A method for monitoring carbon sequestration and sink effect based on multi-region ecological restoration

Abstract

The application discloses a kind of carbon sequestration and carbon sink effect monitoring methods based on multi-region ecological restoration, it is related to ecological restoration technical field.The method steps include: according to the microhabitat unit and microbiota data collection are divided by geomorphic feature;According to the microbiota data, a multiple linear regression equation is established, a first community characteristic coefficient is obtained, the microbiota data is input into the random forest model to obtain the community characteristic importance, and the first community characteristic coefficient is combined by dynamic weighted fusion method to obtain the second community characteristic coefficient;The second community characteristic coefficient is calculated by super efficiency effect analysis method, the community synergistic gain rate is obtained, and the restoration planting type is determined;The first microhabitat geomorphic cluster is obtained by clustering geomorphic data by KMEANS method, and the second microhabitat geomorphic cluster is obtained by correcting through the restoration planting type;The restoration carbon sink intensity is collected and compared with the cluster carbon sink intensity threshold value, and if it is lower than the threshold value, it is early warned, and the monitoring of carbon sink intensity is completed.

Description

Technical Field

[0001] This invention relates to the field of ecological restoration technology, specifically to a method for monitoring the carbon sequestration and enhancement effects of ecological restoration in multiple regions. Background Technology

[0002] Against the backdrop of a global effort to address climate change and advance dual-carbon goals, enhancing the carbon sequestration capacity of terrestrial ecosystems has become a crucial element supporting national strategic needs. Ecological restoration projects, such as reforestation and grassland restoration, mine revegetation, and wetland restoration, serve as important pathways to enhance the carbon sequestration function of ecosystems. Accurate monitoring and scientific assessment of their carbon sequestration effects are of paramount importance for optimizing restoration strategies, calculating carbon sequestration potential, and supporting the healthy development of the carbon trading market. Among these, the scientific quantification of the carbon sequestration intensity of ecosystems under different restoration models is a core foundation for evaluating the effectiveness of restoration projects and guiding future restoration designs.

[0003] In existing technologies, monitoring the carbon sequestration effect of ecological restoration mainly involves estimating the total carbon storage of an ecological area through methods such as ground quadrat surveys and biomass calculations. The degree of carbon sequestration and sink enhancement during ecological restoration is then assessed based on the total carbon storage of the ecological area.

[0004] However, existing technologies rely solely on a single indicator, such as the total carbon storage of an ecological region, without considering the survival relationships between vegetation in different ecological regions or identifying the complex interactions between vegetation communities. Consequently, they cannot fully decompose the specific contributions of different vegetation combinations to carbon sequestration and carbon sink enhancement, leading to often one-sided assessments of carbon sequestration and carbon sink enhancement effects. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method for monitoring the carbon sequestration and enhancement effects of multi-regional ecological restoration, thereby resolving the problems existing in the background technology.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for monitoring the carbon sequestration and enhancement effect based on multi-regional ecological restoration, comprising the following steps: Step S1: Based on the geomorphological features of the ecological restoration area, divide the ecological restoration area into several microhabitat units; collect community data of the microhabitat units to obtain microcommunity data; Step S2: Establish a multiple linear regression equation based on the microbial community data and calculate the first community characteristic coefficient; input the microbial community data into the random forest model and calculate the importance of community characteristics; Step S3: Combine the importance of the community features with the first community feature coefficient using a dynamic weighted fusion method to obtain the second community feature coefficient; perform community synergy calculation on the second community feature coefficient using the overproduction effect analysis method to obtain the community synergy gain rate; Step S4: Based on the community synergistic gain rate and the second community characteristic coefficient, obtain the remediation planting type of the microhabitat unit; collect the geomorphological data of the microhabitat unit, and cluster the geomorphological data of the microhabitat unit using the KMEANS method to obtain the first microhabitat geomorphological cluster; modify the first microhabitat geomorphological cluster using the remediation planting type to obtain the second microhabitat geomorphological cluster. Step S5: Based on the remediation planting type, ecological restoration is carried out on the microhabitat units, and the standard carbon sink intensity after restoration is collected to obtain the restored carbon sink intensity; by analyzing the distribution of restored carbon sink intensity within the second microhabitat landform cluster, the threshold of carbon sink intensity within the cluster is obtained; the restored carbon sink intensity of the microhabitat units in the second microhabitat landform cluster is compared with the threshold of carbon sink intensity within the cluster, and an early warning is issued if it is less than the threshold of carbon sink intensity within the cluster, thereby realizing the monitoring of carbon sequestration and enhancement effect.

[0007] Preferably, the step of establishing a multiple linear regression equation based on the microbial community data and calculating the first community characteristic coefficient includes the following specific steps: Collect tree canopy coverage over N periods Shrub canopy coverage Herbaceous community coverage and standard carbon sequestration intensity, for tree canopy coverage under the N periods. Shrub canopy coverage Herbaceous community coverage Perform a linear transformation to convert it into three main effect terms. , , and 3 interactive items , , :: , , , , , These are the three main effect terms after transformation; after obtaining the main effect terms, interaction terms are constructed using the main effect terms: , , , for The average value, for The average value, for The average value; Combining the three main effect terms , , and 3 interactive items , , Based on the standard carbon sink intensity, a multiple linear regression equation was constructed, and the first community characteristic coefficient was calculated. The first community characteristic coefficient includes: , , , , , , :

[0008] Where Y is the standard carbon sequestration intensity. Based on basic carbon sequestration capacity, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, For the first residual term, ~N(0, ).

[0009] Preferably, the step of inputting the microbial community data into a random forest model to calculate the importance of community features includes the following steps: The microbial community data for N periods , , , , , After Z-score standardization, the following is obtained , , , , , ,Will , , , , , Standard carbon sequestration intensity Y, as the independent variable, is used as the dependent variable and input into the random forest model for training. Decision trees are generated through Bootstrap sampling, resulting in a decision tree set T={ , ,..., ,..., There are a total of M decision trees; For each tree and microbial community data characteristics Calculate the cumulative purity increase for each feature across all split points:

[0010] in, Represents the j-th feature in the microbial community data. The importance of features in the m-th decision tree. For trees The set of splitting points, h is In the context of a split point h, I() is an indicator function, i.e., the characteristic index of the split point h. The index j is 1 when it is equal to the index j of the microbial community data feature, and 0 when they are not. The feature index of the split point h, The reduction in Gini impurity at the splitting point h. = , Let the Gini impurity be the parent node of the split point h. The child node of the split point h, Let be the Gini impurity of the child node `child` of the split point `h`. and , where h represents the number of samples for the parent and child nodes of the split point, and j is the index of the microbial community data features; Then features The importance of community features in the random forest model is as follows:

[0011] in, For the j-th feature of the microbial community data In a random forest model, the importance of community features is given by M, where M is the total number of decision trees and m is the m-th decision tree. Represents the j-th feature in the microbial community data. Importance of community features in the m-th decision tree.

[0012] Preferably, the step of combining the importance of the community features and the first community feature coefficients using a dynamic weighted fusion method to obtain the second community feature coefficients includes the following steps: The second community feature coefficient is obtained by combining the importance of the community features with the first community feature coefficient using a dynamic weighted fusion method.

[0013] in, Let j be the second community characteristic coefficient of the j-th feature in the microbial community data. 0, Let be the first community characteristic coefficient of the j-th feature in the microbial community data. This is the nonlinear intensity factor, with a default value of 0.8. For the j-th feature of the microbial community data The importance of community features in random forest models This represents the mean importance of all features in the microbial community data within the random forest model. This represents the standard deviation of the importance of all features in the microbial community data within the random forest model. This refers to the out-of-bag error of the random forest model. The coefficient of determination is represented by the coefficient of determination in a multiple linear regression model. Based on the second community characteristic coefficient, the modified multiple linear regression equation is obtained.

[0014] Preferably, obtaining the modified multiple linear regression equation based on the second community characteristic coefficients includes the following steps: The multiple linear regression equation is corrected by using the second community characteristic coefficient, resulting in the corrected multiple linear regression equation:

[0015] Where Y is the standard carbon sequestration intensity. Based on basic carbon sequestration capacity, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, This is the second residual term.

[0016] Preferably, the step of performing community synergy calculation on the characteristic coefficients of the second community using the overproduction effect analysis method to obtain the community synergy gain rate includes the following steps: Calculating the carbon sequestration intensity when only trees are present, the shrub crown cover and herbaceous community cover are 0. Substituting this into the modified multiple linear regression equation: , Standard carbon sequestration intensity for trees; Calculating the carbon sink intensity when only shrubs exist, the tree canopy cover and herbaceous community cover are 0. Substituting this into the modified multiple linear regression equation: , Standard carbon sequestration intensity for shrubs; Calculating the carbon sink intensity when only shrubs exist, the tree canopy cover and herbaceous community cover are 0. Substituting this into the modified multiple linear regression equation: , Standard carbon sequestration intensity of vegetation; The community synergy gain rate was obtained by performing community synergy calculations on the characteristic coefficients of the second community using the overproduction effect analysis method.

[0017] Wherein, SCG is the community-coordinated gain rate. This represents the actual standard carbon sequestration intensity. The standard carbon sequestration intensity of trees, Standard carbon sequestration intensity for shrubs, The standard carbon sequestration intensity of vegetation, To prevent division by zero, the value is set to max( ,0.01*( )).

[0018] Preferably, the step of obtaining the remediation planting type of the microhabitat unit based on the community synergistic gain rate and the second community characteristic coefficient includes the following specific steps: When SCG > 0, the actual carbon sequestration intensity of the community is greater than the sum of the individual carbon sequestration intensities, indicating a community synergistic effect within this microhabitat unit. '、 '、 Perform a t-test, retain only the significantly non-zero second community characteristic coefficient, and according to... '、 '、 The significantly non-zero second community characteristic coefficient determines the type of restored vegetation; when SCG < 0, competitive inhibition exists in the community, which has a significant impact on the restoration of vegetation. '、 '、 Perform a t-test, retain only the significantly non-zero second community characteristic coefficient, and according to... '、 '、 The significantly non-zero second community characteristic coefficient determines the type of restored vegetation; when SCG=0, this indicates that the community is at a balance between cooperation and competition, which is beneficial for the restoration of vegetation. '、 '、 '、 '、 '、 Perform a t-test and based on '、 '、 '、 '、 '、 The significantly non-zero second community characteristic coefficient determines the type of vegetation to be restored.

[0019] Preferably, the step of clustering the geomorphic data of the microhabitat units using the KMEANS method to obtain the first microhabitat geomorphic cluster includes the following specific steps: The KMEANS clustering algorithm was used to perform cluster analysis on the above-mentioned geomorphic data. The geomorphic data of all corresponding microhabitat units were Z-score standardized to eliminate the differences in the dimensions and numerical ranges of different indicators. The elbow method was used to analyze the clustering error of the geomorphic data of microhabitat units, and the elbow position where the error descent trend changed from steep to gentle was found to determine the optimal number of clusters K. After determining the number of clusters K, the geomorphic data of microhabitat units were clustered using the KMEANS clustering method. Through iterative calculation, the microhabitat units were clustered to obtain the first microhabitat geomorphic cluster.

[0020] Preferably, the step of modifying the first microhabitat landform cluster using the remediation planting type to obtain the second microhabitat landform cluster includes the following specific steps: Based on the first microhabitat landform cluster, microhabitat units within the first microhabitat landform cluster that are identified as having the same restoration planting type are grouped into one category, thus completing the modification of the first microhabitat landform cluster and obtaining the second microhabitat landform cluster, ultimately resulting in K second microhabitat landform clusters. .

[0021] Preferably, the step of analyzing the distribution of carbon sink intensity within the second microhabitat landform cluster to obtain the intra-cluster carbon sink intensity threshold includes the following specific steps: For each second microhabitat geomorphic cluster, the distribution of remediation carbon sink intensity within the cluster is analyzed to calculate the mean and standard deviation of the remediation carbon sink intensity. The formula for calculating the mean remediation carbon sink intensity is as follows:

[0022] in, denoted as the mean carbon sink intensity for the restoration of the second microhabitat geomorphic cluster, WS represents the number of microhabitat units in the second microhabitat geomorphic cluster, and ws represents the index of the microhabitat unit. The remediation carbon sink intensity of the ws-th microhabitat unit, Let ws be the area of ​​the w-th microhabitat unit; The formula for calculating the standard deviation of carbon sequestration intensity is:

[0023] in, denoted as , where WS is the standard deviation of the carbon sink intensity for the restoration of the second microhabitat geomorphic cluster, WS is the number of microhabitat units in the second microhabitat geomorphic cluster, and ws is the index of the microhabitat unit. The remediation carbon sink intensity of the ws-th microhabitat unit, The average carbon sink intensity for the restoration of the second microhabitat landform cluster; For each second microhabitat landform cluster, the intra-cluster carbon sink intensity threshold is calculated by combining the mean and standard deviation of the restored carbon sink intensity:

[0024] in, The threshold value for intra-cluster carbon sink intensity. This represents the average carbon sequestration intensity for the restoration of the second microhabitat landform cluster. The standard deviation of the carbon sink intensity for the restoration of the second microhabitat landform cluster. This is the threshold adjustment coefficient.

[0025] This invention provides a method for monitoring carbon sequestration and enhancement effects based on multi-regional ecological restoration, involving machine learning and deep learning technologies, which has the following beneficial effects: (1) The second community characteristic coefficient is calculated by combining the importance of community characteristics with the first community characteristic coefficient. The second community characteristic coefficient is based on the first community characteristic coefficient. At the same time, the importance of community characteristics in the random forest model is introduced. The nonlinear contribution of different vegetation characteristics to carbon sequestration is characterized by standardized scores. The fusion ratio is dynamically adjusted by using the random forest bag out-of-bag error and the linear model determination coefficient. The final generated second community characteristic coefficient not only retains interpretable mechanism information, but also accurately captures nonlinear laws, providing a scientific and practical quantitative support for subsequent restoration decisions.

[0026] (2) Based on the community synergistic gain rate (and the second community characteristic coefficient) to determine the remediation planting type of the microhabitat unit, SCG is used as the core judgment indicator. First, the overall effect direction of the community is clarified, and then the specific driving or competitive source is located by combining the second community characteristic coefficient: In the synergistic scenario, only the significantly non-zero second community characteristic coefficient is retained, and according to '、 '、 The significantly non-zero second community characteristic coefficient determines the type of vegetation to be restored; in a competitive scenario, only the second community characteristic coefficient is retained. '、 '、 The significantly non-zero second community characteristic coefficient determines the type of vegetation to be restored. This process completely avoids the drawbacks of a one-size-fits-all approach to restoration, ensuring that the planting type of each microhabitat unit can match its ecological conditions and maximize the synergistic effect of carbon sequestration.

[0027] (3) The first microhabitat landform cluster was modified by the type of restoration planting to obtain the second microhabitat landform cluster. The first microhabitat landform cluster was generated by KMEANS clustering based only on landform data such as altitude, slope, and aspect. Although it achieved the classification of geographical features based on similarity, it did not consider the actual needs of ecological restoration. On the other hand, the second microhabitat landform cluster, based on similar landform features, further classified microhabitat units that are adapted to the same type of restoration planting into one category. This means that units within the same category not only have similar geographical conditions, but also have consistent restoration goals and ecological foundations. This modification makes the subsequent carbon sink intensity analysis more targeted. Attached Figure Description

[0028] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 This is a flowchart of the steps of a method for monitoring carbon sequestration and enhancement effects based on multi-regional ecological restoration proposed in this invention; Figure 2 This is a step hierarchy diagram of obtaining the second community characteristic coefficient in a method for monitoring carbon sequestration and enhancement effects based on multi-regional ecological restoration proposed in this invention; Figure 3 This is a step hierarchy diagram of obtaining the intra-cluster carbon sink intensity threshold in a carbon sequestration and enhancement monitoring method based on multi-regional ecological restoration proposed in this invention. Detailed Implementation

[0030] 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, and 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.

[0031] Please see Figures 1-3 This invention provides a technical solution: a method for monitoring the carbon sequestration and enhancement effect based on multi-regional ecological restoration.

[0032] Step S1: Divide the ecological restoration area into several microhabitat units based on the geomorphological features of the ecological restoration area; collect community data of the microhabitat units to obtain microcommunity data.

[0033] The ecological restoration area is divided into several microhabitat units. The specific division of microhabitat units is mainly based on the significant spatial differences and combination characteristics of four key environmental factors: climate, topography, vegetation, and soil. Climate factors consider local variations in elements such as temperature, precipitation, and light (e.g., the microclimate differences between sunny and shady slopes). Topographic factors focus on analyzing altitude, slope, aspect (which affect light and water distribution), and landform location (e.g., mountain top, slope, foothills, valley bottom). Vegetation factors are based on the spatial distribution patterns of dominant species composition, community structure type (e.g., forest, shrubland, grassland), and vegetation cover (density). Soil factors focus on the spatial variability of soil type (e.g., sandy soil, loam, clay), soil organic matter content, pH value, texture, and thickness. When dividing microhabitat units, the spatial distribution layers of the above factors are overlaid by combining field survey data, high-resolution remote sensing images and geographic information system (GIS) analysis technology to identify continuous spatial ranges with relatively consistent combinations of these factors and distinguishable boundaries with neighboring areas (such as plots with the same slope direction, similar slope, same vegetation community and homogeneous soil type). At the same time, ridgelines, streams, roads or obvious vegetation type boundaries are used as natural separators to determine the natural and ecological boundaries of each unit, ensuring high homogeneity of the internal environment of the unit and clear distinguishable differences in environmental characteristics between units.

[0034] High-resolution (centimeter-level) orthophotos were generated weekly using a drone-borne multispectral camera. These images were then validated using ground quadrats (three 10m × 10m quadrats were set up in each microhabitat unit, and the projected canopy cover of trees, shrubs, and grasses was measured manually). Tree canopy cover was extracted using a deep learning semantic segmentation model (such as U-Net). Shrub canopy coverage Herbaceous community coverage The calculation results are stored in the spatial database after being corrected by ground control points; It should be noted that tree canopy coverage Shrub canopy coverage Herbaceous community coverage The calculation method involves setting up three 10m × 10m (area) units in each microhabitat unit. If a square plot of 100 square meters is used as the unit index, then the tree canopy coverage is... , among which, among which Let be the number of trees in the b-th quadrat. The canopy projection area of ​​the qiao-th tree, and the canopy coverage of the shrubs. Herbaceous community coverage tree canopy coverage The calculation methods are consistent.

[0035] Physiological process data collection: Automatically openable and closed soil respiration chambers (bare surface and root zone) were installed within typical community units to monitor soil carbon dioxide flux density, with data recorded continuously every minute; based on the continuously monitored carbon dioxide flux density (NEE) within the microhabitat unit, flux values ​​exceeding the reasonable range were removed (the normal range for ecosystem carbon dioxide flux density is -10 to 20 μmol). * * The measured CO2 flux was converted to carbon dioxide flux under standard conditions of 0℃ and 101.325 kPa. , ,in, To measure the carbon dioxide flux, For actual measured temperature, The measured air pressure. The negative carbon dioxide flux during the daytime (6:00-18:00) is... <0) Integrate over time at a resolution of seconds to obtain the total carbon accumulation within the period. (Unit: gC / m) 2 ), ; Ultimately Perform unit area and mass conversions to calculate the standard carbon sequestration intensity Y. , where 0.01 is The coefficient for conversion to ha, 1 ha = , The coefficient for converting g to t.

[0036] Environmental covariate collection: A micro-weather station was set up at the center of the unit to obtain the photosynthetically active radiation intensity in real time through a quantum sensor, and the soil volumetric water content was monitored using a soil moisture probe (buried at a depth of 20cm).

[0037] Step S2: Establish a multiple linear regression equation based on the microbial community data and calculate the first community characteristic coefficient; input the microbial community data into the random forest model and calculate the importance of community characteristics.

[0038] For each microhabitat unit, tree canopy coverage is collected over N periods. Shrub canopy coverage Herbaceous community coverage And standard carbon sequestration intensity, based on the tree canopy coverage under the aforementioned N periods. Shrub canopy coverage Herbaceous community coverage Based on the standard carbon sink intensity, a multiple linear regression equation was constructed, and the first community characteristic coefficient was calculated. The first community characteristic coefficient includes: , , , , , , First, we examine each main effect term using partial residual plots. , , A linear relationship with Y, if the residuals of the variable exhibit a non-linear distribution (e.g. If the residual turns positive after 60% reduction, then a logarithmic transformation is performed on the variable: , , , , , The transformed main effect term, avoid =0 is undefined. After obtaining the main effect term, the main effect term is centered, and interaction terms are constructed: , , , for The average value, for The average value, for The average value. Finally, a multiple linear regression equation is constructed:

[0039] Where Y is the standard carbon sequestration intensity. Based on basic carbon sequestration capacity, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, For the first residual term, ~N(0, ).

[0040] It should be noted that this multiple linear regression equation quantifies the main effect term. , , and its interactive items , , Effect on standard carbon sequestration intensity Y: constant term Characterizing the carbon sequestration capacity of non-vegetation background (such as soil microbial activity); coefficients of the main effect term. , , Representing respectively , , The average change in Y for each unit increase. Coefficient of the interaction term. express For every additional unit, The rate of change of the marginal contribution to Y, express For every additional unit, The rate of change of the marginal contribution to Y, express For every additional unit, The rate of change of the marginal contribution to Y.

[0041] It should be noted that the sample size should be determined before constructing the multiple linear regression equation. 60 (sample size) 10 * number of independent variables), then... , , , , , Perform a multicollinearity test on each independent variable and calculate the variance inflation factor for each independent variable. ,in, Indicates The coefficients of determination for an auxiliary regression model with as the dependent variable and the other 5 independent variables as predictors are required. ,like This indicates that the independent variable If severe multicollinearity exists among other independent variables, that independent variable is removed until the VIF of all retained independent variables is less than 10. After constructing the multiple linear regression equation, the residuals are verified using the Shapiro-Wilk test. The normality of the residuals is p-value > 0.05, which means we accept the hypothesis that the residuals follow a normal distribution.

[0042] It should be noted that tree canopy coverage, shrub canopy coverage, and herbaceous community coverage were chosen as core independent variables because these three types of indicators fully characterize the vertical stratification structure of terrestrial ecosystems. The tree layer dominates the construction of long-term carbon sinks and light interception, the shrub layer regulates soil and water conservation and soil carbon accumulation, and the herbaceous layer drives short-term carbon turnover and microclimate regulation. Their combination together determines the biophysical framework for carbon sink formation.

[0043] The microbial community data for N periods , , , , , After Z-score standardization, the following is obtained , , , , , ,Will , , , , , Standard carbon sequestration intensity Y, as the independent variable, is used as the dependent variable and input into the random forest model for training. Decision trees are generated through Bootstrap sampling, resulting in a decision tree set T={ , ,..., ,..., There are a total of M decision trees.

[0044] It should be noted that the number of decision trees, M, is determined by the "OOB error - M curve". First, M is initialized. [10, 1000], with a step size of 10; train a random forest model for each M and calculate the OOB error: , Indicates the number of samples outside the bag. Let M be the predicted values ​​of the i-th out-of-bag sample from M trees. The final decision tree (usually 200-300) is selected based on the minimum out-of-bag error that does not significantly decrease further. The final out-of-bag error corresponding to this determined M is then obtained. .

[0045] For each tree and microbial community data characteristics Calculate the cumulative purity increase for each feature across all split points:

[0046] in, Represents the j-th feature in the microbial community data. The importance of features in the m-th decision tree. For trees The set of splitting points, h is In the context of a split point h, I() is an indicator function, i.e., the characteristic index of the split point h. The index j is 1 when it is equal to the index j of the microbial community data feature, and 0 when they are not. The feature index of the split point h, The reduction in Gini impurity at the splitting point h. = , Let the Gini impurity be the parent node of the split point h. The child node of the split point h, Let be the Gini impurity of the child node `child` of the split point `h`. and , where h represents the number of samples for the parent and child nodes of the split point, and j is the index of the microbial community data features; Then features The importance of community features in the random forest model is as follows:

[0047] in, For the j-th feature of the microbial community data In a random forest model, the importance of community features is given by M, where M is the total number of decision trees and m is the m-th decision tree. Represents the j-th feature in the microbial community data. Importance of community features in the m-th decision tree.

[0048] Step S3: Combine the importance of the community features with the first community feature coefficient using a dynamic weighted fusion method to obtain the second community feature coefficient; perform community synergy calculation on the second community feature coefficient using the overproduction effect analysis method to obtain the community synergy gain rate.

[0049] The second community feature coefficient is obtained by combining the importance of the community features with the first community feature coefficient using a dynamic weighted fusion method.

[0050] in, Let j be the second community characteristic coefficient of the j-th feature in the microbial community data. 0, Let be the first community characteristic coefficient of the j-th feature in the microbial community data. This is the nonlinear intensity factor, with a default value of 0.8. For the j-th feature of the microbial community data The importance of community features in random forest models This represents the mean importance of all features in the microbial community data within the random forest model. This represents the standard deviation of the importance of all features in the microbial community data within the random forest model. This refers to the out-of-bag error of the random forest model. The coefficient of determination is represented by the coefficient of determination in a multiple linear regression model.

[0051] It should be noted that the nonlinear intensity factor , , The maximum coefficient of variation for vegetation cover. =max( ).

[0052] It should be noted that the second community characteristic coefficient dynamically weights and integrates the ecological interpretability of multiple linear regression with the nonlinear predictive ability of random forest: using the linear regression coefficients... As the direction and fundamental strength of ecological functions, the importance of random forest characteristics is standardized by the score. Quantifying its nonlinear contribution ( (Controlling the magnitude of enhancement), and finally through ecological reliability weights. (Based on out-of-bag error and R) 2 (Calculation) Adjust the fusion ratio, , , , For model predictions Compared with the true value The sum of squares of the differences For the true value its mean The sum of squares of the differences. When both models perform poorly (e.g., random forest has a large error and linear model R0), the sum of squares of the differences is used. 2 (very low), then It will be very small (close to 0), at which point... Weak corrections are made to avoid the accumulation of errors. Conversely, if the random forest is reliable (small out-of-bag error) and the linear model has strong explanatory power (R²), then... 2 (High), then When the value is close to 1, a correction is made to maximize it. exist Based on the mechanism, the nonlinear signal is precisely adjusted through the random forest model, which preserves the " The core mechanism of "positive correlation with carbon sinks" also captures the real pattern of "slower growth rate with high coverage." If random forests are reliable but linear models are unreliable, then... Close to 0.5 and If the mechanism is consistent, then the linear coefficients should be adjusted appropriately. If the random forest is reliable but the linear model is highly reliable, then... It's also close to 0.5. and The mechanisms are consistent (both reflecting the linear contribution of shrubs), but due to the unreliability of RF, only tentative corrections were made. The final optimized parameters reflect both the mechanisms (e.g., the carbon sequestration gain for every 1% increase in tree cover) and complex interactions (e.g., the threshold effect of tree-shrub coexistence). This provides a scientific and practical quantitative basis for ecological restoration.

[0053] It should be noted that the multiple linear regression equation is modified using the second community characteristic coefficient, resulting in the modified multiple linear regression equation:

[0054] Where Y is the standard carbon sequestration intensity. Based on basic carbon sequestration capacity, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, This is the second residual term.

[0055] To calculate the community synergistic gain rate for each period, we first need to calculate the carbon sink intensity when each type of vegetation exists in isolation. To calculate the carbon sink intensity when only trees are present, we set the shrub crown cover and herbaceous community cover to 0. When the shrub crown cover and herbaceous community cover are 0, we input the modified multiple linear regression equation: in, The standard carbon sequestration intensity for trees.

[0056] To calculate the carbon sink intensity when only shrubs are present, the tree canopy cover and herbaceous community cover are set to 0. When the tree canopy cover and herbaceous community cover are 0, the modified multiple linear regression equation is used: , The standard carbon sequestration intensity for shrubs.

[0057] To calculate the carbon sink intensity when only shrubs are present, the tree canopy cover and herbaceous community cover are set to 0. When the tree canopy cover and herbaceous community cover are 0, the modified multiple linear regression equation is used: , The standard carbon sequestration intensity of vegetation.

[0058] The community synergy gain rate was obtained by performing community synergy calculations on the characteristic coefficients of the second community using the overproduction effect analysis method.

[0059] Wherein, SCG is the community-coordinated gain rate. This represents the actual standard carbon sequestration intensity. The standard carbon sequestration intensity of trees, Standard carbon sequestration intensity for shrubs, The standard carbon sequestration intensity of vegetation, To prevent division by zero, the value is set to max( ,0.01*( )).

[0060] It should be noted that the community synergy gain rate is obtained by performing community synergy calculations on the characteristic coefficients of the second community using the overproduction effect analysis method. The synergy effect of the community is represented by comparing the relative difference between the actual carbon sink intensity of the community and the sum of the predicted carbon sink values ​​when each vegetation type exists in isolation. A positive SCG value indicates that there is a synergy effect in the community (the carbon sink intensity produced by the coexistence of multiple vegetation types is higher than the carbon sink intensity when each vegetation type exists independently), while a negative value indicates that there is a competitive inhibition effect (the carbon sink intensity produced by the coexistence of multiple vegetation types is lower than the carbon sink intensity when each vegetation type exists independently). The dimensionless nature of this community synergy gain rate makes the ecological restoration effects of different regions and different periods comparable, thus providing a direct and quantitative decision-making basis for optimizing vegetation configuration patterns (such as increasing the proportion of mixed trees and shrubs).

[0061] Step S4: Based on the community synergistic gain rate and the second community characteristic coefficient, obtain the remediation planting type of the microhabitat unit; collect the geomorphological data of the microhabitat unit, and cluster the geomorphological data of the microhabitat unit using the KMEANS method to obtain the first microhabitat geomorphological cluster; modify the first microhabitat geomorphological cluster using the remediation planting type to obtain the second microhabitat geomorphological cluster.

[0062] When determining the planting type, it is necessary to first clarify the community synergistic gain rate (SCG) and the second community characteristic coefficient (SCR). The SCG (Self-Regulatory Coefficient) is used to determine whether the overall synergistic effect of the community is superior to that of individual plants. The second community characteristic coefficient is used to locate the specific vegetation combination that drives synergy or triggers competition. These factors combine to form the basis for planting type selection. The specific application steps are as follows: First, determine the direction of the overall effect based on the SCG. If SCG > 0, it indicates that the actual carbon sequestration intensity of the community is greater than the sum of the individual carbon sequestration intensities of trees, shrubs, and grasses. '、 '、 'conduct The test was performed, and only coefficients that were significantly non-zero (p<0.05) were retained. , ,in, for The estimated standard deviation Let be the cumulative distribution function of the standard normal distribution, if <0.05, then Significant values ​​are included in the decision; otherwise, they are excluded (considered random error). Based on significance... Combining PAR with Design a dynamic density formula, taking "mixed tree and shrub" as an example ( Significant): The baseline density is determined based on the micro-habitat climate zone (e.g., in temperate regions, the baseline density for trees is 150 trees / hectare, and for shrubs it is 300 trees / hectare), and PAR is the photosynthetically active radiation intensity. Here, k represents the soil volumetric water content, and k is the environmental adjustment coefficient, which defaults to 20. and These represent the mean PAR values ​​of microhabitat units, Mean; if SCG < 0, it indicates competition inhibition within the community (actual carbon sink intensity < sum of individual carbon atoms), and the source of competition needs to be located through the second community characteristic coefficient. Similarly, for... '、 '、 Perform a t-test, retaining only significantly non-zero coefficients (p<0.05), and pass the significant... Combining PAR with Design the dynamic density formula (same as above). Consistency) will ultimately lead to a remediation planting type that adapts to the current microhabitat unit, avoids competition, and enhances synergy; if SCG=0, the remediation planting type will be determined primarily based on the feature term that passes the significance test (p<0.05) and has the largest absolute value of the coefficient. For example, if (Tree main effect) and Both the tree-shrub interaction effect and the tree-shrub interaction effect were significant. If the absolute value of ' is the largest, then the restoration type can focus on optimizing the combination of trees and shrubs to guide the microhabitat unit to develop in a positive direction.

[0063] Collect geomorphological data for each microhabitat unit, including data on altitude, slope, aspect, and geomorphic location (such as mountain top, slope, and foothills) obtained through field surveys.

[0064] The KMEANS clustering algorithm was used to perform cluster analysis on the above-mentioned geomorphological data. Z-score standardization was applied to the geomorphological data of all corresponding microhabitat units to eliminate differences in the dimensions and numerical ranges of different indicators. Subsequently, the elbow method was used to analyze the clustering error of the microhabitat unit geomorphological data, identifying the elbow position where the error descent trend gradually decreases, and determining the optimal number of clusters K. After determining the number of clusters K, the KMEANS clustering method was used to cluster the geomorphological data of the microhabitat units. Through iterative calculation, microhabitat units with similar geomorphological characteristics (such as the same slope aspect, similar slope, and the same geomorphological location) were grouped into one class, resulting in the first microhabitat geomorphological cluster. Based on the first microhabitat geomorphological cluster, microhabitat units within the first microhabitat geomorphological cluster that were determined to have the same remediation planting type were grouped into one class, completing the correction of the first microhabitat geomorphological cluster and obtaining the second microhabitat geomorphological cluster. Finally, K second microhabitat geomorphological clusters were obtained. .

[0065] Step S5: Ecologically restore the microhabitat unit using the remediation planting type and collect the standard carbon sink intensity after restoration to obtain the restored carbon sink intensity; analyze the distribution of restored carbon sink intensity within the second microhabitat landform cluster to obtain the cluster carbon sink intensity threshold; compare the restored carbon sink intensity of the microhabitat unit in the second microhabitat landform cluster with the cluster carbon sink intensity threshold, and issue an early warning if it is less than the cluster carbon sink intensity threshold, thereby realizing the monitoring of carbon sequestration and enhancement effect.

[0066] In each cycle, the microhabitat unit is ecologically restored using the aforementioned restoration planting type, and the standard carbon sink intensity after restoration is collected. Outliers in the collected data are then removed using the interquartile range method to obtain the final restored carbon sink intensity.

[0067] For each second microhabitat geomorphic cluster, the distribution of remediation carbon sink intensity within the cluster is analyzed to calculate the mean and standard deviation of the remediation carbon sink intensity. The formula for calculating the mean remediation carbon sink intensity is as follows:

[0068] in, denoted as the mean carbon sink intensity for the restoration of the second microhabitat geomorphic cluster, WS represents the number of microhabitat units in the second microhabitat geomorphic cluster, and ws represents the index of the microhabitat unit. The remediation carbon sink intensity of the ws-th microhabitat unit, Let ws be the area of ​​the ws-th microhabitat unit.

[0069] The formula for calculating the standard deviation of carbon sequestration intensity is:

[0070] in, denoted as , where WS is the standard deviation of the carbon sink intensity for the restoration of the second microhabitat geomorphic cluster, WS is the number of microhabitat units in the second microhabitat geomorphic cluster, and ws is the index of the microhabitat unit. The remediation carbon sink intensity of the ws-th microhabitat unit, This represents the average carbon sink intensity for the restoration of the second microhabitat landform cluster.

[0071] For each second microhabitat landform cluster, the intra-cluster carbon sink intensity threshold is calculated by combining the mean and standard deviation of the restored carbon sink intensity:

[0072] in, The threshold value for intra-cluster carbon sink intensity. This represents the average carbon sequestration intensity for the restoration of the second microhabitat landform cluster. The standard deviation of the carbon sink intensity for the restoration of the second microhabitat landform cluster. This is the threshold adjustment coefficient.

[0073] It should be noted that the threshold adjustment coefficient Based on the coefficient of variation of each secondary microhabitat landform cluster Adjustments should be made if This indicates that the data dispersion is low in the second microhabitat landform cluster. A value of 1.0-1.2 is acceptable; if This indicates that the data dispersion in the second microhabitat landform cluster is moderate. A value of 1.2-1.4 is acceptable; if This indicates that the data dispersion in the second microhabitat landform cluster is moderately high. A value of 1.4-1.6 is acceptable.

[0074] When the remediation carbon sink intensity of a certain microhabitat unit within multiple consecutive monitoring periods (2 or 3 monitoring periods) All were below the intra-cluster carbon sink intensity threshold of their respective second microhabitat geomorphological clusters. This will trigger a dynamic early warning mechanism, which will accurately pinpoint the core factors causing the lag in carbon sequestration growth by tracing back the causal chain analysis results of this unit: if ′、 'or If any value in the cluster is below 1.5 times the standard deviation of the cluster mean for two consecutive periods, it indicates that the unreasonable vegetation combination structure has caused competitive inhibition. If environmental covariates such as soil respiration flux or organic matter content are continuously abnormal, it points to soil condition limitations. If the remediation planting type deviates significantly from the recommended pattern of the actual landform cluster, it suggests that the remediation measures have been inappropriately selected.

[0075] This paper proposes a method for monitoring carbon sequestration and enhancement effects in multi-regional ecological restoration. The overall process revolves around "unit division, data collection, coefficient calculation, type determination, cluster correction, and monitoring and early warning." First, based on four key environmental factors—climate, topography, vegetation, and soil—microhabitat units are divided using field surveys, high-resolution remote sensing, and GIS technology. Community data such as vegetation cover, physiological processes, and environmental covariates are collected. Then, the carbon sequestration coefficient of the first community is calculated using a multiple linear regression equation. Combined with the importance of community features output by a random forest model, the carbon sequestration coefficient of the second community is corrected, and the community synergistic gain rate is calculated. Subsequently, based on these two factors, the restoration planting type for each microhabitat unit is determined. The geomorphological data is clustered using the KMEANS method to obtain the first microhabitat geomorphological cluster. After correction by the restoration planting type, the second microhabitat geomorphological cluster is obtained. Finally, the carbon sequestration intensity after restoration is collected, the carbon sequestration intensity threshold within the cluster is calculated and compared, and an early warning is triggered if it falls below the threshold, thus achieving systematic and precise monitoring of carbon sequestration and enhancement effects.

[0076] A second community feature coefficient was calculated by combining the importance of community features with the first community feature coefficient. The second community feature coefficient is based on the first community feature coefficient and also incorporates the community feature importance from a random forest model. Standardized scores characterize the nonlinear contribution of different vegetation features to carbon sequestration. The fusion ratio is dynamically adjusted using the random forest out-of-bag error and the linear model's coefficient of determination: when both models have high reliability, the correction is moderately strengthened to capture complex interaction effects; when model reliability is low, the correction is weakened to avoid bias. The resulting second community feature coefficient retains interpretable mechanistic information while accurately capturing nonlinear patterns, providing scientific and practical quantitative support for subsequent remediation decisions.

[0077] The second community characteristic coefficient is calculated by combining the importance of community features with the first community characteristic coefficient. The second community characteristic coefficient is based on the first community characteristic coefficient and also incorporates the importance of community features from the random forest model. The standardized scores characterize the nonlinear contribution of different vegetation features to carbon sequestration. The fusion ratio is dynamically adjusted by using the random forest bag out-of-bag error and the linear model determination coefficient. The final generated second community characteristic coefficient retains interpretable mechanistic information and accurately captures nonlinear laws, providing scientific and practical quantitative support for subsequent restoration decisions.

[0078] Based on the community synergistic gain rate (SCG) and the second community characteristic coefficient, the remediation planting type of the microhabitat unit is determined. SCG serves as the core judgment indicator. First, the overall effect direction of the community is clarified, and then the specific driving or competitive source is located by combining the second community characteristic coefficient. In synergistic scenarios, only significantly non-zero second community characteristic coefficients are retained, and based on… '、 '、 The significantly non-zero second community characteristic coefficient determines the type of vegetation to be restored; in a competitive scenario, only the second community characteristic coefficient is retained. '、 '、 The significantly non-zero second community characteristic coefficient determines the type of vegetation to be restored. This process completely avoids the drawbacks of a one-size-fits-all approach to restoration, ensuring that the planting type of each microhabitat unit can match its ecological conditions and maximize the synergistic effect of carbon sequestration.

[0079] The first microhabitat landform cluster was modified by adjusting the planting type to obtain the second microhabitat landform cluster. The first microhabitat landform cluster was generated solely based on geomorphological data such as elevation, slope, and aspect using KMEANS clustering. While this achieved classification based on geographical similarity, it did not consider the actual needs of ecological restoration. The second microhabitat landform cluster, based on similar geomorphological features, further groups microhabitat units adapted to the same planting type for restoration into one category. This means that units within the same cluster not only share similar geographical conditions but also have consistent restoration goals and ecological foundations. This modification makes subsequent carbon sink intensity analysis more targeted.

[0080] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, the phrase "comprising an element defined as..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0081] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for monitoring carbon sequestration and enhancement effects based on multi-regional ecological restoration, characterized in that: Includes the following steps: Step S1: Based on the geomorphological features of the ecological restoration area, divide the ecological restoration area into several microhabitat units; collect community data of the microhabitat units to obtain microcommunity data; The specific division of the microhabitat units is based on the significant spatial differences of four key environmental factors: climate, topography, vegetation, and soil; the microcommunity data includes tree canopy coverage. Shrub canopy coverage Herbaceous community coverage and standard carbon sequestration intensity; Step S2: Establish a multiple linear regression equation based on the microbial community data and calculate the first community characteristic coefficient; input the microbial community data into the random forest model and calculate the importance of community characteristics; Step S3: Combine the importance of the community features with the first community feature coefficient using a dynamic weighted fusion method to obtain the second community feature coefficient; perform community synergy calculation on the second community feature coefficient using the overproduction effect analysis method to obtain the community synergy gain rate; Step S4: Based on the community synergistic gain rate and the second community characteristic coefficient, obtain the remediation planting type of the microhabitat unit; collect the geomorphological data of the microhabitat unit, and cluster the geomorphological data of the microhabitat unit using the KMEANS method to obtain the first microhabitat geomorphological cluster; modify the first microhabitat geomorphological cluster using the remediation planting type to obtain the second microhabitat geomorphological cluster. Step S5: Based on the remediation planting type, ecological restoration is carried out on the microhabitat units, and the standard carbon sink intensity after restoration is collected to obtain the restored carbon sink intensity; by analyzing the distribution of restored carbon sink intensity within the second microhabitat landform cluster, the threshold of carbon sink intensity within the cluster is obtained; the restored carbon sink intensity of the microhabitat units in the second microhabitat landform cluster is compared with the threshold of carbon sink intensity within the cluster, and an early warning is issued if it is less than the threshold of carbon sink intensity within the cluster, thereby realizing the monitoring of carbon sequestration and enhancement effect; The step of establishing a multiple linear regression equation based on the microbial community data and calculating the first community characteristic coefficient includes the following specific steps: Collect tree canopy coverage over N periods Shrub canopy coverage Herbaceous community coverage and standard carbon sequestration intensity, for tree canopy coverage under the N periods. Shrub canopy coverage Herbaceous community coverage Perform a linear transformation to convert it into three main effect terms. , , and 3 interactive items , , : , , , , , These are the three main effect terms after transformation; after obtaining the main effect terms, interaction terms are constructed using the main effect terms: , , , for The average value, for The average value, for The average value; Combining the three main effect terms , , and 3 interactive items , , Based on the standard carbon sink intensity, a multiple linear regression equation was constructed, and the first community characteristic coefficient was calculated. The first community characteristic coefficient includes: , , , , , , : ； Where Y is the standard carbon sequestration intensity. Based on carbon sequestration capacity, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, for The first community characteristic coefficient, For the first residual term, ~N(0, ); The process of inputting the microbial community data into a random forest model and calculating the importance of community features includes the following steps: The microbial community data for N periods , , , , , After Z-score standardization, the following is obtained , , , , , ,Will , , , , , Standard carbon sequestration intensity Y, as the independent variable, is used as the dependent variable and input into the random forest model for training. Decision trees are generated through Bootstrap sampling, resulting in a decision tree set T={ , ,..., ,..., There are a total of M decision trees; For each tree and microbial community data characteristics Calculate the cumulative purity increase for each feature across all split points: ； in, Represents the j-th feature in the microbial community data. The importance of features in the m-th decision tree. For trees The set of split points, h is In the context of a split point h, I() is an indicator function, i.e., the characteristic index of the split point h. The index j is 1 when it is equal to the index j of the microbial community data feature, and 0 when they are not. The feature index of the split point h, The reduction in Gini impurity at the splitting point h. , Let the Gini impurity be the parent node of the split point h. The child node of the split point h, Let be the Gini impurity of the child node `child` of the split point `h`. and , where h represents the number of samples for the parent and child nodes of the split point, and j is the index of the microbial community data features; Then features The importance of community features in the random forest model is as follows: ； in, Let j be the j-th feature of the microbial community data. In a random forest model, the importance of community features is given by M, where M is the total number of decision trees and m is the m-th decision tree. Represents the j-th feature in the microbial community data. Importance of community features in the m-th decision tree; The method of combining the importance of community features with the first community feature coefficient using a dynamic weighted fusion method to obtain the second community feature coefficient includes the following steps: The second community feature coefficient is obtained by combining the importance of the community features with the first community feature coefficient using a dynamic weighted fusion method. ； in, Let j be the second community characteristic coefficient of the j-th feature in the microbial community data. 0, Let be the first community characteristic coefficient of the j-th feature in the microbial community data. It is a nonlinear intensity factor. Let j be the j-th feature of the microbial community data. The importance of community features in random forest models This represents the mean importance of all features in the microbial community data within the random forest model. This represents the standard deviation of the importance of all features in the microbial community data within the random forest model. This refers to the out-of-bag error of the random forest model. The coefficient of determination is represented by the coefficient of determination in a multiple linear regression model. Based on the second community characteristic coefficient, the modified multiple linear regression equation is obtained; The step of performing community synergy calculation on the characteristic coefficients of the second community using the overproduction effect analysis method to obtain the community synergy gain rate includes the following steps: Calculating the carbon sequestration intensity when only trees are present, the shrub crown cover and herbaceous community cover are 0. Substituting this into the modified multiple linear regression equation: , Standard carbon sequestration intensity for trees; Calculating the carbon sink intensity when only shrubs exist, the tree canopy cover and herbaceous community cover are 0. Substituting this into the modified multiple linear regression equation: , Standard carbon sequestration intensity for shrubs; Calculating the carbon sink intensity when only herbaceous plants exist, the tree canopy cover and herbaceous community cover are both 0. Substituting this into the corrected multiple linear regression equation: , Standard carbon sequestration intensity of vegetation; The community synergy gain rate was obtained by performing community synergy calculations on the characteristic coefficients of the second community using the overproduction effect analysis method. ； Wherein, SCG is the community-coordinated gain rate. This represents the actual standard carbon sequestration intensity. The standard carbon sequestration intensity of trees, Standard carbon sequestration intensity for shrubs, The standard carbon sequestration intensity of vegetation, To prevent division by zero, the value is set to max( ,0.01*( )); The step of analyzing the distribution of carbon sink intensity within the second microhabitat landform cluster to obtain the intra-cluster carbon sink intensity threshold includes the following specific steps: For each second microhabitat geomorphic cluster, the distribution of remediation carbon sink intensity within the cluster is analyzed to calculate the mean and standard deviation of the remediation carbon sink intensity. The formula for calculating the mean remediation carbon sink intensity is as follows: ； in, denoted as the mean carbon sink intensity for the restoration of the second microhabitat geomorphic cluster, WS represents the number of microhabitat units in the second microhabitat geomorphic cluster, and ws represents the index of the microhabitat unit. The remediation carbon sink intensity of the ws-th microhabitat unit, Let ws be the area of ​​the w-th microhabitat unit; The formula for calculating the standard deviation of carbon sequestration intensity is: ； in, denoted as , where WS is the standard deviation of the carbon sink intensity for the restoration of the second microhabitat geomorphic cluster, WS is the number of microhabitat units in the second microhabitat geomorphic cluster, and ws is the index of the microhabitat unit. The remediation carbon sink intensity of the ws-th microhabitat unit, The average carbon sink intensity for the restoration of the second microhabitat landform cluster; For each second microhabitat landform cluster, the intra-cluster carbon sink intensity threshold is calculated by combining the mean and standard deviation of the restored carbon sink intensity: ； in, The threshold value for intra-cluster carbon sink intensity. This represents the average carbon sequestration intensity for the restoration of the second microhabitat landform cluster. The standard deviation of the carbon sink intensity for the restoration of the second microhabitat landform cluster. This is the threshold adjustment coefficient.

2. The method for monitoring carbon sequestration and enhancement effects based on multi-regional ecological restoration according to claim 1, characterized in that: The process of obtaining the modified multiple linear regression equation based on the second community characteristic coefficient includes the following steps: The multiple linear regression equation is corrected by using the second community characteristic coefficient, resulting in the corrected multiple linear regression equation: ； Where Y is the standard carbon sequestration intensity. Based on carbon sequestration capacity, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, for The second community characteristic coefficient, This is the second residual term.

3. The method for monitoring carbon sequestration and enhancement effects based on multi-regional ecological restoration according to claim 1, characterized in that: The process of obtaining the remediation planting type for a microhabitat unit based on the community synergistic gain rate and the second community characteristic coefficient includes the following specific steps: When SCG > 0, the actual carbon sequestration intensity of the community is greater than the sum of the individual carbon sequestration intensities, indicating a community synergistic effect within this microhabitat unit. '、 '、 Perform a t-test, retain only the significantly non-zero second community characteristic coefficient, and according to... '、 '、 The significantly non-zero second community characteristic coefficient determines the type of vegetation to be restored; When SCG < 0, competition inhibition exists in the community. '、 '、 Perform a t-test, retain only the significantly non-zero second community characteristic coefficient, and according to... '、 '、 The significantly non-zero second community characteristic coefficient determines the type of restored vegetation; when SCG=0, this indicates that the community is at a balance between cooperation and competition, which is beneficial for the restoration of vegetation. '、 '、 '、 '、 '、 Perform a t-test and based on '、 '、 '、 '、 '、 The significantly non-zero second community characteristic coefficient determines the type of vegetation to be restored.

4. The method for monitoring carbon sequestration and enhancement effects based on multi-regional ecological restoration according to claim 3, characterized in that: The step of clustering the geomorphic data of the microhabitat units using the KMEANS method to obtain the first microhabitat geomorphic cluster includes the following specific steps: The KMEANS clustering algorithm was used to perform cluster analysis on the above-mentioned geomorphic data. The geomorphic data of all corresponding microhabitat units were standardized by Z-score to eliminate the differences in the dimensions and numerical ranges of different indicators. The elbow method was used to analyze the clustering error of the geomorphic data of microhabitat units, and the elbow position where the error decrease trend changed from steep to gentle was found to determine the optimal number of clusters K. After determining the number of clusters K, the geomorphic data of microhabitat units were clustered by the KMEANS clustering method. Through iterative calculation, the microhabitat units were clustered to obtain the first microhabitat geomorphic cluster.

5. A method for monitoring carbon sequestration and enhancement effects based on multi-regional ecological restoration according to claim 4, characterized in that: The process of modifying the first microhabitat landform cluster using the remediation planting type to obtain the second microhabitat landform cluster includes the following specific steps: Based on the first microhabitat landform cluster, microhabitat units within the first microhabitat landform cluster that are identified as having the same restoration planting type are grouped into one category, thus completing the modification of the first microhabitat landform cluster and obtaining the second microhabitat landform cluster, ultimately resulting in K second microhabitat landform clusters. .

Images