A remote sensing monitoring method for soil salinity in the Yellow River Delta integrating vegetation growth information

Abstract

The present invention relates to the field of remote sensing monitoring technology, and specifically to a remote sensing monitoring method for soil salinity in the Yellow River Delta that integrates vegetation growth information, comprising the following steps: acquiring remote sensing image data of the Yellow River Delta region, performing radiation correction, geometric correction and atmospheric correction operations on the remote sensing image data, extracting red light band and near infrared band information in the corrected remote sensing image data to calculate the vegetation index, and constructing a vegetation information image with reference to the vegetation index value. In the present invention, the remote sensing image is accurately corrected to obtain clearer vegetation index information, and then a vegetation information image is constructed. The data accuracy is effectively improved. Regression analysis is performed on the soil salinity observation data and the vegetation information image to identify the relationship between vegetation type and soil salinity. By analyzing the water absorption of vegetation, the soil salt distribution can be effectively inferred.

Description

Technical Field

[0001] The invention relates to the technical field of remote sensing monitoring, and in particular to a remote sensing monitoring method for soil salinity in the Yellow River Delta integrating vegetation growth information. Background Art

[0002] The field of remote sensing monitoring technology refers to a means of observing and monitoring the earth's surface and its environment from the air or space by using remote sensing technology. Remote sensing technology relies on platforms such as satellites, aircraft or drones, and uses electromagnetic waves (such as light, infrared, microwaves, etc.) to detect target areas, thereby obtaining information about ground objects, land use, vegetation cover, water bodies, climate change, etc., and is widely used in agriculture, environmental monitoring, urban planning, and disaster management.

[0003] Among them, the remote sensing monitoring method of soil salinity in the Yellow River Delta is a method of using remote sensing technology to monitor the salt content in the soil in the Yellow River Delta region. This technology evaluates the spatial distribution of soil salinity by analyzing different bands of data in remote sensing images, thereby providing support for land use, agricultural production and ecological protection. Through remote sensing monitoring, information on soil salinity in the region can be obtained in real time and extensively to prevent the impact of soil salinization.

[0004] Existing technologies mainly rely on direct analysis of remote sensing images when dealing with soil salinity monitoring in the Yellow River Delta, lacking in-depth correction and multi-level analysis of data. Although this method can provide extensive monitoring coverage, it is insufficient in data accuracy and detail. Due to the lack of full use of the relationship between vegetation indices and related soil salinity observation data, the understanding of the spatial distribution of soil salinity is relatively rough, and it is difficult to identify subtle abnormal changes and trends. This limitation can lead to misjudgment of the soil salinization process, thereby affecting land use planning and ecological protection strategies. In addition, existing technologies are relatively single in identifying and analyzing abnormal areas, and are unable to promptly detect and respond to the impact of potential environmental changes, limiting adaptability and response speed in dynamic environments. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a remote sensing monitoring method for soil salinity in the Yellow River Delta that integrates vegetation growth information.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a remote sensing monitoring method for soil salinity in the Yellow River Delta integrating vegetation growth information, comprising the following steps:

[0007] S1: Acquire remote sensing image data of the Yellow River Delta region, perform radiation correction, geometric correction and atmospheric correction operations on the remote sensing image data, extract the red light band and near infrared band information in the corrected remote sensing image data to calculate the vegetation index, and construct a vegetation information image with reference to the vegetation index value;

[0008] S2: collecting soil salinity observation data in the Yellow River Delta region, performing regression analysis in combination with the vegetation information image, determining the distribution of vegetation types in each soil salinity concentration range, and generating a relationship diagram between vegetation and soil salinity by mapping vegetation coverage and soil salinity data;

[0009] S3: constructing a geospatial weighted regression model based on the relationship diagram between vegetation and soil salinity, and predicting the soil salinity value, adding the vegetation index value as a regulating factor, adjusting the regression coefficient in the geospatial weighted regression model, and obtaining a salt spatial prediction map by displaying the predicted soil salt concentration distribution in the region;

[0010] S4: performing cluster analysis on the data in the salinity spatial prediction map, identifying abnormal salinity areas that deviate from the soil salinity threshold, analyzing the spatial distribution characteristics of abnormal soil salinity areas, visualizing through a geographic information system, and generating a soil salinity remote sensing monitoring map.

[0011] As a further solution of the present invention, the step of acquiring the vegetation information image is specifically as follows:

[0012] S111: Based on the remote sensing image data of the Yellow River Delta region, radiation correction, geometric correction and atmospheric correction operations are performed, and after correction, red light band and near infrared band data in the image are extracted and normalized to obtain processed red light band and near infrared band data;

[0013] S112: Based on the processed red light band and near infrared band data, the formula is used:

[0014] ,

[0015] Calculate the normalized difference vegetation index value NDVI to obtain vegetation index information;

[0016] Among them, NIR is the reflection value of the near infrared band, and Red is the reflection value of the red light band;

[0017] S113: Based on the vegetation index information, the normalized vegetation index value is graded to divide the image into a plurality of vegetation types or coverage levels, and in the geographic information system, visualization is performed by color mapping to generate a vegetation information image.

[0018] As a further solution of the present invention, the step of obtaining the distribution of vegetation types in each soil salt concentration range is specifically as follows:

[0019] S211: Based on the soil salinity observation data and the vegetation information image, the relationship between soil salinity concentration and vegetation type is obtained, and the influence of soil salinity on the distribution of vegetation types is analyzed by comparing the distribution of vegetation types in multiple salt concentration intervals, so as to obtain the analysis results of vegetation type and salt concentration;

[0020] S212: Based on the analysis results of the vegetation type and salt concentration and according to a preset vegetation index threshold, when the vegetation index in the area is greater than the preset vegetation index threshold, the target area is marked as having high vegetation coverage, and a vegetation type distribution record is obtained.

[0021] As a further solution of the present invention, the steps for obtaining the relationship diagram between vegetation and soil salinity are specifically as follows:

[0022] S221: Based on the vegetation type distribution record, the formula is used by spatial interpolation method:

[0023] ,

[0024] Calculate estimated values ​​for interpolation points , obtain the estimated results of soil salt concentration or vegetation coverage;

[0025] in, is the value of observation point i, is the distance between the interpolation point and the observation point i, p is the power of the distance, and n is the total number of observation points;

[0026] S222: Based on the estimated values ​​of the interpolation points in the soil salt concentration or vegetation coverage estimation result, the spatial expansion of the discrete data points is obtained, and the interpolated data is layered and visualized using a geographic information system to obtain a relationship diagram between vegetation and soil salinity.

[0027] As a further solution of the present invention, the step of obtaining the soil salinity prediction is specifically as follows:

[0028] S311: Based on the data in the relationship diagram between vegetation and soil salinity, the relationship between vegetation coverage and soil salinity concentration in the region is analyzed to determine independent variables and dependent variables, vegetation index values ​​are extracted as adjustment factors, and a geospatial weighted regression model is constructed;

[0029] S312: Based on the data in the geospatial weighted regression model, the formula is used:

[0030] ,

[0031] Calculate the predicted soil salinity value y and obtain the preliminary prediction result of soil salinity in the target area;

[0032] in, is the intercept of the regression model, is the regression coefficient, which reflects the influence of vegetation index on soil salinity prediction, x is the vegetation index value, is the error term.

[0033] As a further solution of the present invention, the steps for obtaining the salinity spatial prediction map are specifically as follows:

[0034] S321: Based on the preliminary prediction result of soil salinity in the target area, according to the soil salinity and vegetation status data of the adjacent areas, identify and analyze the key factors affecting soil salinity, optimize the prediction accuracy of soil salinity by adjusting the regression coefficient, and obtain an optimized prediction result;

[0035] S322: Importing the soil salinity data in the optimized prediction result into a geographic information system, performing a visualization operation of the spatial distribution, extracting the salt concentration information in the area, and generating a salt spatial prediction map.

[0036] As a further solution of the present invention, the steps for obtaining the spatial distribution characteristics of the abnormal soil salinity area are specifically as follows:

[0037] S411: Perform cluster analysis based on the data in the salinity spatial prediction map, using the formula:

[0038] ,

[0039] Calculate the cluster density D, which represents the concentration of salt values ​​in space, and obtain the cluster density value of salt values;

[0040] in, is the salt concentration value of each data point j, is the average value of salt concentration, and m is the total number of salt data points;

[0041] S412: Compare the cluster density value of the salinity value with the preset soil salinity threshold, identify the salt abnormality area that deviates from the soil salinity threshold, and combine the environmental characteristics of the area, including groundwater infiltration, seawater intrusion, natural precipitation and other factors for analysis to generate the soil salinity abnormality area analysis results.

[0042] As a further solution of the present invention, the steps for obtaining the soil salinity remote sensing monitoring map are specifically as follows:

[0043] S421: Based on the analysis result of the abnormal soil salinity area, combined with the real-time acquisition of the remote sensing image data of the current Yellow River Delta area, the remote sensing image of the Yellow River Delta area is updated to obtain the remote sensing image update result;

[0044] S422: Compare the remote sensing image update result with the salinity spatial prediction map, identify the soil salinity abnormal area, mark the location of the soil salinity abnormal area on the geographic information system, and generate a soil salinity remote sensing monitoring map.

[0045] Compared with the prior art, the advantages and positive effects of the present invention are:

[0046] In the present invention, the remote sensing image is accurately corrected to obtain clearer vegetation index information, and then a vegetation information image is constructed. The data accuracy is effectively improved. By combining soil salinity observation data with vegetation information images for regression analysis, the relationship between vegetation type and soil salinity can be identified. By analyzing the water absorption of vegetation, the soil salinity distribution can be effectively inferred, thereby improving the prediction accuracy. The prediction of soil salinity values ​​is further optimized through a geospatial weighted regression model, and the vegetation index is added to the model as a regulating factor, so that the prediction results are closer to the actual situation. The use of spatial clustering algorithms helps to identify abnormal salinity areas, and combines a variety of environmental factors for in-depth analysis to achieve accurate marking and visual display of abnormal soil salinity areas, providing more reliable data support for land use and ecological protection. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of the main steps of the present invention;

[0048] Figure 2 A flow chart of obtaining vegetation information images according to the present invention;

[0049] Figure 3 A flow chart of the present invention for determining the distribution of vegetation types in each soil salt concentration interval;

[0050] Figure 4 A flow chart of the relationship diagram between vegetation and soil salinity obtained by the present invention;

[0051] Figure 5 A flow chart of predicting soil salinity value according to the present invention;

[0052] Figure 6 A flow chart for obtaining a salinity spatial prediction map for the present invention;

[0053] Figure 7 A flow chart for analyzing the spatial distribution characteristics of abnormal soil salinity areas in the present invention;

[0054] Figure 8The present invention is a flow chart for obtaining a remote sensing monitoring map of soil salinity. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0056] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0057] See also Figure 1 The present invention provides a technical solution: a remote sensing monitoring method for soil salinity in the Yellow River Delta integrating vegetation growth information, comprising the following steps:

[0058] S1: Acquire remote sensing image data of the Yellow River Delta region, perform radiation correction, geometric correction and atmospheric correction operations on the remote sensing image data, extract the red light band and near infrared band information in the corrected remote sensing image data, normalize the red light band and near infrared band, calculate the vegetation index based on the normalized red light band and near infrared band, and construct a vegetation information image with reference to the vegetation index value;

[0059] S2: Collect soil salinity observation data in the Yellow River Delta region, perform regression analysis in combination with vegetation information images, determine vegetation types, including the distribution of crops, wetland plants, and natural vegetation in each soil salt concentration range, and identify the relationship between vegetation coverage and soil salinity. For example, through a preset vegetation index threshold, when the vegetation index in the area is greater than the preset vegetation index threshold, the target area is marked as having high vegetation coverage. High vegetation coverage can reduce salt accumulation on the soil surface through the water absorption of vegetation roots, increase the moisture content of the soil, and thus reduce the chance of salt rising to the surface. Therefore, soil salinity distribution is low in areas with high vegetation coverage. Use spatial interpolation to map vegetation coverage and soil salinity data, expand discrete data points to the entire space, and generate a relationship diagram between vegetation and soil salinity.

[0060] S3: Based on the relationship diagram between vegetation and soil salinity, a geospatial weighted regression model is constructed to predict soil salinity. In the geospatial weighted regression model, the vegetation index value is added as an adjustment factor. The impact of soil salinity and vegetation conditions in adjacent areas on the salinity of the target area is referred to, and the regression coefficient in the geospatial weighted regression model is adjusted to optimize the preliminary prediction results of soil salinity. By displaying the predicted soil salt concentration distribution in the region, a salt spatial prediction map is obtained;

[0061] S4: Use spatial clustering algorithm to cluster the data in the salt spatial prediction map, calculate the cluster density of the salt concentration in the current area through the salt value in the area, compare it with the preset soil salt threshold, identify the salt abnormality area that deviates from the soil salt threshold, and analyze the spatial distribution characteristics of the soil salt abnormality area in combination with the regional environmental characteristics, including groundwater infiltration, seawater intrusion, and natural precipitation. Regularly update the remote sensing image data of the Yellow River Delta area and compare it with the salt spatial prediction map, mark the spatial position of the soil salt abnormality area, visualize it through the geographic information system, and generate a soil salt remote sensing monitoring map;

[0062] The vegetation information image includes the normalized pixel values ​​of the red light band and the near-infrared band, the calculated vegetation index value and the vegetation coverage of each area. The relationship diagram between vegetation and soil salinity includes the spatial distribution of various types of vegetation in the soil salt concentration range, the soil salt concentration range division obtained by regression analysis, and the salt adaptability characteristics of various types of vegetation in the region. The salt spatial prediction map includes the salt concentration value predicted by the geospatial weighted regression model, the adjustment parameters of the influence of soil salinity and vegetation index values ​​in adjacent areas on the salt of the target area, and the optimized prediction results. The soil salt remote sensing monitoring map includes the salt abnormality areas obtained by spatial clustering analysis, the salt concentration deviation value compared with the preset soil salt threshold, and the spatial distribution characteristics of the salt abnormality areas.

[0063] See also Figure 2 , the specific steps for obtaining vegetation information images are:

[0064] S111: Based on the remote sensing image data of the Yellow River Delta region, radiation correction, geometric correction and atmospheric correction operations are performed, and after correction, red light band and near infrared band data in the image are extracted and normalized to obtain processed red light band and near infrared band data;

[0065] Acquire remote sensing image data of the Yellow River Delta region, perform radiation correction, geometric correction and atmospheric correction operations on the remote sensing image data, extract the red light band and near infrared band information in the corrected remote sensing image data, and normalize the red light band and near infrared band. During the execution process, first obtain the original image file according to the remote sensing image data. The image is usually in a multi-band data format, and each band represents different spectral information. When performing radiation correction, first correct the radiation error of the image according to the sensor characteristics. This process requires calibration of the characteristics of different sensors, and uses the sensor radiation calibration coefficient to adjust the brightness value of the image. The geometric correction is achieved by comparing the coordinates of the known control points and using the conversion algorithm ( Such as affine transformation or polynomial transformation) to correct the geometric distortion caused by the sensor position or other factors in the image, and then perform atmospheric correction. Usually, an atmospheric correction model (such as the 6S model) is used to correct the atmospheric influence according to the radiation information of the image, the observation angle of the sensor and the spectral characteristics of the atmosphere to ensure that the image can better reflect the real reflection information of the surface. Next, the red light band and the near-infrared band are extracted from the corrected image data. These two bands play an important role in the calculation of vegetation index. After extraction, the pixel values ​​of these two bands are normalized. Normalization is usually performed by subtracting the minimum value of the band from each pixel value and dividing it by the difference between the maximum and minimum values ​​to ensure that the normalized value is between 0 and 1.

[0066] S112: Based on the processed red light band and near infrared band data, the formula is used:

[0067] ,

[0068] Calculate the normalized difference vegetation index value NDVI to obtain vegetation index information;

[0069] Among them, NIR is the reflectance value of the near-infrared band, which usually refers to the spectral data with a wavelength range of 700 to 1300 nanometers. The near-infrared band has a high reflectance to vegetation and can usually effectively reflect the coverage and growth of vegetation. The pixel value of this band is extracted after preprocessing the remote sensing image (such as atmospheric correction, radiation correction, etc.). Red is the reflectance value of the red light band, which usually refers to the spectral data with a wavelength range of 620 to 680 nanometers. The red light band has a low reflectance to vegetation, so this band is more sensitive to the chlorophyll content of vegetation. After preprocessing the remote sensing image, the pixel value of this band can be directly extracted. Indicates the difference between the near-infrared band and the red light band. It represents the sum of the near-infrared band and the red light band, and is used to normalize the reflectance difference. Through the normalization of this value, the NDVI value can be limited to between -1 and +1, making it have a unified measurement standard.

[0070] Taking a certain pixel in an actual remote sensing image as an example, if the reflectance value of the red band (Red) is 0.35, the reflectance value of the near infrared band (NIR) is 0.75.

[0071] ,

[0072] The result shows that the NDVI value is 0.364, indicating that the vegetation growth in the area is at a medium level. Generally speaking, the NDVI value is between 0 and 1, and the larger the value, the more luxuriant the vegetation. For the NDVI value of 0.364, it means that the area has a certain vegetation coverage, but it has not yet reached a highly luxuriant level, which can be used to evaluate vegetation recovery, agricultural growth or ecological environment changes in the area.

[0073] S113: Based on the vegetation index information, the normalized vegetation index value is graded to divide the image into multiple vegetation types or coverage levels, and in the geographic information system, a visualization process is performed by color mapping to generate a vegetation information image;

[0074] After obtaining the vegetation index information, the NDVI value of each pixel can be visualized using the geographic information system. First, the calculated NDVI value is graded. According to different NDVI value ranges, the image is divided into different vegetation types or coverage levels. Usually, areas with NDVI values ​​below 0.2 are marked as no vegetation or sparse vegetation, those between 0.2 and 0.5 are marked as medium vegetation coverage, those between 0.5 and 0.8 are marked as high vegetation coverage, and those above 0.8 are marked as dense vegetation areas. By displaying the levels of different NDVI values ​​in different colors, a clear vegetation information image is formed. Finally, the image is output as needed to complete the visualization of vegetation coverage, which is convenient for further analysis and application.

[0075] See also Figure 3 , the specific steps for determining the distribution of vegetation types in each soil salt concentration range are:

[0076] S211: Based on the soil salinity observation data and the vegetation information image, the relationship between soil salinity concentration and vegetation type is obtained. By comparing the distribution of vegetation types in multiple salt concentration intervals, the influence of soil salinity on the distribution of vegetation types is analyzed, and the analysis results of vegetation type and salt concentration are obtained.

[0077] Soil salinity observation data in the Yellow River Delta region were collected and combined with vegetation information images for regression analysis. First, soil salinity data were obtained. Each soil sample corresponded to the salt concentration of a specific location and was marked according to the spatial coordinates. Subsequently, the obtained soil salinity data were combined with the calculated vegetation information images, which were calculated by the normalized vegetation index values ​​in the remote sensing image data. The normalized vegetation index values ​​reflect the growth status of ground vegetation within a certain range. Next, the relationship between soil salt concentration and vegetation type was analyzed by regression analysis. First, based on the different ranges of the normalized vegetation index values, the vegetation type of each area was determined. Generally, it can be divided into several types according to the normalized vegetation index values, including no vegetation, sparse vegetation, medium vegetation and high vegetation coverage. Then, the soil salt concentration range was grouped, and the distribution of different vegetation types in each group was compared to evaluate the distribution of different vegetation types under different salt concentrations. The purpose of regression analysis is to reveal the effect of soil salinity on the distribution of vegetation types, determine which vegetation types are more significant in different salinity ranges, and provide data support for practical applications.

[0078] S212: Based on the analysis results of vegetation type and salt concentration, according to a preset vegetation index threshold, when the vegetation index in the area is greater than the preset vegetation index threshold, the target area is marked as having high vegetation coverage, and a vegetation type distribution record is obtained;

[0079] To identify the relationship between vegetation coverage and soil salinity, first obtain the vegetation coverage data in the area through remote sensing images. These data are represented by normalized vegetation index values, which reflect the greenness and health of surface vegetation. Next, set a preset vegetation index threshold, which is generally determined based on historical data, vegetation type or specific characteristics of the target area. For example, the normalized vegetation index value threshold can be set to 0.5. When the normalized vegetation index value in the area is greater than the threshold, the area is marked as a high vegetation coverage area. Then, the high vegetation coverage area is compared and analyzed with the soil salinity data. Soil salinity data comes from soil sample collection and analysis. Generally, areas with high soil salt concentration have more serious surface salt accumulation, while areas with high vegetation coverage can reduce the accumulation of salt on the soil surface and promote the retention of soil moisture due to the water absorption of vegetation roots, thereby inhibiting the rise of salt to the surface. Through spatial analysis and data association, it was further confirmed that the soil salinity distribution in areas with high vegetation coverage was lower, and it was concluded that the vegetation in this area had a certain inhibitory effect on soil salinity.

[0080] See also Figure 4 , the steps for obtaining the relationship diagram between vegetation and soil salinity are as follows:

[0081] S221: Through spatial interpolation, based on vegetation type distribution records, using the formula:

[0082] ,

[0083] Calculate estimated values ​​for interpolation points , obtain the estimated results of soil salt concentration or vegetation coverage;

[0084] in, represents the soil salinity or vegetation cover at an unknown location. is the known value of observation point i (salt concentration or vegetation coverage), which is obtained by measuring the salt concentration or analyzing the vegetation coverage through remote sensing images. is the distance between the interpolation point and the observation point i. The direct linear distance between two points can be measured by the geographic information system. p is the power of the distance, usually 2, which reflects the attenuation effect of the distance. When selecting the power p, an optimal value can be determined through a series of experiments and data analysis. For example, a large number of observation point data with known salt concentration and vegetation coverage are collected from the target area, and these data are cross-validated. The data set is divided into a training set and a validation set. Different p values ​​(such as 1, 2, and 3) are used for interpolation calculations. The errors (such as mean square error or absolute error) between the interpolation results and the validation set data under different p values ​​are compared. The p value with the smallest error is selected as the empirical value of the interpolation calculation in the area. This value can be obtained through repeated experiments to ensure that the interpolation result has the highest accuracy and reliability. n is the total number of observation points, which can be obtained by counting the number of observation points.

[0085] For example, there are three observation points, whose soil salinity concentrations are 0.5, 0.7 and 0.6 respectively, and whose distances from the interpolation point are 10, 15 and 20 meters respectively. Substitute them into the formula:

[0086] ,

[0087] The result shows that the soil salt concentration estimated by the interpolation point is 0.568.

[0088] S222: based on the estimated values ​​of the interpolation points in the estimated results of soil salt concentration or vegetation coverage, obtain the spatial expansion of the discrete data points, use the geographic information system to overlay and visualize the interpolated data, and obtain a relationship diagram between vegetation and soil salt;

[0089] First, the spatial interpolation method is used to expand the discrete soil salinity data and vegetation coverage data points in the Yellow River Delta region to the entire target area. The continuous soil salinity concentration distribution and vegetation coverage distribution in the entire area are obtained through interpolation calculation. Next, the interpolated data are layered and visualized using a geographic information system. Specifically, the soil salinity concentration layer and the vegetation coverage layer can be superimposed according to a unified coordinate system to generate a comprehensive relationship diagram. The areas of different colors or tones in the diagram can intuitively show the spatial relationship between vegetation coverage and soil salinity concentration. By observing the changes in color in the diagram, the distribution trend of lower soil salinity in areas with high vegetation coverage can be analyzed. The final relationship diagram can be used for further ecological environmental analysis and land management decisions.

[0090] See also Figure 5 , the specific steps for obtaining the soil salinity prediction are:

[0091] S311: Based on the data in the relationship diagram between vegetation and soil salinity, the relationship between vegetation coverage and soil salt concentration in the region is analyzed to determine the independent and dependent variables, the vegetation index value is extracted as the adjustment factor, and a geospatial weighted regression model is constructed;

[0092] Collect soil salinity data and vegetation cover data, then, based on the geospatial features, select appropriate geospatial weighted regression models using a geographic information system (e.g., ArcGIS or QGIS) to help identify and manage geospatial features. Analyze the relationship between vegetation cover and soil salinity concentration in a region by using statistical analysis software (e.g., statistical libraries in R or Python, such as statsmodels or scikit-learn). For example, in the Yellow River Delta, it was found that areas with high vegetation density (high NDVI values) had lower soil salinity concentrations because vegetation affects soil water redistribution through transpiration, which in turn affects salt accumulation. It helps to determine the independent and dependent variables in the model and extract vegetation index values ​​from them as moderating factors in the model. Use soil salinity and vegetation status data from adjacent areas as references.

[0093] S312: Based on the data in the geospatial weighted regression model, the formula is:

[0094] ,

[0095] Calculate the predicted soil salinity value y and obtain the preliminary prediction result of soil salinity in the target area;

[0096] in, is the intercept of the regression model, i.e. the predicted soil salinity value when the independent variable is zero, which is determined by linear regression analysis. The specific method is as follows: collect historical data sets, including known soil salinity values ​​and corresponding NDVI, and use the least squares method to fit the linear regression model. The least squares method determines the best fit line by minimizing the squared difference between the predicted value and the actual value. The intercept value is part of the regression equation obtained from the least squares calculation, which is calculated using Python's scikitlearn library. is the regression coefficient, which reflects the influence of vegetation index on soil salinity prediction, which is also calculated by the least squares method. In linear regression, It can be solved by the formula: , and are the observation data of vegetation index and soil salinity, and is the average value of vegetation index and soil salinity, x is the vegetation index value, and refers to the normalized vegetation index. is the error term, reflecting the deviation of the model prediction, which is usually determined through deviation analysis of historical data during the model calibration process.

[0097] Taking a certain area as an example, the NDVI value x=0.364, the regression coefficient =1.2, intercept =0.1, the calculation process is as follows:

[0098] ,

[0099] The results show that the soil salinity value of the target area is predicted to be 0.5368.

[0100] See also Figure 6 , the specific steps for obtaining the salt spatial prediction map are:

[0101] S321: Based on the preliminary prediction results of soil salinity in the target area, according to the soil salinity and vegetation status data of the adjacent areas, identify and analyze the key factors affecting soil salinity, optimize the prediction accuracy of soil salinity by adjusting the regression coefficient, and obtain the optimized prediction results;

[0102] First, by analyzing the soil salinity and vegetation status data of adjacent areas, the key factors affecting the soil salinity in these areas are identified, and these factors are used to conduct a detailed assessment of the impact of salinity on the target area. The preliminary predicted soil salinity values ​​are applied in the geospatial weighted regression model, and the regression coefficients are adjusted to optimize the prediction accuracy. In specific operations, for example, the sensitivity of each variable to soil salinity is judged, including increasing or decreasing the value of each independent variable in turn, and observing the degree of influence on the soil salinity prediction results. For example, if increasing the value of an independent variable significantly changes the prediction result, it means that the independent variable is highly sensitive to soil salinity. The sensitivity of different independent variables to the results needs to be considered during adjustment. After optimization, the adjusted regression coefficients are re-entered into the model for calculation, and the adjusted model is verified to ensure that the model's prediction ability has been improved. Combined with the actual observation data of adjacent areas, the model parameters are continuously corrected, and the weights or coefficients of the independent variables are adjusted in the geospatial regression model to ultimately optimize a more accurate soil salinity prediction result.

[0103] S322: importing the soil salinity data in the optimized prediction results into a geographic information system, performing a visualization operation of the spatial distribution, extracting the salt concentration information in the region, and generating a salt spatial prediction map;

[0104] By displaying the predicted soil salt concentration distribution in the region, the predicted soil salt data is first imported into the geographic information system for visualization of the spatial distribution, from which the salt concentration data of each point in the region is extracted, a digital elevation model is created, and a salt concentration distribution map of the geographic space is generated. Then, color scales and contour lines are used to represent the salt concentration levels in different regions in order to clearly present the spatial changes in soil salinity. At the same time, the accuracy of the prediction model is evaluated through cross-validation to ensure that the prediction results of the model are consistent with the actual observed values. Finally, the prediction results are visualized and displayed in the form of a salt spatial prediction map, which provides a basis for soil management and decision-making in the region and obtains a salt spatial prediction map.

[0105] See also Figure 7 ,The specific steps for obtaining the spatial distribution characteristics of soil salinity anomaly areas are:

[0106] S411: Based on the data in the salinity spatial prediction map and cluster analysis, the formula is:

[0107] ,

[0108] Calculate the cluster density D, which represents the concentration of salt values ​​in space, and obtain the cluster density value of salt values;

[0109] in, is the salt concentration value of each data point j, which is extracted through satellite images in the area. is the average value of the salt concentration, which is the sum of the salt values ​​of all data points divided by the total number of data points, and m is the total number of salt data points, which is usually obtained by recording the number of measurements or the number of data records.

[0110] For example, if there are five data points, their values ​​are: 15, 20, 20, 22, 18.

[0111] Calculate the average:

[0112] ,

[0113] Compute the sum of the squared deviations of each data point from the mean:

[0114] ,

[0115] Calculate cluster density:

[0116] ,

[0117] The results show that the cluster density of regional salt concentration is 5.6.

[0118] S412: Compare the cluster density value of the salt value with the preset soil salt threshold, identify the salt abnormality area that deviates from the soil salt threshold, and analyze the environmental characteristics of the area, including groundwater infiltration, seawater intrusion, and natural precipitation, to generate the soil salt abnormality area analysis result;

[0119] After calculating the cluster density of salt concentration in the region as 5.6, the result is compared with the preset soil salinity threshold. Assuming that the preset soil salinity threshold is 3, first, through comparison, it can be found that the cluster density of the current area is higher than the threshold, which indicates that there is salt anomaly in the area and further analysis and intervention are needed. In the process of identifying the salt anomaly area, a detailed analysis is conducted in combination with the environmental characteristics of the area, including: evaluating the impact of groundwater infiltration on salt accumulation through hydrogeological surveys, and using groundwater level monitoring data to determine the direction and rate of groundwater flow to identify areas that cause salinity increase; at the same time, combining marine data and geological maps of coastal areas to evaluate the contribution of seawater intrusion to soil salinity, especially in low-lying areas where seawater backflow occurs, and analyzing the changes in chlorides and other salts by observing the chemical composition of surface and groundwater; in addition, the role of natural precipitation on soil moisture and salt loss is analyzed through meteorological data, and the impact of rainfall frequency and intensity on soil salinity is considered. Through these analyses, a comprehensive distribution map of soil salt anomaly areas is formed, combined with actual geographical characteristics, salt accumulation or loss paths are identified, and corresponding land management and improvement strategies are formulated.

[0120] See also Figure 8 , the specific steps for obtaining soil salinity remote sensing monitoring maps are:

[0121] S421: Based on the analysis results of the abnormal soil salinity area, combined with the real-time acquisition of the current remote sensing image data of the Yellow River Delta area, the remote sensing image of the Yellow River Delta area is updated to obtain the remote sensing image update result;

[0122] Visualization was performed through a geographic information system. The specific operations included importing the Yellow River Delta’s geospatial data, including topography, land use type, and existing salt concentration data, onto a GIS platform. Appropriate visualization parameters were set using GIS tools, such as selecting a color gradient to represent different salt concentration levels, to clearly show the spatial characteristics of salt distribution. Subsequently, remote sensing image data of the Yellow River Delta region were regularly updated to ensure that the images reflected real-time changes.

[0123] S422: Compare the remote sensing image update result with the salinity spatial prediction map, identify the soil salinity abnormal area, mark the location of the soil salinity abnormal area on the geographic information system, and generate a soil salinity remote sensing monitoring map;

[0124] After completing the visualization of the geographic information system, the next step is to compare the latest remote sensing image with the salinity spatial prediction map. The specific operation includes precise spatial registration of the two data through GIS software to ensure the accuracy of the position, and integrating the two data sets in the same coordinate system by using layer overlay. During the analysis process, areas with abnormal soil salinity are identified and marked in space. These abnormal areas can be highlighted by setting different colors or graphic symbols. After marking, a comprehensive soil salinity remote sensing monitoring map is generated, showing the changes in salt concentration and its spatial distribution characteristics, providing visual support for further analysis and decision-making.

[0125] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A remote sensing monitoring method for soil salinity in the Yellow River Delta integrating vegetation growth information, characterized in that: The following steps are involved: Acquire remote sensing image data of the Yellow River Delta region, perform radiation correction, geometric correction and atmospheric correction operations on the remote sensing image data, extract the red light band and near infrared band information in the corrected remote sensing image data to calculate the vegetation index, and construct a vegetation information image with reference to the vegetation index value; Collect soil salinity observation data in the Yellow River Delta region, perform regression analysis in combination with the vegetation information image, determine the distribution of vegetation types in each soil salinity concentration range, and generate a relationship diagram between vegetation and soil salinity by mapping vegetation coverage and soil salinity data; Based on the relationship diagram between vegetation and soil salinity, a geospatial weighted regression model is constructed, and the soil salinity value is predicted. The vegetation index value is added as a regulating factor to adjust the regression coefficient in the geospatial weighted regression model. By displaying the predicted soil salt concentration distribution in the region, a salt spatial prediction map is obtained; Performing cluster analysis on the data in the salinity spatial prediction map, identifying abnormal salinity areas that deviate from the soil salinity threshold, analyzing the spatial distribution characteristics of abnormal soil salinity areas, visualizing through a geographic information system, and generating a soil salinity remote sensing monitoring map; The steps of obtaining the vegetation information image are specifically as follows: Based on the remote sensing image data of the Yellow River Delta region, radiation correction, geometric correction and atmospheric correction operations are performed. After correction, the red light band and near infrared band data in the image are extracted and normalized to obtain the processed red light band and near infrared band data. Based on the processed red light band and near infrared band data, the formula is adopted: ， Calculate the normalized difference vegetation index value NDVI to obtain vegetation index information; Among them, NIR is the reflection value of the near infrared band, and Red is the reflection value of the red light band; Based on the vegetation index information, the normalized vegetation index value is graded to divide the image into multiple vegetation types or coverage levels, and in the geographic information system, a visualization process is performed by color mapping to generate a vegetation information image; The steps for obtaining the distribution of vegetation types in each soil salt concentration range are specifically as follows: Based on the soil salinity observation data and the vegetation information image, the relationship between soil salinity concentration and vegetation type is obtained, and the influence of soil salinity on the distribution of vegetation types is analyzed by comparing the distribution of vegetation types in multiple salt concentration intervals, so as to obtain the analysis results of vegetation type and salt concentration; Based on the analysis results of the vegetation type and salt concentration, according to a preset vegetation index threshold, when the vegetation index in the area is greater than the preset vegetation index threshold, the target area is marked as having high vegetation coverage, and a vegetation type distribution record is obtained; The steps for obtaining the relationship diagram between vegetation and soil salinity are specifically as follows: Through spatial interpolation method, based on the vegetation type distribution record, the formula is adopted: ， Calculate estimated values ​​for interpolation points , obtain the estimated results of soil salt concentration or vegetation coverage; in, is the value of observation point i, is the distance between the interpolation point and the observation point i, p is the power of the distance, and n is the total number of observation points; Based on the estimated values ​​of the interpolation points in the soil salt concentration or vegetation coverage estimation results, the spatial expansion of the discrete data points is obtained, and the interpolated data is layered and visualized using a geographic information system to obtain a relationship diagram between vegetation and soil salinity; The steps for obtaining the soil salinity prediction are as follows: Based on the data in the relationship diagram between vegetation and soil salinity, the relationship between vegetation coverage and soil salinity concentration in the region is analyzed to determine the independent variable and the dependent variable, the vegetation index value is extracted as a regulating factor, and a geospatial weighted regression model is constructed; Based on the data in the geospatial weighted regression model, the formula is used: ， Calculate the predicted soil salinity value y and obtain the preliminary prediction result of soil salinity in the target area; in, is the intercept of the regression model, is the regression coefficient, which reflects the influence of vegetation index on soil salinity prediction, x is the vegetation index value, is the error term; The steps for obtaining the salinity spatial prediction map are specifically as follows: Based on the preliminary prediction results of soil salinity in the target area, according to the soil salinity and vegetation status data of adjacent areas, key factors affecting soil salinity are identified and analyzed, and the prediction accuracy of soil salinity is optimized by adjusting the regression coefficient to obtain an optimized prediction result; Importing the soil salinity data in the optimized prediction results into a geographic information system, performing a visualization operation of the spatial distribution, extracting the salt concentration information in the area, and generating a salt spatial prediction map; The steps for obtaining the spatial distribution characteristics of the soil salinity anomaly area are specifically as follows: Based on the data in the salinity spatial prediction map and cluster analysis, the formula is used: ， Calculate the cluster density D, which represents the concentration of salt values ​​in space, and obtain the cluster density value of salt values; in, is the salt concentration value of each data point j, is the average value of salt concentration, and m is the total number of salt data points; The cluster density value of the salinity value is compared with the preset soil salinity threshold to identify the salt abnormality area that deviates from the soil salinity threshold, and combined with the environmental characteristics of the region, including groundwater infiltration, seawater intrusion, natural precipitation and other factors, to generate the soil salinity abnormality area analysis results.

2. The remote sensing monitoring method for soil salinity in the Yellow River Delta integrating vegetation growth information according to claim 1 is characterized in that: The steps for obtaining the soil salinity remote sensing monitoring map are specifically as follows: Based on the analysis results of the abnormal soil salinity area, combined with the real-time acquisition of the current remote sensing image data of the Yellow River Delta area, the remote sensing image of the Yellow River Delta area is updated to obtain the remote sensing image update results; The remote sensing image update result is compared with the salinity spatial prediction map to identify the soil salinity abnormal area, and the location of the soil salinity abnormal area is marked on the geographic information system to generate a soil salinity remote sensing monitoring map.

Images