Drainage basin water quality prediction method and system based on spatial flow river network topological structure

Abstract

The invention relates to the technical field of water quality monitoring and drainage basin water environment management, in particular to a drainage basin water quality prediction method and system based on a spatial flow river network topological structure, and the method comprises the steps: firstly, determining a monitoring region, obtaining drainage basin digital elevation model (DEM) data and drainage basin boundary data, and then obtaining water quality data in a drainage basin through setting a monitoring point; the method comprises the following steps: firstly, mainly monitoring an area with relatively large water quality change, then obtaining an environment variable in a drainage basin, analyzing and generating an accurate prediction river network and a prediction point in combination with a spatial flow river network topological structure, then predicting water quality parameters by constructing a spatial flow river network topological structure model and combining water quality data of a monitoring point and the environment variable, and finally, predicting the water quality parameters by constructing a spatial flow river network topological structure model. A prediction result is displayed through a geographic information system (GIS) visual tool, real-time monitoring and water quality early warning are facilitated, the technical scheme is suitable for water quality prediction of a large-scale drainage basin, the change of the water quality of the drainage basin can be monitored in real time, and a scientific basis is provided for drainage basin management and pollution prevention and control.

Description

Technical Field

[0001] The present invention relates to the technical field of water quality monitoring and watershed water environment management, and in particular to a watershed water quality prediction method and system based on a spatial river network topology structure. Background Art

[0002] Water quality is an important indicator of the physical, chemical and biological properties of surface water, which has a profound impact on human health and ecosystems. With the intensification of industrial and agricultural activities, such as excessive use of fertilizers and pesticides, and discharge of industrial wastewater, the surface water quality in some areas has dropped sharply, and lake eutrophication has occurred from time to time. At the same time, climate change and the rapid growth of the global population have further exacerbated the risk of shortage of clean drinking water resources. Water quality deterioration and drinking water pollution may cause a series of health problems, especially increasing the risk of liver disease and cardiovascular disease. Therefore, in order to more effectively deal with environmental and human health problems, it is particularly important to effectively manage and predict water quality in the entire river basin. In river and lake ecosystems, the use of models to simulate the physical, chemical and biological processes of water bodies has become a common method for analyzing water quality changes and predicting water quality parameters. However, as a decision support system, the application of models needs to be rigorously evaluated in terms of benefit-cost ratio, time and space scale, data availability, and engineering feasibility.

[0003] Traditional water quality prediction models usually require the uniform deployment of detection points in the river network of the basin, and the analysis and prediction of river water quality based on the mechanism of water quality diffusion. However, in some basins with complex natural environments, diverse river network structures and large areas, it is very difficult and expensive to evenly deploy fixed monitoring points, which to a certain extent limits the accuracy and reliability of water quality prediction for such basins and rivers, resulting in greater uncertainty in the prediction results.

[0004] With the continuous development of water quality analysis technology, some models based on historical data, combined with neural networks and machine learning algorithms, have been used to predict current and future surface water quality conditions. However, such models combining neural networks and machine learning usually require a large amount of historical data for sample training and curve fitting, and rely on the integrity and accuracy of the data. Other models, based on grid search and geostatistical models in professional fields, improve the spatial prediction of river water quality to a certain extent on the basis of limited monitoring points. Although these models have taken the spatial dependence of water quality parameters into consideration, such as simulating the spatial distribution of water quality parameters through grid search or GIS methods, and using spatial autocorrelation analysis based on Euclidean distance, they fail to effectively capture the topological structure of river networks and hydrological connectivity within river basins. Although the physical distance between river sections may be short, from the perspective of river networks, these river sections may cross catchments, be regulated by different dams or come from different sources, and hydrological and ecological conditions may be significantly different, which makes the spatial differences in water quality parameters more obvious. However, existing models often fail to fully consider this spatial heterogeneity when predicting water quality for the entire river basin, and are even unable to quickly generalize the prediction results to every river channel in the basin and visualize them.

[0005] In summary, existing technologies still have obvious limitations in accurately predicting the spatial distribution of water quality and improving watershed water quality. Summary of the invention

[0006] The purpose of the present invention is to provide a method and system for predicting water quality in a river basin based on the spatial river network topology structure, so as to solve the problem that the prior art still has obvious limitations in accurately predicting the spatial distribution of water quality and improving water quality in the river basin.

[0007] To achieve the above object, the present invention provides a method for predicting water quality in a river basin based on a spatial river network topology structure. The method for predicting water quality in a river basin based on a spatial river network topology structure comprises the following steps: S1: Select the watershed area to be predicted according to the actual needs of the watershed, and obtain the DEM data and watershed boundary data of the watershed; S2: After determining the watershed area, it is necessary to set up monitoring points based on the water flow distribution, environmental variable characteristics, and water quality influencing factors in the watershed; S3: At each monitoring point, real-time monitoring equipment or sampling means are installed to obtain real-time data of water quality related indicators, and after pre-processing, the data is stored in the system database; S4: Obtain the environmental variables related to the watershed, and map these environmental variables to various locations in the watershed through data conversion and spatial interpolation processing; S5: Based on the DEM data, boundary data and monitoring points of the watershed, projection conversion and data format conversion processing are performed through the GIS module in the system to generate an accurate predicted river network, and multiple prediction points are evenly distributed and set according to the characteristics of the watershed and the changing laws of water quality parameters; S6: Using the spatial flow river network topology model, the water quality parameters of the monitoring points are used as the response variables, the environmental variables are used as covariates, and the model is constructed by adding the spatial flow river network topology as a driving factor to calculate the values ​​of the water quality parameters of the prediction points; S7: Use GIS tools to visualize the predicted water quality parameter values ​​for each river channel; S8: By comparing and analyzing the prediction results with the data from the actual monitoring points, evaluating the accuracy of the model through cross-validation, and optimizing the model or adding new monitoring points for correction, the accuracy and generalization ability of the model can be improved.

[0008] Wherein, in step S1, the DEM data of the watershed is obtained through satellite remote sensing or lidar technology, and the boundary data of the watershed is obtained through remote sensing images and geographic information system technology.

[0009] Among them, in step S2, the selection criteria of the monitoring points are: the monitoring points should be mainly set up in the main rivers, tributaries and areas that may be polluted in the basin; When setting up monitoring points, full consideration should be given to the hydrological characteristics of the basin, environmentally sensitive areas, and the location of pollution sources; The specific method for setting up monitoring points is: combining the water flow network and river network topology of the basin, and intelligently selecting key areas to set up monitoring points based on the actual hydrological conditions of the basin.

[0010] Wherein, in step S4, the environmental variables include hydrological and meteorological data of forest coverage, land use type, soil type, precipitation, temperature, evaporation, and river flow; Environmental variables are obtained through meteorological station and remote sensing data, and data cleaning, format conversion and spatial interpolation are performed.

[0011] Among them, in step S6, the formula for calculating the value of the water quality parameter at the prediction point is: Y = Xβ+ zTU + zTD + zEU +ε Among them, Y is the vector of response variables, the response variables are the total nitrogen and dissolved oxygen contents at the monitoring points, X is the matrix of covariates, the covariates are environmental variables, β is the parameter vector of each covariate, zTU, zTD and zEU are vectors of random variables with tail-up, tail-down and Euclidean spatial autocorrelation structures, respectively, and ε is the vector of independent random errors.

[0012] Among them, in step S6, the selection of covariates is specifically as follows: by analyzing the correlation between water quality parameters and environmental variables, a regression relationship between environmental variables and water quality parameters is established, and the stepwise regression method is used to screen out the most predictive variable combination from multiple candidate covariates to optimize the accuracy of the prediction model.

[0013] Among them, in step S6, the screening of the spatial flow river network topological structure is specifically as follows: after the regression model is established, the spatial structure related to the spatial flow river network is further selected, and the spatial structure of tail-up, tail-down and Euclidean spatial autocorrelation is selected through random combination to generate all possible combinations and add them to the model. By simulating the interaction between river flow direction and water quality parameters, model training and water quality prediction are promoted.

[0014] Among them, in step S6, the water quality parameter value is calculated specifically as follows: based on the trained regression model and the spatial stream network topology network structure, combined with the environmental data of each prediction point and the water quality information of the monitoring point, the water quality parameter value of each prediction point is calculated.

[0015] The present invention also provides a watershed water quality prediction system based on a spatial stream network topology structure, which is applied to the watershed water quality prediction method based on a spatial stream network topology structure as described above.

[0016] The present invention provides a method and system for predicting water quality in a watershed based on a spatial river network topology structure. First, the monitoring area is determined and the digital elevation model (DEM) data and the boundary data of the watershed are obtained to accurately define the scope of the watershed and provide basic data for subsequent water quality prediction. Then, the water quality data in the watershed is obtained by setting monitoring points, and the areas with large water quality changes are monitored in particular to avoid the shortcomings of uniformly arranging monitoring points in traditional methods. Then, environmental variables in the watershed, including meteorological, soil, precipitation and other variables, are obtained, and an accurate prediction river network and prediction points are generated in combination with the spatial river network topology structure analysis. Finally, by constructing a spatial river network topological structure model, combined with the water quality data of the monitoring points and the environmental variable data, the water quality parameters are predicted. Finally, the prediction results are displayed through GIS visualization tools to facilitate real-time monitoring and water quality early warning. In practical applications, this technical solution intelligently selects key areas for monitoring and prediction based on factors such as the basin's topography, river network structure, and environmental variables, avoiding the problem of traditional methods that cannot effectively consider spatial correlation. It is not only suitable for water quality prediction in large-scale basins, but can also monitor changes in basin water quality in real time, providing a scientific basis for basin management and pollution prevention and control. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a flow chart of the steps of the method for predicting water quality of a watershed based on the spatial river network topology structure provided by the present invention.

[0019] Figure 2 The watershed predicted river network and predicted points are generated by the GIS tool provided by the present invention.

[0020] Figure 3 It is a conceptual diagram of a tail-up space structure and four tail-down space structures provided by the present invention.

[0021] Figure 4 The total nitrogen and dissolved oxygen content of water quality predicted data provided by the present invention and the real data cross-validation scatter point R 2 picture. DETAILED DESCRIPTION

[0022] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0023] See also Figures 1 to 4 The present invention provides a method for predicting water quality of a river basin based on a spatial river network topology structure. The method for predicting water quality of a river basin based on a spatial river network topology structure comprises the following steps: S1: Select the area to be predicted according to the actual needs of the watershed, and obtain the watershed DEM data and watershed boundary data; S2: After determining the watershed area, it is necessary to set up monitoring points based on the water flow distribution, environmental characteristics and water quality influencing factors in the watershed; S3: At each monitoring point, real-time monitoring equipment or sampling means are installed to obtain real-time data of water quality related indicators, and after pre-processing, the data is stored in the system database; S4: Obtain the environmental data related to the watershed, and map these environmental variables to various locations in the watershed through data conversion and spatial interpolation; S5: Based on the basin DEM data, boundary data and the monitoring points that have been set, the GIS module in the system performs projection conversion and data format conversion processing to generate an accurate predicted river network, and evenly distributes and sets multiple prediction points according to the characteristics of the basin and the changing laws of water quality parameters; S6: Using the spatial flow river network topology model, the water quality parameters of the monitoring points are used as the response variables, the environmental variables are used as covariates, and the model is constructed by adding the spatial flow river network topology as a driving factor to calculate the values ​​of the water quality parameters of the prediction points; S7: Use GIS tools to visualize the predicted water quality parameter values ​​for each river channel; S8: By comparing the prediction results with the data from the actual monitoring points and evaluating the accuracy of the model through cross-validation, the model can be optimized or corrected by adding new monitoring points to improve the accuracy and generalization ability of the model.

[0024] In this implementation, first, the monitoring area is determined and the digital elevation model DEM data and basin boundary data of the basin are obtained to accurately define the basin range and provide basic data for subsequent water quality prediction. Then, the water quality data in the basin is obtained by setting monitoring points, focusing on monitoring areas with large water quality changes to avoid the shortcomings of evenly distributed monitoring points in traditional methods. Next, environmental variables in the basin are obtained, including meteorological, soil, precipitation and other variables, and combined with the spatial flow river network topological structure analysis to generate accurate predicted river networks and prediction points. Subsequently, by constructing a spatial flow river network topological structure model, combined with the water quality data and environmental data of the monitoring points, water quality parameters are predicted. Finally, the prediction results are displayed through GIS visualization tools to facilitate real-time monitoring and water quality early warning. In actual applications, this technical solution intelligently selects key areas for monitoring and prediction based on factors such as the basin's topography, river network structure, and environmental variables, avoiding the problem that traditional methods cannot effectively consider spatial correlation. It is not only suitable for water quality prediction in large-scale basins, but also can monitor changes in basin water quality in real time, providing a scientific basis for basin management and pollution prevention and control.

[0025] Furthermore, in step S1, the Yangtze River Basin is selected as the detection area, and the basin DEM data and basin boundary data are obtained. The basin DEM data is obtained through satellite remote sensing or lidar technology, and the basin boundary data is obtained through remote sensing images and geographic information system technology.

[0026] In this embodiment, DEM data is obtained through satellite remote sensing or lidar and other technologies, which can accurately describe the elevation changes of the earth's surface. This data provides a basis for subsequent water flow simulation and watershed hydrological analysis. DEM data usually have different resolutions. The higher the resolution, the more detailed the terrain information obtained, which can help to more accurately define the flow path and direction of water bodies in the watershed. The watershed boundary vector data refers to the spatial data that describes the scope of the watershed, which is usually obtained through remote sensing images, geographic information system (GIS) technology and other means. The boundary data helps to clarify the geographical location and scope of the watershed, and plays an important role in the spatial river network topological structure model.

[0027] Furthermore, in step S2, the selection criteria for monitoring points are as follows: monitoring points should be set mainly in major rivers, tributaries and areas that may be polluted within the basin; When setting up monitoring points, full consideration should be given to the hydrological characteristics of the basin, environmentally sensitive areas, and the location of pollution sources; The specific method for setting up monitoring points is: combining the water flow network and river network topology of the basin, and intelligently selecting key areas to set up monitoring points based on the actual hydrological conditions of the basin.

[0028] Furthermore, Figure 2 The predicted river network and predicted points of the watershed generated by the GIS tool in this technical solution are as follows: Figure 2 As shown, the technical solution generates a predicted river network in the basin through a GIS tool and evenly distributes prediction points in the river network. This step includes steps S3 to S5.

[0029] Furthermore, in step S3, at each monitoring point, real-time monitoring equipment or sampling means are installed to obtain real-time data of water quality related indicators. The data can be obtained by an automated water quality monitoring station, which collects data in real time and uploads it to the monitoring system, or by combining regular sampling with laboratory analysis. After preprocessing, these monitoring data are stored in the system database to provide basic data for subsequent model training and prediction.

[0030] In this embodiment, the water quality evaluation indicators include dissolved oxygen content, pH value, total nitrogen content, nitrate nitrogen content, ammonia nitrogen content, total nitrogen and dissolved oxygen content, etc.

[0031] Furthermore, in step S4, the environmental variables include hydrological and meteorological data of forest coverage, land use type, soil type, precipitation, temperature, evaporation and river flow; the environmental variables are obtained through meteorological stations and remote sensing data, and data cleaning, format conversion and spatial interpolation processing are performed.

[0032] Furthermore, in step S5, based on the basin DEM data, boundary data and the set monitoring points, projection conversion and data format conversion processing are performed through the GIS module to generate a predicted river network. Here, a predicted river network of 1513 rivers in the Yangtze River Basin is generated. In the generated river network, multiple prediction points are evenly distributed and set according to the characteristics of the basin and the changing laws of water quality parameters. These prediction points will play a core role in the subsequent water quality prediction and provide spatially distributed water quality prediction results.

[0033] Furthermore, Figure 3 The following are the concepts of the tail-up space structure and the four tail-down space structures: Figure 3 As shown, the technical solution describes the influence of water flow and the transmission of water quality in the basin by setting a spatial stream network topology, wherein the branching nature of the river system requires the sliding mean function to be split at the confluence of the river, which is accomplished by using a spatial weighting scheme based on basin characteristics such as upstream area or flow. By using a mixed effects modeling framework, the spatial stream network model (SSNM) is able to interpret the variance explained by a set of landscape covariates as fixed effects and the variance explained by different autocovariance functions as random effects. In addition to Euclidean autocorrelation, SSNM can also explicitly quantify the spatial dependence between sites in connected and disconnected flow modes through tail-up and tail-down autocovariance structures, respectively, which includes steps S6 to S8.

[0034] Furthermore, in step S6, the formula for calculating the value of the water quality parameter at the prediction point is: Y = Xβ+ zTU + zTD + zEU +ε Among them, Y is the vector of response variables, the response variables are the total nitrogen and dissolved oxygen contents at the monitoring points, X is the matrix of covariates, the covariates are environmental variables, β is the parameter vector of each covariate, zTU, zTD and zEU are vectors of random variables with tail-up, tail-down and Euclidean spatial autocorrelation structures, respectively, and ε is the vector of independent random errors.

[0035] Furthermore, the selection of covariates is as follows: by analyzing the correlation between water quality parameters and environmental variables, a regression relationship between environmental variables and water quality parameters is established, and the stepwise regression method is used to screen out the most predictive combination of variables from multiple candidate covariates to optimize the accuracy of the prediction model.

[0036] Furthermore, the screening of the spatial flow river network topological structure is as follows: after the regression model is established, the spatial structure related to the spatial flow river network is further selected, and the spatial structures such as tail-up, tail-down and Euclidean spatial autocorrelation are selected through random combination to generate all possible combinations and add them to the model. By simulating the interaction between river flow direction and water quality parameters, the model training and water quality prediction are promoted.

[0037] Furthermore, the water quality parameter values ​​are calculated specifically as follows: based on the trained regression model and the spatial stream network topology structure, the water quality parameter values ​​of each prediction point are calculated in combination with the environmental variables of each prediction point and the water quality information of the monitoring point.

[0038] Furthermore, step S7 is specifically as follows: using GIS tools to visualize the predicted water quality parameter values ​​of each river, the visualization results will show the spatial distribution of water quality in the basin, and users can directly view the water quality information of each river through the GIS interface. All visualization data will be uploaded to the system server for real-time monitoring and further analysis.

[0039] Furthermore, step S8 is specifically as follows: by comparing the predicted results with the data of the actual monitoring points, and evaluating the accuracy of the model through indicators such as the cross-validated R² value, RMSE (root mean square error) and AIC (Akaike Information Criterion), if it is found that there is a large deviation between the predicted value and the observed value, the model can be further optimized or new monitoring points can be added for correction, thereby improving the accuracy and generalization ability of the model.

[0040] Furthermore, Figure 4 In this technical solution, the predicted data of total nitrogen and dissolved oxygen concentration in water quality are cross-validated with the real data. 2 Figure, such as Figure 4 As shown in Figure 1, the results of evaluating the model accuracy through cross-validation are shown. From the prediction results in Table 1, it can be seen that the AIC and RMSE values ​​of the prediction method of the present invention are significantly lower than those of the traditional prediction results that do not consider the spatial stream network topology structure, and R 2 It reached 80%, which is much higher than the traditional method. It can be seen that the prediction accuracy has been significantly improved and the prediction results are very advantageous.

[0041] Table 1 Comparison of total nitrogen prediction results between traditional model and spatial stream network model (SSNM) Model RMSE AIC R2 Spatial Stream Network Model 1.23 2979.71 0.80 Traditional Model 1.92 3638.24 0.59 In summary, the implementation of this technical solution can significantly improve the accuracy of water quality prediction, optimize the layout of monitoring points, reduce the setting of redundant monitoring points, and thus reduce costs. In practical applications, key areas are intelligently selected for monitoring and prediction based on factors such as the basin's topography, river network structure, and environmental variables, avoiding the problem of spatial correlation that cannot be effectively considered in traditional methods. This technical solution is not only suitable for water quality prediction in large-scale basins, but can also monitor changes in basin water quality in real time, providing a scientific basis for basin management and pollution prevention and control.

[0042] What is disclosed above is only a preferred embodiment of the present invention, and it certainly cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. A method for predicting water quality in a river basin based on the spatial river network topology, characterized in that: The steps include: S1: Select the watershed area to be predicted according to the actual needs of the watershed, and obtain the watershed DEM data and watershed boundary data; S2: After determining the watershed area, it is necessary to set up monitoring points based on the water flow distribution, environmental variable characteristics, and water quality influencing factors in the watershed; S3: At each monitoring point, real-time monitoring equipment or sampling means are installed to obtain real-time data of water quality related indicators, and after pre-processing, the data is stored in the system database; S4: Obtain the environmental variables related to the watershed, and map these environmental variables to various locations in the watershed through data conversion and spatial interpolation processing; S5: Based on the basin DEM data, boundary data and the monitoring points that have been set, the GIS module in the system performs projection conversion and data format conversion processing to generate an accurate predicted river network, and evenly distributes and sets multiple prediction points according to the characteristics of the basin and the changing laws of water quality parameters; S6: Using the spatial flow river network topology model, the water quality parameters of the monitoring points are used as the response variables, the environmental variables are used as covariates, and the model is constructed by adding the spatial flow river network topology as a driving factor to calculate the values ​​of the water quality parameters of the prediction points; S7: Use GIS tools to visualize the predicted water quality parameter values ​​for each river channel; S8: By comparing the prediction results with the data from the actual monitoring points and evaluating the accuracy of the model through cross-validation, the model can be optimized or corrected by adding new monitoring points to improve the accuracy and generalization ability of the model.

2. The method for predicting water quality of a river basin based on a spatial river network topological structure according to claim 1, characterized in that: In step S1, the watershed DEM data is obtained through satellite remote sensing or lidar technology, and the watershed boundary data is obtained through remote sensing images and geographic information system technology.

3. The method for predicting water quality of a river basin based on a spatial river network topological structure according to claim 2, characterized in that: In step S2, the selection criteria for monitoring points are as follows: monitoring points should be set up mainly in major rivers, tributaries and areas that may be polluted within the basin; When setting up monitoring points, full consideration should be given to the hydrological characteristics of the basin, environmentally sensitive areas, and the location of pollution sources; The specific method for setting up monitoring points is: combining the water flow network and river network topology of the basin, and intelligently selecting key areas to set up monitoring points based on the actual hydrological conditions of the basin.

4. The method for predicting water quality of a river basin based on a spatial river network topological structure according to claim 3, characterized in that: In step S4, the environmental variables include hydrometeorological data of forest coverage, land use type, soil type, precipitation, temperature, evaporation, and river flow; Environmental variables are obtained through meteorological station and remote sensing data, and data cleaning, format conversion and spatial interpolation are performed.

5. The method for predicting water quality of a river basin based on a spatial river network topology structure according to claim 4, characterized in that: In step S6, the formula for calculating the value of the water quality parameter at the prediction point is: Y = Xβ+ zTU + zTD + zEU +ε Among them, Y is the vector of response variables, the response variables are the total nitrogen and dissolved oxygen contents at the monitoring points, X is the matrix of covariates, the covariates are environmental variables, β is the parameter vector of each covariate, zTU, zTD and zEU are vectors of random variables with tail-up, tail-down and Euclidean spatial autocorrelation structures, respectively, and ε is the vector of independent random errors.

6. The method for predicting water quality of a river basin based on a spatial river network topology structure according to claim 5, characterized in that: In step S6, the selection of covariates is specifically as follows: by analyzing the correlation between water quality parameters and environmental variables, a regression relationship between environmental variables and water quality parameters is established, and the stepwise regression method is used to screen out the most predictive variable combination from multiple candidate covariates to optimize the accuracy of the prediction model.

7. The method for predicting water quality of a river basin based on a spatial river network topological structure according to claim 6, characterized in that: In step S6, the screening of the spatial flow river network topological structure is specifically as follows: after the regression model is established, the spatial structure related to the spatial flow river network is further selected, and the spatial structures of tail-up, tail-down and Euclidean spatial autocorrelation are selected through random combination to generate all possible combinations and add them to the model. By simulating the interaction between river flow direction and water quality parameters, model training and water quality prediction are promoted.

8. The method for predicting water quality of a river basin based on a spatial river network topology structure according to claim 7, characterized in that: In step S6, the water quality parameter value is calculated specifically as follows: based on the trained regression model and the spatial stream network topology structure, the water quality parameter value of each prediction point is calculated in combination with the environmental data of each prediction point and the water quality information of the monitoring point.

9. A watershed water quality prediction system based on spatial river network topology, characterized in that: Applicable to the basin water quality prediction method based on the spatial river network topology structure as described in claim 8.

Images