Regional forest carbon space distribution acquisition method

Abstract

The invention relates to the field of forest carbon sink prediction, in particular to a regional forest carbon spatial distribution acquisition method, which comprises the following steps of: judging forest types, vegetation growth states and soil information by combining satellite remote sensing with unmanned aerial vehicle aerial photography information, calculating forest carbon density through the information, and drawing the forest carbon density on a geographic image through a GIS (Geographic Information System) technology. The forest carbon space distribution diagram is obtained, the problems in the prior art are solved, the calculation precision can be improved by improving the fine granularity, the carbon reserve of a single tree can be estimated more accurately, then the carbon reserve condition of the whole forest is deduced, and the carbon space distribution diagram of the forest is obtained by detecting the air carbon dioxide content of the target area. And whether the vegetation growth condition information of the target area is accurate is determined, so that whether forest carbon density calculation is accurate is verified, and the accuracy of forest carbon spatial distribution drawing is improved.

Description

Technical Field

[0001] The invention relates to the field of forest carbon sink prediction, and in particular to a method for obtaining regional forest carbon spatial distribution. Background Art

[0002] Climate change is a major challenge facing the world today, and forests, as one of the most important terrestrial ecosystems on Earth, play an important role in regulating climate change. Plants in forests absorb carbon dioxide through photosynthesis and convert it into organic matter, thus playing an important role as a carbon sink in the Earth's atmosphere. Therefore, understanding the spatial distribution of forest carbon is of great significance for assessing the impact of climate change and taking corresponding regulatory and management measures. At present, some regional-scale forest carbon spatial distribution models have been proposed, but these models are usually based on a macroscopic scale and fail to reflect the fine-grained spatial distribution of forest carbon.

[0003] Therefore, the present invention proposes a method for obtaining the spatial distribution of regional forest carbon, which solves the problem of macroscopic limitations in the measurement of forest carbon spatial distribution in the prior art, and achieves the effect of improving the granularity of forest carbon spatial distribution calculation and improving the accuracy of forest carbon measurement. Summary of the invention

[0004] In order to solve the problems in the above-mentioned prior art, the first purpose of the present invention is to provide a method for obtaining the spatial distribution of regional forest carbon. By using satellite remote sensing combined with drone aerial photography information, the forest type, vegetation growth status and soil information are judged, and the forest carbon density is calculated through the above information. The forest carbon density is plotted on the geographic image through GIS technology to obtain a forest carbon spatial distribution map, which solves the problems existing in the prior art and achieves the effect of improving the accuracy of forest carbon measurement.

[0005] In order to achieve the above object, the present invention provides a method for obtaining regional forest carbon spatial distribution, comprising: S1: Use satellite remote sensing to obtain remote sensing images of the target area, and obtain spectral feature information, texture feature information and shape feature information of the target area through the image data of the target area; S2: Obtain forest type information of the target area through spectral feature information, texture feature information and shape feature information of the target area; S3: The UAV loads the forest type information of the target area, flies in the target area and takes images of vegetation in the target area, and obtains the growth status information of vegetation in the target area based on the vegetation images of the target area combined with the forest type information; S4: Acquire soil images in the target area through the UAV, and calculate the soil information of the target area according to the soil images; S5: Calculate the carbon density of the target area by combining the vegetation growth status information of the target area and the soil information of the target area; S6: Use a gas analyzer mounted on a drone to detect the carbon dioxide content in the target area, and verify the accuracy of the carbon density in the target area based on the carbon dioxide content in the target area; S7: Use GIS technology to map the carbon density of the target area in geographic space to obtain a forest carbon spatial distribution map.

[0006] As a further improvement of the present invention, the S2 includes: using the red light band and the near infrared band of the spectral characteristics to calculate the normalized vegetation index to obtain vegetation greenness information; using the gray level co-occurrence matrix to calculate the contrast and energy of the remote sensing image of the target area to obtain the texture details and spatial distribution information of the target area; using threshold segmentation to separate the forest area from the background in the remote sensing image of the target area to obtain the segmented area, calculating the area and perimeter of the segmented area, obtaining the shape characteristics of the target area, and obtaining the forest patch compactness of the target area according to the shape characteristics. By calculating the normalized vegetation index and using the reflectivity difference of the red light band and the near infrared band, the distinction between vegetation and other surface cover types is effectively enhanced; the texture characteristics can provide valuable information about tree species composition, forest age and interference degree, which is helpful to further subdivide forest types; the shape characteristics can be used to obtain the forest patch compactness of the target area, and the forest type of the target area can be assisted in judging the forest patch compactness. Coniferous forests have discontinuous forest patches, while broad-leaved forests are continuous forests with few forest patches, so the forest type can be judged.

[0007] As a further improvement of the present invention, S2 further includes using texture details and spatial distribution information to preliminarily determine forest type information of the target area, including: , where S1 represents the first forest type judgment value, Co is the contrast, , En is energy, , where P(i,j) is the normalized frequency of the (i,j)th element when using GLCM to calculate contrast and energy. The first forest type judgment threshold T1 is set. When S1>T1, the forest type information is mixed forest; when S1<T1, the forest type is judged to be a single tree species forest, and then the specific forest type information is obtained through spectral characteristics.

[0008] As a further improvement of the present invention, when the forest type is a single tree species forest, the forest type information is obtained as follows: , where S2 represents the second forest type judgment value, NVDI is the normalized vegetation index, , ρ NIR is the reflectivity in the near-infrared band, ρRED is the reflectivity of the red light band, K is the compactness of the forest patch, , where Sl is the forest area of ​​the target area, Cl is the forest perimeter of the target area, and the second forest type judgment threshold T2 is set. When S2>T2, the forest type information obtained is broad-leaved forest, and when S2<T2, the forest type information obtained is coniferous forest.

[0009] Since the reflectance of coniferous forests and broad-leaved forests in the near-infrared band and the red band is different, the normalized vegetation index of the target area can be calculated through the reflectance of the forest in this frequency band. After confirming whether the forest is a single-species forest, the forest in the target area can be secondary differentiated. The vegetation index is calculated by selecting bands that are sensitive to forest types. Among them, the reflectance of coniferous forests in the red band and the near-infrared band is low, while the reflectance of broad-leaved forests in the red band and the near-infrared band is high. Therefore, the normalized vegetation index can be used to determine the forest type; By initially using texture features to judge forest types, the scope of subsequent spectral feature analysis can be quickly narrowed, and the overall judgment efficiency can be improved. On this basis, combining shape features and spectral features for forest type identification can more comprehensively reflect the actual situation of the forest and improve the accuracy and reliability of judgment. In addition, by adjusting the threshold and feature parameters according to specific needs, it can adapt to the judgment of different regions and different types of forests.

[0010] As a further improvement of the present invention, S3 includes: S31: Combined with the forest type information of the target area, the leaf area L0, trunk diameter D0, trunk height H0 and tree age A0 of the optimal carbon density of the vegetation in the corresponding forest type are loaded into the database; S32: Obtain a vegetation image of the target area, and obtain the leaf area L, trunk diameter D, trunk height H and tree age A of the vegetation in the target area through the vegetation image of the target area; S33: Compare L, D, H and A with L0, D0, H0 and A0 to obtain the growth parameters of vegetation in the target area ; S34: Determine the growth status of vegetation in the target area through the growth parameter G; when G≥0.8, the growth status information of vegetation in the target area is Class I vegetation; when G<0.8 and G≥0.6, the growth status information of vegetation in the target area is Class II vegetation; when G<0.6 and G≥0.2, the growth status information of vegetation in the target area is Class III vegetation; when G<0.2, the growth status information of vegetation in the target area is Class IV vegetation.

[0011] By comparing the actual growth parameters of vegetation in the target area with the benchmark values, the growth status of vegetation can be evaluated more accurately. By comparing the leaf area, trunk diameter, trunk height and age of vegetation in the target area with the above information of optimal carbon density, the contrast between the status of vegetation in the target area and the vegetation under the optimal carbon density can be obtained, and then the impact of vegetation in the target area on carbon density can be calculated. Among them, vegetation from Class I to Class III respectively represents that the growth status of vegetation is gradually reduced, and its carbon absorption capacity and carbon density are also gradually reduced; Class IV vegetation represents ungrown vegetation or aging vegetation that does not meet the forest carbon sink standard, and represents the lowest carbon density. Through the specific classification of vegetation, the fine-grainedness of forest carbon density calculation can be improved, and a more refined forest carbon spatial distribution map can be obtained.

[0012] As a further improvement of the present invention, the S4 comprises: S41: Acquire soil images in the target area through UAV; S42: obtaining soil color, soil texture and soil structure from soil images; S43: Determine soil fertility and moisture conditions based on soil color, soil texture and soil structure to form soil information Si of the target area.

[0013] By evaluating the soil information of the target area, using soil color to analyze the mineral composition, organic matter content and water sealing status of the soil, using soil texture to analyze the water retention capacity, aeration and fertility of the soil, and using soil structure to determine the soil particle size, and combining the above information, it is possible to determine the impact of soil in the target area on forest carbon density and thus obtain the spatial distribution of forest carbon.

[0014] As a further improvement of the present invention, the S5 comprises: S51: preset parameters, when the forest type is mixed forest, the forest type coefficient S=0.8; when the forest type is broad-leaved forest, the forest type coefficient S=1; when the forest type is coniferous forest, the forest type coefficient S=0.5; S52: Preset the effect of the growth state of vegetation in the target area on the carbon density; when the growth state information of the vegetation in the target area is Class I vegetation, the vegetation carbon coefficient Gi=1; when the growth state information of the vegetation in the target area is Class II vegetation, the vegetation carbon coefficient Gi=0.8; when the growth state information of the vegetation in the target area is Class III vegetation, the vegetation carbon coefficient Gi=0.5; when the growth state information of the vegetation in the target area is Class IV vegetation, the vegetation carbon coefficient Gi=0; S53: Calculate carbon density ; Among them, α=0.2, β=0.6, γ=0.2.

[0015] By combining the forest type, vegetation growth status and soil information in the target area, we can more accurately assess the forest carbon density in different regions, types and growth status, and then draw an accurate spatial distribution map.

[0016] As a further improvement of the present invention, in S6, the UAV is equipped with a gas analyzer to detect and obtain the carbon dioxide content in the air of the target area, and sets the maximum threshold value of the carbon dioxide content in the air in the low carbon density area and the minimum threshold value of the carbon dioxide content in the air in the high carbon density area. When Cd>0.6, the target area is obtained as a high carbon density area. At this time, if the carbon dioxide content in the air in the target area is less than the minimum threshold value of the carbon dioxide content in the air, the carbon density calculation of the target area is inaccurate, and the carbon density correction step is entered. Otherwise, the carbon density deep optimization step is entered; when Cd<0.6, the target area is obtained as a low carbon density area. At this time, if the carbon dioxide content in the air in the target area is greater than the maximum threshold value of the carbon dioxide content in the air, the carbon density calculation of the target area is inaccurate, and the carbon density correction step is entered, and the corrected carbon density is output as the final carbon density.

[0017] As a further improvement of the present invention, the carbon density correction step is provided with a correction formula: ; ; Among them, δ(N) is the limiting function, N is the carbon dioxide content, a is the adjustment coefficient, N t is the carbon dioxide content threshold, is the corrected carbon density; when the target area is a high carbon density area and the carbon density correction step is performed, a is a negative value, N t is the minimum threshold of carbon dioxide content; when the target area is a low carbon density area and the carbon density correction step is performed, a is a positive value, N t is the maximum threshold of carbon dioxide content.

[0018] By using a drone equipped with a gas analyzer to conduct actual detection of the carbon dioxide content in the air in the target area, the calculated results of the forest carbon density can be verified. When the actual measured carbon dioxide content in the air in the target area does not match the calculated forest carbon density, it proves that the calculated results of the forest carbon density are inaccurate.

[0019] As a further improvement of the present invention, S7 includes obtaining the geospatial data of the target area from the database, cleaning the carbon density data, removing outliers, matching the cleaned carbon density data with the geospatial data, and importing the matched carbon density data and geospatial data into a GIS project, using the GIS project to interpolate the carbon density data to generate a continuous carbon density distribution map, and adding the carbon density distribution map to a reference layer to generate a forest carbon spatial distribution map.

[0020] As a further improvement of the present invention, the forest carbon spatial distribution map includes carbon density high-value areas, low-value areas and special area markings.

[0021] By using GIS technology to map forest carbon density into geographic images and marking high and low carbon density areas, we can clearly identify areas with high carbon density in the forest, which helps to protect areas with high forest carbon and ensure the continued function of its carbon sink function. It can also clearly identify areas with low carbon density in the forest, which helps to identify potential areas for forest restoration and regeneration.

[0022] The beneficial effects of the present invention are: 1. The present invention introduces a more sophisticated spatial division technology. By using satellite remote sensing to determine forest types and combining it with drone survey area images, the present invention solves the problem of large errors in the calculation of carbon distribution details caused by extensive regional division or grid system in the existing forest carbon spatial distribution calculation, and realizes fine-grained measurement and modeling of forest carbon distribution, thereby more accurately capturing the effects of different tree species, different forest ages, and different soil conditions on carbon storage, thereby greatly improving the accuracy of forest carbon spatial distribution.

[0023] 2. The present invention can improve the calculation accuracy by improving the granularity, and can more accurately estimate the carbon storage of a single tree, and then infer the carbon storage of the entire forest.

[0024] 3. The present invention detects the carbon dioxide content in the air of the target area, and then determines whether the vegetation growth status information in the target area is accurate, thereby verifying whether the forest carbon density calculation is accurate, thereby improving the accuracy of the forest carbon spatial distribution mapping. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a flow chart of a method for obtaining the spatial distribution of regional forest carbon in the present invention; Figure 2 It is a first step-by-step flow chart of a method for obtaining regional forest carbon spatial distribution of the present invention; Figure 3 It is a second step-by-step flow chart of a method for obtaining regional forest carbon spatial distribution of the present invention; Figure 4This is the third sub-step flow chart of a method for acquiring regional forest carbon spatial distribution of the present invention. DETAILED DESCRIPTION

[0026] The specific embodiments of the present invention will be described in detail below. It should be noted that the embodiments described here are only part of the embodiments of the present invention, not all of the embodiments. Based on the following embodiments, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] The embodiment of the present invention provides a method for obtaining regional forest carbon spatial distribution, including: S1: Use satellite remote sensing to obtain remote sensing images of the target area, and obtain spectral feature information, texture feature information and shape feature information of the target area through the image data of the target area; S2: Obtain forest type information of the target area through spectral feature information, texture feature information and shape feature information of the target area; S3: The UAV loads the forest type information of the target area, flies in the target area and takes images of vegetation in the target area, and obtains the growth status information of vegetation in the target area based on the vegetation images of the target area combined with the forest type information; S4: Acquire soil images in the target area through the UAV, and calculate the soil information of the target area according to the soil images; S5: Calculate the carbon density of the target area by combining the vegetation growth status information of the target area and the soil information of the target area; S6: Use a gas analyzer mounted on a drone to detect the carbon dioxide content in the target area, and verify the accuracy of the carbon density in the target area based on the carbon dioxide content in the target area; S7: Use GIS technology to map the carbon density of the target area in geographic space to obtain a forest carbon spatial distribution map.

[0028] In this embodiment, refer to Figure 1-4As shown, the S2 includes calculating the normalized vegetation index using the red band and near infrared band of the spectral characteristics to obtain vegetation greenness information; calculating the contrast and energy of the remote sensing image of the target area using the gray level co-occurrence matrix to obtain the texture details and spatial distribution information of the target area; using threshold segmentation to separate the forest area from the background in the remote sensing image of the target area to obtain the segmented area, calculating the area and perimeter of the segmented area, obtaining the shape characteristics of the target area, and obtaining the forest patch compactness of the target area according to the shape characteristics. By calculating the normalized vegetation index and using the reflectivity difference between the red band and the near infrared band, the distinction between vegetation and other surface cover types is effectively enhanced; the texture characteristics can provide valuable information about tree species composition, forest age and interference degree, which is helpful to further subdivide forest types; the shape characteristics can be used to obtain the forest patch compactness of the target area, and the forest type of the target area can be assisted in judging the forest patch compactness. Coniferous forests have discontinuous forest patches, while broad-leaved forests are continuous forests with few forest patches, so the forest type can be judged.

[0029] In this embodiment, refer to Figure 1-4 As shown, the S2 also includes using texture details and spatial distribution information to preliminarily determine the forest type information of the target area, including: , where S1 represents the first forest type judgment value, Co is the contrast, , En is energy, , where P(i,j) is the normalized frequency of the (i,j)th element when using GLCM to calculate contrast and energy. The first forest type judgment threshold T1 is set. When S1>T1, the forest type information is mixed forest; when S1<T1, the forest type is judged to be a single tree species forest, and then the specific forest type information is obtained through spectral characteristics.

[0030] In this embodiment, refer to Figure 1-4 As shown, when the forest type is a single tree species forest, the forest type information obtained includes: , where S2 represents the second forest type judgment value, NVDI is the normalized vegetation index, , ρ NIR is the reflectivity in the near-infrared band, ρ RED is the reflectivity of the red light band, K is the compactness of the forest patch, , where Sl is the forest area of ​​the target area, Cl is the forest perimeter of the target area, and the second forest type judgment threshold T2 is set. When S2>T2, the forest type information obtained is broad-leaved forest, and when S2<T2, the forest type information obtained is coniferous forest.

[0031] By initially using texture features to judge forest types, the scope of subsequent spectral feature analysis can be quickly narrowed, and the overall judgment efficiency can be improved. On this basis, combining shape features and spectral features for forest type identification can more comprehensively reflect the actual situation of the forest and improve the accuracy and reliability of judgment. In addition, by adjusting the threshold and feature parameters according to specific needs, it can adapt to the judgment of different regions and different types of forests.

[0032] In this embodiment, refer to Figure 1-4 As shown, the S3 includes: S31: Combined with the forest type information of the target area, the leaf area L0, trunk diameter D0, trunk height H0 and tree age A0 of the optimal carbon density of the vegetation in the corresponding forest type are loaded into the database; S32: Obtain a vegetation image of the target area, and obtain the leaf area L, trunk diameter D, trunk height H and tree age A of the vegetation in the target area through the vegetation image of the target area; S33: Compare L, D, H and A with L0, D0, H0 and A0 to obtain the growth parameters of vegetation in the target area ; S34: Determine the growth status of vegetation in the target area through the growth parameter G; when G≥0.8, the growth status information of vegetation in the target area is Class I vegetation; when G<0.8 and G≥0.6, the growth status information of vegetation in the target area is Class II vegetation; when G<0.6 and G≥0.2, the growth status information of vegetation in the target area is Class III vegetation; when G<0.2, the growth status information of vegetation in the target area is Class IV vegetation.

[0033] By comparing the actual growth parameters of vegetation in the target area with the benchmark values, the growth status of vegetation can be evaluated more accurately. By comparing the leaf area, trunk diameter, trunk height and age of vegetation in the target area with the above information of optimal carbon density, the contrast between the status of vegetation in the target area and the vegetation under the optimal carbon density can be obtained, and then the impact of vegetation in the target area on carbon density can be calculated. Among them, vegetation from Class I to Class III respectively represents that the growth status of vegetation is gradually reduced, and its carbon absorption capacity and carbon density are also gradually reduced; Class IV vegetation represents ungrown vegetation or aging vegetation that does not meet the forest carbon sink standard, and represents the lowest carbon density. Through the specific classification of vegetation, the fine-grainedness of forest carbon density calculation can be improved, and a more refined forest carbon spatial distribution map can be obtained.

[0034] In this embodiment, refer to Figure 1-4 As shown, the S4 includes: S41: Use drones to obtain soil images in the target area, and use the drone's high-altitude photography capabilities to quickly and efficiently obtain large-scale soil images in the target area.

[0035] S42: obtaining soil color, soil texture and soil structure from soil images; Soil color reflects information such as the mineral composition, organic matter content, and water status of the soil. Through image processing technology, the RGB value of soil color can be extracted from the soil image for analysis. Soil texture refers to the size, shape, and arrangement of soil particles, which affects the soil's water retention capacity, aeration, and fertility. Through texture analysis algorithms, different texture features can be identified from soil images; soil structure refers to the arrangement and combination of soil particles.

[0036] S43: Determine soil fertility and moisture conditions based on soil color, soil texture and soil structure to form soil information Si of the target area.

[0037] Soil fertility refers to the ability of soil to provide the nutrients needed for plant growth. By analyzing whether the soil color is dark, we can determine whether the soil contains organic matter. By analyzing the soil texture, for example, fine-grained soil usually has a strong ability to retain fertilizer, we can preliminarily judge the soil fertility level. Soil moisture is one of the key factors for plant growth. In terms of soil color, moist soil has a darker color. In terms of soil structure, loose soil has poor water retention capacity. The above information can be used to analyze the soil moisture status.

[0038] The information on soil color, soil texture, soil structure, soil fertility and water conditions obtained from the above analysis is integrated to form the soil information Si of the target area, and the spatial distribution of forest carbon is subsequently evaluated based on the soil information.

[0039] By evaluating the soil information of the target area, using soil color to analyze the mineral composition, organic matter content and water sealing status of the soil, using soil texture to analyze the water retention capacity, aeration and fertility of the soil, and using soil structure to determine the soil particle size, and combining the above information, it is possible to determine the impact of soil in the target area on forest carbon density and thus obtain the spatial distribution of forest carbon.

[0040] In this embodiment, refer to Figure 1-4 As shown, the S5 includes: S51: preset parameters, when the forest type is mixed forest, the forest type coefficient S=0.8; when the forest type is broad-leaved forest, the forest type coefficient S=1; when the forest type is coniferous forest, the forest type coefficient S=0.5; S52: Preset the effect of the growth state of vegetation in the target area on the carbon density; when the growth state information of the vegetation in the target area is Class I vegetation, the vegetation carbon coefficient Gi=1; when the growth state information of the vegetation in the target area is Class II vegetation, the vegetation carbon coefficient Gi=0.8; when the growth state information of the vegetation in the target area is Class III vegetation, the vegetation carbon coefficient Gi=0.5; when the growth state information of the vegetation in the target area is Class IV vegetation, the vegetation carbon coefficient Gi=0; S53: Calculate carbon density ; Among them, α=0.2, β=0.6, γ=0.2.

[0041] By combining the forest type, vegetation growth status and soil information in the target area, we can more accurately assess the forest carbon density in different regions, types and growth status, and then draw an accurate spatial distribution map.

[0042] In this embodiment, refer to Figure 1-4 As shown, as a further improvement of the present invention, in S6, the UAV is equipped with a gas analyzer to detect and obtain the air carbon dioxide content in the target area, and sets the maximum threshold of the air carbon dioxide content in the low carbon density area and the minimum threshold of the air carbon dioxide content in the high carbon density area. When Cd>0.6, the target area is obtained as a high carbon density area. At this time, if the air carbon dioxide content in the target area is less than the minimum threshold of the air carbon dioxide content, the carbon density calculation of the target area is inaccurate, and the carbon density correction step is entered. Otherwise, the carbon density deep optimization step is entered; when Cd<0.6, the target area is obtained as a low carbon density area. At this time, if the air carbon dioxide content in the target area is greater than the maximum threshold of the air carbon dioxide content, the carbon density calculation of the target area is inaccurate, and the carbon density correction step is entered, and the corrected carbon density is output as the final carbon density.

[0043] During the carbon density detection process, since there is a direct correlation between the air carbon dioxide content and the carbon density of the target area, in areas with high carbon density, the plants corresponding to this forest type absorb more air carbon dioxide, so the air carbon dioxide content is relatively low. If the air carbon dioxide content in the target area is higher than the air carbon dioxide content threshold of the high carbon density area, it means that the carbon density calculation of the area is inaccurate. At this time, the carbon density of the target area is recalculated in combination with the air carbon dioxide content.

[0044] In this embodiment, refer to Figure 1-4 As shown, the carbon density correction step is provided with a correction formula: ; ; Among them, δ(N) is the limiting function, N is the carbon dioxide content, a is the adjustment coefficient, N t is the carbon dioxide content threshold, is the corrected carbon density; when the target area is a high carbon density area and the carbon density correction step is performed, a is a negative value, N t is the minimum threshold of carbon dioxide content; when the target area is a low carbon density area and the carbon density correction step is performed, a is a positive value, N tis the maximum threshold of carbon dioxide content. δ(N) is related to carbon dioxide content. When carbon dioxide content exceeds the threshold, the exponential growth relationship of the difference is calculated. The adjustment coefficient a is set according to the forest type of the target area. The forest carbon density classification of the target area is initially calculated, that is, the high carbon density area and the low carbon density area are adjusted positively or negatively. When an error occurs in the detection of a high carbon density area, a is a negative value, otherwise it is a positive value. δ(N) represents the impact of carbon dioxide content on carbon density, which indirectly affects the vegetation carbon coefficient and thus affects the calculation of carbon density in the target area.

[0045] By using a drone equipped with a gas analyzer to conduct actual detection of the carbon dioxide content in the air in the target area, the calculated results of the forest carbon density can be verified. When the actual measured carbon dioxide content in the air in the target area does not match the calculated forest carbon density, it proves that the calculated results of the forest carbon density are inaccurate.

[0046] The S7 includes obtaining the geospatial data of the target area from the database, cleaning the carbon density data, removing outliers, matching the cleaned carbon density data with the geospatial data, and importing the matched carbon density data and geospatial data into a GIS project, interpolating the carbon density data using the GIS project to generate a continuous carbon density distribution map, and adding the carbon density distribution map to a reference layer to generate a forest carbon spatial distribution map.

[0047] In this embodiment, refer to Figure 1-4 As shown, the forest carbon spatial distribution map includes high-value carbon density areas, low-value areas and special area markings.

[0048] By using GIS technology to map forest carbon density into geographic images and marking high and low carbon density areas, we can clearly identify areas with high carbon density in the forest, which helps to protect areas with high forest carbon and ensure the continued function of its carbon sink function. It can also clearly identify areas with low carbon density in the forest, which helps to identify potential areas for forest restoration and regeneration.

[0049] The above shows and describes the basic features, principles and advantages of the present invention. It should be noted that the present invention is not limited to the above embodiments, which are only partial embodiments. Under the premise of not departing from the spirit and scope of the present invention, several improvements and supplements made are considered as the protection scope of the present invention.

Claims

1. A method for obtaining the spatial distribution of regional forest carbon, characterized in that: include: S1: Use satellite remote sensing to obtain remote sensing images of the target area, and obtain spectral feature information, texture feature information and shape feature information of the target area through the image data of the target area; S2: obtaining forest type information of the target area through spectral feature information, texture feature information and shape feature information of the target area, and obtaining forest type coefficient through the forest type information; S3: The UAV loads the forest type information of the target area, flies in the target area and takes images of vegetation in the target area, obtains the growth status information of vegetation in the target area based on the vegetation images of the target area combined with the forest type information, and obtains the vegetation carbon coefficient of the target area through the growth status information; S4: Acquire soil images in the target area through the UAV, and calculate the soil information factor of the target area according to the soil images; S5: Calculate the carbon density of the target area by combining the vegetation carbon coefficient, forest type coefficient and soil information factor of the target area; S6: Use a gas analyzer mounted on a drone to detect the carbon dioxide content in the target area, and verify the accuracy of the carbon density in the target area based on the carbon dioxide content in the target area. When the carbon density in the target area is inaccurate, enter the carbon density correction step, and verify the accuracy of the corrected carbon density again; when the carbon density in the target area is accurate, enter the carbon density deep optimization step to obtain the optimized carbon density value; S7: Use GIS technology to plot the optimized carbon density values ​​of the target area in geographic space to obtain a forest carbon spatial distribution map.

2. A method for obtaining regional forest carbon spatial distribution according to claim 1, characterized in that: The S2 includes calculating the normalized vegetation index using the red light band and the near infrared band of the spectral characteristics to obtain vegetation greenness information; calculating the contrast and energy of the remote sensing image of the target area using the gray level co-occurrence matrix to obtain the texture details and spatial distribution information of the target area; separating the forest area from the background in the remote sensing image of the target area using threshold segmentation to obtain the segmented area, calculating the area and perimeter of the segmented area to obtain the shape characteristics of the target area, and obtaining the forest patch compactness of the target area according to the shape characteristics; The texture details and spatial distribution information are used to preliminarily determine the forest type information in the target area, including: , where S1 represents the first forest type judgment value, Co is the contrast, , En is energy, , where P(i,j) is the normalized frequency of the (i,j)th element when using GLCM to calculate contrast and energy. The first forest type judgment threshold T1 is set. When S1>T1, the forest type information is mixed forest; when S1<T1, the forest type is judged to be a single tree species forest, and then the specific forest type information is obtained through spectral characteristics.

3. The method for obtaining the spatial distribution of regional forest carbon according to claim 2, characterized in that: When the forest type is a single tree species forest, the forest type information obtained includes: , where S2 represents the second forest type judgment value, NVDI is the normalized vegetation index, , ρ NIR is the reflectivity in the near-infrared band, ρ RED is the reflectivity of the red light band, K is the compactness of the forest patch, , where Sl is the forest area of ​​the target area, Cl is the forest perimeter of the target area, and the second forest type judgment threshold T2 is set. When S2>T2, the forest type information obtained is broad-leaved forest, and when S2<T2, the forest type information obtained is coniferous forest.

4. The method for obtaining the spatial distribution of regional forest carbon according to claim 1, characterized in that: The S3 includes: S31: Combined with the forest type information of the target area, the leaf area L0, trunk diameter D0, trunk height H0 and tree age A0 of the optimal carbon density of the vegetation in the corresponding forest type are loaded into the database; S32: Obtain a vegetation image of the target area, and obtain the leaf area L, trunk diameter D, trunk height H and tree age A of the vegetation in the target area through the vegetation image of the target area; S33: Compare L, D, H and A with L0, D0, H0 and A0 to obtain the growth parameters of vegetation in the target area ; S34: Determine the growth status of vegetation in the target area through the growth parameter G; when G≥0.8, the growth status information of vegetation in the target area is Class I vegetation; when G<0.8 and G≥0.6, the growth status information of vegetation in the target area is Class II vegetation; when G<0.6 and G≥0.2, the growth status information of vegetation in the target area is Class III vegetation; when G<0.2, the growth status information of vegetation in the target area is Class IV vegetation.

5. The method for obtaining the spatial distribution of regional forest carbon according to claim 1, characterized in that: The S4 includes: S41: Acquire soil images in the target area through UAV; S42: obtaining soil color, soil texture and soil structure from soil images; S43: Soil fertility and moisture conditions are determined based on soil color, soil texture and soil structure to form a soil information factor Si of the target area.

6. The method for obtaining regional forest carbon spatial distribution according to claim 1, characterized in that: The S5 includes: S51: preset parameters, when the forest type is mixed forest, the forest type coefficient S=0.8; when the forest type is broad-leaved forest, the forest type coefficient S=1; when the forest type is coniferous forest, the forest type coefficient S=0.5; S52: Preset the effect of the growth state of vegetation in the target area on the carbon density; when the growth state information of the vegetation in the target area is Class I vegetation, the vegetation carbon coefficient Gi=1; when the growth state information of the vegetation in the target area is Class II vegetation, the vegetation carbon coefficient Gi=0.8; when the growth state information of the vegetation in the target area is Class III vegetation, the vegetation carbon coefficient Gi=0.5; when the growth state information of the vegetation in the target area is Class IV vegetation, the vegetation carbon coefficient Gi=0; S53: Calculate carbon density ; Among them, α is the proportional weight of soil information factor, β is the proportional weight of vegetation carbon coefficient, and γ is the proportional weight of forest type coefficient.

7. The method for obtaining the spatial distribution of regional forest carbon according to claim 1, characterized in that: In S6, the UAV is equipped with a gas analyzer to detect and obtain the carbon dioxide content in the air of the target area, and sets the maximum threshold value of the carbon dioxide content in the air in the low carbon density area and the minimum threshold value of the carbon dioxide content in the air in the high carbon density area. When Cd>0.6, the target area is obtained as a high carbon density area. At this time, if the carbon dioxide content in the air in the target area is less than the minimum threshold value of the carbon dioxide content in the air, the carbon density calculation of the target area is inaccurate, and the carbon density correction step is entered. Otherwise, the carbon density deep optimization step is entered; when Cd<0.6, the target area is obtained as a low carbon density area. At this time, if the carbon dioxide content in the air in the target area is greater than the maximum threshold value of the carbon dioxide content in the air, the carbon density calculation of the target area is inaccurate, and the carbon density correction step is entered, and the corrected carbon density is output as the final carbon density.

8. The method for obtaining the spatial distribution of regional forest carbon according to claim 7, characterized in that: The carbon density correction step is provided with a correction formula: ； ； Among them, δ(N) is the limiting function, N is the carbon dioxide content, a is the adjustment coefficient, N t is the carbon dioxide content threshold, is the corrected carbon density; when the target area is a high carbon density area and the carbon density correction step is performed, a is a negative value, N t is the minimum threshold of carbon dioxide content; when the target area is a low carbon density area and the carbon density correction step is performed, a is a positive value, N t is the maximum threshold of carbon dioxide content.

9. The method for obtaining regional forest carbon spatial distribution according to claim 1, characterized in that: The S7 includes obtaining the geospatial data of the target area from the database, cleaning the carbon density data, removing outliers, matching the cleaned carbon density data with the geospatial data, and importing the matched carbon density data and geospatial data into the GIS project, interpolating the carbon density data using the GIS project to generate a continuous carbon density distribution map, and adding the carbon density distribution map to the reference layer to generate a forest carbon spatial distribution map, wherein the forest carbon spatial distribution map includes carbon density high value areas, low value areas and special area annotations.

Images