Ecological restoration engineering monitoring and evaluation method and system

Abstract

The present application belongs to the technical field of ecological engineering monitoring and evaluation, and particularly relates to a kind of ecological restoration engineering monitoring and evaluation method and system, the method comprises: collecting the multi-source data of ecological restoration engineering, establishing structured engineering database;Based on database and planning data, compliance audit is carried out;Change detection is carried out using multi-temporal satellite and unmanned aerial vehicle remote sensing data, and dynamic monitoring of engineering medium-term and completion is realized;Finally, the multi-dimensional quantitative evaluation index system covering management, output and effect is constructed and calculated by comprehensively monitoring data, project management information and ground investigation results in the whole cycle, and scientific and comprehensive evaluation of the implementation effect of ecological restoration engineering is realized.The present application deeply integrates process monitoring and effectiveness evaluation, and constructs a technical process covering the whole life cycle of engineering, which effectively improves the standardization level of ecological restoration engineering supervision and the scientificity and objectivity of effectiveness evaluation.

Description

Technical Field

[0001] This invention belongs to the field of ecological engineering monitoring and evaluation technology, specifically relating to a method and system for monitoring and evaluating ecological restoration projects. Background Technology

[0002] With the advancement of science and technology and the significant improvement of social productivity, problems such as rapid population growth, excessive resource consumption, environmental pollution, and ecological damage have become increasingly prominent, making ecological and environmental issues a major concern for the nation and all sectors of society. To comprehensively promote integrated, systematic, and source-based governance of mountains, rivers, forests, fields, lakes, grasslands, and deserts, various regions have actively carried out large-scale ecological restoration projects. However, due to the complexity of the targets of supervision, the wide scope, and the heavy workload, how to effectively supervise and evaluate the effectiveness of ecological restoration work during the implementation of these projects has become an urgent problem for ecological restoration management departments to solve.

[0003] While existing ecological restoration supervision and effectiveness evaluation work has adopted technologies such as remote sensing (RS) and geographic information systems (GIS) in resource and environmental surveys, it is often limited to monitoring a single link or a single indicator. This "seeing the trees but not the forest" evaluation method leads to the separation of project monitoring and supervision from post-project effectiveness evaluation into two independent links. This results in the limitation of one-sided evaluation and seriously ignores the nature of the ecosystem as a "socio-economic-natural" complex. Consequently, the final effectiveness evaluation is difficult to reliably measure the quality of the project and cannot provide objective, accurate, and reliable reference for ecological protection and subsequent project maintenance and restoration. Summary of the Invention

[0004] To address the problems existing in the prior art, the purpose of this invention is to provide a method and system for monitoring and evaluating ecological restoration projects, which effectively improves the standardization of ecological restoration project supervision and the scientific and objective nature of effectiveness evaluation, providing accurate and reliable support for project maintenance and optimization.

[0005] The technical solution of this invention is: A method for monitoring and evaluating ecological restoration projects includes the following steps: The project acquires historical multi-source data information of the construction site and project construction process data. The historical multi-source data information includes geographic data of engineering documents, geographic data of engineering design, land change survey data, ground survey data and original satellite imagery covering the project area. The project construction process data includes at least multi-temporal satellite remote sensing imagery and UAV aerial photography data before and after construction. Based on the original satellite imagery covering the project area, orthorectification, fusion, and light and color homogenization processing are performed using digital elevation models and digital orthophoto data to obtain digital orthophotos for the monitoring period. Based on the engineering design geographic data, monitoring patches are collected on the digital orthophotos for the monitoring period. Land cover classification is performed using a Class Feature Attention network. By combining a spatiotemporal attention change detection network with multi-temporal satellite remote sensing images before and during construction, the vegetation restoration area and soil disturbance range are automatically identified to determine the monitoring quantity and produce project engineering monitoring maps, generating mid-term monitoring data. Based on UAV aerial photography data before and after construction, digital orthophotos, digital surface models, and real-scene 3D models are generated. The Class Feature Attention network is used for land cover classification, and the spatiotemporal attention change detection network is used to detect changes in ground features before and after construction. Through image interpretation, comparison of real-scene 3D models, and analysis of cut and fill, the amount of work to be monitored for completion is determined. Based on field verification and measurement results, monitoring maps are produced and completion monitoring data is generated. Based on mid-term monitoring data, completion monitoring data, geographical data of engineering documents, ground survey data, and land change survey data, a multi-dimensional evaluation index system is constructed, including project management indicators, project output indicators, and project effect indicators. The weights of each indicator are assigned using the analytic hierarchy process (AHP) to achieve a comprehensive evaluation of the effectiveness of ecological restoration projects.

[0006] Preferably, the historical multi-source data information also includes engineering construction scope layer data, historical abandoned mine map patches, three-zone three-line data, and land space ecological restoration planning data; Before generating the mid-term monitoring data, spatial overlay analysis is performed based on the engineering construction scope layer data, historical abandoned mine map patches, land change survey data, three-zone three-line data, and land space ecological restoration planning data to determine the area of ​​the engineering construction scope encroached upon by permanent basic farmland, cultivated land, ecological protection red lines, and land space planning areas. Based on the analysis results of the encroached area, compliance analysis report data and negative list review report data are generated respectively to establish a legal monitoring benchmark scope for the generation of mid-term monitoring data and completion monitoring data. The data on the three zones and three lines include: land and space planning data, ecological protection red line data, and permanent basic farmland data.

[0007] Preferably, the method for comprehensively evaluating the effectiveness of the ecological restoration project includes: Based on mid-term monitoring data, completion monitoring data, geographical data of engineering documents, ground survey data, and land change survey data, multiple tertiary indicators were determined for evaluating the entire construction process. Based on multiple tertiary indicators, clustering and classification are performed to form multiple secondary indicators for evaluating the standardization of the construction process, the quantity and completion status of the project entity, and the ecological, social and economic benefits. These secondary indicators are then clustered and classified again to form primary indicators including project management indicators, project output indicators and project effect indicators. The weights of all primary, secondary, and tertiary indicators are determined. The original scores of all tertiary indicators belonging to the same secondary indicator are weighted and summed to determine the score of the corresponding secondary indicator. The scores of all secondary indicators belonging to the same primary indicator are weighted and summed to determine the score of the corresponding primary indicator. The comprehensive score for evaluating the effectiveness of the ecological restoration project is obtained by weighting and summing all primary indicators.

[0008] Preferably, the method for determining the weights of the primary and secondary indicators includes the following steps: Obtain scaling values ​​for various indicators within the same level from multiple experts; The judgment matrix for each level is determined based on the scaling values ​​of each indicator; Determine the consistency ratio of the judgment matrix. If the consistency ratio exceeds the threshold, return to readjust the judgment matrix until the requirements are met. Based on the judgment matrix of each level, determine its maximum eigenvalue and corresponding eigenvector, and normalize the eigenvector to obtain the initial weights of the corresponding indicators at each level. By combining the initial weights of all levels of indicators given by all experts, the arithmetic mean or geometric mean method is used to aggregate them to obtain the final weights of the corresponding primary and secondary indicators.

[0009] Preferably, the method for determining the weights of the three-level indicators includes: Set corresponding scoring intervals for each of the three-level indicators; Based on the scoring range of each tertiary indicator, and according to the preset scoring rules for each tertiary indicator, scores are assigned to the corresponding tertiary indicators. The scores assigned to each of the three-level indicators are used as their corresponding weight parameters.

[0010] Preferably, the project performance indicators include ecological benefit indicators, social benefit indicators, economic benefit indicators, and service recipient satisfaction indicators; wherein, the ecological benefit indicators include ecosystem quality, ecosystem services, and ecosystem structure, and the ecosystem services include water conservation, soil retention, windbreak and sand fixation, biodiversity conservation, and carbon storage.

[0011] Preferably, the quality of the ecosystem is determined according to the following formula: , In the formula, For the first i Year j Regional ecosystem quality; For the first i Year j Leaf area index and relative density of different zones; For the first i Year j Relative density of vegetation cover in different zones; For the first i Year j The relative density of total primary productivity in the region.

[0012] Preferably, the water conservation capacity is determined according to the following formula: , In the formula, Water conservation capacity; i For the first i Ecosystem type; n The total number of ecosystem types; For the first i The area of ​​ecosystem-like structures; For runoff and rainfall; Surface runoff; This refers to the evaporation rate.

[0013] Preferably, the ecosystem structure is determined according to the following formula: , in, , , In the formula, For ecosystem structure; For ecological land diversity index; The proportion of ecological land area; For the first i The proportion of ecological land landscape area to total ecological land area; m is the number of ecological land landscape types; For the first in the statistics window i line, number j The value of a column cell is 1 when the cell is ecological land, and 0 otherwise; m and n are the total number of rows and columns of the statistics window, respectively.

[0014] An ecological restoration engineering monitoring and evaluation system, used to implement any of the methods described above, comprising: The data acquisition and processing module is used to acquire historical multi-source data information of the construction site and project construction process data. The historical multi-source data information includes geographic data of engineering documents, geographic data of engineering design, land change survey data, ground survey data and original satellite imagery covering the project area. The project construction process data includes at least multi-temporal satellite remote sensing imagery and UAV aerial photography data before and after construction. The mid-term engineering monitoring module is used to orthorectify and fuse raw satellite images covering the project area using digital elevation models and digital orthophoto data, and perform uniform light and color processing to obtain digital orthophoto images for the monitoring period. Based on the engineering design geographic data, monitoring patches are collected on the digital orthophoto images for the monitoring period, and land cover is classified. By combining the spatiotemporal attention change detection network with multi-temporal satellite remote sensing images before and during construction, the vegetation restoration area and soil disturbance range are automatically identified to determine the monitoring quantity and produce project engineering monitoring maps, generating mid-term monitoring data. The project completion monitoring module is used to generate digital orthophotos, digital surface models, and real-scene 3D models based on UAV aerial photography data before and after construction, perform land cover classification, detect changes in ground features before and after construction by combining a spatiotemporal attention change detection network, and generate project completion monitoring data through image interpretation, real-scene 3D model comparison, and cut-and-fill analysis. The effectiveness evaluation module is used to construct a multi-dimensional evaluation index system based on mid-term monitoring data, completion monitoring data, geographical data of engineering documents, ground survey data, and land change survey data. This system includes project management indicators, project output indicators, and project effect indicators. The weights of each indicator are assigned using the analytic hierarchy process (AHP) to achieve a comprehensive evaluation of the effectiveness of ecological restoration projects.

[0015] Compared with existing technologies, the ecological restoration engineering monitoring and evaluation method and system of the present invention have the following beneficial effects: First, this invention integrates multi-source heterogeneous data, including engineering document geographic data, multi-temporal satellite remote sensing, and UAV aerial photography, and utilizes digital elevation models for orthorectification and Class Feature Attention networks for land cover classification. This enables automated, high-precision dynamic monitoring of vegetation restoration and soil disturbance during construction, solving the problem of narrow coverage in traditional verification methods. Second, by constructing a multi-dimensional quantitative evaluation system covering project management, outputs, and effects, and employing the analytic hierarchy process (AHP) for scientific weighting, it avoids the limitations of one-sided evaluation and deeply couples process supervision with effectiveness assessment. This allows for an objective and accurate measurement of the overall project implementation, providing reliable full-cycle technical support for the supervision and decision-making of ecological restoration projects and their subsequent maintenance. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the overall method operation in this embodiment of the invention. Figure 2 This is a flowchart of the engineering data analysis and processing technology in an embodiment of the present invention; Figure 3 This is a schematic diagram of the system framework structure in an embodiment of the present invention; Figure 4 This is a flowchart of the method in an embodiment of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention.

[0019] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0020] See Figures 1 to 4 As shown, in order to provide objective, accurate, and reliable full-cycle technical support for the actual effectiveness, long-term stability, and trend prediction of ecological protection and restoration projects, this embodiment provides a method and system for monitoring and evaluating ecological restoration projects. The core steps of this method include:

[0021] S1. Collect historical multi-source data information and project construction process data of the target project construction site. The historical multi-source data information should include at least the geographic data of project documents, geographic data of project design, layer data of the project construction scope, land use change survey data, ground survey data, historical abandoned mine map patches, data on the "three zones and three lines" (ecological protection red line, ecological protection red line, and permanent basic farmland), and original satellite imagery covering the project area. The "three zones and three lines" data include: land use planning data, ecological protection red line data, and permanent basic farmland data. The project construction process data should include at least multi-temporal satellite remote sensing imagery before and during construction, as well as UAV aerial photography data before and after construction. The geographic data of project documents should include at least project approval data, management data, financial data, and progress records.

[0022] S2. Based on the geographic data of engineering documents, geographic data of engineering design, layer data of engineering construction scope, multi-temporal satellite remote sensing images, and UAV aerial photography data, obtain mid-term monitoring data during the construction process and completion monitoring data after the completion of construction; among which, the completion monitoring data includes at least change detection results, cut and fill analysis results, and field verification data.

[0023] Specifically, the determination of mid-term monitoring data and completion monitoring data includes: S21. Based on the engineering construction scope layer data, spatially overlay it with historical abandoned mine map patches, land change survey data, land space planning data, ecological protection red line data, permanent basic farmland data, and land space ecological restoration planning data. Then, conduct project compliance analysis and negative list review on the overlaid data to generate a compliance analysis report and a negative list review report, in order to identify and eliminate illegal construction areas and establish a legal monitoring benchmark scope.

[0024] S22. Within the legal monitoring baseline range, the original satellite images covering the project area are orthorectified, fused, and processed for uniform illumination and color using digital elevation models and digital orthophoto data to obtain digital orthophotos for the monitoring period. Based on the aforementioned engineering design geographic data, monitoring patches are collected on the digital orthophotos for the monitoring period. Land cover classification is performed using a Class Feature Attention network. By combining a spatiotemporal attention change detection network with multi-temporal satellite remote sensing images before and during construction, the vegetation restoration area and soil disturbance range are automatically identified. This determines the monitoring quantity and produces project engineering monitoring maps, generating mid-term monitoring data.

[0025] S23. Based on UAV aerial photography data before and after construction, generate digital orthophotos, digital surface models, and real-scene 3D models. Use the Class Feature Attention network to classify land cover, and combine it with the spatiotemporal attention change detection network to detect changes in ground features before and after construction. Through image interpretation, comparison of real-scene 3D models, and analysis of cut and fill, determine the amount of work to be monitored as a result. Combine the field verification and measurement results to create monitoring maps and generate as-built monitoring data.

[0026] S3. Based on mid-term monitoring data, completion monitoring data, geographical data of engineering documents, ground survey data, and land change survey data, determine multiple tertiary indicators for evaluating the entire construction process.

[0027] S4. Based on multiple tertiary indicators, clustering and classification are performed to form multiple secondary indicators for evaluating the standardization of the construction process, the quantity and completion status of the project entity, and ecological, social, and economic benefits. All secondary indicators for evaluating the standardization of the construction process are categorized as project management indicators; all secondary indicators for evaluating the quantity and completion status of the project entity are categorized as project output indicators; and all secondary indicators for evaluating ecological, social, and economic benefits are categorized as project effectiveness indicators. Specifically, the indicators at each level of the project management indicators are determined based on project initiation data, management data, financial data, and progress records from the geographical data of the engineering documents; the indicators at each level of the project output indicators are determined based on change detection results, cut and fill analysis results, and field verification data from mid-term and final monitoring data; and the indicators at each level of the project effectiveness indicators are determined based on multi-temporal satellite remote sensing imagery, UAV aerial photography data, ground survey data, and land use change survey data.

[0028] S5. Project management indicators, project output indicators, and project effectiveness indicators are used as primary indicators. A multi-dimensional evaluation indicator system including primary, secondary, and tertiary indicators is constructed. Expert scoring and the analytic hierarchy process (AHP) are used to determine the weight parameters of the primary and secondary indicators, and corresponding scoring intervals are set for each tertiary indicator. Based on the scoring intervals and preset scoring rules, each tertiary indicator is scored. The score of each tertiary indicator is used as its corresponding weight parameter. Then, based on the weight parameters determined for the primary, secondary, and tertiary indicators, a weighted summation is performed at each level to determine the comprehensive evaluation score of the ecological restoration project's implementation effectiveness. Specifically: the original scores of all tertiary indicators belonging to the same secondary indicator are weighted and summed to determine the score of the corresponding secondary indicator; the weighted summation of all secondary indicators belonging to the same primary indicator is performed to determine the score of the corresponding primary indicator; and the weighted summation of all primary indicators is performed to obtain the comprehensive score used to evaluate the implementation effectiveness of the ecological restoration project.

[0029] Furthermore, to make the implementation of this method clearer, the following detailed step-by-step instructions are provided: Step 1: Collection of engineering data and establishment of engineering document geographic database, engineering design geographic database, and engineering construction scope layer data; The collected engineering data underwent data preprocessing. Engineering attributes were filled in according to detailed engineering design drawings and budget documents, generating an engineering document geographic database. An engineering design database was created by governance area, an engineering design geographic database by project type, and a design layer by specific construction measures. By integrating construction scope data collected from various counties and districts, an engineering construction scope layer data was generated. (See also...) Figure 2 As shown.

[0030] Step 2: Perform engineering data analysis and processing based on the generated engineering file geographic database, engineering design geographic database, and engineering construction scope data; In ecological restoration project monitoring, data analysis and processing based on engineering document geodatabases, engineering design geodatabases, and construction scope data requires the integration of Geographic Information System (GIS) technology and ecological professional analysis methods to form a complete process from data integration to output results. First, the data types must be clearly defined. The engineering document geodatabase contains non-spatial data such as design documents, approval documents, and monitoring reports, as well as spatial elements such as engineering locations. The engineering design geodatabase contains spatial data (mostly in CAD, SHP, or GeoDatabase formats) from design drawings, including topography, vegetation planning, and hydrological facilities. The construction scope data is mainly vector surface data (such as construction boundaries and work area divisions), and the coordinate system must be confirmed as CGCS2000 to ensure spatial consistency. The attribute table structure should be standardized, and field names should be unified (e.g., "vegetation type," "project type," etc.) to facilitate subsequent correlation analysis. GIS tools should be used to detect topological errors in the construction scope data, and the design database and construction scope data should be compared to check whether the construction boundaries exceed the design scope.

[0031] Step 3: Based on the analysis and processing results of the engineering data, conduct a pre-project approval review and prepare a compliance analysis report and a negative list review report; Based on the engineering construction scope layer data, a compliance analysis was conducted to obtain the compliance analysis results. Utilizing historical abandoned mine map patches and engineering construction scope layer data, combined with land use change survey data, "three zones and three lines" data, and land space ecological restoration planning data, a distribution map was created through overlay analysis to obtain the compliance analysis results, and a compliance analysis report was compiled. The "three zones and three lines" data includes at least: land space planning data, ecological protection red line data, and permanent basic farmland data.

[0032] Based on the compliance analysis results, the area of ​​historical abandoned mine plots and the scope of the remediation area involving permanent basic farmland, arable land, ecological protection red line area and land space planning area are calculated. A negative list review is conducted, and a negative list review report is prepared to identify and eliminate areas of illegal construction and establish a legal monitoring benchmark scope.

[0033] Step 4: Based on the legal monitoring benchmark range, and using the processed project engineering data, combined with satellite remote sensing images during the project implementation process, satellite remote sensing and machine learning technologies are employed to achieve mid-term monitoring of the ecological restoration project; The mid-term monitoring process includes: collecting Level 0 / 1 raw satellite imagery, RPC models, satellite orbital attitude parameters, sensor model parameters, ground control points (measured GPS control points, high-precision vector ground feature points, and existing high-level DOM control points, evenly distributed across the entire scene, with points required at the edges), DEM digital elevation models (1:10,000 / 1:5,000 high-precision DEMs for the project area, free of holes and gross errors), and reference datums (CGCS2000 National Geodetic Coordinate System, 1985 National Height Datum, Gauss-Krüger 3° zone projection). Using PixelGrid and ERDAS IMAGINE software, the first step is image preprocessing (importing raw satellite data, RPC files, and calibration coefficients into the processing software, unifying them into TIFF / IMG format, checking data integrity, and removing bad rows, bad columns, and noise bands), radiometric calibration, atmospheric correction (eliminating atmospheric scattering, haze, and water vapor effects to obtain the true surface reflectance and ensure color consistency across different time phases), noise removal, and thin cloud shadow processing, etc. The second step involves coarse geometric correction. Utilizing the image's built-in RPC rational polynomial model and satellite orbital attitude data, coarse geometric correction is performed without control points. This initially eliminates satellite attitude distortion, sensor internal distortion, and panoramic distortion, resulting in a coarsely corrected image that forms the basis for subsequent fine correction. The third step involves fine orthorectification. Orthorectification calculations are performed using ground control points and the DEM digital elevation model. Cubic convolution resampling or bilinear interpolation is used to ensure image clarity and smooth edges. A single-scene orthorectified image in CGCS2000 plane coordinates is output. The fourth step involves multi-scene mosaicking, light and color homogenization, and local enhancement. The processed multi-scene images are imported into EPT to generate mosaic lines. Light and color homogenization is applied to the multi-scene images to ensure overall color uniformity, eliminate uneven sunlight illumination and edge vignetting, and eliminate edge color differences, banding, and brightness variations. The overall tone is natural and consistent. All orthorectified single-scene images are stitched together along the mosaic lines to generate a continuous and seamless whole image. The fifth step involves cropping and outputting a digital orthophoto in the CGCS2000 planar coordinate system based on the project scope, and then performing a quality check on the results (planar accuracy, geometric quality, color quality, texture quality, and integrity). Finally, a qualified digital orthophoto for the monitoring period is generated.

[0034] Utilizing project engineering design data (i.e., the engineering design geographic database), the monitoring operation area was delineated and initial monitoring patches were collected on the digital orthophotos of the monitoring period through visual image interpretation. Land cover classification was performed using the Class Feature Attention Network (CAFNet), combined with the Spatiotemporal Attention Change Detection Network (STAB-Net). By comparing multi-temporal satellite remote sensing images before and during construction, the vegetation restoration area and soil disturbance range were automatically identified. The overall network hierarchy is as follows: Input remote sensing DOM image → backbone feature extraction layer → Class Center Construction Module (CCB) → Class Feature Attention Module (CAB) → Classification Decoding Head → Land cover classification results. The backbone network uses ResNet50 / ResNet101 and Swin-T lightweight remote sensing backbones, outputting multi-scale remote sensing shallow texture and deep land cover semantic features, adaptable to land cover categories in the engineering area such as forest, grassland, bare soil, buildings, disturbed surfaces, and cultivated land. The CCB category center construction module takes a backbone as input and outputs a feature map plus a coarse classification prediction probability map. Its principle is to perform soft probability weighted aggregation on each land cover category to generate category-specific feature centers. The CAB category feature attention enhancement module calculates the similarity between single-pixel features and all land cover category centers, generates category-aware attention weights through Softmax, and weightedly fuses category center features to enhance features of similar land cover types, suppress heterogeneous interference, and remove cloud / shadow noise, outputting a high-discriminative remote sensing feature map with enhanced attention. The classification decoding output head samples and restores image resolution, outputting a pixel-by-pixel land cover classification map. The specific steps for land cover classification are as follows: First, construct a remote sensing land cover classification system. The classification categories are set as follows: tree vegetation, shrub vegetation, herbaceous vegetation; artificially restored vegetation areas; undisturbed soil; construction-disturbed exposed surfaces; engineering structures and temporary access roads; water areas, wasteland, and other land uses, totaling 6 categories. Second, dataset creation and model training. Using historical remote sensing interpretation patches and field verification vectors as ground truth, the system expands upon complex samples such as those with shading, seasonal color differences, and those from construction transition periods. These are then used to train the CAFNet with multi-band remote sensing imagery. A category attention mechanism is employed to balance the accuracy of various sample types, addressing the challenge of classifying disturbed features due to limited sample size. The third step involves model inference and classification. Standard DOM images from the monitoring period are input into the trained CAFNet. The network adaptively identifies the feature type of each pixel through category centers and category attention, automatically outputting a refined land cover classification result across the entire region. The fourth step involves post-processing and optimization of the classification results. Classification boundaries are corrected using engineering-collected monitoring patches, small misclassified pixels are removed, feature outlines are smoothed, and raster data is converted to vector format to generate a land cover classification vector layer. The fifth step involves the application of the classification results. The area occupied by each type of land cover is separately calculated and used as semantic prior data for the STAB-Net spatiotemporal attention change detection network, distinguishing between existing vegetation, disturbed bare soil, and newly restored vegetation, providing a basis for feature attribute identification for change assessment.

[0035] Mid-term monitoring involves calculating monitoring quantities, monitoring project implementation, creating project engineering monitoring maps, and compiling mid-term monitoring reports containing multi-source mid-term monitoring data. All areas are calculated based on orthorectified planar projection areas, using the project's established plane coordinate system and elevation datum. Vector plots are automatically counted as pixels / plot area. The first step is soil disturbance area monitoring. This refers to the area affected by construction activities, including damage to surface vegetation, soil disturbance, and changes in surface morphology. The calculation formula is as follows:

[0036] In the formula, The total soil disturbance area is denoted as (㎡ / item). The area of ​​a single automated disturbance vector pattern is determined.

[0037] Statistical Rules: Only areas identified by STAB-Net as having exposed surface and damaged soil due to construction are counted; areas permanently occupied by the project and compliant construction land, as well as existing undisturbed bare rock areas, are excluded; the project design boundary line is overlaid to calculate the actual disturbed area within the boundary line. The second step is vegetation restoration area monitoring. This refers to the newly formed effective vegetation cover area after ecological restoration measures such as soil covering, grass planting, and tree planting are completed during the construction phase. The calculation formula is as follows:

[0038] In the formula, The total area of ​​vegetation restoration during the current phase of the project. Automatically identify the area of ​​vegetation restoration patches in a single area.

[0039] Judgment rules: T1 time phase is bare soil / disturbed surface, T2 time phase is classified by ACFNet as tree / shrub / herbaceous vegetation; natural wild mixed vegetation and original retained native vegetation are excluded; artificial afforestation restoration and herbaceous cover restoration are distinguished as two categories for statistical analysis. The third step is to monitor the compliance of disturbance. (1) The area of ​​disturbance exceeding the scope is calculated as follows: (2) The vegetation restoration completion rate is calculated using the following formula: The fourth step is to monitor land cover structure indicators. This involves directly calculating the area of ​​existing vegetation, untreated bare land, temporary land occupied by the project, and temporary access road occupied by the project area based on ACFNet classification results. The fifth step is to summarize the core monitoring quantities for the mid-term monitoring. This includes the total monitored area of ​​the project (area under red line control); the cumulative total area of ​​soil disturbance, the newly added disturbance area in this period; the area of ​​vegetation restoration completed in this period, and the restoration coverage rate; the remaining area of ​​untreated bare land; and the area of ​​disturbance exceeding the construction permit (violation monitoring indicators).

[0040] Step 5: Conduct as-built monitoring based on satellite imagery data and real-world 3D models before and after construction, combined with project engineering data; Based on digital orthophotos, digital surface models, and real-scene 3D models acquired from UAV aerial photography data before and after construction, land cover classification was performed using the Class Feature Attention Network (CAFNet), and changes in ground features before and after construction were analyzed using the Spatiotemporal Attention Change Detection Network (STAB-Net). Cut and fill volume analysis was conducted through visual interpretation of the images, comparison of the digital surface models, and comparison of the real-scene 3D models to obtain cut and fill volume data. The specific steps are as follows: First, using the raster clipping tool in ArcGIS software, the digital surface model data before and after construction was clipped according to the cut and fill volume vector range data in the construction design drawings. Second, the clipped digital surface model data for both periods were input into the input data field of the ArcGIS software's Spatial Analysis—Surface Analysis tool to calculate the cut and fill volume. Third, the calculation results were statistically analyzed. In the cut and fill raster data obtained in the previous step, raster values ​​greater than zero represent cut volume, and raster values ​​less than zero represent fill volume. All values ​​were summed to calculate the final cut and fill volume of the project. The calculation of monitoring quantities follows the same steps as the mid-term monitoring calculation method. Finally, through field surveys, the status of ecological restoration projects and the measured restoration areas of various engineering projects are verified on-site. Monitoring maps are then created, and a project completion monitoring report is compiled. Based on the multi-source completion monitoring data obtained from the completion monitoring results, a completion monitoring report is written and submitted, achieving full-cycle monitoring.

[0041] Step 6: Based on the above monitoring data (compliance analysis report, negative list review report, mid-term monitoring report, and completion monitoring report), and in conjunction with the project management information, ground survey data, and land use survey data in the aforementioned engineering document geographic database, conduct an assessment of the effectiveness of ecological protection and restoration; Using remote sensing monitoring data and ground survey data from the project, we calculated data such as vegetation cover (FVC), leaf area index (LAI), gross primary productivity (GPP), water conservation, soil conservation, and biomass. We then combined these data with land use survey data to evaluate the effectiveness of ecological protection and restoration. ① Construct an evaluation index system for ecological restoration projects of historically abandoned mines This study evaluates the construction achievements of mine ecological restoration and governance projects from three aspects: "project management," "project output," and "project effectiveness." It employs a combination of qualitative and quantitative analysis, constructing an evaluation index system encompassing 10 secondary indicators and 26 tertiary indicators, including "preliminary management," "business management," "financial management," "quantitative indicators," "quality indicators," "timeliness indicators," "social benefit indicators," "ecological benefit indicators," "economic benefit indicators," and "service recipient satisfaction." The index system design is shown in Table 1.

[0042] Table 1 Evaluation Index System for Ecological Restoration Projects of Historically Abandoned Mines ② Determine the weights of evaluation indicators for ecological restoration projects of abandoned historical mines. We collected public satisfaction questionnaires and expert scoring sheets through questionnaires and expert scoring methods. We used the analytic hierarchy process (AHP) to analyze the weights of the factor and indicator layers. Based on expert suggestions, we revised the indicator weights again and finally determined the weights of the specific indicators.

[0043] ③ Determine the scoring method for each evaluation indicator. The total score for evaluating the implementation effectiveness of ecological restoration projects for historically abandoned mines is set at 100 points. Based on the determined indicator weights, the three secondary indicators—"Project Management Indicators," "Project Output Indicators," and "Project Effectiveness Indicators"—are assigned 23, 35, and 42 points respectively. The specific scoring methods for each indicator are as follows:

[0044] Indicator 1 - Project Establishment Standardization Calculation method: 3.5 points are awarded if all important process documents such as performance targets, implementation plans, and design plans have been submitted to the relevant units for approval and filing in a timely manner; 2 points are awarded if some documents have not been submitted for approval or filing, or if the submission and filing procedures are delayed; and 0 points are awarded if the submission and filing procedures are basically not completed.

[0045] Indicator 2 - Reasonableness of the Implementation Plan Calculation method: If both the organization and implementation method and the schedule are reasonable and feasible, 3.5 points will be awarded; if the design in one aspect is not entirely reasonable, 2 points will be awarded; if the overall or most of the design is unreasonable, 0 points will be awarded.

[0046] Indicator 3 - Soundness of Project Management System Calculation method: If the above-mentioned project management system is sound and can better regulate the implementation process of the project, it will receive 2.5 points; if the relevant management system is not sound, it will receive 0 points.

[0047] Indicator 4 - Project Organization Effectiveness Calculation method: 2.5 points are awarded for projects that are organized in stages and achieve results; 1 point is awarded for projects that lack key links; and 0 points are awarded for projects that are poorly organized and have low management efficiency.

[0048] Indicator 5 - Project Schedule and Content Control Calculation method: If the actual progress of the project (from the start date to the completion of the construction work, which can be obtained from the supervision data) is generally consistent with the expected progress (the schedule proposed in the approved design documents) (within 10%), 3 points will be awarded; if the actual construction period is delayed, but not exceeding 50% of the expected construction period, 1.5 points will be awarded; if the construction period is severely delayed, exceeding 50% of the expected plan, 0 points will be awarded.

[0049] Indicator 6 - Budget Funding Availability Rate Calculation method: The formula for calculating the budgeted funds arrival rate is the actual amount of funds disbursed divided by the budgeted amount of funds. If the arrival rate is ≥80%, 2 points are awarded; if the rate is between 50% and 80%, 1 point is awarded; and if the rate is below 50%, 0.5 points are awarded.

[0050] Indicator 7 - Standardization of Fund Use Calculation method: If the work in the above 3 aspects is in place and in accordance with the standards, 2.5 points will be awarded; if one of the 3 aspects does not meet the corresponding requirements, 1.5 points will be awarded; if two or all three aspects do not meet the relevant requirements, 0 points will be awarded.

[0051] Indicator 8 - Effectiveness of Financial Monitoring Calculation method: If a financial monitoring system is established and necessary monitoring measures or means such as financial inspection are taken during the project implementation, 2 points will be awarded; if a financial monitoring system is established but not fully implemented, 1 point will be awarded; if no financial monitoring system is established, 0 points will be awarded.

[0052] Indicator 9 - Soundness of Financial Management System Calculation method: 1.5 points are awarded for a sound financial management system that is implemented in accordance with the financial management system during project implementation; 0 points are awarded for an unsound management system or one that is not strictly implemented.

[0053] Indicator 10 - Reclamation and utilization rate of historical industrial and mining wasteland in the implementation area Calculation method: The reclamation utilization rate refers to the ratio of the restored land area to the damaged land area, usually expressed as a percentage. A reclamation utilization rate greater than 80% earns 5.5 points; 60%–80% earns 4 points; 40%–60% earns 3 points; 20%–40% earns 2 points; and less than 20% earns 1 point.

[0054] Indicator 11 - Percentage of Project Completion Calculation method: Completion rate is calculated by comparing the completed workload with the total workload. A completion rate greater than 90% earns 6.5 points; 70%–90% earns 5 points; 50%–70% earns 3 points; 30%–50% earns 1 point; and less than 30% earns 0.5 points.

[0055] Index 12 - Stability of the restored geological environment Calculation method: 4 points are awarded if the stability of the geological environment is significantly improved; 2 points are awarded if it is improved to a certain extent; and 0 points are awarded if it is not improved.

[0056] Indicator 13 - Project Acceptance Rate Calculation method: 3.5 points are awarded for a project acceptance pass rate of 95% to 100%; 2 points are awarded for a pass rate of 80% to 95%; and 1 point is awarded for a pass rate of less than 80%.

[0057] Indicator 14 - Harmony between the restored environment and the surrounding ecosystem Calculation method: 3 points are awarded for a significant improvement in harmony with the surrounding ecological environment after restoration; 2 points are awarded for a certain degree of improvement; and 0 points are awarded for no improvement.

[0058] Indicator 15 - Conformity between restored land use and national spatial planning Calculation method: If the land use after restoration meets 100% of the national land use plan, 3.5 points will be awarded; if it does not meet 100%, 0 points will be awarded.

[0059] Indicator 16 - Develop and launch the implementation plan on time. Calculation method: 2.5 points are awarded for projects that prepare and launch implementation plans on time; 1 point is awarded for projects that do not prepare and launch implementation plans on time.

[0060] Indicator 17 - Timeliness of Project Acceptance Calculation method: 2.5 points are awarded for projects that are completed and accepted on time; 1 point is awarded for projects that are not completed and accepted on time.

[0061] Indicator 18 - On-time completion rate of projects Calculation method: Based on the project construction content, 4 points are awarded for an on-time completion rate of 95%~100%; 3 points are awarded for an on-time completion rate of 80%~95%; 2 points are awarded for an on-time completion rate of 70%~80%; 1 point is awarded for an on-time completion rate of 60%~70%; and 0 points are awarded for an on-time completion rate of less than 60%.

[0062] Indicator 19 - Geological Hazards in Mines in the Implementation Area Calculation method: 5 points are awarded if all potential geological hazards in the mining area are eliminated; 2.5 points are awarded if some potential geological hazards in the mining area are eliminated; and 0 points are awarded if no potential geological hazards in the mining area are eliminated.

[0063] Indicator 20 - Ensuring the safety of people's lives and property in the implementation area Calculation method: 5 points are awarded if the safety of people's lives and property is guaranteed in the implementation area; 0 points are awarded if the safety is not guaranteed.

[0064] Indicator 21 - Ecosystem Quality Calculation method: Ecosystem quality (EQI) reflects the overall quality of a regional ecosystem and is constructed from the relative density of vegetation cover (FVC), leaf area index (LAI), and gross primary productivity (GPP).

[0065] , in, For the first i Year j Regional ecosystem quality; For the first i Year j Leaf area index and relative density of different zones; For the first i Year j Relative density of vegetation cover in different zones; For the first i Year j The relative density of total primary productivity in each region. Calculation. EQI According to the ecosystem quality grading standards in the "National Ecological Status Survey and Assessment Technical Specifications - Ecosystem Quality Assessment", ecosystem quality is divided into five levels: excellent, good, medium, low, and poor. EQI The calculation results classify the ecosystem quality of the project area into five levels, with scores of 4, 3, 2, 1, and 0 respectively.

[0066] Indicator 22 - Ecosystem Services Calculation method: Ecosystem services reflect the functions of ecosystems in providing human beings, such as water conservation, soil retention, windbreak and sand fixation, biodiversity protection, and carbon storage. Based on remote sensing and ground survey data, combined with long-term ecosystem monitoring data, the status of ecosystem service functions is quantitatively assessed to understand the changing trends of ecosystem service functions.

[0067] Water conservation capacity is calculated using the water balance equation: , in, Water conservation capacity; i For the firsti Ecosystem type; n The total number of ecosystem types; For the first i The area of ​​ecosystem-like structures; For runoff and rainfall; Surface runoff; This refers to the evaporation rate.

[0068] Soil retention capacity is calculated based on the Modified Soil Loss Equation (RUSLE): Soil retention capacity = Potential soil loss - Actual soil loss Potential soil loss (Ap): Soil loss without vegetation / measures. Ap = R × K × L × S Actual soil loss (Aa): Soil loss with vegetation / measures Aa = R × K × L × S × C × P Soil retention capacity (Sc): Sc = Ap − Aa = R × K × L × S × (1 − C × P) R: Rainfall erosivity factor (MJ·mm / (hm²·h·a), calculated based on rainfall data); K: Soil erodibility factor (t·hm²·h / (hm²·MJ·mm), based on soil texture data); L, S: Slope length and slope factors (dimensionless, calculated from DEM data); C: Vegetation cover factor (dimensionless, related to FVC and LAI). (k is a coefficient); P: Soil and water conservation measures factor (dimensionless, such as P=0.1 for terraced fields and P=1 for bare land).

[0069] The amount of windbreak and sand fixation is calculated using the modified wind erosion equation RWEQ: Windbreak and sand fixation capacity: ,in, Potential wind erosion, This represents the actual wind erosion. In the RWEQ model, wind erosion calculation requires integrating multiple factors such as wind force, soil, vegetation, and topography. The formula is as follows:

[0070] Potential wind erosion: Actual wind erosion: In the formula: The amount of sand fixed is expressed in t / (km²). 2 *a), Potential wind erosion, t / (km²)2 *a), This represents the actual wind erosion amount, in t / (km²). 2 *a), The critical length for wind erosion, in meters. Maximum transfer volume, kg / m³ The potential maximum transfer amount is kg / m³. z The distance at which maximum wind erosion occurs, in meters. WF Climate factor, kg / m For surface roughness factor, EF As a soil erosion factor, SCF As a soil crusting factor, C This refers to vegetation cover factors.

[0071] Carbon storage was determined using the following method: The Carbon module of the InVEST model was used to calculate carbon storage. Based on the principle that "land use type determines carbon pool size," the Carbon module assesses regional carbon storage by summing the total carbon pool (aboveground biomass, belowground biomass, soil carbon, and dead organic matter) for different land use types. By loading the Carbon module into InVEST and inputting multi-period land use data and carbon density data, the change in regional carbon storage can be calculated.

[0072] Biodiversity is determined using the following method: Habitat Quality from the InVEST model is used to indirectly characterize biodiversity; higher habitat quality correlates with richer biodiversity. The Habitat Quality module, based on land use data, assesses habitat resilience to threats, reflecting potential biodiversity levels. In the Habitat Quality module of the InVEST model, multiple periods of land use data and threat source data are input. Threat source data primarily includes data on construction land and roads, allowing for the calculation of regional habitat quality changes.

[0073] Water conservation, soil retention, windbreak and sand fixation, carbon storage, and biodiversity were calculated before and after the implementation of the ecological restoration project. The results were standardized and then summed using an equal-weighted method to construct the ecosystem service value. The ecosystem service value was divided into four levels: excellent, good, medium, and low, with scores of 3, 2, 1, and 0 respectively.

[0074] Indicator 23 - Ecosystem Structure Calculation Method: For ecosystem structure assessment, land use data of the project implementation area before and after construction were used to establish two corresponding land use structure change tables, analyzing the land use changes in the project area before and after construction. The ecosystem structure (ES) in this invention couples ecological land diversity and ecological land area proportion. After standardization, it is constructed using an equal-weighted summation method, with the following formula:

[0075] , Ecological land diversity ignores non-ecological land and only considers ecological land (grassland, water bodies, forests, wetlands, etc.), and is calculated based on the Shannon diversity index. The formula is as follows: , The proportion of ecological land area is calculated by combining different types of ecological land and determining the percentage of total ecological land area within the landscape. The formula is: , In the formula, ES For ecosystem structure; For ecological land diversity index; The proportion of ecological land area; For the first i The proportion of ecological land landscape area to total ecological land area; m is the number of ecological land landscape types; For the first in the statistics window i line, number j The value of a column cell is 1 when the cell is ecological land, and 0 otherwise; m and n are the total number of rows and columns of the statistics window, respectively.

[0076] Calculate before and after construction ES ,analyze ES Change ∆ES ,when ∆ES When >0, 3 points are awarded. ∆ES When =0, get 1 point; when ∆ES If the score is less than 0, you get 0 points.

[0077] Indicator 24 - Remote Sensing Ecological Index Calculation Method: Using Landsat TM / OLI remote sensing satellite data, and based on geographic information technology and related statistical analysis methods, remote sensing satellite images before and after the construction of the ecological restoration project for historical abandoned mines were processed. Ecological indicators reflecting the project area, such as vegetation cover, humidity, dryness, and surface temperature, were retrieved. Based on the selection of indicators and data processing methods for the Remote Sensing Ecological Index (RSEI), a principal component analysis model was used to integrate four index indicators: Normalized Difference Vegetation Index (greenness), humidity, surface temperature (heat), and building index (dryness), to classify the ecological status. According to the "Technical Specification for Ecological Environment Status Assessment" (HJ 192—2015), the remote sensing ecological index calculated before and after the project implementation was divided into five levels: extremely poor (RSEI value 0~0.2), poor (0.2~0.4), moderate (0.4~0.6), good (0.6~0.8), and excellent (0.8~1.0). An RSEI index between 0.0 and 0.2 indicates a poor ecosystem quality level, earning 1 point; between 0.2 and 0.4 indicates a relatively poor ecosystem quality level, earning 2 points; between 0.4 and 0.6 indicates a moderate ecosystem quality level, earning 3 points; between 0.6 and 0.8 indicates a good ecosystem quality level, earning 4 points; and between 0.8 and 1.0 indicates an excellent ecosystem quality level, earning 5 points.

[0078] Indicator 25 - Economic Benefits of Reclaimed Land Calculation method: 10 points for reclamation utilization rate above 90%; 8 points for reclamation utilization rate between 80% and 90%; 7 points for reclamation utilization rate between 70% and 80%; 6 points for reclamation utilization rate between 60% and 70%; 5 points for reclamation utilization rate between 50% and 60%; 4 points for reclamation utilization rate between 40% and 50%; 3 points for reclamation utilization rate between 30% and 40%; 2 points for reclamation utilization rate between 20% and 30%; and 1 point for reclamation utilization rate below 20%.

[0079] Indicator 26 - Public satisfaction in the implementation area Calculation method: 7 points for a satisfaction rate of over 80% among the beneficiaries in the implemented area; 6 points for a satisfaction rate of 70%-80%; 5 points for a satisfaction rate of 60%-70%; 4 points for a satisfaction rate of 50%-60%; 3 points for a satisfaction rate of 40%-50%; 2 points for a satisfaction rate of 30%-40%; 1 point for a satisfaction rate of 10%-20%; and 0 points for a satisfaction rate below 10%.

[0080] Based on the collected data and on-site surveys, the evaluation results of the ecological restoration projects for historical abandoned mines were obtained through the constructed evaluation indicators and methods, and the restoration effectiveness of the ecological restoration projects was comprehensively evaluated.

[0081] Specifically, the scores of indicators at all levels will be weighted and integrated: S61. Determine the base scores for the primary indicators: Preset base scores for the three primary indicators: “Project Management Indicators”, “Project Output Indicators”, and “Project Effectiveness Indicators”. S62. Calculate the scores of the three-level indicators: Based on the scoring rules set for each three-level indicator, determine and calculate the original score of each three-level indicator. S63. Calculate the weighted score of the secondary indicator: sum the original scores of all tertiary indicators belonging to the same secondary indicator to obtain the score of that secondary indicator; S64. Calculate the weighted score of the primary indicator: sum the scores of all secondary indicators belonging to the same primary indicator to obtain the score of the primary indicator. S65. Calculate the overall evaluation score: The scores of all primary indicators are weighted and summed. The sum is the overall score used to evaluate the effectiveness of the ecological restoration project.

[0082] Step 7: Construction of a full-cycle monitoring and effectiveness evaluation platform for mine ecological restoration projects, including data acquisition and processing modules, project review modules, mid-term project monitoring modules, completed project monitoring modules, effectiveness evaluation modules, and a database.

[0083] The data acquisition and processing module collects and processes raw data from ecological restoration projects, establishing an engineering database that includes a geographic database of engineering documents, a geographic database of engineering design, and layer data of the engineering construction scope. The engineering review module performs spatial overlay analysis based on the engineering construction scope layer data, comparing it with historical abandoned mine patches, land use change survey data, "three zones and three lines" data, and land space ecological restoration planning data. It conducts project compliance analysis and negative list review, generating compliance analysis reports and negative list review reports. The mid-term engineering monitoring module acquires satellite remote sensing imagery during project construction, processes it to generate digital orthophotos for the monitoring period; collects monitoring patches on the digital orthophotos based on the engineering design geographic database; classifies and detects land cover changes in the images before and after construction; identifies the extent of vegetation restoration and soil disturbance within the monitoring patches; and generates mid-term monitoring data. The completed project monitoring module generates digital orthophotos, digital surface models, and realistic 3D models based on UAV aerial photography data before and after construction. It detects changes in ground features before and after construction and calculates monitored project quantities by combining image interpretation, 3D model comparison, and cut / fill analysis data, generating completed project monitoring data. The effectiveness evaluation module constructs a multi-dimensional evaluation index system covering project management indicators, project output indicators, and project effect indicators based on compliance analysis reports, negative list review reports, mid-term monitoring data, and completed project monitoring data, combined with project management information, ground survey data, and land use survey data from the project document geodatabase. It quantifies and scores each indicator to comprehensively evaluate the effectiveness of ecological restoration projects. The database stores various data, models, and reports generated during the data acquisition and processing, project review, project monitoring, and effectiveness evaluation modules.

[0084] More detailed: (1) For the overall platform architecture, please refer to Figure 3 As shown, it includes: The monitoring and perception layer integrates multi-source monitoring data such as satellite remote sensing, UAV imagery, video, and lidar, and gathers diverse data including basic geography, spatial planning, socio-economics, and business management to conduct comprehensive and all-round monitoring of natural resources such as mountains, rivers, forests, fields, lakes, grasslands, climate, and biology, as well as human engineering activities.

[0085] Data layer: Integrates project-related data, including basic data, business management data, monitoring data, indicator model data, etc., to provide data support for the system and ensure that data moves as needed.

[0086] The supporting layer establishes intelligent analysis capabilities based on an indicator model system.

[0087] At the application layer, scientific support for information technology construction is provided through business analysis and modeling, and related thematic research. Based on the knowledge reasoning capabilities of indicator models, application modules that run through the entire process of ecological restoration, from pre-assessment to in-process monitoring to post-evaluation, are built to provide full-cycle information technology support services for unified national land space ecological restoration.

[0088] The user layer is used by the natural resources authorities at the provincial, municipal, and county levels.

[0089] (2) Ecological restoration supervision database A unified "data base" for ecological restoration monitoring projects across the province has been established. Based on the Gansu Provincial Territorial Spatial Basic Information Platform, it aggregates current status data, planning data, implementation data, and monitoring data for ecological restoration projects. Building upon this foundation, and leveraging the correlation between data and operational needs, information barriers are broken down to create a unified map of territorial spatial ecological restoration data. This enables comprehensive data management and provides a data foundation for ecological restoration project management, supervision, and oversight.

[0090] (3) Ecological restoration "one map" The system provides an overview of all ecological restoration projects in Gansu Province, including comprehensive land consolidation, mine ecological restoration, and ecological restoration of mountains, rivers, forests, fields, lakes, grasslands, and deserts. It displays a comprehensive project list, basic information, location of map features, and navigation. All project information, ownership information, project progress, and pre-, mid-, and post-restoration stage images, progress data, pictures, videos, documents, and other materials can be clearly and intuitively displayed on the map.

[0091] ①Overall Overview This provides an overview of ecological restoration projects across Gansu Province, including comprehensive land consolidation, mine ecological restoration, and the restoration of mountains, rivers, forests, fields, lakes, grasslands, and deserts. The overview includes statistics on project types, restoration funds, area, and progress, and allows for the tracking of the total number of projects in conjunction with a map.

[0092] ②Comprehensive Inquiry Based on the current status of ecological restoration projects and combined with spatial and operational data, the system allows for coordinate and place-name location searches. It supports querying and displaying information based on project time, stage, keywords, and category, encompassing planning management, project management, comprehensive evaluation, and dynamic monitoring. The system also supports exporting query results, linking image and text information to map features, and automatically locating map features. It provides query and statistical functions for various ecological restoration projects, allowing searches by project name, project number, contract number, and other keywords, supporting fuzzy searches by project number and contract number. Categorized searches are also possible, such as by administrative division, project stage, and year.

[0093] ③Resource Directory The system supports user-selectable layers, control over their display, and allows for layer overlay for easy comparison. It also enables quick view location of layers. Users can view the layer feature list, query features within the list, and highlight them. This includes the following layer data:

[0094] Ecological restoration status data layers: remote sensing image data layers, third national land survey data, land use status change survey data, etc.; Ecological restoration planning data layers include: overall land space planning, "dual" planning, provincial land space ecological restoration planning, ecological protection red line, permanent basic farmland, urban development boundary, primary protection zone of water source area, project planning, etc. Ecological restoration project data: national key projects, provincial key projects, relationships between projects and their sub-projects, project scope, ecological restoration zones, ecological protection and restoration units, scope of sub-projects, etc. Basic geographic data: provincial boundaries, county boundaries, township boundaries, etc.

[0095] ④ Online Inquiry Based on the set query conditions and the drawn query spatial range, the system queries and displays the elements in the specified layer, locates the specific element on the map, displays the associated project information, and exports the project information.

[0096] ⑤ Online review Conduct compliance and duplication reviews of the project, including statistics on the encroachment situation in the "three zones and three lines" and analysis of the encroachment situation with other projects.

[0097] ⑥ Special Topic Analysis Conduct thematic analyses of the project using maps, including current status analysis, planning analysis, and farmland protection analysis, and statistically analyze the distribution of various land types within the project area.

[0098] ⑦ Planning Information By combining map analysis with project planning, the scope and budget of each construction activity can be reviewed. The project can be managed with meticulous detail, marking the location and extent of each planned construction activity on the image map, down to the level of ditches.

[0099] ⑧ Monitoring Information By combining the monitoring records with the map, including reported image information, video information, video information, and monitoring reports, the effectiveness of ecological restoration can be displayed intuitively.

[0100] ⑨ Monitoring and comparison By comparing the planned and monitored land parcels, we can analyze the ecological changes within the project area monitored during the project implementation process and make the project results traceable.

[0101] ⑩ Excavation Analysis The changes in earthwork volume before and after the project were obtained by comparing the digital elevation model (DEM) data produced by using UAV oblique photography with the digital elevation model (DEM) data produced by extracting elevation information from the design topographic map (topography before the project started).

[0102] ⑪ Scene Roaming It provides basic 3D map browsing, allowing users to roam 360 degrees and view a specified location from all angles, facilitating a quick understanding of how the location is viewed from different perspectives. It also offers scene marking and screenshot functions, enabling fast and efficient scene marking and positioning.

[0103] ⑫ Three-dimensional measurement Provides on-map measurement functions to meet the needs of measuring distances and areas in various scenarios. Offers on-map measurement tools that can measure spatial distances, ground-level distances, height differences, areas, and more.

[0104] ⑬ Split-screen browsing Different layout windows are selected based on the number of display layers, and the screen view is evenly distributed with horizontal viewport, vertical viewport, and four views. The image data of each view is displayed and hidden, and the actual data at different stages of the project are compared.

[0105] ⑭ Roller blind comparison The system provides a roll-up analysis tool that supports split-screen mode, allowing interactive comparison of two layers. By rolling up the upper image and comparing it with the lower image, the differences in spatiotemporal data can be intuitively displayed through the roll-up comparison between views, facilitating comparative analysis.

[0106] In summary, the present invention has the following technical effects: This invention, based on multi-temporal remote sensing data, fully utilizes various natural resource survey and monitoring results, planning data, and engineering implementation data. It employs aerospace, airborne, and low-altitude remote sensing technologies, artificial intelligence, and mobile patrol monitoring techniques to classify and monitor the progress and performance of ecological restoration projects within the study area. This achieves refined management of the entire lifecycle of engineering projects. Furthermore, based on the multi-scale characteristics of ecosystems, it constructs an effectiveness evaluation system for ecological protection and restoration of national land space, along with a full-cycle monitoring system to support this effectiveness evaluation. The technological advantages are reflected in the following four aspects:

[0107] (1) Multi-source data fusion and integration: Data acquired by various sensors such as satellite remote sensing, aerial photography, UAV remote sensing, and ground monitoring stations are fused to achieve comprehensive and multi-scale monitoring of the ecological restoration area; (2) High-precision remote sensing monitoring technology: using high-resolution satellite imagery to identify land use changes and vegetation cover in ecological restoration areas, and accurately monitor the implementation effect of the project; (3) Application of Internet of Things and sensor technology: By deploying a large number of Internet of Things sensors in the ecological restoration area, data can be collected in real time to realize dynamic monitoring of the ecological environment; (4) Dynamic monitoring and evaluation system: Establish a dynamic monitoring system to monitor all aspects of the ecological restoration project in real time, promptly identify problems and take corresponding measures. Construct a scientific and reasonable ecological restoration evaluation index system to comprehensively evaluate the project effect from multiple dimensions such as ecology, environment, society, and economy, and fully reflect the actual effectiveness of the ecological restoration project.

[0108] The core technology of this invention is embodied in the following five aspects: (1) A comprehensive and three-dimensional monitoring system of "air, space, and man" has been constructed: a comprehensive dynamic monitoring network has been formed, which includes satellite imagery "seeing from the sky", drones "exploring from the air", cameras "capturing in real time", and mobile APP "verifying on-site", so as to accurately, intuitively and truthfully grasp the progress and effectiveness of ecological restoration projects.

[0109] (2) A monitoring and supervision information platform and a supporting mobile inspection APP have been built: The system integrates functions such as planning and control review, multi-source monitoring data integration, three-dimensional visualization display, quantitative indicator evaluation, mobile inspection and evidence collection, and intelligent dynamic early warning, so as to realize the traceability of project management process and the measurability of implementation progress, and improve the standardization and informatization of monitoring and supervision of ecological restoration projects.

[0110] (3) Improved the full-cycle database for ecological restoration project management: covering the entire cycle of project establishment and design, implementation and restoration, acceptance and evaluation, results management, maintenance and monitoring, integrating multi-source heterogeneous data such as remote sensing monitoring, patrol images, government approvals, management archives, and financial statements, realizing digital management of projects, supporting "menu-style" services, and assisting regulatory departments in making scientific decisions.

[0111] (4) A multi-scale effectiveness evaluation system for land space ecological restoration was constructed: A multi-scale effectiveness evaluation system for land space ecological restoration across the province was constructed, and the effectiveness evaluation of land space ecological protection and restoration across the province was carried out at the regional, ecosystem, and engineering scales. The evaluation model of the Gansu Province Land Space Ecological Restoration Project Management System was improved by combining economic and ecological benefit indicators, a continuous monitoring and supervision model was explored, performance evaluations were conducted on the accepted projects, and relevant algorithms and models for ecological restoration effect evaluation were integrated into the monitoring system, thus improving the system's evaluation function.

[0112] (5) Explored evaluation indicators and technical paths for the implementation effectiveness of ecological restoration projects: In response to the need for monitoring and evaluation of ecological restoration of abandoned mines in the past, we have leveraged the advantages of technologies such as remote sensing and geographic information, based on national ecological protection and mine ecological restoration policies and regulations, and with reference to the current national and industry-related ecological protection technical standards, to conduct research on the overall requirements, workflow, monitoring content, monitoring methods, and evaluation methods of mine ecological restoration monitoring and evaluation. In accordance with the monitoring and evaluation requirements in the "Technical Specification for Monitoring and Evaluation of Mine Land Reclamation and Ecological Restoration" (GBT43935-2024), we have conducted research on the evaluation indicator system for the effectiveness of mine ecological restoration projects, comprehensively analyzed the implementation effect of mine restoration projects, scientifically evaluated the restoration effectiveness, and formed an evaluation model that can be learned from and promoted.

[0113] This invention aims to construct an information-based regulatory system for ecological restoration of national land space. Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for monitoring and evaluating ecological restoration projects, characterized in that, Includes the following steps: Acquire geographic data of engineering documents, geographic data of engineering design, land change survey data, ground survey data, original satellite imagery covering the project area, multi-temporal satellite remote sensing imagery before and during construction, and UAV aerial photography data before and after construction. Based on the original satellite imagery covering the project area, orthorectification, fusion, and light and color homogenization processing are performed using digital elevation models and digital orthophoto data to obtain digital orthophotos for the monitoring period. Based on the engineering design geographic data, monitoring patches are collected on the digital orthophotos for the monitoring period. The Class Feature Attention network is used for land cover classification. By combining the spatiotemporal attention change detection network with multi-temporal satellite remote sensing images before and during construction, the vegetation restoration area and soil disturbance range are identified, the monitoring quantities are determined, and the project engineering monitoring map is produced to generate mid-term monitoring data. Based on UAV aerial photography data before and after construction, digital orthophotos, digital surface models, and real-scene 3D models are generated. The Class Feature Attention network is used for land cover classification, and the spatiotemporal attention change detection network is used to detect changes in ground features before and after construction. Through image interpretation, comparison of real-scene 3D models, and analysis of cut and fill, the amount of work to be monitored as a result is determined. Based on field verification and measurement results, monitoring maps are produced and as-built monitoring data is generated. Based on mid-term monitoring data, completion monitoring data, geographical data of engineering documents, ground survey data, and land change survey data, a multi-dimensional evaluation index system is constructed, including project management indicators, project output indicators, and project effect indicators. The weights of each indicator are assigned using the analytic hierarchy process (AHP) to achieve a comprehensive evaluation of the effectiveness of ecological restoration projects.

2. The method for monitoring and evaluating ecological restoration projects according to claim 1, characterized in that, The historical multi-source data information also includes engineering construction scope layer data, historical abandoned mine map patches, data on the three zones and three lines, and land space ecological restoration planning data; Before generating the mid-term monitoring data, spatial overlay analysis is performed based on the engineering construction scope layer data, historical abandoned mine map patches, land change survey data, three-zone three-line data, and land space ecological restoration planning data to determine the area of ​​the engineering construction scope encroached upon by permanent basic farmland, cultivated land, ecological protection red lines, and land space planning areas. Based on the analysis results of the encroached area, compliance analysis report data and negative list review report data are generated respectively to establish a legal monitoring benchmark scope for the generation of mid-term monitoring data and completion monitoring data. The data on the three zones and three lines include: land and space planning data, ecological protection red line data, and permanent basic farmland data.

3. The method for monitoring and evaluating ecological restoration projects according to claim 2, characterized in that, The method for comprehensively evaluating the effectiveness of the ecological restoration project includes: Based on mid-term monitoring data, completion monitoring data, geographical data of engineering documents, ground survey data, and land change survey data, multiple tertiary indicators were determined for evaluating the entire construction process. Based on multiple tertiary indicators, clustering and classification are performed to form multiple secondary indicators for evaluating the standardization of the construction process, the quantity and completion status of the project entity, and the ecological, social and economic benefits. These secondary indicators are then clustered and classified again to form primary indicators including project management indicators, project output indicators and project effect indicators. The weights of all primary, secondary, and tertiary indicators are determined. The original scores of all tertiary indicators belonging to the same secondary indicator are weighted and summed to determine the score of the corresponding secondary indicator. The scores of all secondary indicators belonging to the same primary indicator are weighted and summed to determine the score of the corresponding primary indicator. The comprehensive score for evaluating the effectiveness of the ecological restoration project is obtained by weighting and summing all primary indicators.

4. The method for monitoring and evaluating ecological restoration projects according to claim 3, characterized in that, The method for determining the weights of the primary and secondary indicators includes the following steps: Obtain scaling values ​​for various indicators within the same level from multiple experts; The judgment matrix for each level is determined based on the scaling values ​​of each indicator; Determine the consistency ratio of the judgment matrix. If the consistency ratio exceeds the threshold, return to readjust the judgment matrix until the requirements are met. Based on the judgment matrix of each level, determine its maximum eigenvalue and corresponding eigenvector, and normalize the eigenvector to obtain the initial weights of the corresponding indicators at each level. By combining the initial weights of all levels of indicators given by all experts, the arithmetic mean or geometric mean method is used to aggregate them to obtain the final weights of the corresponding primary and secondary indicators.

5. The method for monitoring and evaluating ecological restoration projects according to claim 3, characterized in that, The method for determining the weights of the three-level indicators includes: Set corresponding scoring intervals for each of the three-level indicators; Based on the scoring range of each tertiary indicator, and according to the preset scoring rules for each tertiary indicator, scores are assigned to the corresponding tertiary indicators. The scores assigned to each of the three-level indicators are used as their corresponding weight parameters.

6. The method for monitoring and evaluating ecological restoration projects according to claim 1, characterized in that, The project's performance indicators include ecological benefit indicators, social benefit indicators, economic benefit indicators, and service recipient satisfaction indicators. Among them, the ecological benefit indicators include ecosystem quality, ecosystem services, and ecosystem structure. The ecosystem services include water conservation, soil retention, windbreak and sand fixation, biodiversity conservation, and carbon storage.

7. The method for monitoring and evaluating ecological restoration projects according to claim 6, characterized in that, The quality of the ecosystem is determined according to the following formula: ， In the formula, For the first i Year j Regional ecosystem quality; For the first i Year j Leaf area index and relative density of different zones; For the first i Year j Relative density of vegetation cover in different zones; For the first i Year j The relative density of total primary productivity in the region.

8. The method for monitoring and evaluating ecological restoration projects according to claim 6, characterized in that, The water conservation capacity is determined according to the following formula: ， In the formula, Water conservation capacity; i For the first i Ecosystem type; n The total number of ecosystem types; For the first i The area of ​​ecosystem-like structures; For runoff and rainfall; Surface runoff; This refers to the evaporation rate.

9. The method for monitoring and evaluating ecological restoration projects according to claim 6, characterized in that, The structure of the ecosystem is determined according to the following formula: ， in, , ， In the formula, For ecosystem structure; For ecological land diversity index; The proportion of ecological land area; For the first i The proportion of ecological land landscape area to total ecological land area; m is the number of ecological land landscape types; For the first in the statistics window i line, number j The value of a column cell is 1 when the cell is ecological land, and 0 otherwise; m and n are the total number of rows and columns of the statistics window, respectively.

10. An ecological restoration engineering monitoring and evaluation system, used to implement the method described in any one of claims 1-9, characterized in that, include: The data acquisition and processing module is used to acquire historical multi-source data information of the construction site and project construction process data. The historical multi-source data information includes geographic data of engineering documents, geographic data of engineering design, land change survey data, ground survey data and original satellite imagery covering the project area. The project construction process data includes at least multi-temporal satellite remote sensing imagery and UAV aerial photography data before and after construction. The mid-term engineering monitoring module is used to orthorectify and fuse raw satellite images covering the project area using digital elevation models and digital orthophoto data, and perform uniform light and color processing to obtain digital orthophoto images for the monitoring period. Based on the engineering design geographic data, monitoring patches are collected on the digital orthophoto images for the monitoring period, and land cover is classified. By combining the spatiotemporal attention change detection network with multi-temporal satellite remote sensing images before and during construction, the vegetation restoration area and soil disturbance range are automatically identified to determine the monitoring quantity and produce project engineering monitoring maps, generating mid-term monitoring data. The project completion monitoring module is used to generate digital orthophotos, digital surface models, and real-scene 3D models based on UAV aerial photography data before and after construction, perform land cover classification, detect changes in ground features before and after construction by combining a spatiotemporal attention change detection network, and generate project completion monitoring data through image interpretation, real-scene 3D model comparison, and cut-and-fill analysis. The effectiveness evaluation module is used to construct a multi-dimensional evaluation index system based on mid-term monitoring data, completion monitoring data, geographical data of engineering documents, ground survey data, and land change survey data. This system includes project management indicators, project output indicators, and project effect indicators. The weights of each indicator are assigned using the analytic hierarchy process (AHP) to achieve a comprehensive evaluation of the effectiveness of ecological restoration projects.

Images