River water quality prediction method and system based on land utilization dynamic change
By establishing a neural network data set for water quality and land use, using BP neural network and Plus model, river water quality prediction is carried out based on dynamic changes in land use, solving the limitations of traditional water quality prediction methods, and achieving long-term and dynamic prediction of river water quality and water quality management in complex situations.
Patent Information
- Application Number
- CN202411501349.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-25
- Publication Date
- 2025-02-14
AI Technical Summary
Traditional water quality prediction methods are difficult to accurately predict the long-term trend of water quality change, ignore the impact of land use changes on water quality, and are complex in calculations and have great limitations in data acquisition and treatment, making it difficult to deal with water quality prediction and management in complex situations.
By establishing a neural network data set for water quality and land use, using the BP neural network regression model and Plus model, river water quality prediction is carried out based on dynamic changes in land use, and water quality prediction of any monitoring point is achieved.
It realizes long-term and dynamic prediction of river water quality, can effectively respond to water quality changes in complex situations, integrates remote sensing technology, GIS analysis and machine learning models, and improves the accuracy and efficiency of water quality prediction.
Smart Images
Figure CN119443935A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of water quality prediction, and in particular to a method and system for predicting river water quality based on dynamic changes in land use. Background Art
[0002] With the continuous intensification of global climate change and human activities, the changing trends of river water quality have become more and more complex and difficult to predict. Traditional water quality monitoring methods usually rely on regular on-site sampling and laboratory analysis, which is not only time-consuming and labor-intensive, but also difficult to fully reflect the dynamic changes in water quality. At present, traditional water quality prediction methods such as time series methods and mixed models have many disadvantages such as high data quality requirements, complex calculations, and only focusing on the fitting effect of data. It is difficult to accurately predict the long-term trend of water quality changes, ignore the impact of factors such as land use changes on water quality, and cannot effectively respond to water quality changes caused by emergencies. These limitations make traditional methods increasingly limited in water quality prediction and management in complex situations such as rapid urbanization and industrialization.
[0003] With the continuous development of technologies such as geographic information systems, remote sensing technology and mathematical modeling, new means have been provided for water quality monitoring. Remote sensing images can provide large-scale, high-resolution land use data, and GIS can perform accurate spatial analysis on these data. However, existing research and applications are mostly focused on static analysis, and there is little research on the dynamic relationship between land use change and water quality change, lacking systematic methodological guidance. Summary of the invention
[0004] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a river water quality prediction method and system based on dynamic changes in land use, which uses existing big data to establish a water quality and land use neural network data set, and indirectly carries out water quality prediction at any monitoring point by substituting land use change prediction data, so as to solve the limitations of traditional water quality methods in data acquisition and processing.
[0005] To achieve the above object, the present invention provides the following solutions:
[0006] A river water quality prediction method based on dynamic changes in land use, including:
[0007] Counting existing river water quality monitoring sections, and collecting water quality data of the river water quality monitoring sections and elevation data and remote sensing image data of the areas where the river water quality monitoring sections are located;
[0008] Calculating the flow direction, flow rate and river network of the elevation data, and extracting the catchment area with the water quality section as the inclination point;
[0009] Perform unsupervised classification and interpretation on the remote sensing image data of the watershed area to obtain statistical land use data;
[0010] Using a BP neural network regression model to train the water quality data and the land use data to establish a training set;
[0011] Select the water quality section to be predicted, use the Plus model to predict the land use changes in the catchment area of the water quality section to be predicted, substitute the land use change data into the training set, and output the water quality prediction results of the water quality section to be predicted.
[0012] Preferably, the indicators of the water quality data include: dissolved oxygen, permanganate index, ammonia nitrogen, total nitrogen and total phosphorus.
[0013] Preferably, the flow direction, flow rate and river network of the elevation data are calculated, and the catchment area is extracted with the water quality section as the inclination point, including:
[0014] The elevation data is processed using the D8 flow direction algorithm in the GIS software to determine the flow direction data of each pixel point;
[0015] The flow direction data is processed using a flow accumulation tool in a GIS software to determine flow data for each pixel point;
[0016] Using a raster calculator in GIS software to classify the flow data into river networks;
[0017] Using the inclination point calculation tool in the GIS software to calculate the inclination point of the river water quality monitoring section;
[0018] Combining the results of flow, flow direction and river network calculations, the Watershed method in GIS software is used to determine the catchment area with the water quality section as the inclination point.
[0019] Preferably, unsupervised classification interpretation is performed on the remote sensing image data of the catchment area to obtain statistical land use data, including:
[0020] The ISODATA algorithm is used in ENVI software to classify the remote sensing image data of the watershed area to obtain unsupervised classification results;
[0021] The unsupervised classification results are interpreted and verified to determine the land use type represented by each category, and the area ratio of each land use type in the image is counted to obtain the land use data.
[0022] Preferably, the land use types include cultivated land, forest, shrub land, grassland, water body, bare land, construction land and wetland.
[0023] Preferably, the water quality data and the land use data are trained using a BP neural network regression model to establish a training set, including:
[0024] The BP neural network regression model was trained using different water quality indicators as variable Y and land use type data of the catchment area as independent variable X.
[0025] The parameters of the BP neural network regression model are continuously adjusted through the back propagation algorithm to make the BP neural network regression model adapt to the relationship between input data and output data;
[0026] A training set of water quality and land use data is established based on the trained BP neural network regression model.
[0027] Preferably, the elevation data includes DEM30 m data.
[0028] Preferably, the remote sensing image data includes Landsat8 remote sensing image data having a date consistent with the water quality data.
[0029] A river water quality prediction system based on dynamic changes in land use, including:
[0030] A data collection module is used to count existing river water quality monitoring sections, and collect water quality data of the river water quality monitoring sections and elevation data and remote sensing image data of the areas where the river water quality monitoring sections are located;
[0031] A data calculation module is used to calculate the flow direction, flow rate and river network of the elevation data, and extract the water catchment area with the water quality section as the inclination point;
[0032] A data statistics module is used to perform unsupervised classification interpretation on the remote sensing image data of the catchment area and to obtain statistical land use data;
[0033] A training module, used to train the water quality data and the land use data using a BP neural network regression model to establish a training set;
[0034] The prediction module is used to select the water quality section to be predicted, use the Plus model to predict the land use changes in the catchment area of the water quality section to be predicted, substitute the land use change data into the training set, and output the water quality prediction results of the water quality section to be predicted.
[0035] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:
[0036] The present invention provides a method and system for predicting river water quality based on dynamic changes in land use, the method comprising: counting existing river water quality monitoring sections, and collecting water quality data of the river water quality monitoring sections and elevation data and remote sensing image data of the area where the river water quality monitoring sections are located; calculating the flow direction, flow rate and river network of the elevation data, and extracting the catchment area with the water quality section as the inclination point; performing unsupervised classification interpretation on the remote sensing image data of the catchment area, and counting land use data; training the water quality data and the land use data using a BP neural network regression model to establish a training set; selecting a water quality section to be predicted, predicting land use changes in the catchment area of the water quality section to be predicted using the Plus model, and substituting the land use change data into the training set, and outputting the water quality prediction result of the water quality section to be predicted. The present invention uses a large amount of water quality data and land use data to establish a water quality and land use data set, uses land use change prediction data to indirectly carry out water quality prediction, and realizes long-term and dynamic prediction of river water quality. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] 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 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 labor.
[0038] Figure 1 A flow chart of a method provided by an embodiment of the present invention;
[0039] Figure 2 A schematic diagram of a technical route provided for an embodiment of the present invention;
[0040] Figure 3 A schematic diagram of obtaining prediction results provided by an embodiment of the present invention;
[0041] Figure 4 A schematic diagram of a water catchment area of a water quality monitoring point provided by an embodiment of the present invention;
[0042] Figure 5 A statistical diagram of land use type data in a watershed area provided by an embodiment of the present invention;
[0043] Figure 6 A statistical diagram of land use type prediction data for a watershed area of the PLUS model provided in an embodiment of the present invention;
[0044] Figure 7 A schematic diagram of the system structure provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0045] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0046] The purpose of the present invention is to provide a river water quality prediction method and system based on dynamic changes in land use, which uses existing big data to establish a water quality and land use neural network data set, and indirectly carries out water quality prediction at any monitoring point by substituting land use change prediction data, so as to solve the limitations of traditional water quality methods in data acquisition and processing.
[0047] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.
[0048] Figure 1 A flow chart of a method provided by an embodiment of the present invention, such as Figure 1 As shown, the present invention provides a river water quality prediction method based on dynamic changes in land use, comprising:
[0049] Step 100: Count existing river water quality monitoring sections, and collect water quality data of the river water quality monitoring sections and elevation data and remote sensing image data of the areas where the river water quality monitoring sections are located;
[0050] Step 200: Calculate the flow direction, flow rate and river network of the elevation data, and extract the catchment area with the water quality section as the inclination point;
[0051] Step 300: Perform unsupervised classification interpretation on the remote sensing image data of the watershed area to obtain statistical land use data;
[0052] Step 400: using a BP neural network regression model to train the water quality data and the land use data to establish a training set;
[0053] Step 500: Select a water quality section to be predicted, use the Plus model to predict land use changes in the catchment area of the water quality section to be predicted, substitute the land use change data into the training set, and output the water quality prediction result of the water quality section to be predicted.
[0054] like Figure 2 and Figure 3As shown, a river water quality prediction method based on dynamic changes in land use in this embodiment includes: counting existing river water quality monitoring sections, collecting water quality data of the sections and elevation data and remote sensing image data of the areas where they are located; using GIS to calculate the flow direction, flow rate and river network of the elevation data, and extracting the catchment area with the water quality section as the inclination point; using ENVI software to perform unsupervised classification and interpretation on the catchment area remote sensing image data, and counting the land use data; using the BP neural network regression model to train the water quality data and land use data to establish a training set; selecting the water quality section that needs to be predicted, using the Plus model to predict the land use changes in the section catchment area and substituting the land prediction data into the training set, and outputting the water quality prediction result of the selected section.
[0055] Furthermore, this embodiment can pre-collect water quality data of different water quality monitoring sections on the river and remote sensing images and elevation data of the areas where they are located, interpret and extract land use data in the catchment area of the water quality monitoring section through GIS and ENVI software, and use BP neural network to establish a training set of land use and water quality data. Subsequently, any section of the river can be selected, and the PLUS model can be used to predict the land use data in the catchment area of the section, and the land use prediction data can be substituted into the training set to output the water quality prediction data. The following first introduces the data involved in the prediction process, and then introduces the training process of the water quality prediction model. In this embodiment, any monitoring section on the river is selected, and the collected data includes three categories: water quality data, remote sensing image data, and elevation data.
[0056] Specifically, statistics on existing water quality sections include all national and provincial monitoring sections; water quality data statistics include the total daily data to date, and indicators include: dissolved oxygen, permanganate index, ammonia nitrogen, total nitrogen, and total phosphorus; elevation data include national DEM30 m data; remote sensing images include Landsat8 remote sensing image data consistent with the water quality data date.
[0057] Optionally, the specific implementation steps of establishing the water quality section catchment area in this embodiment are as follows: Figure 4 As shown, including:
[0058] A1: Use the "D8 flow direction algorithm" in GIS software to process the elevation data to determine the flow direction of each pixel point;
[0059] A2: Use the flow accumulation tool in the GIS software to process the A1 flow direction data to determine the flow at each pixel point;
[0060] A3: Use the raster calculator in GIS software to classify the river network based on the flow data in A2;
[0061] A4: Use the inclination point calculation tool in the GIS software to calculate the inclination point of the water quality monitoring section selected in S1;
[0062] A5: Combining the results of flow, flow direction and river network calculation, use the Watershed method in GIS software to determine the catchment area with the water quality section as the inclination point.
[0063] Furthermore, the specific implementation steps for collecting land use data of the catchment area are as follows: Figure 5 As shown, including:
[0064] B1: Use ISODATA algorithm in ENVI software to classify watershed remote sensing image data;
[0065] B2: Interpret and verify the unsupervised classification results, determine the land use type represented by each category, and count the area proportion of each land type in the image.
[0066] Among them, land use types include: cultivated land, forest, shrubland, grassland, water body, bare land, construction land and wetland.
[0067] Furthermore, the steps to build a training set of water quality and land use data are as follows:
[0068] C1: Different water quality indicators are used as variables Y, and the land use type data of the catchment area is used as the independent variable X (X ≥ 1) for training;
[0069] C2: Continuously adjust network parameters through the back-propagation algorithm so that the model can adapt to the relationship between input data and output data;
[0070] C3: Establish water quality and land use data training set.
[0071] Optionally, based on the existing land use data, the spatiotemporal autoregressive model in the Plus model is used to simulate and predict the dynamic changes of land in the water quality section catchment area of the embodiment, and the prediction data of different land use types are statistically analyzed. Figure 6 shown.
[0072] Specifically, the predicted data of different land use types in the cross-section catchment area are used as the independent variable X* (X*≥1), and X* is substituted into the established training set to obtain the water quality prediction data.
[0073] Corresponding to the above method, such as Figure 7 As shown, this embodiment also provides a river water quality prediction system based on dynamic changes in land use, including:
[0074] A data collection module is used to count existing river water quality monitoring sections, and collect water quality data of the river water quality monitoring sections and elevation data and remote sensing image data of the areas where the river water quality monitoring sections are located;
[0075] A data calculation module is used to calculate the flow direction, flow rate and river network of the elevation data, and extract the water catchment area with the water quality section as the inclination point;
[0076] A data statistics module is used to perform unsupervised classification interpretation on the remote sensing image data of the catchment area and to obtain statistical land use data;
[0077] A training module, used to train the water quality data and the land use data using a BP neural network regression model to establish a training set;
[0078] The prediction module is used to select the water quality section to be predicted, use the Plus model to predict the land use changes in the catchment area of the water quality section to be predicted, substitute the land use change data into the training set, and output the water quality prediction results of the water quality section to be predicted.
[0079] The beneficial effects of the present invention are as follows:
[0080] (1) The present invention uses a large amount of water quality data and land use data to establish a water quality and land use data set, and uses land use change prediction data to indirectly carry out water quality prediction, thereby achieving long-term and dynamic prediction of river water quality;
[0081] (2) The present invention uses land use change and water quality training sets to carry out water quality prediction at any water quality monitoring point in any river;
[0082] (3) The present invention realizes dynamic prediction of river water quality by integrating remote sensing technology, GIS analysis and machine learning models.
[0083] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.
[0084] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of the present invention. At the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.
Claims
1. A river water quality prediction method based on dynamic changes in land use, characterized in that: include: Counting existing river water quality monitoring sections, and collecting water quality data of the river water quality monitoring sections and elevation data and remote sensing image data of the areas where the river water quality monitoring sections are located; Calculating the flow direction, flow rate and river network of the elevation data, and extracting the catchment area with the water quality section as the inclination point; Perform unsupervised classification and interpretation on the remote sensing image data of the watershed area to obtain statistical land use data; Using a BP neural network regression model to train the water quality data and the land use data to establish a training set; Select the water quality section to be predicted, use the Plus model to predict the land use changes in the catchment area of the water quality section to be predicted, substitute the land use change data into the training set, and output the water quality prediction results of the water quality section to be predicted.
2. The method for predicting river water quality based on dynamic changes in land use according to claim 1, characterized in that: The indicators of the water quality data include: dissolved oxygen, permanganate index, ammonia nitrogen, total nitrogen and total phosphorus.
3. The method for predicting river water quality based on dynamic changes in land use according to claim 1, characterized in that: The flow direction, flow rate and river network are calculated for the elevation data, and the catchment area is extracted with the water quality section as the inclination point, including: The elevation data is processed using the D8 flow direction algorithm in the GIS software to determine the flow direction data of each pixel point; The flow direction data is processed using a flow accumulation tool in a GIS software to determine flow data for each pixel point; Using a raster calculator in GIS software to classify the flow data into river networks; Using the inclination point calculation tool in the GIS software to calculate the inclination point of the river water quality monitoring section; Combining the results of flow, flow direction and river network calculations, the Watershed method in GIS software is used to determine the catchment area with the water quality section as the inclination point.
4. The method for predicting river water quality based on dynamic changes in land use according to claim 1, characterized in that: The remote sensing image data of the watershed area is interpreted by unsupervised classification to obtain land use data, including: The ISODATA algorithm is used in ENVI software to classify the remote sensing image data of the watershed area to obtain unsupervised classification results; The unsupervised classification results are interpreted and verified to determine the land use type represented by each category, and the area ratio of each land use type in the image is counted to obtain the land use data.
5. The method for predicting river water quality based on dynamic changes in land use according to claim 1, characterized in that: The land use types include cultivated land, forest, shrubland, grassland, water body, bare land, construction land and wetland.
6. The method for predicting river water quality based on dynamic changes in land use according to claim 1, characterized in that: The water quality data and the land use data are trained using a BP neural network regression model to establish a training set, including: The BP neural network regression model was trained using different water quality indicators as variable Y and land use type data of the catchment area as independent variable X. The parameters of the BP neural network regression model are continuously adjusted through the back propagation algorithm to make the BP neural network regression model adapt to the relationship between input data and output data; A training set of water quality and land use data is established based on the trained BP neural network regression model.
7. The method for predicting river water quality based on dynamic changes in land use according to claim 1, characterized in that: The elevation data includes DEM 30 m data.
8. The method for predicting river water quality based on dynamic changes in land use according to claim 1, characterized in that: The remote sensing image data includes Landsat8 remote sensing image data with a date consistent with the water quality data.
9. A river water quality prediction system based on dynamic changes in land use, characterized in that: include: A data collection module is used to count existing river water quality monitoring sections, and collect water quality data of the river water quality monitoring sections and elevation data and remote sensing image data of the areas where the river water quality monitoring sections are located; A data calculation module is used to calculate the flow direction, flow rate and river network of the elevation data, and extract the water catchment area with the water quality section as the inclination point; A data statistics module is used to perform unsupervised classification interpretation on the remote sensing image data of the catchment area and to obtain statistical land use data; A training module, used to train the water quality data and the land use data using a BP neural network regression model to establish a training set; The prediction module is used to select the water quality section to be predicted, use the Plus model to predict the land use changes in the catchment area of the water quality section to be predicted, substitute the land use change data into the training set, and output the water quality prediction results of the water quality section to be predicted.