Ecological environment remote sensing data analysis method

Abstract

The invention discloses an ecological environment remote sensing data analysis method, and belongs to the technical field of ecological environment. The data analysis method comprises the following steps: collecting remote sensing image data of a monitored water area, and carrying out atmospheric correction and geometric correction on the remote sensing image data; utilizing a co-occurrence matrix method to extract water body spectral texture features so as to identify spectral texture feature wavebands related to water body turbidity, chlorophyll content and dissolved oxygen; constructing a water quality parameter inversion model, training and verifying the model by using field sampling data, and optimizing the model; regularly acquiring new remote sensing image data of a monitoring area, inputting the new remote sensing image data into the optimized water quality parameter inversion model, and monitoring spatial and temporal distribution changes of water pollution parameters; determining a key area according to an inversion result, and obtaining high-temporal-spatial-resolution supplementary data of the key area by using the unmanned aerial vehicle; and collecting meteorological and hydrological station observation data, and comprehensively analyzing the source and flow direction of pollutants in combination with the inverted water quality parameter spatial and temporal distribution information.

Description

Technical Field

[0001] The present application relates to the technical field of ecological environment, and more specifically, to a method for analyzing remote sensing data of ecological environment. Background Art

[0002] With the increasing prominence of global environmental problems, the monitoring and protection of the ecological environment have become increasingly important. As a means of non-contact and rapid acquisition of large-scale surface information, remote sensing technology has been widely used in ecological environment monitoring and management. Through the data obtained by remote sensing technology, it is possible to achieve coverage of large-scale areas, provide rich surface information, and provide important data support for ecological environment protection and management.

[0003] Traditional methods for analyzing remote sensing data of ecological environment mainly rely on single remote sensing parameters or indicators, such as surface reflectance, vegetation index, etc. This method can reflect the basic situation of the surface to a certain extent, but it is difficult to provide deeper and more comprehensive information, so the limitations are relatively large. In addition, traditional methods are less efficient in processing big data and are difficult to handle problems such as multi-source data fusion and spatio-temporal information fusion.

[0004] In view of the limitations of traditional methods, in recent years, some studies have proposed methods for analyzing remote sensing data of ecological environment based on multi-source data. These methods integrate remote sensing data from different sources, such as optical remote sensing data, radar remote sensing data, etc., and combine ground observation data and model simulation results to achieve deeper and more comprehensive analysis of the ecological environment. However, the existing methods still have problems such as low data processing efficiency and large data integration difficulty, and it is difficult to meet the requirements of ecological environment monitoring and management.

[0005] Therefore, there is an urgent need for a method that can comprehensively utilize multi-source remote sensing data for ecological environment analysis. This method should have the ability to efficiently process big data, realize the fusion and analysis of multi-source data, and provide more accurate and comprehensive information support for ecological environment monitoring and management. Summary of the Invention

[0006] In order to overcome a series of defects existing in the prior art, the purpose of the present application is to provide a method for analyzing remote sensing data of ecological environment in view of the above problems, including the following steps:

[0007] Step 1: Collect remote sensing image data of the monitored water area and perform atmospheric correction and geometric correction on it;

[0008] Step 2: Use the co-occurrence matrix method to extract the spectral texture features of the water body to identify the spectral texture feature bands related to water turbidity, chlorophyll content, and dissolved oxygen;

[0009] Step 3: Based on the extracted characteristic bands, construct a water quality parameter inversion model using the support vector machine algorithm, train and validate the model with on-site sampling data, and optimize the model.

[0010] Step 4: Regularly obtain new remote sensing image data of the monitoring area, input it into the optimized water quality parameter inversion model, and monitor the spatio-temporal distribution changes of water pollution parameters.

[0011] Step 5: According to the inversion results, determine the key areas, and use drones to obtain high spatio-temporal resolution supplementary data of the key areas.

[0012] Step 6: Collect meteorological and hydrological station observation data, combine with the spatio-temporal distribution information of the inverted water quality parameters, and comprehensively analyze the sources and flows of pollutants.

[0013] Further, Step 2 includes the following steps:

[0014] Preprocess the water body spectral image, including grayscaling, denoising, and contrast enhancement, to improve the recognizability of texture features.

[0015] Calculate the gray-level co-occurrence matrix based on the distance and angle of pixel pairs.

[0016] Extract texture features from the gray-level co-occurrence matrix to reflect the uniformity, clarity, and complexity of the image texture.

[0017] Analyze the correlation between the extracted texture features and water turbidity, chlorophyll content, and dissolved oxygen.

[0018] According to the results of feature analysis, identify the spectral texture feature bands with the strongest correlation with water turbidity, chlorophyll content, and dissolved oxygen.

[0019] Further, the calculation formula of the gray-level co-occurrence matrix is:

[0020]

[0021] where Δx and Δy represent the distances of the pixel pair in the horizontal and vertical directions respectively, i and j represent the gray levels, N is the total number of pixel pairs that meet the conditions, and M is the size of the image.

[0022] Further, use the Pearson correlation coefficient to analyze the correlation between the extracted texture features and water turbidity, chlorophyll content, and dissolved oxygen. The specific formula is:

[0023]

[0024] where x i and y i are the standardized texture feature values and water quality parameter values respectively, and It is their average value.

[0025] Furthermore, step 3 includes the following steps:

[0026] Collect on-site sampling data, including measured values of water quality parameters and corresponding spectral data;

[0027] Use the selected characteristic bands as input and the measured values of water quality parameters as output to construct a water quality parameter inversion model based on the support vector machine algorithm;

[0028] Use 80% of the on-site sampling data as the training set and optimize the model parameters through the cross-validation method;

[0029] Use the remaining 20% of the on-site sampling data as the test set to evaluate the performance of the model;

[0030] According to the verification results, use the genetic algorithm to automatically optimize the model parameters and improve the inversion accuracy.

[0031] Furthermore, using the genetic algorithm to automatically optimize the model parameters includes the following steps:

[0032] Step 1, define the fitness function as the reciprocal of the mean square error;

[0033] Step 2, randomly generate an initial population, and each individual represents a set of model parameters;

[0034] Step 3, calculate the fitness of each individual in the population;

[0035] Step 4, select excellent individuals according to the fitness to enter the next generation;

[0036] Step 5, generate new individuals through crossover operations;

[0037] Step 6, perform mutation operations on the new individuals to introduce new genetic diversity;

[0038] Step 7, replace the individuals with lower fitness in the current population with the newly generated individuals;

[0039] Step 8, termination condition: check whether the termination condition is met. If not, return to step 3.

[0040] Furthermore, step 4 includes the following steps:

[0041] Regularly use satellites or drones to obtain new remote sensing image data of the monitoring area and perform necessary preprocessing on the obtained remote sensing images;

[0042] Extract the spectral texture feature bands with the strongest correlations with water turbidity, chlorophyll content, and dissolved oxygen from the preprocessed image data and input them into the water quality parameter inversion model;

[0043] Run the model to invert the water turbidity, chlorophyll content, and dissolved oxygen parameters. The specific formula is as follows: Among them, C x is the water quality parameter, R rs1 and R rs2 are the remote sensing reflectances in different bands, and a and b are the model coefficients;

[0044] Use Kriging interpolation method to analyze the inversion results, monitor the spatio-temporal distribution changes of water pollution parameters, and generate pollution distribution maps. The specific formula is as follows: Among them, Z(x0) is the value at the position to be estimated, λ i is the weight coefficient, and Z(x i ) is the value at the known position.

[0045] Furthermore, step 5 includes the following steps:

[0046] Analyze the output results of the water quality parameter inversion model to identify areas with abnormal or significantly changing pollution indicators;

[0047] Based on the inversion results, historical pollution data, and topographic and geomorphic information, comprehensively judge and determine the key areas. The specific formula is as follows: Among them, w i is the weight of data source i, and X i is the observed value of data source i;

[0048] Develop a detailed drone flight plan for the key areas, including flight path, altitude, time, and frequency, to ensure efficient collection of the required data;

[0049] According to the flight plan, use the drone to fly in the key areas to obtain supplementary data with high spatio-temporal resolution;

[0050] Input the processed drone data as supplementary information into the water quality parameter inversion model to improve the accuracy and reliability of the model. Among them, the accuracy is evaluated using the error propagation formula. The specific formula is as follows: Among them, σ model and σ data are the standard errors of the model and the data respectively;

[0051] Use the drone data to verify the inversion results to ensure the accuracy of the pollution parameters in the key areas;

[0052] Integrate the drone data and the inversion results to form a more detailed pollution distribution map.

[0053] Furthermore, step 6 includes the following steps:

[0054] Step 6.1: Regularly collect observational data from meteorological and hydrological stations, including rainfall, temperature, humidity, wind speed, water level, and flow rate;

[0055] Step 6.2: Integrate the collected meteorological and hydrological data with the spatio-temporal distribution information of the water quality parameters inverted in Steps 4 and 5;

[0056] Step 6.3: Use statistical and geographic information system models, combine meteorological, hydrological, and water quality data, and establish a model for pollutant sources and flow directions;

[0057] Step 6.4: Run the model, trace the sources and flow directions of pollutants, and identify key pollution sources and affected areas;

[0058] Step 6.5: Evaluate the impact degree of different pollution sources on the water body, as well as the propagation paths and scopes of pollutants in the water body;

[0059] Step 6.6: Visualize the analysis results using GIS software, and create pollutant flow direction maps and hotspot distribution maps.

[0060] Furthermore, Step 6.3 includes the following steps:

[0061] Analyze the time series of meteorological and hydrological data to identify the periodicity and trends of the data;

[0062] Conduct correlation analysis between the water quality parameters inverted in Steps 4 and 5 and the meteorological and hydrological data;

[0063] Consider the contribution rates and influences of different pollution sources, as well as the roles of environmental factors, and establish a statistical model for pollutant sources and flow directions;

[0064] Use GIS software for spatial analysis to identify key pollution sources and affected areas.

[0065] Compared with the prior art, the beneficial effects of this application are as follows.

[0066] This application improves the efficiency and accuracy of water quality parameter inversion, realizes the spatio-temporal monitoring ability of water body pollution, provides high-resolution data support, and can comprehensively analyze pollution sources and flow directions at the same time, providing an important technical means for water environmental protection and water resource management. Brief Description of the Drawings

[0067] Figure 1 It is a flowchart of an ecological environment remote sensing data analysis method disclosed in an embodiment of this application. Detailed Embodiments

[0068] To make the objectives, technical solutions, and advantages of the present invention more clear, the following will describe the technical solutions in the embodiments of the present invention in more detail with reference to the accompanying drawings in the embodiments of the present invention. In the drawings, the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The described embodiments are some, but not all, of the embodiments of the present invention.

[0069] All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0070] The embodiments described below with reference to the accompanying drawings and directional terms are exemplary and are intended to explain the present invention and should not be construed as limiting the present invention.

[0071] As Figure 1 shown, an ecological environment remote sensing data analysis method includes the following steps:

[0072] Step 1: Collect remote sensing image data of the monitored water area and perform atmospheric correction and geometric correction on it;

[0073] Step 2: Use the co-occurrence matrix method to extract water body spectral texture features to identify spectral texture feature bands related to water turbidity, chlorophyll content, and dissolved oxygen;

[0074] Step 3: Based on the extracted feature bands, use the support vector machine algorithm to construct a water quality parameter inversion model, train and verify the model using on-site sampling data, and optimize the model;

[0075] Step 4: Regularly obtain new remote sensing image data of the monitored area, input it into the optimized water quality parameter inversion model, and monitor the spatio-temporal distribution changes of water pollution parameters;

[0076] Step 5: According to the inversion results, determine the key areas and use drones to obtain high spatio-temporal resolution supplementary data of the key areas;

[0077] Step 6: Collect meteorological and hydrological station observation data, combine with the spatio-temporal distribution information of the inverted water quality parameters, and comprehensively analyze the sources and flows of pollutants.

[0078] In this embodiment, by collecting remote sensing image data and performing atmospheric correction and geometric correction, accurate water body information can be obtained. The spectral texture features extracted by the co-occurrence matrix method can identify the characteristic bands related to water turbidity, chlorophyll content, and dissolved oxygen, thereby realizing the precise monitoring of water quality. Based on the extracted characteristic bands, a water quality parameter inversion model constructed using the support vector machine algorithm can achieve the real-time inversion of water quality parameters. By regularly obtaining new remote sensing image data of the monitoring area, the spatio-temporal distribution changes of water pollution parameters can be monitored in real time. According to the key areas determined by the inversion results, drones can be used to obtain supplementary data with high spatio-temporal resolution, thereby realizing the efficient tracking of pollution sources. Combining the collected meteorological and hydrological station observation data, the sources and flow directions of pollutants can be comprehensively analyzed. Generally speaking, this method can achieve the precise monitoring and real-time inversion of water pollution, improve the efficiency and accuracy of water quality monitoring, and is of great significance for environmental protection and pollution control. At the same time, this method can also provide valuable data and references for related scientific research work.

[0079] Further, step 2 includes the following steps:

[0080] Preprocess the water body spectral image, including grayscale conversion, noise removal, and contrast enhancement, to improve the recognizability of texture features; Grayscale conversion is to convert a color image into a grayscale image, which can reduce the computational complexity. Noise removal is to eliminate random noise in the image and improve the image quality. Contrast enhancement is to make the details of the image more obvious.

[0081] Calculate the gray-level co-occurrence matrix based on the distance and angle of pixel pairs; The gray-level co-occurrence matrix is a statistical method used to measure the texture features of an image. It considers the spatial relationship between pixel pairs and constructs a matrix by calculating the frequency distribution of pixel pairs with specific distances and angles.

[0082] Extract texture features from the gray-level co-occurrence matrix to reflect the uniformity, clarity, and complexity of the image texture; Common texture features include contrast, correlation, energy, and homogeneity, etc.

[0083] Analyze the correlation between the extracted texture features and water turbidity, chlorophyll content, and dissolved oxygen; This step is to find out which texture features have a strong correlation with water turbidity, chlorophyll content, and dissolved oxygen. This is usually achieved through statistical analysis methods.

[0084] According to the results of feature analysis, identify the spectral texture feature bands with the strongest correlation with water turbidity, chlorophyll content, and dissolved oxygen. This step is to determine the texture feature bands that can best reflect the water body properties. These bands can be used for subsequent water body monitoring and analysis.

[0085] Furthermore, the calculation formula of the gray-level co-occurrence matrix is:

[0086]

[0087] where Δx and Δy respectively represent the distances of the pixel pair in the horizontal and vertical directions, i and j represent the gray levels, N is the total number of pixel pairs that meet the conditions, and M is the size of the image.

[0088] Furthermore, the Pearson correlation coefficient analysis is used to extract the correlations between the texture features and water turbidity, chlorophyll content, and dissolved oxygen. The specific formula is:

[0089]

[0090] where x i and y i are respectively the normalized texture feature values and water quality parameter values, and are their average values.

[0091] Furthermore, step 3 includes the following steps:

[0092] Collect on-site sampling data, including the measured values of water quality parameters and the corresponding spectral data;

[0093] Use the selected characteristic bands as the input and the measured values of water quality parameters as the output to construct a water quality parameter inversion model based on the support vector machine algorithm;

[0094] Use 80% of the on-site sampling data as the training set and optimize the model parameters through the cross-validation method;

[0095] Use the remaining 20% of the on-site sampling data as the test set to evaluate the performance of the model;

[0096] According to the verification results, use the genetic algorithm to automatically optimize the model parameters to improve the inversion accuracy.

[0097] Furthermore, using the genetic algorithm to automatically optimize the model parameters includes the following steps:

[0098] Step 1, define the fitness function as the reciprocal of the mean square error;

[0099] Step 2, randomly generate the initial population, and each individual represents a set of model parameters;

[0100] Step 3, calculate the fitness of each individual in the population;

[0101] Step 4, select excellent individuals according to the fitness to enter the next generation;

[0102] Step 5, generate new individuals through the crossover operation;

[0103] Step 6, perform mutation operation on the new individuals to introduce new genetic diversity;

[0104] Step 7, replace the individuals with lower fitness in the current population with the newly generated individuals;

[0105] Step 8, termination condition: check whether the termination condition is met. If not, return to Step 3.

[0106] Furthermore, Step 4 includes the following steps:

[0107] Regularly use satellites or drones to obtain new remote sensing image data of the monitoring area, and perform necessary preprocessing on the obtained remote sensing images;

[0108] Extract the spectral texture feature bands with the strongest correlation with water turbidity, chlorophyll content, and dissolved oxygen from the preprocessed image data, and input them into the water quality parameter inversion model;

[0109] Run the model to invert the water turbidity, chlorophyll content, and dissolved oxygen parameters. The specific formula is: where C x is the water quality parameter, R rs1 and R rs2 are the remote sensing reflectances of different bands, and a and b are the model coefficients;

[0110] Use Kriging interpolation method to analyze the inversion results, monitor the spatio-temporal distribution changes of water pollution parameters, and generate pollution distribution maps. The specific formula is: where Z(x0) is the value at the position to be estimated, λ i is the weight coefficient, and Z(x i ) is the value at the known position.

[0111] Furthermore, Step 5 includes the following steps:

[0112] Analyze the output results of the water quality parameter inversion model, and identify the areas where the pollution indicators are abnormal or change significantly;

[0113] Based on the inversion results, historical pollution data, and topographic and geomorphic information, comprehensively judge and determine the key areas. The specific formula is: where w i is the weight of data source i, and X i is the observed value of data source i;

[0114] Develop a detailed drone flight plan for the key areas, including flight path, altitude, time, and frequency, to ensure that the required data can be collected efficiently;

[0115] According to the flight plan, use drones to fly in key areas to obtain supplementary data with high spatio-temporal resolution;

[0116] Input the processed drone data as supplementary information into the water quality parameter inversion model to improve the accuracy and reliability of the model. Among them, the accuracy is evaluated using the error propagation formula, and the specific formula is: Among them, where σ model and σ data are the standard errors of the model and the data respectively;

[0117] Use the drone data to verify the inversion results to ensure the accuracy of the pollution parameters in the key areas;

[0118] Integrate the drone data and the inversion results to form a more detailed pollution distribution map.

[0119] Furthermore, step 6 includes the following steps:

[0120] Step 6.1, Regularly collect observational data from meteorological and hydrological stations, including rainfall, temperature, humidity, wind speed, water level, and flow rate. These data are the basis for establishing models of pollutant sources and flow directions.

[0121] Step 6.2, Integrate the collected meteorological and hydrological data with the spatio-temporal distribution information of the water quality parameters inverted in steps 4 and 5; This step is to integrate the meteorological and hydrological data and the water quality parameter data to form a complete data set, providing data support for subsequent model establishment and analysis.

[0122] Step 6.3, Use statistical and geographic information system models, combined with meteorological and hydrological and water quality data, to establish models of pollutant sources and flow directions; This model can be used to predict and track the sources and flow directions of pollutants.

[0123] Step 6.4, Run the model to track the sources and flow directions of pollutants and identify key pollution sources and affected areas;

[0124] Step 6.5, Evaluate the impact degree of different pollution sources on the water body, as well as the propagation paths and scopes of pollutants in the water body; This information is very important for formulating effective water quality protection and improvement measures.

[0125] Step 6.6, Visualize the analysis results using GIS software to create pollution flow maps and hotspot distribution maps. These charts can intuitively display the sources, flow directions, and distributions of pollutants, helping people better understand and solve water quality problems.

[0126] Furthermore, step 6.3 includes the following steps:

[0127] Analyze the time series of meteorological and hydrological data to identify the periodicity and trends of the data;

[0128] Perform correlation analysis on the water quality parameters inverted in steps 4 and 5 with the meteorological and hydrological data;

[0129] Consider the contribution rates and influences of different pollution sources, as well as the roles of environmental factors, and establish a statistical model for the sources and flows of pollutants;

[0130] Use GIS software for spatial analysis to identify key pollution sources and affected areas.

Claims

1. A method for remote sensing data analysis of ecological environment, characterized in that It includes the following steps: Step 1: Collect remote sensing image data of the monitored water area and perform atmospheric correction and geometric correction on it; Step 2: Use the co-occurrence matrix method to extract the spectral texture features of water bodies to identify the spectral texture feature bands related to water turbidity, chlorophyll content, and dissolved oxygen; Step 3: Based on the extracted feature bands, use the support vector machine algorithm to construct a water quality parameter inversion model, train and validate the model using on-site sampling data, and optimize the model; Step 4: Regularly obtain new remote sensing image data of the monitored area, input it into the optimized water quality parameter inversion model, and monitor the spatio-temporal distribution changes of water pollution parameters; Step 5: According to the inversion results, determine the key areas and use drones to obtain high spatio-temporal resolution supplementary data of the key areas; Step 6: Collect meteorological and hydrological station observation data, combine with the spatio-temporal distribution information of the inverted water quality parameters, and comprehensively analyze the sources and flow directions of pollutants.

2. The method for analyzing remote sensing data of ecological environment according to claim 1, characterized in that Step 2 includes the following steps: Preprocess the water body spectral image, including grayscale conversion, noise reduction, and contrast enhancement, to improve the recognizability of texture features; Calculate the gray-level co-occurrence matrix based on the distance and angle of pixel pairs; Extract texture features from the gray-level co-occurrence matrix to reflect the uniformity, clarity, and complexity of the image texture; Analyze the correlation between the extracted texture features and water turbidity, chlorophyll content, and dissolved oxygen; According to the results of feature analysis, identify the spectral texture feature bands with the strongest correlation with water turbidity, chlorophyll content, and dissolved oxygen.

3. The method for analyzing remote sensing data of ecological environment according to claim 2, characterized in that The calculation formula of the gray-level co-occurrence matrix is: where Δx and Δy respectively represent the distances of pixel pairs in the horizontal and vertical directions, i and j represent gray levels, N is the total number of pixel pairs that meet the conditions, and M is the size of the image.

4. The method for analyzing remote sensing data of ecological environment according to claim 3, characterized in that Use the Pearson correlation coefficient to analyze the correlation between the extracted texture features and water turbidity, chlorophyll content, and dissolved oxygen. The specific formula is: where x i and y i are the texture feature value and water quality parameter value after standardization respectively, and are their average values.

5. The method for analyzing remote sensing data of ecological environment according to claim 1, wherein Step 3 includes the following steps: Collect on-site sampling data, including measured values of water quality parameters and corresponding spectral data; Use the selected feature bands as input and the measured values of water quality parameters as output to construct a water quality parameter inversion model based on the support vector machine algorithm; Use 80% of the on-site sampling data as the training set and optimize the model parameters through the cross-validation method; Use the remaining 20% of the on-site sampling data as the test set to evaluate the performance of the model; According to the verification results, use the genetic algorithm to automatically optimize the model parameters to improve the inversion accuracy.

6. The method for analyzing remote sensing data of ecological environment according to claim 5, characterized in that Using the genetic algorithm to automatically optimize the model parameters includes the following steps: Step 1: Define the fitness function as the reciprocal of the mean square error; Step 2: Randomly generate an initial population, and each individual represents a set of model parameters; Step 3: Calculate the fitness of each individual in the population; Step 4: Select excellent individuals according to the fitness to enter the next generation; Step 5: Generate new individuals through crossover operations; Step 6: Perform mutation operations on the new individuals to introduce new genetic diversity; Step 7: Replace the individuals with lower fitness in the current population with the newly generated individuals; Step 8: Termination condition: Check whether the termination condition is met. If not, return to Step 3.

7. A method for analyzing remote sensing data of the ecological environment according to claim 1, characterized in that, Step 4 includes the following steps: Regularly use satellites or drones to obtain new remote sensing image data of the monitoring area, and perform necessary preprocessing on the obtained remote sensing images; Extract the spectral texture feature bands with the strongest correlation with water turbidity, chlorophyll content, and dissolved oxygen from the preprocessed image data, and input them into the water quality parameter inversion model; Run the model to invert the water turbidity, chlorophyll content, and dissolved oxygen parameters. The specific formula is as follows: Among them, C x is the water quality parameter, R rs1 and R rs2 are the remote sensing reflectances of different bands, and a and b are the model coefficients; Use Kriging interpolation method to analyze the inversion results, monitor the spatio-temporal distribution changes of water pollution parameters and generate pollution distribution maps. The specific formula is as follows: Among them, Z(x0) is the value at the position to be estimated, and λ i is the weight coefficient, and Z(x i ) is the value at the known position.

8. The method for analyzing remote sensing data of ecological environment according to claim 1, wherein Step 5 includes the following steps: Analyze the output results of the water quality parameter inversion model to identify areas with abnormal or significantly changing pollution indicators; Based on the inversion results, historical pollution data, and topographic and geomorphic information, comprehensively judge and determine the key areas. The specific formula is as follows: where w i is the weight of data source i, and X i is the observed value of data source i; Develop a detailed drone flight plan for key areas, including flight paths, altitudes, times, and frequencies, to ensure efficient collection of the required data; According to the flight plan, use the drone to fly in key areas to obtain supplementary data with high spatio-temporal resolution; The processed UAV data is input into the water quality parameter inversion model as supplementary information to improve the accuracy and reliability of the model. Among them, the accuracy is evaluated using the error propagation formula, and the specific formula is as follows: where σ model and σ data are the standard errors of the model and the data respectively; Use the drone data to verify the inversion results to ensure the accuracy of the pollution parameters in key areas; Integrate the drone data and the inversion results to form a more detailed pollution distribution map.

9. A method for analyzing remote sensing data of ecological environment according to claim 1, characterized in that, Step 6 includes the following steps: Step 6.1, Regularly collect observational data from meteorological and hydrological stations, including rainfall, temperature, humidity, wind speed, water level, and flow rate; Step 6.2, Integrate the collected meteorological and hydrological data with the spatio-temporal distribution information of the water quality parameters inverted in Steps 4 and 5; Step 6.3, Use statistical and geographic information system models, combined with meteorological and hydrological and water quality data, to establish a model of pollutant sources and flow directions; Step 6.4, Run the model to track the sources and flow directions of pollutants, and identify key pollution sources and affected areas; Step 6.5, Evaluate the impact degree of different pollution sources on the water body, as well as the propagation paths and scopes of pollutants in the water body; Step 6.6, Visualize the analysis results using GIS software to create a pollution flow map and a hot spot distribution map.

10. The method for analyzing remote sensing data of ecological environment according to claim 9, wherein Step 6.3 includes the following steps: Analyze the time series of meteorological and hydrological data to identify the periodicity and trends of the data; Conduct a correlation analysis of the water quality parameters inverted in Steps 4 and 5 with the meteorological and hydrological data; Consider the contribution rates and influences of different pollution sources, as well as the roles of environmental factors, to establish a statistical model of pollutant sources and flow directions; Use GIS software for spatial analysis to identify key pollution sources and affected areas.

Images