Landscape ecological security pattern optimization method and system based on scale hierarchical reverse deduction method

Abstract

The application discloses a landscape ecological security pattern optimization method and system based on a scale hierarchical reverse deduction method, and belongs to the technical field of landscape ecological planning and ecological network construction. The method comprises the following steps: delimiting an ecological buffer zone boundary by coupling an InVEST model, and generating an ecological landscape component structure of multiple spatial scales; constructing an integrality and connectivity evaluation index system, calculating comprehensive scores of each scale by principal component analysis, calculating a second derivative of a curve after three times spline interpolation of the scores, and taking a scale corresponding to a maximum value of the derivative as an optimal ecological source reference scale; reversely mapping and identifying an ecological source by using the optimal scale component; constructing a comprehensive ecological resistance surface by fusing explicit ecological resistance and implicit ecological resistance, and generating an ecological corridor network based on a minimum cumulative resistance model; and identifying spatial positions of ecological nodes, determining node construction scales by reverse deduction of secondary scales, generating a planning construction boundary, and determining a construction type. The application realizes objective and quantitative construction of an ecological network in a whole process.

Description

Technical Field

[0001] This invention relates to the field of landscape ecological planning and ecological network construction technology, and in particular to a method and system for optimizing landscape ecological security patterns based on scale-level back deduction. Background Technology

[0002] Optimizing the landscape ecological security pattern is a core technical means to coordinate urban development and ecological protection, and enhance the connectivity of regional ecosystems. Current mainstream technologies follow a basic paradigm of "source area identification—resistance surface construction—corridor generation—node location," with the combination of scale-based inversion and the minimum cumulative resistance (MCR) model being the most widely used. However, existing technologies have the following significant drawbacks: (1) The determination of the optimal scale relies on subjective experience and lacks quantitative standards. Existing judgment methods mostly use visual interpretation or simple inflection point identification when determining the optimal scale, and the judgment results of different researchers vary significantly. Although some studies have attempted to combine principal component analysis for comprehensive evaluation, the selection of the optimal scale is still at the level of qualitative description of "inflection point of the scoring curve" and lacks reproducible quantitative judgment rules.

[0003] (2) Ecological nodes can only identify locations, but cannot quantify construction scale and spatial form. Although existing studies have mentioned the discussion of the scale and shape of ecological nodes, a systematic method for deducing node morphology has not been formed. Most studies can only determine the spatial location of ecological nodes, but cannot quantify the construction area, spatial shape and suitable construction type of nodes, making it difficult to directly implement the planning results.

[0004] (3) Lack of specific technical constraints for urban ecological buffer zones. The general-scale back-inference method does not couple ecosystem service assessment to delineate the control boundaries of buffer zones, making it difficult to meet the precise implementation needs of special plans such as urban ecological buffer zones.

[0005] (4) The method for constructing resistance surfaces needs improvement. Existing studies have proposed a method to distinguish between explicit and implicit ecological resistance characteristics and to construct a comprehensive ecological resistance surface by combining Kriging interpolation and spatial principal component analysis. However, the ecological resistance assignment parameters vary significantly in different regions, and there is a lack of adaptation schemes for regional ecosystem characteristics. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of existing general-scale back-inference methods, such as lack of buffer zone constraints, subjective scale determination, inability to quantify node morphology, and poor adaptability of resistance parameters. It provides a landscape ecological security pattern optimization method and system based on scale-level back-inference, which is coupled with ecosystem service assessment, has buffer zone-specific constraints, and is quantitatively objective throughout the entire process. By introducing second-order derivatives to quantitatively determine the optimal scale, establishing a complete inference process for secondary-scale node morphology, and coupling the InVEST model to delineate buffer zone boundaries, the invention achieves the objective and quantitative construction of the ecological network from the source area to the nodes.

[0007] To achieve the above objectives, this invention provides a method for optimizing landscape ecological security patterns based on scale-level back-inference, comprising the following steps: Step S1: Delineate the boundaries of the ecological buffer zone in the study area based on the ecosystem service assessment, extract land use raster data, classify ecological landscape types according to the ecological service value per unit area, and generate a multi-level spatial scale ecological landscape component structure. Step S2: Construct an evaluation system that includes two types of indicators: integrity and connectivity. Standardize and positively process all indicators. Calculate the comprehensive score for each scale through principal component analysis. Fit the score curve with cubic spline interpolation and then calculate the second derivative. Use the scale corresponding to the maximum value of the second derivative as the optimal reference scale for the ecological source area. Step S3: Map the boundaries of the ecological landscape components at the optimal scale back to the original land use data, identify natural ecological patches and patches to be ecologically transformed, integrate them to form a complete ecological source area and determine the minimum threshold for the construction scale of the regional ecological source area. Step S4: Construct a comprehensive ecological resistance surface that integrates explicit and implicit ecological resistance; Step S5: Based on the ecological source area, and combining the comprehensive ecological resistance surface and the minimum cumulative resistance model, generate an ecological corridor network; Step S6: Identify the spatial location of ecological nodes in the ecological corridor network, determine the construction scale and spatial form of ecological nodes through reverse extrapolation at the secondary scale, generate the planning and construction boundary of nodes, and determine the construction type.

[0008] Preferably, step S1 specifically includes: S11. Input high-resolution remote sensing interpretation of land use data of the study area, run the InVEST model to quantify the regional ecosystem service function, use the natural split point method to classify the ecological service value level per unit area, and classify forest land, shrubland, grassland and water area with value higher than the classification threshold as ecological landscape type. S12. Generate continuous multi-level scale grid units from 50m to 1200m with a preset gradient step size. The scale is pushed up using a unified majority judgment rule: when the ecological landscape area ratio in a single-scale grid unit is greater than 50%, it is marked as an ecological scale unit. The set of connected ecological scale units constitutes an ecological landscape component. S13. Based on the ecosystem service value classification layer output by the InVEST model, two buffer zone control boundaries are delineated: a 200-500m core isolation zone around the city and a 500-1000m outer conservation zone. The width of the buffer zone is determined based on the average dispersal distance of the main indicator species in the study area, serving as a spatial constraint for the subsequent ecological network layout.

[0009] Preferably, in step S2, the evaluation system and the optimal scale determination rule are as follows: S21. The overall indicators include cohesion, patch cohesion index, adjacency ratio, average proximity index, and landscape separation; the connectivity indicators include landscape component number, maximum component number, connectivity index, and percentage increase in connectivity; the three negative indicators, average proximity index, landscape separation, and landscape component number, are converted to positive values ​​by taking their negative values. S22. Perform Z-score standardization on all indicators at all scales to eliminate dimensional differences. The principal component extraction criteria are: eigenvalue > 1 and cumulative variance contribution rate ≥ 85%. S23. Construct a comprehensive connectivity scoring formula: In the formula For comprehensive scoring, Principal component scores, The weights of the variance contribution rates of each principal component to the total cumulative contribution rate are determined; the second derivative is calculated after fitting a smooth curve using cubic spline interpolation for discrete scale scoring points, and the scale corresponding to the maximum value of the second derivative is taken as the optimal reference scale for the ecological source area.

[0010] Preferably, in step S3, the reverse deduction process for the ecological source area is as follows: S31. Convert the optimal scale ecological landscape component raster into a vector boundary and spatially overlay it with the original high-precision land use raster layer. S32. The original ecological landscape patches inside the vector boundary constitute the main body of the ecological source area. The cultivated land and scattered construction land inside the boundary are designated as patches to be ecologically transformed. The spatial merging of the two types of patches forms the complete ecological source area. S33. Traverse all vector ecological source area patches, extract the minimum value as the minimum scale threshold for regional ecological source area construction, and simultaneously calculate the total area and spatial distribution of all patches to be ecologically transformed.

[0011] Preferably, in step S4, the comprehensive ecological resistance surface construction step is as follows: S41. Based on the characteristics of the ecosystem in the study area, assign ecological resistance coefficients to each landscape type and generate explicit ecological resistance grids. S42. Extract the geometric centroids of all land use patches as interpolation sampling points, use ordinary Kriging interpolation and a spherical model for the semi-variogram model to simulate the spatial diffusion effect of ecological resistance and generate a hidden ecological resistance surface. S43. Combine the explicit ecological resistance grid with the implicit ecological resistance grid, perform spatial principal component analysis, and extract the first principal component grid as the comprehensive ecological resistance surface. The cumulative variance contribution rate of the first principal component should be ≥98%.

[0012] Preferably, in step S5, the rules for constructing the ecological corridor network are as follows: S51. The source dataset is calculated using all vector patches of the ecological source areas as the minimum cumulative resistance. The ecological resistance coefficient of the area inside the ecological source area is assigned to 0. The boundary of the ecological source area is used as the start and end boundary of the corridor path. S52. Calculate the cumulative resistance value of the entire grid using the minimum cumulative resistance model. The model calculation formula is as follows: ; In the formula: For arbitrary grid cells To the nearest ecological source The minimum cumulative resistance value, This is the function for calculating minimum cumulative resistance. For grid cells To the ecological source area The spatial straight-line distance of the boundary For grid cells The ecological resistance coefficient within the ecological source area The value is 0. This represents the total number of grid cells. This represents the total number of ecological source areas; S53. Extract the path with the minimum cumulative resistance between any two ecological source areas, and the set of all paths constitutes an ecological corridor network; according to the type of substrate traversed by the corridor, the ecological corridors are divided into forest corridors, cultivated land corridors, water corridors, construction land corridors and unused land corridors, and the minimum construction width of the corridors is uniformly set at 100m.

[0013] Preferably, in step S6, the method for identifying ecological node locations is as follows: S61, Vulnerable nodes of corridor-ridgeline: The integrated ecological resistance grid is equivalent to a digital elevation model. The resistance ridgeline is extracted using GIS hydrological analysis tools. The intersection of the resistance ridgeline and the ecological corridor vector line is designated as the vulnerable node of corridor-ridgeline. S62. Corridor Intersection Key Nodes: Spatial points where the vector line segments of two or more ecological corridors intersect are designated as corridor-corridor key nodes.

[0014] Preferably, in step S6, the quantitative deduction step of the spatial morphology of ecological nodes is as follows: S63. Using a fixed decreasing step size of 50-100m, generate at least 3 levels of secondary scale unit grids within the range smaller than the optimal source area scale; reuse the principal component comprehensive scoring method in step S2, and select the secondary scale with the highest comprehensive score as the optimal reference scale for ecological nodes. S64. The minimum construction area of ​​the ecological node is taken as 1 / 2 of the area of ​​a single grid at the optimal reference scale of the ecological node. The minimum search radius of the ecological node is calculated based on the formula for calculating the area of ​​a circle. S65. Construct a linear network dataset using all ecological corridors. Using each ecological node as the center and the minimum search radius as the threshold, perform network service area analysis using network distance as the impedance parameter to generate irregular polygons that follow the corridor direction, which serve as the planning and construction boundaries for ecological nodes.

[0015] Preferably, in step S6, the criteria for determining the type of ecological node construction are as follows: Ecological nodes that simultaneously meet the following three conditions are designated as aquatic ecological nodes, and all others are uniformly designated as forest ecological nodes: ① The ecological node center point falls within an area ≥ 50 hm² 2 Within the complete catchment basin vector range; ②The Strahler classification of the water system where the ecological node is located is ≥3; ③ The ecological node is located at the outlet vector point of the corresponding water catchment basin.

[0016] A landscape ecological security pattern optimization system based on scale-level back-inference method includes: The multi-scale component generation module takes ecological buffer zone boundaries and land use raster as input and outputs multi-level spatial scale ecological landscape component raster. The optimal scale evaluation module takes into account the calculation results of multi-scale component raster and landscape index, and outputs the optimal source area reference scale and the secondary optimal scale of ecological nodes through cubic spline interpolation and second derivative calculation. The ecological source area inversion module takes the optimal scale vector boundary and the original land use data as input and outputs the ecological source area vector layer and the minimum construction scale threshold. The resistance surface construction module takes as input the centroid of land use patches and the ecological resistance assignment parameters, and outputs a comprehensive ecological resistance raster. The corridor network generation module takes ecological source geological cores and comprehensive resistance grids as inputs and outputs classified ecological corridor vector networks. The ecological node simulation module takes as input corridor network, integrated resistance grid and regional hydrological vector data, and outputs ecological node locations, planning and construction boundaries and node classification results.

[0017] Therefore, the landscape ecological security pattern optimization method and system based on scale-level back-inference method described above have the following advantages: (1) Objective quantification of optimal scale determination: For the first time, cubic spline interpolation and second derivative are introduced into the optimal scale determination. The scale corresponding to the maximum value of the second derivative is used as the reference scale of the optimal ecological source area. This overcomes the subjective defects of traditional methods that rely on visual interpretation or simple inflection point identification. The determination results are reproducible. (2) Complete quantitative deduction of ecological node morphology: For the first time, a complete node morphology deduction chain was established, which includes "secondary scale generation → principal component evaluation → grid area 1 / 2 to determine scale → network service area analysis to determine shape". The output results include node location, construction area, polygon boundary and construction type, which can directly support the implementation of the project. (3) Buffer zone constraints coupled with ecosystem service assessment: The value of ecosystem services is quantified by the InVEST model and directly used to delineate the two-layer buffer zone control boundary between the core isolation zone and the outer conservation zone around the city, so that the layout of the ecological network has a clear ecological basis; (4) Accurate construction of comprehensive ecological resistance surface: The implicit ecological resistance surface is constructed by using ordinary Kriging interpolation (spherical model). The explicit and implicit resistance surfaces are objectively integrated through spatial principal component analysis, which is closer to the resistance distribution of real ecological processes.

[0018] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0019] Figure 1 This is an overall flowchart of the landscape ecological security pattern optimization method based on scale-level back-inference method in an embodiment of the present invention; Figure 2 This is a module structure diagram of the landscape ecological security pattern optimization system based on the scale-level back deduction method in an embodiment of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0021] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0022] Example like Figures 1-2 As shown in the figure, this embodiment proposes a landscape ecological security pattern optimization method based on scale-level back-inference method. The specific implementation steps are as follows: Step S1: Based on the ecosystem service assessment, delineate the boundaries of the ecological buffer zone in the study area and generate a multi-level spatial scale structure of ecological landscape components: This embodiment studies the central urban area of ​​Lhasa and its surrounding ecological buffer zone, covering a total area of ​​approximately 600 km². 2 The original land use data has a resolution of 5m and is divided into 7 categories: forest land, shrubland, grassland, water area, cultivated land, construction land, and unused land.

[0023] S11. Delineation of Ecological Landscape Types: Input high-resolution remote sensing interpreted land use data of the study area, run the InVEST model to quantify the regional ecosystem service functions, and use the natural split point method to divide the ecological service value per unit area into 5 levels. Forest land, shrubland, grassland and water area of ​​level 4 and above are designated as ecological landscape types.

[0024] S12. Generation of Multi-Scale Components: Using a preset gradient step size (50m as the fixed gradient step size in this embodiment), nine scale grid units of 50m, 100m, 200m, 400m, 500m, 600m, 800m, 1000m, and 1200m are generated. A unified majority rule is used for scale advancement: a single-scale grid unit is marked as an ecological scale unit when the ecological landscape area accounts for >50%, and connected ecological scale units constitute an ecological landscape component. The preset gradient step size is mainly determined based on the research scale of the ecosystem in the study area.

[0025] S13. Delineation of Buffer Zone Control Boundaries: Based on the ecosystem service value grading layer output by the InVEST model, two layers of buffer zone control boundaries are delineated: a 200–500m ring-city core isolation zone and a 500–1000m outer conservation zone. In this embodiment, the width of the buffer zone is determined based on the average dispersal distance (200–500m) of major indicator species in Lhasa (such as the black-necked crane and Tibetan fox), serving as a spatial constraint for subsequent ecological network layout. In specific operations, the ring-city core isolation zone is used as the spatial priority area for ecological source area selection, and the outer conservation zone is used as the extension area for ecological corridor layout.

[0026] Step S2: Construct an evaluation system and determine the optimal reference scale for the ecological source area through cubic spline interpolation and second derivative: S21. Construction of the Evaluation Index System: Nine indicators were selected from two dimensions: integrity and connectivity. Indicators for integrity include Aggregation (AI), Cohesion Index (COHESION), Adjacency Ratio (IJI), Average Proximity Index (MPI), and Division. Indicators for connectivity include Landscape Component Number (NP), Maximum Component Patch Number (MPS), Connectivity Index (CONNECT), and Percentage Increase in Connectivity (PC). The three negative indicators—MPI, Division, and NP—were converted to positive values.

[0027] S22. Principal Component Analysis: Z-score standardization was performed on the index data at 9 scales to eliminate dimensional differences. The criteria for principal component extraction were eigenvalues ​​> 1 and cumulative variance contribution rate ≥ 85%. In this example, 3 principal components were extracted, with a cumulative variance contribution rate of 90.01%.

[0028] S23. Comprehensive Scoring and Second Derivative Determination: Constructing a Comprehensive Connectivity Scoring Formula: ; In the formula: For comprehensive scoring, Principal component scores, The weight of each principal component's variance contribution rate relative to the total cumulative contribution rate.

[0029] For discrete-scale scoring points (50m, 100m, 200m, 400m, 500m, 600m, 800m, 1000m, and 1200m corresponding to 9 scoring values), cubic spline interpolation is used to fit a smooth curve, and the second derivative is calculated. The scale corresponding to the maximum value of the second derivative is the optimal reference scale for the ecological source area. In this embodiment, the 500m scale corresponds to the maximum value of the second derivative, and 500m is determined as the optimal reference scale for the ecological source area. The specific implementation of cubic spline interpolation is as follows: the scale value is used as the independent variable. The overall score is the dependent variable. Construct cubic polynomials between adjacent scale nodes The natural boundary conditions of continuous function values, continuous first derivative, and continuous second derivative at the nodes are satisfied. Then, the second derivative is calculated for the smoothed curve after interpolation. ,Pick The maximum value corresponding to value.

[0030] Step S3: Optimal scale reverse mapping to identify ecological source areas and construction scale thresholds: S31. Vector Boundary Conversion and Overlay: Convert the 500m scale ecological landscape component raster into a vector boundary and spatially overlay it with the original 5m resolution land use raster layer.

[0031] S32. Ecological Source Area Integration: The original ecological landscape patches (woodland, shrubland, grassland, and water area) within the vector boundary constitute the main body of the ecological source area; cultivated land and scattered construction land within the boundary are designated as patches to be ecologically transformed. The spatial merging of the two types of patches forms the complete ecological source area.

[0032] S33. Determining the scale threshold: Traverse the area of ​​all vector ecological source patches, extract the minimum value as the minimum scale threshold for the construction of regional ecological source areas, and simultaneously calculate the total area and spatial distribution of all patches to be converted into ecological areas.

[0033] This embodiment identified a total of 18 ecological source areas, including 8 primary ecological source areas and 10 secondary ecological source areas; the smallest ecological source area is 38.5 hm². 2 This serves as the minimum scale threshold for the construction of ecological source areas in Lhasa; 11 non-ecological patches requiring conversion were identified, with a total area of ​​20.26 hm². 2 They are mainly distributed in areas with weak connections between ecological source areas.

[0034] Step S4: Construct a comprehensive ecological resistance surface: S41. Visible Ecological Resistance Surface: Based on the ecosystem characteristics of the study area and the service value per unit area of ​​the plateau ecosystem, ecological resistance coefficients are assigned to different landscape types. The resistance value ranges from 1 to 100, with higher values ​​indicating higher resistance: water bodies have a resistance value of 1 (minimum), forests 15, shrublands 20, grasslands 30, cultivated land 60, unused land 80, and construction land 100 (maximum resistance). A visible ecological resistance grid is generated based on the assigned values.

[0035] S42. Latent Ecological Resistance Surface: The geometric centroids of all land use patches are extracted as interpolation sampling points. Ordinary Kriging interpolation and a spherical model are used for the semi-variogram model to simulate the spatial diffusion effect of ecological resistance, generating a latent ecological resistance surface. The resistance values ​​of the latent ecological resistance surface have a smoother transition and can reflect potential ecologically vulnerable areas that are ignored by the explicit resistance surface.

[0036] S43. Ecological Resistance Surface Fusion: The explicit and implicit ecological resistance grids are combined by band analysis, and spatial principal component analysis is performed to extract the first principal component grid as the comprehensive ecological resistance surface. In this embodiment, the cumulative variance contribution rate of the first principal component is 98.12%, which meets the requirement of ≥98%.

[0037] Step S5: MCR model generates ecological corridor network: S51. The source dataset is calculated using all vector patches of the ecological source areas as the minimum cumulative resistance. The ecological resistance coefficient of the area inside the ecological source area is assigned to 0. The boundary of the ecological source area is used as the start and end boundary of the corridor path. S52. Calculate the cumulative resistance value of the entire grid using the minimum cumulative resistance model. The model calculation formula is as follows: ; In the formula: For arbitrary grid cells To the nearest ecological source The minimum cumulative resistance value, This is the function for calculating minimum cumulative resistance. For grid cells To the ecological source area The spatial straight-line distance of the boundary For grid cells The ecological resistance coefficient within the ecological source area The value is 0. This represents the total number of grid cells. This represents the total number of ecological source areas; S53. Extract the path with the minimum cumulative resistance between any two ecological source areas, and the set of all paths constitutes an ecological corridor network; according to the type of substrate traversed by the corridor, the ecological corridors are divided into forest corridors, cultivated land corridors, water corridors, construction land corridors and unused land corridors, and the minimum construction width of the corridors is uniformly set at 100m.

[0038] This embodiment generated a total of 17 ecological corridors, with a total length of approximately 68 km. Based on the type of landscape substrate traversed by the corridors, they were divided into three categories: cultivated land corridors, shrub and grass corridors, and construction land corridors, with cultivated land corridors accounting for 82.35%.

[0039] Step S6: Identification of ecological node locations, quantitative spatial morphology deduction, and determination of construction type: S61-S62, Ecological Node Location Identification: Vulnerable nodes along corridor-ridgelines: The integrated ecological resistance grid is equivalent to a digital elevation model. GIS hydrological analysis tools are used to extract resistance ridgelines (i.e., the paths with the greatest cumulative resistance). The intersection points of these resistance ridgelines and the ecological corridor vector lines are designated as vulnerable nodes along the corridor-ridgelines. These nodes are ecologically vulnerable points on the ecological corridor and require priority development.

[0040] Corridor intersection key nodes: Spatial points where the vector lines of two or more ecological corridors intersect are designated as corridor-to-corridor key nodes. These nodes are critical areas for the exchange of matter and energy in ecological flows.

[0041] This embodiment identifies a total of 11 ecological nodes, including 8 vulnerable nodes along the corridor-ridgeline and 3 key nodes at corridor intersections.

[0042] S63-S65, Quantitative extrapolation of the spatial morphology of ecological nodes: S63. Determination of the optimal reference scale for secondary scales: Using a fixed decreasing step size of 100m, generate three-level secondary scale unit grids of 400m, 300m, and 200m within an interval less than 500m (optimal source site scale). Reusing the principal component analysis scoring method from step S2, calculate the comprehensive connectivity score for each of the three secondary scales, and select the secondary scale with the highest comprehensive score as the optimal reference scale for the ecological node. In this embodiment, the 400m scale has the highest comprehensive score and is therefore determined as the optimal reference scale for the ecological node.

[0043] S64. Minimum Construction Scale and Search Radius Calculation: Based on the optimal reference scale of ecological nodes (400m), the area of ​​a single grid cell is (400m × 400m = 160,000m²). 2 =16hm 2 Half of the area is designated as the minimum construction area for ecological nodes, i.e., 8 hm². 2 Based on the formula for the area of ​​a circle The minimum search radius of the ecological node is calculated. m.

[0044] S65. Planning and Construction Boundary Generation: Construct a linear network dataset with all ecological corridors, use each ecological node as the center and a minimum search radius of 160m as the threshold, and use network distance as the impedance parameter to perform network service area analysis to generate irregular polygons that follow the corridor direction, which serve as the planning and construction boundaries of the ecological nodes.

[0045] In this embodiment, 11 ecological nodes correspond to 11 construction areas, with each node covering an area of ​​approximately 4.5 to 18 hectares. 2 The total area is approximately 76 hectares. 2 .

[0046] Determination of Ecological Node Construction Type: Based on hydrological analysis data of the study area, ecological nodes that simultaneously meet the following three conditions are designated as aquatic ecological nodes, and all others are uniformly designated as forest ecological nodes: ① The ecological node center point falls within an area ≥ 50 hm² 2 Within the complete catchment basin vector range; ②The Strahler classification of the water system where the ecological node is located is ≥3; ③ The ecological node is located at the outlet vector point of the corresponding water catchment basin.

[0047] In this embodiment, two ecological nodes meet the conditions for water-type construction (planned to be constructed in the form of artificial wetlands and ponds), while the remaining nine are forest-type ecological nodes (in the plateau scene, shrubs are the main vegetation).

[0048] like Figure 2 As shown, this embodiment provides a landscape ecological security pattern optimization system based on scale-level backward inference and node morphology quantitative inference to implement the above method, including: Multi-scale component generation module: Input 5m land use raster and InVEST buffer zone boundary, output nine-level scale ecological landscape component raster. This module is responsible for performing all the functions of step S1.

[0049] Optimal Scale Evaluation Module: Input the nine-level scale raster and landscape index calculation results, and output the optimal source area reference scale of 500m and the secondary optimal scale of ecological nodes of 400m through cubic spline interpolation and second derivative calculation. This module is responsible for performing all the functions of step S2.

[0050] Ecological source area inversion module: Input 500m vector boundary and original land use data, output 18 ecological source area vector layers, 38.5hm 2 Minimum construction threshold. This module is responsible for performing all the functions of step S3.

[0051] Resistance surface construction module: Input the centroid of land use patches and the assigned parameters of plateau ecological resistance, use ordinary kriging interpolation (spherical model) and spatial principal component analysis to output a comprehensive ecological resistance raster. This module is responsible for performing all the functions of step S4.

[0052] Corridor Network Generation Module: Input ecological source geological cores and comprehensive resistance grids, output 17 classified ecological corridor vector networks. This module is responsible for performing all the functions of step S5.

[0053] Ecological Node Projection Module: Input corridor network, integrated resistance grid, and regional hydrological vector data; output 11 ecological node locations, polygonal planning and construction boundaries, and water / forest land classification results. This module is responsible for performing all the functions of step S6.

[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for optimizing landscape ecological security patterns based on scale-level back-inference, characterized in that, Includes the following steps: Step S1: Delineate the boundaries of the ecological buffer zone in the study area based on the ecosystem service assessment, extract land use raster data, classify ecological landscape types according to the ecological service value per unit area, and generate a multi-level spatial scale ecological landscape component structure. Step S2: Construct an evaluation system that includes two types of indicators: integrity and connectivity. Standardize and positively process all indicators. Calculate the comprehensive score for each scale through principal component analysis. Fit the score curve with cubic spline interpolation and then calculate the second derivative. Use the scale corresponding to the maximum value of the second derivative as the optimal reference scale for the ecological source area. Step S3: Map the boundaries of the ecological landscape components at the optimal scale back to the original land use data, identify natural ecological patches and patches to be ecologically transformed, integrate them to form a complete ecological source area and determine the minimum threshold for the construction scale of the regional ecological source area. Step S4: Construct a comprehensive ecological resistance surface that integrates explicit and implicit ecological resistance; Step S5: Based on the ecological source area, and combining the comprehensive ecological resistance surface and the minimum cumulative resistance model, generate an ecological corridor network; Step S6: Identify the spatial location of ecological nodes in the ecological corridor network, determine the construction scale and spatial form of ecological nodes through reverse extrapolation at the secondary scale, generate the planning and construction boundary of nodes, and determine the construction type.

2. The landscape ecological security pattern optimization method based on scale-level back-inference method according to claim 1, characterized in that: Step S1 specifically includes: S11. Input high-resolution remote sensing interpretation of land use data of the study area, run the InVEST model to quantify the regional ecosystem service function, use the natural split point method to classify the ecological service value level per unit area, and classify forest land, shrubland, grassland and water area with value higher than the classification threshold as ecological landscape type. S12. Generate continuous multi-level scale grid units from 50m to 1200m with a preset gradient step size. The scale is pushed up using a unified majority judgment rule: when the ecological landscape area ratio in a single-scale grid unit is greater than 50%, it is marked as an ecological scale unit. The set of connected ecological scale units constitutes an ecological landscape component. S13. Based on the ecosystem service value classification layer output by the InVEST model, two buffer zone control boundaries are delineated: a 200-500m core isolation zone around the city and a 500-1000m outer conservation zone. The width of the buffer zone is determined based on the average dispersal distance of the main indicator species in the study area, serving as a spatial constraint for the subsequent ecological network layout.

3. The method for optimizing landscape ecological security patterns based on scale-level back-inference as described in claim 1, characterized in that: In step S2, the evaluation system and the optimal scale determination rule are as follows: S21. The overall indicators include cohesion, patch cohesion index, adjacency ratio, average proximity index, and landscape separation; the connectivity indicators include landscape component number, maximum component number, connectivity index, and percentage increase in connectivity; the three negative indicators, average proximity index, landscape separation, and landscape component number, are converted to positive values ​​by taking their negative values. S22. Perform Z-score standardization on all indicators at all scales to eliminate dimensional differences. The principal component extraction criteria are: eigenvalue > 1 and cumulative variance contribution rate ≥ 85%. S23. Construct a comprehensive connectivity scoring formula: In the formula For comprehensive scoring, Principal component scores, The weights of each principal component's variance contribution rate relative to the total cumulative contribution rate; After fitting a smooth curve using cubic spline interpolation to the discrete scale scoring points, the second derivative is calculated, and the scale corresponding to the maximum value of the second derivative is taken as the optimal reference scale for the ecological source area.

4. The method for optimizing landscape ecological security patterns based on scale-level back-inference as described in claim 1, characterized in that: In step S3, the reverse deduction process for the ecological source area is as follows: S31. Convert the optimal scale ecological landscape component raster into a vector boundary and spatially overlay it with the original high-precision land use raster layer. S32. The original ecological landscape patches inside the vector boundary constitute the main body of the ecological source area. The cultivated land and scattered construction land inside the boundary are designated as patches to be ecologically transformed. The spatial merging of the two types of patches forms the complete ecological source area. S33. Traverse all vector ecological source area patches, extract the minimum value as the minimum scale threshold for regional ecological source area construction, and simultaneously calculate the total area and spatial distribution of all patches to be ecologically transformed.

5. The method for optimizing landscape ecological security patterns based on scale-level back-inference as described in claim 1, characterized in that: In step S4, the steps for constructing the integrated ecological resistance surface are as follows: S41. Based on the characteristics of the ecosystem in the study area, assign ecological resistance coefficients to each landscape type and generate explicit ecological resistance grids. S42. Extract the geometric centroids of all land use patches as interpolation sampling points, use ordinary Kriging interpolation and a spherical model for the semi-variogram model to simulate the spatial diffusion effect of ecological resistance and generate a hidden ecological resistance surface. S43. Combine the explicit ecological resistance grid with the implicit ecological resistance grid, perform spatial principal component analysis, and extract the first principal component grid as the comprehensive ecological resistance surface. The cumulative variance contribution rate of the first principal component should be ≥98%.

6. The method for optimizing landscape ecological security patterns based on scale-level back-inference as described in claim 1, characterized in that: In step S5, the rules for constructing the ecological corridor network are as follows: S51. The source dataset is calculated using all vector patches of the ecological source areas as the minimum cumulative resistance. The ecological resistance coefficient of the area inside the ecological source area is assigned to 0. The boundary of the ecological source area is used as the start and end boundary of the corridor path. S52. Calculate the cumulative resistance value of the entire grid using the minimum cumulative resistance model. The model calculation formula is as follows: ； In the formula: For arbitrary grid cells To the nearest ecological source The minimum cumulative resistance value, This is the function for calculating minimum cumulative resistance. For grid cells To the ecological source area The spatial straight-line distance of the boundary For grid cells The ecological resistance coefficient within the ecological source area The value is 0. This represents the total number of grid cells. This represents the total number of ecological source areas; S53. Extract the path with the minimum cumulative resistance between any two ecological source areas, and the set of all paths constitutes an ecological corridor network; according to the type of substrate traversed by the corridor, the ecological corridors are divided into forest corridors, cultivated land corridors, water corridors, construction land corridors and unused land corridors, and the minimum construction width of the corridors is uniformly set at 100m.

7. The method for optimizing landscape ecological security patterns based on scale-level back-inference as described in claim 1, characterized in that: In step S6, the method for identifying ecological node locations is as follows: S61, Vulnerable nodes of corridor-ridgeline: The integrated ecological resistance grid is equivalent to a digital elevation model. The resistance ridgeline is extracted using GIS hydrological analysis tools. The intersection of the resistance ridgeline and the ecological corridor vector line is designated as the vulnerable node of corridor-ridgeline. S62. Key nodes of corridor intersection: Spatial points where the vector lines of two or more ecological corridors intersect are designated as key nodes between corridors.

8. The method for optimizing landscape ecological security patterns based on scale-level back-inference as described in claim 1, characterized in that: In step S6, the quantitative deduction steps for the spatial morphology of ecological nodes are as follows: S63. Using a fixed decreasing step size of 50-100m, generate at least 3 levels of secondary scale unit grids within the range smaller than the optimal source area scale; reuse the principal component comprehensive scoring method in step S2, and select the secondary scale with the highest comprehensive score as the optimal reference scale for ecological nodes. S64. The minimum construction area of ​​the ecological node is taken as 1 / 2 of the area of ​​a single grid at the optimal reference scale of the ecological node. The minimum search radius of the ecological node is calculated based on the formula for calculating the area of ​​a circle. S65. Construct a linear network dataset using all ecological corridors. Using each ecological node as the center and the minimum search radius as the threshold, perform network service area analysis using network distance as the impedance parameter to generate irregular polygons that follow the corridor direction, which serve as the planning and construction boundaries for ecological nodes.

9. The method for optimizing landscape ecological security patterns based on scale-level back-inference as described in claim 1, characterized in that: In step S6, the criteria for determining the type of ecological node construction are as follows: Ecological nodes that simultaneously meet the following three conditions are designated as aquatic ecological nodes, and all others are uniformly designated as forest ecological nodes: ① The ecological node center point falls within an area ≥ 50 hm² 2 Within the complete catchment basin vector range; ②The Strahler classification of the water system where the ecological node is located is ≥3; ③ The ecological node is located at the outlet vector point of the corresponding water catchment basin.

10. A landscape ecological security pattern optimization system based on scale-level back-inference method, used to implement the landscape ecological security pattern optimization method based on scale-level back-inference method as described in any one of claims 1-9, characterized in that, include: The multi-scale component generation module takes ecological buffer zone boundaries and land use raster as input and outputs multi-level spatial scale ecological landscape component raster. The optimal scale evaluation module takes into account the calculation results of multi-scale component raster and landscape index, and outputs the optimal source area reference scale and the secondary optimal scale of ecological nodes through cubic spline interpolation and second derivative calculation. The ecological source area inversion module takes the optimal scale vector boundary and the original land use data as input and outputs the ecological source area vector layer and the minimum construction scale threshold. The resistance surface construction module takes as input the centroid of land use patches and the ecological resistance assignment parameters, and outputs a comprehensive ecological resistance raster. The corridor network generation module takes ecological source geological cores and comprehensive resistance grids as inputs and outputs classified ecological corridor vector networks. The ecological node simulation module takes as input corridor network, integrated resistance grid and regional hydrological vector data, and outputs ecological node locations, planning and construction boundaries and node classification results.

Images