A method and system for assessing the health of an island ecosystem

Abstract

This application relates to a method and system for assessing the health of island ecosystems. The method collects and standardizes multi-source ecological data from islands to generate an evaluation index matrix and assigns initial static weights. It then constructs a niche network and relationship matrix that characterizes the nonlinear interactions between indicators. Based on this, it calculates a comprehensive niche overlap index among the indicators and constructs an overlap matrix. Subsequently, based on the degree of deviation of each indicator's actual state and utilizing the conflict or synergistic relationships between indicators revealed by the overlap matrix, it dynamically adjusts the initial weights—that is, when an indicator deteriorates, it increases the weight of conflicting indicators and decreases the weight of synergistic indicators. Finally, it uses the dynamic weights to calculate a comprehensive health index and determine the health level. This overcomes the shortcomings of traditional fixed-weight evaluation methods that cannot reflect the dynamic relationships between indicators within the ecosystem, achieving a more accurate, sensitive, and ecologically sound assessment of the health status of island ecosystems.

Description

Technical Field

[0001] This invention belongs to the field of environmental monitoring and ecological assessment, and in particular relates to a method and system for assessing the health of island ecosystems. Background Technology

[0002] As a special and fragile ecological unit, the accurate assessment of the health status of island ecosystems is crucial for ecological protection and sustainable management. Currently, the assessment of island ecosystem health generally employs a multi-indicator comprehensive evaluation system. This involves selecting a series of indicators reflecting environmental quality, biological ecology, and landscape patterns, assigning fixed weights to each indicator through expert scoring, analytic hierarchy process (AHP), or entropy methods, and finally calculating a weighted comprehensive health index. While this method provides an operational framework in practical applications, it has a fundamental methodological limitation: it assumes that the evaluation indicators are independent of each other or that there is only a simple linear correlation, thus assigning each indicator a static and unchanging importance weight. However, a real ecosystem is a complex adaptive system with non-linear and dynamic interactions and feedback mechanisms among its internal elements. For example, tourism development (socioeconomic indicator) may conflict with the retention rate of natural coastlines (landscape indicator); while increased vegetation cover (biological indicator) may have a synergistic effect with soil carbon sequestration capacity (environmental indicator). These complex relationships of "cooperation" and "conflict" can change in strength or even direction depending on external pressures or changes in the internal state of the system.

[0003] Existing static weighted evaluation models are completely unable to capture such dynamic interactions. This leads to a situation where, when certain key system indicators change drastically due to stress, the evaluation model cannot accurately reflect the evolution of the actual impact of this change on the overall system health. The evaluation results may be distorted or lagging, failing to reveal the intrinsic driving forces of system imbalance and early risk signals, thus weakening the supporting value of the evaluation results for precise management decisions. Although existing research recognizes the importance of interactions between indicators, how to effectively quantify this complex, nonlinear dynamic relationship and organically integrate it into the weight allocation logic of health evaluation remains an unresolved technical challenge.

[0004] Therefore, there is an urgent need for a new evaluation method that can transcend the static weighting model and dynamically reflect the complex interactions between indicators within the ecosystem, so as to improve the ecological mechanism conformity, state identification sensitivity, and management decision support of the evaluation. Summary of the Invention

[0005] Therefore, it is necessary to provide a method and system for assessing the health of island ecosystems to address the aforementioned technical issues.

[0006] Firstly, this application provides a method for assessing the health of an island ecosystem, including:

[0007] S1. Collect multi-source ecological data of the target island, standardize the multi-source ecological data, and generate a standardized index value matrix composed of multiple evaluation indicators; assign initial static weights to each evaluation indicator in the standardized index value matrix and generate an initial static weight vector.

[0008] S2. Construct an indicator niche network to characterize the nonlinear interaction relationships among all evaluation indicators, and construct an indicator relationship matrix based on the indicator niche network.

[0009] S3. Based on the indicator relationship matrix, calculate the comprehensive niche overlap index between every two evaluation indicator nodes in the indicator niche network, and construct the niche overlap matrix based on the comprehensive niche overlap index.

[0010] S4. Based on the deviation between the current state value and the corresponding ideal health state value of each evaluation indicator in the standardized index value matrix, and combined with the niche overlap relationship between the evaluation indicators represented in the niche overlap matrix, perform dynamic weight correction on the initial static weight vector to generate a dynamic weight vector.

[0011] The dynamic weight adjustment is used to: when the current state value of any evaluation indicator indicates a deterioration in health status, increase the weight of other evaluation indicators that conflict with the corresponding evaluation indicator, and decrease the weight of other evaluation indicators that cooperate with the corresponding evaluation indicator.

[0012] S5. Based on the dynamic weight vector, perform weighted summation on the standardized index value matrix to obtain the comprehensive health index of the target island's ecosystem.

[0013] S6. Based on the preset health level threshold, determine the ecosystem health level of the target island according to the comprehensive health index of the island ecosystem.

[0014] Secondly, this application also provides an island ecosystem health assessment system for implementing the method described in the first aspect, the system comprising:

[0015] The ecological data preprocessing module is used to collect multi-source ecological data of the target island, standardize the multi-source ecological data, generate a standardized index value matrix composed of multiple evaluation indicators, and assign initial static weights to each evaluation indicator in the standardized index value matrix to generate an initial static weight vector.

[0016] The indicator relationship modeling module is used to construct an indicator niche network to characterize the nonlinear interaction relationships between all evaluation indicators, and to construct an indicator relationship matrix based on the indicator niche network.

[0017] The niche overlap analysis module is used to calculate the comprehensive niche overlap index between every two evaluation index nodes in the index niche network based on the index relationship matrix, and to construct the niche overlap matrix based on the comprehensive niche overlap index.

[0018] The dynamic weight optimization module is used to perform dynamic weight correction on the initial static weight vector based on the deviation between the current state value of each evaluation indicator in the standardized index value matrix and the corresponding ideal health state value, and in combination with the niche overlap relationship between the evaluation indicators represented in the niche overlap matrix, to generate a dynamic weight vector.

[0019] The dynamic weight adjustment is used to: when the current state value of any evaluation indicator indicates a deterioration in health status, increase the weight of other evaluation indicators that conflict with the corresponding evaluation indicator, and decrease the weight of other evaluation indicators that cooperate with the corresponding evaluation indicator.

[0020] The health index calculation module is used to perform weighted summation calculation on the standardized index value matrix based on the dynamic weight vector to obtain the comprehensive health index of the island ecosystem of the target island.

[0021] The health level determination module is used to determine the ecosystem health level of a target island based on a preset health level threshold and the island ecosystem health comprehensive index.

[0022] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement a method for assessing the health of an island ecosystem as described in the first aspect.

[0023] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for assessing the health of an island ecosystem as described in the first aspect.

[0024] The aforementioned method and system for assessing the health of island ecosystems collects and standardizes multi-source ecological data from islands to generate an evaluation index matrix and assigns initial static weights. This leads to the construction of a niche network and relationship matrix that characterizes the nonlinear interactions between indicators. Based on this, a comprehensive niche overlap index is calculated, and an overlap matrix is ​​constructed. Subsequently, the initial weights are dynamically adjusted according to the actual deviation of each indicator and the conflict or synergistic relationships revealed by the overlap matrix—that is, when an indicator deteriorates, the weight of conflicting indicators is increased, and the weight of synergistic indicators is decreased. Finally, the comprehensive health index is calculated using the dynamic weights, and the health level is determined. This overcomes the shortcomings of traditional fixed-weight evaluation methods that cannot reflect the dynamic relationships between indicators within the ecosystem, achieving a more accurate, sensitive, and ecologically sound assessment of the health status of island ecosystems. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 A flowchart illustrating a method for assessing the health of an island ecosystem provided by this invention;

[0027] Figure 2 This is a schematic diagram illustrating the process of constructing a niche overlap matrix in one optional embodiment of the present invention;

[0028] Figure 3 This is a schematic diagram of the structure of an island ecosystem health assessment system provided by the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0030] refer to Figure 1 The application presents a flowchart illustrating a method for assessing the health of an island ecosystem, which includes the following steps:

[0031] S1. Collect multi-source ecological data of the target island, standardize the multi-source ecological data, and generate a standardized index value matrix composed of multiple evaluation indicators; assign initial static weights to each evaluation indicator in the standardized index value matrix and generate an initial static weight vector.

[0032] Specifically, the first step is to collect multi-source ecological data for the target island. The data collection scope covers both natural and social dimensions. The natural dimension includes data related to environmental quality, biological ecology, and landscape ecology, while the social dimension includes data related to human well-being. Environmental quality data is obtained through standardized testing, biological ecology data is obtained through a combination of field surveys and remote sensing interpretation, and social dimension data is collected from statistical departments, medical institutions, and other channels. After data collection, preprocessing is performed to remove outliers and supplement missing data. Outlier removal uses statistical methods to filter extreme values, and missing data is supplemented using interpolation or analogy with neighboring similar units to ensure data integrity and reliability.

[0033] To eliminate the impact of dimensional differences among various evaluation indicators, the preprocessed data is standardized to ensure a uniform range for all indicators, facilitating subsequent calculations. For positive indicators (higher values ​​indicate greater ecological health, such as vegetation cover and biodiversity index), an extreme value standardization approach is used, normalizing the data by calculating the original values ​​against the extreme values ​​within the indicator's range. For negative indicators (higher values ​​indicate less ecological health, such as heavy metal content and landscape fragmentation), a reverse extreme value calculation is employed to ensure that all standardized indicators positively reflect ecological health conditions. Based on this standardization logic, the formula for calculating the standardized positive indicators is as follows: ,in, This represents the standardized value of the j-th indicator in the i-th evaluation unit. This represents the original value of the j-th indicator in the i-th evaluation unit. This represents the minimum value of the j-th indicator among all evaluation units. This represents the maximum value of the j-th indicator across all evaluation units. The formula for standardizing negative indicators is: The meanings of each parameter are completely consistent with the standardized formula for positive indices. Similarly, this represents the standardized value of the j-th index in the i-th evaluation unit. This is the original value of the indicator. The maximum value of the indicator. This represents the minimum value of the indicator. Through the above processing, all indicator values ​​fall within the same range. Based on this, a standardized indicator value matrix is ​​constructed, where the number of rows corresponds to the number of evaluation units, and the number of columns corresponds to the number of evaluation indicators.

[0034] The initial static weight allocation for the indicators adopted an expert consultation method, covering four core dimensions: environmental quality, biological ecology, landscape ecology, and social dimensions. This constitutes a three-tiered indicator system framework. The first-level indicators include environmental quality, biological ecology, landscape ecology, and social dimensions. Each first-level indicator is further subdivided into second-level indicators, and the second-level indicators are further broken down into specific tertiary evaluation indicators. Senior experts in oceanography, environmental science, and ecology were invited to score the importance of each indicator according to a unified score range. The scoring process incorporated the actual impact of each dimension indicator on the health of the island ecosystem. For example, in the biological ecology dimension, biodiversity-related indicators play a core supporting role in ecosystem stability, and in the social dimension, human activity-related indicators have a correlation with ecosystem sustainability. After collecting the scoring results, extreme outliers were removed, and the mean of the remaining scores for each indicator was calculated. Then, the mean of all indicators was normalized to ensure that the sum of the weights is 1. Based on the above weight allocation and normalization logic, the formula for calculating the initial static weight normalization is as follows: ,in, This represents the initial static weight of the j-th evaluation index. Let represent the average expert score for the j-th indicator after preprocessing, and n represent the total number of evaluation indicators. This represents the sum of the average expert scores for all evaluation indicators after preprocessing. Based on this formula, the initial static weights for each indicator are calculated, and an initial static weight vector is constructed, with the vector dimension matching the number of evaluation indicators.

[0035] S2. Construct an indicator niche network to characterize the nonlinear interaction relationships among all evaluation indicators, and construct an indicator relationship matrix based on the indicator niche network.

[0036] Specifically, the first step is to define the niche dimension of the indicators. Considering the common synergistic and conflicting relationships among indicators in island ecosystems, the niche dimension can be defined as resource utilization, functional performance, and environmental response, respectively representing the indicator's resource occupation characteristics, ecosystem service contribution capacity, and environmental disturbance adaptation capacity. Synergistic relationships refer to mutually reinforcing associations between indicators, such as increased vegetation cover promoting enhanced soil carbon sequestration. Conflicting relationships refer to mutually inhibiting associations, such as increased human tourism activity potentially leading to a decline in the retention rate of natural coastlines. The niche width of each indicator in each dimension needs to be quantified through the resource utilization ratio. The resource utilization ratio refers to the proportion of resource occupation or utilization of the corresponding resource in a given niche dimension, determined based on indicator monitoring data and ecosystem characteristic analysis. A larger niche width indicates a wider range of resource utilization within the ecosystem. The specific calculation of niche width can be performed using the following formula: ,in, This represents the niche width of the j-th evaluation index. This represents the resource utilization ratio of the j-th evaluation index in the k-th niche dimension. This represents the total number of niche dimensions. This represents the sum of squares of the resource utilization ratios of the j-th indicator across all niche dimensions.

[0037] Identifying the interactions between indicators can be achieved through a combination of literature review, nonlinear correlation analysis, and expert consultation. This involves distinguishing between synergistic relationships (mutual promotion), conflicting relationships (mutual inhibition), and no significant interaction. Nonlinear correlation analysis uses mutual information to quantify the degree of correlation. The strength of interactions can be calculated using the grey relational analysis method, which is suitable for multi-factor, nonlinear data characteristics in ecosystems and effectively quantifies the strength of interactions between indicators. Based on the core logic of the grey relational analysis method, the formula for calculating the interaction strength is:

[0038]

[0039] in, This represents the grey relational degree (i.e., the interaction strength) between the i-th evaluation index and the j-th evaluation index. This represents the resolution coefficient, used to adjust the resolution of correlation calculations, and can take a value of 0.5. This represents the standardized value of the i-th indicator in the k-th evaluation unit. This represents the standardized value of the j-th indicator in the k-th evaluation unit. This represents the minimum difference between the standardized values ​​of two indicators across all evaluation units. This represents the maximum value of the standardized difference between the two indicators across all evaluation units. This represents the absolute difference between the standardized values ​​of the two indicators in the k-th evaluation unit.

[0040] Based on the analysis results of niche breadth, interaction relationships, and intensity, an indicator niche network is constructed (nodes represent indicators, and edges represent interaction relationships and intensity), and an indicator relationship matrix is ​​then constructed accordingly. Matrix elements quantify the interaction relationships and intensity between indicators; synergistic relationships are represented by positive grey relational values, conflict relationships by negative grey relational values, and no significant interaction is represented by 0. The expression for the elements of the indicator relationship matrix is ​​as follows:

[0041]

[0042] in, This represents the quantified value of the interaction relationship between the i-th and j-th evaluation indicators. This represents the grey relational degree between the i-th and j-th indicators, indicating a collaborative relationship. equal In conflict relationships equal When there is no significant interaction The value is 0.

[0043] S3. Based on the indicator relationship matrix, calculate the comprehensive niche overlap index between every two evaluation indicator nodes in the indicator niche network, and construct the niche overlap matrix based on the comprehensive niche overlap index.

[0044] Specifically, the basic niche overlap is used to quantify the degree of resource utilization overlap between two indicators along the niche dimension. A higher degree of overlap indicates that the indicator niches are closer and their resource utilization needs are more similar. The calculation can use the Piaka niche overlap formula, which quantifies overlap characteristics based on the resource utilization ratio of the indicators. Combining the above quantitative logic, the formula for calculating the basic niche overlap is:

[0045]

[0046] in, This represents the basic niche overlap between the i-th and j-th evaluation indicators. This represents the resource utilization ratio of the i-th indicator in the k-th niche dimension. p represents the resource utilization ratio of the j-th indicator in the k-th niche dimension, and p represents the total number of niche dimensions. This represents the sum of the products of the resource utilization ratios of the two indicators across all niche dimensions. This represents the sum of squares of the resource utilization ratios of the i-th indicator across all niche dimensions. It represents the sum of squares of the resource utilization ratios of the j-th indicator across all niche dimensions, and the square root is the product of the sums of squares of the resource utilization ratios of the two indicators.

[0047] The comprehensive niche overlap index considers both the basic niche overlap and the intensity of interactions between indicators, providing a more comprehensive characterization of the degree of mutual influence among indicators. The calculation involves multiplying the basic niche overlap by the absolute values ​​of the elements in the indicator relationship matrix. Based on this calculation logic, the formula for the comprehensive niche overlap index is as follows: ,in, This represents the comprehensive niche overlap index between the i-th and j-th evaluation indicators. This represents the basic niche overlap between the i-th and j-th indicators. This represents the absolute value of the quantitative value of the interaction relationship between the i-th and j-th indicators (i.e., the absolute value of the corresponding element in the indicator relationship matrix).

[0048] For each pair of indicators, a comprehensive niche overlap index is calculated. The overlap between an indicator and itself is 1 (resource utilization ratios are completely identical, indicating the strongest interaction). Based on the calculation results, a niche overlap matrix is ​​constructed. The matrix dimension matches the number of evaluation indicators, and the matrix elements are the comprehensive niche overlap indices of the corresponding two indicators. The expression for the elements of the niche overlap matrix is:

[0049]

[0050] in, This represents the comprehensive niche overlap index between the i-th and j-th evaluation indicators, where i and j are the same indicator. When i and j are different indicators The matrix, derived from the above formula for calculating the comprehensive niche overlap index, provides a quantitative basis for subsequent dynamic weight correction.

[0051] S4. Based on the deviation between the current state value and the corresponding ideal health state value of each evaluation indicator in the standardized index value matrix, and combined with the niche overlap relationship between the evaluation indicators represented in the niche overlap matrix, perform dynamic weight correction on the initial static weight vector to generate a dynamic weight vector.

[0052] The dynamic weight adjustment is used to: increase the weight of other evaluation indicators that conflict with the corresponding evaluation indicator and decrease the weight of other evaluation indicators that cooperate with the corresponding evaluation indicator when the current state value of any evaluation indicator indicates a deterioration in health status.

[0053] Specifically, the ideal health status value for each indicator is first determined. This ideal value is based on general standards and benchmarks for ecosystem health assessment, specifically combining historical best monitoring data and the average values ​​of corresponding indicators from similar healthy islands. Further calibration through expert review ensures that the value accurately represents the indicator's value when it is in an ideal health state. The evaluation standards and benchmarks must align with the unique characteristics of the island ecosystem, fully considering core features such as land-sea interaction and ecological vulnerability. For example, the ideal value for biological and ecological indicators should reflect the survival and reproduction needs of unique island species, while the ideal value for social dimension indicators should balance the synergy between human development and ecological protection. The ideal value must be standardized using methods consistent with data standardization to ensure it falls within the same range as the indicator's current standardized state value, facilitating the calculation of deviations.

[0054] The degree of deviation between the current state of the indicator and the ideal health state is used to quantify the degree of health deterioration. The greater the deviation, the worse the health state. Exceeding a preset threshold triggers dynamic weight adjustment. The specific calculation of the deviation degree uses a relative deviation quantification method, and the corresponding formula is:

[0055]

[0056] in, This indicates the degree of deviation of the j-th evaluation indicator. This represents the standardized value of the current state of the j-th evaluation index. This represents the standardized value of the ideal health status for the j-th evaluation indicator. This represents the absolute value of the difference between the current standardized value and the ideal standardized value of the j-th indicator.

[0057] Dynamic weight correction requires adjusting the initial static weights based on the degree of deviation and interaction between indicators. The correction coefficient for indicators that do not significantly deteriorate is 1 (weights remain unchanged). For significantly deteriorating indicators, the weights of synergistic indicators need to be reduced (mutual promotion leads to the transmission of deterioration), while the weights of conflicting indicators need to be increased (mutual inhibition weakens the conflict). The correction coefficient is calculated by combining the comprehensive niche overlap index and the degree of deviation. After clarifying the correction logic, the formula for calculating the weight correction coefficient falls into three cases: when there is no significant interaction... ; Collaborative relationships In conflict relationships .in, This represents the weight adjustment coefficient for the j-th evaluation indicator. This represents the comprehensive niche overlap index between the i-th significantly deteriorating indicator and the j-th indicator. The correction coefficient represents the degree of deviation of the i-th significantly deteriorating indicator. When there is a synergistic relationship, the correction coefficient is less than 1 (reducing the weight). When there is a conflict relationship, the correction coefficient is greater than 1 (increasing the weight). When there is no significant interaction, the correction coefficient is 1 (the weight remains unchanged).

[0058] After the correction coefficients are calculated, the initial static weights are corrected and normalized to ensure that the sum of all index weights is 1, thus generating a dynamic weight vector. To achieve the weight normalization goal, the formula for calculating the dynamic weights is as follows: ,in, This represents the dynamic weight of the j-th evaluation indicator. This represents the initial static weight of the j-th indicator. This represents the weight adjustment coefficient for the j-th indicator, and n represents the total number of evaluation indicators. This represents the sum of the products of the initial static weights of all indicators and their corresponding correction coefficients.

[0059] S5. Based on the dynamic weight vector, perform weighted summation on the standardized index value matrix to obtain the comprehensive health index of the target island's ecosystem.

[0060] Specifically, the Comprehensive Index of Island Ecosystem Health is used to comprehensively quantify the overall health level of target islands. Its core principle is to multiply the standardized values ​​of each indicator by their dynamic weights, sum them using a weighted average, and then perform a scaling transformation. This makes the index more intuitive and easier to understand, and reflects the impact of the dynamic interactions between indicators on overall health. Based on this comprehensive calculation logic, the formula for calculating the comprehensive index is:

[0061]

[0062] in, This represents the comprehensive health index of the island ecosystem for the i-th evaluation unit, where n represents the total number of evaluation indicators. This represents the standardized value of the j-th indicator in the i-th evaluation unit. This represents the dynamic weight of the j-th indicator. This represents the sum of the standardized values ​​of all indicators in the i-th evaluation unit multiplied by their corresponding dynamic weights. Multiplying by 100 converts the comprehensive index scale into a fixed interval for easier subsequent health level determination. In the formula, It is the core quantitative indicator that characterizes the health level of the island ecosystem in the i-th evaluation unit. The larger the value, the higher the health level. The value of n is determined by the construction results of the evaluation indicator system, which covers all core evaluation indicators in both natural and social dimensions. These are standardized indicator values, which have eliminated dimensional differences and ensured that different types of indicators can be directly used in calculations. The indicator weights are dynamically adjusted, taking into full account the nonlinear interactions between indicators and the dynamic changes in the health status of indicators, and can accurately reflect the importance of each indicator under the current system status. It is a basic health value that integrates the contributions of all indicators. By multiplying it by 100, a scale conversion is achieved, making the index result more in line with conventional evaluation habits.

[0063] During the calculation process, it is essential to ensure that the standardized indicator value matrix matches the dimensions of the dynamic weight vector, and that the standardized indicator values ​​accurately correspond to their respective dynamic weights, avoiding mismatches. After the calculation is completed, the reliability of the results is verified in conjunction with the actual ecological conditions of the target islands. If islands with good health conditions correspond to higher comprehensive indices, and islands with poor health conditions correspond to lower comprehensive indices, then the calculation results are reliable. If deviations exist, it is necessary to retrospectively check the entire process of data standardization, weight correction, and weighted summation to identify potential errors and ensure that the comprehensive index accurately represents the true health level of the island's ecosystem.

[0064] S6. Based on the preset health level threshold, determine the ecosystem health level of the target island according to the comprehensive health index of the island ecosystem.

[0065] Specifically, the preset health level thresholds are used to convert the comprehensive index into intuitive health levels, employing a five-level classification method (very healthy, healthy, sub-healthy, unhealthy, and pathological). This classification method aligns with the general norms of ecosystem health assessment, and the grading standards fully integrate the structural and functional characteristics of island ecosystems, clearly defining the ecosystem state boundaries corresponding to different health levels. Threshold determination combines expert consultation with statistical analysis. First, ecological health monitoring and evaluation data from multiple typical islands are collected. Cluster analysis is used to initially divide the level intervals. Then, considering the actual needs of island ecological management, experts in relevant fields are organized to verify and adjust the preliminary intervals, ensuring the scientific validity and practicality of the threshold division.

[0066] Based on the above threshold setting logic, the expression for dividing the health level threshold can be as follows:

[0067]

[0068] in, This represents the comprehensive health index of the island ecosystem. The ranges for each interval were determined through expert review and statistical analysis, accurately distinguishing the characteristics of ecosystems at different health states. In the formula, The value ranges from [0, 100], and the magnitude of the value directly corresponds to the ecosystem health level; when When the value is in the range [80, 100], it is considered to be in a very healthy state, corresponding to excellent environmental quality, rich biodiversity, and sustainable ecosystem services. When the value is in the range [60, 80), it is considered to be in a healthy state, corresponding to good environmental quality and relatively stable biological community structure. When the value is in the range [40, 60), it is considered to be in a sub-healthy state, corresponding to moderate environmental quality and average biodiversity. When the value is in the range [20, 40), it is considered to be in an unhealthy state, corresponding to poor environmental quality and limited ecosystem services. When the value is in the range [0, 20), it is considered to be in a pathological state, corresponding to poor environmental quality and severely degraded ecosystem services.

[0069] When determining the health level of a target island, its comprehensive index is extracted. The results are compared with preset thresholds, and supplemented by sub-indices from both natural and social dimensions. The calculation methods for the sub-indices are consistent with those for the comprehensive index, integrating the standardized values ​​and dynamic weights of the corresponding dimension indicators. After the health level is determined, an evaluation report is generated that includes the level determination results, indicator shortcomings, key influencing factors, and management recommendations, providing a scientific basis for the precise protection and sustainable management of island ecosystems.

[0070] The aforementioned method for assessing the health of island ecosystems collects and standardizes multi-source ecological data from islands to generate an evaluation index matrix and assigns initial static weights. It then constructs a niche network and relationship matrix that characterizes the nonlinear interactions between indicators. Based on this, it calculates a comprehensive niche overlap index among the indicators and constructs an overlap matrix. Subsequently, based on the degree of deviation of each indicator's actual state and utilizing the conflict or synergistic relationships between indicators revealed by the overlap matrix, it dynamically adjusts the initial weights—that is, when an indicator deteriorates, it increases the weight of conflicting indicators and decreases the weight of synergistic indicators. Finally, it uses the dynamic weights to calculate the comprehensive health index and determine the health level. This overcomes the shortcomings of traditional fixed-weight evaluation methods that cannot reflect the dynamic relationships between indicators within the ecosystem, achieving a more accurate, sensitive, and ecologically sound assessment of the health status of island ecosystems.

[0071] In one optional embodiment, an indicator niche network is constructed to characterize the nonlinear interaction relationships among all evaluation indicators, and an indicator relationship matrix is ​​constructed based on the indicator niche network, including the following steps:

[0072] S11. For each evaluation indicator in the standardized indicator value matrix, define the niche center and niche width of the evaluation indicator; whereby the niche center represents the optimal value of the evaluation indicator under ideal health conditions, and the niche width represents the sensitivity of the evaluation indicator to changes in its value.

[0073] Specifically, the niche center represents the optimal value of the evaluation index under ideal health conditions. This value is determined by combining the ideal health conditions standard of the index to ensure that it can accurately reflect the value when the index performs its optimal ecological function. The niche width represents the sensitivity of the evaluation index to changes in value. The larger the width, the more sensitive the index value is to external disturbances and the wider the range of allowable value fluctuations. Conversely, the index value is relatively stable and less sensitive to disturbances.

[0074] The formula for calculating the niche center is: ,in, This represents the niche center of the j-th evaluation index. This represents the standardized value of the ideal health status of the j-th evaluation indicator. The quantification of niche breadth is based on the variation characteristics and sensitivity analysis of indicator values, combined with resource utilization ratios, and is calculated using the following formula: ,in, This represents the niche width of the j-th evaluation index. Let represent the resource utilization ratio of the j-th evaluation indicator in the k-th niche dimension (i.e., the proportion of the corresponding resource occupied or utilized by the indicator in this dimension), and p represent the total number of niche dimensions. This represents the sum of squares of the resource utilization ratios across all niche dimensions for the j-th indicator. The coefficient of variation represents the standardized value of the j-th evaluation index, which is used to quantify the dispersion of the index value and further characterize the sensitivity to numerical changes.

[0075] S12. Based on the niche center and niche width, combined with ecological theory, long-term series monitoring data contained in multi-source ecological data, and pre-acquired expert knowledge, determine the interaction strength and direction between any two evaluation indicators.

[0076] Specifically, this step, based on the established niche center and niche width, combines multi-dimensional criteria to determine the interaction strength and direction between any two evaluation indicators, ensuring the scientific rigor and accuracy of the interaction characterization. Ecological theory provides the core basis for judging interaction relationships; for example, it derives synergistic or conflicting associations between indicators based on theories of species symbiosis and competition in ecosystems. Long-term series monitoring data included in multi-source ecological data refers to dynamic monitoring data of indicators covering multiple monitoring periods. By analyzing the correlation of the changing trends of two indicator values ​​in different periods, the dynamic characteristics of the interaction are quantified. Pre-acquired expert knowledge is used to calibrate and verify interaction relationships, combining the practical understanding of the correlation of island ecosystem indicators by domain experts to correct potential biases in data-driven analysis.

[0077] The intensity of the interaction was quantified using the grey relational analysis method that integrates niche characteristics. This method comprehensively considers the degree of niche center deviation and the synergy of numerical changes in both indicators. The calculation formula is as follows:

[0078]

[0079] in, This represents the interaction strength between the k-th evaluation index and the j-th evaluation index. This represents the niche center of the k-th evaluation index. This represents the niche center of the j-th evaluation index. This represents the absolute deviation between the two indicator niche centers. This represents the minimum deviation of all indicators from the niche center. This represents the maximum value of all indicators' deviation from the niche center; This represents the adjustment coefficient, which can take a value of 0.5; This represents the grey relational degree between the standardized values ​​of two indicators calculated based on long-term series monitoring data. The direction of action is determined through comprehensive analysis: if the two indicators show consistent trends (e.g., both increasing or decreasing simultaneously), and are determined to be mutually reinforcing based on ecological theory and expert knowledge, then the direction of action is synergistic and marked as positive; if the two indicators show opposite trends (one increasing and the other decreasing), and are determined to be mutually inhibiting, then the direction of action is conflicting and marked as negative.

[0080] S13. Construct an index relationship matrix based on the interaction strength and direction between all pairs of evaluation indicators.

[0081] In the index relationship matrix, the element value in the k-th row and j-th column is used to characterize the direct influence strength of the k-th evaluation index on the j-th evaluation index; a positive element value indicates a synergistic effect, a negative element value indicates a conflicting effect, and the absolute value of the element value represents the influence strength.

[0082] Specifically, this step constructs an indicator relationship matrix based on the interaction strength and direction between all paired evaluation indicators, achieving a structured representation of the interactions between indicators. The dimension of this matrix is ​​consistent with the total number of evaluation indicators. The element value in the k-th row and j-th column of the matrix is ​​specifically used to characterize the direct influence strength of the k-th evaluation indicator on the j-th evaluation indicator. The positive or negative value of the element corresponds to the direction of influence: a positive value indicates a synergistic effect between the two indicators, meaning that optimization of the k-th indicator will promote the improvement of the j-th indicator; a negative value indicates a conflicting effect between the two indicators, meaning that deterioration of the k-th indicator will inhibit the performance of the j-th indicator. The absolute value of the element directly represents the strength of the influence; the larger the absolute value, the more significant the direct influence of the k-th indicator on the j-th indicator.

[0083] The expression for the elements of the indicator relationship matrix is: ,in, This represents the value of the element in the k-th row and j-th column of the index relationship matrix. This indicates the sign of the direction of the effect of the k-th evaluation index on the j-th evaluation index, in the case of synergistic effect. When conflict occurs , This represents the interaction strength between the k-th evaluation indicator and the j-th evaluation indicator. Using this expression, the interaction attributes of each pair of indicators can be transformed into matrix element values, ultimately forming a complete indicator relationship matrix that comprehensively presents the direct influence characteristics between all indicators.

[0084] S14. For the index relationship matrix, remove the elements whose absolute value of the interaction strength is less than the preset relationship strength threshold to obtain the index relationship matrix that represents the core nonlinear interaction relationship.

[0085] Specifically, this step involves filtering the interactions within the constructed indicator relationship matrix, removing redundant correlations to focus on key impacts. The determination of the preset relationship strength threshold requires expert consultation and statistical analysis, referencing the distribution characteristics of interaction strengths across all indicators, and eliminating correlations with extremely low intensity and negligible impact on the overall ecosystem health assessment. In practice, the absolute value of each element in the indicator relationship matrix is ​​evaluated. If the absolute value is less than the preset relationship strength threshold, it is considered that there is no significant core interaction between the corresponding two indicators, and the element value is set to 0. If the absolute value is greater than or equal to the preset relationship strength threshold, the element value is retained, thus obtaining the optimized indicator relationship matrix.

[0086] The optimized index relationship matrix retains only the elements representing the core nonlinear interaction relationships, which can effectively simplify the complexity of subsequent niche network analysis, while ensuring that core correlation information is not lost. This makes the constructed index niche network more targeted and provides accurate basic data support for subsequent comprehensive niche overlap index calculation and dynamic weight correction.

[0087] refer to Figure 2 In one optional embodiment, based on the indicator relationship matrix, the comprehensive niche overlap index between every two evaluation indicator nodes in the indicator niche network is calculated, and a niche overlap matrix is ​​constructed based on the comprehensive niche overlap index, including the following steps:

[0088] S21. Based on the direct influence intensity in the indicator relationship matrix, construct the direct influence intensity matrix.

[0089] Specifically, the indicator relationship matrix clearly defines the direct influence strength and direction of action between each pair of evaluation indicators. The direct influence strength matrix is ​​a structured extraction and solidification of this core information. Its matrix dimension is completely consistent with the total number of evaluation indicators. The element value in the k-th row and j-th column of the matrix directly corresponds to the direct influence strength of the k-th evaluation indicator on the j-th evaluation indicator in the indicator relationship matrix. Moreover, the positive or negative value of the element value is consistent with the direction of action (positive value is synergistic effect, negative value is conflicting effect). The absolute value of the element value represents the magnitude of the direct influence strength.

[0090] During the construction process, there is no need to perform additional transformation on the direct influence intensity in the indicator relationship matrix. It is only necessary to completely replicate the core elements representing the direct influence intensity in the indicator relationship matrix into the direct influence intensity matrix according to the original indicator correspondence, ensuring the one-to-one correspondence between matrix elements and indicator pairs, while retaining the positive and negative attributes and numerical values ​​of the elements, providing accurate data support for the subsequent calculation of the direct influence weight ratio.

[0091] S22. Based on network propagation theory, construct a propagation model for calculating indirect impact; wherein, the propagation model quantifies the strength of the indirect effect of one evaluation indicator on another evaluation indicator through other intermediate evaluation indicators by simulating the process of propagation and attenuation of the impact along the path in the indicator niche network.

[0092] Specifically, this step, based on network propagation theory, constructs a propagation model for calculating indirect influence. The core objective is to quantify the strength of the indirect effect of one evaluation indicator on another through other intermediate evaluation indicators. Network propagation theory here views the indicator niche network as a complex propagation system. The influence between evaluation indicators can be transmitted through paths within the network. During propagation, the strength of the influence decreases as the propagation path lengthens and intermediate indicators become involved. The propagation model quantifies the strength of indirect influence by simulating this propagation and attenuation process.

[0093] The core simulation elements of the propagation model include: first, the identification of propagation paths, i.e., all complete paths connecting the source indicator (e.g., indicator k) to the target indicator (e.g., indicator j) through one or more intermediate indicators are included in the propagation path set; second, the rules for the transmission of influence direction, i.e., the sign of the weight of each edge on the path (corresponding to the direct influence strength element value in the indicator relationship matrix) is transmitted through multiplication, ultimately determining the overall direction of the indirect influence of the entire path; and third, the rules for the attenuation of influence strength, i.e., the longer the propagation path (number of edges), the more significant the attenuation of influence strength, and the attenuation rate is controlled by a preset attenuation coefficient. By integrating the above elements, the propagation model can accurately output the indirect influence contribution of each propagation path, and then summarize the overall indirect influence strength between the two indicators.

[0094] S23. For any two evaluation indicators j and k, calculate the comprehensive niche overlap index based on the direct influence intensity matrix and the indirect influence intensity output by the propagation model; wherein, the formula for calculating the comprehensive niche overlap index is:

[0095]

[0096] in, This represents the comprehensive niche overlap index between evaluation index j and evaluation index k; This indicates the strength of the direct influence of evaluation index k on evaluation index j in the index relationship matrix; The weighting coefficient represents the direct impact, with a value range of (0,1); the second term of the formula This represents the average indirect influence strength from evaluation index k to evaluation index j; This represents the set of all propagation paths from evaluation index k to evaluation index j; Represents a set One of the transmission paths; Indicates the transmission path The product of the numerical signs of all edge weights is the result of multiplying the results together. The product of the results represents the overall direction of the path's influence through its numerical sign, and the edge weights are the corresponding element values ​​in the index relation matrix. ; Indicates the propagation path The number of sides; This represents the path length attenuation coefficient. This represents the total number of propagation paths from evaluation index k to evaluation index j.

[0097] Specifically, this step calculates the comprehensive niche overlap index for any two evaluation indicators j and k, combining the direct impact intensity in the direct impact intensity matrix and the indirect impact intensity output by the propagation model. This index simultaneously encompasses the overlap characteristics of both direct and indirect associations, providing a more comprehensive characterization of the niche association degree between indicators. A weighted summation method is used to fuse the direct impact intensity and the average indirect impact intensity according to preset weights, where the weight of direct impact is determined by a coefficient. The weight of indirect influence from regulation is... This is to balance the contributions of the two types of influences to the comprehensive niche overlap index.

[0098] In the formula for calculating the comprehensive niche overlap index, The comprehensive niche overlap index represents the evaluation index j and evaluation index k. Its value directly reflects the comprehensive degree of niche overlap between the two indicators. The closer the value is to 1, the more significant the niche overlap between the two indicators and the more critical their mutual influence. This represents the direct influence strength of evaluation index k on evaluation index j in the index relationship matrix (i.e., the direct influence strength matrix). The positive or negative value of the value indicates the direction of the direct action (positive for synergy, negative for conflict), and the absolute value indicates the magnitude of the direct influence strength. The weighting coefficient represents the direct impact, with a value range of (0,1). The specific value can be determined through expert consultation or statistical calibration based on the characteristics of the island ecosystem and the correlation between indicators. For example, when the direct correlation between indicators is more significant... The value can be between 0.6 and 0.8; when the indirect association is more significant, The possible values ​​are 0.2 to 0.4.

[0099] The second term of the formula This represents the average indirect influence strength from evaluation index k to evaluation index j, where This represents the set of all propagation paths from evaluation index k to evaluation index j. The set includes all complete paths connected by intermediate indices, but excludes directly connected paths (directly connected paths are already reflected by the first item's direct influence strength). Represents a set A propagation path in the process, with each path corresponding to a unique intermediate indicator connection sequence; This represents the product of the numerical signs of all edge weights along the propagation path p. The edge weights are the corresponding element values ​​in the index relation matrix. The sign (positive or negative) of the product result directly indicates the overall direction of the indirect influence of the path. If the product result is positive, it indicates that the indirect influence of the path is cooperative; if it is negative, it indicates conflict. Indicates the propagation path The number of edges, that is, the total number of connecting edges between indicators in the path, the more edges there are, the longer the path is; This represents the path length attenuation coefficient, which is a positive number and is used to adjust the attenuation rate of the indirect influence intensity caused by the path length. The larger the value, the slower the decay, and the more significant the indirect contribution of long paths. The smaller the value, the faster the decay, and the more prominent the indirect contribution of the short path. The specific value can be determined by combining the verification results of the propagation model. This represents the total number of propagation paths from evaluation index k to evaluation index j, i.e., the set. The number of propagation paths included is used to sum up the indirect impact contributions of all paths and then average them to obtain the average indirect impact intensity per path.

[0100] In the specific calculation, we first traverse all propagation paths from index k to index j using the propagation model, and then calculate the propagation path for each path. and Multiplying the two values ​​yields the indirect impact contribution of a single path; summing the contribution values ​​of all paths gives the total indirect impact contribution; finally, dividing the total contribution value by the total number of paths gives the total indirect impact contribution. The average indirect impact intensity is obtained. Then, the direct impact intensity and the average indirect impact intensity are weighted and summed according to the formula to obtain the comprehensive niche overlap index of evaluation index j and k.

[0101] S24. Traverse all evaluation index pairs and repeat step S23 to generate a complete niche overlap matrix.

[0102] Specifically, by iterating through all evaluation index pairs and repeating the calculation process in step S23, a complete niche overlap matrix is ​​generated. The iteration process needs to cover all pairwise combinations of evaluation index pairs, including combinations of an index itself (i.e., the case where j=k). For the comprehensive niche overlap index of an index itself, since its niches are completely overlapping, it is directly set to 1. For different index pairs (j≠k), the corresponding values ​​are obtained according to the calculation logic in S23. .

[0103] All calculated comprehensive niche overlap indices are entered into a matrix according to the index of the evaluation indicators. The number of rows and columns in the matrix equals the total number of evaluation indicators. The element value in the j-th row and k-th column of the matrix is ​​the comprehensive niche overlap index between evaluation indicator j and evaluation indicator k, thus forming a complete niche overlap matrix. This matrix can systematically and intuitively display the comprehensive characteristics of niche overlap among all evaluation indicators, providing a core quantitative basis for subsequent dynamic weight adjustment and ensuring that the weight adjustment can fully take into account the direct and indirect mutual influences between indicators.

[0104] In one optional embodiment, based on the deviation between the current state value and the corresponding ideal health state value of each evaluation indicator in the standardized indicator value matrix, and in conjunction with the niche overlap relationship between evaluation indicators represented in the niche overlap matrix, a dynamic weight correction is performed on the initial static weight vector to generate a dynamic weight vector, including:

[0105] S31. Extract the current state value of each evaluation indicator from the standardized indicator value matrix, calculate the absolute difference between the current state value of each evaluation indicator and the corresponding niche center, and use it as the state deviation of each evaluation indicator.

[0106] Specifically, this step extracts the current state value of each evaluation indicator and calculates its state deviation, quantifying the difference between the current state and the ideal health state. First, the current state value of each evaluation indicator is extracted from the standardized indicator value matrix. This value has been dimensionless and can be directly used for cross-indicator comparative analysis. Since the niche center (i.e., the optimal value of the evaluation indicator under ideal health, which has been standardized) accurately represents the core value of the indicator's ideal health state, the state deviation is determined by calculating the absolute difference between the current state value of the indicator and the corresponding niche center. The larger the absolute difference, the more serious the deviation of the indicator's current state from the ideal health state, and the worse its health condition.

[0107] The formula for calculating the deviation from the state is as follows: ,in, This represents the degree of deviation of the j-th evaluation index. This represents the current state value extracted from the standardized index value matrix for the j-th evaluation index. This represents the niche center of the j-th evaluation index (the optimal value of the ideal health state after standardization). This represents the absolute difference between the current state value of the j-th evaluation indicator and the center of the ecological niche. This value directly reflects the degree of deviation of the indicator's health status.

[0108] S32. Mark all evaluation indicators whose state deviation is greater than the preset deviation threshold as problem indicators, and form a problem indicator set from the problem indicators.

[0109] Specifically, this step filters problematic indicators by setting a pre-defined deviation threshold, focusing on indicators of health deterioration that require special attention. The determination of the pre-defined deviation threshold requires expert consultation and consideration of the actual needs of ecosystem management. It also involves referencing the statistical distribution characteristics of the deviations of all evaluation indicators to set a reasonable critical value. When the deviation of an indicator exceeds this critical value, it indicates a significant deterioration in its health status, and it should be marked as a problematic indicator.

[0110] In practice, the deviation of each evaluation indicator from its state is compared one by one. Deviation threshold from preset ,like If so, the indicator is determined to be a problem indicator; if If the indicator is deemed to be in a basically ideal health state and not a problematic indicator, then it is determined that the health status of the indicator basically meets the ideal requirements and is not a problematic indicator. All marked problematic indicators are then integrated to form a problematic indicator set. ,in This indicates the total number of problematic indicators. This represents the t-th problem indicator. .

[0111] S33. For each evaluation indicator, calculate the weight correction amount of the evaluation indicator based on the niche overlap index of the evaluation indicator and all problem indicators in the problem indicator set in the niche overlap matrix, the state deviation of the problem indicator, and the initial static weight of the problem indicator in the initial static weight vector.

[0112] Specifically, this step calculates the weight adjustment for each evaluation indicator. The core logic is to adjust the weight ratio of each indicator based on the health deterioration characteristics of the problem indicator and the niche overlap between indicators. The calculation of the weight adjustment requires considering three core factors: first, the niche overlap index between the evaluation indicator and each problem indicator in the problem indicator set (reflecting the closeness of their mutual influence); second, the state deviation of the problem indicator (reflecting the severity of its health deterioration); and third, the initial static weight of the problem indicator (reflecting its fundamental importance in the evaluation system). These three factors jointly determine the magnitude and direction of the adjustment. The closer the niche overlap, the more severe the deviation of the problem indicator, and the higher the initial weight of the problem indicator, the greater the weight adjustment magnitude for the corresponding evaluation indicator. The formula for calculating the weight adjustment is:

[0113]

[0114] in, This represents the weight adjustment amount for the j-th evaluation indicator. This represents the relationship between the j-th evaluation indicator and the t-th problem indicator. The niche overlap index (taken from the niche overlap matrix) between the two indicates that the larger the value, the closer the mutual influence between them. Represents the t-th problem indicator Degree of deviation from the state; Represents the t-th problem indicator The initial static weights corresponding to the initial static weight vector; This represents the correction coefficient, with a value range of (0,1). It is used to control the overall magnitude of the correction amount to avoid over-correction or under-correction. Its specific value is determined in advance through expert demonstration. Represents a set of problem indicators The total number of problematic indicators.

[0115] The direction of the correction is determined by the interaction direction between the indicators (reflected by the direction of the interaction associated with the niche overlap index): if the evaluation indicator and the problem indicator are synergistic (the direction of the interaction corresponding to the niche overlap index is positive), then the deterioration of the health of the problem indicator will inhibit the evaluation indicator, and the weight correction will be negative (reducing the weight of the evaluation indicator); if the two are in conflict (the direction of the interaction corresponding to the niche overlap index is negative), then the deterioration of the health of the problem indicator will weaken the inhibition of the evaluation indicator, and the weight correction will be positive (increasing the weight of the evaluation indicator).

[0116] S34. Add the initial static weight of each evaluation indicator to the corresponding weight correction amount to obtain the original adjusted weight.

[0117] Specifically, this step involves adding the initial static weight of each evaluation indicator to its corresponding weight correction to obtain the original adjusted weight, thus completing the dynamic adjustment of the basic weights. The original adjusted weights directly inherit the fundamental importance of the initial static weights while incorporating the correction magnitude brought about by the correlation characteristics of the problem indicators. The calculation formula is as follows: ,in, This represents the original adjusted weight of the j-th evaluation indicator. This represents the initial static weight of the j-th evaluation index. This represents the weight adjustment amount for the j-th evaluation indicator.

[0118] The calculation process must ensure that the initial static weights and the weight correction amounts have the same numerical dimensions. The weights are adjusted directly through addition and subtraction operations. If the original adjusted weights are negative (due to the correction amount being negative and its absolute value being greater than the initial static weights), the negative value is retained and subsequently converted into a reasonable weight value through normalization processing to ensure the completeness and rationality of the weight adjustment.

[0119] S35. Normalize the original adjusted weights of all evaluation indicators to generate a dynamic weight vector.

[0120] Specifically, this step normalizes the original adjusted weights of all evaluation indicators to ensure that the sum of the adjusted weights is 1, ultimately generating a dynamic weight vector. The purpose of normalization is to transform the original adjusted weights into values ​​that conform to the weight allocation rules, making the weight proportions of each indicator reasonable and directly usable for subsequent calculation of the comprehensive health index.

[0121] The formula for normalization is as follows: ,in, This represents the dynamic weight of the j-th evaluation index after normalization. This represents the original adjusted weight of the j-th evaluation indicator, and n represents the total number of evaluation indicators. This represents the sum of the original adjusted weights of all evaluation indicators. After calculating the dynamic weight of each evaluation indicator using this formula, all dynamic weights are integrated according to the indicator's index to form a dynamic weight vector. This vector can accurately reflect the dynamic importance of each evaluation indicator under the current ecosystem status, providing a core weight basis for the subsequent calculation of the comprehensive health index of the island ecosystem.

[0122] In an optional embodiment, for each evaluation index, based on the niche overlap index between the evaluation index and all problem indices in the problem index set, the state deviation of the problem index, and the initial static weight corresponding to the problem index in the initial static weight vector, the weight correction amount of the evaluation index is calculated, including:

[0123] S41. For the target evaluation index of the current correction amount to be calculated, traverse each problem index in the problem index set and query the comprehensive niche overlap index between the target evaluation index and the problem index from the niche overlap matrix.

[0124] Specifically, this step involves matching the target evaluation indicator whose correction amount needs to be calculated with the comprehensive niche overlap index of all problem indicators, providing foundational data for subsequent relationship type determination and correction amount calculation. First, it clarifies that the current calculation object is a single target evaluation indicator (denoted as j). Then, it iterates through each problem indicator (denoted as k) in the problem indicator set Q, ensuring that all problem indicators within the set are covered without omitting any relationships. Since the comprehensive niche overlap index between all evaluation indicator pairs is pre-stored in the niche overlap matrix, for each pair of target evaluation indicator j and problem indicator k, the corresponding value can be directly retrieved from this matrix to obtain their comprehensive niche overlap index. This value fully characterizes the degree of overlap in the ecological niches of the two and the direction of their interaction.

[0125] S42. Based on the preset relationship strength threshold, determine the relationship type between the target evaluation index and the problem index, and determine the corresponding correction direction coefficient based on the relationship type; where, if If so, it is determined to be a relationship of strong conflict. ;like If so, it is determined to be a strong cooperative relationship. Otherwise, confirm. ;in, This represents the comprehensive niche overlap index between the target evaluation index j and the problem index k. To preset the relationship strength threshold, This represents the correction direction coefficient of the target evaluation index j corresponding to the problem index k.

[0126] Specifically, this step determines the relationship type between the target evaluation indicator and the problem indicator based on a preset relationship strength threshold, and determines the corresponding correction direction coefficient accordingly. The core is to clarify the direction of the problem indicator's weight correction on the target evaluation indicator (increase, decrease, or no effect). The preset relationship strength threshold θ is consistent with the threshold used in the indicator relationship matrix optimization to ensure the consistency of relationship type judgment. Its specific value is determined through expert consultation and ecosystem indicator correlation characteristic analysis, and is used to distinguish between three relationship types: strong conflict, strong synergy, and no significant correlation.

[0127] The specific judgment logic is as follows: If the comprehensive niche overlap index of the target evaluation index j and the problem index k is... If the two are in strong conflict, then the deterioration of the health of the problem indicator k will weaken its inhibitory effect on the target evaluation indicator j, and the weight of the target evaluation indicator j needs to be increased. Therefore, the correction direction coefficient is determined. ;like If the two indicators are determined to have a strong synergistic relationship, then the deterioration of the health of the problem indicator k will have a ripple effect on the target evaluation indicator j, and the weight of the target evaluation indicator j needs to be reduced. Therefore, the correction direction coefficient is determined. ;like In Within the specified interval, it is determined that there is no significant correlation between the two indicators, and the weight of the problem indicator k has no significant impact on the weight of the target evaluation indicator j. Therefore, the correction direction coefficient is determined. .in, This represents the correction direction coefficient of the target evaluation index j corresponding to the problem index k. Its value directly determines the positive or negative direction of the single correction component, and thus affects the adjustment direction of the target evaluation index weight.

[0128] S43. Based on the state deviation of the problem indicator and the initial static weight corresponding to the problem indicator in the initial static weight vector, combined with the correction direction coefficient and the comprehensive niche overlap index, calculate the single correction component of the target evaluation indicator caused by the problem indicator; wherein, the calculation formula for the single correction component is:

[0129]

[0130] in, This represents the single-item correction component of the target evaluation index j due to the problem index k; Indicates the global adjustment coefficient; Indicates the comprehensive niche overlap index The absolute value; This indicates the degree of deviation of the problem indicator k from its state. This represents the initial static weight of the problem index k.

[0131] Specifically, this step, based on the above relationship judgment results, combines the state deviation of the problematic indicator, the initial static weight, and the comprehensive niche overlap index to calculate the single-item correction component of the target evaluation indicator caused by the problematic indicator. The core is to quantify the specific impact of a single problematic indicator on the weight of the target evaluation indicator. Multiple factors are considered in the calculation: the correction direction coefficient determines the adjustment direction; the absolute value of the comprehensive niche overlap index represents the strength of the correlation between the two; the state deviation of the problematic indicator represents the severity of its health deterioration; the initial static weight of the problematic indicator represents its fundamental importance in the evaluation system; and a global adjustment coefficient regulates the overall correction magnitude to avoid over- or under-correction.

[0132] In the calculation formula of the single-term correction component, λ represents the single-item correction component of the target evaluation index j caused by the problem index k. Its positive or negative value corresponds to the direction of weight adjustment (positive value increases weight, negative value decreases weight), and the absolute value corresponds to the adjustment range. λ represents the global adjustment coefficient, with a value range of (0,1). It is used to uniformly control the overall range of all single-item correction components to ensure that the corrected weight is within a reasonable range. Its specific value is determined through expert demonstration and verification of the evaluation system. This indicates the correction direction coefficient (with values ​​of +1, -1, or 0) for the target evaluation index j corresponding to the problem index k, which directly determines the positive or negative direction of a single correction component. This represents the comprehensive niche overlap index between the target evaluation index j and the problem index k. The absolute value of the index represents the strength of the correlation between the two; the larger the absolute value, the more significant the influence of the problem index on the weight of the target evaluation index. This indicates the degree of deviation of the problem indicator k from its state, representing the severity of the health deterioration of the problem indicator. The larger the value, the greater the correction to the weight of the target evaluation indicator. This represents the initial static weight of the problem indicator k in the initial static weight vector, which characterizes the fundamental importance of the problem indicator itself in the evaluation system. The higher the weight, the more prominent its influence on the weight of the target evaluation indicator.

[0133] In specific calculations, the above parameters are directly multiplied according to the formula to obtain the single-item correction component of the target evaluation index j caused by the individual problem index k. For example, when the target evaluation index and the problem index have a strong conflicting relationship ( ), with a relatively strong correlation ( Large values), serious deviations in problem indicators ( Large value) and high initial weight ( When the value is large, A large positive value indicates that the weight of the target evaluation indicator needs to be significantly increased.

[0134] S44. Sum the individual correction components of the target evaluation index relative to all problem indicators to obtain the weight correction amount of the target evaluation index; wherein, the formula for calculating the weight correction amount is:

[0135]

[0136] in, This represents the weight adjustment amount for the target evaluation index j. This is a set of problem indicators.

[0137] Specifically, this step involves summing the individual correction components of the target evaluation indicator relative to all problem indicators to obtain the final weight correction amount for the target evaluation indicator. The core of this step is to integrate the comprehensive influence of all problem indicators on the weight of the target evaluation indicator. Since the target evaluation indicator may be affected by multiple problem indicators simultaneously, and the direction and magnitude of the influence of each problem indicator are different, the summation operation is used to superimpose all individual correction components to obtain the overall correction effect.

[0138] In the formula for calculating the weight adjustment, This represents the final weight adjustment amount for the target evaluation index j, with its positive or negative value corresponding to the overall adjustment direction and its absolute value corresponding to the overall adjustment magnitude; Q represents the set of problem indicators. This represents iterating through each problem indicator k in the set of problem indicators. Through this summation calculation, the influence of all problem indicators on the target evaluation indicator weights can be integrated, ultimately yielding a precise weight correction amount. This provides a core basis for subsequent calculations of the original adjustment weights and the generation of the dynamic weight vector.

[0139] For each evaluation indicator, the calculation process of S41-S44 is repeated to obtain the weight correction amount for each evaluation indicator, ensuring that the subsequent dynamic weight correction can fully adapt to the correlation characteristics of each indicator and the current health status of the ecosystem.

[0140] The aforementioned method for assessing the health of island ecosystems collects and standardizes multi-source ecological data from islands to generate an evaluation index matrix and assigns initial static weights. It then constructs a niche network and relationship matrix that characterizes the nonlinear interactions between indicators. Based on this, it calculates a comprehensive niche overlap index among the indicators and constructs an overlap matrix. Subsequently, based on the degree of deviation of each indicator's actual state and utilizing the conflict or synergistic relationships between indicators revealed by the overlap matrix, it dynamically adjusts the initial weights—that is, when an indicator deteriorates, it increases the weight of conflicting indicators and decreases the weight of synergistic indicators. Finally, it uses the dynamic weights to calculate the comprehensive health index and determine the health level. This overcomes the shortcomings of traditional fixed-weight evaluation methods that cannot reflect the dynamic relationships between indicators within the ecosystem, achieving a more accurate, sensitive, and ecologically sound assessment of the health status of island ecosystems.

[0141] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0142] Based on the same inventive concept, this application also provides a system for implementing the island ecosystem health assessment method described above. The solution provided by this system is similar to the implementation scheme described in the above method; therefore, the specific limitations of one or more island ecosystem health assessment system embodiments provided below can be found in the limitations of the island ecosystem health assessment method described above, and will not be repeated here.

[0143] In one exemplary embodiment, such as Figure 3 As shown, an island ecosystem health assessment system 30 is provided to implement the methods in the above-described method embodiments. The system includes:

[0144] The ecological data preprocessing module 31 is used to collect multi-source ecological data of the target island, standardize the multi-source ecological data, generate a standardized index value matrix composed of multiple evaluation indicators, and assign initial static weights to each evaluation indicator in the standardized index value matrix to generate an initial static weight vector.

[0145] The indicator relationship modeling module 32 is used to construct an indicator niche network to characterize the nonlinear interaction relationship between all evaluation indicators, and to construct an indicator relationship matrix based on the indicator niche network.

[0146] The niche overlap analysis module 33 is used to calculate the comprehensive niche overlap index between every two evaluation index nodes in the index niche network based on the index relationship matrix, and to construct the niche overlap matrix based on the comprehensive niche overlap index.

[0147] The dynamic weight optimization module 34 is used to perform dynamic weight correction on the initial static weight vector based on the deviation between the current state value of each evaluation indicator in the standardized index value matrix and the corresponding ideal health state value, and in combination with the niche overlap relationship between the evaluation indicators represented in the niche overlap matrix, to generate a dynamic weight vector.

[0148] The dynamic weight adjustment is used to: increase the weight of other evaluation indicators that conflict with the corresponding evaluation indicator and decrease the weight of other evaluation indicators that cooperate with the corresponding evaluation indicator when the current state value of any evaluation indicator indicates a deterioration in health status.

[0149] The health index calculation module 35 is used to perform weighted summation calculation on the standardized index value matrix based on the dynamic weight vector to obtain the comprehensive health index of the island ecosystem of the target island.

[0150] The health level determination module 36 is used to determine the ecosystem health level of the target island based on the preset health level threshold and the comprehensive health index of the island ecosystem.

[0151] Embodiments of this application also provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the aforementioned method embodiments.

[0152] Embodiments of this application also provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described method embodiments.

[0153] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0154] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.

Claims

1. A method for assessing the health of an island ecosystem, characterized in that, The method includes: S1. Collect multi-source ecological data of the target island, standardize the multi-source ecological data, and generate a standardized index value matrix composed of multiple evaluation indicators; assign initial static weights to each evaluation indicator in the standardized index value matrix to generate an initial static weight vector. S2. Construct an indicator niche network to characterize the nonlinear interaction relationships among all the evaluation indicators, and construct an indicator relationship matrix based on the indicator niche network; S3. Based on the index relationship matrix, calculate the comprehensive niche overlap index between every two evaluation index nodes in the index niche network, and construct a niche overlap matrix based on the comprehensive niche overlap index. S4. Based on the deviation between the current state value and the corresponding ideal health state value of each evaluation indicator in the standardized index value matrix, and in combination with the niche overlap relationship between the evaluation indicators represented in the niche overlap matrix, perform dynamic weight correction on the initial static weight vector to generate a dynamic weight vector. The dynamic weight correction is used to: when the current state value of any of the evaluation indicators indicates a deterioration in health status, increase the weight of other evaluation indicators that conflict with the corresponding evaluation indicator, and decrease the weight of other evaluation indicators that cooperate with the corresponding evaluation indicator. S5. Based on the dynamic weight vector, perform a weighted summation calculation on the standardized index value matrix to obtain the comprehensive health index of the island ecosystem of the target island; S6. Based on a preset health level threshold, determine the ecosystem health level of the target island according to the comprehensive health index of the island ecosystem.

2. The method according to claim 1, characterized in that, The construction of an indicator niche network to characterize the nonlinear interaction relationships among all the evaluation indicators, and the construction of an indicator relationship matrix based on the indicator niche network, includes: S11. For each evaluation index in the standardized index value matrix, define the niche center and niche width of the evaluation index; wherein, the niche center represents the optimal value of the evaluation index under ideal health conditions, and the niche width represents the sensitivity of the evaluation index to numerical changes. S12. Based on the niche center and the niche width, combined with ecological theory, long-term series monitoring data contained in the multi-source ecological data, and pre-acquired expert knowledge, determine the interaction strength and direction between any two of the evaluation indicators respectively. S13. Construct an index relationship matrix based on the interaction strength and direction between all pairs of evaluation indices; The element value in the k-th row and j-th column of the indicator relationship matrix is ​​used to characterize the direct influence strength of the k-th evaluation indicator on the j-th evaluation indicator; the positive element value indicates a synergistic effect, the negative element value indicates a conflicting effect, and the absolute value of the element value represents the influence strength. S14. For the index relationship matrix, remove the elements whose absolute value of the interaction strength is less than a preset relationship strength threshold to obtain the index relationship matrix that represents the core nonlinear interaction relationship.

3. The method according to claim 1, characterized in that, The step of calculating the comprehensive niche overlap index between every two evaluation index nodes in the index niche network based on the index relationship matrix, and constructing a niche overlap matrix based on the comprehensive niche overlap index, includes: S21. Based on the direct influence intensity in the index relationship matrix, construct the direct influence intensity matrix; S22. Based on network propagation theory, a propagation model for calculating indirect effects is constructed; wherein, the propagation model quantifies the strength of the indirect effect of one evaluation indicator on another evaluation indicator through other intermediate evaluation indicators by simulating the process of propagation and attenuation of the influence along the path in the indicator niche network. S23. For any two evaluation indicators j and k, calculate the comprehensive niche overlap index based on the direct influence intensity matrix and the indirect influence intensity output by the propagation model; wherein, the formula for calculating the comprehensive niche overlap index is: in, The comprehensive niche overlap index represents the sum of evaluation index j and evaluation index k. This indicates the strength of the direct influence of evaluation index k on evaluation index j in the index relationship matrix; The weighting coefficient represents the direct impact, with a value range of (0,1); the second term of the formula This represents the average indirect influence strength from evaluation index k to evaluation index j; This represents the set of all propagation paths from evaluation index k to evaluation index j; Represents a set One of the transmission paths; Indicates the transmission path The product of the numerical signs of all edge weights is the result of multiplying the results together. The product of these results represents the overall direction of the path's influence through its numerical sign. The edge weights are the corresponding element values ​​in the index relationship matrix. ; Indicates the propagation path The number of sides; This represents the path length attenuation coefficient. This represents the total number of propagation paths from evaluation index k to evaluation index j; S24. Traverse all evaluation index pairs and repeat step S23 to generate a complete niche overlap matrix.

4. The method according to claim 2, characterized in that, The step of dynamically correcting the initial static weight vector based on the deviation between the current state value and the corresponding ideal health state value of each evaluation indicator in the standardized index value matrix, and in conjunction with the niche overlap relationship between the evaluation indicators represented in the niche overlap matrix, to generate a dynamic weight vector includes: S31. Extract the current state value of each evaluation indicator from the standardized indicator value matrix, and calculate the absolute difference between the current state value of each evaluation indicator and the corresponding niche center as the state deviation of each evaluation indicator. S32. Mark all the evaluation indicators whose state deviation is greater than a preset deviation threshold as problem indicators, and form a problem indicator set by the problem indicators; S33. For each evaluation indicator, based on the niche overlap index of the evaluation indicator in the niche overlap matrix and all the problem indicators in the problem indicator set, the state deviation of the problem indicator, and the initial static weight of the problem indicator in the initial static weight vector, calculate the weight correction amount of the evaluation indicator. S34. Add the initial static weight of each evaluation index to the corresponding weight correction amount to obtain the original adjusted weight; S35. Normalize the original adjusted weights of all the evaluation indicators to generate the dynamic weight vector.

5. The method according to claim 4, characterized in that, For each evaluation indicator, based on the niche overlap index between the evaluation indicator in the niche overlap matrix and all the problem indicators in the problem indicator set, the state deviation of the problem indicator, and the initial static weight corresponding to the problem indicator in the initial static weight vector, the weight correction amount of the evaluation indicator is calculated, including: S41. For the target evaluation index of the current correction amount to be calculated, traverse each problem index in the problem index set, and query the comprehensive niche overlap index between the target evaluation index and the problem index from the niche overlap matrix. S42. Based on the preset relationship strength threshold, determine the relationship type between the target evaluation index and the problem index, and determine the corresponding correction direction coefficient based on the relationship type; wherein, if If so, it is determined to be a relationship of strong conflict. ;like If so, it is determined to be a strong cooperative relationship. Otherwise, confirm. ;in, The comprehensive niche overlap index represents the relationship between the target evaluation index j and the problem index k. The preset relationship strength threshold, This indicates that the correction direction coefficient corresponds to the problem index k in the target evaluation index j. S43. Based on the state deviation of the problem indicator and the initial static weight corresponding to the problem indicator in the initial static weight vector, combined with the correction direction coefficient and the comprehensive niche overlap index, calculate the single-item correction component of the target evaluation indicator caused by the problem indicator; wherein, the calculation formula of the single-item correction component is: in, This refers to the single-item correction component generated by the target evaluation index j due to the problem index k; Indicates the global adjustment coefficient; The comprehensive niche overlap index represents the total ecological niche overlap index. The absolute value; The deviation of the state described by the problem index k; The initial static weights representing the problem index k; S44. Sum the individual correction components of the target evaluation index relative to all the problem indices to obtain the weight correction amount of the target evaluation index; wherein, the formula for calculating the weight correction amount is: in, The weight adjustment amount represents the target evaluation index j. This refers to the set of indicators for the aforementioned problem.

6. A system for assessing the health of an island ecosystem, used to implement the method described in any one of claims 1 to 5, characterized in that, The system includes: An ecological data preprocessing module is used to collect multi-source ecological data of the target island, standardize the multi-source ecological data, generate a standardized index value matrix composed of multiple evaluation indicators, and assign initial static weights to each evaluation indicator in the standardized index value matrix to generate an initial static weight vector. The indicator relationship modeling module is used to construct an indicator niche network to characterize the nonlinear interaction relationship between all the evaluation indicators, and to construct an indicator relationship matrix based on the indicator niche network. The niche overlap analysis module is used to calculate the comprehensive niche overlap index between every two evaluation index nodes in the index niche network based on the index relationship matrix, and to construct a niche overlap matrix based on the comprehensive niche overlap index. The dynamic weight optimization module is used to perform dynamic weight correction on the initial static weight vector based on the deviation between the current state value and the corresponding ideal health state value of each evaluation indicator in the standardized index value matrix, and in combination with the niche overlap relationship between the evaluation indicators represented in the niche overlap matrix, to generate a dynamic weight vector. The dynamic weight correction is used to: when the current state value of any of the evaluation indicators indicates a deterioration in health status, increase the weight of other evaluation indicators that conflict with the corresponding evaluation indicator, and decrease the weight of other evaluation indicators that cooperate with the corresponding evaluation indicator. The health index calculation module is used to perform weighted summation calculation on the standardized index value matrix based on the dynamic weight vector to obtain the comprehensive health index of the island ecosystem of the target island. The health level determination module is used to determine the ecosystem health level of the target island based on a preset health level threshold and the comprehensive health index of the island ecosystem.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 5.

Images