Park scale vegetation carbon sink amount calculation method based on visible light vegetation index

Abstract

The invention discloses a park scale vegetation carbon sink calculation method based on a visible light vegetation index, and the method comprises the following steps: obtaining a visible light orthographic image of a target park, and calculating a blue-green ratio index GBRI of the pre-processed visible light orthographic image, extracting and converting raster data of a vegetation coverage area in an integral pixel of the visible light orthographic image, and inputting the raster data into a visible light vegetation index calculation model to obtain a visible light vegetation index in the range of the target park; collecting the specific information of the vegetation to obtain the carbon reserve of the vegetation in each sample plot and the carbon sequestration of the vegetation in each sample plot; establishing a park vegetation carbon reserve estimation model by using the carbon reserve of the vegetation in each sample plot and the corresponding different visible light indexes in each sample plot; establishing a park vegetation carbon sequestration amount estimation model by using the carbon reserves of the vegetation in each sample plot and the corresponding different visible light indexes in each sample plot; according to the method, efficient and accurate calculation of the park vegetation carbon sequestration is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of park vegetation carbon sink calculation, and more particularly to a park-scale vegetation carbon sink calculation method based on visible light vegetation index. Background Art

[0002] Domestic and foreign research on carbon sinks is mainly distributed in natural ecosystems such as forests, grasslands, wetlands, farmlands, and oceans, as well as artificial ecosystems, and focuses on exploring the process of carbon absorption and storage. Research on carbon sink calculations mainly focuses on the summary and analysis of calculation methods and quantitative calculations at different scales.

[0003] Regarding the research on carbon sink calculation methods, scholars He Ying and Zhao Lin reviewed the forest-scale carbon sink calculation methods, and compared the advantages and disadvantages of methods such as biomass method, stock method, biomass inventory method, eddy correlation method, relaxation eddy accumulation method, eddy covariance method, and box method; Guo Jing and other scholars conducted a comparative analysis of the advantages and disadvantages and application scope of the sample plot inventory method; Cao Jixin and other scholars introduced remote sensing technology and models into the forest carbon sink estimation method, reducing the consumption of manpower and material resources.

[0004] Regarding the quantitative calculation of carbon sinks, the current research mainly focuses on biomass, carbon storage and carbon fixation. Michael Swoish and other scholars calculated the biomass of farmland crops based on satellite images and vegetation indices, and compared the accuracy of different satellite images and vegetation indices in calculating farmland biomass; Tran Ngoc KhanhNi and other scholars used satellite remote sensing images combined with atmospheric correction, supervised classification and other technologies to calculate the biomass and canopy coverage of submerged plants; Fang Jingyun and other scholars estimated the national terrestrial vegetation carbon storage based on resource inventory statistical data, meteorological data and satellite remote sensing data; Zhou Yurong and other scholars estimated the carbon storage and carbon balance flux of major forest ecosystems in my country based on forest statistical data, and obtained the average carbon density value of my country's forest ecosystems and their components; Gao Yan and other scholars inverted the aboveground vegetation biomass of Shengjin Lake grassland based on drones and satellite remote sensing images, constructed a biomass regression model and completed the biomass calculation. Liu Yanhui and other scholars constructed the green index EXG and color index CIVE based on drone images, estimated the grassland vegetation coverage through vegetation index, and calculated the grassland biomass through regression equations; Zhang Zhengjian and other scholars constructed a regression model of biomass and multiple visible light vegetation indices based on drone visible light images and ground measured samples, estimated the biomass of each model and analyzed the differences in calculation accuracy of each model. Scholar Wu Jie measured the net photosynthetic rate of leaves of various vegetation in Shenzhen Special Economic Zone, and used the assimilation method to complete the calculation of vegetation carbon fixation at the urban scale. Shi Hongwen and other scholars measured the photosynthetic rate of dominant garden tree species in Wuhan City, and concluded that the carbon fixation and oxygen release capacity of trees and deciduous vegetation is better than that of shrubs and evergreen vegetation. Scholars Sun Sujing and Geng Shipeng used the assimilation method and standard forest land data to estimate the carbon fixation of green space at the park scale. Based on the former's research, Zhang Chenyue divided the park's greening structure into three types: tree, shrub and grass type, shrub and grass type, and lawn type, and used the same method to complete the estimation of the carbon fixation of the park's green space. Gong Wen and other scholars used the plant carbon sink and consumption balance method, plant photosynthesis equation method, national forest carbon storage estimation method and other methods to calculate carbon sinks at the park scale from multiple angles.

[0005] In summary, current carbon sink research focuses on macro or meso scales such as countries, provinces, and cities, and there are fewer studies on small scales such as parks. Compared with natural forest vegetation, artificial vegetation in parks is not concentrated enough in species diversity, and previous carbon sink calculation methods have failed to effectively solve the contradiction between calculation accuracy and human and material resource consumption. Remote sensing includes satellite remote sensing and drone low-altitude remote sensing. Remote sensing estimation methods can reduce the consumption of manpower and material resources, but satellite remote sensing data has lower accuracy and satellite coverage than drone remote sensing data, and cannot effectively solve the problem of carbon sink estimation at the park scale.

[0006] Therefore, in terms of carbon sink calculation methods in the existing technology, current carbon sink calculations are mostly targeted at macro or meso scales such as countries, provinces, and cities, and there is a lack of efficient and accurate calculation methods for vegetation carbon sinks at small scales such as parks. Compared with artificial forests, park vegetation has the problem of uneven distribution of vegetation diversity, and compared with natural forests, there is a lack of regular forest census data. Applying previous carbon sink calculation methods to park-scale carbon sink calculation research is time-consuming and labor-intensive, and satellite remote sensing data has lower accuracy and incomplete satellite coverage compared to UAV remote sensing data, and cannot effectively solve the problem of carbon sink measurement at the park scale. Summary of the invention

[0007] The purpose of the present invention is to provide a method for calculating the carbon sink of vegetation at a park scale based on the visible light vegetation index. Based on the collected visible light index of vegetation, a park scale vegetation visible light index-carbon sink calculation model is constructed, thereby realizing efficient and accurate calculation of the carbon sink of vegetation at the park scale, thereby solving the deficiencies existing in the above-mentioned prior art.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A method for calculating the carbon sink of vegetation at the park scale using a visible light vegetation index comprises the following steps:

[0010] S1. Obtain a visible light orthographic image of a target park, perform preprocessing, calculate the blue-green ratio index GBRI of the preprocessed visible light orthographic image, obtain different visible light information data, and input the data into a visible light vegetation index calculation model to obtain a visible light vegetation index within the range of the target park;

[0011] S2. Select several sample plots of equal area from the target park, and collect specific information of vegetation in the sample plots of equal area;

[0012] S3, inputting the collected specific information of vegetation into the vegetation biomass calculation model to obtain the biomass of vegetation in each sample plot;

[0013] The biomass of vegetation in each plot is input into the vegetation biomass-carbon storage calculation model to obtain the carbon storage of vegetation in each plot;

[0014] The growth parameters and leaf area index of the vegetation in each plot were input into the vegetation biomass-carbon sequestration calculation model to obtain the carbon sequestration of the vegetation in each plot;

[0015] S4, dividing a number of sample plots of equal area into training samples and verification samples, wherein the training samples and verification samples respectively include: the carbon storage of vegetation in each sample plot, the carbon fixation of vegetation in each sample plot, and different visible light indexes corresponding to each sample plot;

[0016] Using the carbon storage of vegetation in each plot and the corresponding different visible light indexes in each plot, a vegetation carbon storage estimation model for the park was established;

[0017] Using the carbon storage of vegetation in each plot and the corresponding different visible light indexes in each plot, a model for estimating the carbon sequestration of vegetation in the park was established;

[0018] Use the park vegetation carbon storage estimation model to estimate the carbon storage in the target park;

[0019] The carbon sequestration amount in the target park is estimated using the park vegetation carbon sequestration estimation model.

[0020] According to the specific embodiment provided by the present invention, the present invention discloses the following technical effects: the present invention obtains the visible light orthophoto image of the target park range through the unmanned aerial vehicle, and calculates the visible light vegetation index of the vegetation in the target park range; collects the vegetation information of each sample plot and calculates the carbon sink of the sample plot vegetation; constructs the visible light index-carbon sink calculation model of the park vegetation based on correlation analysis and regression analysis (including two sub-models for carbon storage and carbon fixation calculation), and then uses the vegetation visible light index-carbon sink calculation model to perform correlation analysis on the carbon fixation, carbon storage and different visible light indexes in each sample plot, and obtains the park vegetation carbon storage estimation model and the park vegetation carbon fixation estimation model. Through the park vegetation carbon storage estimation model and the park vegetation carbon fixation estimation model in the present invention, it is possible to realize the carbon sink measurement of the park green space while ensuring the calculation accuracy and efficiency, reduce the workload of manually collecting vegetation information to a certain extent, and provide data support for the evaluation of the park vegetation carbon sink and the carbon balance calculation of the park. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] 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 or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0022] The following is a further description of the method for calculating the park-scale vegetation carbon sink based on the visible light vegetation index in the present invention in conjunction with the accompanying drawings;

[0023] Figure 1 It is a schematic diagram of the overall process of the method for calculating the carbon sink of vegetation at the park scale based on the visible light vegetation index provided by the present invention;

[0024] Figure 2 It is a visible light orthographic image in the method for calculating the carbon sink of vegetation at the park scale based on the visible light vegetation index provided by the present invention;

[0025] Figure 3 It is the blue-green ratio index GBRI in the park-scale vegetation carbon sink calculation method based on the visible light vegetation index provided by the present invention;

[0026] Figure 4 It is the vegetation pixel reclassification in the park-scale vegetation carbon sink calculation method based on the visible light vegetation index provided by the present invention;

[0027] Figure 5 It is the surface element shape of the vegetation coverage area in the park-scale vegetation carbon sink calculation method based on the visible light vegetation index provided by the present invention;

[0028] Figure 6 It is a general diagram of the survey plot arrangement in the park-scale vegetation carbon sink calculation method based on the visible light vegetation index provided by the present invention;

[0029] Figure 7 It is a detailed diagram of sample plot arrangement of some areas in the method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index provided by the present invention;

[0030] Figure 8 It is the vegetation carbon storage of each sample plot in the survey sample plot in the park-scale vegetation carbon sink calculation method based on the visible light vegetation index provided by the present invention;

[0031] Fig. 9 It is the carbon sequestration amount of vegetation in each sample plot of the survey sample plot in the park-scale vegetation carbon sink calculation method based on the visible light vegetation index provided by the present invention;

[0032] Fig.10 The present invention provides a method for calculating the carbon sink of vegetation at the park scale based on the visible light vegetation index, which includes a correlation analysis between the carbon storage of vegetation in each sample plot in the sample plot area and the visible light vegetation index. Fig.10 Among them, (a) is the correlation analysis between the carbon storage of vegetation in each sample plot and the overgreening index EXG; (b) is the correlation analysis between the carbon storage of vegetation in each sample plot and the visible light band difference vegetation index; (c) is the correlation analysis between the carbon storage of vegetation in each sample plot and the super green and super red difference index EXGR; (d) is the correlation analysis between the carbon storage of vegetation in each sample plot and the red, green and blue vegetation index RGBVI; (e) is the correlation analysis between the carbon storage of vegetation in each sample plot and the super green, red and blue difference index EGRBDI; (f) is the correlation analysis between the carbon storage of vegetation in each sample plot and the improved red-green vegetation index MGRVI.

[0033] Fig.11 It is a regression model of vegetation carbon storage and various visible light vegetation indices in the park-scale vegetation carbon sink calculation method of the visible light vegetation index provided by the present invention; Fig.11Among them, (a) is the regression model of vegetation carbon storage and overgreen index EXG; (b) is the regression model of vegetation carbon storage and visible light band difference vegetation index VDVI; (c) is the regression model of vegetation carbon storage and super green and super red difference index EXGR;

[0034] Fig.12 The method for calculating the carbon sink of vegetation at the park scale using the visible light vegetation index provided by the present invention is a correlation analysis between the carbon sequestration of vegetation in each sample plot in the sample plot area and the visible light vegetation index; Fig.12 In the figure, (a) is the correlation analysis between the carbon fixation of vegetation in each sample plot and the overgreening index EXG; (b) is the correlation analysis between the carbon fixation of vegetation in each sample plot and the super-green and super-red difference index EXGR; (c) is the correlation analysis between the carbon fixation of vegetation in each sample plot and the green-red ratio index GRRI; (d) is the correlation analysis between the carbon fixation of vegetation in each sample plot and the blue-green ratio index GBRI; (e) is the correlation analysis between the carbon fixation of vegetation in each sample plot and the vegetation color index CIVE; (f) is the correlation analysis between the carbon fixation of vegetation in each sample plot and the visible light band difference vegetation index VDVI; (g) is the correlation analysis between the carbon fixation of vegetation in each sample plot and the visible light band difference vegetation index VDVI. Correlation analysis between carbon fixation and red-green-blue vegetation index RGBVI; (h) correlation analysis between carbon fixation of vegetation in each sample plot and improved red-green vegetation index MGRVI; (i) correlation analysis between carbon fixation of vegetation in each sample plot and super green red-blue difference index EGRBDI; (j) correlation analysis between carbon fixation of vegetation in each sample plot and normalized red-green difference index NGRDI; (k) correlation analysis between carbon fixation of vegetation in each sample plot and normalized blue-green difference index NGBDI; (l) correlation analysis between carbon fixation of vegetation in each sample plot and enhanced normalized green-blue difference index E-NGBD;

[0035] Fig.13 It is a regression model of vegetation carbon fixation and various visible light vegetation indices in the park-scale vegetation carbon sink calculation method of the visible light vegetation index provided by the present invention; Fig.13 Among them, (a) is the regression model of vegetation carbon fixation and overgreen index EXG; (b) is the regression model of vegetation carbon fixation and super green and super red difference index EXGR; (c) is the regression model of vegetation carbon fixation and visible light band difference vegetation index VDVI;

[0036] Fig.14 It is the calculation result of the park-scale super-green and super-red difference index EXGR in the park-scale vegetation carbon sink calculation method of the visible light vegetation index provided by the present invention;

[0037] Fig.15 It is the calculation result of the park-scale super-green and super-red differential index EXGR resampling in the park-scale vegetation carbon sink calculation method of the visible light vegetation index provided by the present invention. DETAILED DESCRIPTION

[0038] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0039] In order to better understand the purpose, structure and function of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings.

[0040] like Figure 1 As shown, the present invention provides a method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index, comprising the following steps:

[0041] 1. Construct a visible light index-carbon sink calculation model, which includes: a visible light index-carbon storage calculation sub-model of vegetation in the target park and a visible light index-carbon fixation calculation sub-model of vegetation in the target park.

[0042] S1. Obtain a visible light orthographic image of the target park, perform preprocessing, and calculate the blue-green ratio index GBRI of the preprocessed visible light orthographic image, such as Figure 3 As shown;

[0043] The overall pixels of the visible light orthophoto image are classified based on the blue-green ratio index GBRI, such as Figure 4 As shown in the figure, the raster data of the vegetation coverage area in the whole pixel is extracted, and the raster data of the vegetation coverage area is converted to obtain the surface element shape of the vegetation coverage area, such as Figure 5 As shown, it is used as a mask to obtain different visible light information data in the vegetation coverage area;

[0044] Based on different visible light information data, the data is input into the visible light vegetation index calculation model to obtain the visible light vegetation index within the scope of the target park;

[0045] It should be noted that the visible light vegetation index involved in the present invention is an index reflecting the growth status of vegetation formed by combining vegetation information obtained by remote sensing technology based on the spectral characteristics of vegetation in the visible light region. The visible light vegetation index is widely used in global and regional land cover, vegetation classification and environmental change, primary productivity analysis, crop and forage yield estimation, drought monitoring and other aspects. Based on the low-altitude remote sensing technology of unmanned aerial vehicles, the vegetation indices that can be used for visible light image vegetation extraction include: over-green index EXG, super-green and super-red difference index EXGR, green-red ratio index GRRI, blue-green ratio index GBRI, vegetation color index CIVE, visible light band difference vegetation index VDVI, red-green-blue vegetation index RGBVI, improved red-green vegetation index MGRVI, super-green-red-blue difference index EGRBDI, normalized red-green difference index NGRDI, normalized blue-green difference index NGBDI, enhanced normalized green-blue difference index E-NGBD and other 12 categories. The present invention summarizes the above 12 types of indicators, sorts out the calculation models of each visible light vegetation index and their value ranges, as shown in Table 1, and calculates the 12 visible light vegetation indices of the regional range according to formulas (1) to (12) in Table 1.

[0046] Table 1 Calculation models and value ranges of various visible light vegetation indices

[0047]

[0048] Note: R is the red channel value of the image; G is the green channel value of the image; B is the blue channel value of the image; r is the normalized result of the red channel; g is the normalized result of the green channel;

[0049] b is the normalized result of the blue light channel, where:

[0050] r+g+b=1, the range of r, g, b is [0,1].

[0051] S2. Randomly select a number of sample plots of equal area from the target park, and collect vegetation-specific information of the sample plots of equal area;

[0052] It should be noted that: according to the size of the target park, n sample plots (80) of a certain size (10m×10m) were randomly selected, such as Figure 6 and Figure 7 As shown, Figure 6 The red dots in the middle are schematic diagrams of the locations of some of the sample plots; Figure 7 There are 20 sample plots in the Figure 7 Middle red dot position.

[0053] S3, inputting the collected vegetation specific information into the vegetation biomass calculation model to obtain the biomass of the vegetation in each sample plot;

[0054] The biomass of vegetation in each sample plot is input into the vegetation biomass-carbon storage calculation model to obtain the carbon storage of vegetation within each sample plot;

[0055] The vegetation growth parameters and leaf area index of each plot obtained by field sampling measurements were input into the vegetation biomass-carbon sequestration calculation model to obtain the carbon sequestration of vegetation within the plot.

[0056] S4. Constructing a visible light index-carbon sink calculation model, wherein the visible light index-carbon sink calculation model includes a visible light index-carbon storage calculation sub-model of vegetation in the target park and a visible light index-carbon fixation calculation sub-model of vegetation in the target park;

[0057] The visible light index-carbon storage calculation sub-model of vegetation in the target park is constructed by performing correlation analysis and regression analysis using the carbon storage of vegetation in each sample plot and the corresponding different visible light indices in each sample plot;

[0058] The visible light index-carbon fixation calculation submodel of the vegetation in the target park is constructed by performing correlation analysis and regression analysis on the carbon fixation of the vegetation in each sample plot and the corresponding different visible light indices in each sample plot.

[0059] The visible light index-carbon fixation calculation sub-model is specifically as follows: resampling the visible light vegetation index within each sample plot; based on the resampling calculation results, using scatter plots and Pearson correlation analysis methods to perform correlation analysis on the vegetation carbon storage in each sample plot and different visible light vegetation indices, determining the top three types of visible light vegetation indices with the strongest correlation, and performing regression analysis. The correlation results are as follows: Fig.10 shown.

[0060] In actual applications, statistical software (such as SPSS) will automatically calculate the P value and give the "sig" value in the output results. You do not need to calculate the P value manually, but you need to understand how the P value is calculated based on the selected test and data, and how to judge the significance of the results based on the P value and the preset significance level.

[0061] In the step S1, the visible light orthographic image is preprocessed, specifically: strong interference factors in the orthographic image are eliminated by cropping;

[0062] The strong interference factors include water surface, aquatic plants and track and field pavement.

[0063] In the step S1, the visible light orthophoto image is classified based on the blue-green ratio index GBRI, and the raster data of the vegetation coverage area is extracted. Specifically, the overall pixels of the image are reclassified into vegetation pixels and non-vegetation image sources based on the GBRI index to extract the raster data of the vegetation coverage area.

[0064] In step S2, a number of sample plots of equal area are randomly selected from the target park, specifically: 80 sample plots of 10m×10m are randomly selected according to the size of the target park;

[0065] The specific information of vegetation in the sample plots is collected, specifically: tree information of the sample plots is collected, including the type, diameter at breast height, crown width and plant height of the tree, as shown in Table 2. Table 2 is Figure 7 Information on trees in the area shown in;

[0066] The shrub and herb information of the sample plots, including the base diameter, crown width, plant height and area of ​​shrubs and herbs, are shown in Table 3. Figure 7 Information on shrubs and herbs for the area shown.

[0067] Table 2 Information on tree parameters in each plot of YFBZ-1

[0068]

[0069]

[0070] Table 3 Information on shrub and herbaceous parameters in each plot of YFBZ-1

[0071]

[0072] The vegetation biomass calculation model in step S3 is shown in formula (13), and the biomass of vegetation in each plot is obtained, specifically:

[0073] B Q,i =B A,i +B B,i +B C,i (13)

[0074] Where B Q,i ——vegetation biomass of the ith survey plot; B A,i ——the biomass of trees in the ith survey plot; B B,i ——the biomass of shrubs in the i-th survey plot; B C,i ——The biomass of herbs in the ith survey plot.

[0075] It should be noted that the calculation methods for biomass of trees, shrubs and herbs are shown in Table 4.

[0076] Table 4 Allometric growth equations of biomass of each vegetation type

[0077]

[0078]

[0079] The vegetation biomass-carbon storage calculation model in step S3 is specifically:

[0080] CS Q,i =B Q,i ×α (25)

[0081] In the above formula, CS Q,i is the vegetation carbon storage of the i-th survey plot; B Q,i ——Vegetation biomass of the ith survey plot; α is the carbon content of vegetation.

[0082] It should be noted that: after the above calculation, the calculation results of vegetation carbon storage in each sample plot of the four survey plots are obtained, such as Figure 8 shown.

[0083] The vegetation biomass-carbon fixation calculation model in step S3 is specifically:

[0084] C s,q =C sa,q +C sb,q +C sc,q ; (26)

[0085] In the above formula, C s,q is the annual carbon fixation of vegetation in the survey plot; C sa,q is the annual carbon fixation of trees in the survey plot; C sb,q is the annual carbon sequestration of shrubs in the survey plot; C sc,q To investigate the annual carbon sequestration of vegetation in the sample plot;

[0086]

[0087] In the above formula, C sa,q is the annual carbon fixation of trees in the survey plot; C sa,d is the daily net carbon fixation per unit leaf area of ​​the ath tree in the survey plot; L a is the leaf area index of the ath tree in the survey plot; C A,a is the crown area of ​​the ath tree in the survey plot; d is the number of carbon fixation days per year;

[0088] Among them, C A =(π×C1×C2) / 4; C1 is the east-west width of the crown, C2 is the north-south width of the crown;

[0089]

[0090] In the above formula, C sb,q is the annual carbon sequestration of shrubs in the survey plot; C sb,d is the mass of daily net CO2 fixed per unit leaf area of ​​the bth shrub in the survey plot; L b is the leaf area index of the bth shrub in the survey plot; C A,b is the crown area of ​​the bth shrub in the survey plot; d is the number of carbon fixation days per year;

[0091] Among them, C A =(π×C1×C2) / 4; C1 is the east-west width of the crown, C2 is the north-south width of the crown;

[0092]

[0093] In the above formula, C sc,q is the annual carbon fixation of herbs in the survey plot; C sc,d is the daily net CO2 fixation mass per unit leaf area of ​​the cth herb in the survey plot; L c is the leaf area index of the cth herb in the survey plot; S c is the planting area of ​​the cth sample plot in the survey sample plot; d is the number of carbon fixation days per year.

[0094] After the above-mentioned vegetation biomass-carbon fixation calculation model is used to calculate the carbon fixation of vegetation within the sample plot, the calculation results are as follows: Fig. 9 shown.

[0095] It should be noted that the correlation analysis between carbon storage and visible light vegetation index was carried out to determine the visible light vegetation index with strong correlation with vegetation carbon storage. After scatter analysis and correlation coefficient matrix analysis, the top three types of visible light vegetation indices with strong correlation were determined for regression analysis, which were the overgreen index EXG (r = 0.878, p < 0.01), the visible light band difference index VDVI (r = 0.881, p < 0.01), and the super green and super red difference index EXGR (r = 0.885, p < 0.01). Fig.10 And as shown in Table 5.

[0096] Table 5 Correlation coefficient matrix between vegetation carbon storage and visible light vegetation index in each plot area

[0097]

[0098] Note: ** indicates significant correlation at the 0.01 level (two-tailed).

[0099] The present invention further provides a technical solution for establishing a park vegetation carbon storage estimation model based on the above method:

[0100] Divide a number of sample plots of equal area into two parts, one part as training samples and the other part as verification samples, wherein the training samples and the verification samples respectively include: the carbon storage of vegetation in each sample plot, the carbon fixation of vegetation in each sample plot and different visible light indexes corresponding to each sample plot;

[0101] 2. Construction process of the park vegetation carbon storage estimation model and the park vegetation carbon fixation estimation model:

[0102] 1. (a) The construction process of the park vegetation carbon storage estimation model is as follows:

[0103] The carbon storage of vegetation in each plot and the corresponding different visible light indexes in each plot were respectively input into the visible light index-carbon storage calculation sub-model of the target park vegetation, and based on the correlation analysis, the visible light index-carbon storage calculation sub-model with the highest determination coefficient was obtained as the park vegetation carbon storage estimation model;

[0104] Based on the correlation analysis, the determination coefficients and F values ​​in each visible light index-carbon storage calculation sub-model were compared. The model with the highest determination coefficient was the park vegetation carbon storage estimation model, which is as follows:

[0105] The 80 measured sample plots were randomly divided into two parts, and 75% of the samples (60) were selected to be input into each visible light index-carbon storage calculation sub-model to obtain the park vegetation carbon storage estimation model, and the remaining 25% (20) were used to verify the performance of the park vegetation carbon storage estimation model. Fig.11 As shown in Table 6, based on correlation analysis, the determination coefficient R of the visible light index-carbon storage calculation submodel is compared. 2 and F value, determination coefficient R 2 The highest is the regression model of EXGR (R 2 =0.785), and the F significance test shows that the regression model of EXGR (F=104.059, sig<0.001) rejects the null hypothesis and the regression model is statistically significant. Therefore, the EXGR regression model is determined to be used as the vegetation carbon storage estimation model in the park.

[0106] The carbon storage estimation model is specifically as follows: by performing regression analysis on the visible light vegetation index and the measured sample carbon storage, the following is obtained:

[0107]

[0108] In the above formula, CSt is the carbon storage of the entire campus vegetation (Kg); l is a vegetation pixel, and m is the total number of vegetation pixels in the drone image of the entire campus; is the visible light vegetation index value corresponding to the lth pixel; β0, β1, β2 and β3 correspond to the constant term and the coefficients of the first, second and third terms respectively.

[0109] (b) Verification process of the park vegetation carbon storage estimation model:

[0110] 60 modeling samples and 20 validation sample data were used to verify the calculation accuracy of the park vegetation carbon storage estimation model established above, and the verification results are shown in Table 7. As can be seen from the table, the differences in RMSE, MAE and MRE between the EXGR regression model modeling samples and the validation samples are not significant, and the estimation model is shown in the following formula.

[0111] Y EXGR =-790.28x 2 +4489.60x+737.31 (30)

[0112] In the above formula, Y EXGR is the annual carbon storage of vegetation in the park (Kg), and x is the super green and super red difference index EXGR.

[0113] Table 7 Validation indicators of vegetation carbon storage regression model based on EXGR

[0114]

[0115] The park vegetation visible light index-carbon fixation calculation sub-model was constructed by correlation analysis and regression analysis based on the different visible light indices and vegetation carbon fixation of each sample site.

[0116] First, the correlation analysis between carbon fixation and visible light vegetation index was conducted to determine the visible light vegetation index that showed a strong correlation with vegetation carbon storage. After scatter analysis and correlation coefficient matrix analysis, the top three types of visible light vegetation indices with strong correlation were determined for regression analysis, namely, the super green and super red difference index EXGR (r = 0.831, p < 0.01), the overgreen index EXG (r = 0.822, p < 0.01), and the visible light band difference index VDVI (r = 0.815, p < 0.01). Fig.12 and as shown in Table 8.

[0117] Table 8 Correlation coefficient matrix between vegetation carbon fixation and visible light vegetation index in each sample area

[0118]

[0119] Note: ** indicates significant correlation at the 0.01 level (two-tailed).

[0120] 2. (a) Construction process of the park vegetation carbon sequestration estimation model:

[0121] The carbon storage of vegetation in each plot and the corresponding different visible light indexes in each plot were input into the visible light index-carbon fixation calculation sub-model of vegetation in each park, and based on the correlation analysis, the visible light index-carbon fixation calculation sub-model with the highest determination coefficient was obtained as the park vegetation carbon fixation estimation model;

[0122] The visible light index-carbon fixation calculation sub-model with the highest determination coefficient obtained in step S4 is the park vegetation carbon fixation estimation model, specifically: linear regression analysis is performed on the vegetation carbon fixation in the sample plot, and the visible light vegetation index that shows a strong correlation with the vegetation carbon storage is determined to obtain each vegetation carbon fixation regression model, and the determination coefficient and F value in each vegetation carbon fixation regression model are compared. The vegetation carbon fixation regression model with the highest determination coefficient is the park vegetation carbon fixation estimation model, specifically:

[0123] The 80 measured samples were randomly divided into two parts, and 75% of the samples (60) were selected as training samples for modeling; the remaining 25% (20) were used for model verification. Fig.13 As shown in Table 9, based on correlation analysis, by comparing the determination coefficient R 2 and F value, determination coefficient R 2 The highest is the regression model of EXGR (R 2 =0.785), and the F significance test shows that the EXGR regression model (F=79.802, sig<0.001) rejects the null hypothesis and the regression model is statistically significant. Therefore, the EXGR regression model is determined to be used as the estimation model for vegetation carbon storage.

[0124] Table 9 Results of vegetation carbon sequestration estimation model

[0125]

[0126] (b) Validate the model for estimating carbon sequestration in vegetation in the park;

[0127] 60 modeling samples and 20 validation sample data were used to verify the calculation accuracy of the park vegetation carbon storage estimation model established above. The verification results are shown in Table 10. As can be seen from the table, the differences in RMSE, MAE and MRE between the EXGR regression model modeling samples and the validation samples are not significant. The estimation model is shown in the following formula. EXGR is the annual carbon fixation of vegetation in the park (Kg), and x is the super green and super red difference index EXGR.

[0128] Y EXGR =432.2x 2 +174.8x+176.8 (31)

[0129] Table 11 Verification indicators of vegetation carbon sequestration regression model based on EXGR

[0130]

[0131] 3. Calculation of carbon sequestration in park vegetation;

[0132] The calculation of park vegetation carbon sink includes the calculation of park vegetation carbon storage and the calculation of park vegetation carbon fixation. Among them, the park vegetation carbon storage is calculated by sub-modeling the park vegetation visible light index-carbon storage calculation sub-model obtained in S4 into the visible light index of vegetation in the whole area obtained in S1; the park vegetation carbon fixation is calculated by sub-modeling the park vegetation visible light index-carbon fixation calculation sub-model obtained in S4 into the visible light index of vegetation in the whole area obtained in S1.

[0133] (1) Calculation of the visible light vegetation index in the park

[0134] First, the EXGR visible light vegetation index is calculated based on the overall image of the park (see Table 1, Formula 2 for details). The calculation results are as follows: Fig.14 shown.

[0135] (2) The resampling command was executed on the calculation results of each visible light vegetation index of the entire park. The pixel size was consistent with the sampling survey sample size, using a unit pixel size of 10m*10m. The resampling technology used was the NEAREST nearest neighbor method, and the following was obtained: Fig.15 The resampling calculation results are shown.

[0136] (3) By extracting the resampling result values ​​of all the super-green and super-red differential indexes EXGR (see Formula 2 in Table 1 for details) of the vegetation coverage area at the scale of the park, the blue-green ratio index GBRI (see Formula 4 in Table 1 for details) is used to obtain the surface element shape of the overall vegetation coverage area of ​​the park. The super-green and super-red differential index EXGR raster data corresponding to the vegetation coverage area can be obtained by executing the tool command of extracting information by mask on the resampling calculation results of the super-green and super-red differential index EXGR through GIS software. Then, all the values ​​of the super-green and super-red differential index EXGR corresponding to the vegetation coverage area of ​​the park can be extracted through the raster conversion tool and the table conversion Excel tool. The carbon storage of the park vegetation is calculated according to formula (30). After calculation, the overall carbon storage of the park is 13050032.87 kg, that is, 13050.03 t. The carbon sequestration of the park vegetation is calculated according to formula (31). After calculation, the annual carbon sequestration of the overall vegetation in the park is 1886972.39 kg, that is, 1886.97 t.

[0137] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for calculating vegetation carbon sink at park scale based on visible light vegetation index, characterized in that: The following steps are involved: S1. Obtain a visible light orthographic image of a target park, perform preprocessing, calculate the blue-green ratio index GBRI of the preprocessed visible light orthographic image, obtain different visible light information data, and input the data into a visible light vegetation index calculation model to obtain a visible light vegetation index within the range of the target park; S2. Select several sample plots of equal area from the target park, and collect specific information of vegetation in the sample plots of equal area; S3, inputting the collected specific information of vegetation into the vegetation biomass calculation model to obtain the biomass of vegetation in each sample plot; The biomass of vegetation in each plot is input into the vegetation biomass-carbon storage calculation model to obtain the carbon storage of vegetation in each plot; The growth parameters and leaf area index of the vegetation in each plot were input into the vegetation biomass-carbon sequestration calculation model to obtain the carbon sequestration of the vegetation in each plot; S4, dividing a number of sample plots of equal area into training samples and verification samples, wherein the training samples and verification samples respectively include: the carbon storage of vegetation in each sample plot, the carbon fixation of vegetation in each sample plot, and different visible light indexes corresponding to each sample plot; Using the carbon storage of vegetation in each plot and the corresponding different visible light indexes in each plot, a vegetation carbon storage estimation model for the park was established; Using the carbon storage of vegetation in each plot and the corresponding different visible light indexes in each plot, a model for estimating the carbon sequestration of vegetation in the park was established; Use the park vegetation carbon storage estimation model to estimate the carbon storage in the target park; The carbon sequestration amount in the target park is estimated using the park vegetation carbon sequestration estimation model.

2. The method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index according to claim 1 is characterized in that: In the step S1, a visible light orthographic image of the target park is obtained and preprocessed, specifically: strong interference factors in the visible light orthographic image of the target park are eliminated by cropping; The strong interference factors include water surface, aquatic plants and track and field pavement.

3. The method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index according to claim 1 is characterized in that: In step S1, the blue-green ratio index GBRI of the pre-processed visible light orthographic image is calculated to obtain different visible light information data, specifically: Based on the blue-green ratio index (GBRI), the overall pixels of the visible light orthophoto are reclassified into vegetation pixels and non-vegetation image sources, and the raster data of the vegetation coverage area in the overall pixels are extracted. The surface element shape of the vegetation coverage area is converted and used as a mask to obtain different visible light information data in the vegetation coverage area.

4. The method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index according to claim 1 is characterized in that: In step S2, a plurality of plots of equal area are selected from the target park, specifically: According to the size of the target park, multiple plots of a set area are randomly selected; The specific information of vegetation collected in the sample plots is specifically: collecting tree information in the sample plots, including the type, diameter at breast height, crown width and plant height of the tree; The shrub information and herb information of the sample plot are collected, and the shrub information and herb information respectively include: the area, base diameter, crown width, plant height and area of ​​the shrub.

5. The method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index according to claim 1 is characterized in that: The vegetation biomass calculation model in step S3 is specifically: B Q,i =B A,i +B B,i +B C,i Where B Q,i ——vegetation biomass of the ith survey plot; B A,i ——the biomass of trees in the ith survey plot; B B,i ——the biomass of shrubs in the i-th survey plot; B C,i ——The biomass of herbs in the ith survey plot.

6. The method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index according to claim 1 is characterized in that: The vegetation biomass-carbon storage calculation model in step S3 is specifically: CS Q,i =B Q,i ×α In the above formula, CS Q,i is the vegetation carbon storage of the i-th survey plot; B Q,i ——Vegetation biomass of the ith survey plot; α is the carbon content of vegetation.

7. The method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index according to claim 1 is characterized in that: The vegetation biomass-carbon fixation calculation model in step S3 is specifically: C s,q =C sa,q +C sb,q +C sc,q ; In the above formula, C s,q is the annual carbon fixation of vegetation in the survey plot; C sa,q is the annual carbon fixation of trees in the survey plot; C sb,q is the annual carbon sequestration of shrubs in the survey plot; C sc,q To investigate the annual carbon sequestration of vegetation in the sample plot; In the above formula, C sa,q is the annual carbon fixation of trees in the survey plot; C sa,d is the daily net carbon fixation per unit leaf area of ​​the ath tree in the survey plot; L a is the leaf area index of the ath tree in the survey plot; C A,a is the crown area of ​​the ath tree in the survey plot; d is the number of carbon fixation days per year; Among them, C A =(π×C1×C2) / 4; C1 is the east-west width of the crown, C2 is the north-south width of the crown; In the above formula, C sb,q is the annual carbon sequestration of shrubs in the survey plot; C sb,d is the mass of daily net CO2 fixed per unit leaf area of ​​the bth shrub in the survey plot; L b is the leaf area index of the bth shrub in the survey plot; C A,b is the crown area of ​​the bth shrub in the survey plot; d is the number of carbon fixation days per year; Among them, C A =(π×C1×C2) / 4; C1 is the east-west width of the crown, C2 is the north-south width of the crown; In the above formula, C sc,q is the annual carbon fixation of herbs in the survey plot; C sc,d is the daily net CO2 fixation mass per unit leaf area of ​​the cth herb in the survey plot; L C is the leaf area index of the cth herb in the survey plot; S c is the planting area of ​​the cth sample plot in the survey sample plot; d is the number of carbon fixation days per year.

8. The method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index according to claim 1 is characterized in that: In step S4, the carbon storage of vegetation in each plot and the corresponding different visible light indexes in each plot are used to establish a park vegetation carbon storage estimation model, which is specifically: Constructing a visible light index-carbon sink calculation model, wherein the visible light index-carbon sink calculation model includes a visible light index-carbon storage calculation sub-model of vegetation in the target park and a visible light index-carbon fixation calculation sub-model of vegetation in the target park; The carbon storage of vegetation in each plot and the corresponding different visible light indexes in each plot were respectively input into the visible light index-carbon storage calculation sub-model of vegetation in the target park, and based on the correlation analysis, the visible light index-carbon storage calculation sub-model with the highest determination coefficient was obtained as the carbon storage estimation model of vegetation in the park; Using the carbon storage of vegetation in each plot and the corresponding different visible light indexes in each plot, a model for estimating the carbon sequestration of vegetation in the park was established. Specifically: The carbon storage of vegetation in each sample plot and the corresponding different visible light indices in each sample plot were respectively input into the visible light index-carbon fixation calculation sub-model of vegetation in the target park, and based on correlation analysis, the visible light index-carbon fixation calculation sub-model with the highest determination coefficient was obtained as the park vegetation carbon fixation estimation model.

9. The method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index according to claim 8 is characterized in that: The method for constructing the visible light index-carbon storage calculation sub-model of vegetation in the target park is specifically as follows: Resample the visible light vegetation index within each plot; Based on the resampling calculation results, the scatter plot and Pearson correlation analysis method were used to analyze the correlation between different visible light indices and the carbon storage of vegetation in each plot, and the top three types of visible light vegetation indices with the strongest correlation were determined. The method for constructing the visible light index-carbon fixation calculation submodel of the vegetation in the target park is specifically as follows: resampling the visible light vegetation index within each sample plot; based on the resampling calculation results, using a scatter plot and Pearson correlation analysis method to perform correlation analysis on the vegetation carbon storage in each sample plot and the different visible light vegetation indices in each sample plot, and determining the top three visible light vegetation indices with the strongest correlation.

10. The method for calculating the vegetation carbon sink at the park scale based on the visible light vegetation index according to claim 8, characterized in that: The visible light index-carbon storage calculation sub-model with the highest determination coefficient is used as the park vegetation carbon storage estimation model, specifically: Based on the correlation analysis, the determination coefficients and F values ​​in each visible light index-carbon storage calculation sub-model were compared, and the visible light index-carbon storage calculation sub-model with the highest determination coefficient was the park vegetation carbon storage estimation model; The visible light index-carbon fixation calculation sub-model with the highest determination coefficient is the park vegetation carbon fixation estimation model, specifically: A linear regression analysis was performed on the carbon fixation of vegetation in the sample plot to determine the visible light vegetation index that showed a strong correlation with vegetation carbon storage, and each visible light index-carbon fixation calculation sub-model was obtained. The determination coefficient and F value in each visible light index-carbon fixation calculation sub-model were compared, and the visible light index-carbon fixation calculation sub-model with the highest determination coefficient was the vegetation carbon fixation estimation model for the park.

Images