A method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data

Abstract

The present invention discloses a method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data, comprising the following processing steps: preparing satellite remote sensing data; calculating the normalized vegetation index pixel by pixel; and, in step three, grouping the normalized vegetation index according to the accuracy of the leaf area index, fitting the probability distribution curve of each group of normalized vegetation index using a normal distribution function, wherein the normalized vegetation index at the point of maximum distribution probability is the normalized vegetation index corresponding to each leaf area index; and inverting the high-resolution leaf area index. The present invention determines the distribution patterns of vegetation index and leaf area index based on low-resolution satellite remote sensing data, constructing a simple, flexible, and easy-to-use high-resolution leaf area index inversion method, reducing the workload of ground measurements. Linear interpolation is used to calculate the leaf area index, reducing errors caused by imperfect fitting curves and improving estimation accuracy.

Description

Technical Field

[0001] The present invention relates to a data inversion method, in particular to a method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data. Background Art

[0002] The Leaf Area Index (LAI) characterizes the complexity of vegetation's vertical structure and is a key variable in describing biophysical changes and canopy structure in terrestrial ecosystems. It directly influences vegetation transpiration efficiency, photosynthesis, and energy balance. It is often used to drive meteorological, ecological, and terrestrial primary productivity models, as well as growth models. Its spatial resolution directly impacts the accuracy of model outputs. Existing satellite remote sensing LAI products, such as MODIS (500 m × 500 m), AVHRR LAI (8 km × 8 km), and GLOBMAP (8 km × 8 km), have fixed and coarse spatial resolutions. Due to the spatial heterogeneity of surface vegetation types, a coarse spatial resolution can result in multiple vegetation types within the pixel coverage area. Consequently, remote sensing LAI products fail to accurately reflect surface vegetation information, severely impacting the simulation precision and accuracy of physical process models driven by LAI. Therefore, there is an urgent need to improve the spatial resolution of LAI derived from satellite remote sensing to reduce errors caused by vegetation distribution heterogeneity.

[0003] Existing technologies primarily employ two methods for inverting leaf area index. One is the empirical model method, which assumes a strong correlation between the leaf area index and the vegetation index calculated based on surface reflectance. The leaf area index is estimated by establishing a functional relationship between the two. The leaf area index data used to fit the empirical model parameters can be obtained primarily through surface measurements and model simulations. The empirical model method simplifies the complex photon transmission process within the canopy and is simple, flexible, and efficient, achieving high accuracy within a small area. However, this method requires a large amount of data as a statistical basis. Another method for inverting the leaf area index is the physical model method, which simulates the radiation transmission process within the canopy based on the photon transmission theory of the vegetation canopy. Based on the outgoing radiation at the upper boundary of the vegetation, the optical and structural properties of the vegetation are inverted. Commonly used physical models include PROSAIL, which is based on physical optics and is applicable to a wider range of vegetation types and spatial ranges. However, some problems remain, such as the large number of model parameters, the difficulty in obtaining some parameters, and the long calculation time, making it difficult to apply to large-area inversions. Summary of the Invention

[0004] In order to solve the shortcomings of the above technologies, the present invention provides a method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: a method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data, comprising the following processing steps:

[0006] Step 1: Prepare satellite remote sensing data;

[0007] Step 2: Calculate the normalized vegetation index pixel by pixel based on the prepared satellite remote sensing data;

[0008] Step 3: Determine the normalized vegetation index corresponding to the leaf area index:

[0009] The NDVIs were grouped according to the accuracy of the LAI, and the probability distribution curve of each group of NDVIs was fitted using the normal distribution function. The NDVI at the point with the maximum distribution probability was the NDVI corresponding to each LAI.

[0010] Step 4: Invert high-resolution leaf area index.

[0011] Furthermore, the prepared data include but are not limited to surface reflectance products from low-resolution satellite remote sensing, leaf area index products from low-resolution satellite remote sensing, and surface reflectance product data from high-resolution remote sensing.

[0012] Furthermore, in step 2, the empirical relationship between the leaf area index and the normalized vegetation index is used to invert the high-resolution leaf area index based on the normalized vegetation index calculated based on the high-resolution satellite remote sensing surface reflectance data. The calculation method of the normalized vegetation index is shown in formula (1):

[0013]

[0014] Among them, R nir is the reflectance value in the near-infrared band, R red is the reflectance value of the red light band.

[0015] Furthermore, in step three, the grouping method is: since the range of the leaf area index is 0-6.9 with an accuracy of 0.1, and each leaf area index corresponds to multiple normalized vegetation indices, the leaf area index data are divided into 69 groups according to their accuracy.

[0016] Furthermore, the normalized vegetation index corresponding to each group of leaf area index obeys the normal distribution, and the normal distribution function is as follows:

[0017]

[0018] Where, μ is the mean of each group of normalized vegetation index, σ is the variance of each group of normalized vegetation index;

[0019] The normal distribution function of Formula 2 is used to fit the probability distribution function of each group of normalized vegetation indices.

[0020] Furthermore, in step 4, a scatter plot of the normalized difference vegetation index and the leaf area index is drawn to determine the location of the high-resolution normalized difference vegetation index, and the high-resolution leaf area index is calculated using a linear interpolation method.

[0021] Furthermore, a scatter plot is drawn based on the leaf area index of low-resolution satellite remote sensing and the normalized vegetation index calculated based on surface reflectivity. The coordinates of each point are (NDVI i ,LAI i ), assuming that the distance between two adjacent points is a straight line; high-resolution leaf area index LAI H The calculation formula is as follows:

[0022]

[0023] Among them, LAI H Represents high-resolution leaf area index, NDVI H Represents the high-resolution normalized difference vegetation index, which is calculated based on high-resolution satellite remote sensing surface reflectance; Represents the points (NDVI H , LAI H ) Normalized difference vegetation index between two adjacent points; Respectively represent the points (NDVI H , LAI H )Leaf area index of two adjacent points, i=1,2,...,68.

[0024] The present invention determines the distribution patterns of vegetation index and leaf area index based on low-resolution satellite remote sensing data, constructs a simple, flexible and easy-to-operate high-resolution leaf area index inversion method, reduces the workload of ground measurement, and improves inversion efficiency; on the other hand, linear interpolation is used to calculate the leaf area index, which reduces the error caused by imperfect fitting curves, improves estimation accuracy, and realizes the inversion of high-resolution leaf area index in different regions by this method. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is the overall technical flow chart of the present invention.

[0026] Figure 2 This is a pixel-by-pixel normalized vegetation index and leaf area index distribution diagram in an embodiment of the present invention.

[0027] Figure 3 This is a scatter plot of the normalized vegetation index and the leaf area index in an embodiment of the present invention.

[0028] Figure 4This is a graph showing the spatial distribution of leaf area index with a spatial resolution of 10m in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0030] In view of the fact that the spatial resolution of current leaf area index (LAI) satellite remote sensing products is generally low, the problem of mixed pixels is prone to occur, that is, multiple plant types exist in one pixel, which seriously affects the accuracy of model simulation and prediction. The present invention proposes a method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data. Based on the distribution law of low-resolution normalized difference vegetation index (NDVI) and leaf area index (LAI), the linear interpolation method is used, and the normalized difference vegetation index (NDVI) calculated from the surface reflectance of high-resolution satellite remote sensing is used as input to estimate the high-resolution leaf area index (LAI).

[0031] like Figure 1 As shown, the overall process of the present invention is as follows:

[0032] Step 1: Preparation of satellite remote sensing data:

[0033] The prepared data include low-resolution satellite remote sensing surface reflectance products, low-resolution satellite remote sensing leaf area index products, high-resolution remote sensing surface reflectance products, etc.

[0034] Step 2: Calculation of normalized vegetation index:

[0035] According to formula (1), the normalized vegetation index is calculated pixel by pixel based on low-resolution and high-resolution satellite remote sensing surface reflectance product data;

[0036] Step 3: Determine the normalized vegetation index corresponding to the leaf area index:

[0037] The NDVIs were grouped according to the accuracy of the LAI, and the probability distribution curve of each group of NDVIs was fitted using the normal distribution function. The NDVI at the point with the maximum distribution probability was the NDVI corresponding to each LAI.

[0038] Step 4: High-resolution leaf area index inversion:

[0039] A scatter plot of normalized difference vegetation index and leaf area index was drawn to determine the location of high-resolution normalized difference vegetation index, and high-resolution leaf area index was calculated using linear interpolation.

[0040] According to the above processing flow, the specific processing process of the method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data of the present invention is as follows:

[0041] Since the leaf area index (LAI) and the normalized difference vegetation index (NDVI) have a strong positive correlation, and the relationship between the two is fixed within a certain area, the high-resolution leaf area index can be inverted using the empirical relationship between the leaf area index and the normalized difference vegetation index, which is calculated based on the high-resolution satellite remote sensing surface reflectance data. The calculation method of the normalized difference vegetation index is shown in formula (1):

[0042]

[0043] Among them, R nir is the reflectance value in the near-infrared band, R red is the reflectance value of the red light band.

[0044] Since the range of leaf area index is 0-6.9 with an accuracy of 0.1, and each leaf area index corresponds to multiple normalized vegetation indices, the leaf area index data are divided into 69 groups according to their accuracy. The normalized vegetation index corresponding to each group of leaf area index obeys the normal distribution. The normal distribution function is as follows:

[0045]

[0046] Where, μ is the mean of each group of normalized vegetation index, σ is the variance of each group of normalized vegetation index;

[0047] The normal distribution function shown in Formula 2 is used to fit the probability distribution function of each group of normalized vegetation indices. The normalized vegetation index at the maximum value of the probability distribution function is the normalized vegetation index corresponding to each leaf area index.

[0048] There is a correlation between the leaf area index and the normalized vegetation index. If the leaf area index is estimated by fitting the empirical function of the two, the fitting curve may overestimate or underestimate, increasing the estimation error of the leaf area index. Therefore, the linear interpolation method is considered to estimate the high-resolution leaf area index. That is, a scatter plot is drawn based on the leaf area index of low-resolution satellite remote sensing and the normalized vegetation index calculated based on surface reflectivity. The coordinates of each point are (NDVI i ,LAI i ), assuming that there is a straight line between two adjacent points. High-resolution leaf area index (LAI) H ) is calculated as follows:

[0049]

[0050] Among them, LAI H Represents high-resolution leaf area index, NDVI H Represents the high-resolution normalized difference vegetation index, which is calculated based on high-resolution satellite remote sensing surface reflectance; Represents the points (NDVI H , LAI H ) Normalized difference vegetation index between two adjacent points; Respectively represent the points (NDVI H , LAI H )Leaf area index of two adjacent points, i=1, 2, ..., 68.

[0051] [Example]

[0052] The method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data disclosed in the present invention is further described below with reference to specific application examples.

[0053] In this example, Ordos City, Inner Mongolia is taken as the research object. MODIS satellite remote sensing surface reflectance data (MOD09A1) with a spatial resolution of 500m from January 2012 to December 2021 is used to calculate the normalized vegetation index of Ordos City pixel by pixel. The low-resolution satellite remote sensing leaf area index product is selected from the MODIS satellite remote sensing leaf area index product (MOD15A2H) with a spatial resolution of 500m. The normalized vegetation index is grouped according to the accuracy of the leaf area index, and the normal distribution function is used to fit the probability distribution of the normalized vegetation index, as shown in Figure 2. Figure 2 The NDVI with the largest distribution probability in each group is the NDVI corresponding to each leaf area index. A scatter plot of leaf area index and NDVI is drawn. Figure 3 The surface reflectance data of Sentinel-2A satellite remote sensing from January to December 2019 were used to calculate the normalized vegetation index with a spatial resolution of 10m for each pixel. The corresponding leaf area index was calculated using linear interpolation method to obtain the spatial distribution map of the leaf area index with a spatial resolution of 10m in Ordos City, as shown in the figure. Figure 4 shown.

[0054] It can be seen that compared with the existing technology, the present invention has the following technical advantages:

[0055] 1) Based on low-resolution satellite remote sensing data, the present invention reveals the distribution patterns of the normalized vegetation index and leaf area index in the region. Determining the distribution patterns of the vegetation index and leaf area index based on low-resolution satellite remote sensing data reduces the workload of ground observations and has obvious advantages in areas where observation data is scarce.

[0056] 2) The technology of the present invention is simple and easy to operate, which improves the efficiency of vegetation leaf area index inversion;

[0057] 3) The present invention adopts linear interpolation method to estimate the regional high-resolution leaf area index based on the normalized vegetation index calculated by high-resolution satellite remote sensing surface reflectance. This not only improves the estimation accuracy, but also the method can be applied to the inversion of high-resolution leaf area index in different regions.

[0058] The above embodiments are not limitations of the present invention, and the present invention is not limited to the above examples. Any changes, modifications, additions or substitutions made by technicians in this technical field within the scope of the technical solution of the present invention also fall within the scope of protection of the present invention.

Claims

1. A method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data, characterized by: The following processing steps are included: Step 1: Prepare satellite remote sensing data; Step 2: Calculate the normalized vegetation index pixel by pixel based on the prepared satellite remote sensing data; Step 3: Determine the normalized vegetation index corresponding to the leaf area index: The NDVIs were grouped according to the accuracy of the LAI, and the probability distribution curve of each group of NDVIs was fitted using the normal distribution function. The NDVI at the point with the maximum distribution probability was the NDVI corresponding to each LAI. Step 4: Invert high-resolution leaf area index; A scatter plot of normalized difference vegetation index and leaf area index was drawn to determine the location of high-resolution normalized difference vegetation index, and high-resolution leaf area index was calculated using linear interpolation.

2. The method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data according to claim 1, characterized in that: In step 1, the prepared data include but are not limited to low-resolution satellite remote sensing surface reflectance products, low-resolution satellite remote sensing leaf area index products, and high-resolution remote sensing surface reflectance product data.

3. The method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data according to claim 1, characterized in that: In the second step, the high-resolution leaf area index is inverted by using the empirical relationship between the leaf area index and the normalized vegetation index, which is calculated based on the surface reflectance data of high-resolution satellite remote sensing. The calculation method of the normalized vegetation index is shown in formula (1): （1）， in, is the reflectance value in the near-infrared band, is the reflectance value of the red light band.

4. The method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data according to claim 3, characterized in that: In step three, the grouping method is: since the range of the leaf area index is 0-6.9 and the precision is 0.1, and each leaf area index corresponds to multiple normalized vegetation indices, the leaf area index data are divided into 69 groups according to their precision.

5. The method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data according to claim 4, characterized in that: The normalized vegetation index corresponding to each group of leaf area index obeys the normal distribution, and the normal distribution function is as follows: （2）， in, is the mean of each group of normalized difference vegetation index, is the variance of each group of normalized vegetation indices; The normal distribution function of Formula 2 is used to fit the probability distribution function of each group of normalized vegetation indices.

6. The method for inverting vegetation leaf area index based on high-resolution satellite remote sensing data according to claim 5, characterized in that: A scatter plot is drawn based on the leaf area index of low-resolution satellite remote sensing and the normalized vegetation index calculated based on surface reflectivity. The coordinates of each point are ( NDVI i , LAI i ), assuming that there is a straight line between two adjacent points; high-resolution leaf area index LAI H The calculation formula is as follows: （3）， in, represents the high-resolution leaf area index, Represents the high-resolution normalized difference vegetation index, which is calculated based on high-resolution satellite remote sensing surface reflectance; 、 Respectively represent points ( , LAI H ) Normalized difference vegetation index between two adjacent points; 、 Respectively represent the points ( , LAI H ) Leaf area index of two adjacent points, i =1,2,...,68.

Images