Tea garden ecosystem health evaluation method and system based on landscape multifunctionality

Abstract

The invention discloses a tea garden ecosystem health evaluation method and system based on landscape multifunctionality. The method belongs to the technical field of tea garden management and ecological environment evaluation, and comprises the following steps: determining tea garden ecological system health evaluation indexes in four aspects based on an'vitality-organization power-restoring power 'framework and in combination with an ecological system service theory; a three-layer hierarchical structure model is constructed through an analytic hierarchy process, index weights are calculated in combination with an entropy weight method, feature importance sorting is performed on indexes by adopting a random forest algorithm, and an evaluation index system is constructed; carrying out dimension reduction processing on the indexes by utilizing a principal component analysis method and calculating weights; and based on the evaluation index and the weight after dimension reduction processing, calculating a comprehensive health score of the tea garden ecosystem by adopting a weighted summation method, and performing grading. By integrating various theories and methods, the health condition of the tea garden ecosystem is evaluated comprehensively and objectively, a scientific basis is provided for management and protection of the tea garden, and sustainable development of the tea garden ecosystem is promoted.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of tea garden management and ecological environment evaluation, and more particularly to a tea garden ecosystem health evaluation method and system based on landscape multifunctionality. BACKGROUND

[0002] As an important agricultural ecosystem, the sustainable development of tea garden involves multiple dimensions such as economy, ecology, landscape aesthetics, and social culture. In the traditional tea garden evaluation system, the technical means mainly focus on yield, economic benefit, and soil fertility, etc. This evaluation mode takes single economic output as the core, and seriously lacks comprehensive consideration of the landscape multifunctionality of tea garden, resulting in the serious underestimation of the ecosystem service function of tea garden. For example, the traditional evaluation often ignores the ecological value of tea garden vegetation in soil and water conservation, biological habitat construction, and the social and cultural value of tea garden landscape in leisure tourism and cultural heritage. In recent years, with the deepening of the concept of ecological agriculture and the development of landscape ecology theory, the landscape functions of tea garden, such as soil and water conservation, biodiversity protection, and leisure tourism, have gradually attracted the attention of academia and industry. However, the existing evaluation methods still have significant defects: first, they mostly use single indicators for evaluation, which cannot fully reflect the sustainable development level of tea garden in multiple dimensions such as ecology, economy, and society; second, they lack a comprehensive evaluation index system, making it difficult to systematically assess the internal relationship between the health of tea garden ecosystem and the landscape multifunctionality. For example, the existing soil fertility evaluation index can only reflect the soil nutrient status, but cannot reflect the ecological function of soil and surrounding landscape elements under the synergistic effect. Therefore, it is an urgent problem for those skilled in the art to provide an ecosystem health evaluation method and system that can comprehensively consider the landscape multifunctionality of tea garden. SUMMARY

[0003] Therefore, the present application provides a tea garden ecosystem health evaluation method and system based on landscape multifunctionality, which comprehensively and scientifically evaluates the ecological, economic, social, and landscape functions of tea garden by constructing an "organization force-vigor-resilience-ecosystem service" evaluation framework and a comprehensive evaluation index system, and provides a scientific basis for the health management and sustainable use of tea garden ecosystem.

[0004] To achieve the above-mentioned purpose, the present application adopts the following technical solutions: On the one hand, the present application provides a tea garden ecosystem health evaluation method based on landscape multifunctionality, comprising: Based on the "vigor-organization force-resilience" framework, combined with the theory of ecosystem service, the evaluation indexes of tea garden ecosystem health in four aspects of vigor, organization force, resilience, and ecosystem service are determined; A three-layer hierarchical structure model is constructed by an analytic hierarchy process, index weight calculation is performed in combination with an entropy method, a random forest algorithm is used to sort feature importance of the tea garden ecosystem health evaluation indexes, and a tea garden ecosystem health evaluation index system is constructed based on the feature importance sorting; Obtaining evaluation index data and performing standardization processing on the evaluation index data; Performing dimensionality reduction processing on the evaluation indexes by using a principal component analysis method and calculating weights; Calculating the dimensionality-reduced evaluation index values by using the standardization-processed evaluation index data; Based on the evaluation indexes and the weights obtained after the dimensionality reduction processing, a health comprehensive score of the tea garden ecosystem is calculated by using a weighted summation method; According to the health comprehensive score, the health status of the tea garden ecosystem is graded.

[0005] Preferably, a three-layer hierarchical structure model is constructed by an analytic hierarchy process, index weight calculation is performed in combination with an entropy method, a random forest algorithm is used to sort feature importance of the tea garden ecosystem health evaluation indexes, and a tea garden ecosystem health evaluation index system is constructed based on the feature importance sorting, including: Taking tea garden ecosystem health evaluation as a target layer, first-level indexes are divided; According to the characteristics of the tea garden ecosystem, the target layer is decomposed into four first-level indexes; For each first-level index, second-level indexes are further refined; The target layer, the first-level indexes and the second-level indexes are constructed into a three-layer hierarchical structure model according to a hierarchical relationship; The original data is standardized to make the data in the same order of magnitude.

[0006] According to the standardized data, the information entropy of each index is calculated; Based on the information entropy, the redundancy is determined; According to the redundancy, the weight coefficient of each index is calculated; The weight obtained by the analytic hierarchy process and the weight coefficient obtained by the entropy method are integrated to obtain the comprehensive weight of each index; Based on the comprehensive weight, the indexes are screened to obtain a preliminary index set; A random forest algorithm is used to construct a model, and the random forest model is trained using training data; Based on the trained random forest model, the importance score of each feature variable is obtained; According to the importance score, each evaluation index is sorted, and a key index is screened from the preliminary index set by setting a threshold; Based on the key index, a tea garden ecosystem health evaluation index system is constructed.

[0007] Preferably, the determined evaluation indexes of tea garden ecosystem health in four aspects of vitality, organization, resilience and ecosystem services include: The vitality aspect includes net primary productivity and leaf area index; The organization aspect includes spread index, aggregation index, cohesion index, Shannon diversity index, tea tree canopy vertical heterogeneity index; The resilience aspect includes associated plant species richness, tea tree genetic diversity index, land use type resilience coefficient; The ecosystem services include water conservation capacity, soil conservation amount, tea production, and fresh leaf phenol-ammonia ratio.

[0008] Preferably, the principal component analysis method is used for dimension reduction processing of the evaluation indexes, and the weights are calculated, including: The covariance matrix is calculated, and the elements of the covariance matrix are: ; Wherein, Z ij represents the standardized value of the ith sample on the jth index; The matrix form can be represented as: ; Wherein, Z T is the transpose of the standardized matrix Z; The eigenvalue decomposition is performed on the covariance matrix : ; Wherein: is the eigenvector matrix, v j is the unit orthogonal vector corresponding to the direction of the jth principal component; V T is the transpose of the standardized matrix V; is a p×p diagonal matrix, and the elements on the diagonal are the eigenvalues of , representing the variance corresponding to the principal component; The variance contribution rate of the jth principal component is calculated: ; Wherein, λ j is the jth eigenvalue, ; The cumulative variance contribution rate is calculated: ; The smallest k is selected to make the cumulative contribution rate reach a preset threshold; For the kth original variable, the absolute value of its load on all m principal components is calculated: ; Wherein: v kj represents the load of the kth variable on the jth principal component; The principal components are constructed, and the expression of the jth principal component Y j is: ; Wherein, is the jth eigenvector, and Z is the standardized matrix; The first k eigenvectors are combined into a transformation matrix , and the principal component matrix is: ; Wherein, each row corresponds to the coordinates of a sample in the k-dimensional principal component space; For the kth index, the formula chain is summarized as: ; Wherein: m is the number of principal components that meet the cumulative contribution rate requirement.

[0009] Preferably, the evaluation index value after dimension reduction is calculated by using the standardized evaluation index data, including: The land restoration coefficient calculation formula is: ; Wherein, RC i represents the restoration coefficient of the ith grid, NDVI i represents the NDVI value of the ith grid, NDVI_mean j represents the NDVI average value of the land class k where the ith grid is located, RC k represents the restoration coefficient of the set land class k ; According to the revised universal soil loss equation, the actual soil erosion amount and the potential soil erosion amount are calculated respectively, and the difference between the actual soil erosion amount and the potential soil erosion amount is the soil conservation amount; The water yield of the ecological service InVEST model is used to evaluate the change of the water source conservation function. The tea production amount calculation formula is: ; Wherein, represents the tea production amount corresponding to each grid, represents the leaf area index of each grid, represents the total leaf area index corresponding to the tea garden, representing the total annual tea production; The formula for calculating the fresh leaf ammonia-phenol ratio is: ; TP refers to the total phenolic content; and TA refers to the total free amino acid content.

[0010] Preferably, based on the evaluation indexes and weights obtained after dimension reduction processing, the formula for calculating the health comprehensive score of the tea garden ecosystem is calculated by using the weighted summation method: ; wherein, W represents the weight value of the index; a, b,..., x are the standardized data of the original indexes.

[0011] Preferably, the health comprehensive score is used to grade the health status of the tea garden ecosystem, including: when HI is greater than or equal to 0.8, the health status of the tea garden ecosystem is determined to be healthy; when 0.6 is less than HI and HI is less than 0.8, the health status of the tea garden ecosystem is determined to be sub-healthy; when HI is less than 0.6, the health status of the tea garden ecosystem is determined to be unhealthy.

[0012] On the other hand, the present application also provides a tea garden ecosystem health evaluation system based on landscape multi-functionality, comprising: an index collection module for determining tea garden ecosystem health evaluation indexes in four aspects of vitality, organizational force, resilience and ecosystem service based on the "vitality-organizational force-resilience" framework and combining the ecosystem service theory; an index system construction module for constructing a three-layer hierarchical structure model by using the analytic hierarchy process, calculating the index weights by using the entropy weight method, sorting the feature importance of the tea garden ecosystem health evaluation indexes by using the random forest algorithm, and constructing the tea garden ecosystem health evaluation index system based on the feature importance sorting; a data collection and processing module for obtaining evaluation index data and performing standardized processing on the evaluation index data; a dimension reduction module for performing dimension reduction processing on the evaluation indexes by using the principal component analysis method and calculating the weights; a calculation module for calculating the dimension-reduced evaluation index values by using the standardized evaluation index data; a comprehensive evaluation module for calculating the health comprehensive score of the tea garden ecosystem by using the weighted summation method based on the evaluation indexes and weights obtained after dimension reduction processing; a grading module for grading the health status of the tea garden ecosystem according to the health comprehensive score.

[0013] Compared with the prior art, the application provides a tea garden ecosystem health evaluation method and system based on landscape multifunctionality, which has the following beneficial effects: (1) Based on the "vigor-organization-restoration" framework and combined with the ecosystem service theory, the evaluation indexes are determined from four key aspects, various characteristics and functions of the tea garden ecosystem are comprehensively considered, and a solid theoretical basis is provided for accurately evaluating the health status of the tea garden ecosystem.

[0014] (2) The three-layer hierarchical structure model is constructed by using the analytic hierarchy process, and the index weight is calculated by combining the entropy weight method, so that the weight determination process considers both the expert experience and the objectivity based on data information, and the rationality and accuracy of the weight distribution are improved.

[0015] (3) The random forest algorithm is used to sort the feature importance of the evaluation indexes, which can filter out the key indexes that have greater influence on the health status of the tea garden ecosystem, so as to construct a more simplified and efficient evaluation index system, and avoid the interference of too many irrelevant indexes.

[0016] (4) After the evaluation index data is standardized, the principal component analysis method is used for dimension reduction processing, which not only eliminates the dimension influence, but also simplifies the data structure and reduces the calculation complexity, while retaining most of the effective information of the original data, which is beneficial to more accurate calculation and analysis in the subsequent process.

[0017] (5) The weighted summation method is used to calculate the health comprehensive score of the tea garden ecosystem, and the health status of the tea garden ecosystem is classified according to the score, so as to realize the quantitative evaluation of the health status of the tea garden ecosystem, and facilitate the managers to intuitively understand the health status of the tea garden and formulate corresponding management strategies. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the drawings needed to be used in the embodiment or prior art description will be briefly introduced. Obviously, the drawings in the following description are only embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of the provided drawings.

[0019] Figure 1 The flowchart provided by the application.

[0020] Figure 2 The technical roadmap of the application.

[0021] Figure 3 The structural schematic diagram provided by the application. DETAILED DESCRIPTION

[0022] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative work are within the protection scope of the present application.

[0023] In the present application, the tea garden ecosystem health evaluation method based on landscape multifunctionality applies the "vigor-organization-resilience-ecosystem service" ecosystem health evaluation framework to the tea garden ecosystem. In the traditional VORS framework, the selection of indexes depends on land use data, while in the present evaluation method, the selection of indexes pays more attention to distinguishing the differences and health conditions of the tea garden ecosystem.

[0024] Figure 2 The flow of the tea garden ecosystem health evaluation method based on landscape multifunctionality is presented. The figure discloses that the VORS model is an evolution based on the traditional VOR model. In addition to the ability to evaluate the impact of the outside world on the ecosystem, the "ecosystem service" dimension is added, so that the VORS model can evaluate the function of the ecosystem. The figure explains in detail the application of each index (such as primary productivity, leaf area index, aggregation index, etc.) in the evaluation process, and shows how to reflect the function of ecosystem service through adjusting the service and supply service.

[0025] Specifically, the embodiment of the present application discloses a tea garden ecosystem health evaluation method based on landscape multifunctionality, as shown in Figures 1-2 The embodiment of the present application discloses a tea garden ecosystem health evaluation method based on landscape multifunctionality, as shown in Based on the "vigor-organization-resilience" framework, combined with the ecosystem service theory, the tea garden ecosystem health evaluation indexes in the four aspects of vigor, organization, resilience and ecosystem service are determined.

[0026] Specifically, based on the "vigor-organization-resilience" framework, combined with the theory of ecosystem services, an evaluation index system of tea garden ecosystem health is constructed from four aspects of vigor, organization, resilience and ecosystem services. Vigor refers to the metabolic capacity and productivity of the ecosystem, including primary productivity and other indicators; organization refers to the structure and composition of the ecosystem, including species diversity, community structure and other indicators; resilience refers to the ability of the ecosystem to resist disturbance and recover, including recovery rate, resistance and other indicators; ecosystem services refer to the various benefits provided by the ecosystem to humans, including supply services (such as tea yield, quality), regulation services (such as water conservation, soil conservation), support services (such as biodiversity maintenance) and cultural services (such as landscape aesthetic value) and other services.

[0027] According to the actual situation, primary productivity, species richness, tea yield, water conservation and other indicators are selected to construct the evaluation index system. Using the VORS evaluation framework, an evaluation index system of tea garden ecosystem health is constructed (Table 1), which measures the health status of tea garden ecosystem in the dimensions of function, structure, process and service.

[0028] Table 1 Evaluation index of certain ecosystem health

[0029] A three-level hierarchical structure model is constructed by the analytic hierarchy process, and the entropy weight method is used to calculate the index weight. The random forest algorithm is used to sort the feature importance of the evaluation index of tea garden ecosystem health, and the evaluation index system of tea garden ecosystem health is constructed based on the feature importance sorting; The evaluation index data is obtained and standardized; The principal component analysis method is used to reduce the dimension of the evaluation index and calculate the weight; The evaluation index value after dimension reduction is calculated based on the standardized evaluation index data; Based on the evaluation index and weight obtained after dimension reduction, the weighted sum method is used to calculate the comprehensive health score of tea garden ecosystem; According to the health comprehensive score, the health status of tea garden ecosystem is classified.

[0030] Further, a three-level hierarchical structure model is constructed by the analytic hierarchy process, combined with the entropy weight method for index weight calculation, and the random forest algorithm is used to sort the feature importance of the evaluation index of tea garden ecosystem health, and the evaluation index system of tea garden ecosystem health is constructed based on the feature importance sorting, including: Taking the evaluation of tea garden ecosystem health as the target layer, the first-level indicators are divided; According to the characteristics of tea garden ecosystem, the target layer is decomposed into four first-level indicators; For each primary indicator, further refine the secondary indicators; Construct a three-level hierarchical structure model according to the hierarchical relationship of target layer, primary indicators and secondary indicators; Standardize the original data to the same order of magnitude for subsequent calculations.

[0031] According to the standardized data, calculate the information entropy of each indicator; Determine the redundancy based on information entropy; Calculate the weight coefficient of each indicator according to the redundancy; Integrate the weight obtained by AHP and the weight coefficient obtained by entropy weight method to get the comprehensive weight of each indicator; Based on the comprehensive weight, screen the indicators to get the preliminary indicator set; Use the random forest algorithm to build the model, and use the training data to train the random forest model; Based on the trained random forest model, get the importance score of each feature variable; According to the importance score, sort the evaluation indicators, and set a threshold to screen the key indicators from the preliminary indicator set; Based on the key indicators, construct the tea garden ecosystem health evaluation index system.

[0032] Specifically, (1) the specific process of constructing a three-level hierarchical structure model based on AHP: Determine the target layer: The target layer is "tea garden ecosystem health evaluation". This is the ultimate goal of the evaluation work, and also the starting point of building a hierarchical structure model.

[0033] Divide the primary indicators: According to the characteristics of tea garden ecosystem, the target layer is divided into four primary indicators: Vigor, Organization, Resilience, and Ecosystem Service. These four primary indicators comprehensively cover all key aspects of tea garden ecosystem.

[0034] Refine the secondary indicators: For each primary indicator, further refine the specific secondary indicators. For example: Vigor: includes net primary productivity (NPP) and leaf area index (LAI) under ecosystem productivity.

[0035] Organization: including CONTAG, AI, AWMPFD, COHESION under landscape structure stability, and SHDI under community diversity, as well as vertical heterogeneity index of tea tree canopy.

[0036] Resilience: including associated plant species richness and tea tree genetic diversity index under biodiversity support, and land use type resilience coefficient under anti-interference ability.

[0037] Ecosystem Service: including water conservation and soil conservation under regulation service, and tea yield and fresh leaf phenol-ammonia ratio (quality core index) under supply service.

[0038] ‌Building a hierarchical structure model: Organize the target layer, first-level indicators and second-level indicators according to the hierarchical relationship to form a clear three-layer hierarchical structure model. This model provides a framework and foundation for subsequent evaluation work.

[0039] (2) The specific process of combining entropy weight method to calculate the weight of indicators: ‌Data standardization: Since the dimensions and orders of magnitude of each indicator may be different, the original data need to be standardized first to be on the same order of magnitude, which is convenient for subsequent calculation.

[0040] ‌Calculate information entropy: According to the standardized data, calculate the information entropy of each indicator. Information entropy is an important indicator to measure the dispersion degree of indicator data. The smaller the information entropy, the greater the dispersion degree of the indicator, and the more information it provides.

[0041] ‌Determine redundancy: Redundancy is inversely proportional to information entropy, that is, the smaller the information entropy, the greater the redundancy. Redundancy reflects the importance of the indicator in the evaluation system.

[0042] ‌Calculate weight coefficient: Calculate the weight coefficient of each indicator according to the redundancy. The size of the weight coefficient directly reflects the importance of the indicator in the evaluation system.

[0043] ‌Comprehensive weight: Combine the weight obtained by AHP with the weight obtained by entropy weight method to get the comprehensive weight of each indicator. This comprehensive weight more objectively and accurately reflects the importance of the indicator in the tea garden ecosystem health evaluation system.

[0044] (3) The specific process of using random forest algorithm to sort the feature importance of tea garden ecosystem health evaluation indicators: Data Preparation: Organize the relevant data for tea garden ecosystem health assessment into a format suitable for input into the random forest algorithm. This includes the target variable (i.e., the result of the tea garden ecosystem health assessment) and various evaluation indicators as feature variables.

[0045] Building Random Forest Model: Use the random forest algorithm to build the model. During the construction process, multiple decision trees are constructed by randomly sampling the data and features, forming a random forest.

[0046] Training Model: Train the random forest model using the prepared data. During the training process, the model learns the relationship between each feature variable and the target variable.

[0047] Feature Importance Evaluation: After training, the random forest model gives an importance score for each feature variable. This score is calculated based on the contribution of the feature variable in the model's prediction process.

[0048] Sorting and Filtering: Sort the various evaluation indicators according to their importance scores. By setting a threshold or selecting indicators with higher importance scores, we can filter out the key indicators that have the greatest impact on the tea garden ecosystem health assessment.

[0049] Specifically, the determined evaluation indicators for the four aspects of vitality, organizational capacity, resilience, and ecosystem services of tea garden ecosystem health include: The vitality dimension selects specific indicators from the ecosystem productivity level, mainly including net primary productivity of vegetation and leaf area index. The organizational capacity dimension selects specific indicators from the landscape structure stability level and community diversity level. The landscape structure stability level mainly includes contagion index, aggregation index, area-weighted mean patch fractal dimension index, and cohesion index. The community diversity level mainly includes Shannon diversity index and tea tree canopy vertical heterogeneity index.

[0050] The resilience dimension selects specific indicators from the biodiversity support level and disturbance resistance level. The biodiversity support level mainly includes associated plant species richness and tea tree genetic diversity index. The disturbance resistance level mainly includes land use type resilience coefficient.

[0051] The ecosystem service dimension selects specific indicators from the regulation service level and supply service level. The regulation service level mainly includes water conservation and soil conservation. The supply service level mainly includes tea yield and tea quality.

[0052] In order to eliminate the influence of variable dimension difference on covariance calculation, data standardization (centering and scaling) is performed: Calculate the mean and standard deviation of each variable: Let the original data matrix be (n samples, p variables), the mean of the jth variable is: ; In the formula, x ij represents the original value of the ith sample on the jth index.

[0053] The standard deviation is: ; Standardized data: subtract the mean from each data point and divide by the standard deviation to get the standardized matrix Z: ; At this time, the mean of each variable of Z is 0 and the variance is 1.

[0054] Further, the principal component analysis method is used to reduce the dimension of the evaluation index and calculate the weight, including: Calculate the covariance matrix: The elements of the covariance matrix are: ; Among them: Z ij represents the standardized value of the ith sample on the jth index.

[0055] The matrix form can be expressed as: ; Among them, Z T is the transpose of the standardized matrix Z (Note: if the sample is population data, use n as the denominator; if it is sample data, use n-1 unbiased estimate).

[0056] The specific method of eigenvalue decomposition is: Eigenvalue decomposition is performed on the covariance matrix : ; Among them: V T is the transpose of the standardized matrix V.

[0057] is the diagonal matrix of p x p , the elements on the diagonal are the eigenvalues of , representing the variance of the corresponding principal component; is the eigenvector matrix, v j is an orthonormal vector corresponding to the direction of the jth principal component. V is a p × p orthogonal matrix, and each column of it junit eigenvector. Eigenvector v j is the loading vector of the jth principal component. v kj represents the loading (coefficient) of the kth original variable on the jth principal component.

[0058] The principle of selecting the number of principal components: retain enough variance contribution rate (such as cumulative contribution rate ≥ 80%).

[0059] Calculate the variance contribution rate of the jth principal component: ; where λ j is the j th eigenvalue. (because the sum of the diagonal elements of the correlation matrix ∑ is p ).

[0060] Cumulative variance contribution rate: ; Select the smallest k to make the cumulative contribution rate reach the preset threshold (such as 0.8). The number of principal components selected ( k < p ).

[0061] Calculate the comprehensive loading of each original variable: For the kth original variable, calculate the absolute value of its loading on all m principal components (usually take the first m principal components whose cumulative contribution rate exceeds a certain threshold, such as 80% or 85%) and weight average according to the variance contribution rate of the corresponding principal component: ; where: v kj represents the k th variable on the j th principal component (from the eigenvector matrix V ), and v kj Take the absolute value because the positive and negative signs of the load usually only represent the directional relationship between the variable and the principal component (positive / negative correlation), and the absolute value size represents the influence degree. The weight pays attention to the influence strength. m represents the number of principal components selected ( m < p ). The key is that the weight calculation should be based on all principal components ( m = p ) or principal components that explain most of the variance ( m cumulative contribution rate > threshold). Use all principal components ( m = pAll information is theoretically used, but the calculation is slightly large; the small variance part is ignored by selecting partial principal components. In practice, all principal components or principal components with cumulative contribution rate close to 100% are often taken.

[0062] Construction of principal components: Principal component expression: the jth principal component Y j is a linear combination of original standardized variables: ; Wherein is the eigenvector of the jth principal component, j is the standardized matrix. Z

[0063] Principal component matrix: the first k eigenvectors are combined into a transformation matrix , then the principal component matrix is: ; Where each row corresponds to the coordinates of a sample in the k-dimensional principal component space. k Summary formula chain (for the jth index):

[0064] k ; Wherein: m Usually take p or the number of principal components that meet the cumulative contribution rate requirement.

[0065] Further, the evaluation index values after dimension reduction are calculated by using the standardized evaluation index data, including: Vigor (V) represents the metabolism or primary productivity of the ecosystem, which is measured by net primary productivity (NPP) and leaf area index (LAI).

[0066] Organization (O) refers to the stability of the landscape pattern of the ecosystem, including landscape connectivity (LC), important patch connectivity (IC) and landscape heterogeneity (LH). LC represents the connectivity of the entire ecosystem landscape, and in this embodiment, the contagion index (CONTAG), aggregation index (AI) are selected to measure the connectivity level of the ecosystem landscape; IC is the connectivity of the patch type which has a decisive influence on the health of the ecosystem. For the tea garden in this embodiment, the patch types that affect the health of the ecosystem are mainly woodland and water area, which play an important role in water conservation, maintenance of biodiversity and soil conservation, etc. The cohesion index (COHESION) is used to measure; LH represents the diversity of landscape types and combination ways in the landscape system, and the area-weighted average patch fractal dimension (AWMPFD) and Shannon diversity index (SHDI) are used to measure.​​​

[0067] Resilience (R) represents the ability of the ecosystem to recover after external disturbance. The closer the land use type is to the natural ecosystem, the stronger the resilience. The closer the land use type is to the natural ecosystem, the easier it is to recover to the original state after external disturbance. According to existing research, the resilience coefficient (RC) of different land use types is set according to the difficulty of recovery (Table 2), and the difference within the same type is modified using NDVI: ; In the formula, RC i represents the resilience coefficient of the i grid to be calculated, NDVI i represents the NDVI value of the i grid, NDVI_mean j represents the NDVI average value of the i grid, and RC k represents the resilience coefficient of the set land class. k k

[0068] Table 2 Resilience coefficient of different land use types Ecosystem services include: (1) Soil conservation: According to the Revised Universal Soil Loss Equation (RUSLE), the actual soil erosion amount and the potential soil erosion amount are calculated respectively, and the difference between them is the soil conservation amount. The SDR module of the InVEST model is used to estimate the soil conservation service. First, the potential soil loss amount is calculated, and the formula is as follows: ; In the formula: S e0x is the potential soil loss amount of grid x; R x is the rainfall erosion factor of grid x; K x is the soil erodibility factor of grid x; and LS x is the slope length factor of grid x.

[0069] Then, the actual soil loss amount is calculated, and the formula is as follows ; In the formula: USLE x is the actual soil loss amount of grid x; R x , K x , LS x have the same meaning and calculation method as above; and C x ​​is the vegetation cover and management factor of grid x; P x is the soil conservation measure factor of grid x.

[0070] The amount of soil loss reduction is equal to the potential amount of soil loss (S e0x ) minus the actual amount of soil loss (USLE x ): ; Considering the interception and deposition of sediment by vegetation during transport, the formula is modified by the sediment transport ratio, and the soil conservation service supply calculation formula is as follows: ; In the formula: SEC i is the amount of soil conservation of grid x; SDR i is the sediment transport ratio, which is the proportion of sediment reaching the outlet section of the watershed to the total amount of soil erosion in the watershed.

[0071] (2) Water conservation: The water yield (WaterYield) in the ecological service InVEST model for land ecosystem service function evaluation is used to evaluate the change of water conservation function. The water supply of each landscape type is calculated, as well as the influence of land cover change on annual surface water yield, and the watershed water yield is calculated according to the principle of water balance, the formula is as follows: ; ; ; ; In the formula: Y xj is the annual water yield of land use type j in grid unit x; AET xj is the actual evapotranspiration; P x is the precipitation; for arbor forest land, shrub land, grassland and farmland, the actual evapotranspiration is calculated based on the algorithm proposed by the water-heat coupling balance assumption; for water body and construction land, the actual evapotranspiration is directly calculated by potential evapotranspiration, and the upper limit is precipitation. ω is a physical parameter representing natural climate and soil properties, which is related to effective water content (AWC) of vegetation. Z is a seasonal constant that can reflect the spatial pattern of precipitation.

[0072] (3) Tea production: The significant correlation between leaf area index (LAI) and crop yield indicates that this index can not only accurately describe crop growth status but also effectively predict yield, thus having wide application value in the field of agricultural science. Based on county-level leaf area index and tea production data released by national statistical agencies, a tea production distribution map was drawn. In this embodiment, remote sensing imagery data acquired using Google Earth technology in 2020 was analyzed to accurately identify and locate the distribution areas of tea gardens. Based on this, the tea production of the region was further calculated. First, precise cropping was performed according to the LAI data of each township. Second, LAI attribute tables for each township were generated and exported, and then the total LAI value of each township was calculated and summarized. Finally, the tea production of each township was calculated separately using the following formula, and then the results were integrated. The calculation method is shown in the formula: ; in, This represents the tea production corresponding to each grid (30 meters × 30 meters). The leaf area index represents the area of ​​each grid cell. The total leaf area index represents the area of ​​a specific tea garden. This represents the total annual tea production of a certain region.

[0073] (4) Fresh leaf acetaminophen ratio: The phenol-to-amino acid ratio is a core indicator for evaluating the relative content of phenolic substances and amino acids in fresh tea leaves (such as fresh tea tree leaves). It is often used to characterize the biochemical characteristics and quality potential of agricultural products such as tea.

[0074] Phenolic substances: mainly including catechins (such as EGCG), flavonoids, etc., which dominate the bitterness and astringency of tea and its antioxidant activity.

[0075] Amino acids: Mainly theanine and glutamic acid, which give fresh leaves a fresh and sweet taste.

[0076] The higher the phenol-to-amino acid ratio, the more pronounced the bitterness and astringency of the fresh leaves; the lower the ratio, the more prominent the freshness and crispness. This indicator is often used to assess the suitability of tea for processing (such as the suitability for processing green or black tea) and to grade its quality.

[0077] The formula for calculating the phenol-to-amino acid ratio is: ; TP refers to the total phenolic content, usually expressed as "milligram gallic acid equivalents / gram dry weight (mg GAE / g)" or "mg / g". TA refers to the total free amino acid content, commonly expressed as "milligram theanine equivalents / gram dry weight (mg / g)" or directly as mass concentration.

[0078] Further, based on the evaluation indexes and weights obtained after dimension reduction processing, the calculation formula of the health comprehensive score of the tea garden ecosystem is calculated by using the weighted summation method: ; Wherein, W represents the weight value of the index; a, b,..., x is the standardized data of the original index.

[0079] Preferably, the health status of the tea garden ecosystem is graded according to the health comprehensive score, including: When HI is greater than or equal to 0.8, the health status of the tea garden ecosystem is determined to be healthy; When 0.6 is less than HI and is greater than 0.8, the health status of the tea garden ecosystem is determined to be sub-healthy; When HI is less than 0.6, the health status of the tea garden ecosystem is determined to be unhealthy.

[0080] Furthermore, the ecosystem health at the grid scale of a tea garden was calculated and its spatial distribution pattern was analyzed. The influence of soil erosion modulus and groundwater depth on the spatial heterogeneity of ecosystem health was explored through geographically weighted regression (GWR). Based on this, a structural equation model was used to analyze the correlation between ecosystem health and economic development and ecological policy. The specific analysis process was as follows: (1) Ecosystem health, economic development, and ecological policy were taken as latent variables; (2) The indicators for measuring the latent variables were taken as observed variables. The observed variables for the ecosystem are: vitality, organization, resilience, and ecosystem services; the observed variables for economic development are: ① Net present value of comprehensive tea garden income (reflecting the economic benefits of the tea garden throughout its entire life cycle, affected by factors such as tea age, labor costs, and raw tea prices); ② Labor cost ratio (labor prices are negatively correlated with net present value; excessively high labor costs can lead to operating losses); ③ Organic / low-carbon certification coverage (certified tea gardens can increase profits by reducing pesticide input and production costs); ④ Tea garden land utilization rate (intercropping with green manure (such as soybeans and rapeseed) can improve soil fertility and indirectly reduce fertilizer costs); ⑤ Potential value of carbon sequestration (the carbon sequestration capacity of ecological tea gardens (e.g., 43.01 t)). CO2 / ha); convertible into carbon trading revenue); the observed variables for ecological policies are: ① the proportion of ecological tea gardens (under the policy promotion, the proportion of ecological tea gardens in Fujian Province has reached 83%, reflecting the policy implementation strength); ② the rate of organic fertilizer replacing chemical fertilizer (promoting the "organic fertilizer + green manure" model (such as intercropping soybeans for nitrogen fixation); reducing soil acidification); ③ the coverage rate of green prevention and control technologies (the application rate of technologies such as biological control and physical control (over 60% nationwide); reducing pesticide residues); ④ the number of ecological restoration projects (such as returning tea plantations to forests, vegetation restoration, etc., to improve the surrounding environment of tea gardens); ⑤ the completeness of soil and water conservation measures (mountain tea gardens need to build ring roads, terraces, and other facilities to reduce soil erosion). Based on the path analysis results of the structural equation model, the correlation between ecosystem health, economic development, and ecological policies is explored. The analysis results show that the tea garden has problems such as declining species diversity and insufficient resilience. It is recommended to take ecological restoration measures and increase vegetation coverage to improve the health status of the tea garden ecosystem.

[0081] On the other hand, the present invention also provides a tea garden ecosystem health assessment system based on landscape multifunctionality, such as Figure 3 As shown, it includes: The indicator collection module is used to determine the four aspects of tea garden ecosystem health evaluation indicators: vitality, organization, resilience, and ecosystem services, based on the "vitality-organization-resilience" framework and combined with ecosystem service theory. The index system construction module is configured to construct a three-layer hierarchical structure model by using the analytic hierarchy process, calculate the index weight by combining the entropy weight method, sort the feature importance of the tea garden ecosystem health evaluation index by using the random forest algorithm, and construct a tea garden ecosystem health evaluation index system based on the feature importance sorting; The data acquisition and processing module is configured to acquire the evaluation index data and perform standardization processing on the evaluation index data; The dimension reduction module is configured to perform dimension reduction processing on the evaluation index by using the principal component analysis method and calculate the weight; The calculation module is configured to calculate the dimension-reduced evaluation index value by using the standardization-processed evaluation index data; The comprehensive evaluation module is configured to calculate the health comprehensive score of the tea garden ecosystem based on the evaluation index and the weight obtained after the dimension reduction processing by using the weighted summation method; The grading module is configured to grade the health status of the tea garden ecosystem according to the health comprehensive score.

[0082] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts of each embodiment can be referred to each other.

[0083] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to the embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for evaluating the health of a tea garden ecosystem based on its multifunctional landscape, characterized in that, include: Based on the "vitality-organization-resilience" framework and combined with ecosystem service theory, four indicators for evaluating the health of tea garden ecosystems were determined: vitality, organization, resilience, and ecosystem services. A three-level hierarchical structure model was constructed using the analytic hierarchy process (AHP), and the index weights were calculated using the entropy weight method. The random forest algorithm was used to rank the feature importance of the tea garden ecosystem health evaluation indicators, and a tea garden ecosystem health evaluation indicator system was constructed based on the feature importance ranking. Acquire evaluation index data and standardize the evaluation index data; Principal component analysis was used to reduce the dimensionality of the evaluation indicators and the weights were calculated. Calculate the dimensionality-reduced evaluation index values ​​using standardized evaluation index data; Based on the evaluation indicators and weights obtained after dimensionality reduction, the weighted summation method is used to calculate the comprehensive health score of the tea garden ecosystem. The health status of the tea garden ecosystem is classified based on the comprehensive health score.

2. The method for evaluating the health of a tea garden ecosystem based on landscape multifunctionality according to claim 1, characterized in that, A three-level hierarchical structure model was constructed using the analytic hierarchy process (AHP). Entropy weighting was used to calculate the index weights. A random forest algorithm was employed to rank the feature importance of the tea garden ecosystem health evaluation indicators. Based on this ranking, a tea garden ecosystem health evaluation indicator system was constructed, including: Using the health assessment of the tea garden ecosystem as the target layer, primary indicators are defined; Based on the characteristics of the tea garden ecosystem, the target layer is broken down into four primary indicators; For each primary indicator, further refine it into secondary indicators; The target layer, the first-level indicators, and the second-level indicators are constructed into a three-level hierarchical structure model according to their hierarchical relationship; Standardize the raw data to bring them to the same order of magnitude; Calculate the information entropy of each indicator based on the standardized data; Redundancy is determined based on the information entropy; Calculate the weight coefficient for each indicator based on the redundancy; The weights obtained by the analytic hierarchy process (AHP) and the weight coefficients obtained by the entropy weight method are combined to obtain the comprehensive weight of each indicator. The indicators are screened based on the comprehensive weights to obtain a preliminary set of indicators; The model is built using the random forest algorithm, and the random forest model is trained using training data; The importance score of each feature variable is obtained based on the trained random forest model; The evaluation indicators are ranked according to the importance scores, and key indicators are selected from the initial set of indicators by setting a threshold. Based on the aforementioned key indicators, a health evaluation index system for tea garden ecosystems was constructed.

3. The method for evaluating the health of a tea garden ecosystem based on landscape multifunctionality according to claim 1, characterized in that, The established indicators for evaluating the health of tea garden ecosystems, encompassing four aspects: vitality, organization, resilience, and ecosystem services, include: Vitality measures include net primary productivity and leaf area index; Organizational strength includes the spread index, aggregation index, cohesion index, Shannon diversity index, and vertical heterogeneity index of tea canopy; In terms of resilience, factors include the richness of associated plant species, the genetic diversity index of tea trees, and the resilience coefficient of land use types; Ecosystem services include water conservation capacity, soil retention, tea production, and the phenol-to-amino acid ratio in fresh leaves.

4. The method for evaluating the health of a tea garden ecosystem based on landscape multifunctionality according to claim 1, characterized in that, Principal component analysis is used to reduce the dimensionality of the evaluation index and calculate its weights, including: Calculate the covariance matrix. The elements are: ； in, This represents the standardized value of the i-th sample on the j-th indicator; The matrix form can be represented as: ； Among them, Z T It is the transpose of the normalized matrix Z; For the covariance matrix Perform eigenvalue decomposition: ； in: Let v be the eigenvector matrix. j V is an orthogonal vector corresponding to the direction of the j-th principal component; T It is the transpose of the normalized matrix V; for p×p A diagonal matrix, elements on the diagonal yes eigenvalues, , representing the variance of the corresponding principal component; Calculate the variance contribution rate of the j-th principal component: ； Where, λ j It is the first j 1 eigenvalue, ; Calculate the cumulative variance contribution rate: ； Choose the smallest k such that the cumulative contribution rate reaches the preset threshold; For the k-th original variable, calculate its value across all... m Absolute values ​​of loadings on each principal component: ； in: v kj Indicates the first k The variable in the first... j Loadings on each principal component; Construct principal components, the j-th principal component Y j The expression is: ； in, For the first j 1 eigenvector Z It is a standardized matrix; The transformation matrix is ​​formed from the first k eigenvectors. Then the principal component matrix is: ； Each row corresponds to one sample. k Coordinates in the principal component space; For the k The formula chain for summarizing the indicators is as follows: ； in: m The principal components required to meet the cumulative contribution rate requirement.

5. The method for evaluating the health of a tea garden ecosystem based on landscape multifunctionality according to claim 3, characterized in that, The dimensionality-reduced evaluation index values ​​are calculated using standardized evaluation index data, including: The formula for calculating the land resilience coefficient is: ； Among them, RC i NDVI represents the restoring force coefficient of the i-th grid cell. i This represents the NDVI value of the i-th grid cell. NDVI_mean k Indicates the type of the location of the i-th grid cell. k NDVI average, RC k Indicates the set land category k The restoring force coefficient; The actual soil erosion and potential soil erosion are calculated according to the modified general soil loss equation. The difference between the actual soil erosion and potential soil erosion is the soil retention. We used the water yield from the InVEST model for assessing terrestrial ecosystem service functions to evaluate changes in water conservation function. The formula for calculating tea production is: ； in, This represents the tea yield corresponding to each grid cell. The leaf area index represents the area of ​​each grid cell. The total leaf area index represents the area of ​​the tea plantation. Represents the total annual tea production; The formula for calculating the acetaminophen-to-aminophenol ratio in fresh leaves is: ； TP refers to the total phenolic content; TA refers to the total free amino acid content.

6. The method for evaluating the health of a tea garden ecosystem based on landscape multifunctionality according to claim 1, characterized in that, Based on the evaluation indicators and weights obtained after dimensionality reduction, the formula for calculating the comprehensive health score of the tea garden ecosystem using the weighted summation method is as follows: ； Where W represents the weight value of the indicator; a, b, ..., x are the standardized data of the original indicator.

7. The method for evaluating the health of a tea garden ecosystem based on landscape multifunctionality according to claim 6, characterized in that, The health status of the tea garden ecosystem is graded based on the comprehensive health score, including: When HI ≥ 0.8, the tea garden ecosystem is considered to be in good health. When 0.6 ≤ HI < 0.8, the tea garden ecosystem is judged to be in a sub-healthy state. When HI < 0.6, the tea garden ecosystem is considered unhealthy.

8. A tea garden ecosystem health assessment system based on landscape multifunctionality, characterized in that, include: Indicator collection module: Used to determine the four aspects of tea garden ecosystem health evaluation indicators—vitality, organization, resilience, and ecosystem services—based on the "vitality-organization-resilience" framework and combined with ecosystem service theory; The indicator system construction module is used to construct a three-level hierarchical structure model through the analytic hierarchy process, calculate the indicator weights by combining the entropy weight method, and rank the feature importance of the tea garden ecosystem health evaluation indicators using the random forest algorithm. Based on the feature importance ranking, the tea garden ecosystem health evaluation indicator system is constructed. The data acquisition and processing module is used to acquire evaluation index data and perform standardization processing on the evaluation index data; The dimensionality reduction module is used to reduce the dimensionality of the evaluation index using principal component analysis and to calculate the weights. The calculation module is used to calculate the dimensionality-reduced evaluation index values ​​using standardized evaluation index data; The comprehensive evaluation module is used to calculate the comprehensive health score of the tea garden ecosystem based on the evaluation indicators and weights obtained after dimensionality reduction, using a weighted summation method. The grading module is used to grade the health status of the tea garden ecosystem based on the comprehensive health score.

Images