A comprehensive prediction method for resource and environmental carrying capacity

Abstract

This invention provides a comprehensive prediction method for resource and environmental carrying capacity, relating to the field of hub boxes. Its key features include acquiring multi-source spatiotemporal data on natural resources, ecological environment, socio-economic factors, and spatial geography of the target area, and performing spatiotemporal alignment and standardization; and constructing a PSR framework covering resource supply and environmental pollution carrying capacity. The advantages of this invention are: it overcomes the limitations of traditional static, single-dimensional evaluation by dynamically simulating human-land interaction feedback through bidirectional coupling of system dynamics and multi-agent systems; it introduces spatial gridding and geographically weighted regression to accurately characterize the spatial heterogeneity of resource and environmental elements; it constructs an inter-regional ecological-economic-social interaction model to quantify external impacts and improve the synergy and realism of predictions; and it combines multi-scenario simulation, uncertainty analysis, and online Bayesian calibration to achieve accurate identification and dynamic early warning of overload risks, providing scientific support for land spatial planning and sustainable development decision-making.

Description

Technical Field

[0001] This invention relates to the field of resource and environmental data processing and prediction technology, and in particular to a comprehensive prediction method for resource and environmental carrying capacity. Background Technology

[0002] Resource and environmental carrying capacity refers to the maximum capacity and sustainable carrying capacity of natural resources, ecological environment, land, water resources, atmospheric capacity, and other factors within a certain period and region to continuously support population size, economic and social development, urban construction, and industrial activities. It considers both the upper limit of the supply of resources such as water, land, and minerals, and the bottom line of ecological environment pollution absorption, self-repair, and ecosystem stability. It is an important basic threshold for measuring whether the scale of regional development is reasonable, whether the development of land space is appropriate, and whether the coordinated and sustainable development of humans and nature can be achieved.

[0003] Currently, existing resource and environmental carrying capacity assessment technologies generally employ static indicators or single dimensions (such as considering only water or land resources) for evaluation, which fails to fully reflect the complexity and dynamic evolution characteristics of the resource and environmental system. Firstly, traditional methods neglect the dynamic changes of resource and environmental elements in time and space, lacking the ability to simulate and predict carrying capacity changes under different future development scenarios. Secondly, most assessment models treat regions as homogeneous units, failing to fully characterize the spatial heterogeneity of resource and environmental elements and ignoring the radiation and interaction effects between regions in ecological, economic, and social aspects.

[0004] Therefore, there is an urgent need for a comprehensive carrying capacity evaluation method that can integrate multi-source data, couple multi-dimensional resource and environmental factors, and possess spatiotemporal dynamic simulation capabilities. Summary of the Invention

[0005] The purpose of this invention is to provide a comprehensive prediction method for resource and environmental carrying capacity, which solves the problem that existing resource and environmental carrying capacity evaluation technologies, due to the use of static indicators and single-dimensional assessment, are unable to reflect the spatiotemporal dynamics, spatial heterogeneity, and regional interactions of the resource and environmental system.

[0006] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows:

[0007] A comprehensive prediction method for resource and environmental carrying capacity includes: S1: acquiring multi-source spatiotemporal data of natural resources, ecological environment, socio-economic conditions and spatial geography of the target area, and performing spatiotemporal alignment and standardization;

[0008] S2: Based on the PSR framework, construct a multi-dimensional evaluation index system covering resource supply, environmental pollution control, ecological support and economic and social development. Based on this, set the state variables, rate variables and auxiliary variables of each subsystem, establish causal feedback loop equations within and between subsystems, and form a system dynamics model.

[0009] S3: Spatial gridding of the region is performed, and spatial heterogeneity of resource and environmental elements is extracted by geographic weighted regression (GWR) and spatial autocorrelation analysis. Local regression coefficients are generated as spatial correction factors and embedded into the state variable equations corresponding to the system dynamics model.

[0010] S4: Construct a multi-agent model, where each agent represents a government, enterprise, or resident within a grid cell; input the macroscopic state variables output by the system dynamics model at each time step as environmental constraints into the multi-agent model, which simulates resource utilization and pollution emission decisions according to preset behavior rules, and feeds back the aggregated total demand and total emissions to the system dynamics model to achieve bidirectional dynamic coupling and construct a dynamic simulation and prediction model.

[0011] S5: Establish an inter-regional ecological-economic-social interaction model to quantify the impact of neighboring regions on the target area in terms of resource supply, pollution transfer, and economic radiation, and integrate it into the prediction model as an external input;

[0012] S6: Set up multiple development scenarios and drive the model by adjusting parameters such as population growth rate, industrial upgrading rate, and water-saving technology adoption rate to simulate the spatiotemporal evolution of carrying capacity under different scenarios;

[0013] S7: Calculate the comprehensive bearing capacity index of each grid, analyze its probability distribution using kernel density estimation, identify the spatial center, dispersion trend and directional evolution of high and low value areas using the standard deviation ellipse method, and determine the overload or critical overload area.

[0014] S8: Conduct Monte Carlo uncertainty analysis and Morris sensitivity test to assess model robustness and identify key driving factors, and support online Bayesian calibration based on newly added observation data;

[0015] S9: Overlay the carrying capacity index with the territorial spatial planning zoning to identify risk areas within the red line or outside the development boundary, and output differentiated early warning and control recommendations;

[0016] S10: Outputs a load-bearing capacity level zoning map and an index time-series variation curve, using grids or administrative regions as units.

[0017] Furthermore, the multi-source spatiotemporal data in step S1 is obtained by fusing remote sensing images, ground monitoring stations, statistical yearbooks, and social perception data;

[0018] Natural resource data includes total water resources, precipitation, surface water and groundwater reserves, and land use / cover types.

[0019] Ecological and environmental data include water quality monitoring data, PM2.5, and other data. 2.5 Related to SO2 concentration and the boundary of ecological red line areas;

[0020] Socioeconomic data include population size and density, GDP, and the proportion of the three industries.

[0021] The spatial geographic data includes a 30-meter resolution digital elevation model and a 1:100,000 river system network.

[0022] Furthermore, in step S2, the weights of each indicator are assigned by combining the entropy weight method and the Delphi expert scoring method. The formula for calculating the combined weight is as follows:

[0023]

[0024] in, For adjustment coefficients, The weights obtained by the entropy weight method, The weights are normalized for the expert scores.

[0025] Furthermore, the spatial meshing in step S3 employs an adaptive partitioning strategy:

[0026] In areas with a population density greater than 500 people / km², a GDP density greater than 100 million yuan / km², or located in national-level ecological function zones, a 1 km × 1 km high-resolution grid is adopted.

[0027] The remaining areas use a 5 km × 5 km low-resolution grid.

[0028] Furthermore, in step S4, the behavioral rules of the government intelligent agent include: when the comprehensive carrying capacity index of the grid cell is lower than 0.6 for two consecutive time steps, the negative list for industry access is automatically triggered;

[0029] The behavioral rules for enterprise intelligent agents include: dynamically adjusting the production scale based on the unit GDP water consumption threshold and the sewage discharge cost function of the grid in which they are located;

[0030] The behavioral rules for the resident intelligent agent include: adjusting the intensity of domestic water use based on the per capita available water resources.

[0031] Furthermore, the inter-regional ecological-economic-social multi-factor interaction model in step S5 includes:

[0032] The economic radiation sub-model based on the improved gravity model has the following formula for calculating the strength of economic linkages: ,in For the region economic scale, For road network distance, The attenuation coefficient;

[0033] Based on distributed hydrological model A pollution transport sub-model coupled with an atmospheric Gaussian diffusion model;

[0034] based on The model-based ecosystem service flow sub-model is used to quantify the flow of water conservation and soil retention services between regions.

[0035] Furthermore, in the pollution transport sub-model, the amount of transboundary water pollutant input... The calculation formula is:

[0036]

[0037] in, Let K be the flow velocity at section k. This represents the measured concentration of the pollutant. The migration conversion coefficient was calibrated using historical monitoring data; atmospheric transboundary deposition was calculated using a Gaussian plume model, and the diffusion parameters were corrected using the least squares method with measured data from more than five surrounding atmospheric monitoring stations.

[0038] Furthermore, the various future development scenarios in step S6 include:

[0039] Baseline scenario: Keep historical trend parameters unchanged;

[0040] Ecological priority scenario: Increase the ecological land protection rate by 10% and restrict the expansion of water-intensive industries;

[0041] Priority scenario for economic development: GDP growth rate increases to 7% annually, while the improvement in water resource utilization efficiency slows down;

[0042] Coordinated development scenario: Simultaneously promote a 5% annual increase in the adoption rate of water-saving technologies and raise the proportion of investment in ecological restoration to 3%.

[0043] Furthermore, the online Bayesian calibration in step S8 employs a particle filter algorithm, using newly added quarterly remote sensing vegetation index NDVI and monthly water quality monitoring data as observations to dynamically update the key parameters of the ecological support subsystem in the system dynamics model.

[0044] Furthermore, the method can be applied to the preparation of provincial or municipal territorial spatial master plans, dynamic assessment of ecological protection red lines, preliminary environmental carrying capacity assessment of major linear projects (such as high-speed railways and oil pipelines), or annual diagnosis of the sustainable development capacity of urban agglomerations.

[0045] The beneficial effects of this invention are as follows:

[0046] (i) Breaking through the limitations of traditional static, single-dimensional evaluation methods, a dynamic simulation and prediction model with bidirectional coupling of system dynamics and multi-agent systems is constructed. By combining macroscopic system dynamics models with microscopic multi-agent behavioral rules, the adaptive decision-making behavior of governments, enterprises, and residents under different carrying capacity states can be dynamically reflected, realizing the interactive feedback simulation between resource and environmental systems and human activities. At the same time, the introduction of spatial gridding, geographically weighted regression, and spatial autocorrelation analysis effectively characterizes the spatial heterogeneity of resource and environmental elements and improves the model's ability to identify differences within regions.

[0047] (ii) It fully considers the ecological-economic-social interactions between regions and establishes a multi-factor regional interaction model that includes economic radiation, pollution transmission, and ecosystem service flow. By quantifying the impact of neighboring regions on the target area in terms of resource supply, pollution transfer, and economic radiation, and integrating these impacts as external inputs into the prediction model, it overcomes the shortcomings of traditional methods that treat regions as isolated systems, making carrying capacity prediction more regionally coordinated and realistically reasonable.

[0048] (III) This invention possesses a comprehensive uncertainty analysis and model calibration mechanism, supporting multi-scenario simulation and dynamic updates. Through Monte Carlo uncertainty analysis, Morris sensitivity testing, and online Bayesian calibration based on particle filtering, it can assess model robustness, identify key driving factors, and dynamically update model parameters using newly added observation data. Combining kernel density estimation, standard deviation ellipse method, and overlay analysis of territorial spatial planning zones, it can accurately identify overloaded or critically overloaded areas, output differentiated early warning and control recommendations, and is widely applicable to practical scenarios such as territorial spatial planning, ecological red line assessment, and preliminary assessment of major projects. Attached Figure Description

[0049] Figure 1 This is an overall flowchart of a comprehensive prediction method for resource and environmental carrying capacity according to the present invention.

[0050] Figure 2 This is a diagram of the bidirectional coupling architecture of the multi-agent system of the present invention. Detailed Implementation

[0051] To make the content of this invention easier to understand, the technical solutions in the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings.

[0052] like Figure 1As shown, a comprehensive prediction method for resource and environmental carrying capacity includes: S1: acquiring multi-source spatiotemporal data of natural resources, ecological environment, socio-economic conditions, and spatial geography of the target area, and performing spatiotemporal alignment and standardization; S2: constructing a multi-dimensional evaluation index system based on the PSR framework, covering resource supply, environmental pollution, ecological support, and socio-economic development, and setting state variables, rate variables, and auxiliary variables for each subsystem accordingly, establishing causal feedback loop equations within and between subsystems, and forming a system dynamics model; S3: spatially gridding the region, using geographically weighted regression (GWR) and spatial autocorrelation analysis to extract the spatial heterogeneity of resource and environmental elements, generating local regression coefficients as spatial correction factors, and embedding them into the state variable equations corresponding to the system dynamics model; S4: constructing a multi-agent model, where each agent represents the government, enterprises, or residents within the grid unit; inputting the macroscopic state variables output by the system dynamics model at each time step as environmental constraints into the multi-agent model, which simulates resource utilization and pollution emission decisions according to preset behavioral rules, and feeding back the aggregated total demand and total emissions to the system dynamics model. The system achieves two-way dynamic coupling and constructs a dynamic simulation and prediction model; S5: Establishes an inter-regional ecological-economic-social interaction model, quantifies the impact of neighboring regions on the target area in terms of resource supply, pollution transfer, and economic radiation, and integrates it into the prediction model as an external input; S6: Sets multiple development scenarios, drives the model by adjusting parameters such as population growth rate, industrial upgrading rate, and water-saving technology popularization rate, and simulates the spatiotemporal evolution of carrying capacity under different scenarios; S7: Calculates the comprehensive carrying capacity index of each grid, uses kernel density estimation to analyze its probability distribution, uses the standard deviation ellipse method to identify the spatial center, dispersion trend, and directional evolution of high and low value areas, and determines overloaded or critically overloaded areas; S8: Conducts Monte Carlo uncertainty analysis and Morris sensitivity test to evaluate the robustness of the model and identify key driving factors, and supports online Bayesian calibration based on newly added observation data; S9: Overlays the carrying capacity index with the territorial spatial planning zoning, identifies risk areas within the red line or outside the development boundary, and outputs differentiated early warning and control suggestions; S10: Outputs carrying capacity level zoning maps and index time series change curves using grids or administrative regions as units.

[0053] like Figure 1 As shown, the multi-source spatiotemporal data in step S1 is obtained through the fusion of remote sensing imagery, ground monitoring stations, statistical yearbooks, and social perception data; among them, natural resource data includes total water resources, precipitation, surface water and groundwater reserves, and land use / cover type; ecological and environmental data includes water quality section monitoring data, PM2.5, and PM2.5. 2.5 The data includes SO2 concentration, ecological red line boundaries; socioeconomic data including population size and density, GDP, and the proportion of the three industries; and spatial geographic data including a 30-meter resolution digital elevation model and a 1:100,000 river system network.

[0054] like Figure 1 As shown, in step S2, the weights of each indicator are assigned by combining the entropy weight method and the Delphi expert scoring method. The formula for calculating the combined weight is:

[0055]

[0056] in, For adjustment coefficients, The weights obtained by the entropy weight method, The weights are normalized for the expert scores.

[0057] like Figure 2 As shown, the spatial gridding in step S3 adopts an adaptive partitioning strategy: in areas with a population density greater than 500 people / km², a GDP density greater than 100 million yuan / km², or located in national-level ecological function zones, a 1 km × 1 km high-resolution grid is used; in other areas, a 5 km × 5 km low-resolution grid is used. In step S4, the behavioral rules for the government agent include: when the comprehensive carrying capacity index of its grid unit is lower than 0.6 for two consecutive time steps, the negative list for industrial access is automatically triggered; the behavioral rules for the enterprise agent include: dynamically adjusting the production scale based on the unit GDP water consumption threshold and the sewage discharge cost function of its grid; the behavioral rules for the resident agent include: adjusting the intensity of domestic water use based on the per capita available water resources.

[0058] like Figure 1 As shown, the inter-regional ecological-economic-social multi-factor interaction model in step S5 includes: an economic radiation sub-model based on an improved gravity model, whose formula for calculating the intensity of economic linkages is: ,in For the region economic scale, For road network distance, The attenuation coefficient is based on a distributed hydrological model. A pollution transport sub-model coupled with an atmospheric Gaussian diffusion model; based on The model-based ecosystem service flow sub-model is used to quantify the flow of water conservation and soil retention services between regions.

[0059] In the pollution transport sub-model, the amount of transboundary water pollutant input... The calculation formula is:

[0060]

[0061] in, Let K be the flow velocity at section k. This represents the measured concentration of the pollutant. The migration conversion coefficient was calibrated using historical monitoring data; atmospheric transboundary deposition was calculated using a Gaussian plume model, and the diffusion parameters were corrected using the least squares method with measured data from more than five surrounding atmospheric monitoring stations.

[0062] like Figure 2 As shown, the various future development scenarios in step S6 include: Baseline scenario: maintaining historical trend parameters unchanged; Ecological priority scenario: increasing the ecological land protection rate by 10% and restricting the expansion of water-intensive industries; Economic development priority scenario: increasing the average annual GDP growth rate to 7% and slowing down the improvement of water resource utilization efficiency; Coordinated development scenario: simultaneously promoting a 5% annual increase in the popularization rate of water-saving technologies and increasing the proportion of ecological restoration investment to 3%. The online Bayesian calibration in step S8 uses a particle filter algorithm, with newly added quarterly remote sensing vegetation index NDVI and monthly water quality monitoring data as observations, dynamically updating the key parameters of the ecological support subsystem in the system dynamics model. This method is applied to the preparation of provincial or municipal territorial spatial master plans, dynamic assessment of ecological protection red lines, preliminary environmental carrying capacity assessment of major linear projects (such as high-speed railways and oil pipelines), or annual diagnosis of the sustainable development capacity of urban agglomerations.

[0063] In the implementation process, firstly, multi-source spatiotemporal data of the target area is acquired and standardized; then, a multi-dimensional evaluation index system is constructed based on the PSR framework, and a system dynamics model is established; next, the region is adaptively spatially gridded, and spatial heterogeneity is extracted using geographic weighted regression to correct the model; on this basis, a multi-agent model (simulating the behavior of government, enterprises, and residents) is coupled with the system dynamics model to achieve two-way dynamic interaction; simultaneously, an inter-regional interaction model is constructed to quantify the impact of external resources, pollution, and the economy; subsequently, multiple development scenarios are set to drive the model for simulation and prediction; overloaded areas are identified by calculating the comprehensive carrying capacity index and combining it with spatial statistical methods; then, uncertainty and sensitivity analysis is used to evaluate the robustness of the model and support online calibration; finally, the results are overlaid with the national land spatial planning to output risk warnings, control suggestions, and visualization results.

[0064] Example: Taking the revision of the overall land use plan of a provincial capital city as an example, the city can apply the method of this patent: First, integrate multi-source data such as satellite remote sensing, environmental monitoring stations, and statistical yearbooks; then, construct an indicator system including water resources, atmospheric environment, ecological land use, and industrial layout based on the PSR framework, and establish a system dynamics model; subsequently, use a 1km×1km high-resolution grid to divide the built-up area into grids, and use geographic weighted regression to extract the resource and environmental characteristics of each grid; on this basis, construct a multi-agent model composed of the government, enterprises, and residents to simulate how the various stakeholders respond to changes in carrying capacity and provide feedback under different scenarios such as "ecological priority" or "coordinated development"; at the same time, quantify the cross-border impact of surrounding cities on the city through rivers, atmosphere, and economic ties; finally, predict the evolution of carrying capacity in various regions over the next 15 years through the model, accurately identify the risk of overloading in the western industrial cluster and the southern new city, and propose a negative list for industrial access, key areas for ecological restoration, and suggestions for optimizing the layout of infrastructure, providing scientific support for planning decisions.

[0065] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A comprehensive prediction method for resource and environmental carrying capacity, characterized in that, Includes the following steps: S1: Acquire multi-source spatiotemporal data on natural resources, ecological environment, socio-economic conditions, and spatial geography of the target area, and perform spatiotemporal alignment and standardization; S2: Based on the PSR framework, construct a multi-dimensional evaluation index system covering resource supply, environmental pollution control, ecological support and economic and social development. Based on this, set the state variables, rate variables and auxiliary variables of each subsystem, establish causal feedback loop equations within and between subsystems, and form a system dynamics model. S3: Spatial gridding of the region is performed, and spatial heterogeneity of resource and environmental elements is extracted by geographic weighted regression (GWR) and spatial autocorrelation analysis. Local regression coefficients are generated as spatial correction factors and embedded into the state variable equations corresponding to the system dynamics model. S4: Construct a multi-agent model, where each agent represents a government, enterprise, or resident within a grid cell; input the macroscopic state variables output by the system dynamics model at each time step as environmental constraints into the multi-agent model, which simulates resource utilization and pollution emission decisions according to preset behavior rules, and feeds back the aggregated total demand and total emissions to the system dynamics model to achieve bidirectional dynamic coupling and construct a dynamic simulation and prediction model. S5: Establish an inter-regional ecological-economic-social interaction model to quantify the impact of neighboring regions on the target area in terms of resource supply, pollution transfer, and economic radiation, and integrate it into the prediction model as an external input; S6: Set up multiple development scenarios and drive the model by adjusting parameters such as population growth rate, industrial upgrading rate, and water-saving technology adoption rate to simulate the spatiotemporal evolution of carrying capacity under different scenarios; S7: Calculate the comprehensive bearing capacity index of each grid, analyze its probability distribution using kernel density estimation, identify the spatial center, dispersion trend and directional evolution of high and low value areas using the standard deviation ellipse method, and determine the overload or critical overload area. S8: Conduct Monte Carlo uncertainty analysis and Morris sensitivity test to assess model robustness and identify key driving factors, and support online Bayesian calibration based on newly added observation data; S9: Overlay the carrying capacity index with the territorial spatial planning zoning to identify risk areas within the red line or outside the development boundary, and output differentiated early warning and control recommendations; S10: Outputs a load-bearing capacity level zoning map and an index time-series variation curve, using grids or administrative regions as units.

2. The comprehensive prediction method for resource and environmental carrying capacity according to claim 1, characterized in that, The multi-source spatiotemporal data in step S1 is obtained by fusing remote sensing images, ground monitoring stations, statistical yearbooks and social perception data. Natural resource data includes total water resources, precipitation, surface water and groundwater reserves, and land use / cover type. Ecological and environmental data include water quality monitoring data, PM2.5, and other data. 2.5 Related to SO2 concentration and the boundary of ecological red line areas; Socioeconomic data include population size and density, GDP, and the proportion of the three industries. Spatial geographic data includes a 30-meter resolution digital elevation model and a 1:100,000 river system network.

3. The comprehensive prediction method for resource and environmental carrying capacity according to claim 2, characterized in that, In step S2, the weights of each indicator are assigned by combining the entropy weight method and the Delphi expert scoring method. The formula for calculating the combined weight is as follows: in, For adjustment coefficients, The weights obtained by the entropy weight method, The weights are normalized for the expert scores.

4. The comprehensive prediction method for resource and environmental carrying capacity according to claim 1, characterized in that, The spatial meshing in step S3 adopts an adaptive partitioning strategy: In areas with a population density greater than 500 people / km², a GDP density greater than 100 million yuan / km², or located in national-level ecological function zones, a 1 km × 1 km high-resolution grid is adopted. The remaining areas use a 5 km × 5 km low-resolution grid.

5. The comprehensive prediction method for resource and environmental carrying capacity according to claim 4, characterized in that, In step S4, the behavioral rules of the government intelligent agent include: when the comprehensive carrying capacity index of the grid cell is lower than 0.6 for two consecutive time steps, the negative list for industry access is automatically triggered. The behavioral rules for enterprise intelligent agents include: dynamically adjusting the production scale based on the unit GDP water consumption threshold and the sewage discharge cost function of the grid in which they are located; The behavioral rules for the resident intelligent agent include: adjusting the intensity of domestic water use based on the per capita available water resources.

6. The comprehensive prediction method for resource and environmental carrying capacity according to claim 1, characterized in that, The inter-regional ecological-economic-social multi-factor interaction model in step S5 includes: The economic radiation sub-model based on the improved gravity model has the following formula for calculating the strength of economic linkages: ,in For the region economic scale, For road network distance, The attenuation coefficient; Based on distributed hydrological model A pollution transport sub-model coupled with an atmospheric Gaussian diffusion model; based on The model-based ecosystem service flow sub-model is used to quantify the flow of water conservation and soil retention services between regions.

7. The comprehensive prediction method for resource and environmental carrying capacity according to claim 6, characterized in that, In the pollution transport sub-model, the amount of transboundary water pollutant input... The calculation formula is: in, Let K be the flow velocity at section k. This represents the measured concentration of the pollutant. The migration conversion coefficient was calibrated using historical monitoring data; atmospheric transboundary deposition was calculated using a Gaussian plume model, and the diffusion parameters were corrected using the least squares method with measured data from more than five surrounding atmospheric monitoring stations.

8. The comprehensive prediction method for resource and environmental carrying capacity according to claim 7, characterized in that, The various future development scenarios in step S6 include: Baseline scenario: Keep historical trend parameters unchanged; Ecological priority scenario: Increase the ecological land protection rate by 10% and restrict the expansion of water-intensive industries; Priority scenario for economic development: GDP growth rate increases to 7% annually, while the improvement in water resource utilization efficiency slows down; Coordinated development scenario: Simultaneously promote a 5% annual increase in the adoption rate of water-saving technologies and raise the proportion of investment in ecological restoration to 3%.

9. The comprehensive prediction method for resource and environmental carrying capacity according to claim 1, characterized in that, The online Bayesian calibration in step S8 uses a particle filter algorithm, with the newly added quarterly remote sensing vegetation index NDVI and monthly water quality monitoring data as observations, to dynamically update the key parameters of the ecological support subsystem in the system dynamics model.

10. The comprehensive prediction method for resource and environmental carrying capacity according to claim 9, characterized in that, The method is applied to the preparation of provincial or municipal territorial spatial master plans, dynamic assessment of ecological protection red lines, preliminary environmental carrying capacity assessment of major linear projects (such as high-speed railways and oil pipelines), or annual diagnosis of the sustainable development capacity of urban agglomerations.

Images