A land space optimization method based on cross-scale planning objectives and ecological resilience feedback

Abstract

The present invention relates to a land space optimization method that integrates cross-scale planning objectives and ecological resilience feedback. First, based on the big data platform GEE, the random forest algorithm is used to perform long-term land space function classification and analyze driving factors. Planning texts from multiple administrative levels are collected, planning objectives are analyzed, and the target parameter group is obtained using the equal spacing method. Combined with the "double evaluation" results, a model is built to simulate the future land space structure. On this basis, the vegetation structure is analyzed, and the long-term resilience indicators of the global and local grids are simulated based on biophysical processes. Finally, the changing trends of all resilience indicators are analyzed. If there is a trend shift, the parameters are fed back and iterated until all resilience indicators remain stable or show a positive trend, the cycle is ended, and the output is output. This method quantifies for the first time the feedback relationship between planning objectives at different levels and ecosystem resilience, successfully achieving the effective integration of sustainability and resilience, and improving the scientific nature of land space structure optimization and spatial decision-making.

Description

Technical Field

[0001] The present invention relates to a land space optimization method based on cross-scale planning objectives and ecological resilience feedback under the support of geographic information big data, and in particular to a land space scenario optimization simulation method that measures changes in ecological resilience under global climate change and feeds back to planning objective adjustments. Background Art

[0002] The spatial structure of land and space is a complex structure resulting from the interaction between people and land. It refers to the spatial organization of the areas, proportions, and interactions between different land and space development and protection functions. This includes spatial agglomeration, spatial differentiation, topological structure, and local and remote interactions. It can be understood as a complex structure organically combining the structure of natural ecosystems and the spatial organization of human society. Against the backdrop of global climate change and the promotion of sustainable development, optimizing the sustainability and resilience of the integrated spatial structure of land and space is a hot topic in land and space planning and natural ecological management.

[0003] Optimizing national land spatial structure consists of three basic elements: an objective function, decision variables, and constraints. Previous approaches to optimizing national land spatial structure typically treated natural systems as static constraints, and only considered sustainability and efficiency in economic, social, or ecological dimensions at the objective level. For example, in 2023, Yang Xuedi et al. published an article titled "Optimizing Land Use Functional Layout in the Lanxi Urban Agglomeration with Ecological Security" in Volume 43, Issue 7 of the Acta Ecologica Sinica. The article simulated national land spatial functions and structures under different socioeconomic planning objectives to achieve the optimization goal of minimizing ecological land occupation. Zhai Chenyang et al. published an article titled "Evolution and Scenario Simulation of Multidimensional Well-being in the Poyang Lake Region Based on System Dynamics" in Volume 41, Issue 8 of the Acta Ecologica Sinica. The article simulated the evolution of national land spatial functions under different development scenarios to explore optimal scenarios that maximize human well-being in the study area. As climate change becomes increasingly recognized and attracts widespread attention, ecosystem resilience has been confirmed as a key factor in maintaining ecosystem sustainability. This has been incorporated into natural resource and ecosystem management and applied to national land spatial pattern optimization. Consequently, the objective function, constraints, and decision variables of national land spatial structure optimization have undergone fundamental changes. For example, previous objective functions have shifted from focusing on a single dimension of sustainability, such as economic efficiency, social welfare, or ecosystem service functions at a single scale, to now considering economic and social sustainability goals and ecosystem resilience at multiple scales simultaneously.

[0004] Substantial progress has been made in measuring ecosystem resilience. For example, by leveraging long-term remote sensing or monitoring data, the variance, skewness, and peak variation of data such as leaf area index and NPP within the time series can be used to measure the resilience of ecosystems such as forests, grasslands, and lakes. Recently, several methods and models have emerged that simulate the changes in ecosystems following disturbances from the perspective of disturbance response, measuring the overall resilience of ecosystems by measuring the domain of change in system state variables, the time it takes for the system to recover, and the residual impact on the system's steady state. However, no technical methods or models currently exist that incorporate ecosystem resilience into the optimization of national spatial structure, particularly for future predictions, by establishing a feedback relationship between the evolution of national spatial structure driven by planning objectives and ecosystem resilience. The key challenge lies in integrating socioeconomic sustainability goals with ecosystem resilience across scales—that is, maintaining global and local ecosystem resilience while achieving future regional socioeconomic sustainability goals.

[0005] In response to the above difficulties, the present invention proposes a land space optimization method based on cross-scale planning objectives and ecological resilience feedback, so as to maximize the realization of socio-economic planning objectives at multiple levels while maintaining global and local ecosystem resilience. This technical method establishes an adaptive feedback relationship between cross-scale planning objectives, land space structure evolution and ecosystem resilience, simulates the evolution of land space structure under different planning objectives, measures the feedback of future long-term simulation of land space structure on ecosystem resilience under the background of climate change, and achieves a trade-off between the sustainability and resilience of land space structure. This technology is suitable for land space structure optimization and land space planning, especially the trade-off and management of ecosystem resilience and socio-economic sustainability objectives at different administrative levels under climate change and human interference. Summary of the Invention

[0006] The technical problem to be solved by the present invention is: in view of the problem that previous land space structure optimization simulation methods cannot quantify the feedback of different planning objectives on ecosystem resilience under climate change, and cannot incorporate it into the future land space structure optimization, a land space optimization method with cross-scale planning objectives and ecological resilience feedback is proposed to simultaneously achieve the socio-economic sustainability goals of the future land space structure and the resilience of the ecosystem.

[0007] To solve the above technical problems, the present invention proposes a land space optimization method based on cross-scale planning objectives and ecological resilience feedback, which includes the following steps:

[0008] Step 1: Classify the land space functions within the set time range in a long time series and analyze and summarize the factors affecting the evolution of land space structure;

[0009] Step 2: Obtain the planning objectives of the study area at each administrative level and construct a planning objective database to determine the functional types of the study area and the implementation priority of the functional types. The planning objective database at least includes the planning objective value, the objective change rate, the current value, and the evolution trend;

[0010] Step 3: Set the initial planning target parameter group and optimization step size;

[0011] Based on the planning target database obtained in step 2, based on the current supply values ​​of the parameters of each functional type in the study area, according to the target change rates in the planning of different administrative levels, and according to the target change rates that evolve within a specific planning period, the potential target parameter values ​​that each functional type will reach at a specific time node under the planning requirements of the government at different administrative levels are calculated respectively. The potential target parameter values ​​of different administrative levels at the same time node are combined. For the same functional type, the potential target parameter values ​​of all administrative levels are taken as a union to determine the target parameter optimization interval of the functional type. According to the importance of the functional type in the study area, the optimization step size of the functional type is selected within the range of 2%-20% of the corresponding target parameter optimization interval. Based on the optimization step size, the target parameter optimization interval is divided into several endpoint values ​​using the equal interval segmentation method to obtain the endpoint values. The endpoints of different functional types are combined to obtain a target parameter group. In combination with the implementation priority of the functional types obtained in step 2, the target parameter groups are sorted from high to low based on the endpoint values.

[0012] Step 4: Simulation of the future evolution of national land spatial structure;

[0013] By setting the target parameter group that the future land space needs to meet, the existing constraints in the region, and the driving factors of the regional land space structure evolution, the modeling method is used to simulate the evolution process of the land space function and structure. The specific steps include:

[0014] 4.1. Target parameter setting: From all target parameter groups obtained in step 3, select the target parameter groups in order of ranking, take the target parameter group of the target year as the end point, and the data of the starting year corresponding to the target parameter group as the starting point, interpolate the target parameter values ​​for the specified time step, obtain the target parameter values ​​of each time node, and generate a time series of the target parameter group as the target parameter input model in the simulation process; at the same time, with the help of the current land space structure data and the demand satisfaction spatial pattern data, use the spatial analysis method to calculate the average value of each land space use type's ability to meet each target, and use it as the target parameter satisfaction ability input model;

[0015] 4.2. Constraint setting: Using the resource and environmental carrying capacity of national land space planning and the results of national land space development suitability assessment, two types of spatial constraints are set: an upper development limit and a lower protection limit. The upper development limit is obtained by removing the areas most suitable for agricultural production, animal husbandry production, and ecological protection from the areas with the highest suitability for urban development. The lower protection limit is obtained by taking the union of the areas with the highest suitability for agricultural production and the areas with the highest importance for ecological protection.

[0016] 4.3. Driving force factor setting: Using the ROC test, we analyzed the correlation between the historical evolution of the spatial structure of the study area and geographical elements. We selected geographical elements with an AUC greater than 0.85 as driving factors for the evolution of spatial function and structure. Using multiple linear regression, we combined the spatial structure of the starting year to analyze the quantitative relationship between the evolution of spatial function and structure and the driving factors.

[0017] 4.4. Complex system modeling and simulation: Using the driving force factors analyzed in step 4.3 and the quantitative relationship between the evolution of the land space functional structure and the driving force factors, based on the spatial atlas data of the driving force factors, the driving force weights of different driving factors are combined with the current land space functional structure and the land space use type's ability to meet the functional type, and the weighted summation is used to obtain the land space functional structure evolution probability atlas. The calculation formula is as follows:

[0018]

[0019] Among them, Prob a，t is the probability of evolution of land space functional structure of land space use type a at time node t, o i The implementation priority of function type i, Supply i，t Demand is the demand parameter that can satisfy function type i at time node t. i，t is the target parameter of function type i at time node t, w j is the quantitative weight of driving factor j, D j is the specific value of the driving force factor j, n k，t is the number of pixels of land space utilization type k at time node t, cap i，k is the supply capacity of a single land type k to functional type i, m, n and N are the number of functional types, the number of driving factors and the number of land space use types respectively;

[0020] In each grid unit of the land space functional structure, the land space utilization type with the highest evolution probability is selected as the evolution result; according to the land space structure of the current simulation results and the ability of land space utilization type to meet the functional type, the demand parameter Supply that the land space structure functional type can meet is calculated.i :

[0021] Supply i =∑n j *cap i，j

[0022] Among them, Supply i is the parameter that can meet the requirements of function type i, n j is the number of pixels of land class j, cap i，j is the supply capacity of a single land type j to functional type i, i = 1, 2, ..., N, where N represents the number of functional types; the demand parameter is compared with the target parameter of the current time node. When the ratio of the maximum demand difference of a single functional type to the demand parameter is less than 1%, and the ratio of the demand difference of all functional types to the sum of the demand parameters of all functional types is less than 0.5%, the national land space structure simulation of the current time node is completed and the national land space structure simulation of the next time node is carried out. Otherwise, the current step is repeated until the requirements of the current time node are met;

[0023] Step 5: Measure ecosystem resilience indicators under climate change. The specific method is as follows:

[0024] Based on the land space structure simulation results at each time node obtained in step 4, the area proportion of each land space use type at the grid scale is calculated. Combined with the vegetation structure proportion calculated by field survey, the long-term time series data of the vegetation type proportion of each grid unit is obtained. The calculation formula for the vegetation type proportion within the grid unit is as follows:

[0025]

[0026] Among them, PFT i，t is the proportion of vegetation type i in grid g at time node t, Cov j，g，t is the area proportion of land space use type j in grid g at time node t, Prop i，j is the proportion of vegetation type i in land use type j;

[0027] Based on the long-term climate change data of temperature, precipitation, sunshine days, evapotranspiration, and surface accumulated temperature, and the regional resource data of soil organic carbon, soil sandy loam, soil carbon and nitrogen content, soil pH, and soil plant cell deposition, the long-term vegetation type proportion data of all grid cells in the study area were used as the ecosystem vegetation structure, and the long-term climate change data and regional resource data were used as the nutrients for biological competition. The biophysical process model LPJ-GUESS (Lund-Potsdam-Jena General Ecosystem Simulator), simulate the evolution of ecosystem status during the spatial evolution of the land in the study area, generate the long-term evolution process of the future ecosystem status and vegetation structure, obtain the long-term ecosystem parameter results of NPP, LAI, river runoff, ecosystem carbon flux, vegetation carbon sink, vegetation biomass, and vegetation organic carbon at the grid scale as local-scale resilience indicators, and summarize all long-term ecosystem parameter results to the entire study area as a global-scale resilience indicator;

[0028] Step 6: Determine the conditions for triggering the loop feedback;

[0029] Based on the long-term ecological parameter results on the global and local grids in step 5, the sustainability and resilience of the system are determined according to the following steps:

[0030] 6.1. Global and local scale ecological status assessment: Use the following algorithm to determine the evolution trend of ecological status parameter results at the global scale and at the local scale:

[0031] ①Set the minimum trend length to t limit , select a moving window with a time length of , which moves backward from the starting point of the time series. The value range of the time length , is 20%-40% of the time span from the starting year to the target year;

[0032] ② Take the maximum and minimum values ​​of the ecological state parameter results in the moving window a i with a j , the corresponding time nodes are i and j respectively; if ij≥t limit , it is considered that in the time series T = {j, j+1, ..., i}, the time series has an upward trend; if ji≥t limit , it is considered that within the time series T = {i, i + 1, ..., j}, the time series has a downward trend; otherwise, there is no evolution trend within the moving window;

[0033] ③ Take the later of time nodes i and j as the starting point of the moving window and repeat step ② until the time series iteration is completed;

[0034] If any ecological state parameter time series data has both an upward and downward trend, select the next target parameter group and repeat step 4; if there are no two opposite trends, output the current result as the optimization result of national land space function and pattern;

[0035] Step 7: Summarize and visualize the optimization results.

[0036] This method utilizes big data processing to optimize the spatial structure of national territory over long time periods. By constructing a space-time cube and using segmentation techniques, combined with forest-based parameter simulation of different functional land uses and spatial simulation methods based on adaptive cellular automata, it can rapidly and accurately simulate the evolution of national territory's spatial functional structure over long time periods. This method is suitable for optimizing the spatial structure of regional functions over long time periods, high density, and large areas, and is particularly well-suited for simulating, analyzing, and sustainably regulating national territory's spatial functions and structures in various scenarios for future national territory spatial planning.

[0037] This paper utilizes a random forest algorithm, based on multi-source big data, to achieve long-term spatial functional classification and driver analysis. Based on multi-administrative planning objectives and through the coupling of the CLUMondo multivariate linear optimization model with the LPJ-GUESS biophysical process model, it effectively simulates the long-term functional structure evolution and ecosystem state changes of national land space under different planning objectives, accurately identifying potential ecosystem unsustainability and resilience risks under planning objectives. This method quantifies for the first time the feedback relationship between planning objectives at different levels and ecosystem resilience, successfully integrating regional-scale sustainability objectives and resilience indicators, and improving the scientific nature of national land space structure optimization and spatial decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] The present invention will be further described below with reference to the accompanying drawings.

[0039] Figure 1 This is an overview of the study area of ​​the embodiment of the present invention.

[0040] Figure 2 It is a technical flow chart of the method of the present invention.

[0041] Figure 3 This is a diagram of the land space structure simulation results of an embodiment of the present invention.

[0042] Figure 4 This is a simulation result diagram of the proportion of functional land in the national land space according to an embodiment of the present invention.

[0043] Figure 5 This is a diagram of the global system simulation results of ecosystem state parameters according to an embodiment of the present invention.

[0044] Figure 6 This is a diagram of the local system simulation results of ecosystem state parameters according to an embodiment of the present invention.

[0045] Figure 7 This is the final summary visualization result diagram of the embodiment of the present invention. DETAILED DESCRIPTION

[0046] The experimental data used in this experiment include: Landsat series images from 1987 to 2019, surface reflectance products, 30m resolution; VIIRS night light data (Nighttime Day / Night Band Composites Version 1), 500m resolution; ChinaClim long-term temperature and precipitation monitoring and forecast data, 1000m resolution; WorldPop long-term population data, 100m resolution; MODIS satellite NPP remote sensing inversion products, 500m resolution; digital elevation data (SRTM), 30m resolution; regional traffic network data for 2020; administrative division data of towns and administrative villages for 2020.

[0047] The study area is the Huangshui River Basin in Qinghai Province, located at 36°02′~37°28′N, 100°42′~103°04′E, with a drainage area of ​​approximately 16,200 km. 2 It is located in the transition zone between the Qinghai-Tibet Plateau and the Loess Plateau, and is an important tributary of the upper Yellow River and one of its main water sources (such as Figure 1 The Huangshui River Basin, comprising the Huangshui River mainstream and its tributary, the Beichuan River, encompasses three prefecture-level administrative regions: Xining City, Haidong City, and Haibei Prefecture. With an altitude ranging from 1,606 to 4,871 meters, with an average elevation of 2,300 meters, the basin receives an average annual precipitation of approximately 300 to 500 mm, suffers from frequent droughts, and is ecologically fragile and highly heterogeneous. Therefore, the Huangshui River Basin exhibits some characteristics of both a plateau basin and a semi-arid region, making it a typical ecologically fragile basin. Furthermore, the Huangshui River Basin boasts the highest population density and the most concentrated economy on the entire Qinghai-Tibet Plateau, occupying less than 1% of the plateau's area but home to approximately 25% of the total population and economy. Over the past 30 years, climate change and human expansion have significantly impacted the Huangshui River Basin, manifested in a slight increase in average annual precipitation and the continuous expansion of irrigation water diversion networks. Population and economic activity have further concentrated in the Huangshui River Valley and upstream, leading to a 3.6-fold increase in urban construction land, creating increasingly significant challenges for agricultural and animal husbandry production and ecological conservation.

[0048] This method classifies long-term imagery on the geographic information big data platform Google Earth Engine (GEE). JavaScript is used to implement spatiotemporal consistency correction, cube segmentation and infilling, and prediction of land use parameters for different functional areas based on integrated forests. Spatial simulation of land functions is achieved using the CLUMondo model, and ecosystem resilience measurement is implemented using the LPJ-GUESS model. All other operations are handled locally in Python, version 3.6.

[0049] The embodiment of the present invention provides a method for simulating the function and structure of land space in a long time series, comprising the following steps: Figure 2 As shown:

[0050] Step 1: Conduct a long-term classification of national land space functions over the past twenty years and analyze and summarize the factors influencing the evolution of national land space structure.

[0051] Leveraging the Google Earth Engine big data cloud platform, and utilizing Landsat long-term satellite imagery, population data, VIIRS nighttime light data, MODIS NPP inversion products, Landsat NDVI inversion products, and Landsat NDWI inversion products, a long-term classification of the Huangshui River Basin's territorial spatial functions was conducted from a regional functional perspective, ultimately yielding a territorial spatial functional classification from 1989 to 2020. Based on a regional functional etiology analysis, influencing factors were selected as independent variables from three dimensions: natural carrying capacity, socioeconomic spatial organization, and regional interaction. Using partial correlation analysis, a preliminary selection of 15 factors influencing the evolution of territorial spatial structure was conducted: altitude, slope, NPP, temperature, precipitation, soil sandiness, soil clayiness, soil organic carbon, soil moisture content, distance to irrigation canals, road accessibility, railway accessibility, urban accessibility, cost distance to Xining, and GDP.

[0052] Step 2: Obtain planning objectives for the study area at each administrative level and construct a planning objective database. Determine the functional types and implementation priorities for the study area. The planning objective database includes at least the planning objective value, target change rate, current value, and evolution trend.

[0053] In this step, we collected data on planning documents related to the Huangshui River Basin through methods such as collecting public documents and conducting interviews with government departments. Ultimately, we collected several planning documents with different focuses, including the basin water resources plan, the Lanxi urban agglomeration development plan, and provincial and municipal land and space plans. Using meta-analysis, map interpretation, and semantic extraction, we analyzed the sustainability positioning and planning goals of the Huangshui River Basin at different administrative levels: national, Qinghai Province, Xining City, Haidong City, and their subordinate districts and counties. This generated a planning target database, identifying a series of planning target ranges for the Huangshui River Basin by 2050, including a population urbanization rate of 72%-85% and a livestock carrying capacity growth rate of -0.55%-1.15%. We also compared the Huangshui River Basin's dominant national ecological protection function with its dominant urban development function at the prefecture and city levels, clarifying that the Huangshui River Basin's primary sustainable functions are population accommodation, food production, livestock carrying capacity, and ecological services. We prioritized ecological protection over agricultural development.

[0054] Step 3: Set the initial planning target parameter group and optimization step size.

[0055] Based on the planning target database obtained in step 2, based on the current supply values ​​of the functional type parameters of the Huangshui River Basin, and according to the target change rates in 18 planning documents at four levels, namely the Lanxi Urban Agglomeration, Qinghai Province, Xining City, Haidong City, and Haibei Prefecture, the target parameters for 2050 and 2100 were calculated according to the target change rates of short-term, medium-term, and long-term evolution, respectively. The target parameters of all administrative levels were combined to determine the optimized range of the target parameters for the functional types. Target parameter groups such as the basin population carrying capacity of 3.5403 million to 5.0810 million and grain production of 22.1307 million to 28.4962 million tons in 2050 were calculated. Based on the importance of the functional types in the study area, an optimization step size of 10% was selected. Based on the optimization step size, the target parameter optimization range was divided into several endpoint values ​​using the equal interval segmentation method. The endpoints of different functional types were combined to obtain target parameter groups. Combined with the implementation priority of the functional types obtained in step 2, the target parameter groups were sorted from high to low based on the endpoint values.

[0056] Step 4: Simulate the future evolution of the national land spatial structure.

[0057] By setting the target parameter group that the future land space needs to meet, the existing constraints in the region, and the driving factors of the regional land space structure evolution, the modeling method is used to simulate the evolution process of the land space function and structure. The specific steps include:

[0058] 4.1 Target parameter setting; Among all target parameter groups, the target parameter groups are selected in order according to the ranking, and the logarithmic function fitting method is used to interpolate the scenario parameter time series with a time step of 5 years from the starting year data of 2020, the target parameter groups of 2050 and 2100, which are used as the target parameter input model in the simulation process; at the same time, with the help of the 2020 land space structure data and the demand satisfaction spatial pattern data, we resample the demand and supply raster data to 90m and spatially match it with the land space utilization data raster. Using the table-based zoning statistics tool in ArcGIS, according to different land space utilization types, the average value of the demand and supply raster in different types of space is counted respectively. Through this method, the average unit pixel satisfaction of the four types of demand, population carrying capacity, grain production, animal husbandry, and ecological services, of different utilization types is calculated as the target parameter satisfaction capacity input model.

[0059] 4.2 Constraint Setting: Using the resource and environmental carrying capacity of national land space planning and the results of the national land space development suitability assessment, two types of spatial constraints are set: an upper development limit and a lower protection limit. The upper development limit is obtained by removing the areas most suitable for agricultural production, animal husbandry, and ecological protection from the areas with the highest suitability for urban development. The lower protection limit is obtained by taking the union of the areas with the highest suitability for agricultural production and the areas with the highest importance for ecological protection.

[0060] 4.3 Driving Factor Design: Using the ROC test, we analyzed the correlation between the evolution of the land spatial structure in the Huangshui River Basin from 1990 to 2020 and 15 geographic factors: altitude, slope, NPP, temperature, precipitation, soil sandiness, soil clayiness, soil organic carbon, soil moisture, distance to irrigation canals, road accessibility, railway accessibility, urban accessibility, cost distance to Xining, and GDP. We selected geographic factors with an AUC greater than 0.85 as driving factors for the evolution of land spatial functional structure. We determined the driving effects of DEM, slope, temperature, precipitation, NPP, soil sandiness, soil clayiness, and distance to irrigation canals on cultivated land, and the driving effects of NPP, precipitation, DEM, soil organic carbon, and soil moisture on forests. Using multiple linear regression, we combined the land spatial structure in the starting year to analyze the quantitative relationship between land spatial function and structural evolution and the driving factors. We determined the correlation between land use generation probability and different driving factors, which served as the driving force weights of the driving factors, and derived the correlation polynomial between the two.

[0061] 4.4 Complex system modeling and simulation;

[0062] Using the quantitative relationship between the driving factors analyzed in step 4.3 and the evolution of the land space functional structure and the driving factors, based on the spatial atlas data of the driving factors, the driving force weights of different driving factors are combined with the current land space functional structure and the land space use type's ability to meet the functional type, and the weighted summation is used to obtain the probability atlas of the evolution of the land space functional structure. The calculation formula is as follows:

[0063]

[0064] Among them, Prob a，t is the probability of evolution of land space functional structure of land space use type a at time node t, o i The implementation priority of function type i, Supply i，t Demand is the demand parameter that can satisfy function type i at time node t. i，t is the target parameter of function type i at time node t, w j is the quantitative weight of driving factor j, D j is the specific value of the driving force factor j, n k，tis the number of pixels of land space utilization type k at time node t, cap i，k is the supply capacity of a single land type k to functional type i, m, n and N are the number of functional types, the number of driving factors and the number of land space use types respectively;

[0065] In each grid unit of the land space functional structure, the land space utilization type with the highest evolution probability is selected as the evolution result; according to the land space structure of the current simulation results and the ability of land space utilization type to meet the functional type, the demand parameter Supply that the land space structure functional type can meet is calculated. i :

[0066] Supply i =∑n j *cap i，j

[0067] Among them, Supply i is the parameter that can meet the requirements of function type i, n j is the number of pixels of land class j, cap i，j is the supply capacity of a single land type j to functional type i, i = 1, 2, ..., N, where N represents the number of functional types; the demand parameter is compared with the target parameter of the current time node. When the ratio of the maximum demand difference of a single functional type to the demand parameter is less than 1%, and the ratio of the demand difference of all functional types to the sum of the demand parameters of all functional types is less than 0.5%, the national land space structure simulation of the current time node is completed and the national land space structure simulation of the next time node is carried out. Otherwise, the current step is repeated until the requirements of the current time node are met;

[0068] The output scenario finally obtained according to this patent shows that in terms of land space function and structure, the function is more concentrated and efficient, the concentration and contiguousness of the spatial structure is more significant and more consistent with the land function suitability pattern ( Figure 4-Figure 5 High-density settlements have become the main type of population accommodation, with their area share increasing from 21.80% in 2020 to 41.27% in 2100. Cultivated land has gradually concentrated in the middle reaches, with the proportion of cultivated land in the upper reaches decreasing by 52.04%. High-yield cultivated land has become the dominant type, with its proportion increasing from 47.90% to 53.85%. High-NPP grasslands have always carried more than half of the livestock volume. The area of ​​natural land in areas with higher ecological vulnerability has increased by 17.32%, of which the two most evolved vegetation communities, woodland and high-cover grassland, have increased by 254.08 km. 2 .

[0069] Step 5: Measure ecosystem resilience indicators under climate change.

[0070] Based on the land space structure simulation results of each time node obtained in step 4, we used 0.5°*0.5° grid units to count the area proportion of each land space use type at the grid scale, and combined with the vegetation structure proportion counted in the field survey to obtain long-term data on the proportion of vegetation types in each grid unit; based on the long-term climate change data of temperature, precipitation, sunshine days, and surface accumulated temperature, and the regional resource data of soil organic carbon, soil sandy loam, soil carbon and nitrogen content, soil pH, and soil plant cell deposition, the biophysical process model LPJ-GUESS was used to obtain the long-term ecosystem parameter results of NPP, LAI, and river runoff of the nine grids in the Huangshui River Basin as local-scale resilience indicators; and the sum of NPP, river runoff, and the mean of LAI were counted, and all long-term ecosystem parameter results were summarized to the entire Huangshui River Basin as a global-scale resilience indicator.

[0071] Through the optimization simulation of this patent, the final ecological state of the output scenario maintains an upward trend of sustainability and resilience ( Figure 6-Figure 7 In the Huangshui River Basin, global NPP and LAI increased by 0.84% ​​and 0.92% by 2100, respectively. River runoff has been declining for over 30 years, with the absolute value of the linear fit coefficient below 0.25. Meanwhile, local NPP and LAI increased by at least 0.60% and 0.67%, respectively. River runoff remained relatively stable after 2030, with most grids showing a downward trend, and the average annual rate of change across all grids remaining between -0.40% and 0.15%.

[0072] Step 6: Determine the conditions for triggering the loop feedback.

[0073] Based on the long-term ecological parameter results on the global and local grids in step 5, the sustainability and resilience of the system are determined according to the following steps:

[0074] 6.1. Global and local scale ecological status assessment: Use the following algorithm to determine the evolution trend of ecological status parameter results at the global scale and at the local scale:

[0075] ① Set the minimum trend length to 10 years and select a moving window with a time length of 20 years, which moves backward from the starting point of the time series;

[0076] ② Take the maximum and minimum values ​​of the ecological state parameter results in the moving window a i with a i , the corresponding time nodes are i and j respectively; if ij≥t limit , it is considered that in the time series T = {j, j+1, ..., i}, the time series has an upward trend; if ji≥t limit, it is considered that within the time series T = {i, i + 1, ..., j}, the time series has a downward trend; otherwise, there is no evolution trend within the moving window;

[0077] ③ Take the later of time nodes i and j as the starting point of the moving window and repeat step ② until the time series iteration is completed;

[0078] If any ecological state parameter time series data has both an upward and a downward trend, the next target parameter group is selected and step 4 is repeated; if there are no two opposite trends, the current result is output as the result of the optimization of the national land space function and pattern; in this embodiment, the Huangshui River Basin shows the consistency of the ecological parameter evolution trend at both the global and local scales ( Figure 6-Figure 7 At the global scale, the evolution of NPP and LAI in the Huangshui River Basin showed only an upward trend, with the time periods being 2020 to 2070 and 2020 to 2060, respectively, while river runoff showed only a downward trend. At the local scale, NPP and LAI in all grid cells also showed only an upward trend after 2020, while river runoff showed a downward trend in six grids, and no significant evolutionary trend was observed in the remaining three grids. Both indicate that basin optimization will enhance its biodiversity protection and reduce the risk of soil and water loss.

[0079] Step 7: Summarize and visualize the optimization results to identify key areas for sustainability and resilience management. The output results include the adjusted planning target parameter set and the future trend line of the ecological resilience parameters associated with the national land spatial pattern. Areas where ecological state parameters are prone to rising or falling during the adaptation feedback process are identified as key areas for sustainability and resilience management, and the results are visualized on a single map. In this case, local grids 2 and 3 are identified as agricultural areas for soil erosion renewal, grid 5 is identified as an urbanized area at risk of ecological diversity loss, and grid 7 is identified as an agricultural-pastoral transition area where ecosystem function and quality are declining (such as Figure 7 shown).

[0080] Step 8: Scenario comparison and optimization results analysis

[0081] To verify the optimization effectiveness of this method, three scenarios, including the initial scenario (BAU) and the intermediate optimization process (SUS) in the optimization scenario parameter group, were selected for comparison with the final optimization output result (SUS-RES). The optimization effects of the national land space function and structure between each scenario were measured through the evolution of NPP, LAI, river runoff, ecosystem carbon flux, vegetation carbon sink, vegetation biomass quality, vegetation organic carbon ecosystem state parameters, land use intensity, and total area of ​​functional land. Scenario comparison was also carried out to analyze the advantages of the final optimization output result over other scenarios.

[0082] Compared with the BAU scenario, the SUS-RES scenario optimizes the state of ecological parameters at the global scale. In this scenario, NPP and LAI continue to rise rapidly, while the growth of NPP and LAI slows down significantly after 2030, and NPP continues to decline at a rate of 0.15% after 2060. Compared with the SUS scenario, the SUS-RES scenario better optimizes the state of the ecosystem in the local system. The growth rate of NPP and LAI in all grid units is always higher than 0.60%, while the changes in NPP and LAI in grids 2 and 3 in the SUS scenario are always lower than 0.10% after 2060. Compared with the above two comparison scenarios, the SUS-RES scenario maintains the fastest possible social development while maintaining the continuous enhancement of the ecological state of the Huangshui River Basin system, and the sustainability and resilience of the system are optimized and enhanced.

[0083] In addition to the above embodiments, the present invention may also have other implementations. Any technical solution formed by equivalent replacement or equivalent transformation falls within the scope of protection required by the present invention.

Claims

1. A land space optimization method based on cross-scale planning objectives and ecological resilience feedback, including the following steps: Step 1: Classify the land space functions within the set time range in a long time series and analyze and summarize the factors affecting the evolution of land space structure; Step 2: Obtain the planning objectives of the study area at each administrative level and build a planning objective database to determine the functional types of the study area and the implementation priority of the functional types. The planning objective database at least includes the planning target value, target change rate, current value, and evolution trend; Step 3: Set the initial planning target parameter group and optimization step size; Based on the planning target database obtained in step 2, based on the current supply values ​​of the parameters of each functional type in the study area, according to the target change rates in the planning of different administrative levels, and according to the target change rates that evolve within a specific planning period, the potential target parameter values ​​that each functional type will reach at a specific time node under the planning requirements of the government at different administrative levels are calculated. Combined with the potential target parameter values ​​of different administrative levels at the same time node, for the same functional type, the potential target parameter values ​​of all administrative levels are taken as the union to determine the target parameter optimization interval of the functional type; According to the importance of the functional type in the study area, the optimization step size of the functional type is selected within the range of 2%-20% of the corresponding target parameter optimization interval. Based on the optimization step size, the target parameter optimization interval is divided into several endpoint values ​​using the equal interval segmentation method. The endpoints of different functional types are combined to obtain the target parameter group. Combined with the implementation priority of the functional types obtained in step 2, the target parameter groups are sorted from high to low based on the endpoint values; Step 4: Simulation of the future evolution of national land spatial structure; By setting the target parameter group that the future land space needs to meet, the existing constraints in the region, and the driving factors of the regional land space structure evolution, the modeling method is used to simulate the evolution process of the land space function and structure. The specific steps include: 4.

1. Target parameter setting: From all target parameter groups obtained in step 3, select the target parameter groups in order of ranking, take the target parameter group of the target year as the end point, and the data of the starting year corresponding to the target parameter group as the starting point, interpolate the target parameter values ​​for the specified time step, obtain the target parameter values ​​of each time node, and generate a time series of the target parameter group as the target parameter input model in the simulation process; at the same time, with the help of the current land space structure data and the demand satisfaction spatial pattern data, use the spatial analysis method to calculate the average value of each land space use type's ability to meet each target, and use it as the target parameter satisfaction ability input model; 4.

2. Constraint setting: Using the resource and environmental carrying capacity of national land space planning and the results of national land space development suitability assessment, two types of spatial constraints are set: an upper development limit and a lower protection limit. The upper development limit is obtained by removing the areas most suitable for agricultural production, animal husbandry production, and ecological protection from the areas with the highest suitability for urban development. The lower protection limit is obtained by taking the union of the areas with the highest suitability for agricultural production and the areas with the highest importance for ecological protection. 4.

3. Driving force factor setting: Using the ROC test, we analyzed the correlation between the historical evolution of the spatial structure of the study area and geographical elements. We selected geographical elements with an AUC greater than 0.85 as driving factors for the evolution of spatial function and structure. Using multiple linear regression, we combined the spatial structure of the starting year to analyze the quantitative relationship between the evolution of spatial function and structure and the driving factors. 4.

4. Complex system modeling and simulation: Using the driving force factors analyzed in step 4.3 and the quantitative relationship between the evolution of the land space functional structure and the driving force factors, based on the spatial atlas data of the driving force factors, the driving force weights of different driving factors are combined with the current land space functional structure and the land space use type's ability to meet the functional type, and the weighted summation is used to obtain the land space functional structure evolution probability atlas. The calculation formula is as follows: Among them, Prob a,t is the probability of evolution of land space functional structure of land space use type a at time node t, o i The implementation priority of function type i, Supply i,t Demand is the demand parameter that can satisfy function type i at time node t. i,t is the target parameter of function type i at time node t, w j is the quantitative weight of driving factor j, D j is the specific value of the driving force factor j, n k,t is the number of pixels of land space utilization type k at time node t, cap i,k is the supply capacity of a single land type k to functional type i, m, n and N are the number of functional types, the number of driving factors and the number of land space use types respectively; In each grid unit of the land space functional structure, the land space utilization type with the highest evolution probability is selected as the evolution result; according to the land space structure of the current simulation results and the ability of land space utilization type to meet the functional type, the demand parameter Supply that the land space structure functional type can meet is calculated. i : Supply i =∑n j *cap i,j Among them, Supply i is the parameter that can meet the requirements of function type i, n j is the number of pixels of land class j, cap i,j is the supply capacity of a single land type j to functional type i, i = 1, 2, …, N, where N represents the number of functional types; the demand parameter is compared with the target parameter of the current time node. When the ratio of the maximum demand difference of a single functional type to the demand parameter is less than 1%, and the ratio of the demand difference of all functional types to the sum of the demand parameters of all functional types is less than 0.5%, the national land space structure simulation of the current time node is completed and the national land space structure simulation of the next time node is carried out. Otherwise, the current step is repeated until the requirements of the current time node are met; Step 5: Measure ecosystem resilience indicators under climate change. The specific method is as follows: Based on the land space structure simulation results at each time node obtained in step 4, the area proportion of each land space use type at the grid scale is calculated. Combined with the vegetation structure proportion calculated by field survey, the long-term time series data of the vegetation type proportion of each grid unit is obtained. The calculation formula for the vegetation type proportion within the grid unit is as follows: Among them, PFT i,t is the proportion of vegetation type i in grid g at time node t, Cov j,g,t is the area proportion of land space use type j in grid g at time node t, Prop i,j is the proportion of vegetation type i in land use type j; Based on the long-term climate change data of temperature, precipitation, sunshine days, evapotranspiration, and surface accumulated temperature, and the regional resource data of soil organic carbon, soil sandy loam, soil carbon and nitrogen content, soil pH, and soil plant cell deposition, the long-term vegetation type proportion data of all grid cells in the study area were used as the ecosystem vegetation structure, and the long-term climate change data and regional resource data were used as nutrients for biological competition. The biophysical process model LPJ-GUESS, which simulates the material competition of ecosystem vegetation, was used to simulate the evolution of ecosystem state in the process of land space evolution in the study area, and generate the long-term evolution process of future ecosystem state and vegetation structure. The long-term ecosystem parameter results of NPP, LAI, river runoff, ecosystem carbon flux, vegetation carbon sink, vegetation biomass, and vegetation organic carbon at the grid scale were obtained as local-scale resilience indicators. All long-term ecosystem parameter results were summarized to the entire study area as global-scale resilience indicators. Step 6: Determine the conditions for triggering the loop feedback; Based on the long-term ecological parameter results on the global and local grids in step 5, the sustainability and resilience of the system are determined according to the following steps: 6.

1. Global and local scale ecological status assessment: Use the following algorithm to determine the evolution trend of ecological status parameter results at the global scale and at the local scale: ①Set the minimum trend length to t limit , select a moving window with a time length of l, which moves backward from the starting point of the time series, where the value range of the time length l is 20%-40% of the time span from the starting year to the target year; ② Take the maximum and minimum values ​​of the ecological state parameter results in the moving window a i with a j , the corresponding time nodes are i and j respectively; if ij≥t limit , it is considered that in the time series T={j,j+1,…,i}, the time series has an upward trend; if ji≥t limit , it is considered that within the time series T = {i,i+1,…,j}, the time series has a downward trend; otherwise, there is no evolution trend within the moving window; ③ Take the later of time nodes i and j as the starting point of the moving window and repeat step ② until the time series iteration is completed; If any ecological state parameter time series data has both an upward and downward trend, select the next target parameter group and repeat step 4; if there are no two opposite trends, output the current result as the optimization result of national land space function and pattern; Step 7: Summarize and visualize the optimization results.

2. The land space optimization method based on cross-scale planning objectives and ecological resilience feedback according to claim 1 is characterized by: The specific method of step 1 is as follows: with the help of Google Earth Engine big data cloud platform, using long-term satellite image data, population data, VIIRS night light data, MODIS NPP inversion products, Landsat NDVI inversion products, and Landsat NDWI inversion products, the long-term classification of the regional functions of the study area's land space is carried out from the perspective of regional functions. Based on the analysis of regional function etiology, influencing factors are selected as independent variables from the three dimensions of natural carrying capacity, socio-economic spatial organization, and regional interaction. The partial correlation analysis method is used to preliminarily select the influencing factors that affect the evolution of the land space structure.

3. The land space optimization method based on cross-scale planning objectives and ecological resilience feedback according to claim 1 is characterized by: The specific method of step 2 is as follows: collect planning text data related to the study area through public document collection and interviews with government departments; use meta-analysis, map interpretation, and semantic extraction methods to analyze the sustainability positioning and planning goals of the study area at different administrative levels, namely national, provincial, and local, and generate a planning goal database at different administrative levels; compare the sustainability positioning of the study area at different administrative levels, determine the functional type of the study area, and analyze the implementation priority of the functional type of the study area.

4. The land space optimization method based on cross-scale planning objectives and ecological resilience feedback according to claim 1 is characterized by: The specific details of step 7 are as follows: the output results include the adjusted planning target parameter set, the future trend line of the ecological resilience parameters associated with the national land spatial pattern, and the areas where the ecological state parameters are prone to rising or falling during the adaptation feedback process are identified as key areas for sustainability and resilience management, and the results are visualized on a graph.

5. The land space optimization method based on cross-scale planning objectives and ecological resilience feedback according to claim 1 is characterized by: It also includes step 8, scenario comparison and optimization result analysis: The upper limit and lower limit of scenario parameters in the optimization scenario parameter group were selected for comparison with the final optimization results; the optimization effects of the land space function and structure between each scenario were measured through the evolution of NPP, LAI, river runoff, ecosystem carbon flux, vegetation carbon sink, vegetation biomass quality, vegetation organic carbon ecosystem state parameters and land use intensity, and the total area of ​​functional land. The scenarios were compared to analyze the advantages of the final optimization output results compared with other scenarios.

Images