Carbon sink monitoring system based on current carbon sink monitoring methodology

Abstract

The embodiment of the invention discloses a carbon sink monitoring system based on an existing carbon sink monitoring methodology, and the system comprises a data module which is used for obtaining the whole forest data in a to-be-monitored region; wherein the whole forest data is obtained based on remote sensing technology monitoring; the biomass calculation module is used for calculating by combining the whole forest data with the verified biomass remote sensing model to obtain biomass data; the data processing module is used for performing estimation processing on the biomass data in combination with a constructed regression model for carbon reserve estimation so as to obtain carbon reserves; subtracting the carbon reserves at the starting time and the ending time to obtain corresponding carbon sinks; the method has the advantages that multiple technologies based on satellite remote sensing are integrated, a calculation method is provided for key parameter variables in existing methodology, the data acquisition difficulty is reduced, and more complete and comprehensive monitoring is provided for forest land carbon sink monitoring.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon sink monitoring, and in particular to a carbon sink monitoring system based on current carbon sink monitoring methodology. Background Art

[0002] Currently, the Voluntary Greenhouse Gas Emission Reduction Project Methodology Afforestation Carbon Sink (CCER-14-001-V01)

[0003] (referred to as "methodology") is the methodology for forest afforestation carbon sinks. Different research fields have proposed different measurement methods for carbon sinks, including calculation methods based on net primary productivity, monitoring methods based on changes in oxygen concentration, and measurement methods based on eddy covariance technology. In the current methodology for monitoring forest carbon sinks, the carbon stock change method is mainly used for calculation. This method requires measuring the project start time, project end time and related forest tree parameter values ​​(area, tree height, breast diameter, number of trees, hectare stocking), etc. When obtaining these parameter values, there are problems such as long measurement period, waste of more manpower and material resources, low degree of intelligence, and high measurement cost.

[0004] Although there are some monitoring schemes that rely on models, when they are applied, the relevant parameters are selected through theoretical analysis or experience, which has certain limitations and is not scientific and perfect, resulting in the defect of low accuracy.

[0005] In view of this, a carbon sink monitoring system is proposed. Based on the computational architecture of the methodology, it integrates a variety of satellite remote sensing-based technologies to provide calculation methods for key variables in the methodology, thereby reducing the difficulty of data acquisition and providing a more complete and comprehensive monitoring solution for carbon sink monitoring in forest land. Summary of the invention

[0006] In view of the technical defects mentioned in the background technology, an object of the embodiments of the present invention is to provide a carbon sink monitoring system based on the current carbon sink monitoring methodology to solve the problems raised in the above background technology.

[0007] To achieve the above objectives, an embodiment of the present invention provides a carbon sink monitoring system based on the current carbon sink monitoring methodology, the system comprising:

[0008] A data module is used to obtain the whole forest data in the area to be monitored; wherein the whole forest data is obtained based on remote sensing technology monitoring;

[0009] A biomass calculation module, used to calculate the whole forest data in combination with the verified biomass remote sensing model to obtain biomass data;

[0010] Data processing module for:

[0011] The biomass data is combined with the constructed regression model for carbon storage estimation to perform estimation processing to obtain carbon storage;

[0012] Subtract the carbon storage at the start time from that at the end time to obtain the corresponding carbon sink.

[0013] As a specific implementation of the present application, the whole forest data includes forest area, number of trees, volume per unit area, diameter at breast height of each tree and height of each tree.

[0014] As a specific implementation method of the present application, the forest area is obtained by monitoring using remote sensing technology, and the specific steps include:

[0015] Acquire multispectral satellite images of the area to be monitored;

[0016] Select forest sample points;

[0017] Classification using classifiers or deep learning;

[0018] Extract forest land distribution data based on classification results;

[0019] The forest area is calculated according to the forest distribution data.

[0020] As a specific implementation of the present application, the estimation of the number of trees and the height of the trees includes:

[0021] First, the acquired aerial images are denoised; then, an object-oriented, deep learning segmentation method is used to identify the tree crown width in the monitoring area, and the number of crown widths is the number of trees;

[0022] The DEM and the digital surface model DSM are then used to construct the surface information of the forest, and then the canopy height model and the canopy width information are used to obtain the height of each tree; wherein the canopy height model is obtained by subtracting the DSM from the DEM.

[0023] As a specific implementation of the present application, the steps of obtaining the diameter at breast height of each tree include:

[0024] UAV multispectral image acquisition;

[0025] Ground sample area division and sampling;

[0026] Image denoising processing;

[0027] Extract the crown width and crown area based on the processed images;

[0028] Model building and validation;

[0029] The DBH was estimated based on the obtained model.

[0030] As a specific implementation of the present application, the extraction of crown width and crown area based on the processed image includes:

[0031] Use image segmentation technology to segment the tree crown area in the image;

[0032] Then, the crown width and crown area are extracted from the segmented crown region; wherein the crown width is calculated based on the segmentation boundary; and the crown area is obtained by calculating the number of pixels in the segmented region and converting it into the actual area.

[0033] As a specific implementation of the present application, the accumulation amount is obtained by specifically steps including:

[0034] Acquire multispectral satellite imagery;

[0035] Delineate sample plots and obtain forest inventory data;

[0036] Satellite image preprocessing;

[0037] Calculate vegetation index;

[0038] Build regression models;

[0039] Model validation and stock estimation.

[0040] As a specific implementation of the present application, the biomass data is based on multispectral satellite images and ground inventory data, and uses vegetation indices and verified biomass remote sensing models to estimate the forest biomass in the monitoring area.

[0041] As a specific implementation of the present application, the estimation of carbon reserves specifically includes:

[0042] Obtain multispectral satellite images of the monitoring area;

[0043] Divide the sample area, obtain the forest inventory data in the sample area, and use the preset biomass formula to calculate the biomass of the sample area, and use the preset carbon storage formula to calculate the carbon storage of the sample area;

[0044] Preprocess the multispectral satellite images and then calculate the vegetation index;

[0045] Using the carbon storage data per unit area of ​​a set proportion in the sample area, and constructing multiple regression models with at least one of the vegetation indices;

[0046] The remaining proportion of sample area data was used as validation, and the model with the highest determination coefficient and the smallest error was selected as the regression model for carbon storage estimation, thereby calculating the carbon storage in multispectral satellite images within the monitoring area.

[0047] The technical solution provided by the embodiment of the present invention has the following beneficial effects:

[0048] 1. Based on the carbon sink calculation under the methodological framework, a carbon sink monitoring framework was constructed, and the relevant parameters and calculations are scientific and authoritative;

[0049] 2. The framework of the present invention incorporates a variety of basic data monitoring methods, calculates the data required in the methodology, and can meet the user's monitoring needs in terms of accuracy, breadth, cost, etc.;

[0050] 3. It incorporates a variety of satellite remote sensing technologies to provide calculation methods for key parameter variables in current methodologies, reduces the difficulty of data acquisition, and provides more complete and comprehensive monitoring for forest carbon sinks. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the specific implementation of the present invention or the technical solution in the prior art, the drawings required for use in the specific implementation or the description of the prior art are briefly introduced below.

[0052] Figure 1 is a block diagram of a carbon sink monitoring system based on current carbon sink monitoring methodology provided by an embodiment of the present invention;

[0053] Figure 2 It is a schematic diagram of a framework of a carbon sink monitoring methodology provided by an embodiment of the present invention;

[0054] Figure 3 It is a schematic diagram of a process for calculating a diameter at breast height provided by an embodiment of the present invention;

[0055] Figure 4 It is a processing schematic diagram of accumulation amount calculation provided by an embodiment of the present invention;

[0056] Figure 5 It is a processing schematic diagram of carbon storage calculation provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0059] DEM (Digital Elevation Model): DEM for short is a digital simulation of the ground terrain through limited terrain elevation data. It is a physical ground model that represents the ground elevation in the form of a set of ordered numerical arrays.

[0060] DSM (Digital Surface Model): DSM for short is an extension of the Digital Elevation Model (DEM for short). It not only contains the elevation information of the surface, but also records the height information of surface objects such as buildings and trees.

[0061] Please refer to Figures 1 to 5 , an embodiment of the present invention provides a carbon sink monitoring system based on the current carbon sink monitoring methodology, the system comprising:

[0062] A data module is used to obtain the whole forest data in the area to be monitored; wherein the whole forest data is obtained based on remote sensing technology monitoring;

[0063] A biomass calculation module, used to calculate the whole forest data in combination with the verified biomass remote sensing model to obtain biomass data;

[0064] Data processing module for:

[0065] The biomass data is combined with the constructed regression model for carbon storage estimation to perform estimation processing to obtain carbon storage;

[0066] Subtract the carbon storage at the start time from that at the end time to obtain the corresponding carbon sink.

[0067] In this embodiment, the whole forest data includes forest area, number of trees, stock volume per unit area, diameter at breast height of each tree and height of each tree; forest inventory data (forestry statistical data) of sample plots in the study area can also be obtained through the Forestry Bureau and other departments or field surveys, and the data includes area, number of trees, stock volume per unit area, diameter at breast height of each tree, height of each tree, etc.

[0068] In this embodiment, the forest area is monitored using remote sensing technology, and the specific steps include:

[0069] Acquire multispectral satellite images of the area to be monitored;

[0070] Select forest sample points;

[0071] Classification using classifiers or deep learning;

[0072] Extract forest land distribution data based on classification results;

[0073] The forest area is calculated according to the forest distribution data.

[0074] That is, the number of forest pixels in the study area is n, and then the forest area area is calculated according to the spatial resolution p of the image. The calculation formula is area = p*p*n. The specific steps are as follows:

[0075] Acquiring multispectral satellite imagery

[0076] Data source selection: Select a multispectral satellite image data source suitable for the research area and research purpose, such as Landsat series, Sentinel series, etc.;

[0077] Image preprocessing: Perform necessary preprocessing on the acquired satellite images, including radiation correction, atmospheric correction, geometric correction, cropping, etc., to ensure the quality and accuracy of the images.

[0078] Obtain forest sample points through field survey data or expert experience

[0079] Field research: Through field investigation, representative forest sample points are selected in the study area, and their location, type and other information are recorded.

[0080] Expert experience: If field surveys are not feasible or costly, you can rely on the knowledge and experience of experts to determine the location and attributes of forest sample points.

[0081] Supervised classification or deep learning classification

[0082] Sample selection: Based on field survey data or expert experience, representative forest and non-forest land samples are selected from satellite images.

[0083] Classification method:

[0084] ①Supervised classification: Use the selected samples to train the classifier (such as maximum likelihood classifier, support vector machine, etc.), then classify the entire image to extract the forest distribution data.

[0085] ② Deep learning classification: Using deep learning technology (such as convolutional neural network (CNN)) to extract and classify images can usually achieve higher classification accuracy.

[0086] Extracting forest land distribution data

[0087] Post-processing of classification results: Post-process the classification results, such as removing small spots and smoothing boundaries, to improve the accuracy of forest distribution data.

[0088] Forest pixel quantity statistics: Count the number of forest pixels n in the classification results.

[0089] Calculate forest area

[0090] Spatial resolution p: Get the spatial resolution p (usually in meters) of the satellite image based on its metadata.

[0091] Area calculation: The forest area is calculated using the formula area = p*p*n. Here, p*p represents the actual ground area (square meters) represented by a pixel, and n represents the number of forest pixels.

[0092] Result analysis and verification

[0093] Result analysis: The extracted forest distribution data and calculated forest area were analyzed to evaluate their rationality and accuracy.

[0094] Verification: Other independent data sources (such as higher-resolution satellite imagery, aerial imagery, ground-truth data, etc.) can also be used to verify the results to ensure their reliability.

[0095] Through the above process, we can effectively utilize multispectral satellite images and field survey data or expert experience, extract the forest distribution data in the study area through supervised classification or deep learning classification methods, and calculate the forest area based on this; when calculating the area, the above functional steps can be combined into an area calculation model as shown in the attached figure.

[0096] When implemented, the estimation of the number and height of the trees includes:

[0097] First, the acquired aerial images are denoised; then, an object-oriented, deep learning segmentation method is used to identify the tree crown width in the monitoring area, and the number of crown widths is the number of trees;

[0098] The DEM and the digital surface model DSM are then used to construct the surface information of the forest, and then the canopy height model and the canopy width information are used to obtain the height of each tree; wherein the canopy height model is obtained by subtracting the DSM from the DEM.

[0099] In this embodiment, a drone equipped with a laser radar is used to obtain an aerial image of the area. The specific steps are as follows:

[0100] (1) Image denoising

[0101] Purpose: Since LiDAR data may be affected by a variety of noise sources (such as atmospheric conditions, instrument errors, ground reflections, etc.) during the acquisition process, denoising is an important step in data preprocessing to improve the accuracy and reliability of the data.

[0102] method:

[0103] ① Filtering algorithm: Commonly used filtering algorithms include Gaussian filtering, median filtering, bilateral filtering, etc. These algorithms can help remove random noise in point cloud data.

[0104] ②Statistical filtering: Identify and remove outliers based on the statistical characteristics of point cloud data (such as distance, density, etc.).

[0105] ③ Ground point separation: The ground points are separated from the vegetation points through algorithms (such as cloth simulation algorithm, progressive morphological filtering, etc.), which helps with subsequent processing.

[0106] (2) Tree crown identification

[0107] Purpose: To identify the crown width of each tree, which is a key step in estimating the number and distribution of trees.

[0108] method:

[0109] ① Object-oriented classification: Objects (such as tree crowns) are formed by segmenting point cloud data, and classified based on the geometric features of the objects (such as shape, size, height, etc.) and reflection intensity.

[0110] ② Deep learning: Use convolutional neural networks (CNN) or other deep learning models to process point cloud data and automatically identify tree crowns by training the model. This method usually requires a large amount of labeled data for training to build the crown width extraction model shown in the attached figure.

[0111] (3) Using DEM and DSM to construct forest surface information

[0112] Purpose: DEM (digital elevation model) represents the terrain surface, while DSM (digital surface model) includes the height information of terrain and vegetation surface. The difference between the two can be used to obtain the surface information of forests.

[0113] method:

[0114] Subtracting DSM from DEM: Directly subtract DEM from DSM to obtain the vegetation height model (CHM, canopy height model), which reflects the height of vegetation (mainly tree crowns) above the ground.

[0115] Canopy segmentation: Based on CHM and combined with the previously identified tree crown positions, the canopy height information of each tree is further segmented.

[0116] (4) Calculate tree height

[0117] Purpose: To calculate the height of each tree using a canopy height model and canopy width information.

[0118] method:

[0119] ① Direct measurement: In CHM, for each tree crown, the vertical distance between its highest point and the ground is measured, which is the tree height.

[0120] ②Statistical analysis: For complex situations, it may be necessary to conduct statistical analysis in combination with the geometric characteristics of the tree crown (such as shape, volume, etc.) to more accurately estimate the tree height.

[0121] in conclusion

[0122] By using drones equipped with LiDAR systems to survey and monitor forest resources, combined with image denoising, object-oriented or deep learning segmentation, and DEM and DSM analysis, key information such as tree crown width, quantity, and tree height can be obtained efficiently and accurately.

[0123] Furthermore, the step of obtaining the diameter at breast height of each tree comprises:

[0124] UAV multispectral image acquisition;

[0125] Ground sample area division and sampling;

[0126] Image denoising processing;

[0127] Extract the crown width and crown area based on the processed images;

[0128] Model building and validation;

[0129] The DBH was estimated based on the obtained model.

[0130] The extraction of crown width and crown area based on the processed image includes:

[0131] Use image segmentation technology to segment the tree crown area in the image;

[0132] Then, the crown width and crown area are extracted from the segmented crown region; wherein the crown width is calculated based on the segmentation boundary; and the crown area is obtained by calculating the number of pixels in the segmented region and converting it into the actual area.

[0133] In this embodiment, the model for estimating the diameter at breast height based on the obtained model is the diameter at breast height prediction model in the accompanying drawings;

[0134] (1) UAV multispectral image acquisition

[0135] UAV flight: UAVs equipped with multispectral sensors were used to fly over the study area to obtain high-resolution images in multiple bands.

[0136] Image stitching: stitch multiple images taken by drones to form a complete image of the study area.

[0137] (2) Ground sample area division and sampling

[0138] Sample area division: Sample areas should be divided reasonably according to research needs and characteristics of the study area. Sample areas should be representative and cover different forest types and growth conditions.

[0139] Data collection: In each plot, field measurements are taken of each tree to record information such as crown width (usually the diameter in the east-west and north-south directions), crown area (measured with the assistance of other tools such as laser rangefinders or GPS), and diameter at breast height (the diameter of the trunk at 1.3 meters from the ground).

[0140] (3) Image denoising

[0141] Denoising algorithm: Denoise the multispectral images acquired by the drone to reduce the impact of noise on subsequent processing. Commonly used denoising methods include median filtering, Gaussian filtering, etc.

[0142] (4) Extraction of crown width and crown area

[0143] Image segmentation: Use the watershed algorithm or other image segmentation techniques to segment the tree crown area in the image. The watershed algorithm is particularly suitable for processing objects with obvious boundaries, such as tree crowns.

[0144] Feature extraction: Extract the crown width (calculated based on the segmentation boundary) and crown area (directly calculate the number of pixels in the segmented area and convert it into the actual area) from the segmented crown area.

[0145] (5) Model construction and verification

[0146] Data division: The collected ground sample data is divided into two parts, 80% for model training and 20% for model verification.

[0147] Model fitting: Use the crown width, crown area and DBH data in the training data set to construct multiple regression models (such as linear regression, polynomial regression, nonlinear regression, etc.), so that the model should be able to predict DBH as a function of crown width and / or crown area.

[0148] Model validation: The performance of each model was evaluated using the validation dataset and the coefficient of determination (R 2 The model with the highest score is taken as the final model. The closer the coefficient of determination is to 1, the stronger the predictive ability of the model is.

[0149] (6) Estimation of DBH

[0150] Image calculation: The diameter at breast height of each tree in the study area was estimated using the final selected model and the crown width and crown area information extracted from the drone imagery.

[0151] Result analysis: Perform statistical analysis on the estimation results to assess their accuracy and reliability. If necessary, fine-tune or retrain the model to improve estimation accuracy.

[0152] In this embodiment, the accumulation amount is obtained by specifically steps including:

[0153] Acquire multispectral satellite imagery;

[0154] Delineate sample plots and obtain forest inventory data;

[0155] Satellite image preprocessing;

[0156] Calculate vegetation index;

[0157] Build regression models;

[0158] Model validation and stock estimation.

[0159] Specifically, (1) obtain multispectral satellite images of the study area.

[0160] Select appropriate multispectral satellite data sources, such as Landsat series, Sentinel series, etc., to ensure that the time, resolution and coverage of the images meet research needs.

[0161] (2) Divide sample areas and obtain forest inventory data

[0162] The sample plots were rationally divided within the study area to ensure that they were representative and covered different forest stand types and growth conditions.

[0163] Conduct a ground inventory of each sample area and record the volume per unit area (i.e. the total volume of trees per unit area), which is the key data for estimating the volume.

[0164] (3) Satellite image preprocessing

[0165] Perform radiometric calibration on multispectral satellite images and convert the DN value of the image into radiometric brightness value or reflectivity.

[0166] Perform atmospheric correction to eliminate the influence of atmosphere on surface reflectivity and obtain more realistic surface reflectance information.

[0167] Reproject the images to the same geographic coordinate system and projection.

[0168] Perform image registration to ensure spatial consistency between multi-temporal or multi-source images.

[0169] (4) Calculation of vegetation index

[0170] According to the band information of multispectral images, the Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Soil Adjusted Vegetation Index (SAVI), Modified Soil Adjusted Vegetation Index (MSAVI), etc. are calculated. These indices are important indicators reflecting the growth status and coverage of vegetation.

[0171]

[0172] EVI takes into account the influence of atmospheric and soil background. The calculation formula is relatively complex and usually contains additional coefficients and terms.

[0173]

[0174] SAVI and MSAVI are improvements on NDVI, aiming to reduce the impact of soil background on vegetation indices.

[0175]

[0176]

[0177] ⑤⑥⑦⑧In the equations, NIR is the near infrared band, RED is the red band, and BLUE is the blue band;

[0178] (5) Constructing a regression model

[0179] Use 80% of the unit area volume data in the sample area and the corresponding vegetation index (one or more of NDVI, EVI, SAVI, MSAVI) to build a regression model. You can choose linear regression, polynomial regression, nonlinear regression and other models to try.

[0180] By comparing the coefficient of determination (R 2 ), root mean square error (RMSE) and other statistical indicators, and the model with the best performance is selected as the final model for accumulation estimation; the final model for accumulation estimation here is the accumulation remote sensing model in the attached figure.

[0181] (6) Model verification and accumulation estimation

[0182] The remaining 20% ​​of the sample data are used to verify the selected model to ensure the accuracy and reliability of its prediction results.

[0183] Apply the validated model to the multispectral satellite imagery of the entire study area and calculate the stock volume of all forest pixels in the image. This usually involves substituting the vegetation index value into the regression model and calculating the corresponding stock volume value.

[0184] (7) Results analysis and discussion

[0185] Conduct statistical analysis on the estimation results, including the spatial distribution of accumulation volume and calculation of total volume.

[0186] The estimated results were compared with the ground inventory data to assess the estimation accuracy and applicability of the model.

[0187] in conclusion

[0188] Through the above process, the forest stock in the study area can be estimated using vegetation indices and regression models based on multispectral satellite images and ground inventory data.

[0189] In this embodiment, the biomass data is based on multispectral satellite images and ground inventory data, and uses vegetation index and a verified biomass remote sensing model to estimate the forest biomass in the monitoring area.

[0190] Specifically, 80% of the biomass data per unit area in the sample area were used to build a regression model with one or more indexes such as NDVI, EVI, SAVI, MSAVI, etc., and the remaining 20% ​​of the sample point data were used for verification. The model with the highest determination coefficient was selected as the regression model for biomass estimation, and the biomass of all forest pixel points in the study area calculated using images was used.

[0191] The carbon stock estimation specifically includes:

[0192] Obtain multispectral satellite images of the monitoring area;

[0193] Divide the sample area, obtain the forest inventory data in the sample area, and use the preset biomass formula to calculate the biomass of the sample area, and use the preset carbon storage formula to calculate the carbon storage of the sample area;

[0194] Preprocess the multispectral satellite images and then calculate the vegetation index;

[0195] Using the carbon storage data per unit area of ​​a set proportion in the sample area, and constructing multiple regression models with at least one of the vegetation indices;

[0196] The remaining proportion of sample area data was used as validation, and the model with the highest determination coefficient and the smallest error was selected as the regression model for carbon storage estimation, thereby calculating the carbon storage in multispectral satellite images within the monitoring area.

[0197] It should be noted that the biomass formula is as follows:

[0198] Bio=(a*DBH b *H c )*N*area ①

[0199] Bio=(a+b*V)*area②

[0200] ①Wherein Bio is biomass;

[0201] DBH is the diameter at breast height, H is the tree height, N is the number of trees per unit area, and area is the area;

[0202] a, b, and c are empirical coefficients;

[0203] ②In the formula, Bio is biomass;

[0204] V is the volume per unit area, and area is the area;

[0205] The carbon storage formula is as follows:

[0206] T=Bio*CF ③

[0207] ③Where T is carbon storage;

[0208] Bio is biomass;

[0209] CF is the carbon content of biomass, which is an empirical coefficient;

[0210] Carbon sink calculation

[0211] Carbon storage is used to calculate carbon sinks, and the calculation formula is as follows:

[0212] Tc n =T n -T n-1 ④

[0213] ④TC n is the carbon sink in the nth year;

[0214] T n is the carbon storage in year n.

[0215] (1) Obtain multispectral satellite images of the study area

[0216] Select a multispectral satellite data source that covers the study area, such as the Landsat series, Sentinel series, etc., to ensure that the resolution, timestamp, and coverage of the image meet the research requirements.

[0217] (2) Divide sample areas and obtain forest inventory data

[0218] Representative sample plots were divided in the study area to ensure that they could reflect the characteristics of different forest stand types, ages, growth conditions, etc.

[0219] Conduct a detailed ground inventory of each sample area, record basic information such as tree height, diameter at breast height, number of trees, and use the biomass formula (which is usually based on an empirical formula or model based on factors such as tree height and diameter at breast height) to calculate the biomass of the sample area.

[0220] The biomass is then converted to carbon stock using a carbon stock formula (which is usually based on biomass or a carbon content factor for a specific tree species). Carbon stock formulas may vary by tree species, region, and purpose of the study.

[0221] (3) Multispectral satellite image preprocessing

[0222] The acquired satellite images are radiometrically calibrated to convert the original DN values ​​into radiation brightness or reflectivity.

[0223] Perform atmospheric correction to eliminate the influence of atmosphere on surface reflectivity and obtain more realistic surface information.

[0224] Reproject all images to the same geographic coordinate system and projection.

[0225] Perform image registration to ensure the spatial consistency of images from different times or different sensors.

[0226] (4) Calculation of vegetation index

[0227] The band information of the preprocessed multispectral image is used to calculate the Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI), Soil Adjusted Vegetation Index (SAVI) and Modified Soil Adjusted Vegetation Index (MSAVI), etc. These indices are important indicators reflecting the growth status and coverage of vegetation.

[0228]

[0229] EVI takes into account the influence of atmospheric and soil background. The calculation formula is relatively complex and usually contains additional coefficients and terms.

[0230]

[0231] SAVI and MSAVI are improvements on NDVI, aiming to reduce the impact of soil background on vegetation indices.

[0232]

[0233]

[0234] ⑤⑥⑦⑧In the equations, NIR is the near infrared band, RED is the red band, and BLUE is the blue band;

[0235] (5) Construction of carbon storage estimation regression model

[0236] A regression model is constructed using 80% of the unit area carbon storage data in the sample area and the corresponding vegetation index (one or more of NDVI, EVI, SAVI, MSAVI). Different regression methods (such as linear regression, polynomial regression, machine learning algorithms, etc.) can be tried to find the best model.

[0237] The performance of the model is usually evaluated by calculating the coefficient of determination (R 2 ), root mean square error (RMSE) and other statistical indicators.

[0238] Use methods such as cross-validation or leave-one-out validation to ensure the stability and generalization ability of the model.

[0239] The model with the highest determination coefficient and the smallest error is selected as the final carbon storage estimation model; the carbon storage estimation model here is the carbon storage remote sensing model in the attached figure.

[0240] (6) Model validation and carbon stock estimation

[0241] The remaining 20% ​​of the sample data were used to validate the selected model to ensure its predictive ability on unknown data.

[0242] The validated model was applied to the multispectral satellite images of the entire study area to calculate the carbon storage of all forest pixels in the image. The vegetation index value was substituted into the regression model to calculate the corresponding carbon storage value.

[0243] Through the above process, the forest carbon stock in the study area can be estimated using vegetation indices and regression models based on multispectral satellite images and ground inventory data.

[0244] Based on the attached Figure 4 and 5 It can be seen that when the accumulation volume and carbon storage are applied, the corresponding model is first derived based on the sample area data, and then the model is applied to the satellite data of the whole forest data of the entire area for calculation and processing.

[0245] This scheme builds a carbon sink monitoring framework based on carbon sink calculation under the methodological framework, and the relevant parameters and calculations are scientific and authoritative;

[0246] The framework of the present invention incorporates a variety of basic data monitoring methods, calculates the data required in the methodology, and can meet the user's monitoring needs in terms of accuracy, breadth, cost, etc.;

[0247] It incorporates a variety of satellite remote sensing-based technologies to provide calculation methods for key parameter variables in current methodologies, reduces the difficulty of data acquisition, and provides more complete and comprehensive monitoring of forest carbon sinks.

[0248] In the embodiments provided in the present application, it should be understood that the disclosed system can also be implemented in other ways. The system embodiments described above are merely schematic. It should also be noted that in some alternative implementations, the functions marked in the boxes can also occur in an order different from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0249] In addition, the functional modules in the various embodiments of the present invention may be integrated together to form an independent part, or each module may exist independently, or two or more modules may be integrated to form an independent part.

[0250] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various changes and variations. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

[0251] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A carbon sink monitoring system based on current carbon sink monitoring methodology, characterized in that: The system comprises: A data module is used to obtain the whole forest data in the area to be monitored; wherein the whole forest data is obtained based on remote sensing technology monitoring; A biomass calculation module, used to calculate the whole forest data in combination with the verified biomass remote sensing model to obtain biomass data; Data processing module for: The biomass data is combined with the constructed regression model for carbon storage estimation to perform estimation processing to obtain carbon storage; Subtract the carbon storage at the start time from that at the end time to obtain the corresponding carbon sink.

2. A carbon sink monitoring system based on current carbon sink monitoring methodology as claimed in claim 1, characterized in that: The whole forest data includes forest area, number of trees, volume per unit area, diameter at breast height of each tree and height of each tree.

3. A carbon sink monitoring system based on current carbon sink monitoring methodology as claimed in claim 2, characterized in that: The forest area is monitored using remote sensing technology, and the specific steps include: Acquire multispectral satellite images of the area to be monitored; Select forest sample points; Classification using classifiers or deep learning; Extract forest land distribution data based on classification results; The forest area is calculated according to the forest distribution data.

4. A carbon sink monitoring system based on current carbon sink monitoring methodology as claimed in claim 2, characterized in that: The estimation of the number and height of the trees includes: First, the acquired aerial images are denoised; then, an object-oriented, deep learning segmentation method is used to identify the tree crown width in the monitoring area, and the number of crown widths is the number of trees; The DEM and the digital surface model DSM are then used to construct the surface information of the forest, and then the canopy height model and the canopy width information are used to obtain the height of each tree; wherein the canopy height model is obtained by subtracting the DSM from the DEM.

5. A carbon sink monitoring system based on current carbon sink monitoring methodology as claimed in claim 2, characterized in that: The specific steps of obtaining the diameter at breast height of each tree include: UAV multispectral image acquisition; Ground sample area division and sampling; Image denoising processing; Extract the crown width and crown area based on the processed images; Model building and validation; The DBH was estimated based on the obtained model.

6. A carbon sink monitoring system based on current carbon sink monitoring methodology as claimed in claim 5, characterized in that: The extraction of crown width and crown area based on the processed image includes: Use image segmentation technology to segment the tree crown area in the image; Then, the crown width and crown area are extracted from the segmented crown region; wherein the crown width is calculated based on the segmentation boundary; and the crown area is obtained by calculating the number of pixels in the segmented region and converting it into the actual area.

7. A carbon sink monitoring system based on current carbon sink monitoring methodology as claimed in claim 2, characterized in that: The accumulation amount is obtained by specifically comprising the following steps: Acquire multispectral satellite imagery; Delineate sample plots and obtain forest inventory data; Satellite image preprocessing; Calculate vegetation index; Build regression models; Model validation and stock estimation.

8. A carbon sink monitoring system based on current carbon sink monitoring methodology as claimed in claim 7, characterized in that: The biomass data is based on multispectral satellite images and ground inventory data, using vegetation indices and validated biomass remote sensing models to estimate forest biomass within the monitoring area.

9. A carbon sink monitoring system based on current carbon sink monitoring methodology as claimed in any one of claims 2 to 8, characterized in that: The carbon stock estimation specifically includes: Obtain multispectral satellite images of the monitoring area; Divide the sample area, obtain the forest inventory data in the sample area, and use the preset biomass formula to calculate the biomass of the sample area, and use the preset carbon storage formula to calculate the carbon storage of the sample area; Preprocess the multispectral satellite images and then calculate the vegetation index; Using the carbon storage data per unit area of ​​a set proportion in the sample area, and constructing multiple regression models with at least one of the vegetation indices; The remaining proportion of sample area data was used as validation, and the model with the highest determination coefficient and the smallest error was selected as the regression model for carbon storage estimation, thereby calculating the carbon storage in multispectral satellite images within the monitoring area.

Images