Mangrove forest wetland soil carbon reserve assessment method, device and equipment and storage medium

Abstract

The invention discloses a mangrove forest wetland soil carbon reserve assessment method, device and equipment and a storage medium, and the method comprises the steps: obtaining remote sensing image data, and carrying out the ground feature coverage classification of a to-be-assessed region; the hyperspectral reflectivity data of soil at different depths are actually measured by using a ground feature spectrometer; performing layered evaluation on the actually measured hyperspectral data by combining a soil density evaluation model and an organic carbon content evaluation model to obtain the soil density and the organic carbon content of each sampling point at different depths; determining soil carbon reserves at different depths of each sampling point based on the soil density and the organic carbon content at different depths of each sampling point; and carrying out spatial interpolation processing on the soil carbon reserves at different depths of each sampling point to obtain a carbon reserve distribution diagram of the to-be-evaluated region. According to the method, the sample points are determined by using the remote sensing image data, and the soil carbon reserves at different depths are determined by combining the hyperspectral data actually measured on site of the sample points with the evaluation model, so that comprehensive and accurate evaluation of the carbon reserves of the whole ecological system is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of soil carbon storage estimation, and in particular to a mangrove wetland soil carbon storage assessment method, device, equipment and storage medium. Background Art

[0002] With the intensification of global climate change, the monitoring and assessment of carbon storage has become an important research direction in the fields of environmental science and ecological protection. Traditional methods for studying mangrove soil carbon storage mainly include ground sampling, laboratory analysis, and statistical model estimation. However, ground sampling and laboratory analysis are very time-consuming and labor-intensive, making them difficult to apply widely across large mangrove areas. Secondly, laboratory analysis usually requires the use of large amounts of chemical reagents, which may be harmful to the environment and human health. Finally, traditional statistical models are unable to effectively address the spatial heterogeneity of soil carbon storage, especially in complex wetland ecosystems such as mangroves, where the spatial variation characteristics of soil are obvious. A single statistical model often cannot meet the needs of accurate estimation.

[0003] In recent years, remote sensing technology has been increasingly used in the monitoring and assessment of soil carbon stocks due to its rapid, non-destructive, and wide-area coverage. Hyperspectral remote sensing can obtain spectral information of ground objects in different wavelengths. This spectral information can reflect the content characteristics of organic matter, minerals, and other components in the soil. By analyzing these spectral characteristics, the carbon content in the soil can be indirectly inferred. However, mangrove soils are clearly stratified. The surface soil is significantly affected by tides and vegetation cover, while the carbon content in the deeper soil layers is more stable and less susceptible to external interference. Traditional soil carbon stock estimation methods typically focus only on the surface soil and ignore the carbon storage in the deeper soil layers, resulting in low accuracy in assessing the overall carbon storage of mangrove ecosystems. Summary of the Invention

[0004] Based on this, it is necessary to provide a mangrove wetland soil carbon storage assessment method, device, equipment and storage medium to address the above technical problems, so as to solve at least one of the above technical problems.

[0005] The present invention provides a method for assessing carbon storage in mangrove wetland soil, comprising:

[0006] Acquiring remote sensing image data of the area to be evaluated and performing data preprocessing on the remote sensing image data;

[0007] Based on the data preprocessing results, the land cover type of the area to be assessed is classified to determine the sampling points within the area to be assessed;

[0008] Use a portable ground feature spectrometer to conduct on-site hyperspectral measurements of sampling points within the assessment area to obtain soil hyperspectral reflectance data at different depths at each sample point;

[0009] Based on the soil hyperspectral reflectance data, a soil density assessment model and an organic carbon content assessment model are used to perform a hierarchical assessment on different sampling points in the area to be assessed, to obtain the soil density and organic carbon content at different depths at each sampling point;

[0010] Based on the soil density and organic carbon content at different depths in each sampling point, the soil carbon storage at different depths in each sampling point was determined;

[0011] The soil carbon storage at different depths in each sampling point is spatially interpolated to obtain a carbon storage distribution map of the area to be assessed.

[0012] Optionally, according to a mangrove wetland soil carbon storage assessment method provided by the present invention, determining the soil carbon storage at different depths at each sampling point based on the soil density and organic carbon content at different depths at each sampling point includes:

[0013] For each layer depth at any sampling point:

[0014] determining a target volume based on the depth and a preset grid cell;

[0015] The soil carbon storage at the depth is determined based on the soil density, the organic carbon content, and the target volume.

[0016] Optionally, according to a mangrove wetland soil carbon storage assessment method provided by the present invention, based on the soil hyperspectral reflectance data, a soil density assessment model and an organic carbon content assessment model are used to perform a hierarchical assessment on different sampling points in the area to be assessed, to obtain the soil density and organic carbon content at different depths at each sampling point, including:

[0017] Preprocessing and feature extraction are performed on the soil hyperspectral reflectance data to obtain hyperspectral reflectance feature data;

[0018] Inputting the hyperspectral reflectance characteristic data into the soil density assessment model to assess the soil density at different depths in the sampling point, thereby obtaining the soil density at different depths in the sampling point;

[0019] The hyperspectral reflectance characteristic data is input into the organic carbon content evaluation model to evaluate the organic carbon content at different depths in the sampling point to obtain the organic carbon content at different depths in the sampling point.

[0020] Optionally, according to a mangrove wetland soil carbon storage assessment method provided by the present invention, the spatial interpolation processing of the soil carbon storage at different depths in each sampling point to obtain the carbon storage distribution map of the area to be assessed includes:

[0021] Determining the land cover type of the area to be assessed based on the remote sensing image data;

[0022] Determine the geospatial coordinates of each sampling point;

[0023] The land cover type of the area to be assessed, the geographic spatial coordinates of each sampling point, and the soil carbon storage at different depths are subjected to three-dimensional spatial interpolation processing to obtain a carbon storage distribution map of the area to be assessed.

[0024] Optionally, according to a mangrove wetland soil carbon storage assessment method provided by the present invention, the soil density assessment model and the organic carbon content assessment model are trained based on the following steps:

[0025] Acquire remote sensing image samples of the target area and classify the land cover types of the target area to determine test sampling points according to different land cover types;

[0026] Determine the soil density and organic carbon content of soil samples at different sampling depths at each test sampling point under different land cover types;

[0027] Preprocessing and feature extraction are performed on soil hyperspectral reflectance samples at different sampling depths measured by a portable ground feature spectrometer to obtain hyperspectral reflectance feature samples;

[0028] For any of the test sampling points:

[0029] Training the soil density assessment model based on the hyperspectral reflectance characteristic samples of the test sampling points and the soil densities corresponding to different sampling depths;

[0030] The organic carbon content assessment model is trained based on the hyperspectral reflectance characteristic samples of the test sampling points and the organic carbon contents corresponding to different sampling depths.

[0031] Optionally, according to a mangrove wetland soil carbon storage assessment method provided by the present invention, determining the soil density and organic carbon content of soil samples corresponding to different sampling depths at each test sampling point under different land cover types includes:

[0032] For any of the soil samples:

[0033] Drying the soil sample until the mass of the soil sample no longer changes;

[0034] determining the soil mass of the soil sample after drying;

[0035] determining a soil density of the soil sample based on the mass of the dried soil and a preset volume of the trepanation drill;

[0036] The air-dried soil sample is mixed with an excess of potassium dichromate solution and concentrated sulfuric acid. After sufficient reaction, the soil sample is titrated with a standard ferrous sulfate solution to calculate the organic carbon content in the soil sample.

[0037] Optionally, according to a mangrove wetland soil carbon storage assessment method provided by the present invention, the soil density assessment model is trained based on the hyperspectral reflectance characteristic samples of the test sampling points and the soil densities corresponding to different sampling depths, including:

[0038] Inputting the hyperspectral reflectance characteristic sample, the sampling depth and the corresponding soil density into a soil density assessment model, and obtaining a soil density assessment result corresponding to the sampling depth output by the soil density assessment model;

[0039] Determining a model loss value based on the soil density corresponding to the sampling depth and the soil density assessment result;

[0040] The soil density estimation model is trained based on the model loss value.

[0041] The present invention also provides a mangrove wetland soil carbon storage assessment device, comprising:

[0042] An acquisition module is used to acquire remote sensing image data of the area to be evaluated and perform data preprocessing on the remote sensing image data;

[0043] The classification module is used to classify the land cover types of the area to be assessed based on the data preprocessing results and determine the sampling points within the area to be assessed;

[0044] Soil hyperspectral reflectance measurement module, used to use a portable ground feature spectrometer to perform on-site hyperspectral measurements of sampling points within the assessment area, and obtain soil hyperspectral reflectance data at different depths at each sample point;

[0045] An evaluation module is used to perform a hierarchical evaluation of different sampling points in the area to be evaluated based on the soil hyperspectral reflectance data using a soil density evaluation model and an organic carbon content evaluation model, respectively, to obtain soil density and organic carbon content at different depths at each sampling point;

[0046] a determination module, for determining the soil carbon storage at different depths in each sampling point based on the soil density and organic carbon content at different depths in each sampling point;

[0047] The interpolation module is used to perform spatial interpolation processing on the soil carbon storage at different depths in each sampling point to obtain a carbon storage distribution map of the area to be assessed.

[0048] The present invention also provides a portable ground feature spectrometer, which is used to measure soil hyperspectral reflectance data, and the soil hyperspectral reflectance data is used to implement the above-mentioned mangrove wetland soil carbon storage assessment method.

[0049] The present invention also provides a computer device comprising a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the above-mentioned mangrove wetland soil carbon storage assessment method is implemented.

[0050] The present invention also provides one or more readable storage media storing computer-readable instructions, which, when executed by a processor, implement the above-mentioned mangrove wetland soil carbon storage assessment method.

[0051] The above-mentioned mangrove wetland soil carbon stock assessment method, device, equipment and storage medium include: obtaining remote sensing image data of the area to be assessed and performing data preprocessing on the remote sensing image data; based on the data preprocessing results, classifying the land cover type of the area to be assessed to determine the sampling points in the area to be assessed; using a portable ground feature spectrometer to perform on-site hyperspectral measurement of the sampling points in the area to be assessed to obtain soil hyperspectral reflectance data at different depths of each sample point; based on the soil hyperspectral reflectance data, using a soil density assessment model and an organic carbon content assessment model to perform a stratified assessment on different sampling points in the area to be assessed to obtain soil density and organic carbon content at different depths in each sampling point; based on the soil density and organic carbon content at different depths in each sampling point, determining the soil carbon stock at different depths in each sampling point; performing spatial interpolation processing on the soil carbon stock at different depths in each sampling point to obtain a carbon stock distribution map of the area to be assessed. The present invention determines the sample points through remote sensing image data, combines the hyperspectral data measured on-site at the sample points, the soil density assessment model and the organic carbon content assessment model, and conducts a stratified assessment of different sampling points to obtain the soil density and organic carbon content at different depths in each sampling point. This can effectively capture the spatial heterogeneity of carbon content in the soil and improve the accuracy of the estimation. Then, based on the soil density and organic carbon content at different depths, the soil carbon storage at different depths in the mangrove ecosystem is calculated, thereby generating a carbon storage distribution map of the mangrove ecosystem and realizing a comprehensive and accurate assessment of the carbon storage of the entire ecosystem. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0053] Figure 1 This is a flow chart of a method for assessing soil carbon storage in mangrove wetlands according to an embodiment of the present invention;

[0054] Figure 2 This is a schematic structural diagram of a mangrove wetland soil carbon storage assessment device according to one embodiment of the present invention;

[0055] Figure 3 FIG. 1 is a schematic diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0057] The terms used in one or more embodiments of the present invention are for the purpose of describing specific embodiments only and are not intended to limit one or more embodiments of the present invention. The singular forms "a", "the" and "the" used in one or more embodiments of the present invention are also intended to include plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used in one or more embodiments of the present invention refers to and includes any or all possible combinations of one or more associated listed items.

[0058] In one embodiment, specifically, Figure 1 As shown, Figure 1 FIG1 is a flow chart of a method for assessing carbon storage in mangrove wetland soil according to an embodiment of the present invention. The method provides a method for assessing carbon storage in mangrove wetland soil according to an embodiment of the present invention, comprising the following steps:

[0059] Step S11, obtaining remote sensing image data of the area to be evaluated and performing data preprocessing on the remote sensing image data;

[0060] It should be noted that the areas to be assessed can be mangrove wetlands in various regions. High-resolution multispectral imagery data is acquired as a remote sensing data source to ensure high quality and consistency. In one specific example, image acquisition requires cloud cover below 10%, and is performed during low tide or similar tidal conditions to reduce tidal interference with surface features and ensure that the image data accurately reflects the distribution characteristics of surface vegetation. To effectively eliminate sensor systematic errors and atmospheric effects, the image data can be preprocessed through radiometric calibration, atmospheric correction, and geometric correction. For example, ENVI software can be used to perform radiometric calibration on the image data, converting the digital magnitude (DN) of the image data into radiance or reflectance to eliminate sensor effects and ensure physical comparability of the images. The FLAASH (Fast Line-of-sight Atmospheric Analysis of Spectral Hypercubes) module is used to perform atmospheric correction on the image data to eliminate the effects of scattering and absorption by atmospheric molecules and aerosols, making the radiance values ​​in the image more closely resemble the actual surface conditions. In terms of geometric correction, known landmarks or high-precision georeferenced images are selected as references, and combined with high-precision digital elevation models (DEMs), the images are geometrically corrected to ensure the spatial position consistency between different images.

[0061] Step S12: Based on the data preprocessing results, the land cover type of the area to be evaluated is classified to determine the sampling points within the area to be evaluated;

[0062] Specifically, based on the data after data preprocessing, use an appropriate classification algorithm (such as supervised classification or unsupervised classification) to classify land cover types. By extracting spectral features (such as vegetation index, soil index, etc.) from hyperspectral data, different land feature types (such as mangroves, wetlands, grasslands, etc.) are identified. After the classification is completed, the accuracy of the classification results is verified to ensure that the classification accuracy meets the requirements, and it is verified through ground measured data or calibration points. Then, based on the land cover classification results and environmental characteristics, the spectral features extracted from the image data (such as vegetation index, soil index, etc.) and spatial analysis methods (such as cluster analysis) are used to select areas with obvious spectral characteristics and uniform spatial distribution as sampling points.

[0063] Step S13, using a portable ground feature spectrometer to perform on-site hyperspectral measurement of sampling points within the area to be assessed, and obtaining soil hyperspectral reflectance data at different depths at each sampling point;

[0064] Specifically, sampling is performed on selected sampling points, and hyperspectral reflectance measurements are performed on-site using a portable ground feature spectrometer to obtain soil hyperspectral reflectance data at different depths at each sampling point.

[0065] Step S14, based on the soil hyperspectral reflectance data, using the soil density assessment model and the organic carbon content assessment model to perform a hierarchical assessment on different sampling points in the assessment area, to obtain the soil density and organic carbon content at different depths at each sampling point;

[0066] It should be noted that in order to enable the soil density assessment model and the organic carbon content assessment model to assess soil information at different depths, the model needs to be able to learn and capture the impact of depth on soil density. Therefore, during the model training process, the depth of the soil sample is input into the model as an additional feature for training to help the model better understand and predict the changes in soil density at different depths. The soil density assessment model is obtained by training based on remote sensing image samples and soil density of soil samples corresponding to different sampling depths in each test sampling point measured experimentally. The organic carbon content assessment model is obtained by training based on remote sensing image samples and organic carbon content of soil samples corresponding to different sampling depths in each test sampling point measured experimentally. The specific training process of the model is specifically described in the following embodiments and will not be repeated here.

[0067] Specifically, the soil hyperspectral reflectance data is preprocessed. Optionally, the spectral data corresponding to the soil hyperspectral reflectance data has a wavelength range of 350-2500 nm. Since the signal-to-noise ratio of the 350-400 nm and 2450-2500 nm bands is low and the data quality is poor, these two bands are removed, and the data from 400-2450 nm is retained for subsequent analysis. During the processing, the spectrum is smoothed, for example, using a Savitzky-Golay filter with a filter window size of 11 to reduce high-frequency noise in the spectral curve and make the spectral curve smoother.

[0068] After data preprocessing, feature extraction is performed on the hyperspectral data to obtain remote sensing feature data. For example, principal component analysis (PCA) can be used to reduce the dimensionality of hyperspectral data. PCA extracts the principal components from hyperspectral data through linear transformations, reducing the data's dimensionality while preserving the majority of the spectral information. Optionally, PCA can be used to extract the first 10 principal components as feature inputs, reducing feature redundancy and model complexity, thereby improving the efficiency and accuracy of subsequent modeling.

[0069] Furthermore, the soil hyperspectral reflectance characteristic data was input into the soil density assessment model to evaluate the soil density at different depths within the sampling point, and the soil density at different depths within the sampling point was obtained. In addition, the hyperspectral reflectance characteristic data was input into the organic carbon content assessment model to evaluate the organic carbon content at different depths within the sampling point, and the organic carbon content at different depths within the sampling point was obtained. The algorithm model was used to estimate the soil density and organic carbon content at different soil depths.

[0070] Step S15, determining the soil carbon storage at different depths in each sampling point based on the soil density and organic carbon content at different depths in each sampling point;

[0071] Specifically, the following steps are performed for each depth of any sampling point:

[0072] The area to be assessed is divided into several grid cells according to a preset grid size. The target volume is determined based on the depth corresponding to the sampling point and the preset grid cell. The soil carbon storage at that depth is then determined based on the soil density, organic carbon content, and target volume. The soil carbon storage is calculated as follows:

[0073] C=ρ×SOC×V

[0074] Where C is carbon storage (kg), ρ is soil density (kg / m 3 ), SOC is the organic carbon content (%), V is the target volume (m 3 For each sampling point, the soil carbon storage at each depth in the sampling point was accumulated to obtain the total carbon storage of the sampling point.

[0075] Step S16: Perform spatial interpolation processing on the soil carbon storage at different depths in each sampling point to obtain a carbon storage distribution map of the area to be assessed.

[0076] It should be noted that to comprehensively assess the carbon storage distribution of mangrove ecosystems, the estimated soil carbon storage data for different land cover types and at different depths were integrated through three-dimensional spatial interpolation. Specifically, based on remote sensing imagery, the land cover type of the assessed area was determined; the geospatial coordinates of each sampling point were determined, where the geospatial coordinates include longitude, latitude, and depth. Three-dimensional spatial interpolation was performed on the land cover type of the assessed area, the geospatial coordinates of each sampling point, and the soil carbon storage at different depths to obtain a carbon storage distribution map for the assessed area.

[0077] The generated carbon storage distribution map was visualized using three-dimensional analysis tools such as ArcGIS Pro. Using various visualization methods, such as cross-sectional and volumetric views, the changes in carbon storage at different depths and in different regions within the mangrove ecosystem were displayed, providing an intuitive understanding of the spatial distribution of carbon storage within the mangrove ecosystem. Furthermore, spatial statistical tools were used to analyze carbon storage distribution, including calculating statistical quantities such as the mean and variance of carbon storage within each region to assess the carbon sink function of the mangrove ecosystem. Spatial comparisons of carbon storage across regions can identify areas with high and low carbon storage, providing a scientific basis for decision-making in ecological conservation and management.

[0078] The embodiment of the present invention uses the above-mentioned scheme to perform a hierarchical evaluation of remote sensing image data from different sampling points by utilizing a soil density assessment model and an organic carbon content assessment model, thereby obtaining soil density and organic carbon content at different depths in each sampling point. This can effectively capture the spatial heterogeneity of carbon content in the soil and improve the accuracy of the estimation. Furthermore, based on the soil density and organic carbon content at different depths, the soil carbon storage at different depths in the mangrove ecosystem is calculated, thereby generating a carbon storage distribution map of the mangrove ecosystem based on the soil carbon storage at different depths, thereby achieving a comprehensive and accurate assessment of the carbon storage of the entire ecosystem.

[0079] In one embodiment of the present invention, the soil density estimation model and the organic carbon content estimation model are trained based on the following steps:

[0080] Acquire remote sensing image samples of the target area and classify the land cover types of the target area to determine the test sampling points according to different land cover types; determine the soil density and organic carbon content of soil samples corresponding to different sampling depths in each test sampling point under different land cover types; preprocess and extract features of the soil hyperspectral reflectance samples at different sampling depths measured by a portable land feature spectrometer to obtain hyperspectral reflectance feature samples; for any test sampling point: train a soil density assessment model based on the remote sensing feature samples of the test sampling point and the soil density corresponding to different sampling depths; train an organic carbon content assessment model based on the remote sensing feature samples of the test sampling point and the organic carbon content corresponding to different sampling depths.

[0081] It should be noted that the model architecture of the soil density and organic carbon content assessment models includes an input layer, three hidden layers, and an output layer. The hidden layers contain 50, 30, and 10 neurons, respectively. The hidden layers use the tansig activation function, and the output layer uses a linear activation function.

[0082] Specifically, remote sensing image samples of the target area are obtained, and then the land cover type of the target area is classified, wherein the land cover classification is mainly to identify mangrove vegetation and other land features (such as bare land, water bodies, etc.) so as to sample soil in different types of areas and ensure the comprehensiveness and representativeness of sample collection. In one embodiment, a supervised classification method is used to classify the land cover of remote sensing image samples. Optionally, a classification model is trained in combination with prior field sample point data. The features used include the original band data of the multispectral image and various vegetation indices (such as NDVI), water body indices (such as NDWI) and image texture participation extracted therefrom. These indices can enhance the spectral differences and spatial texture characteristics between land feature types and improve classification accuracy.

[0083] Furthermore, based on land cover type, test sampling points were identified within representative areas, such as mangrove vegetation, bare land, and surrounding water bodies. The selection of sampling points should consider spatial uniformity and representativeness to ensure that the collected soil samples fully reflect the soil characteristics underlying the different land features within the entire study area. For example, sampling locations were designed using the ArcGIS platform to ensure that a sufficient number of sampling points were included for each land cover type. During sampling, soil was collected in layers using a circular drill. For example, ten layers of soil were sampled at each test sampling point, with each layer sampling at a depth of 10 cm. Sampling standards were strictly adhered to during collection to avoid damage to the soil structure or sample contamination due to human factors. Before and after each sampling session, sampling tools were thoroughly cleaned with anhydrous alcohol to prevent cross-contamination. After sampling, soil samples were placed in sealed bags and labeled with the sampling number, sampling point, sampling depth, date, and specific location (including longitude, latitude, and altitude) of each test sampling point. This information was then entered into a database for subsequent analysis. Following sampling, hyperspectral reflectance measurements of the soil at each sampling point were performed on-site using a portable surface spectrometer. This step can obtain detailed hyperspectral information of the soil, providing important data support for subsequent analysis of soil organic carbon content, soil density, etc.

[0084] Furthermore, the experiment measured the soil density and organic carbon content of soil samples at different sampling depths at each test sampling point under different land cover types. The specific process is as follows: For any soil sample: weigh the wet weight of each layer of soil sample and record it. Then, dry the soil sample, for example, by placing it in a 60°C oven for 48 hours until the sample mass no longer changes. Determine the mass of the dried soil sample; then, based on the dried soil mass and the preset volume of the auger, determine the soil density of the soil sample. Calculate the soil density using the following formula:

[0085]

[0086] Among them, m 干 is the mass of soil after drying (g), V 已知 is the known volume of the trepan drill (cm 3 ). ρ is the soil density.

[0087] In addition, the air-dried soil samples were mixed with excess potassium dichromate solution and concentrated sulfuric acid. After sufficient reaction, they were titrated with a standard ferrous sulfate solution to calculate the organic carbon content in the soil samples.

[0088] Furthermore, the soil hyperspectral reflectance samples are preprocessed and feature extracted to obtain hyperspectral reflectance feature samples. The remote sensing image preprocessing and feature extraction process is similar to the preprocessing and feature extraction process in step S12 above and will not be repeated here. Furthermore, for any test sampling point: a soil density assessment model is trained based on the remote sensing feature samples at the test sampling point and the soil density corresponding to different sampling depths; and an organic carbon content assessment model is trained based on the remote sensing feature samples at the test sampling point and the organic carbon content corresponding to different sampling depths.

[0089] It should be noted that soil density may have different distribution characteristics at different depths, while hyperspectral data primarily reflects spectral information of the surface layer (or the entire soil). Hyperspectral data itself cannot directly identify soil depth information. Therefore, during model training, in order to enable the MLP model to understand this sampling depth information, in one embodiment, sampling depth (e.g., soil depth from the surface layer to a deeper layer) is used as an input variable to help the model understand the changing trend of soil density. Sampling depth can be used as a numerical feature along with remote sensing feature samples to input into the soil density assessment model. The soil density assessment model outputs a soil density assessment result corresponding to the sampling depth; then, based on the soil density corresponding to the sampling depth and the soil density assessment result, a model loss value is determined; and based on the model loss value, the soil density assessment model is trained. The training process of the organic carbon content assessment model is similar to that of the soil density assessment model described above and will not be repeated here. Thus, the soil density assessment model and the organic carbon content assessment model can be used to perform a hierarchical assessment of remote sensing image data from different sampling points, obtaining soil density and organic carbon content at different depths at each sampling point.

[0090] The embodiment of the present invention uses the above-mentioned scheme to train a soil density assessment model in combination with soil hyperspectral reflectance feature samples and soil density corresponding to different sampling depths; and trains an organic carbon content assessment model based on hyperspectral reflectance feature samples and organic carbon content corresponding to different sampling depths. Thus, the soil density assessment model and the organic carbon content assessment model can be used to perform a hierarchical assessment of remote sensing image data of different sampling points, thereby calculating the soil carbon storage at different depths in the sampling points, and realizing a comprehensive and accurate assessment of the carbon storage of the entire ecosystem.

[0091] In one embodiment of the present invention, spatial interpolation processing is performed on the soil carbon storage at different depths in each sampling point to obtain a carbon storage distribution map of the area to be assessed, including:

[0092] Based on remote sensing image data, the land cover type of the area to be assessed is determined; the geographic spatial coordinates of each sampling point are determined; and three-dimensional spatial interpolation processing is performed on the land cover type of the area to be assessed, the geographic spatial coordinates of each sampling point, and the soil carbon storage at different depths to obtain a carbon storage distribution map of the area to be assessed.

[0093] Specifically, based on remote sensing image data, the land cover type of the area to be evaluated is determined, wherein the land cover classification process is specifically described in the above embodiment and will not be repeated here. In addition, it is necessary to determine the geographic spatial coordinates corresponding to the longitude, latitude and depth of each sampling point in the area to be evaluated. During the interpolation process, the raster data of the land cover type is resampled to the same resolution as the interpolation result. The specific spatial interpolation processing process is as follows: the spatial relationship between each sampling point is evaluated by spatial autocorrelation analysis (Moran's I index) to determine the spatial aggregation or dispersion of the data. The higher the Moran's I index value, the stronger the data is in spatial positive correlation, that is, the spatial variation is relatively smooth, and a Gaussian model is used for interpolation; for areas with lower Moran's I index values, it indicates that the data has higher spatial heterogeneity and drastic variations. A spherical model is used to capture these complex variation characteristics to more accurately reflect the spatial autocorrelation and variation patterns of carbon storage.

[0094] Determine key parameters for interpolation, such as the search radius and resolution. For example, based on the density of sampling points and the spatial variation of carbon stocks, and ensuring that each interpolation point has sufficient neighbors, set the search radius between 100 and 500 meters to ensure interpolation accuracy and computational efficiency. Based on the scale and accuracy requirements of the study area, set the interpolation resolution to 10 meters to ensure that the interpolation results display sufficient detail and maintain computational rationality.

[0095] Furthermore, spatial interpolation of soil carbon storage at different depths within the mangrove ecosystem was performed. Based on the selected semivariogram model and set parameters, ArcGIS Pro calculated carbon storage estimates at each location, generating a carbon storage distribution map for the assessed area. During the interpolation process, supervised interpolation was limited to the area covered by the mangrove ecosystem to prevent the interpolation results from extending beyond the study area and introducing unnecessary errors and uncertainties.

[0096] Through the above scheme, the embodiment of the present invention generates a carbon storage distribution map of the mangrove ecosystem based on the soil carbon storage at different depths in each sampling point, thereby achieving a comprehensive and accurate assessment of the carbon storage of the entire ecosystem.

[0097] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0098] In one embodiment, a mangrove wetland soil carbon storage assessment device is provided, which corresponds to the mangrove wetland soil carbon storage assessment method in the above embodiment. Figure 2 As shown, Figure 2 FIG. 1 is a schematic structural diagram of a mangrove wetland soil carbon storage assessment device according to an embodiment of the present invention. The mangrove wetland soil carbon storage assessment device comprises:

[0099] An acquisition module 21 is used to acquire remote sensing image data of the area to be evaluated and perform data preprocessing on the remote sensing image data;

[0100] A classification module 22 is used to classify the land cover type of the area to be evaluated based on the data preprocessing results and determine the sampling points in the area to be evaluated;

[0101] The soil hyperspectral reflectance measurement module 23 is used to perform on-site hyperspectral measurement of sampling points in the assessment area using a portable ground feature spectrometer to obtain soil hyperspectral reflectance data at different depths at each sampling point;

[0102] An evaluation module 24 is configured to perform a hierarchical evaluation of different sampling points in the evaluation area using a soil density evaluation model and an organic carbon content evaluation model based on the soil hyperspectral reflectance data, thereby obtaining soil density and organic carbon content at different depths at each sampling point.

[0103] a determination module 25 for determining the soil carbon storage at different depths in each sampling point based on the soil density and organic carbon content at different depths in each sampling point;

[0104] The interpolation module 26 is used to perform spatial interpolation processing on the soil carbon storage at different depths in each sampling point to obtain a carbon storage distribution map of the area to be assessed.

[0105] The determination module 25 is further configured to:

[0106] For each layer depth at any sampling point:

[0107] Determine the target volume based on the depth and pre-set grid cells;

[0108] Determine soil carbon storage at depth based on soil density, organic carbon content, and target volume.

[0109] The evaluation module 24 is also used to:

[0110] Preprocess and extract features of remote sensing image data to obtain remote sensing feature data;

[0111] The remote sensing feature data are input into the soil density assessment model to assess the soil density at different depths in the sampling point, thereby obtaining the soil density at different depths in the sampling point;

[0112] The remote sensing characteristic data are input into the organic carbon content assessment model to evaluate the organic carbon content at different depths in the sampling points and obtain the organic carbon content at different depths in the sampling points.

[0113] The interpolation module 26 is further configured to:

[0114] Determine the land cover type of the area to be assessed based on remote sensing image data;

[0115] Determine the geospatial coordinates of each sampling point;

[0116] The land cover type of the area to be assessed, the geographic coordinates of each sampling point, and the soil carbon storage at different depths were spatially interpolated to obtain a carbon storage distribution map of the area to be assessed.

[0117] The mangrove wetland soil carbon storage assessment device also includes:

[0118] Obtain remote sensing image samples of the target area and classify the land cover types of the target area to determine the test sampling points according to different land cover types;

[0119] Determine the soil density and organic carbon content of soil samples at different sampling depths at each test sampling point under different land cover types;

[0120] Preprocessing and feature extraction are performed on soil hyperspectral reflectance samples at different sampling depths measured by a portable ground feature spectrometer to obtain hyperspectral reflectance feature samples;

[0121] For any test sampling point:

[0122] A soil density assessment model is trained based on the hyperspectral reflectance characteristic samples of the test sampling points and the soil density corresponding to different sampling depths;

[0123] An organic carbon content assessment model is trained based on the hyperspectral reflectance characteristic samples of the test sampling points and the organic carbon content corresponding to different sampling depths.

[0124] The mangrove wetland soil carbon storage assessment device also includes:

[0125] For any soil sample:

[0126] Dry the soil sample until the mass of the soil sample no longer changes;

[0127] Determine soil quality after drying soil samples;

[0128] Determine the soil density of the soil sample based on the mass of the dried soil and the volume of the pre-set trepanation drill;

[0129] The air-dried soil sample was mixed with an excess of potassium dichromate solution and concentrated sulfuric acid. After sufficient reaction, it was titrated with a standard ferrous sulfate solution to calculate the organic carbon content in the soil sample.

[0130] The mangrove wetland soil carbon storage assessment device also includes:

[0131] Inputting remote sensing feature samples, sampling depths and their corresponding soil densities into a soil density assessment model, and obtaining a soil density assessment result corresponding to the sampling depth output by the soil density assessment model;

[0132] Determine the model loss value based on the soil density corresponding to the sampling depth and the soil density assessment results;

[0133] Based on the model loss value, a soil density estimation model is trained.

[0134] The specific definitions of the mangrove wetland soil carbon stock assessment device can be found in the definitions of the mangrove wetland soil carbon stock assessment method above and will not be repeated here. Each module in the aforementioned mangrove wetland soil carbon stock assessment device can be implemented in whole or in part through software, hardware, or a combination thereof. Each of these modules can be embedded in or independent of a processor within a computer device in hardware form, or stored in a computer device memory in software form, allowing the processor to call and execute the corresponding operations of each module.

[0135] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 3 As shown, Figure 3This is a schematic diagram of a computer device in one embodiment of the present invention. The computer device includes a processor, a memory, a network interface, and a database connected via a device bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a readable storage medium and an internal memory. The readable storage medium stores an operating device, computer-readable instructions, and a database. The internal memory provides an environment for the operation of the operating device and computer-readable instructions in the readable storage medium. The database of the computer device is used to store data related to the method for assessing carbon storage in mangrove wetland soils. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer-readable instructions are executed by the processor, a method for assessing carbon storage in mangrove wetland soils is implemented. The readable storage medium provided in this embodiment includes a non-volatile readable storage medium and a volatile readable storage medium.

[0136] In one embodiment, a computer device is provided. The computer device may be a terminal device, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, and a network interface connected via a device bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a readable storage medium. The readable storage medium stores computer-readable instructions. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer-readable instructions are executed by the processor, a method for assessing soil carbon storage in mangrove wetlands is implemented. The readable storage medium provided in this embodiment includes a non-volatile readable storage medium and a volatile readable storage medium.

[0137] In one embodiment, a portable ground feature spectrometer is provided, which is used to measure soil hyperspectral reflectance data. The soil hyperspectral reflectance data is used to implement the steps of the above-mentioned mangrove wetland soil carbon storage assessment method.

[0138] In one embodiment, a computer device is provided, comprising a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the steps of the above-mentioned method for assessing soil carbon storage in mangrove wetlands are implemented.

[0139] In one embodiment, a readable storage medium is provided, which stores computer-readable instructions. When the computer-readable instructions are executed by a processor, the steps of the above-mentioned mangrove wetland soil carbon storage assessment method are implemented. A person skilled in the art will understand that all or part of the processes in the above-mentioned embodiment method can be implemented by instructing related hardware through computer-readable instructions. The computer-readable instructions can be stored in a non-volatile readable storage medium or a volatile readable storage medium. When the computer-readable instructions are executed, they can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0140] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0141] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A method for assessing soil carbon storage in mangrove wetlands, characterized in that: include: Acquire remote sensing image data of the area to be evaluated, and perform data preprocessing on the remote sensing image data; Based on the data preprocessing results, the land cover type of the area to be evaluated is classified to determine the sampling points in the area to be evaluated; Use a portable ground spectrometer to conduct on-site hyperspectral measurements of sampling points in the assessment area to obtain soil hyperspectral reflectance data at different depths at each sample point; Based on the soil hyperspectral reflectance data, a soil density assessment model and an organic carbon content assessment model are used to perform a hierarchical assessment on different sampling points in the area to be assessed, so as to obtain soil density and organic carbon content at different depths in each sampling point; Based on the soil density and organic carbon content at different depths in each sampling point, the soil carbon storage at different depths in each sampling point was determined; The soil carbon storage at different depths in each sampling point is spatially interpolated to obtain a carbon storage distribution map of the area to be assessed.

2. The method for assessing soil carbon storage in mangrove wetlands according to claim 1, characterized in that: The method of determining the soil carbon storage at different depths in each sampling point based on the soil density and organic carbon content at different depths in each sampling point includes: For each layer depth at any sampling point: Determining a target volume based on the depth and a preset grid cell; The soil carbon stock at the depth is determined based on the soil density, the organic carbon content, and the target volume.

3. The method for assessing soil carbon storage in mangrove wetlands according to claim 1, characterized in that: Based on the soil hyperspectral reflectance data, a soil density assessment model and an organic carbon content assessment model are used to perform a hierarchical assessment on different sampling points in the area to be assessed, and the soil density and organic carbon content at different depths in each sampling point are obtained, including: Preprocessing and feature extraction are performed on the soil hyperspectral reflectance data to obtain hyperspectral reflectance feature data; Inputting the hyperspectral reflectance characteristic data into the soil density assessment model to assess the soil density at different depths in the sampling point, thereby obtaining the soil density at different depths in the sampling point; The hyperspectral reflectance characteristic data is input into the organic carbon content assessment model to assess the organic carbon content at different depths in the sampling point, thereby obtaining the organic carbon content at different depths in the sampling point.

4. The method for assessing soil carbon storage in mangrove wetlands according to claim 1, characterized in that: The spatial interpolation processing of the soil carbon storage at different depths in each sampling point is performed to obtain the carbon storage distribution map of the area to be assessed, including: Determining the land cover type of the area to be assessed based on the remote sensing image data; Determine the geospatial coordinates of each sampling point; The land cover type of the area to be assessed, the geographic spatial coordinates of each sampling point and the soil carbon storage at different depths are subjected to three-dimensional spatial interpolation processing to obtain a carbon storage distribution map of the area to be assessed.

5. The method for assessing soil carbon storage in mangrove wetlands according to claim 1, characterized in that: The soil density assessment model and the organic carbon content assessment model are trained based on the following steps: Acquire remote sensing image samples of the target area, and classify the land cover types of the target area to determine test sampling points according to different land cover types; Determine the soil density and organic carbon content of soil samples corresponding to different sampling depths in each test sampling point under different land cover types; Preprocess and extract features of soil hyperspectral reflectance samples at different sampling depths measured by a portable ground feature spectrometer to obtain hyperspectral reflectance feature samples; For any of the test sampling points: Training the soil density assessment model based on the hyperspectral reflectance characteristic samples of the test sampling points and the soil densities corresponding to different sampling depths; The organic carbon content assessment model is trained based on the high spectral reflectance characteristic samples of the test sampling points and the organic carbon contents corresponding to different sampling depths.

6. The method for assessing soil carbon storage in mangrove wetlands according to claim 5, characterized in that: The determining of the soil density and organic carbon content of soil samples corresponding to different sampling depths at each test sampling point under different land cover types includes: For any of the soil samples: Drying the soil sample until the mass of the soil sample no longer changes; determining the soil mass of the soil sample after drying; Determining the soil density of the soil sample based on the mass of the dried soil and the volume of a preset auger; The air-dried soil sample is mixed with an excess of potassium dichromate solution and concentrated sulfuric acid, and after sufficient reaction, it is titrated with a standard ferrous sulfate solution to calculate the organic carbon content in the soil sample.

7. The method for assessing soil carbon storage in mangrove wetlands according to claim 5, characterized in that: The soil density assessment model is trained based on the hyperspectral reflectance characteristic samples of the test sampling points and the soil densities corresponding to different sampling depths, including: Inputting the hyperspectral reflectance characteristic sample, the sampling depth and the corresponding soil density into a soil density assessment model, and obtaining a soil density assessment result corresponding to the sampling depth output by the soil density assessment model; Determining a model loss value based on the soil density corresponding to the sampling depth and a soil density assessment result; Based on the model loss value, the soil density assessment model is trained.

8. A mangrove wetland soil carbon storage assessment device, characterized in that: include: An acquisition module is used to acquire remote sensing image data of the area to be evaluated and perform data preprocessing on the remote sensing image data; A classification module is used to classify the land cover type of the area to be evaluated based on the data preprocessing results and determine the sampling points in the area to be evaluated; Soil hyperspectral reflectance measurement module, used to use a portable ground feature spectrometer to perform on-site hyperspectral measurement of sampling points in the assessment area, and obtain soil hyperspectral reflectance data at different depths of each sample point; An evaluation module is used to perform a hierarchical evaluation on different sampling points in the area to be evaluated based on the soil hyperspectral reflectance data using a soil density evaluation model and an organic carbon content evaluation model to obtain soil density and organic carbon content at different depths in each sampling point; A determination module, for determining the soil carbon storage at different depths in each sampling point based on the soil density and organic carbon content at different depths in each sampling point; The interpolation module is used to perform spatial interpolation processing on the soil carbon storage at different depths in each sampling point to obtain a carbon storage distribution map of the area to be evaluated.

9. A computer device comprising a memory, a processor, and computer-readable instructions stored in the memory and executed on the processor, characterized in that: When the processor executes the computer-readable instructions, the method for assessing soil carbon storage in mangrove wetlands as described in any one of claims 1 to 7 is implemented.

10. A readable storage medium having computer readable instructions stored thereon, characterized in that: When the computer-readable instructions are executed by a processor, the method for assessing soil carbon storage in mangrove wetlands as described in any one of claims 1 to 7 is implemented.

Images