A collaborative inversion method and system for leaf area index and chlorophyll concentration based on multi-source data assimilation

Through the multi-source data assimilation and the dual vegetation index lookup table of the PROSAIL model, combined with the standard deviation weighting algorithm, the leaf area index and chlorophyll concentration are coordinated, which solves the inversion uncertainty and timing monitoring problems in the existing technology, and achieves high Precision monitoring of vegetation parameters.

CN118711070BActive Publication Date: 2024-11-22NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411181930.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-11-22
Estimated Expiration
2044-08-27

AI Technical Summary

Technical Problem

There is uncertainty in the existing leaf area index and chlorophyll concentration inversion methods. Single parameter inversion ignores the mutual interference of the two parameters on the remote sensing signal, and is limited by the time and spatial resolution of remote sensing satellites, making it difficult to achieve effective timing. monitor.

Method used

Using a multi-source data assimilation method, by downloading and fusing remote sensing images with high temporal resolution and high spatial resolution, the PROSAIL model is used to construct a dual vegetation index lookup table, combined with standard deviation weighting algorithm, and collaboratively invert leaf area index and chlorophyll concentration, reduce uncertainty and improve inversion accuracy.

Benefits of technology

The coordinated inversion of leaf area index and chlorophyll concentration is achieved, which reduces the uncertainty of individual inversion, improves the inversion accuracy, and can effectively monitor the timing changes of vegetation parameters.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118711070B_ABST
    Figure CN118711070B_ABST
Patent Text Reader

Abstract

The invention discloses a leaf area index and chlorophyll concentration collaborative inversion method and system based on multi-source data assimilation, including the following steps: downloading and fusing remote sensing images with high temporal resolution and high spatial resolution to generate long-time series multispectral canopy reflectance data; using the PROSAIL model to construct a dual vegetation index lookup table, matching it with the multispectral canopy reflectance data, and obtaining simulated data of the leaf area index; measuring at multiple measurement points in the sample plot to obtain measured data of the leaf area index; performing data assimilation on the simulated data of the leaf area index and the measured data through a standard deviation weighted algorithm to obtain assimilated data of the leaf area index; substituting the assimilated data into the dual vegetation index lookup table to realize collaborative inversion of chlorophyll concentration, and obtaining the spatial distribution and time variation characteristics of the chlorophyll concentration. The invention can perform collaborative inversion of the leaf area index and the chlorophyll concentration, and reduce the uncertainty during separate inversion.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of collaborative inversion of vegetation parameters, and in particular relates to a collaborative inversion method and system for leaf area index and chlorophyll concentration based on multi-source data assimilation. Background Art

[0002] Vegetation is the hub for the exchange of carbon, water and energy between the earth's surface and the atmosphere. In order to obtain accurate vegetation information in a timely manner, it is necessary to continuously monitor vegetation ecological parameters. Traditional field measurement methods are limited by time, space and human resources, so how to use remote sensing data for measurement more efficiently has become a research hotspot. The shortwave remote sensing signals obtained from vegetation contain information on vegetation structural parameters and leaf biochemical parameters. Leaf area index is an important structural parameter of the vegetation canopy, reflecting the coverage of vegetation and the density of leaves. Chlorophyll concentration is an important biochemical parameter of vegetation, reflecting the nutrient status and photosynthetic capacity of vegetation. Accurate estimation of leaf area index and chlorophyll concentration is of great significance for a comprehensive understanding of the status of regional and global ecosystems.

[0003] There are two main methods for inverting leaf area index and chlorophyll concentration: methods based on empirical models and methods based on physical models. The method based on empirical models is to invert by establishing an empirical relationship between vegetation index and leaf area index or chlorophyll concentration. This method is simple and easy to implement, but it is affected by factors such as region, crop type, soil background, etc., and is not universal and stable. The method based on physical models is to obtain leaf area index or chlorophyll concentration by establishing a canopy radiation transfer equation and inversely transforming the spectral reflectance. This method has a theoretical basis, but is affected by factors such as model parameters and observation angles, has high computational complexity, and low inversion accuracy. At the same time, both methods only consider the inversion of a single parameter, ignoring the mutual interference of the two vegetation parameters, leaf area index and chlorophyll concentration, on remote sensing signals, which also brings uncertainty to the inversion results. At the same time, affected by the temporal resolution and spatial resolution of remote sensing satellites, effective time series monitoring has not been achieved.

[0004] Therefore, it is urgent to solve the above problems. Summary of the invention

[0005] Purpose of the invention: The first purpose of the present invention is to provide a collaborative inversion method for leaf area index and chlorophyll concentration based on multi-source data assimilation, which assimilates remote sensing images and ground measured values, and can perform collaborative inversion of leaf area index and chlorophyll concentration, reducing the uncertainty during separate inversion.

[0006] The second object of the present invention is to provide a leaf area index and chlorophyll concentration collaborative inversion system based on multi-source data assimilation.

[0007] Technical solution: To achieve the above objectives, the present invention discloses a collaborative inversion method of leaf area index and chlorophyll concentration based on multi-source data assimilation, comprising the following steps:

[0008] (1) Download and fuse remote sensing images with high temporal resolution and high spatial resolution to generate long-term multispectral canopy reflectance data; high temporal resolution means that remote sensing images are obtained at least once a day, and high spatial resolution means that the spatial resolution is no greater than 10 meters;

[0009] (2) Use the PROSAIL model to construct a dual vegetation index lookup table and match it with the multispectral canopy reflectance data to obtain simulated data of leaf area index;

[0010] (3) Conduct measurements at multiple measurement points in the sample plot to obtain actual data on leaf area index;

[0011] (4) Assimilate the simulated data and measured data of leaf area index using the standard deviation weighted algorithm to obtain the assimilated data of leaf area index;

[0012] (5) Substitute the assimilated data into the dual vegetation index lookup table to achieve collaborative inversion of chlorophyll concentration and obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration.

[0013] Wherein, step (1) comprises the following steps:

[0014] (1.1) Select a satellite with the required temporal and spatial resolution as the remote sensing image source and download the remote sensing image;

[0015] Select two satellites as remote sensing image sources, one of which is a remote sensing image source with high temporal resolution, and the other is a remote sensing image source with high spatial resolution;

[0016] (1.2) Perform spatiotemporal fusion synthesis of remote sensing images obtained at different times and from different remote sensing image sources, that is, according to the time and longitude and latitude coordinates of the two satellites, the data at the same geographical location at the same time are matched in units of image pixels, and the data of the two satellites are fused based on the ratio of different times and different pixels, ultimately obtaining remote sensing image data with high temporal resolution and high spatial resolution;

[0017] (1.3) The fused remote sensing image is preprocessed, including radiation correction, atmospheric correction and geometric correction, to obtain accurate multispectral canopy reflectance data.

[0018] Preferably, step (2) comprises the following steps:

[0019] (2.1) The expression of PROSAIL model is:

[0020] ,

[0021] in is the canopy reflectance, represents the PROSAIL model operator, is the blade structural parameter, is the chlorophyll concentration, is the equivalent water thickness, is the dry matter content, is the leaf area index, is the average leaf inclination angle, is the hotspot parameter, is the soil dryness coefficient, is the soil hardness coefficient, To observe the zenith angle, is the solar zenith angle, is the relative azimuth between the sun and the observation, and the variable parameters are , , , , and , the other parameters are fixed values;

[0022] (2.2) According to the parameter range and step size of the variable parameters, different values ​​of different variable parameters are input into the PROSAIL model to obtain the vegetation canopy reflectance data, that is, to obtain the simulated reflectance curve;

[0023] (2.3) resampling the band of any satellite in the remote sensing image source selected in step (1.1) to obtain vegetation canopy reflectance data matching each band of the satellite;

[0024] The band response function is:

[0025] ,

[0026] in, For satellite The equivalent remote sensing reflectivity of the band, and is the band range of the band, To measure the hyperspectral remote sensing reflectance, for Spectral responsivity at wavelength;

[0027] (2.4) Two vegetation indices are selected, one of which is sensitive to chlorophyll concentration and the other is sensitive to leaf area index. The two selected vegetation indices are chlorophyll reflectance absorption conversion index and TCARI and Optimize Soil Adjusted Vegetation Index OSAVI ;

[0028] Chlorophyll reflectance absorption conversion index TCARI The corresponding vegetation index Soil Adjusted Vegetation Index OSAVI The corresponding vegetation index The values ​​are:

[0029] ,

[0030] Among them, B3, B4, B5, and B8A correspond to the reflectivity values ​​of the simulated reflectivity curve in the four bands of B3, B4, B5, and B8A, respectively, and represent the reflectivity values ​​at the central wavelengths of 560nm, 665nm, 705nm, and 865nm, respectively;

[0031] Obtain all vegetation indices corresponding to the simulated reflectance curve in step (2.2) Values ​​and vegetation indices Values ​​are arranged in ascending order, and a dual vegetation index lookup table is constructed, that is, an M*M two-dimensional table is constructed, and the horizontal rows of the table represent the vegetation index. Value, vertical row represents vegetation index The corresponding leaf area index Chlorophyll concentration Fill in the corresponding squares. Each grid point represents , and the mapping relationship between leaf area index and chlorophyll concentration; matching the multispectral canopy reflectance data obtained in step (1) with the dual vegetation index lookup table to obtain the simulated data of leaf area index; if there are multiple leaf area index values ​​in a grid point, the average is taken as the final leaf area index value; if there is no corresponding leaf area index value in a grid point, the LAI value in the nearest grid point is taken as the leaf area index value in the grid point.

[0032] Furthermore, step (3) includes the following steps:

[0033] (3.1) Select measurement points within the sample plot, including the four corners and the center point of the sample plot;

[0034] (3.2) Use a canopy analysis instrument to measure the leaf area index at each measurement point, record the measured values, and obtain the actual measured data of the leaf area index.

[0035] Further, step (4) includes the following steps:

[0036] (4.1) Calculate the standard deviation of simulated data and the standard deviation of the measured data , determine the weight coefficient of the simulated data based on the standard deviation and the weight coefficient of the measured data ;

[0037] ,

[0038] In the formula, is the current time value, is the model value, is the measured value, is the average value of the simulation value, is the average value of the measured values;

[0039] (4.2) Apply the standard deviation weighted method to fuse the simulated data and measured data to obtain the assimilated data of leaf area index;

[0040] ,

[0041] In the formula, is the assimilation value.

[0042] Preferably, step (5) comprises the following steps:

[0043] (5.1) Use the leaf area index assimilated data obtained in step (4) as input and substitute it into the dual vegetation index lookup table again;

[0044] Arrange the assimilated data of leaf area index from small to large and divide them into M parts on average. On the basis of the double vegetation index lookup table, introduce the third dimension, that is, divide it into M*M*M squares, each square represents , , right The mapping relationship;

[0045] (5.2) Using the mapping relationship between leaf area index and chlorophyll concentration in the dual vegetation index lookup table, collaborative inversion is performed to obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration.

[0046] Based on the same inventive concept, the present invention also provides a leaf area index and chlorophyll concentration collaborative inversion system based on multi-source data assimilation, comprising:

[0047] The canopy reflectance generation module is used to download and fuse remote sensing images with high temporal resolution and high spatial resolution to generate long-time multispectral canopy reflectance data; high temporal resolution means that remote sensing images are obtained at least once a day, and high spatial resolution means that the spatial resolution is no greater than 10 meters;

[0048] The dual vegetation index lookup table module is used to construct a dual vegetation index lookup table using the PROSAIL model, match it with the multispectral canopy reflectance data, and obtain the simulated data of the leaf area index;

[0049] A module for obtaining measured data is used to perform measurements at multiple measuring points in a sample plot to obtain measured data of leaf area index;

[0050] The leaf area index inversion module is used to assimilate the simulated data and measured data of the leaf area index through the standard deviation weighted algorithm to obtain the assimilated data of the leaf area index;

[0051] The chlorophyll concentration inversion module is used to substitute the assimilated data into the dual vegetation index lookup table to achieve the coordinated inversion of chlorophyll concentration and obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration.

[0052] Among them, in the canopy reflectance generation module, a satellite with the required temporal resolution and spatial resolution is selected as the remote sensing image source, and the remote sensing image is downloaded;

[0053] Select two satellites as remote sensing image sources, one of which is a remote sensing image source with high temporal resolution, and the other is a remote sensing image source with high spatial resolution;

[0054] The remote sensing images obtained at different times and from different remote sensing image sources are synthesized in time and space, that is, according to the time and longitude and latitude coordinates of the two satellites, the data at the same geographical location at the same time are matched in units of image pixels, and the data of the two satellites are fused by the ratio of different times and different pixels, and finally remote sensing image data with high temporal resolution and high spatial resolution are obtained;

[0055] The fused remote sensing images are preprocessed, including radiation correction, atmospheric correction and geometric correction, to obtain accurate multispectral canopy reflectance data.

[0056] Preferably, the expression of the PROSAIL model in the dual vegetation index lookup table module is:

[0057] ,

[0058] Among them is Canopy reflectance, represents the PROSAIL model operator, is the blade structural parameter, is the chlorophyll concentration, is the equivalent water thickness, is the dry matter content, is the leaf area index, is the average leaf inclination angle, is the hotspot parameter, is the soil dryness coefficient, is the soil hardness coefficient, To observe the zenith angle, is the solar zenith angle, is the relative azimuth between the sun and the observation, and the variable parameters are , , , , and , the other parameters are fixed values;

[0059] According to the parameter range and step size of the variable parameters, different values ​​of different variable parameters are input into the PROSAIL model to obtain the vegetation canopy reflectance data, that is, to obtain the simulated reflectance curve;

[0060] Resample the band of any satellite in the selected remote sensing image source to obtain the vegetation canopy reflectance data matching each band of the satellite;

[0061] The band response function is:

[0062] ,

[0063] in, For satellite The equivalent remote sensing reflectivity of the band, and is the band range of the band, To measure the hyperspectral remote sensing reflectance, for Spectral responsivity at wavelength;

[0064] Two vegetation indices are selected, one of which is sensitive to chlorophyll concentration and the other is sensitive to leaf area index. The two selected vegetation indices are chlorophyll reflectance absorption conversion index TCARI and Optimize Soil Adjusted Vegetation Index OSAVI ;

[0065] Chlorophyll reflectance absorption conversion index TCARI The corresponding vegetation index Soil Adjusted Vegetation Index OSAVI The corresponding vegetation index The values ​​are:

[0066] ,

[0067] Among them, B3, B4, B5, and B8A correspond to the reflectivity values ​​of the simulated reflectivity curve in the four bands of B3, B4, B5, and B8A, respectively, and represent the reflectivity values ​​at the central wavelengths of 560nm, 665nm, 705nm, and 865nm, respectively;

[0068] Obtain all vegetation indices corresponding to the simulated reflectivity curve Values ​​and vegetation indices Values ​​are arranged in ascending order, and a dual vegetation index lookup table is constructed, that is, an M*M two-dimensional table is constructed, and the horizontal rows of the table represent the vegetation index. Value, vertical row represents vegetation index The corresponding leaf area index Chlorophyll concentration Fill in the corresponding squares. Each grid point represents , The mapping relationship between the leaf area index and chlorophyll concentration is established; the obtained multispectral canopy reflectance data is matched with the dual vegetation index lookup table to obtain the simulated data of the leaf area index; if there are multiple leaf area index values ​​in a grid point, the average is taken as the final leaf area index value; if there is no corresponding leaf area index value in a grid point, the LAI value in the nearest grid point is taken as the leaf area index value in the grid point.

[0069] Furthermore, the obtained leaf area index assimilation data is used as input in the chlorophyll concentration inversion module and substituted into the dual vegetation index lookup table again;

[0070] Arrange the assimilated data of leaf area index from small to large and divide them into M parts on average. On the basis of the double vegetation index lookup table, introduce the third dimension, that is, divide it into M*M*M squares, each square represents , , right The mapping relationship;

[0071] The mapping relationship between leaf area index and chlorophyll concentration in the dual vegetation index lookup table was used for collaborative inversion to obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration.

[0072] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: first, the present invention determines two vegetation indices, which are respectively sensitive to leaf area index and chlorophyll concentration, avoiding the uncertainty of inversion under a single vegetation index; second, the present invention uses data assimilation to construct the relationship between remote sensing data observations and ground measurements, making the inversion of leaf area index more accurate; third, the present invention uses leaf area index to collaboratively invert chlorophyll concentration, taking into account the influence of leaf area index on chlorophyll concentration, and the inversion of chlorophyll concentration will also be more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is a schematic diagram of the process of the present invention;

[0074] Figure 2 It is a schematic diagram of the dual vegetation index lookup table in the present invention. DETAILED DESCRIPTION

[0075] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0076] Example 1: Figure 1 As shown, this embodiment discloses a collaborative inversion method of leaf area index and chlorophyll concentration based on multi-source data assimilation, comprising the following steps:

[0077] (1) Download and fuse remote sensing images with high temporal resolution and high spatial resolution to generate long-term multispectral canopy reflectance data; high temporal resolution means that remote sensing images are obtained at least once a day, and high spatial resolution means that the spatial resolution is no greater than 10 meters;

[0078] Step (1) includes the following steps:

[0079] (1.1) Select a remote sensing image source with the required temporal and spatial resolutions and download the remote sensing image;

[0080] The selected remote sensing image sources are MODIS and Sentinel-2. MODIS is a high temporal resolution remote sensing image source, which obtains remote sensing images once a day, but its resolution is only 1000 meters; Sentinel-2 is a high spatial resolution remote sensing image source with a resolution of 10 meters, but it obtains remote sensing images once every 5 days.

[0081] (1.2) Temporally and spatially fuse remote sensing images acquired at different times and from different remote sensing image sources to fill in data gaps and improve data continuity;

[0082] Spatiotemporal fusion: Based on the time and longitude and latitude coordinates of MODIS and Sentinel-2, the data of the same geographical location at the same time are matched in units of image pixels. The data of MODIS and Sentinel-2 are fused based on the ratio of different times and different pixels, and finally remote sensing image data with a spatial resolution of 10 meters and a temporal resolution of 1 day are obtained;

[0083] (1.3) Preprocess the fused remote sensing images, including radiation correction, atmospheric correction and geometric correction, to obtain accurate multispectral canopy reflectance data;

[0084] Radiation correction, geometric correction and atmospheric correction preprocessing are common remote sensing image processing methods. This embodiment uses ENVI5.6 software to complete the above processing;

[0085] (2) Use the PROSAIL model to construct a dual vegetation index lookup table and match it with the multispectral canopy reflectance data to obtain simulated data of leaf area index;

[0086] Step (2) includes the following steps:

[0087] (2.1) The PROSAIL model is a canopy reflectance model. The canopy reflectance data can be obtained by assigning values ​​to the model parameters. The expression of the PROSAIL model is:

[0088] ,

[0089] in is the canopy reflectance, represents the PROSAIL model operator, is the blade structural parameter, is the chlorophyll concentration, is the equivalent water thickness, is the dry matter content, is the leaf area index, is the average leaf inclination angle, is the hotspot parameter, is the soil dryness coefficient, is the soil hardness coefficient, To observe the zenith angle, is the solar zenith angle, is the relative azimuth between the sun and the observation, as shown in Table 1; the variable parameters are , , , , and , the remaining parameters are fixed values, which can be determined according to existing literature and empirical values, as shown in Table 2;

[0090] ,

[0091] ,

[0092] (2.2) According to the parameter range and step size of the variable parameters, different values ​​of different variable parameters are input into the PROSAIL model to obtain the vegetation canopy reflectance data of 400nm-2500nm with an interval of 1nm, that is, 37440 simulated reflectance curves are obtained;

[0093] (2.3) Resample the bands of any satellite in the remote sensing image source selected in step (1.1) to obtain vegetation canopy reflectance data matching each band of the satellite. The Sentinel-2 satellite bands include the bands shown in Table 3;

[0094] ,

[0095] The band response function is:

[0096] ,

[0097] in, For satellite The equivalent remote sensing reflectivity of the band, and is the band range of the band, To measure the hyperspectral remote sensing reflectance, for Spectral responsivity at wavelength; After the above steps, the simulated reflectivity curve is matched with the Sentinel-2 satellite band;

[0098] (2.4) Considering the mutual influence of leaf area index and chlorophyll concentration in the spectral range, two vegetation indices are selected, one of which is sensitive to chlorophyll concentration and the other is sensitive to leaf area index. The two selected vegetation indices are chlorophyll reflectance absorption conversion index and TCARI and Optimize Soil Adjusted Vegetation Index OSAVI ;

[0099] Calculate the chlorophyll reflectance absorption conversion index TCARI The green edge and red edge bands within the spectral range are required, which are often used to measure the absorption of chlorophyll to red wavelength light and are sensitive to chlorophyll content; while the optimized soil adjusted vegetation index OSAVI and leaf area index LAI Closely related to and not affected by soil background; Chlorophyll reflectance absorption conversion index TCARI The corresponding vegetation index Soil Adjusted Vegetation Index OSAVI The corresponding vegetation index The values ​​are:

[0100] ,

[0101] Among them, B3, B4, B5, and B8A correspond to the reflectivity values ​​of the simulated reflectivity curve in the four bands of B3, B4, B5, and B8A, respectively, and represent the reflectivity values ​​at the central wavelengths of 560nm, 665nm, 705nm, and 865nm, respectively;

[0102] Obtain all vegetation indices corresponding to the 37440 simulated reflectance curves in step (2.2) Values ​​and vegetation indices The values ​​are sorted in ascending order, and a dual vegetation index lookup table is constructed based on this, such as Figure 2 As shown, a 200*200 two-dimensional table is designed, and the horizontal rows of the table represent the vegetation index. Value, vertical row represents vegetation index Each vegetation index is divided into 200 parts, corresponding to a step range of 0.05, and its corresponding leaf area index is Chlorophyll concentration LCCFill in the corresponding squares. Each grid point represents , and leaf area index, and the mapping relationship between leaf area index and leaf area index, thereby establishing a dual vegetation index lookup table; matching the multispectral canopy reflectance data obtained in step (1) with the dual vegetation index lookup table to obtain the simulated data of leaf area index; if there are multiple leaf area index values ​​in a grid point, the average is taken as the final leaf area index value; if there is no corresponding leaf area index value in a grid point, the leaf area index in the nearest grid point is taken The value is taken as the leaf area index value in the grid point;

[0103] (3) Use canopy analysis instruments to measure at multiple points in the sample plot to obtain actual data on leaf area index;

[0104] Step (3) includes the following steps:

[0105] (3.1) Select measurement points within the sample plot, including the four corners and the center point of the sample plot, to ensure the representativeness and accuracy of the measurement results;

[0106] (3.2) Using a canopy analysis instrument to measure the leaf area index at each measuring point, recording the measured values, and obtaining the measured data of the leaf area index; the leaf area index value in this embodiment does not need to be measured every day, but the more days the measurement is performed, the higher the accuracy of the subsequent assimilation will be;

[0107] (4) Assimilate the simulated data and measured data using the standard deviation weighted method to obtain the assimilated data of leaf area index;

[0108] Step (4) includes the following steps:

[0109] (4.1) Calculate the standard deviation of simulated data and the standard deviation of the measured data , determine the weight coefficient of the simulated data based on the standard deviation and the weight coefficient of the measured data ;

[0110] ,

[0111] In the formula, is the current time value, is the model value, is the measured value, is the average value of the simulation value, is the average value of the measured values;

[0112] (4.2) Apply the standard deviation weighted method to fuse the simulated data and measured data to obtain the assimilated data of leaf area index;

[0113] ,

[0114] In the formula, is the assimilation value.

[0115] (5) Substitute the assimilated data into the dual vegetation index lookup table to achieve collaborative inversion of chlorophyll concentration and obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration;

[0116] The step (5) includes the following steps:

[0117] (5.1) Use the leaf area index assimilated data obtained in step (4) as input and substitute it into the dual vegetation index lookup table again;

[0118] The assimilated data of leaf area index are arranged in order from small to large and divided into 200 equal parts. The third dimension is introduced on the basis of the double vegetation index lookup table, that is, it is divided into 200*200*200 squares, each square represents , , right The mapping relationship avoids the problem of information overlap in the two-dimensional plane;

[0119] (5.2) Using the mapping relationship between leaf area index and chlorophyll concentration in the dual vegetation index lookup table, collaborative inversion is performed to obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration.

[0120] Example 2: Figure 1 As shown, this embodiment discloses a leaf area index and chlorophyll concentration collaborative inversion system based on multi-source data assimilation, including:

[0121] The canopy reflectance generation module is used to download and fuse remote sensing images with high temporal resolution and high spatial resolution to generate long-time multispectral canopy reflectance data; high temporal resolution means that remote sensing images are obtained at least once a day, and high spatial resolution means that the spatial resolution is no greater than 10 meters;

[0122] In the canopy reflectance generation module, select a remote sensing image source with the required temporal and spatial resolutions and download the remote sensing image;

[0123] The selected remote sensing image sources are MODIS and Sentinel-2. MODIS is a high temporal resolution remote sensing image source, which obtains remote sensing images once a day, but its resolution is only 1000 meters; Sentinel-2 is a high spatial resolution remote sensing image source with a resolution of 10 meters, but it obtains remote sensing images once every 5 days.

[0124] The remote sensing images acquired at different times and from different remote sensing image sources are synthesized in time and space to fill in data gaps and improve data continuity;

[0125] Spatiotemporal fusion: Based on the time and longitude and latitude coordinates of MODIS and Sentinel-2, the data of the same geographical location at the same time are matched in units of image pixels. The data of MODIS and Sentinel-2 are fused based on the ratio of different times and different pixels, and finally remote sensing image data with a spatial resolution of 10 meters and a temporal resolution of 1 day are obtained;

[0126] Preprocess the fused remote sensing images, including radiation correction, atmospheric correction and geometric correction, to obtain accurate multispectral canopy reflectance data;

[0127] Radiation correction, geometric correction and atmospheric correction preprocessing are common remote sensing image processing methods. This embodiment uses ENVI5.6 software to complete the above processing;

[0128] Dual vegetation index lookup table module, used to construct dual vegetation index lookup table using PROSAIL model and match it with multispectral canopy reflectance data to obtain simulated data of leaf area index;

[0129] The PROSAIL model in the dual vegetation index lookup table module is a canopy reflectance model. The canopy reflectance data can be obtained by assigning values ​​to the model parameters. The expression of the PROSAIL model is:

[0130] ,

[0131] in is the canopy reflectance, represents the PROSAIL model operator, is the blade structural parameter, is the chlorophyll concentration, is the equivalent water thickness, is the dry matter content, is the leaf area index, is the average leaf inclination angle, is the hotspot parameter, is the soil dryness coefficient, is the soil hardness coefficient, To observe the zenith angle, is the solar zenith angle, is the relative azimuth between the sun and the observation, as shown in Table 1; the variable parameters are , , , , and , the remaining parameters are fixed values, which can be determined according to existing literature and empirical values, as shown in Table 2;

[0132] According to the parameter range and step size of the variable parameters, different values ​​of different variable parameters were input into the PROSAIL model to obtain the vegetation canopy reflectance data with an interval of 1 nm from 400 nm to 2500 nm, that is, 37440 simulated reflectance curves were obtained;

[0133] According to the band of any satellite in the selected remote sensing image source, the vegetation canopy reflectance data matching each band of the satellite are obtained. The bands included in the Sentinel-2 satellite band are shown in Table 3;

[0134] The band response function is:

[0135] ,

[0136] in, For satellite The equivalent remote sensing reflectivity of the band, and is the band range of the band, To measure the hyperspectral remote sensing reflectance, for Spectral responsivity at wavelength; After the above steps, the simulated reflectivity curve is matched with the Sentinel-2 satellite band;

[0137] Considering the mutual influence of leaf area index and chlorophyll concentration in the spectral range, two vegetation indices are selected, one of which is sensitive to chlorophyll concentration and the other is sensitive to leaf area index. The two selected vegetation indices are chlorophyll reflectance absorption conversion index and chlorophyll absorption conversion index. TCARI and Optimize Soil Adjusted Vegetation Index OSAVI ;

[0138] Calculate the chlorophyll reflectance absorption conversion index TCARI The green edge and red edge bands within the spectral range are required, which are often used to measure the absorption of chlorophyll to red wavelength light and are sensitive to chlorophyll content; while the optimized soil adjusted vegetation index OSAVI and leaf area index LAI Closely related to and not affected by soil background; Chlorophyll reflectance absorption conversion index TCARI The corresponding vegetation index Soil Adjusted Vegetation Index OSAVI The corresponding vegetation index The values ​​are:

[0139] ,

[0140] Among them, B3, B4, B5, and B8A correspond to the reflectivity values ​​of the simulated reflectivity curve in the four bands of B3, B4, B5, and B8A, respectively, and represent the reflectivity values ​​at the central wavelengths of 560nm, 665nm, 705nm, and 865nm, respectively;

[0141] Obtain all vegetation indices corresponding to 37440 simulated reflectivity curves Values ​​and vegetation indices The values ​​are sorted in ascending order, and a dual vegetation index lookup table is constructed based on this, such as Figure 2 As shown, a 200*200 two-dimensional table is designed, and the horizontal rows of the table represent the vegetation index. Value, vertical row represents vegetation index Each vegetation index is divided into 200 parts, corresponding to a step range of 0.05, and its corresponding leaf area index is Chlorophyll concentration LCC Fill in the corresponding squares. Each grid point represents , and leaf area index, and the mapping relationship between leaf area index and leaf area index, thereby establishing a dual vegetation index lookup table; matching the multispectral canopy reflectance data obtained in step (1) with the dual vegetation index lookup table to obtain the simulated data of leaf area index; if there are multiple leaf area index values ​​in a grid point, the average is taken as the final leaf area index value; if there is no corresponding leaf area index value in a grid point, the leaf area index in the nearest grid point is taken The value is taken as the leaf area index value within the grid point.

[0142] The module for obtaining measured data is used to measure at multiple points in the sample plot using a canopy analysis instrument to obtain measured data of the leaf area index;

[0143] In the module for obtaining measured data, measurement points are selected within the sample plot, including the four corners and the center point of the sample plot, to ensure the representativeness and accuracy of the measurement results;

[0144] The leaf area index is measured at each measuring point using a canopy analysis instrument, and the measured values ​​are recorded to obtain the measured data of the leaf area index. The leaf area index value in this embodiment does not need to be measured every day, but the more days the measurement is performed, the higher the accuracy of the subsequent assimilation will be.

[0145] The leaf area index inversion module is used to assimilate the simulated data and the measured data by the standard deviation weighted method to obtain the assimilated data of the leaf area index;

[0146] Calculation of standard deviation of simulated data in the leaf area index inversion module and the standard deviation of the measured data , determine the weight coefficient of the simulated data based on the standard deviation and the weight coefficient of the measured data ;

[0147] ,

[0148] In the formula, is the current time value, is the model value, is the measured value, is the average value of the simulation value, is the average value of the measured values;

[0149] The standard deviation weighted method is used to fuse the simulated data and measured data to obtain the assimilated data of leaf area index;

[0150] ,

[0151] In the formula, is the assimilation value.

[0152] The chlorophyll concentration inversion module is used to substitute the assimilated data into the dual vegetation index lookup table to achieve the coordinated inversion of chlorophyll concentration and obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration;

[0153] The obtained leaf area index assimilation data is used as input in the chlorophyll concentration inversion module and substituted into the dual vegetation index lookup table again;

[0154] The assimilated data of leaf area index are arranged in order from small to large and divided into 200 equal parts. The third dimension is introduced on the basis of the double vegetation index lookup table, that is, it is divided into 200*200*200 squares, each square represents , , right The mapping relationship avoids the problem of information overlap in the two-dimensional plane;

[0155] The mapping relationship between leaf area index and chlorophyll concentration in the dual vegetation index lookup table was used for collaborative inversion to obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration.

Claims

1. A collaborative inversion method for leaf area index and chlorophyll concentration based on multi-source data assimilation, characterized in that: The following steps are involved: (1) Download and fuse remote sensing images with high temporal resolution and high spatial resolution to generate long-term multispectral canopy reflectance data; high temporal resolution means that remote sensing images are obtained at least once a day, and high spatial resolution means that the spatial resolution is no greater than 10 meters; (2) Use the PROSAIL model to construct a dual vegetation index lookup table and match it with the multispectral canopy reflectance data to obtain simulated data of leaf area index; The step (2) comprises the following steps: (2.1) The expression of PROSAIL model is: , in is the canopy reflectance, represents the PROSAIL model operator, is the blade structural parameter, is the chlorophyll concentration, is the equivalent water thickness, is the dry matter content, is the leaf area index, is the average leaf inclination angle, is the hotspot parameter, is the soil dryness coefficient, is the soil hardness coefficient, To observe the zenith angle, is the solar zenith angle, is the relative azimuth between the sun and the observation, and the variable parameters are , , , , and , the other parameters are fixed values; (2.2) According to the parameter range and step size of the variable parameters, different values ​​of different variable parameters are input into the PROSAIL model to obtain the vegetation canopy reflectance data, that is, to obtain the simulated reflectance curve; (2.3) resampling the band of any satellite in the remote sensing image source selected in step (1.1) to obtain vegetation canopy reflectance data matching each band of the satellite; The band response function is: , in, For satellite The equivalent remote sensing reflectivity of the band, and is the band range of the band, To measure the hyperspectral remote sensing reflectance, for Spectral responsivity at wavelength; (2.4) Two vegetation indices are selected, one of which is sensitive to chlorophyll concentration and the other is sensitive to leaf area index. The two selected vegetation indices are chlorophyll reflectance absorption conversion index and TCARI and Optimize Soil Adjusted Vegetation Index OSAVI ; Chlorophyll reflectance absorption conversion index TCARI The corresponding vegetation index Soil Adjusted Vegetation Index OSAVI The corresponding vegetation index The values ​​are: , Among them, B3, B4, B5, and B8A correspond to the reflectivity values ​​of the simulated reflectivity curve in the four bands of B3, B4, B5, and B8A, respectively, and represent the reflectivity values ​​at the central wavelengths of 560nm, 665nm, 705nm, and 865nm, respectively; Obtain all vegetation indices corresponding to the simulated reflectance curve in step (2.2) Values ​​and vegetation indices Values ​​are arranged in ascending order, and a dual vegetation index lookup table is constructed, that is, an M*M two-dimensional table is constructed, and the horizontal rows of the table represent the vegetation index. Value, vertical row represents vegetation index The corresponding leaf area index Chlorophyll concentration Fill in the corresponding squares. Each grid point represents , and the mapping relationship between leaf area index and chlorophyll concentration; matching the multispectral canopy reflectance data obtained in step (1) with the dual vegetation index lookup table to obtain the simulated data of leaf area index; if there are multiple leaf area index values ​​in a grid point, the average is taken as the final leaf area index value; if there is no corresponding leaf area index value in a grid point, the LAI value in the nearest grid point is taken as the leaf area index value in the grid point; (3) Conduct measurements at multiple measurement points in the sample plot to obtain actual data on leaf area index; (4) Assimilate the simulated data and measured data of leaf area index using the standard deviation weighted algorithm to obtain the assimilated data of leaf area index; (5) Substitute the assimilated data into the dual vegetation index lookup table to achieve collaborative inversion of chlorophyll concentration and obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration.

2. The collaborative inversion method of leaf area index and chlorophyll concentration based on multi-source data assimilation according to claim 1, characterized in that: The step (1) comprises the following steps: (1.1) Select a satellite with the required temporal and spatial resolution as the remote sensing image source and download the remote sensing image; Select two satellites as remote sensing image sources, one of which is a remote sensing image source with high temporal resolution, and the other is a remote sensing image source with high spatial resolution; (1.2) Perform spatiotemporal fusion synthesis of remote sensing images obtained at different times and from different remote sensing image sources, that is, according to the time and longitude and latitude coordinates of the two satellites, the data at the same geographical location at the same time are matched in units of image pixels, and the data of the two satellites are fused based on the ratio of different times and different pixels, ultimately obtaining remote sensing image data with high temporal resolution and high spatial resolution; (1.3) The fused remote sensing image is preprocessed, including radiation correction, atmospheric correction and geometric correction, to obtain accurate multispectral canopy reflectance data.

3. The collaborative inversion method of leaf area index and chlorophyll concentration based on multi-source data assimilation according to claim 1, characterized in that: The step (3) comprises the following steps: (3.1) Select measurement points within the sample plot, including the four corners and the center point of the sample plot; (3.2) Use a canopy analysis instrument to measure the leaf area index at each measurement point, record the measured values, and obtain the actual measured data of the leaf area index.

4. The collaborative inversion method of leaf area index and chlorophyll concentration based on multi-source data assimilation according to claim 1, characterized in that: The step (4) comprises the following steps: (4.1) Calculate the standard deviation of simulated data and the standard deviation of the measured data , determine the weight coefficient of the simulated data based on the standard deviation and the weight coefficient of the measured data ; , In the formula, is the current time value, is the model value, is the measured value, is the average value of the simulation value, is the average of the measured values; (4.2) Apply the standard deviation weighted method to fuse the simulated data and measured data to obtain the assimilated data of leaf area index; , In the formula, is the assimilation value.

5. The collaborative inversion method of leaf area index and chlorophyll concentration based on multi-source data assimilation according to claim 1, characterized in that: The step (5) includes the following steps: (5.1) Use the leaf area index assimilated data obtained in step (4) as input and substitute it into the dual vegetation index lookup table again; Arrange the assimilated data of leaf area index from small to large and divide them into M parts on average. On the basis of the double vegetation index lookup table, introduce the third dimension, that is, divide it into M*M*M squares, each square represents , , right The mapping relationship of (5.2) Using the mapping relationship between leaf area index and chlorophyll concentration in the dual vegetation index lookup table, collaborative inversion is performed to obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration.

6. A leaf area index and chlorophyll concentration collaborative inversion system based on multi-source data assimilation, characterized in that: include: The canopy reflectance generation module is used to download and fuse remote sensing images with high temporal resolution and high spatial resolution to generate long-term multispectral canopy reflectance data; High temporal resolution means that remote sensing images are acquired at least once a day, and high spatial resolution means that the spatial resolution is no greater than 10 meters; The dual vegetation index lookup table module is used to construct a dual vegetation index lookup table using the PROSAIL model and match it with the multispectral canopy reflectance data to obtain simulated data of the leaf area index; The expression of the PROSAIL model in the dual vegetation index lookup table module is: , Among them is Canopy reflectance, represents the PROSAIL model operator, is the blade structural parameter, is the chlorophyll concentration, is the equivalent water thickness, is the dry matter content, is the leaf area index, is the average leaf inclination angle, is the hotspot parameter, is the soil dryness coefficient, is the soil hardness coefficient, To observe the zenith angle, is the solar zenith angle, is the relative azimuth between the sun and the observation, and the variable parameters are , , , , and , the other parameters are fixed values; According to the parameter range and step size of the variable parameters, different values ​​of different variable parameters are input into the PROSAIL model to obtain the vegetation canopy reflectance data, that is, to obtain the simulated reflectance curve; Resample the band of any satellite in the selected remote sensing image source to obtain the vegetation canopy reflectance data matching each band of the satellite; The band response function is: , in, For satellite The equivalent remote sensing reflectivity of the band, and is the band range of the band, To measure the hyperspectral remote sensing reflectance, for Spectral responsivity at wavelength; Two vegetation indices are selected, one of which is sensitive to chlorophyll concentration and the other is sensitive to leaf area index. The two selected vegetation indices are chlorophyll reflectance absorption conversion index TCARI and Optimize Soil Adjusted Vegetation Index OSAVI ; Chlorophyll reflectance absorption conversion index TCARI The corresponding vegetation index Soil Adjusted Vegetation Index OSAVI The corresponding vegetation index The values ​​are: , Among them, B3, B4, B5, and B8A correspond to the reflectivity values ​​of the simulated reflectivity curve in the four bands of B3, B4, B5, and B8A, respectively, and represent the reflectivity values ​​at the central wavelengths of 560nm, 665nm, 705nm, and 865nm, respectively; Obtain all vegetation indices corresponding to the simulated reflectivity curve Values ​​and vegetation indices Values ​​are arranged in ascending order, and a dual vegetation index lookup table is constructed, that is, an M*M two-dimensional table is constructed, and the horizontal rows of the table represent the vegetation index. Value, vertical row represents vegetation index The corresponding leaf area index Chlorophyll concentration Fill in the corresponding squares. Each grid point represents , The mapping relationship between the leaf area index and chlorophyll concentration is established; the multispectral canopy reflectance data obtained is matched with the dual vegetation index lookup table to obtain the simulated data of the leaf area index; if there are multiple leaf area index values ​​in a grid point, the average is taken as the final leaf area index value; if there is no corresponding leaf area index value in a grid point, the LAI value in the nearest grid point is taken as the leaf area index value in the grid point; A module for obtaining measured data is used to perform measurements at multiple measuring points in a sample plot to obtain measured data of leaf area index; The leaf area index inversion module is used to assimilate the simulated data and measured data of the leaf area index through the standard deviation weighted algorithm to obtain the assimilated data of the leaf area index; The chlorophyll concentration inversion module is used to substitute the assimilated data into the dual vegetation index lookup table to achieve the coordinated inversion of chlorophyll concentration and obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration.

7. The leaf area index and chlorophyll concentration collaborative inversion system based on multi-source data assimilation according to claim 6, characterized in that: In the canopy reflectance generation module, a satellite with required temporal resolution and spatial resolution is selected as a remote sensing image source, and the remote sensing image is downloaded; Select two satellites as remote sensing image sources, one of which is a remote sensing image source with high temporal resolution, and the other is a remote sensing image source with high spatial resolution; The remote sensing images obtained at different times and from different remote sensing image sources are synthesized in time and space, that is, according to the time and longitude and latitude coordinates of the two satellites, the data at the same geographical location at the same time are matched in units of image pixels, and the data of the two satellites are fused by the ratio of different times and different pixels, and finally remote sensing image data with high temporal resolution and high spatial resolution are obtained; The fused remote sensing images are preprocessed, including radiation correction, atmospheric correction and geometric correction, to obtain accurate multispectral canopy reflectance data.

8. The leaf area index and chlorophyll concentration collaborative inversion system based on multi-source data assimilation according to claim 6, characterized in that: The chlorophyll concentration inversion module uses the obtained leaf area index assimilation data as input and substitutes it into the dual vegetation index lookup table again; Arrange the assimilated data of leaf area index from small to large and divide them into M parts on average. On the basis of the double vegetation index lookup table, introduce the third dimension, that is, divide it into M*M*M squares, each square represents , , right The mapping relationship of The mapping relationship between leaf area index and chlorophyll concentration in the dual vegetation index lookup table was used for collaborative inversion to obtain the spatial distribution and temporal variation characteristics of chlorophyll concentration.

Citation Information

Patent Citations

  • Mining area ecological environment evolution driving factor weight quantitative analysis method

    CN113128134A

  • Method for rapidly monitoring and determining drought condition of summer corn on basis of chlorophyll content by using unmanned aerial vehicle

    WO2024061159A1