A Methodology for Evaluating the Effectiveness of Coastal Wetland Ecosystem Restoration Based on Nature Solutions

By constructing an evaluation method for the effectiveness of coastal wetland ecosystem restoration through automatic analysis and standardized processing of multi-source data, the problems of strong subjectivity and poor repeatability of evaluation results in existing technologies have been solved. This method enables the quantification, repeatability, and cross-project comparison of coastal wetland ecosystem restoration projects, thereby improving the objectivity and efficiency of the evaluation.

CN123117536APending Publication Date: 2026-11-17XIAMEN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611324320.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-28
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies lack quantifiable and calculable assessment indicators for the restoration of coastal wetland ecosystems. The assessment results are highly subjective, have poor repeatability, and are difficult to process multi-source data, making it difficult to achieve process evaluation throughout the entire life cycle and cross-project comparisons.

Method used

A method for evaluating the effectiveness of coastal wetland ecosystem restoration based on nature-based solutions is constructed. Through automatic parsing, standardization, and weighted fusion of multi-source data, an NbS implementation effectiveness index that can be compared horizontally is output. Outliers are detected by the median absolute deviation criterion, and qualitative indicators are automatically recommended by TF-IDF text similarity. A comprehensive index is generated by hierarchical weighted aggregation.

Benefits of technology

It enables objective, repeatable, and quantifiable assessment of coastal wetland ecosystem restoration projects, improves the objectivity and consistency of assessment results, covers process control and social value evaluation throughout the entire life cycle, supports cross-project comparisons, and improves the efficiency and reliability of assessments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123117536A_ABST
    Figure CN123117536A_ABST
Patent Text Reader

Abstract

This invention discloses a method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions, belonging to the field of ecological monitoring and management technology. This method constructs a full life-cycle assessment system covering six criterion layers: analysis and diagnosis, solution formulation, solution implementation, social governance, ecological effectiveness, and sustainable development. It uses the median absolute deviation criterion to detect outliers in monitoring data. Quantitative indicators are normalized to scores of 0-3 using a quantitative calculation model and piecewise mapping function. Qualitative indicators are automatically graded and human-machine collaboratively reviewed through TF-IDF text similarity matching combined with a confidence mechanism. The NbS implementation effectiveness index (NSI) is calculated through hierarchical weighted aggregation to automatically determine the assessment level and output diagnostic suggestions. This invention enables objective, repeatable, and horizontally comparable quantitative assessment of coastal wetland restoration projects and is applicable to the evaluation of restoration effectiveness for various types of coastal wetlands, including seagrass beds, mangroves, and coastal salt marshes.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of ecological monitoring and management technology, specifically relating to a method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions. Background Technology

[0002] Coastal wetlands are globally important blue carbon ecosystems, mainly including seagrass beds, mangroves, and coastal salt marshes, which have significant carbon sequestration and storage functions, biodiversity maintenance capabilities, and fishery resource support capacity. Therefore, ecological restoration has become a core means to curb degradation and restore ecosystem functions, and conducting full life-cycle effectiveness assessments of restoration projects is an objective basis for acceptance, blue carbon project audits, and self-improvement by restoration teams. The Nature-based Solutions (NbS) global standard published by the International Union for Conservation of Nature (IUCN) is an important framework for assessing whether ecological restoration projects conform to the concept of nature solutions. However, existing assessment techniques have many shortcomings: First, the standard only provides a general qualitative framework and lacks specific, quantifiable, and calculable indicators for specific ecosystems such as coastal wetlands. The assessment is highly dependent on human judgment, which is subjective and has poor repeatability. Second, traditional assessments of the effectiveness of coastal wetland restoration often focus on short-term ecological results such as target vegetation survival rate and coverage, ignoring process evaluations throughout the entire life cycle, including preliminary diagnosis, program development, implementation and management, social governance, and long-term sustainability. It also lacks quantifiable expressions for dimensions such as social welfare and adaptive management. Third, there is a lack of standardized comprehensive indices that can aggregate multi-dimensional results into a single value and support cross-project and cross-stage horizontal comparisons.

[0003] At the data processing level, there are still technical obstacles to processing multi-source data throughout the entire lifecycle of restoration projects: First, the project data sources are heterogeneous and vary in form, including both numerical ecological monitoring data and natural language text management documents. The two have different dimensions, scales, and expressions, and there is a lack of technical means to automatically parse, standardize, and integrate them into a single quantitative result; Second, ecological monitoring data inevitably contains outliers and missing values, and direct calculation without identification will significantly reduce the reliability of the assessment results; Third, manually reading data and scoring each item against standards results in inconsistent standards among different assessors, making the process difficult to repeat and unable to support cross-project horizontal comparisons.

[0004] Therefore, there is an urgent need to construct an evaluation method that quantifies the NbS framework and covers the entire remediation lifecycle, so as to achieve automatic parsing of multi-source heterogeneous data, outlier detection, standardized mapping and multi-criteria weighted fusion, and output an objective, repeatable and horizontally comparable quantitative effectiveness index. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes a method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions. This method transforms the global NbS standard into a quantifiable indicator system specific to coastal wetlands. It automatically analyzes, standardizes, and weights multi-source heterogeneous data, outputting a horizontally comparable NbS implementation effectiveness index, thereby achieving an objective and repeatable quantitative assessment of the entire lifecycle of restoration projects.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A method for evaluating the effectiveness of coastal wetland ecosystem restoration based on nature-based solutions includes the following steps: S1. Collect multi-source data throughout the entire life cycle of the coastal wetland restoration project. First, perform format parsing and data cleaning on the multi-source data. Then, perform outlier detection on the measured monitoring data. Mark the monitoring values ​​that are identified as outliers and remove or correct them according to preset rules. Mark and supplement missing values. S2. Standardize the quantitative and qualitative indicators separately to obtain standardized scores; S3. Retrieve the pre-stored weight values ​​of the six criterion layers and their subordinate evaluation indicators. Aggregate the standardized scores of each indicator under the same criterion layer according to its weight to form the criterion layer evaluation value. Then calculate the NbS implementation effectiveness index using the following formula: , Wherein, NSI is the NbS implementation effectiveness index, and the NSI value ranges from 0 to 3; i is the criterion layer number; n is the total number of criterion layers; The weights of the i-th criterion layer; The evaluation value of the i-th criterion layer is denoted as criterion layer ; the six criterion layers are analysis and diagnosis, scheme formulation, scheme implementation, social governance, ecological effectiveness, and sustainable development. S4. Based on the NbS implementation effectiveness index, the evaluation level is automatically determined according to the preset range, and a visual chart of the score of each criterion level, a detailed table of sub-indicator scores, and a diagnostic suggestion report are generated.

[0007] Preferably, in step S1, the multi-source data includes baseline survey data, restoration project implementation data, ecological monitoring data, community questionnaire survey data, and project management documents, and the restoration objects include one or more of seagrass beds, mangroves, or coastal salt marshes; The outlier detection uses the median absolute deviation criterion. For a monitoring value sequence of the same indicator, the median and the median absolute deviation are calculated using the following formula: , Where MAD is the median absolute deviation; The median is a function of the median; x is the monitored value; M is the median. When satisfied When the value x is outlier, the monitored value is marked as an outlier; where λ is a preset deviation factor, with a default value of 3.

[0008] Preferably, the specific process of step S2 is as follows: For quantitative indicators, the original feature values ​​of each quantitative indicator are first calculated from the measured monitoring data according to the preset quantitative calculation model, and then the original feature values ​​are normalized to a standardized score of 0 to 3 through a segmented mapping function; For qualitative indicators, the text of the project management document is segmented into words to construct a text feature vector, and the similarity is calculated one by one with the feature vectors of each level description text stored in the level description library. Based on the principle of maximum similarity and the confidence threshold, a level assignment of 0 to 3 points is automatically recommended, and then a manual review instruction is received to update and obtain the final standardized score.

[0009] Preferably, in step S2, the quantitative calculation model of the quantitative indicator includes: The area verification rate is calculated using the following formula: , in, For area verification rate; This refers to the actual area repaired in the project. The planned repair area for the project; The survival rate of the target vegetation is calculated using the following formula: , Where L represents the survival rate of the target vegetation; y represents the monitoring patch number; and Y represents the total number of monitoring patches. The number of target plants surviving in the y-th monitoring patch; The total number of target plants planted or transplanted within the y-th monitoring patch; The target vegetation cover is calculated using the following formula: , in, Target vegetation cover; Let y be the vertically projected area of ​​the target vegetation canopy within the y-th monitoring patch; Let y be the total area of ​​the y-th monitoring patch; The relative density of vegetation is calculated using the following formula: , in, Relative density of vegetation; Target vegetation density for the restoration area; The vegetation density corresponding to the natural reference area or historical baseline; The relative organic carbon burial rate is calculated using the following formula: , in, This represents the relative organic carbon burial rate. The rate of organic carbon burial in the remediation area; The organic carbon burial rate is the rate corresponding to the natural reference area or historical baseline.

[0010] Preferably, in step S2, the quantitative calculation model for the quantitative index further includes a comprehensive macrobenthic diversity index, calculated using the following formula: , , , , , Where d is the Margalef richness index; S is the number of benthic animal species; N is the total number of benthic animal individuals; H′ is the Shannon-Wiener diversity index; and j is the benthic animal species number. The proportion of individuals of the j-th benthic species to the total number of individuals; This is the standardized value of the Margalef richness index; This serves as the benchmark value for the Margalef richness index; This is the standardized value of the Shannon-Wiener diversity index; This serves as the benchmark value for the Shannon-Wiener diversity index. The function is for finding the minimum value; This is a comprehensive index of macrobenthic diversity; α+β=1, with default values ​​of α=0.4 and β=0.6. The weight parameters can be configured and adjusted.

[0011] Preferably, in step S2, the piecewise mapping function maps the original feature values ​​of the quantitative indicator to the preset four-level threshold of the indicator. The standardized score is normalized to 0 to 3, with the specific segmentation rules as follows: When a ≤ t0, G = 0; When t0 < a ≤ t1, G = (a - t0) / (t1 - t0); When t1< a ≤t2, G=1+(a-t1) / (t2-t1); When t2 < a ≤ t3, G = 2 + (a - t2) / (t3 - t2); when a > t3, G = 3, that is, the upper limit saturation is reached. Where a is the original characteristic value of the quantitative indicator; G is the standardized score; for relative indicators, t3 corresponds to the indicator reaching or exceeding the level of the natural or planned benchmark, and saturation occurs when it exceeds t3, and the standardized score is the full score of 3 points.

[0012] Preferably, in step S2, the automatic recommendation and assignment process for the qualitative indicators is as follows: Construct a feature vector from the level description text of level k. TF-IDF weighted text feature vectors are constructed for relevant paragraphs in the project management document. Where k is the level number in the level description text, k=0,1,2,3; Calculate eigenvectors With text feature vectors The cosine similarity is calculated using the following formula: , in, For feature vectors With text feature vectors Cosine similarity; Let L2 be the norm of the vector; The level with the highest similarity is selected as the optimal recommendation level. The calculation formula is as follows: , in, The highest recommended level; To select the level with the highest similarity; The confidence level for automatic grading of qualitative indicators is calculated using the following formula: , Where conf is the confidence level of the qualitative indicator that is automatically graded; The feature vector of the text to be evaluated Description vector of the optimal recommendation level standard Cosine similarity; The second highest similarity among all remaining levels after excluding the optimal recommendation level; When the confidence level (conf) of the automatic classification of a qualitative indicator is less than the preset threshold (θ), the qualitative indicator is marked as requiring manual review. When the confidence level (conf) of the automatic classification of a qualitative indicator is greater than or equal to the preset threshold (θ), the score Gq=k* corresponding to the optimal recommendation level k* is used as the automatic recommended standardized score of the qualitative indicator. For qualitative indicators marked as requiring manual review, a manual review instruction is received, and the manually confirmed score of 0, 1, 2, or 3 is used as the final standardized score of the qualitative indicator. For those not manually corrected, the automatic recommended standardized score is used as the final standardized score.

[0013] Preferably, in step S3, the weight values ​​include the criterion layer weight, the local weight of the first-level indicator, and the local weight of the second-level indicator; The default weights for the six criteria layers are as follows: the default weight for analysis and diagnosis is 0.2056, the default weight for scheme formulation is 0.1691, the default weight for scheme implementation is 0.1748, the default weight for social governance is 0.1150, the default weight for ecological effectiveness is 0.2373, and the default weight for sustainable development is 0.0983. The specific process of the weighted aggregation is as follows: First, the standardized scores of the secondary indicators are weighted and summed according to the local weights of the secondary indicators to obtain the evaluation value of the primary indicator; then, the evaluation values ​​of the primary indicators are weighted and summed according to the local weights of the primary indicators to obtain the evaluation value of the corresponding criterion layer.

[0014] Preferably, in step S4, the rule for determining the evaluation level is: When NSI ≥ 2.25, it is determined to be highly compliant with NbS; When 1.50 ≤ NSI < 2.25, it is determined to be basically in compliance with NbS; When 0.75 ≤ NSI < 1.50, it is determined to be partially in compliance with NbS; When NSI < 0.75, it is determined that it does not meet the NbS requirement.

[0015] Preferably, step S4 is followed by a comparative analysis step, which involves comparing the results of this NSI assessment with the calculation results of the standard NbS self-assessment tool, outputting a report on the total score difference, the differences of each criterion level and the classification of the reasons for the differences, and comparing the results with a historical database of similar projects to generate project rankings and optimization suggestions.

[0016] By adopting the above technical solution, the present invention has the following beneficial effects: 1. This invention transforms the globally applicable NbS standard into 28 quantifiable assessment indicators adapted to coastal wetland ecosystems, replacing the traditional purely manual qualitative judgment mode. This significantly improves the objectivity, consistency, and repeatability of the assessment results, solves the problem of the lack of ecosystem-specific quantitative implementation methods for the NbS standard, and realizes the quantitative implementation of the NbS standard for coastal wetlands.

[0017] 2. The evaluation system of this invention covers six major stages: analysis and diagnosis, scheme formulation, scheme implementation, social governance, ecological effectiveness, and sustainable development. It runs through the entire life cycle of a project from early planning to long-term operation and maintenance. It not only evaluates the final ecological results, but also focuses on process control and social value. It can accurately locate the weak links in each stage of the project and provide direct targeted directions for management optimization.

[0018] 3. This invention supports the automatic parsing and unified standardization of structured numerical monitoring data and unstructured text management documents; it employs the MAD median absolute deviation criterion for outlier detection, relying on robust statistical properties to avoid extreme value interference, thereby improving the quality of monitoring data and the reliability of evaluation results, and overcoming the technical obstacles of inconsistent formats, different dimensions, and difficulty in fusion and calculation of multi-source data. Simultaneously, this invention uses a four-level piecewise linear mapping function to continuously map the original indicator values ​​to standardized scores of 0-3 points. Compared to the traditional integer-level lookup table scoring method, this preserves the continuous change information of the indicators, provides a more refined gradient characterization of the repair effectiveness, and can more sensitively reflect the phased improvement effect of the project.

[0019] 4. The qualitative indicators of this invention use TF-IDF text similarity for automatic initial judgment, and combine it with confidence threshold to automatically transfer low confidence results to manual review. This not only improves the automation efficiency of qualitative indicator evaluation, but also ensures the accuracy of evaluation in complex scenarios through manual backup, thus achieving a balance between efficiency and reliability.

[0020] 5. This invention outputs a unified NbS implementation effectiveness index (NSI) through hierarchical weighted aggregation, establishing a standardized quantitative evaluation benchmark. This index can be directly used for horizontal comparison of effectiveness between different restoration projects and between different stages of the same project, providing intuitive and comparable quantitative evidence for acceptance, blue carbon project audits, and self-optimization by restoration teams. Furthermore, the index weights, level description library, and segmented thresholds all support configuration and adjustment, adapting to the restoration assessment needs of different types of coastal wetlands such as mangroves, seagrass beds, and coastal salt marshes. The emphasis of indicators can also be adjusted according to different assessment purposes, making the method highly portable. Attached Figure Description

[0021] Figure 1 This is a flowchart of the method of the present invention.

[0022] Figure 2 This is a radar chart example of the evaluation values ​​of the six criteria layers in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0024] like Figure 1 and Figure 2 As shown. A method for evaluating the effectiveness of coastal wetland ecosystem restoration based on nature-based solutions includes the following steps: S1. Collect multi-source data throughout the entire life cycle of the coastal wetland restoration project. First, perform format parsing and data cleaning on the multi-source data. Then, perform outlier detection on the measured monitoring data. Mark the monitoring values ​​that are identified as outliers and remove or correct them according to preset rules. Mark and supplement missing values. In step S1, the multi-source data includes baseline survey data, restoration project implementation data, ecological monitoring data, community questionnaire survey data, and project management documents. The restoration objects include one or more of seagrass beds, mangroves, or coastal salt marshes. The outlier detection uses the median absolute deviation criterion. For a monitoring value sequence of the same indicator, the median and the median absolute deviation are calculated using the following formula: , Where MAD is the median absolute deviation; The median is a function of the median; x is the monitored value; M is the median. When satisfied When the value x is outlier, the monitored value is marked as an outlier; where λ is a preset deviation factor, with a default value of 3. S2. Standardize the quantitative and qualitative indicators separately to obtain standardized scores; The specific process of step S2 is as follows: For quantitative indicators, the original feature values ​​of each quantitative indicator are first calculated from the measured monitoring data according to the preset quantitative calculation model, and then the original feature values ​​are normalized to a standardized score of 0 to 3 through the segmented mapping function; For qualitative indicators, the text of the project management document is segmented into words to construct a text feature vector, and the similarity is calculated one by one with the feature vectors of each level description text in the pre-stored level description library. Based on the principle of maximum similarity and the confidence threshold, a level assignment of 0 to 3 points is automatically recommended, and then a manual review instruction is received to update and obtain the final standardized score; In step S2, the quantitative calculation model for the quantitative indicator includes: The area verification rate is calculated using the following formula: , in, For area verification rate; This refers to the actual area repaired in the project. The planned repair area for the project; The survival rate of the target vegetation is calculated using the following formula: , Where L represents the survival rate of the target vegetation; y represents the monitoring patch number; and Y represents the total number of monitoring patches. The number of target plants surviving in the y-th monitoring patch; The total number of target plants planted or transplanted within the y-th monitoring patch; The target vegetation cover is calculated using the following formula: , in, Target vegetation cover; Let y be the vertically projected area of ​​the target vegetation canopy within the y-th monitoring patch; Let y be the total area of ​​the y-th monitoring patch; The relative density of vegetation is calculated using the following formula: , in, Relative density of vegetation; Target vegetation density for the restoration area; The vegetation density corresponding to the natural reference area or historical baseline; The relative organic carbon burial rate is calculated using the following formula: , in, This represents the relative organic carbon burial rate. The rate of organic carbon burial in the remediation area; The organic carbon burial rate corresponding to the natural reference area or historical baseline; In step S2, the quantitative calculation model for the quantitative indicators also includes a comprehensive index of macrobenthic diversity, calculated using the following formula: , , , , , Where d is the Margalef richness index; S is the number of benthic animal species; N is the total number of benthic animal individuals; H′ is the Shannon-Wiener diversity index; and j is the benthic animal species number. The proportion of individuals of the j-th benthic species to the total number of individuals; This is the standardized value of the Margalef richness index; This serves as the benchmark value for the Margalef richness index; This is the standardized value of the Shannon-Wiener diversity index; This serves as the benchmark value for the Shannon-Wiener diversity index. The function is for finding the minimum value; This is a comprehensive index of macrobenthic diversity; α+β=1, with default values ​​of α=0.4 and β=0.6, and the weight parameters can be configured and adjusted. In step S2, the piecewise mapping function modifies the original feature values ​​of the quantitative indicator according to the preset four-level threshold of the indicator. The standardized score is normalized to 0 to 3, with the specific segmentation rules as follows: When a ≤ t0, G = 0; When t0 < a ≤ t1, G = (a - t0) / (t1 - t0); When t1< a ≤t2, G=1+(a-t1) / (t2-t1); When t2 < a ≤ t3, G = 2 + (a - t2) / (t3 - t2); when a > t3, G = 3, that is, the upper limit saturation is reached. Where a is the original characteristic value of the quantitative indicator; G is the standardized score; for relative indicators, t3 corresponds to the indicator reaching or exceeding the level of the natural or planned benchmark, and saturation occurs when it exceeds t3, and the standardized score is the full score of 3. In step S2, the specific process of automatically recommending and assigning values ​​to the qualitative indicators is as follows: Construct a feature vector from the level description text of level k. TF-IDF weighted text feature vectors are constructed for relevant paragraphs in the project management document. Where k is the level number in the level description text, k=0,1,2,3; Calculate eigenvectors With text feature vectors The cosine similarity is calculated using the following formula: , in, For feature vectors With text feature vectors Cosine similarity; Let L2 be the norm of the vector; The level with the highest similarity is selected as the optimal recommendation level. The calculation formula is as follows: , in, The highest recommended level; To select the level with the highest similarity; The confidence level for automatic grading of qualitative indicators is calculated using the following formula: , Where conf is the confidence level of the qualitative indicator that is automatically graded; The feature vector of the text to be evaluated Description vector of the optimal recommendation level standard Cosine similarity; The second highest similarity among all remaining levels after excluding the optimal recommendation level; When the confidence level (conf) of the automatic classification of a qualitative indicator is less than the preset threshold (θ), the qualitative indicator is marked as requiring manual review. When the confidence level (conf) of the automatic classification of a qualitative indicator is greater than or equal to the preset threshold (θ), the score Gq=k* corresponding to the optimal recommendation level k* is used as the automatic recommended standardized score of the qualitative indicator. For qualitative indicators marked as requiring manual review, a manual review instruction is received, and the manually confirmed score of 0, 1, 2, or 3 is used as the final standardized score of the qualitative indicator. For those not manually corrected, the automatic recommended standardized score is used as the final standardized score. S3. Retrieve the pre-stored weight values ​​of the six criterion layers and their subordinate evaluation indicators. Aggregate the standardized scores of each indicator under the same criterion layer according to its weight to form the criterion layer evaluation value. Then calculate the NbS implementation effectiveness index using the following formula: , Wherein, NSI is the NbS implementation effectiveness index, and the NSI value ranges from 0 to 3; i is the criterion layer number; n is the total number of criterion layers; The weights of the i-th criterion layer; The evaluation value of the i-th criterion layer is denoted as criterion layer ; the six criterion layers are analysis and diagnosis, scheme formulation, scheme implementation, social governance, ecological effectiveness, and sustainable development. In step S3, the weight values ​​include the criterion layer weights, the local weights of the first-level indicators, and the local weights of the second-level indicators; wherein, the 6 criterion layers, 20 first-level indicators, and 28 second-level indicators are shown in Table 1; Table 1: Evaluation Index System for the Effectiveness of Coastal Wetland Ecosystem Restoration

[0025] The default weights for the six criteria layers are as follows: the default weight for analysis and diagnosis is 0.2056, the default weight for scheme formulation is 0.1691, the default weight for scheme implementation is 0.1748, the default weight for social governance is 0.1150, the default weight for ecological effectiveness is 0.2373, and the default weight for sustainable development is 0.0983. The specific process of the weighted aggregation is as follows: First, the standardized scores of the secondary indicators are weighted and summed according to the local weights of the secondary indicators to obtain the evaluation value of the primary indicator; then, the evaluation values ​​of the primary indicators are weighted and summed according to the local weights of the primary indicators to obtain the evaluation value of the corresponding criterion layer; wherein... S4. Based on the NbS implementation effectiveness index, the evaluation level is automatically determined according to the preset range, and a visual chart of the score of each criterion level, a detailed table of sub-indicator scores, and a diagnostic suggestion report are generated. In step S4, the rule for determining the evaluation level is as follows: When NSI ≥ 2.25, it is determined to be highly compliant with NbS; When 1.50 ≤ NSI < 2.25, it is determined to be basically in compliance with NbS; When 0.75 ≤ NSI < 1.50, it is determined to be partially in compliance with NbS; When NSI < 0.75, it is determined that it does not meet the NbS requirement; Step S4 is followed by a comparative analysis step, which involves comparing the results of this NSI assessment with the calculation results of the standard NbS self-assessment tool, outputting a report on the total score difference, the differences at each criterion level, and the classification of the reasons for the differences, and comparing the results with a historical database of similar projects to generate project rankings and optimization suggestions. Example 1

[0026] like Figure 1 As shown, this embodiment uses the *Haliotis diversicolor* seagrass bed restoration project in the Hepu Dugong National Nature Reserve in Guangxi as the evaluation object. The actual restoration area of ​​the project is 2.18 hm². 2 The grass block method was used for transplantation, with a restoration period of 12 months. The effectiveness was evaluated according to the following steps: Step S1: Multi-source data acquisition and preprocessing

[0027] Data import: Import project baseline survey report, restoration plan, construction records, 6 months of ecological monitoring data (seagrass survival rate, coverage, stem and branch density, macrobenthic animal samples, sediment carbon burial data), 476 community questionnaire survey data, as well as project management documents such as funding contracts, patrol records, and publicity materials; the system automatically parses structured Excel / CSV data and unstructured text data.

[0028] Outlier detection: For the numerical sequences of each monitoring indicator, outlier detection is performed using the median absolute deviation criterion, with a deviation factor λ=3. Identified outliers are marked and removed according to preset rules, and missing values ​​are marked and supplemented to ensure the reliability of subsequent calculations. Step S2: Indicator Standardization Mapping Process

[0029] (1) Calculation and standardization of quantitative indicators The original characteristic values ​​of each quantitative indicator were calculated from the measured monitoring data. Combined with the preset four-level thresholds, the scores were normalized to a standardized score of 0-3 using a piecewise mapping function. The calculation results for each indicator are as follows: Seagrass survival rate: measured value L=97.7%, threshold set at t0=50%, t1=70%, t2=85%, t3=95%; since L>t3, the upper limit saturation is reached, and the standardized score G=3.0 points.

[0030] Seagrass cover: measured value C=50%, threshold set at t0=30%, t1=50%, t2=70%, t3=90%; since C=t1, the standardized score G=1.0.

[0031] Relative vegetation density: measured value =273%, set t3=100%, super natural baseline saturation, standardized score G=3.0 points.

[0032] Relative organic carbon burial rate: measured value =91%, thresholds are set at t0=40%, t1=60%, t2=80%, t3=90%; because >t3, standardized score G=3.0.

[0033] Area verification rate: Actual measured value =109%, exceeding the planned baseline saturation, standardized score G=3.0 points.

[0034] The Macrobenthic Diversity Index (BMDI) is calculated from benthic animal monitoring data, including Margalef richness (d) and Shannon-Wiener diversity (H′). After standardization with the baseline values ​​of the natural reference area, the raw BMDI value is obtained by weighted summation at α=0.4 and β=0.6. After segmented threshold mapping, the standardized score is G=1.0.

[0035] (2) Automatic classification and assignment of qualitative indicators After segmenting the project management document into words, a TF-IDF text feature vector is constructed. Cosine similarity is then calculated between this vector and the standard descriptive vectors for levels 0-3 in the level description library. Levels are automatically recommended, and confidence scores are calculated. The confidence threshold θ = 0.15 is set. Social challenge identification: The document only mentions "habitat degradation and biodiversity loss", which has the highest similarity to the 1-point level description, and the confidence level is greater than the threshold. The final standardized score is 1 point. Clarity of human well-being goals: The document does not mention any quantitative well-being goals, which is most similar to the description of the 0-point level. The confidence level is greater than the threshold, and the final standardized score is 0. Adaptive management mechanism: The document has no records related to dynamic adjustment mechanism, has the highest similarity to the description of 0-point level, has a confidence level greater than the threshold, and the final standardized score is 0. Community outreach: The project has media coverage and has the highest similarity to the 3-point rating description, with a confidence level greater than the threshold, and the final standardized score is 3 points; the remaining qualitative indicators were automatically graded and assigned according to the same rules, and all low-confidence indicators were manually reviewed by experts to confirm the final standardized score. Step S3: Hierarchical weighted aggregation and comprehensive index calculation

[0036] Retrieve the preset three-tier weighting system of secondary indicators, primary indicators, and criteria, and aggregate them layer by layer with weights: The scores of the secondary indicators are weighted and summed according to the local weights of the secondary indicators to obtain the evaluation value of each primary indicator; The evaluation values ​​of the first-level indicators are weighted and summed according to their local weights to obtain the evaluation values ​​of the six criterion layers: Analysis and diagnosis: 1.7883 points; Solution formulation: 2.4032 points; Implementation score: 1.9167; Social governance: 2.6364 points; Ecological effectiveness: 2.0104 points; Sustainable development: 1.2698 points; Based on the default weights of the criteria layer, the NbS implementation effectiveness index (NSI) is calculated as follows: NSI = 0.2056 × 1.7883 + 0.1691 × 2.4032 + 0.1748 × 1.9167 + 0.1150 × 2.6364 + 0.2373 × 2.0104 + 0.0983 × 1.2698 = 2.0140. Step S4: Level Determination and Result Output

[0037] Level determination: NSI=2.0140, which is in the range of 1.50≤NSI<2.25, and is determined to be basically in line with NbS.

[0038] Visualization output: Automatically generates radar charts of scores for six criterion layers, such as... Figure 2 As shown.

[0039] Diagnostic Recommendation Generation: The system automatically identifies weak points in the project and outputs optimization suggestions: The sustainable development dimension scored the lowest, with weaknesses concentrated in policy and regulatory alignment, quantification of human well-being, and experience dissemination. It is recommended to supplement well-being indicators such as community income growth and job creation, and strengthen the mechanism for summarizing and disseminating restoration experience. The analysis and diagnosis dimension scored low, and it is recommended to strengthen social needs research, driving factor analysis, and risk assessment in the early stages of the project. In the implementation dimension, the adaptive management mechanism scored 0, and it is recommended to embed decision trigger thresholds and dynamic adjustment rules into the monitoring plan to improve management resilience.

[0040] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for evaluating the effectiveness of coastal wetland ecosystem restoration based on nature-based solutions, characterized in that, Includes the following steps: S1. Collect multi-source data throughout the entire life cycle of the coastal wetland restoration project. First, perform format parsing and data cleaning on the multi-source data. Then, perform outlier detection on the measured monitoring data. Mark the monitoring values ​​that are identified as outliers and remove or correct them according to preset rules. Mark and supplement missing values. S2. Standardize the quantitative and qualitative indicators separately to obtain standardized scores; S3. Retrieve the pre-stored weight values ​​of the six criterion layers and their subordinate evaluation indicators. Aggregate the standardized scores of each indicator under the same criterion layer according to its weight to form the criterion layer evaluation value. Then calculate the NbS implementation effectiveness index using the following formula: , Wherein, NSI is the NbS implementation effectiveness index, and the NSI value ranges from 0 to 3; i is the criterion layer number; n is the total number of criterion layers; The weights of the i-th criterion layer; The evaluation value of the i-th criterion layer is denoted as criterion layer ; the six criterion layers are analysis and diagnosis, scheme formulation, scheme implementation, social governance, ecological effectiveness, and sustainable development. S4. Based on the NbS implementation effectiveness index, the evaluation level is automatically determined according to the preset range, and a visual chart of the score of each criterion level, a detailed table of sub-indicator scores, and a diagnostic suggestion report are generated.

2. The method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions as described in claim 1, characterized in that: In step S1, the multi-source data includes baseline survey data, restoration project implementation data, ecological monitoring data, community questionnaire survey data, and project management documents. The restoration objects include one or more of seagrass beds, mangroves, or coastal salt marshes. The outlier detection uses the median absolute deviation criterion. For a monitoring value sequence of the same indicator, the median and the median absolute deviation are calculated using the following formula: , Where MAD is the median absolute deviation; The median is a function of the median; x is the monitored value; M is the median. When satisfied When the value x is outlier, the monitored value is marked as an outlier; where λ is a preset deviation factor, with a default value of 3.

3. The method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions as described in claim 1, characterized in that, The specific process of step S2 is as follows: For quantitative indicators, the original feature values ​​of each quantitative indicator are first calculated from the measured monitoring data according to the preset quantitative calculation model, and then the original feature values ​​are normalized to a standardized score of 0 to 3 through the segmented mapping function; For qualitative indicators, the text of the project management document is segmented into words to construct a text feature vector, and the similarity is calculated one by one with the feature vectors of each level description text stored in the level description library. Based on the principle of maximum similarity and the confidence threshold, a level assignment of 0 to 3 points is automatically recommended, and then a manual review instruction is received to update and obtain the final standardized score.

4. The method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions as described in claim 3, characterized in that, In step S2, the quantitative calculation model for the quantitative indicator includes: The area verification rate is calculated using the following formula: , in, For area verification rate; This refers to the actual area repaired in the project. The planned repair area for the project; The survival rate of the target vegetation is calculated using the following formula: , Where L represents the survival rate of the target vegetation; y represents the monitoring patch number; and Y represents the total number of monitoring patches. The number of target plants surviving in the y-th monitoring patch; The total number of target plants planted or transplanted within the y-th monitoring patch; The target vegetation cover is calculated using the following formula: , in, Target vegetation cover; Let y be the vertically projected area of ​​the target vegetation canopy within the y-th monitoring patch; Let y be the total area of ​​the y-th monitoring patch; The relative density of vegetation is calculated using the following formula: , in, Relative density of vegetation; Target vegetation density for the restoration area; The vegetation density corresponding to the natural reference area or historical baseline; The relative organic carbon burial rate is calculated using the following formula: , in, This represents the relative organic carbon burial rate. The rate of organic carbon burial in the remediation area; The organic carbon burial rate is the rate corresponding to the natural reference area or historical baseline.

5. The method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions as described in claim 4, characterized in that, In step S2, the quantitative calculation model for the quantitative indicators also includes a comprehensive index of macrobenthic diversity, calculated using the following formula: , , , , , Where d is the Margalef richness index; S is the number of benthic animal species; N is the total number of benthic animal individuals; H′ is the Shannon-Wiener diversity index; and j is the benthic animal species number. The proportion of individuals of the j-th benthic species to the total number of individuals; This is the standardized value of the Margalef richness index; This serves as the benchmark value for the Margalef richness index; This is the standardized value of the Shannon-Wiener diversity index; This serves as the benchmark value for the Shannon-Wiener diversity index. The function is for finding the minimum value; This is a comprehensive index of macrobenthic diversity; α+β=1, with default values ​​of α=0.4 and β=0.

6. The weight parameters can be configured and adjusted.

6. The method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions as described in claim 5, characterized in that, In step S2, the piecewise mapping function modifies the original feature values ​​of the quantitative indicator according to the preset four-level threshold of the indicator. The standardized score is normalized to 0 to 3, with the specific segmentation rules as follows: When a ≤ t0, G = 0; When t0 < a ≤ t1, G = (a - t0) / (t1 - t0); When t1< a ≤t2, G=1+(a-t1) / (t2-t1); When t2 < a ≤ t3, G = 2 + (a - t2) / (t3 - t2); When a > t3, G = 3, which means the upper limit of saturation has been reached; Where a is the original characteristic value of the quantitative indicator; G is the standardized score; for relative indicators, t3 corresponds to the indicator reaching or exceeding the level of the natural or planned benchmark, and saturation occurs when it exceeds t3, and the standardized score is the full score of 3 points.

7. The method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions as described in claim 3, characterized in that, In step S2, the specific process of automatically recommending and assigning values ​​to the qualitative indicators is as follows: Construct a feature vector from the level description text of level k. TF-IDF weighted text feature vectors are constructed for relevant paragraphs in the project management document. Where k is the level number in the level description text, k=0,1,2,3; Calculate eigenvectors With text feature vectors The cosine similarity is calculated using the following formula: , in, For feature vectors With text feature vectors Cosine similarity; Let L2 be the norm of the vector; The level with the highest similarity is selected as the optimal recommendation level. The calculation formula is as follows: , in, The highest recommended level; To select the level with the highest similarity; The confidence level for automatic grading of qualitative indicators is calculated using the following formula: , Where conf is the confidence level of the qualitative indicator that is automatically graded; The feature vector of the text to be evaluated Description vector of the optimal recommendation level standard Cosine similarity; The second highest similarity among all remaining levels after excluding the optimal recommendation level; When the confidence level (conf) of the automatic classification of a qualitative indicator is less than the preset threshold (θ), the qualitative indicator is marked as requiring manual review. When the confidence level (conf) of the automatic classification of a qualitative indicator is greater than or equal to the preset threshold (θ), the score Gq=k* corresponding to the optimal recommendation level k* is used as the automatic recommended standardized score of the qualitative indicator. For qualitative indicators marked as requiring manual review, a manual review instruction is received, and the manually confirmed score of 0, 1, 2, or 3 is used as the final standardized score of the qualitative indicator. For those not manually corrected, the automatic recommended standardized score is used as the final standardized score.

8. The method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions as described in claim 1, characterized in that: In step S3, the weight values ​​include the criterion layer weight, the local weight of the first-level indicator, and the local weight of the second-level indicator; The default weights for the six criteria layers are as follows: the default weight for analysis and diagnosis is 0.2056, the default weight for scheme formulation is 0.1691, the default weight for scheme implementation is 0.1748, the default weight for social governance is 0.1150, the default weight for ecological effectiveness is 0.2373, and the default weight for sustainable development is 0.0983. The specific process of the weighted aggregation is as follows: First, the standardized scores of the secondary indicators are weighted and summed according to the local weights of the secondary indicators to obtain the evaluation value of the primary indicator; then, the evaluation values ​​of the primary indicators are weighted and summed according to the local weights of the primary indicators to obtain the evaluation value of the corresponding criterion layer.

9. The method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions as described in claim 1, characterized in that, In step S4, the rule for determining the evaluation level is as follows: When NSI ≥ 2.25, it is determined to be highly compliant with NbS; When 1.50 ≤ NSI < 2.25, it is determined to be basically in compliance with NbS; When 0.75 ≤ NSI < 1.50, it is determined to be partially in compliance with NbS; When NSI < 0.75, it is determined that it does not meet the NbS requirement.

10. The method for evaluating the effectiveness of coastal wetland ecosystem restoration based on natural solutions as described in claim 1, characterized in that, Step S4 is followed by a comparative analysis step, which involves comparing the results of this NSI assessment with the calculation results of the standard NbS self-assessment tool, outputting a report on the total score difference, the differences at each criterion level, and the classification of the reasons for the differences, and comparing the results with a historical database of similar projects to generate project rankings and optimization suggestions.