Forest stand structure regulation and control method for improving carbon sequestration capacity of pine forest

Abstract

The invention discloses a forest stand structure regulation and control method for improving the carbon sequestration capacity of a pine forest, and particularly relates to the technical field of forestry ecological carbon sink regulation and control. Obtaining forest stand structure parameters and soil fertility parameters, and constructing a structure vector and a soil fertility vector; inputting the model into a carbon sink influence factor response model, and extracting a sensitive structure factor set C1; identifying grid units which do not meet a critical value based on C1, and delimiting a carbon sink structure bottleneck block P; performing regulation and control measures such as directional thinning, multi-layer building and structure intermediate cutting on the forest stand in the P block; dynamic evaluation is carried out through the change delta C of the carbon sink increase amount, and if the standard is not reached, the model weight is updated to form closed-loop optimization; the method realizes intelligent identification and accurate improvement of the pine forest carbon sequestration capability, and has the advantages of high operability, wide adaptability, remarkable carbon sequestration effect and the like.

Description

Technical Field

[0001] This invention relates to the field of forestry ecological carbon sequestration regulation technology, specifically to a method for regulating forest stand structure to improve the carbon sequestration capacity of pine forests. Background Technology

[0002] Pine forests, as a typical type of coniferous forest, are widely distributed in Northeast, Southwest, and the hilly areas of southern my country, and are an important component of forest carbon sinks. Due to their rapid growth, low wood density, and rapid biomass accumulation, pine forests have long been considered to have good carbon sequestration potential. However, recent field monitoring data show that although pine forests in some areas have high canopy closure, their carbon sequestration efficiency is low, with carbon sink per unit area far below the predictions of the national forestry carbon storage model. Furthermore, some degraded forests are even showing a trend of carbon source conversion, posing a potential obstacle to achieving regional forestry carbon neutrality goals.

[0003] Analysis revealed that current artificial management of pine forests mainly focuses on a single tree species and uniform density, neglecting the synergistic effect of different tree ages, diameter at breast height (DBH) structures, and forest spatial structures on carbon sequestration efficiency. Especially in areas with poor soil phosphorus and frequent climate disturbances (such as periodic droughts), mono-structure pine forests often experience canopy closure, intensified underground competition, and stagnant biomass growth in the later stages of growth, severely limiting the further enhancement of carbon sequestration potential. Summary of the Invention

[0004] The purpose of this invention is to provide a method for regulating forest stand structure to improve the carbon sequestration capacity of pine forests, thereby addressing the shortcomings of the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for regulating forest stand structure to improve the carbon sequestration capacity of pine forests, comprising:

[0006] S100. Obtain the stand structure parameters of the target pine forest plot, including diameter at breast height (DBH) distribution, tree height distribution, number of forest layers, average crown width and average spacing, and construct the stand structure feature vector V1.

[0007] S200. Soil carbon, nitrogen and phosphorus content was tested in the target plot to obtain forest land fertility parameter set V2, which includes soil organic carbon content, available phosphorus content, available potassium content and pH value.

[0008] S300. Input V1 and V2 into the preset pine forest carbon sink impact factor response model and output the carbon sink sensitive structural factor set C1.

[0009] S400. Based on C1, identify forest stand structural units in the target pine forest plot that do not meet the critical value, and determine the carbon sink structure bottleneck block P.

[0010] S500. Regulate the stand structure units within each P block, including adjusting the standard deviation of diameter at breast height (DBH) in areas with narrow DBH distribution through directional thinning; artificially introducing different-aged pine species or naturally regenerated species to form multi-layered forest structures in single-layer forest areas; and implementing thinning to reduce canopy overlap in stands where the ratio of average crown width to average spacing is less than a set threshold.

[0011] Preferably, the input of V1 and V2 into the preset pine forest carbon sequestration impact factor response model includes:

[0012] A carbon sink impact factor response model based on the random forest regression algorithm was constructed. The structural parameters in historical sample plot data and the measured carbon sink growth were used for training, and a variable importance ranking mechanism was established.

[0013] Using the obtained V1 and V2 as model inputs, the annual carbon sink growth per unit area of ​​the target sample plot is predicted, and the contribution of each structural parameter to the carbon sink change is calculated.

[0014] Carbon sink sensitive factors are extracted based on structural parameters whose contribution is greater than a preset threshold, forming a set C1, which includes the structural indicators that most significantly affect carbon sequestration efficiency and their critical value ranges.

[0015] Preferably, based on C1, identifying stand structure units in the target pine forest plot that do not meet the critical value and determining the carbon sink structure bottleneck block P includes:

[0016] Using the critical value range of each sensitive factor in C1 as the structural screening criterion, a multi-factor spatial filtering rule is constructed.

[0017] The target plot is divided into several regular grid cells, and the corresponding structural feature vector V1′ is recalculated for each cell;

[0018] By comparing V1′ with C1 one by one, grid cells that do not meet all critical interval conditions are identified.

[0019] Adjacent units that do not meet the conditions are clustered to form a continuous spatial block P, which serves as the bottleneck block of the carbon sink structure to be regulated.

[0020] Preferably, the step of comparing V1′ with C1 one by one to identify mesh cells that do not meet all critical interval conditions includes:

[0021] The structural feature sub-vector V1′ of each grid cell is mapped one-to-one with the critical value range of the sensitivity factor in C1 according to its dimension;

[0022] Set logical matching rules. When any dimension indicator does not fall into the corresponding critical interval, the unit is determined to be a structurally substandard unit.

[0023] Generate a structure matching matrix M, where a value of 0 indicates that the element does not meet the critical requirements, and a value of 1 indicates that it does.

[0024] Summarize the cell numbers of all cells with an M value of 0 and record their spatial coordinates as the target set for cluster analysis.

[0025] Preferably, the regulation of forest stand structural units within each P block includes:

[0026] For areas where the standard deviation of diameter at breast height (DBH) distribution is less than the set target value, targeted thinning is used to remove individuals with high DBH concentration, thereby expanding the DBH distribution towards greater diversity.

[0027] In plots identified as single-layer forest structures, construct multi-layer forest structures;

[0028] For stand areas where the ratio of average canopy width to average spacing is less than 0.65, thinning should be implemented to reduce densely overlapping canopies.

[0029] Preferably, after the adjustment is completed, V1 is collected again, and the annual carbon sequestration growth ΔC before and after the adjustment is compared. If ΔC does not reach the set increase threshold T, return to step S400 to update C1 and adjust again until ΔC≥T.

[0030] Preferably, after the forest stand structure regulation is completed, the carbon sequestration effect is verified and dynamically optimized, including:

[0031] After re-collecting the forest stand structure parameters after regulation, an updated structural feature vector DV1 was constructed and input into the carbon sink impact factor response model to calculate the annual carbon sink growth per unit area after regulation.

[0032] By comparing the increase in carbon sink before regulation with the result after regulation, the increase in carbon sequestration ΔC is obtained.

[0033] Set a threshold T for improving carbon fixation. When ΔC is less than T, retrain the importance weights of the sensitivity factors in C1.

[0034] Based on the updated C1 results, the bottleneck block P is re-identified and structural control is performed to form a closed-loop optimization cycle until ΔC reaches or exceeds the T value.

[0035] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0036] 1. This invention provides a method for regulating the structure of pine forest stands based on a carbon sink response model. This method can accurately identify key structural factors limiting carbon sequestration efficiency and, through zonal regulation, model feedback, and dynamic optimization mechanisms, achieve a systematic transformation of the forest stand structure from a "carbon sequestration bottleneck" to a "high-efficiency carbon sink" state. This method overcomes the limitations of traditional forestry management that relies solely on empirical thinning and static structural analysis.

[0037] 2. This invention achieves full-process quantitative control from factor selection and bottleneck identification to regulation assessment by constructing a set of carbon sink sensitive factors and a spatial structure screening mechanism, combined with a random forest regression algorithm and closed-loop feedback optimization path. It is adaptable to different forest ages and soil fertility backgrounds, demonstrating high versatility and promotional value. Simultaneously, this method can effectively increase the annual carbon sink growth per unit area, contributing to the achievement of regional carbon neutrality goals. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0039] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] For examples, please refer to Figure 1 As shown in this embodiment, a method for regulating stand structure to improve the carbon sequestration capacity of pine forests includes:

[0042] S100. Obtain the stand structure parameters of the target pine forest plot, including diameter at breast height (DBH) distribution, tree height distribution, number of forest layers, average crown width and average spacing, and construct the stand structure feature vector V1.

[0043] S200. Soil carbon, nitrogen and phosphorus content was tested in the target plot to obtain forest land fertility parameter set V2, which includes soil organic carbon content, available phosphorus content, available potassium content and pH value.

[0044] S300. Input V1 and V2 into the preset pine forest carbon sink impact factor response model and output the carbon sink sensitive structural factor set C1, where C1 contains the structural indicators that most significantly affect carbon sequestration efficiency and their critical value ranges.

[0045] S400. Based on C1, identify forest stand structural units in the target pine forest plot that do not meet the critical value, and determine the carbon sink structure bottleneck block P.

[0046] S500. Regulate the stand structure units within each P block, including adjusting the standard deviation of diameter at breast height (DBH) in areas with narrow DBH distribution through directional thinning; artificially introducing different-aged pine species or naturally regenerated species to form multi-layered forest structures in single-layer forest areas; and implementing thinning to reduce canopy overlap in stands where the ratio of average crown width to average spacing is less than a set threshold.

[0047] S600. After the adjustment is completed, V1 is collected again, and the annual carbon sequestration growth ΔC before and after the adjustment is compared. If ΔC does not reach the set increase threshold T, return to step S400 to update C1 and adjust again until ΔC≥T.

[0048] In a preferred embodiment of the present invention, step S100 is first performed: obtaining the stand structure parameters of the target pine forest sample plot to comprehensively characterize the spatial structure of the stand and the tree growth distribution characteristics, providing basic data support for subsequent identification and regulation of carbon sink structure bottlenecks. Specifically, the stand structure parameters include the following five core indicators:

[0049] Diameter at breast height (DBH): By setting up standard fixed quadrats (e.g., 20m×20m) in the sample plot, the diameter of the trunk at all heights at breast height (1.3m) in the quadrats is measured using a diameter measuring ruler, and a histogram of the frequency distribution of DBH is plotted.

[0050] Tree height distribution: The total height of each sample tree was measured using a laser altimeter or high pole measurement method, and the distribution range and standard deviation of tree height were analyzed to determine the vertical layer differences of the forest stand.

[0051] Forest layer number: Based on visual inspection and standing tree stratification, forest layers (such as main layer, sublayer, understory, etc.) are divided according to the difference in canopy height and the degree of shading, and the number of forest layers and their biomass distribution are recorded.

[0052] Average crown width: The crown projection method is used to measure the crown width length of the sample trees in the main axis direction and vertical direction, calculate the average value, and then take the average value in the sampling plot to characterize the crown coverage capacity of the forest stand;

[0053] Average spacing: Using GPS or a total station, the spatial coordinates of the sample trees are located, the Euclidean distance between any two trees is calculated, and the average value is obtained to reflect the stand density and spatial configuration.

[0054] After the above data collection was completed, the structural parameters were dimensionless by a standardization method (such as z-score standardization), and a structural feature vector V1 = [v1, v2, v3, v4, v5] was constructed, where: v1 represents the standard deviation or coefficient of variation of the diameter at breast height distribution; v2 represents the average value or height difference of the tree height distribution; v3 represents the number of forest layers; v4 represents the average crown width; and v5 represents the average tree spacing or its reciprocal (i.e., density index).

[0055] After completing the construction of the forest stand structure feature vector V1, proceed to step S200: conduct field testing of the content of key nutrients such as soil carbon, nitrogen, and phosphorus in the target pine forest plots, aiming to comprehensively assess the soil fertility status of the forest land and its potential impact on carbon sequestration capacity.

[0056] The soil testing described in this step mainly targets the mineral soil layer within the top 0-20cm of the forest land. Sampling points are set using either the "plum blossom method" or the "S-shaped distribution method" to ensure representativeness and uniformity. Five soil sampling points are set up in each quadrat. After sampling, the samples are mixed to prepare a composite sample, which will be used as the test sample. The test indicators include:

[0057] Soil organic carbon content (SOC): determined by potassium dichromate-sulfuric acid external heating oxidation method or direct determination by elemental analyzer, in g / kg, representing the carbon storage basis of forest soil;

[0058] Available phosphorus content: Phosphorus was extracted using the Olsen method and then determined by colorimetric method. The unit is mg / kg, which reflects the plant's ability to absorb phosphorus.

[0059] Available potassium content: determined by flame photometry after extraction with 1 mol / L ammonium acetate, in mg / kg. It is a key element for regulating cell turgor pressure during pine growth.

[0060] Soil pH value: The soil pH value was measured using a 1:2.5 water-to-soil suspension method and a pH meter to reflect the soil acidity and alkalinity environment, which has a significant impact on the nutrient absorption capacity of pine tree roots.

[0061] After completing the above tests, the soil index data were standardized to construct a forest land fertility parameter set V. 2＝ [c1,c2,c3,c4], where: c1 represents the unit soil organic carbon content; c2 represents the unit available phosphorus content; c3 represents the unit available potassium content; and c4 represents the soil pH value (usually between 4.5 and 6.5 is the suitable growth range for pine trees).

[0062] After completing the construction of the forest stand structure feature vector V1 and the forest land fertility parameter set V2, proceed to step S300: input V1 and V2 together into the preset pine forest carbon sink influencing factor response model to explore the forest stand structure factors that mainly affect carbon sink efficiency, and finally output the carbon sink sensitive structure factor set C1.

[0063] The carbon sink impact factor response model is a multi-factor nonlinear prediction model built based on the Random Forest Regression algorithm. It features strong anti-overfitting ability and high variable importance identification capability, making it suitable for modeling the response of complex interactive variables to carbon sinks in forest ecosystems. The model construction steps include:

[0064] A representative historical dataset of pine forest plots was collected. Each sample in the dataset contains structural parameters (i.e., the 5 types of features in V1), soil parameters (i.e., the 4 types of indicators in V2), and the corresponding annual carbon sink growth per unit area (in tons of carbon per hectare per year).

[0065] Using structural parameters and soil parameters as independent variables and carbon sequestration growth as dependent variable, a random forest regression model was constructed using the Scikit-learn machine learning library in Python. The number of trees was set to 100-300, and 10-fold cross-validation was used to optimize the model stability.

[0066] Using the model's built-in Gini importance scoring mechanism, the importance of each input variable is ranked to obtain its contribution score to carbon sequestration changes, and the output variable importance vector W = [w1, w2, ..., w n ], where w i This represents the contribution score of the i-th factor, ranging from 0 to 1, where n is the total number of factors.

[0067] Inputting V1 and V2 corresponding to the target plot into the model will output the predicted annual carbon sequestration growth per unit area of ​​the plot, while also returning the relative contribution of each input variable. For example, if the contribution of the diameter at breast height (DBH) distribution coefficient of variation to the prediction result is 0.22, while the contribution of soil available phosphorus content is 0.07, then the former is more sensitive to the carbon sequestration response.

[0068] To screen the structural factors that most significantly affect carbon sink efficiency, this invention sets a contribution threshold of 0.15, meaning that only structural parameters with a contribution score greater than 0.15 are retained, forming a set of carbon sink sensitive structural factors C1. This set includes not only the names of key factors but also the corresponding critical value ranges obtained from model analysis, such as "the carbon sink growth efficiency is highest when the coefficient of variation of diameter at breast height distribution is between 0.25 and 0.35".

[0069] The resulting C1 set serves as the input basis for subsequent structural bottleneck identification steps, used to determine which regions do not meet the characteristics of a good structure, thereby identifying target regulation locations and improving forest stand carbon sequestration efficiency.

[0070] After obtaining the set of carbon sink sensitive structural factors C1, proceed to step S400. Based on the indicators in C1 and their critical value ranges, identify forest stand structural units in the target pine forest plot that do not meet the carbon sequestration optimization conditions, and finally delineate the carbon sink structural bottleneck block P as the target for subsequent regulation.

[0071] First, the sensitive factors extracted from C1 (such as diameter at breast height (DBH) coefficient of variation, average crown width, and number of canopies) and their corresponding critical intervals are used as the judgment criteria to construct a multi-factor spatial filtering rule set F. Each rule R k Given a sensitivity factor i and its ideal interval [min i ,max i The composition satisfies the condition that the minimum value is 0. i ≤v i ≤max i .

[0072] Subsequently, the target pine forest plot was divided into several regular grid units, with a recommended grid side length of 10 meters, forming a spatial analysis grid G ​​= {g1, g2, ..., g n Each grid cell g j As the basic unit of analysis, the local structural feature vector V is recalculated using the observation data of all sample trees within it. 1′j To ensure the accuracy of subsequent comparisons, the dimensions are kept consistent with the original vector V1.

[0073] Next, the V1′ vector of each unit is compared dimension-by-dimensionally with each sensitivity factor interval defined in C1 to establish the index matching logic: if v in any dimension i ′ Not fell into [min i ,max i If the range is within a certain range, the grid cell is determined to be a substandard cell; otherwise, it is a compliant cell.

[0074] To achieve rapid batch recognition, a structure matching matrix M is constructed, with dimensions of N rows × 1 column (where N is the total number of grid cells). Each element m in the matrix... j Corresponding to a grid cell g j If g j If all sensitivity factor thresholds are met, a value of 1 is assigned; otherwise, a value of 0 is assigned. That is: m j =1 indicates that the unit structure meets the standard; m j =0 indicates that the unit structure does not meet the standard.

[0075] All m jThe unit numbers with a value of 0 are summarized, and their corresponding two-dimensional spatial coordinates are extracted to form a set of structurally substandard units S = {ga, gb, ..., gk}, which serves as the input target for subsequent spatial clustering.

[0076] Finally, the grid cells in set S are subjected to adjacency clustering. The 8-neighborhood method or the DBSCAN density clustering algorithm is used to aggregate spatially adjacent or densely distributed non-compliant cells into continuous spatial blocks. Each block is labeled P. j This indicates a bottleneck area with carbon sink structure limitations, and its spatial boundaries and location serve as the core target area for subsequent regulation strategies.

[0077] Through the above-mentioned fine-grained spatial identification and clustering process, it is possible to effectively locate key areas where carbon sequestration efficiency is limited due to forest stand structure deviating from the optimal range.

[0078] After identifying the bottleneck block P of the carbon sink structure, proceed to step S500: for each forest stand structure unit within block P, implement differentiated structural regulation strategies based on its specific structural deviation type to enhance its carbon sink response capability.

[0079] First, the standard deviation of diameter at breast height (DBH) distribution in each grid cell within block P is recalculated to assess structural diversity. If the DBH standard deviation in a cell is less than a preset threshold σ0 (e.g., 6.0 cm), it indicates excessive homogenization of the stand, a single structure, and limited ability of the trees to utilize light, water, and nutrients in a stratified manner.

[0080] For the aforementioned areas with narrow diameter at breast height (DBH), targeted thinning will be implemented. Thinning will primarily target trees concentrated within the dominant DBH range, reducing highly overlapping individuals of the same age and diameter to free up space for retaining and regenerating trees. The thinning ratio will be controlled between 15% and 25% to ensure that the standard deviation of DBH distribution is increased without excessively reducing total biomass. After thinning, plot data must be recollected to verify whether the standard deviation of DBH has reached the target value σ1 (e.g., ≥8.5 cm).

[0081] Secondly, for areas identified as single-layer forest structures—that is, areas with fewer than two forest layers and no obvious understory vegetation below the main canopy—multi-layer forest structure construction measures should be implemented. Specific methods include:

[0082] Young pine seedlings are artificially introduced into forest gaps or areas with low canopy closure, and the age of the introduced seedlings differs from that of the main forest layer tree species by no less than 10 years.

[0083] By utilizing existing natural regeneration individuals, competing vegetation (such as herbs and shrubs) is cleared away, improving their survival rate and growth, and promoting their evolution into understory forest.

[0084] The formation of the multi-layer structure can improve the utilization efficiency of the vertical space of the forest stand, enhance the carbon storage capacity and the system stability.

[0085] Thirdly, for the forest stand areas within the P block where the ratio of the average crown width to the average spacing is less than 0.65, it indicates a relatively high degree of crown overlap, resulting in uneven light distribution and limited water evaporation. Structural thinning needs to be implemented. The specific method is as follows: First, preferentially remove the standing trees with an overly large crown width and a crown overlap area with adjacent tree crowns exceeding 50%; control the thinning intensity not to exceed 20% to maintain the continuity of the forest stand; the adjusted goal is to make the ratio of the crown width to the spacing reach or be higher than 0.75, forming a more reasonable horizontal structure configuration and improving the microclimate and light conditions within the forest.

[0086] After completing the regulation of the carbon sink structure bottleneck block P, in order to verify its regulation effect and achieve dynamic feedback optimization, the present invention sets up a closed-loop evaluation mechanism: By re-collecting the forest stand structure parameters and evaluating the change in the annual carbon sequestration increment, it is judged whether the expected improvement threshold is met. If not, the set of sensitive factors C1 in the response model is updated, and the regulation is re-executed until the carbon sink improvement effect meets the target requirements.

[0087] Specifically, first, within 1 to 2 growing seasons after the regulation of each P block is completed, the forest stand structure parameters are re-collected to construct an updated structural feature vector, denoted as DV1. The dimension of DV1 is the same as the original vector V1, including five indicators: diameter at breast height distribution, tree height distribution, number of forest layers, average crown width, and average spacing, ensuring consistent docking with the model structure.

[0088] DV1 and the forest land fertility parameter V2 obtained in the previous step are jointly input into the existing carbon sink impact factor response model to predict the annual carbon sequestration increment per unit area in the current state of this sample plot, denoted as M2, with the unit of tons of carbon per hectare per year.

[0089] At the same time, the predicted value M1 of the carbon sequestration increment before regulation in the historical record is called, and the carbon sink improvement amount ΔC is calculated. Its calculation method is: ΔC = M2 - M1; In order to judge whether the regulation effect meets the standard, the present invention sets a carbon sequestration improvement threshold T, and the value of T can be adjusted according to different forest types and fertility conditions. The recommended range is 0.6 to 1.0 tons of carbon per hectare per year. If ΔC ≥ T, it means that the current round of regulation has achieved the goal and the process ends; otherwise, it is regarded as insufficient structural response and enters the feedback optimization process.

[0090] When ΔC < T, enter the model retraining process: Use the V1, DV1, V2 before and after the regulation of the current sample plot and the corresponding carbon sequestration increments M1 and M2 to form new sample data, add it to the historical training data set, and use the incremental training mechanism (IncrementalLearning) to update the original random forest regression model, re-calculate the contribution degree weights of each structural factor, and generate an updated set of sensitive factors C1'.

[0091] The factor weights and critical intervals in the newly generated C1′ may differ from those in the original C1, indicating that the model has a more accurate understanding of the structural response mechanism. Subsequently, following the described step S400, a new bottleneck block P′ is re-identified, and structural regulation is performed again.

[0092] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for regulating stand structure to improve the carbon sequestration capacity of pine forests, characterized in that: include: S100. Obtain the stand structure parameters of the target pine forest plot, including diameter at breast height (DBH) distribution, tree height distribution, number of forest layers, average crown width and average spacing, and construct the stand structure feature vector V1. S200. Soil carbon, nitrogen and phosphorus content was tested in the target plot to obtain forest land fertility parameter set V2, which includes soil organic carbon content, available phosphorus content, available potassium content and pH value. S300. Input V1 and V2 into the preset pine forest carbon sink impact factor response model and output the carbon sink sensitive structural factor set C1. S400. Based on C1, identify forest stand structural units in the target pine forest plot that do not meet the critical value, and determine the carbon sink structure bottleneck block P. S500. Regulate the stand structure units within each P block, including adjusting the standard deviation of diameter at breast height (DBH) in areas with narrow DBH distribution through directional thinning; artificially introducing different-aged pine species or naturally regenerated species to form multi-layered forest structures in single-layer forest areas; and implementing thinning to reduce canopy overlap in stands where the ratio of average crown width to average spacing is less than a set threshold.

2. The method for regulating forest structure to improve carbon sequestration capacity of pine forests according to claim 1, characterized in that: The input of V1 and V2 into the preset pine forest carbon sequestration impact factor response model includes: A carbon sink impact factor response model based on the random forest regression algorithm was constructed. The structural parameters in historical sample plot data and the measured carbon sink growth were used for training, and a variable importance ranking mechanism was established. Using the obtained V1 and V2 as model inputs, the annual carbon sink growth per unit area of ​​the target sample plot is predicted, and the contribution of each structural parameter to the carbon sink change is calculated. Carbon sink sensitive factors are extracted based on structural parameters whose contribution is greater than a preset threshold, forming a set C1, which includes the structural indicators that most significantly affect carbon sequestration efficiency and their critical value ranges.

3. The method for regulating stand structure to improve carbon sequestration capacity of pine forests according to claim 1, characterized in that: Among them, based on C1 identifies stand structure units in the target pine forest plot that do not meet the critical value, and determines the carbon sink structure bottleneck block P, including: Using the critical value range of each sensitive factor in C1 as the structural screening criterion, a multi-factor spatial filtering rule is constructed. The target plot is divided into several regular grid cells, and the corresponding structural feature vector V1′ is recalculated for each cell; By comparing V1′ with C1 one by one, grid cells that do not meet all critical interval conditions are identified. Adjacent units that do not meet the conditions are clustered to form a continuous spatial block P, which serves as the bottleneck block of the carbon sink structure to be regulated.

4. The method for regulating stand structure to improve carbon sequestration capacity of pine forests according to claim 3, characterized in that: The process of comparing V1′ with C1 one by one to identify mesh cells that do not meet all critical interval conditions includes: The structural feature sub-vector V1′ of each grid cell is mapped one-to-one with the critical value range of the sensitivity factor in C1 according to its dimension; Set logical matching rules. When any dimension indicator does not fall into the corresponding critical interval, the unit is determined to be a structurally substandard unit. Generate a structure matching matrix M, where a value of 0 indicates that the element does not meet the critical requirements, and a value of 1 indicates that it does. Summarize the cell numbers of all cells with an M value of 0 and record their spatial coordinates as the target set for cluster analysis.

5. The method for regulating stand structure to improve carbon sequestration capacity of pine forests according to claim 1, characterized in that: The regulation of forest stand structure units within each P block includes: For areas where the standard deviation of diameter at breast height (DBH) distribution is less than the set target value, targeted thinning is used to remove individuals with high DBH concentration, thereby expanding the DBH distribution towards greater diversity. In plots identified as single-layer forest structures, construct multi-layer forest structures; For stand areas where the ratio of average canopy width to average spacing is less than 0.65, thinning should be implemented to reduce densely overlapping canopies.

6. The method for regulating stand structure to improve carbon sequestration capacity of pine forests according to claim 1, characterized in that: After the adjustment is completed, V1 is collected again, and the annual carbon sequestration growth ΔC before and after the adjustment is compared. If ΔC does not reach the set increase threshold T, return to step S400 to update C1 and adjust again until ΔC≥T.

7. The method for regulating stand structure to improve carbon sequestration capacity of pine forests according to claim 6, characterized in that: After the forest stand structure regulation is completed, the carbon sequestration effect will be verified and dynamically optimized, including: After re-collecting the forest stand structure parameters after regulation, an updated structural feature vector DV1 was constructed and input into the carbon sink impact factor response model to calculate the annual carbon sink growth per unit area after regulation. By comparing the increase in carbon sink before regulation with the result after regulation, the increase in carbon sequestration ΔC is obtained. Set a threshold T for improving carbon fixation. When ΔC is less than T, retrain the importance weights of the sensitivity factors in C1. Based on the updated C1 results, the bottleneck block P is re-identified and structural control is performed to form a closed-loop optimization cycle until ΔC reaches or exceeds the T value.

Images