Remote sensing estimation method for grassland above-ground biomass

Abstract

The invention belongs to the technical field of remote sensing, and provides a grassland above-ground biomass remote sensing estimation method, which comprises the following steps of: collecting meteorological remote sensing data to be analyzed, and predicting a biomass estimation model; the construction process of the biomass estimation model comprises environmental database integration, interpolation processing, cutting and mean value calculation, remote sensing data processing, vegetation index pixel value acquisition, pixel data acquisition, biomass correlation coefficient calculation, feature data set acquisition, feature data set grouping and biomass estimation model training. According to the method, the meteorological data is interpolated, so that the meteorological data volume is enriched, and the robustness of the model is improved; by cutting and calibrating temperature, rainfall, remote sensing data and the like, the diversity and fineness of training data are improved, and the accuracy of the model is increased; and through the biomass correlation coefficient, the training data is optimized, so that the model training is more efficient and more reliable.

Description

Technical Field

[0001] The invention relates to the field of remote sensing technology, and in particular to a remote sensing estimation method for grassland biomass. Background Art

[0002] Grasslands are one of the most widespread vegetation types on Earth and play a vital role in the study of global carbon cycling and primary production.

[0003] Grassland above ground biomass, abbreviated as AGB, is the basic basis for determining the carrying capacity of pastures and the primary condition for determining livestock production. Rapid and accurate estimation of grassland above ground biomass is a hot issue in grassland ecosystem research. The traditional method of measuring grassland above ground biomass is to set up sample plots in the study area and conduct sampling, usually by cutting, drying and weighing vegetation. These traditional measurement methods require a lot of manpower, material resources and time costs, and are affected by sample plot selection and biomass sampling, and have certain limitations. Remote sensing technology has been widely used in grassland above ground biomass estimation and monitoring. It obtains large-scale surface information through satellite, aerial and other remote sensing data. The existing technology has proposed some grassland above ground biomass estimation methods based on remote sensing data. These methods combine remote sensing data, such as multispectral, hyperspectral and other data, to improve the accuracy and robustness of the estimation.

[0004] However, the resolution of these satellite remote sensing data is low, which will have a great impact on data extraction and cause the estimation results to be too biased. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a remote sensing estimation method for grassland aboveground biomass, which enriches the amount of meteorological data and improves the robustness of the model by interpolating meteorological data; improves the diversity and precision of training data and increases the accuracy of the model by cropping and calibrating temperature, rainfall, remote sensing data, etc.; optimizes training data through biomass correlation coefficient to make model training more efficient and reliable.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] A remote sensing estimation method for grassland aboveground biomass, comprising:

[0008] Collect historical meteorological data and remote sensing data for the grassland to be estimated;

[0009] Inputting the historical meteorological data to be analyzed and the remote sensing data to be analyzed into a preset biomass estimation model to obtain a target biomass estimation result;

[0010] The process of constructing the biomass estimation model includes:

[0011] Collect environmental information of a preset area and integrate it to obtain an environmental database; the environmental database includes: preset high-resolution remote sensing data, preset biomass historical data and preset meteorological data; the preset meteorological data includes: preset historical surface temperature data, preset precipitation data and DEM elevation data;

[0012] Performing interpolation processing on the preset meteorological data to obtain continuous meteorological data;

[0013] Clipping and averaging the preset historical surface temperature data and the preset precipitation data to obtain meteorological data pixel values;

[0014] Mosaicing, radiation calibration, atmospheric correction and cropping are performed on the preset high-resolution remote sensing data to obtain remote sensing image data corresponding to the preset area;

[0015] Extracting the vegetation index of the remote sensing image data to obtain a vegetation index pixel value;

[0016] Extracting pixel data of the points to be analyzed according to the meteorological data pixel values ​​and the vegetation index pixel values; the pixel data includes: the meteorological data pixel values ​​and the vegetation index pixel values ​​of all the points to be analyzed;

[0017] According to the pixel data, biomass correlation calculation is performed to obtain a biomass correlation coefficient;

[0018] Optimizing the pixel data using the biomass correlation coefficient to obtain a feature data set;

[0019] The feature data set is randomly divided into a training subset and a test subset in a ratio of 2:1;

[0020] The meteorological data pixel values ​​in the training subset are used as the input of the preset random forest regression model, the vegetation index pixel values ​​in the training subset are used as the output of the preset random forest regression model, the root mean square error and relative error are used as the loss function of the preset random forest regression model, the preset random forest regression model is trained, and the trained biomass estimation model is obtained.

[0021] Preferably, the vegetation index pixel value includes: normalized vegetation index, enhanced vegetation index, difference vegetation index, ratio vegetation index and soil adjusted vegetation index.

[0022] Preferably, the preset historical surface temperature data and the preset precipitation data are clipped and averaged to obtain meteorological data pixel values, including:

[0023] The projection mode of the preset high-resolution remote sensing data is changed to WGS1984 to obtain projection data;

[0024] Determine the coordinate range of the preset area according to the projection data to obtain a preset projection range;

[0025] The preset historical surface temperature data is clipped and averaged according to the preset projection range to obtain a temperature pixel value.

[0026] Preferably, the calculation formula of the biomass correlation coefficient is: R xy is the correlation coefficient about x and y; x is the pixel data; y is the preset biomass historical data; n is the number of sampling points; is the mean value of the pixel data; is the mean value of the preset biomass historical data; x i is the pixel data of the i-th sampling point; y i is the preset biomass historical data of the ith sampling point.

[0027] Preferably, the calculation formula of the root mean square error is: The calculation formula of the relative error is: RE is the calculation result of the root mean square error; RMSE is the calculation result of the relative error; x i is the actual biomass value of the i-th sampling point; n is the total number of the sampling points; Δx is the predicted biomass value.

[0028] Preferably, the tool for performing difference processing on the preset meteorological data is a geographic statistical analysis tool in ArcGIS software.

[0029] Preferably, extracting the vegetation index of the remote sensing image data to obtain the vegetation index pixel value includes:

[0030] The band type is determined, and band algebra-band operation is used to extract information from the remote sensing image data to obtain the vegetation index pixel value.

[0031] Preferably, extracting pixel data of the point to be analyzed according to the meteorological data pixel value and the vegetation index pixel value includes:

[0032] Counting the longitude and latitude coordinates of all the points to be analyzed to obtain a point file;

[0033] According to the point file, the spatial analyst tool in ArcGIS software is used to extract the pixel values ​​of the meteorological data and the pixel values ​​of the vegetation index to obtain the pixel data.

[0034] The present invention discloses the following technical effects:

[0035] The present invention provides a remote sensing estimation method for grassland biomass. By interpolating meteorological data, the problem of discontinuity of meteorological data is solved, and the meteorological data is enriched. By clipping and calibrating temperature, rainfall, remote sensing data, etc., the problem of single training data is solved, and the diversity and precision of training data are improved. Through the biomass correlation coefficient, the problem of low quality of training data is solved, and the training data is optimized. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] 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.

[0037] Figure 1 A schematic diagram of a remote sensing estimation process of grassland aboveground biomass provided by an embodiment of the present invention;

[0038] Figure 2 A logical schematic diagram of a random forest regression model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] 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.

[0040] The purpose of the present invention is to provide a remote sensing estimation method for grassland biomass, which enriches the amount of meteorological data and improves the robustness of the model by interpolating meteorological data; improves the diversity and precision of training data and increases the accuracy of the model by cutting and calibrating temperature, rainfall, remote sensing data, etc.; optimizes training data through biomass correlation coefficient, making model training more efficient and reliable.

[0041] 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.

[0042] Figure 1 A schematic diagram of a remote sensing estimation process of grassland aboveground biomass provided by an embodiment of the present invention is shown in FIG. Figure 1As shown, the present invention provides a remote sensing estimation method for grassland aboveground biomass, comprising:

[0043] Step 100: Collect historical meteorological data to be analyzed and remote sensing data to be analyzed of the grassland to be estimated;

[0044] Step 200: inputting the historical meteorological data to be analyzed and the remote sensing data to be analyzed into a preset biomass estimation model to obtain a target biomass estimation result;

[0045] The process of building the biomass estimation model includes:

[0046] Collect environmental information of the preset area and integrate it to obtain an environmental database; the environmental database includes: preset high-resolution remote sensing data, preset biomass historical data and preset meteorological data; the preset meteorological data includes: preset historical surface temperature data, preset precipitation data and DEM elevation data;

[0047] Perform interpolation processing on preset meteorological data to obtain continuous meteorological data;

[0048] The preset historical surface temperature data and the preset precipitation data are clipped and averaged to obtain the pixel value of the meteorological data;

[0049] Mosaic, radiometrically calibrate, atmospherically correct and crop the preset high-resolution remote sensing data to obtain remote sensing image data corresponding to the preset area;

[0050] Extract the vegetation index of remote sensing image data and obtain the vegetation index pixel value;

[0051] Extracting pixel data of the points to be analyzed according to the pixel values ​​of the meteorological data and the pixel values ​​of the vegetation index; the pixel data includes: the pixel values ​​of the meteorological data and the pixel values ​​of the vegetation index of all the points to be analyzed;

[0052] According to the pixel data, the biomass correlation is calculated to obtain the biomass correlation coefficient;

[0053] The biomass correlation coefficient is used to optimize the pixel data to obtain the characteristic data set;

[0054] The feature data set is randomly divided into a training subset and a test subset in a ratio of 2:1;

[0055] The pixel values ​​of meteorological data in the training subset are used as the input of the preset random forest regression model, and the pixel values ​​of vegetation index in the training subset are used as the output of the preset random forest regression model. The root mean square error and relative error are used as the loss function of the preset random forest regression model. The preset random forest regression model is trained to obtain a trained biomass estimation model.

[0056] Optionally, the vegetation index pixel value includes: normalized vegetation index, enhanced vegetation index, difference vegetation index, ratio vegetation index and soil adjusted vegetation index.

[0057] Furthermore, the preset historical surface temperature data and the preset precipitation data are clipped and averaged to obtain the meteorological data pixel values, including:

[0058] Change the projection mode of the preset high-resolution remote sensing data to WGS1984 to obtain the projection data;

[0059] Determine the coordinate range of the preset area according to the projection data to obtain a preset projection range;

[0060] The preset historical surface temperature data is clipped and averaged according to the preset projection range to obtain the temperature pixel value.

[0061] Specifically, the calculation formula of the biomass correlation coefficient is: R xy is the correlation coefficient between x and y; x is the pixel data; y is the preset biomass historical data; n is the number of sampling points; is the mean value of pixel data; is the mean of the preset biomass historical data; x i is the pixel data of the i-th sampling point; y i is the preset biomass historical data of the ith sampling point.

[0062] Specifically, the calculation formula of the root mean square error is: The calculation formula for relative error is: RE is the calculation result of root mean square error; RMSE is the calculation result of relative error; x i is the actual biomass value of the i-th sampling point; n is the total number of the sampling points; Δx is the predicted biomass value.

[0063] Optionally, the tool for performing difference processing on preset meteorological data is a geographic statistical analysis tool in ArcGIS software.

[0064] Specifically, extract the vegetation index of remote sensing image data to obtain the vegetation index pixel value, including:

[0065] Determine the band type, and use band algebra-band operation to extract information from remote sensing image data to obtain vegetation index pixel values.

[0066] Furthermore, pixel data of the point to be analyzed is extracted according to the pixel value of the meteorological data and the pixel value of the vegetation index, including:

[0067] Count the longitude and latitude coordinates of all points to be analyzed and obtain a point file;

[0068] According to the point file, the spatial analyst tool in ArcGIS software was used to extract the pixel values ​​of meteorological data and vegetation index pixel values ​​to obtain pixel data.

[0069] Specifically, obtain the historical remote sensing surface temperature data and meteorological data of the target area: obtain meteorological data through a meteorological station, including the average temperature and average precipitation data from July to August 2022. The above meteorological data require spatial interpolation processing. Because these data are discrete in space, they need to be made spatially continuous. In ArcGIS software, spatial interpolation method is performed on each type of meteorological data. Find Geostatistical Analyst Tools in Arctoolbox (toolbox) in ArcGIS software, and then find the interpolation analysis tool in the geographic statistical analysis tool, and then use the tool to import discrete meteorological data to the corresponding coordinates to generate spatially continuous meteorological data for the study area.

[0070] Furthermore, the temperature and precipitation data need to be processed as follows:

[0071] 1) Clipping, you need to use ArcGIS to clip it to the size of the study area. First change its projection mode to WGS1984, then select Extract by Rectangle in the Extraction Analysis in ArcGIS Spatial Analysis, and enter the coordinate range of the study area rectangle. Clip the surface temperature data of the study area size.

[0072] 2) Calculate the pixel value. Find the raster calculator in ArcGIS, input the above average surface temperature remote sensing data, enter the code in the formula, divide the pixel value by 100 and then subtract 273.15 to convert the pixel value to degrees Celsius (℃). After the above steps, the average surface temperature remote sensing data of the study area is obtained. The precipitation data is processed in the same way as the temperature data. In this way, the remote sensing data of precipitation can be obtained.

[0073] Optionally, obtain historical data on grassland biomass in the target area: obtain grassland biomass data for 2022 through a monitoring station, which includes longitude and latitude and grassland biomass.

[0074] Furthermore, obtain various vegetation index data of the target area: the obtained high-resolution remote sensing data is preprocessed, including mosaicking, radiation calibration, atmospheric correction, clipping and other operations, to obtain the remote sensing image data of the study area, and then use ENVI and ArcGIS software to extract various vegetation indexes from the processed remote sensing images. The following takes the extraction of normalized vegetation index (NDVI) as an example. Open ENVI 5.3 version software, import the preprocessed remote sensing image into ENVI5.3 software, click Bandmath tool (Band Algebra-Band Operation), then enter (float(b5)-b4) / (b5+b4), click formula, and then select b4 and b5 bands as infrared and near-infrared bands respectively. Click Choose a storage path and click OK to get the NDVI image map. Note that the NDVI value range should be between [-1,1]. The same method can be used to obtain image maps of other vegetation indices.

[0075] Specifically, based on the target grassland biomass survey area, the historical meteorological data, vegetation index and high-resolution remote sensing data pixel values ​​of the sampling points are obtained: Through the above steps, the spatially continuous meteorological data and ten-day vegetation index data of the study area are obtained. In ArcGIS, click File, Add Data, Add XY Data in turn, and find the Feature to Point tool. Import the longitude and latitude coordinates of all grassland biomass sampling points to generate a point file. Then click Toolbox, SpatialAnalyst, Extract Analysis in turn. Find the Value Extraction to Point tool. Import the point file and the spatially continuous ten-day raster data of various meteorological and surface temperature values ​​to be extracted, and then obtain various pixel data of the grassland biomass measurement points. It contains the pixel values ​​of the meteorological data in the study area, and the pixel values ​​of the vegetation index are used for subsequent calculations.

[0076] Furthermore, based on the aboveground biomass of grassland in the target area, the pixel characteristic data of historical meteorology, vegetation index and high-resolution remote sensing data are obtained: the correlation between the pixel values ​​of meteorological data in the study area and the various vegetation indices in the study area and the aboveground biomass of grassland in the study area is analyzed, and the corresponding characteristic data of meteorological data pixel values ​​and various vegetation indices are screened out. The correlation between the meteorological data pixel values ​​and the vegetation index pixel values ​​and the aboveground biomass of grassland at each point is calculated to obtain the correlation coefficient between the meteorological pixel data and the vegetation index pixel data and the aboveground biomass of grassland. In this way, characteristic data with high correlation is screened out. Some irrelevant characteristic data are eliminated to reduce the amount of useless data input, and the pixel characteristic data of historical meteorology, vegetation index and high-resolution remote sensing data are obtained. In this step, when calculating the correlation coefficient, the Pearson correlation coefficient can be used to describe the linear correlation between the meteorological and vegetation index pixel data and the aboveground biomass of grassland. The calculation formula is as follows:

[0077]

[0078] R in the formula xy Is the correlation coefficient between x and y. It is used to indicate the correlation between two coefficients. xy The value of R is between [-1, 1], xy If it is greater than 0, it means that there is a positive correlation between the two factors; if it is less than 0, it means that there is a negative correlation between the two factors; if it is equal to 0, it means that there is no correlation between the two factors. xy The closer the absolute value of R is to 1, the closer the connection is. xy The closer the absolute value of is to 0, the weaker the connection is. In the formula, n is the number of sampling points for grassland aboveground biomass, x is the pixel value of each meteorological data and the pixel value of vegetation index, and y is the aboveground biomass value of grassland at each point. and It is the corresponding pixel value of each meteorological data, the pixel value of vegetation index and the aboveground biomass value of grassland at each point.

[0079] Optionally, a training model is constructed based on the aboveground biomass of grassland in the target area and the pixel feature data of historical meteorological, vegetation index and high-resolution remote sensing data to obtain a grassland aboveground biomass estimation model; the pixel feature data of the corresponding coordinate points of each meteorological, vegetation index and high-resolution remote sensing data and the aboveground biomass data of grassland at each coordinate point are randomly divided into two groups according to the number of coordinate points, respectively referred to as training group feature data and test group feature data, and the number of training groups accounts for 2 / 3, and the number of test groups accounts for 1 / 3. The pixel meteorological, vegetation index and remote sensing feature data of the corresponding coordinates of the training group and the aboveground biomass of grassland at the corresponding coordinates are input into the constructed model to train the model. This embodiment selects to use a random forest regression model, which is constructed based on a decision tree regression model. Random forest randomly samples (with replacement) multiple times from the training data set, and each sampled data subset is used to train a decision tree. In this way, a forest containing multiple decision trees is formed. When constructing each tree, a feature subset is randomly selected to find the best split point instead of using all features. This increases the diversity of the model and reduces the risk of overfitting. For regression problems, random forests calculate the predicted mean of all trees to obtain the final prediction result. The characteristics of the random forest algorithm are strong resistance to overfitting, ability to process high-dimensional data, and high robustness to noise data and outliers. The random forest regression model is constructed using MATLAB software, with temperature, precipitation, and various vegetation indices as input feature data and grassland aboveground biomass as output value. The model diagram is shown in the figure below. Figure 2 .

[0080] Furthermore, the model is verified using the test group data. The test data is input into the model, and the obtained grassland biomass is calculated and verified with the grassland biomass of the test group data. In this embodiment, the root mean square error RMSE and relative error RE are used to verify the model. The calculation formulas of the root mean square error RMSE and the relative error RE are as follows:

[0081]

[0082] In the formula, x i is the grassland aboveground biomass value estimated by the model, x j is the actual grassland biomass value in the test group data. In the formula, n is the number of grassland biomass sampling points. These error values ​​obtained by calculation are compared with the preset threshold. When the error is less than the preset value, it is considered that the model for remote sensing estimation of grassland biomass meets the conditions. Large-scale remote sensing monitoring of grassland biomass can be carried out in this area. Otherwise, the model structure parameters will be further modified to make it meet the conditions.

[0083] Specifically, the aboveground biomass of grassland is estimated based on the historical meteorological data, vegetation index and high-resolution remote sensing data of the region based on the aboveground biomass of grassland: the aboveground biomass of grassland model based on the obtained high-resolution remote sensing can be used for the estimation of aboveground biomass of grassland in the region in the subsequent time, wherein various meteorological data can be replaced by remote sensing data, and the prediction accuracy of the model will be greatly enhanced after the replacement by remote sensing data. This method can greatly reduce the workload and number of investigators. It can achieve large-scale, fast, and cost-saving investigation, and has better results than traditional manual investigation.

[0084] The beneficial effects of the present invention are as follows:

[0085] The present invention enriches the amount of meteorological data and improves the robustness of the model by interpolating meteorological data; improves the diversity and precision of training data and increases the accuracy of the model by clipping and calibrating temperature, rainfall, remote sensing data, etc.; optimizes training data through biomass correlation coefficient, making model training more efficient and reliable.

[0086] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0087] 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 remote sensing estimation method for grassland aboveground biomass, characterized in that: include: Collect historical meteorological data and remote sensing data for the grassland to be estimated; Inputting the historical meteorological data to be analyzed and the remote sensing data to be analyzed into a preset biomass estimation model to obtain a target biomass estimation result; The process of constructing the biomass estimation model includes: Collect environmental information of a preset area and integrate it to obtain an environmental database; the environmental database includes: preset high-resolution remote sensing data, preset biomass historical data and preset meteorological data; the preset meteorological data includes: preset historical surface temperature data, preset precipitation data and DEM elevation data; Performing interpolation processing on the preset meteorological data to obtain continuous meteorological data; Clipping and averaging the preset historical surface temperature data and the preset precipitation data to obtain meteorological data pixel values; Mosaicing, radiation calibration, atmospheric correction and cropping are performed on the preset high-resolution remote sensing data to obtain remote sensing image data corresponding to the preset area; Extracting the vegetation index of the remote sensing image data to obtain a vegetation index pixel value; Extracting pixel data of the points to be analyzed according to the meteorological data pixel values ​​and the vegetation index pixel values; the pixel data includes: the meteorological data pixel values ​​and the vegetation index pixel values ​​of all the points to be analyzed; According to the pixel data, biomass correlation calculation is performed to obtain a biomass correlation coefficient; Optimizing the pixel data using the biomass correlation coefficient to obtain a feature data set; The feature data set is randomly divided into a training subset and a test subset in a ratio of 2:1; The meteorological data pixel values ​​in the training subset are used as the input of the preset random forest regression model, the vegetation index pixel values ​​in the training subset are used as the output of the preset random forest regression model, the root mean square error and relative error are used as the loss function of the preset random forest regression model, the preset random forest regression model is trained, and the trained biomass estimation model is obtained.

2. The remote sensing estimation method for grassland aboveground biomass according to claim 1, characterized in that: The vegetation index pixel values ​​include: normalized vegetation index, enhanced vegetation index, difference vegetation index, ratio vegetation index and soil adjusted vegetation index.

3. The method for remote sensing estimation of aboveground biomass of grassland according to claim 1, characterized in that: The preset historical surface temperature data and the preset precipitation data are clipped and averaged to obtain meteorological data pixel values, including: The projection mode of the preset high-resolution remote sensing data is changed to WGS1984 to obtain projection data; Determine the coordinate range of the preset area according to the projection data to obtain a preset projection range; The preset historical surface temperature data is clipped and averaged according to the preset projection range to obtain a temperature pixel value.

4. The method for remote sensing estimation of grassland aboveground biomass according to claim 1, characterized in that: The calculation formula of the biomass correlation coefficient is: R xy is the correlation coefficient about x and y; x is the pixel data; y is the preset biomass historical data; n is the number of sampling points; x is the mean of the pixel data; y is the mean of the preset biomass historical data; x i is the pixel data of the i-th sampling point; y i is the preset biomass historical data of the ith sampling point.

5. The method for remote sensing estimation of aboveground biomass of grassland according to claim 1, characterized in that: The calculation formula of the root mean square error is: The calculation formula of the relative error is: RE is the calculation result of the root mean square error; RMSE is the calculation result of the relative error; x i is the actual biomass value of the i-th sampling point; n is the total number of the sampling points; Δx is the predicted biomass value.

6. The method for remote sensing estimation of grassland aboveground biomass according to claim 1, characterized in that: The tool for performing difference processing on the preset meteorological data is the geographic statistical analysis tool in ArcGIS software.

7. The method for remote sensing estimation of aboveground biomass of grassland according to claim 1, characterized in that: Extracting the vegetation index of the remote sensing image data to obtain the vegetation index pixel value includes: The band type is determined, and band algebra-band operation is used to extract information from the remote sensing image data to obtain the vegetation index pixel value.

8. The method for remote sensing estimation of grassland aboveground biomass according to claim 1, characterized in that: Extracting pixel data of the point to be analyzed according to the meteorological data pixel value and the vegetation index pixel value includes: Counting the longitude and latitude coordinates of all the points to be analyzed to obtain a point file; According to the point file, the spatial analyst tool in ArcGIS software is used to extract the pixel values ​​of the meteorological data and the pixel values ​​of the vegetation index to obtain the pixel data.

Images