a spatiotemporally continuous and physically reasonable standard precipitation dataset.

[0052] In one specific embodiment, the process of performing step S102 may specifically include the following steps:

[0053] (1) Multiply the precipitation data in the standard hydrological dataset with the topographic elevation information to calculate the water yield per unit area, divide by the runoff time to obtain the initial flow data; spatially overlay the initial flow data to generate a runoff map;

[0054] (2) Extract the confluence path from the water flow convergence map, calculate the cumulative flow based on the flow direction relationship matrix, multiply by the slope coefficient, divide by the critical threshold, and output the coordinates of the sub-basin boundary points; connect the coordinates of the sub-basin boundary points into a line to generate sub-basin division data;

[0055] (3) Using the sub-basin division data as the boundary, extract the area ratio of land use type of each sub-basin, multiply it by the permeability coefficient, and calculate the underlying surface characteristic value; combine the underlying surface characteristic value with soil parameters to obtain the hydrological response parameters;

[0056] (4) Classify and statistically analyze the hydrological response parameters, calculate the area weight of different response types, multiply by the hydrological characteristic coefficient, and output the distribution of hydrological response units; superimpose the distribution of hydrological response units with the sub-basin boundary to generate spatial data of response units;

[0057] (5) Calculate the runoff coefficient of each unit using the spatial data of the response unit, multiply the runoff coefficient by the precipitation, divide by the area to obtain the unit runoff; perform spatial integration on the unit runoff and output the watershed water volume data.

[0058] (6) Combine the watershed water volume data with the hydrological parameter relationship formula, calculate the hydrological response coefficient matrix, establish the water balance equation, and generate the basic hydrological calculation dataset.

[0059] Specifically, precipitation data from the standard hydrological dataset is processed and combined with topographic elevation information to calculate water yield per unit area. Initial flow data is obtained by multiplying the precipitation value of each grid cell by the corresponding elevation correction factor and then dividing by the runoff time of that grid. Runoff time is the time required for water to flow from a point to the outlet section, calculated using slope and flow path length. Spatially overlaying these initial flow data generates a runoff map reflecting the water flow convergence across the entire region.

[0060] Based on the water flow convergence map, confluence paths are extracted, and the D8 algorithm is used to determine the flow direction of each grid cell. A flow direction relationship matrix records the water flow transmission relationship between each grid cell and its eight neighboring grid cells, and cumulative flow is calculated. Multiplying the cumulative flow by a slope coefficient reflecting the degree of topographic relief, and then dividing by a critical threshold determined based on the characteristics of the study area, yields the coordinates of the sub-basin boundary points. These boundary points are vectorized and connected to form closed boundary lines, forming the sub-basin delineation data. After the sub-basin delineation is completed, land use information within each sub-basin needs to be extracted. Land use types include different categories such as forest land, cultivated land, and grassland, each with its specific permeability coefficient. By calculating the area proportion of each land type and multiplying it by the corresponding permeability coefficient, a comprehensive parameter reflecting the underlying surface characteristics is obtained. This underlying surface characteristic value is then combined with soil physical parameters (such as porosity, bulk density, and permeability) to generate hydrological response parameters.

[0061] The hydrological response parameters were processed using a classification and statistical method. The parameters were categorized into levels based on their distribution characteristics, and the area weight of each level was calculated. This weight was then multiplied by a coefficient reflecting the characteristics of the hydrological process to obtain the spatial distribution of the hydrological response units. This distribution information was then overlaid with the previously obtained sub-basin boundary data to form a spatial dataset of the response units.

[0062] Based on the spatial data of the response units, the runoff coefficient for each unit is calculated. The runoff coefficient is the proportion of rainfall converted into surface runoff, which is influenced by factors such as land use type, soil properties, and topographic slope. Multiplying the runoff coefficient by the rainfall and dividing by the unit area yields the runoff per unit area. Spatially integrating the runoff per unit area for all units provides the water volume data for the entire watershed.

[0063] The relationship between watershed water volume data and various hydrological parameters is integrated to calculate the hydrological response coefficient matrix. The water balance equation, considering the relationships between precipitation, evaporation, infiltration, and surface runoff, can be expressed as:

[0064]

[0065] in: This is the water balance value; Let be the precipitation correction factor for the i-th unit; Let be the precipitation in the i-th unit; Let be the evaporation coefficient of the i-th unit; Let be the evaporation amount of the i-th unit; Let be the infiltration coefficient of the i-th unit; Let be the infiltration amount of the i-th unit; Let be the runoff coefficient of the i-th unit; Let be the runoff volume of the i-th unit; Let be the area of ​​the i-th unit; Let be the coefficient of the j-th groundwater exchange term; Let be the j-th groundwater exchange volume.

[0066] For example, hydrological processes are analyzed. Initial flow is calculated based on precipitation and DEM data, and major confluence paths are determined through cumulative flow analysis. During sub-basin delineation, the study area is divided into multiple sub-basins considering topographic features. In land use analysis, the distribution of various land types is extracted, and hydrological response parameters are determined in conjunction with soil survey data. The hydrological characteristic data of the study area are obtained by solving the water balance equation, providing a foundation for subsequent water quality simulation. The accuracy and continuity of data processing are crucial throughout the entire hydrological unit delineation and water balance calculation process. Standard hydrological datasets serve as the basic data source, containing time-series precipitation data that has undergone pre-processing and standardization. Topographic elevation information comes from the Digital Elevation Model (DEM), reflecting the three-dimensional topographic features of the study area. When multiplying these two types of data, a spatial overlay analysis method is used to ensure spatial registration and unit consistency.

[0067] The analysis of the water flow convergence process employed a cumulative flow calculation method. A flow direction matrix was constructed, with the flow direction of each grid cell determined by the direction of maximum slope. The cumulative flow was calculated using a recursive algorithm, progressively accumulating the contributing area from upstream to downstream. The determination of the critical threshold considered the size of the study area and the characteristics of the river system, optimizing the threshold value through trial and error. The sub-basin boundaries were determined using a tracing algorithm, starting from the watershed and tracing along contour lines until the point of confluence with other tributaries. Land use types were extracted using remote sensing image interpretation results, including different types such as forest, farmland, grassland, and water bodies. The permeability coefficient, an important parameter reflecting the underlying surface's ability to infiltrate water, was determined through field testing or by referencing standard values. The calculation of underlying surface characteristic values ​​comprehensively considered factors such as surface cover, soil structure, and topographic slope, forming a comprehensive index.

[0068] The division of hydrological response units is a multi-factor comprehensive analysis process. Cluster analysis of hydrological response parameters is performed to determine the main response types. The area weight is calculated based on the spatial distribution characteristics of each type, reflecting the importance of different response types in the study area. Hydrological characteristic coefficients are obtained from the analysis of historical hydrological data, reflecting the runoff generation and confluence characteristics of different regions. Runoff generation calculation is the core of hydrological simulation. The determination of runoff generation coefficients needs to consider multiple influencing factors, including rainfall intensity, anterior soil moisture content, and vegetation cover. The calculation of unit runoff generation adopts a distributed method to fully consider the impact of spatial heterogeneity. A grid overlay method is used in the spatial integration process to ensure the spatial continuity of the calculation results.

[0069] The construction of the hydrological response coefficient matrix reflects the interrelationships among various hydrological elements. Each element in the matrix represents the transformation relationship between different hydrological processes, and its specific value is determined through parameter calibration. The water balance equation is established following the principle of water conservation, considering all water input and output terms within the system.

[0070] Taking a large reservoir basin as an example, ten years of precipitation data and 90-meter resolution DEM data were collected for the basin. Initial flow data were calculated through spatial analysis. Based on cumulative flow analysis, the confluence paths of major rivers were determined, and the entire basin was divided into multiple sub-basins based on a critical catchment area threshold of 500 square kilometers. Land use analysis showed that the study area mainly included three types: forest, farmland, and grassland. The permeability coefficient of each type was determined through field sampling. The division of hydrological response units considered three main factors: slope, soil, and vegetation, forming spatial distribution data. The runoff generation coefficient was calculated based on rainfall intensity and underlying surface characteristics, and the reliability of the calculation results was verified through measured data. By solving the water balance equation, hydrological cycle process data were obtained.

[0071] In one specific embodiment, the process of executing step S103 may specifically include the following steps:

[0072] (1) First, multiply the daily precipitation data in the basic hydrological calculation dataset by the basin area, subtract the product of daily evaporation and daily runoff, divide by the total water volume, and calculate the daily water balance value; then, accumulate the daily water balance value monthly to obtain the monthly water volume regulation data.

[0073] (2) Calculate the monthly water conservation coefficient by multiplying the monthly average soil infiltration rate by the monthly water volume regulation data, subtracting the monthly groundwater recharge, and then dividing by the total monthly precipitation; multiply the monthly water conservation coefficient by the underlying surface characteristic value to obtain the initial value of water conservation.

[0074] (3) Calculate the daily average value of the initial value of water conservation capacity, compare the calculation results with the upper limit of soil water holding capacity, and determine the effective conservation depth; calculate the water conservation flux based on the effective conservation depth and extract water conservation characteristic data;

[0075] (4) Classify and statistically analyze the water conservation characteristics data according to cultivated land, forest land and grassland, and calculate the average annual conservation amount of different land use types; standardize the average annual conservation amount to obtain the land use conversion rate;

[0076] (5) Based on the land use conversion rate, multiply the changes in the area of ​​various types of land use by their corresponding conservation contributions to obtain the total annual conservation change; perform standard deviation analysis on the total annual conservation change to output the water conservation stability index.

[0077] (6) Substitute the water conservation stability index into the water conservation capacity assessment function to calculate the regional water conservation capacity value and generate water conservation assessment data.

[0078] Specifically, the basic hydrological calculation dataset includes daily precipitation data, watershed area data, daily evaporation data, and daily runoff data. By processing and calculating these basic data, water conservation assessment data is obtained, thereby enabling accurate simulation of water quality and quantity.

[0079] Calculating the daily water balance value requires multiplying the daily precipitation data by the catchment area to obtain the total regional precipitation. Then, the product of daily evaporation and daily runoff is subtracted from this total precipitation; this step reflects the regional water budget. Dividing the result by the total water volume yields the daily water balance value. The daily water balance value reflects the balance of water input and output within a single day. These daily data are then accumulated by month to form monthly water regulation data, providing a foundation for subsequent analysis.

[0080] After obtaining monthly water regulation data, the impact of soil infiltration and groundwater recharge needs to be considered. The monthly average soil infiltration rate is an important indicator of soil's water absorption capacity. Multiplying the monthly water regulation data by the monthly average soil infiltration rate, subtracting the monthly groundwater recharge, and then dividing by the total monthly precipitation yields the monthly water conservation coefficient. This coefficient reflects the region's water regulation capacity on a monthly scale. Multiplying the monthly water conservation coefficient by the underlying surface characteristic value yields the initial value of water conservation capacity. The underlying surface characteristic value includes the influence of factors such as land cover type and vegetation status on water conservation capacity.

[0081] Calculating the daily average value of the initial water conservation capacity is to obtain more detailed time-scale characteristics. The calculated daily average value is compared with the upper limit of soil water holding capacity to determine the effective conservation depth. The effective conservation depth refers to the depth range within the soil layer that can effectively store and regulate water. Based on this depth range, the water conservation flux, i.e., the amount of water passing through a unit area per unit time, is calculated, thereby extracting water conservation characteristic data.

[0082] Water conservation characteristic data needs to be categorized and statistically analyzed according to different land use types. These mainly include three types: cultivated land, forest land, and grassland. The average annual water conservation capacity for each land use type is then calculated. This average annual water conservation capacity data is then standardized to obtain the land use conversion rate. Standardization eliminates the influence of different units and magnitudes, making the data comparable.

[0083] Based on the land use conversion rate, and combining the changes in the area of ​​various land uses with their corresponding water conservation contributions, the total annual change in water conservation capacity is calculated. This step considers the impact of land use change on water conservation capacity. A standard deviation analysis is performed on the total annual change in water conservation capacity to output a water conservation stability index, reflecting the stability of the region's water conservation capacity.

[0084] Finally, the water conservation stability index is substituted into the water conservation capacity assessment function to calculate the regional water conservation capacity. The expression for the water conservation capacity assessment function is: in: This indicates the region's water conservation capacity value; , , These represent the weighting coefficients for vegetation cover, soil properties, and topographic conditions, respectively. This indicates the degree of root system development of the i-th vegetation type; This represents the coverage of the i-th vegetation. This represents the growth status index of the i-th vegetation. This represents the water-holding capacity of soil type j; This represents the permeability of soil type j; This represents the thickness coefficient of the j-th type of soil; This represents the slope factor of the k-th terrain unit; This represents the slope aspect factor of the k-th terrain unit; The elevation factor of the k-th topographic unit is represented by ; n, m, and l represent the vegetation type, soil type, and number of topographic units, respectively.

[0085] This function comprehensively considers the impact of three main factors—vegetation cover, soil properties, and topographic conditions—on water conservation capacity. By calculating the regional water conservation capacity values, it generates water conservation assessment data, providing a scientific basis for water quality and quantity regulation.

[0086] For example, data on daily precipitation, evaporation, and runoff for the entire watershed are collected annually. Daily-scale water balance calculations yield basic data reflecting the water budget. Monthly summaries are then compiled, and the water conservation coefficient is calculated based on soil infiltration characteristics and groundwater recharge. The impact of dynamic changes in different land use areas within the watershed on water conservation capacity is analyzed. Finally, a water conservation capacity assessment function is used to comprehensively evaluate the watershed's water conservation capacity. For instance, an area with high vegetation cover, good soil structure, and moderate topographic relief will have a relatively high water conservation capacity value, reflecting its strong water conservation function. The entire assessment process forms a data processing chain with close logical connections between each step, ensuring the scientific validity and accuracy of the assessment results.

[0087] In one specific embodiment, the process of executing step S104 may specifically include the following steps:

[0088] (1) Multiply the water source conservation assessment data with the inflow cross-section of the reservoir, subtract the average daily evaporation loss of the reservoir, divide by the cross-sectional area, and calculate the inflow velocity of the reservoir; combine the inflow velocity of the reservoir with the pollutant concentration monitoring value to obtain the initial pollution load.

[0089] (2) Calculate the pollutant diffusion per unit time by multiplying the initial pollution load by the water body diffusion coefficient, adding the pollutant degradation coefficient, and dividing by the cross-sectional area of ​​the water flow; accumulate the pollutant diffusion per unit time by hour to obtain the daily-scale diffusion flux.

[0090] (3) Perform spatial stratification calculation on the daily diffusion flux, multiply the concentration values ​​of the surface, middle and bottom layers by the corresponding weighting coefficients, and calculate the vertical mixing coefficient; based on the vertical mixing coefficient, deduce the horizontal diffusion intensity and output water quality change prediction data;

[0091] (4) Based on the water quality change prediction data, extract the pollutant concentration change gradient, multiply the concentration gradient value by the diffusion coefficient, divide by the characteristic length, and calculate the convection diffusion term; substitute it into the pollutant degradation reaction term to construct a diffusion equation set;

[0092] (5) Configure boundary conditions for the diffusion equations, substitute the inflow boundary concentration value, outflow boundary concentration value and shore concentration value into the solution to obtain the steady-state concentration field; superimpose the steady-state concentration field with the hydrodynamic field to output the dynamic concentration distribution;

[0093] (6) Based on the dynamic concentration distribution, calculate the spatial diffusion range of pollutants, determine the area ratio of the area exceeding the standard, and establish water quality distribution data.

[0094] Specifically, by analyzing water source conservation assessment data and water flow characteristics, accurate simulation of pollutant diffusion processes can be achieved. Water source conservation assessment data includes key information such as water source conservation capacity and water balance status. By combining this data with the inflow rate at the reservoir cross-section, pollutant diffusion analysis is conducted.

[0095] The calculation of inflow velocity involves multiple hydrological parameters. Multiplying the water conservation assessment data by the inflow cross-sectional flow rate reflects the total volume of water entering the reservoir. Subtracting the average daily evaporation loss from this value accounts for the impact of water loss. Dividing the result by the cross-sectional area yields the inflow velocity, a crucial driving force for pollutant transport and diffusion. Combining the inflow velocity with pollutant concentration monitoring values ​​calculates the initial pollution load, reflecting the total pollutant load characteristics.

[0096] The initial pollution load gradually distributes spatially through water diffusion. The water diffusion coefficient is an important parameter describing the ability of pollutants to diffuse in water, and it is related to factors such as water temperature and turbulence intensity. Multiplying the initial pollution load by the water diffusion coefficient and adding the pollutant degradation coefficient considers the natural degradation process of pollutants. Dividing the result by the cross-sectional area of ​​the flow yields the amount of pollutant diffusion per unit time. By accumulating the pollutant diffusion amount per unit time on an hourly scale, the daily diffusion flux is obtained, reflecting the diffusion pattern of pollutants within a day. Aquatic pollutants exhibit stratified characteristics in the vertical direction. The daily diffusion flux needs to consider the concentration distribution of the surface, middle, and bottom water layers. Different water layers have different weighting coefficients, reflecting the importance of each layer in the pollutant diffusion process. By multiplying the concentration values ​​of each layer by their corresponding weighting coefficients, the vertical mixing coefficient is calculated, describing the degree of mixing of pollutants in the vertical direction. Based on the vertical mixing coefficient, the horizontal diffusion intensity is estimated, and water quality change prediction data is output.

[0097] Water quality change prediction data includes spatial distribution information of pollutant concentrations. By extracting the pollutant concentration change gradient, the spatial concentration change trend of pollutants is analyzed. The concentration gradient value is multiplied by the diffusion coefficient and divided by the characteristic length to calculate the convection-diffusion term. Simultaneously considering the degradation reaction of pollutants, a diffusion equation set is constructed. The expression of the diffusion equation set is: in: Represents the pollutant concentration field; Represents a time variable; Represents the diffusion coefficient tensor; Represents the velocity vector; Represents source and sink functions; Let represent the second-order diffusion coefficient of the i-th spatial dimension; Represents the i-th spatial coordinate; Represents the kinetic coefficient of the j-th reaction; The concentration of the j-th reaction component is represented by ; n represents the spatial dimension; and m represents the number of reaction components.

[0098] Solving the diffusion equations requires appropriate boundary conditions. A boundary condition system is established by configuring concentration values ​​at the inlet, outlet, and shoreline boundaries. Solving this system yields a steady-state concentration field, reflecting the spatial distribution of pollutants at equilibrium. Superimposing the steady-state concentration field with the hydrodynamic field provides a dynamic concentration distribution considering water flow. This dynamic concentration distribution is fundamental for assessing the extent of pollutant impact. By analyzing the areas where pollutant concentrations exceed standard limits, the percentage of areas exceeding limits is determined, thus establishing water quality distribution data. This data provides crucial information for water quality management and pollution control.

[0099] For example, hydrological data and pollutant monitoring data from the inflow section are collected, including flow rate, water level, and pollutant concentration. Initial conditions are determined through flow velocity calculation and pollution load analysis. The migration and transformation patterns of pollutants are analyzed by combining water diffusion and degradation processes. Considering the stratification characteristics of the water body, a vertical mixing model is established. The spatial distribution characteristics of pollutants are obtained by solving the diffusion equations. The entire analysis process forms a data processing chain, ensuring the accuracy and reliability of the simulation results.

[0100] In one specific embodiment, the process of executing step S105 may specifically include the following steps:

[0101] (1) Divide the water quality distribution data into monitoring grids according to the cross-sectional location, multiply by the corresponding water quality parameter weights, divide by the total number of grids, and calculate the water quality status index; perform time series sampling on the water quality status index to obtain the water quality change sequence;

[0102] (2) For the water quality change sequence, calculate the concentration difference between adjacent time points, divide the difference by the time interval, multiply by the diffusion coefficient to obtain the pollution diffusion rate; superimpose the pollution diffusion rate with the water flow direction data to output the diffusion direction vector;

[0103] (3) Based on the diffusion direction vector, extract the trajectory of the pollution plume, associate the coordinates of the trajectory points with the time series, and calculate the rate of change of the diffusion range; analyze the pollution load flux based on the rate of change of the diffusion range, and generate pollution diffusion situation data;

[0104] (4) Classify the pollution diffusion trend data, calculate the mean and standard deviation of pollutant concentration at each level, multiply the mean concentration by the weighting coefficient, and establish a classification evaluation index; compare the classification evaluation index with the water quality standard and output the exceedance risk coefficient.

[0105] (5) Construct a standard for classifying early warning levels using the risk coefficient of exceeding the standard, correspond the risk coefficient range with the early warning level, calculate the trigger probability of each level, and form an early warning indicator system; set graded response conditions based on the early warning indicator system and generate early warning threshold data.

[0106] Specifically, water quality distribution data includes information such as pollutant concentration and spatial distribution. By processing this data through gridding and time-series analysis, a water quality monitoring and early warning system is established. Gridding of water quality distribution data is the foundation for constructing the monitoring system. The monitoring section area is divided into several grid units, each representing a specific monitoring point. Each grid unit is assigned a corresponding water quality parameter weight, reflecting the importance of that monitoring point. By multiplying the water quality data of each grid unit by its corresponding weight and dividing by the total number of grids, a water quality status index is calculated. This index is a comprehensive indicator measuring the overall water quality status. The water quality status index is sampled at fixed time intervals to form time-series data, reflecting the dynamic changes in water quality status.

[0107] The concentration change is obtained by calculating the concentration difference between adjacent sampling times. Dividing this difference by the sampling time interval yields the concentration change rate. Multiplying the concentration change rate by the diffusion coefficient calculates the pollution diffusion rate, reflecting the speed of pollutant diffusion in space. The pollution diffusion rate is overlaid with water flow direction data to output a diffusion direction vector, describing the directional characteristics of pollutant transport. The extraction of the pollution plume's migration trajectory is based on the diffusion direction vector. By tracking the diffusion direction vector over continuous time periods, the movement path of the pollution plume is determined. The spatial coordinates of the trajectory points are correlated with the corresponding time series to calculate the rate of change of the diffusion range over time. The rate of change of the diffusion range reflects the dynamic characteristics of the pollution plume diffusion process. Based on this rate of change, the pollutant transport flux is analyzed, generating data reflecting the overall situation of pollution diffusion.

[0108] The situational data is categorized into different levels to determine varying degrees of pollution. For each level, the mean and standard deviation of pollutant concentration data are calculated to assess the pollution characteristics of that level. The mean concentration is multiplied by a corresponding weighting coefficient to establish a tiered evaluation index system. By comparing these evaluation indicators with water quality standards, exceedance risk coefficients are calculated to quantify the pollution risk level at different levels. The warning level classification is based on the exceedance risk coefficient. By dividing the risk coefficient into intervals, a correspondence between the risk coefficient and the warning level is established. The trigger probability for each warning level is calculated to construct a warning index system. Based on this system, response conditions for different levels are set, and the trigger thresholds for each warning level are defined, forming warning threshold data.

[0109] For example, by deploying monitoring grids at key cross-sections, water quality data is collected in real time. The collected data is then processed into a grid to calculate the water quality status index. Through continuous monitoring, water quality change sequences are obtained. Based on the water quality change data, pollutant diffusion characteristics are analyzed, and the migration trajectory of pollution plumes is tracked. The degree of pollution is assessed in a graded manner, and an early warning triggering mechanism is established. The entire monitoring and early warning process forms a data analysis chain, with each link closely connected from data collection to early warning triggering, ensuring the timeliness and accuracy of monitoring and early warning.

[0110] In one specific embodiment, the process of executing step S106 may specifically include the following steps:

[0111] (1) Align the warning threshold data with the measured data according to the time nodes, calculate the prediction deviation value at the corresponding time, multiply it by the weight coefficient to obtain the weighted error sequence; perform statistical analysis on the weighted error sequence and output the correction coefficient;

[0112] (2) Correct the warning threshold using the correction coefficient, calculate the correlation coefficient between the corrected value and the measured value, divide it by the number of samples to obtain the fitting accuracy value; classify the fitting accuracy value according to the monitoring section to generate the spatial accuracy distribution.

[0113] (3) Based on the spatial accuracy distribution, calculate the prediction accuracy of each monitoring section, multiply the accuracy value by the section importance coefficient, divide by the total number of sections, and output the section weight coefficient; compare the section weight coefficient with the accuracy threshold to establish accuracy evaluation data;

[0114] (4) Classify the accuracy assessment data, count the proportion of data in different accuracy ranges, calculate the credibility coefficient of each level, and generate a classification evaluation matrix; associate the classification evaluation matrix with the monitoring threshold and output the classification monitoring standard.

[0115] (5) Based on the graded monitoring standard, calculate the early warning triggering conditions for each monitoring section, and divide the difference between the monitored value and the standard value by the allowable deviation to obtain the early warning discrimination coefficient; classify the early warning discrimination coefficient to determine the monitoring response level;

[0116] (6) Match the monitoring response level with the water quality category, calculate the monitoring frequency requirements for each level, and generate water quality and quantity monitoring results.

[0117] Specifically, the comparative analysis of warning threshold data and measured data is fundamental to evaluating monitoring accuracy. By aligning the two types of data according to time nodes, temporal consistency is ensured, providing an accurate data foundation for subsequent analysis. After alignment, the prediction deviation value at the corresponding time point is calculated, reflecting the degree of difference between the prediction result and the actual situation. Considering the varying importance of data at different time periods, a weighting coefficient is introduced, which is multiplied by the deviation value to obtain a weighted error sequence. Statistical analysis is performed on this sequence, including calculating statistics such as mean and standard deviation, and outputting correction coefficients for data correction. The application of correction coefficients aims to improve the accuracy of warning thresholds. Data correction is achieved by multiplying the original warning threshold by the correction coefficient. To verify the correction effect, the corrected values ​​need to be compared with the measured values, and the correlation coefficient is calculated. Dividing the correlation coefficient by the sample size yields the fitting accuracy value, which reflects the degree of agreement between the corrected data and the measured data. The fitting accuracy values ​​are categorized and organized according to different monitoring sections to generate accuracy distribution data reflecting spatial distribution characteristics.

[0118] Analyzing spatial accuracy distribution data is crucial for evaluating monitoring effectiveness. The prediction accuracy of each monitoring section is calculated, reflecting the monitoring reliability of that specific section. Considering the different roles of various sections within the overall monitoring network, a section importance coefficient is introduced. Multiplying the accuracy value by this coefficient and dividing by the total number of sections yields a section weight coefficient reflecting the relative importance of each section. By comparing these section weight coefficients with pre-set accuracy thresholds, an accuracy assessment data system is established. The hierarchical processing of accuracy assessment data is fundamental to establishing a monitoring mechanism. The assessment data is graded, and the distribution of data within each accuracy interval is statistically analyzed, calculating the proportion of data from different intervals to the total. Based on these statistical results, a reliability coefficient for each level is calculated, reflecting the reliability of data at different levels. This information is integrated to form a hierarchical evaluation matrix, which includes the accuracy characteristics of each level. By correlating the hierarchical evaluation matrix with monitoring thresholds, hierarchical monitoring standards are generated.

[0119] The application of tiered monitoring standards is reflected in specific monitoring practices. By calculating the early warning trigger conditions for each monitoring section, the actual monitored values ​​are compared with the standard values ​​to calculate the difference. Dividing this difference by the allowable deviation range yields the early warning discrimination coefficient, which is a crucial basis for determining whether an early warning needs to be triggered. The early warning discrimination coefficients are then tiered to determine the required monitoring response level under different circumstances. Finally, a correspondence is established between the monitoring response level and water quality categories, clarifying the required monitoring frequency for different water quality categories. Based on the characteristics of water quality categories and regulatory requirements, corresponding monitoring frequency requirements are formulated, resulting in water quality and quantity monitoring results. These results include not only monitoring data but also management information such as monitoring frequency and early warning levels.

[0120] For example, after obtaining 30 consecutive days of measured data and early warning system prediction data for a certain monitoring section, the data is aligned over time. For instance, the predicted value at 8:00 AM on January 1st is 0.85 mg / L, and the measured value is 0.82 mg / L, resulting in a prediction deviation of 0.03 mg / L. Considering this time point falls during the morning peak water usage period, a weighting coefficient of 1.5 is set, resulting in a weighted error of 0.045 mg / L. Similar processing is performed on the 30 days of data to obtain a weighted error sequence, and a correction coefficient of 1.08 is output through statistical analysis. This correction coefficient is then applied to correct the original early warning threshold. For example, the original early warning threshold is 1.0 mg / L, which is corrected to 1.08 mg / L. The correlation coefficient between the corrected value and the measured value is calculated to be 0.92. With a sample size of 30, the fitting accuracy is calculated to be 0.0307. The same analysis was also performed on cross-sections at different locations upstream of the monitoring section, such as 500 meters, 1000 meters, and 1500 meters, and the fitting accuracy values ​​were obtained as 0.0312, 0.0298, and 0.0320, respectively, forming a spatial accuracy distribution.

[0121] Based on the accuracy distribution of these cross-sections, the prediction accuracy is calculated. If the predicted value and the measured value of this cross-section are within the allowable range for 27 out of 30 days, the accuracy is 90%. Since this cross-section is located upstream of the water intake, the cross-section importance coefficient is set to 1.2, and the total number of cross-sections is 10, resulting in a cross-section weight coefficient of 0.108. This weight coefficient is compared with the set accuracy threshold of 0.1 to establish accuracy assessment data. The accuracy assessment data is then statistically graded into intervals such as above 90%, 80-90%, and 70-80%, and different levels of reliability coefficients are calculated. For example, the reliability coefficient for the above 90% level is 0.95, and for the 80-90% level it is 0.85. These data are combined to form a graded evaluation matrix, which is then correlated with the monitoring threshold to form a graded monitoring standard.

[0122] According to the tiered monitoring standards, the early warning trigger conditions for the monitoring sections are calculated. For example, if a monitoring value is 0.95 mg / L, the standard value is 1.0 mg / L, and the allowable deviation is 0.1 mg / L, the calculated early warning discrimination coefficient is -0.5. This coefficient is then used to classify the response level, determining it to be Level II. Finally, monitoring frequency requirements are established based on the response level. For example, Level I response requires monitoring every 2 hours, Level II response every 4 hours, and Level III response every 6 hours, forming monitoring specifications. This establishes a dynamic monitoring system that is continuously optimized based on measured data.

[0123] The above describes the water quality and quantity accurate simulation method based on the SWAT model in the embodiments of this application. The following describes the water quality and quantity accurate simulation system based on the SWAT model in the embodiments of this application. Please refer to [link / reference]. Figure 2 One embodiment of the water quality and quantity accurate simulation system based on the SWAT model in this application includes:

[0124] The acquisition module is used to analyze the spatiotemporal characteristics of multi-source precipitation and evapotranspiration data through dual-parameter trend testing and slope evaluation, acquire initial hydrological data, process the initial hydrological data using a triple configuration algorithm, and generate a standard hydrological dataset.

[0125] The partitioning module is used to perform regional hydrological partitioning based on the standard hydrological dataset and combined with topographic elevation information, generate sub-basin partitioning data, calculate the hydrological response units of the sub-basins based on the sub-basin partitioning data, and construct a basic hydrological calculation dataset.

[0126] The calculation module is used to perform water balance calculations using the basic hydrological calculation dataset, extract water conservation characteristic data, evaluate the water conservation characteristic data with land use change, and generate water conservation assessment data.

[0127] The analysis module is used to perform pollutant diffusion analysis of the inflow water body based on the water source conservation assessment data, output water quality change prediction data, integrate the water quality change prediction data and solve the pollutant diffusion equation to establish water quality distribution data.

[0128] The monitoring module is used to monitor the reservoir's operational status based on the water quality distribution data, obtain pollution diffusion trend data, design an early warning indicator system based on the pollution diffusion trend data, and determine early warning threshold data.

[0129] The calibration module is used to compare and calibrate the warning threshold data with the measured data, establish accuracy assessment data, construct a hierarchical monitoring mechanism based on the accuracy assessment data, and output water quality and quantity monitoring results.

[0130] Through the collaborative efforts of the aforementioned components, the spatiotemporal characteristics of multi-source precipitation and evapotranspiration data were analyzed using dual-parameter trend testing and slope assessment. A triple-configuration algorithm was employed to process the initial hydrological data, ensuring its accuracy and reliability. Regional hydrological delineation was performed based on standard hydrological datasets and topographic elevation information, generating sub-basin delineation data and calculating hydrological response units, thus improving the spatial accuracy of hydrological analysis. Water balance calculations were conducted using basic hydrological calculation datasets to extract water conservation characteristic data and evaluate it in conjunction with land use change data, achieving a precise assessment of water conservation capacity. Pollutant diffusion analysis of inflow water was conducted based on water conservation assessment data, and water quality distribution data was established by solving pollutant diffusion equations, improving the accuracy of water quality assessment. Reservoir operation status monitoring was implemented based on water quality distribution data, acquiring pollution diffusion trend data and designing an early warning indicator system, enhancing the real-time nature of water quality monitoring. By comparing and correcting early warning threshold data with measured data, accuracy assessment data was established, and a hierarchical monitoring mechanism was constructed, ensuring the reliability of monitoring results.

[0131] This application also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the water quality and quantity accurate simulation method based on the SWAT model.

[0132] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system and units described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0133] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0134] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for accurate simulation of water quality and quantity based on the SWAT model, characterized in that, The accurate water quality and quantity simulation method based on the SWAT model includes: The spatiotemporal characteristics of multi-source precipitation and evapotranspiration data are analyzed by dual-parameter trend test and slope evaluation to obtain initial hydrological data. The initial hydrological data are then processed by a triple configuration algorithm to generate a standard hydrological dataset. Based on the standard hydrological dataset, regional hydrological division is carried out in combination with topographic elevation information to generate sub-basin division data. Based on the sub-basin division data, hydrological response units of sub-basins are calculated to construct a basic hydrological calculation dataset. Water balance calculations are performed using the basic hydrological calculation dataset to extract water conservation characteristic data. The water conservation characteristic data is then evaluated in conjunction with land use change to generate water conservation assessment data. Based on the water source conservation assessment data, pollutant diffusion analysis of inflow water is performed, water quality change prediction data is output, the water quality change prediction data is integrated and the pollutant diffusion equation is solved to establish water quality distribution data. Based on the water quality distribution data, reservoir operation status monitoring is implemented to obtain pollution diffusion trend data. An early warning index system is designed based on the pollution diffusion trend data, and early warning threshold data is determined. This includes: dividing the water quality distribution data into monitoring grids according to cross-sectional locations, multiplying by the corresponding water quality parameter weights, and dividing by the total number of grids to calculate a water quality status index; performing time-series sampling on the water quality status index to obtain a water quality change sequence; calculating the concentration difference between adjacent time points for the water quality change sequence, dividing the difference by the time interval, and multiplying by the diffusion coefficient to obtain the pollution diffusion rate; superimposing the pollution diffusion rate with water flow direction data to output a diffusion direction vector; and using the diffusion direction vector as a basis... Extract the migration trajectory of the pollution plume, correlate the coordinates of the trajectory points with the time series, and calculate the rate of change of the diffusion range; analyze the pollution load flux based on the rate of change of the diffusion range to generate the pollution diffusion situation data; classify the pollution diffusion situation data, calculate the mean and standard deviation of pollutant concentrations at each level, multiply the mean concentration by a weighting coefficient, and establish a classification evaluation index; compare the classification evaluation index with water quality standards to output the exceedance risk coefficient; construct a warning level classification standard using the exceedance risk coefficient, correspond the risk coefficient interval with the warning level, calculate the trigger probability of each level, and form a warning indicator system; set classification response conditions based on the warning indicator system to generate the warning threshold data. The process involves comparing and correcting the warning threshold data with measured data to establish accuracy assessment data. A tiered monitoring mechanism is then constructed based on this accuracy assessment data to output water quality and quantity monitoring results. This includes: aligning the warning threshold data and measured data according to time nodes, calculating the prediction deviation value at the corresponding time, multiplying it by a weighting coefficient to obtain a weighted error sequence; performing statistical analysis on the weighted error sequence to output correction coefficients; using the correction coefficients to correct the warning thresholds, calculating the correlation coefficient between the corrected values ​​and the measured values, dividing it by the sample size to obtain the fitting accuracy value; classifying the fitting accuracy values ​​according to monitoring sections to generate a spatial accuracy distribution; and calculating the prediction accuracy rate for each monitoring section based on the spatial accuracy distribution, multiplying the accuracy rate value by... The importance coefficient of a cross-section is divided by the total number of cross-sections to output the cross-section weight coefficient. This cross-section weight coefficient is compared with the accuracy threshold to establish the accuracy assessment data. The accuracy assessment data is then graded, and the proportion of data in different accuracy intervals is statistically analyzed. The reliability coefficient for each level is calculated to generate a graded evaluation matrix. This graded evaluation matrix is ​​correlated with the monitoring threshold to output the graded monitoring standard. Based on the graded monitoring standard, the early warning triggering conditions for each monitoring cross-section are calculated. The difference between the monitored value and the standard value is divided by the allowable deviation to obtain the early warning discrimination coefficient. The early warning discrimination coefficient is then graded to determine the monitoring response level. The monitoring response level is correlated with the water quality category, and the monitoring frequency requirements for each level are calculated to generate the water quality and quantity monitoring results.

2. The method for accurate simulation of water quality and quantity based on the SWAT model according to claim 1, characterized in that, The process involves analyzing the spatiotemporal characteristics of multi-source precipitation and evapotranspiration data using dual-parameter trend testing and slope evaluation to obtain initial hydrological data. This initial hydrological data is then processed using a triple-configuration algorithm to generate a standard hydrological dataset, including: The time weighting coefficients of the multi-source precipitation and evapotranspiration data are extracted, and weighted filtering is used to eliminate random disturbances and generate a time series matrix. The least squares operation is performed on the time series matrix to calculate the mutation detection threshold, a trend test sequence is constructed based on the MK statistic, and the significance level is output. Based on the significance level as the benchmark parameter, the time series rate of change is calculated in segments, the trend change is determined by the Sen slope assessment, and the spatiotemporal characteristic parameters are generated by quantile correction. Based on the spatiotemporal characteristic parameters as constraints, Kriging spatial interpolation is performed to calculate the spatial autocorrelation coefficient, and the initial hydrological data is output through co-evolution analysis. Heteroscedasticity test matrix operation is performed on the initial hydrological data, outliers are removed by combining Mahalanobis distance discrimination method, and Bayesian estimation is used to supplement the data; The processed hydrological data undergoes triple configuration weight calculation, adaptive fusion iterative operation, and normalization processing to generate the standard hydrological dataset.

3. The method for accurate simulation of water quality and quantity based on the SWAT model according to claim 1, characterized in that, The process involves dividing the region into hydrological sub-basins based on the standard hydrological dataset and topographic elevation information, generating sub-basin sub-data, calculating hydrological response units for each sub-basin based on the sub-basin sub-data, and constructing a basic hydrological calculation dataset, including: The precipitation data in the standard hydrological dataset is multiplied by the topographic elevation information to calculate the water yield per unit area. This yield is then divided by the runoff time to obtain the initial flow data. The initial flow data is then spatially overlaid to generate a runoff map. The confluence path is extracted from the water flow convergence map, the cumulative flow is calculated based on the flow direction relationship matrix, multiplied by the slope coefficient, divided by the critical threshold, and the coordinates of the sub-basin boundary points are output; the coordinates of the sub-basin boundary points are connected into a line to generate the sub-basin division data; Using the sub-basin division data as boundaries, the area ratio of land use type in each sub-basin is extracted, multiplied by the permeability coefficient, and the underlying surface characteristic value is calculated; the underlying surface characteristic value is combined with soil parameters to obtain hydrological response parameters; The hydrological response parameters are classified and statistically analyzed. The area weights of different response types are calculated, multiplied by the hydrological characteristic coefficients, and the distribution of hydrological response units is output. The distribution of hydrological response units is superimposed on the sub-basin boundary to generate spatial data of response units. The runoff coefficient of each unit is calculated using the spatial data of the response unit. The runoff coefficient is multiplied by the precipitation and divided by the area to obtain the unit runoff. The unit runoff is then spatially integrated to output the watershed water volume data. By combining the watershed water volume data with the hydrological parameter relationships, the hydrological response coefficient matrix is ​​calculated, the water balance equation is established, and the basic hydrological calculation dataset is generated.

4. The method for accurate simulation of water quality and quantity based on the SWAT model according to claim 1, characterized in that, The process of performing water balance calculations using the basic hydrological dataset, extracting water conservation characteristic data, and then evaluating the water conservation characteristic data in conjunction with land use change to generate water conservation assessment data includes: First, multiply the daily precipitation data in the basic hydrological calculation dataset by the catchment area, subtract the product of daily evaporation and daily runoff, and divide by the total water volume to calculate the daily water balance value; then, accumulate the daily water balance value monthly to obtain the monthly water volume regulation data. The monthly water conservation coefficient is calculated by multiplying the monthly average soil infiltration rate by the monthly water volume regulation data, subtracting the monthly groundwater recharge, and then dividing by the total monthly precipitation. The monthly water conservation coefficient is then multiplied by the underlying surface characteristic value to obtain the initial value of water conservation capacity. The initial value of water conservation capacity is calculated as a daily average, and the calculation result is compared with the upper limit of soil water holding capacity to determine the effective conservation depth; the water conservation flux is calculated based on the effective conservation depth, and the water conservation characteristic data is extracted. The water conservation characteristic data are classified and statistically analyzed according to cultivated land, forest land, and grassland, and the average annual conservation volume of different land use types is calculated; the average annual conservation volume is standardized to obtain the land use conversion rate. Based on the land use conversion rate, the changes in the area of ​​various types of land use are multiplied by their corresponding conservation contribution to obtain the total annual conservation change; standard deviation analysis is performed on the total annual conservation change to output the water conservation stability index. The water conservation stability index is substituted into the water conservation capacity assessment function to calculate the regional water conservation capacity value and generate the water conservation assessment data.

5. The method for accurate simulation of water quality and quantity based on the SWAT model according to claim 1, characterized in that, Based on the water source conservation assessment data, the process involves conducting pollutant diffusion analysis on the incoming water body, outputting water quality change prediction data, integrating the water quality change prediction data, solving the pollutant diffusion equation, and establishing water quality distribution data, including: Multiply the water source conservation assessment data by the inflow cross-sectional flow rate, subtract the average daily evaporation loss of the reservoir, and divide by the cross-sectional area to calculate the inflow water velocity; combine the inflow water velocity with the pollutant concentration monitoring value to obtain the initial pollution load. The initial pollution load is multiplied by the water body diffusion coefficient, and the pollutant degradation coefficient is added. The result is then divided by the cross-sectional area of ​​the water flow to calculate the pollutant diffusion rate per unit time. The pollutant diffusion rate per unit time is accumulated hourly to obtain the daily-scale diffusion flux. The daily-scale diffusion flux is spatially stratified and calculated. The concentration values ​​of the surface, middle and bottom layers are multiplied by the corresponding weighting coefficients to calculate the vertical mixing coefficient. Based on the vertical mixing coefficient, the horizontal diffusion intensity is estimated, and water quality change prediction data is output. Based on the water quality change prediction data, the pollutant concentration change gradient is extracted, the concentration gradient value is multiplied by the diffusion coefficient, divided by the characteristic length, and the convection diffusion term is calculated; the pollutant degradation reaction term is substituted into the equations to construct a diffusion equation set. Boundary conditions are configured for the diffusion equations, and the concentration values ​​at the inlet boundary, outlet boundary, and shore are substituted into the solution to obtain the steady-state concentration field. The steady-state concentration field is then superimposed with the hydrodynamic field to output the dynamic concentration distribution. Based on the dynamic concentration distribution, the spatial diffusion range of pollutants is calculated, the area ratio of the exceeding area is determined, and the water quality distribution data is established.

6. A water quality and quantity accurate simulation system based on the SWAT model, used to implement the water quality and quantity accurate simulation method based on the SWAT model as described in any one of claims 1-5, characterized in that, The water quality and quantity accurate simulation system based on the SWAT model includes: The acquisition module is used to analyze the spatiotemporal characteristics of multi-source precipitation and evapotranspiration data through dual-parameter trend testing and slope evaluation, acquire initial hydrological data, process the initial hydrological data using a triple configuration algorithm, and generate a standard hydrological dataset. The partitioning module is used to perform regional hydrological partitioning based on the standard hydrological dataset and combined with topographic elevation information, generate sub-basin partitioning data, calculate the hydrological response units of the sub-basins based on the sub-basin partitioning data, and construct a basic hydrological calculation dataset. The calculation module is used to perform water balance calculations using the basic hydrological calculation dataset, extract water conservation characteristic data, evaluate the water conservation characteristic data with land use change, and generate water conservation assessment data. The analysis module is used to perform pollutant diffusion analysis of the inflow water body based on the water source conservation assessment data, output water quality change prediction data, integrate the water quality change prediction data and solve the pollutant diffusion equation to establish water quality distribution data. The monitoring module is used to monitor the reservoir's operational status based on the water quality distribution data and acquire pollution diffusion trend data. Based on the pollution diffusion trend data, it designs an early warning indicator system and determines early warning threshold data, including: dividing the water quality distribution data into monitoring grids according to cross-sectional locations, multiplying by the corresponding water quality parameter weights, and dividing by the total number of grids to calculate a water quality status index; performing time-series sampling on the water quality status index to obtain a water quality change sequence; calculating the concentration difference between adjacent time points for the water quality change sequence, dividing the difference by the time interval, and multiplying by the diffusion coefficient to obtain the pollution diffusion rate; superimposing the pollution diffusion rate with water flow direction data to output a diffusion direction vector; and using the diffusion direction vector... Based on this, the migration trajectory of the pollution plume is extracted, and the coordinates of the trajectory points are correlated with the time series to calculate the rate of change of the diffusion range. Based on the rate of change of the diffusion range, the pollution load flux is analyzed to generate the pollution diffusion situation data. The pollution diffusion situation data is classified, and the mean and standard deviation of pollutant concentrations at each level are calculated. The mean concentration is multiplied by a weighting coefficient to establish a classification evaluation index. The classification evaluation index is compared with water quality standards to output the exceedance risk coefficient. Using the exceedance risk coefficient, a warning level classification standard is constructed, and the risk coefficient interval is mapped to the warning level. The trigger probability of each level is calculated to form a warning indicator system. Based on the warning indicator system, classification response conditions are set to generate the warning threshold data. The correction module is used to compare and correct the warning threshold data with the measured data, establish accuracy assessment data, construct a hierarchical monitoring mechanism based on the accuracy assessment data, and output water quality and quantity monitoring results. This includes: aligning the warning threshold data and measured data according to time nodes, calculating the prediction deviation value at the corresponding time, multiplying it by a weighting coefficient to obtain a weighted error sequence; performing statistical analysis on the weighted error sequence and outputting correction coefficients; correcting the warning threshold using the correction coefficients, calculating the correlation coefficient between the corrected value and the measured value, dividing it by the sample size to obtain a fitting accuracy value; classifying the fitting accuracy value according to monitoring sections to generate a spatial accuracy distribution; and calculating the prediction accuracy rate for each monitoring section based on the spatial accuracy distribution. Multiply the value by the cross-section importance coefficient, divide by the total number of cross-sections, and output the cross-section weight coefficient; compare the cross-section weight coefficient with the accuracy threshold to establish the accuracy assessment data; classify the accuracy assessment data, statistically analyze the data proportion of different accuracy intervals, calculate the reliability coefficient of each level, and generate a classification evaluation matrix; associate the classification evaluation matrix with the monitoring threshold to output the classification monitoring standard; based on the classification monitoring standard, calculate the early warning triggering conditions for each monitoring cross-section, calculate the difference between the monitored value and the standard value, divide by the allowable deviation, and obtain the early warning discrimination coefficient; classify the early warning discrimination coefficient to determine the monitoring response level; correspond the monitoring response level to the water quality category, calculate the monitoring frequency requirements for each level, and generate the water quality and quantity monitoring results.

7. A computer-readable storage medium storing instructions thereon, characterized in that, When the instruction is executed by the processor, it implements the method for accurate simulation of water quality and quantity based on the SWAT model as described in any one of claims 1-5.

Images