A simulation method and system for the evolution of vulnerability in sandy areas

Abstract

This application relates to a method and system for simulating the evolution of vulnerability in sandy areas. The method includes the following steps: obtaining the natural environment data of the research area; setting constraint conditions based on the social, economic, and ecological construction needs of the research area according to the natural environment data; optimizing the land use structure based on the constraint conditions to meet the land use quantity structure; simulating the changes in the vegetation, social economy, and water resource data development of the research area based on the determination of the land use quantity structure to construct a model; calibrating, validating, and conducting sensitivity analysis of key parameters on the model, continuously iteratively optimizing the model, and promoting changes in vulnerability assessment indicators to achieve the simulation of vulnerability in the research area. The method for simulating the evolution of vulnerability in sandy areas of this application can achieve the simulation of vulnerability evolution under different scenarios, in order to formulate the best solutions that meet the needs of various stakeholders for different regions and different objectives. This method can reduce costs and improve work efficiency.

Description

Technical Field

[0001] This application relates to the technical field of environmental vulnerability processing and analysis, and particularly to a method and system for simulating the evolution of vulnerability in sandy areas. Background Art

[0002] Environmental vulnerability is the sensitive response and self-recovery ability of an environmental system relative to external disturbances at a specific spatio-temporal scale, and is the result of the combined action of natural attributes and human interference behaviors. The vulnerability of the ecological environment has both natural and human factors. Natural factors include geological structure, geomorphic characteristics, surface composition materials, regional hydrological characteristics, biological group types, and climate factors. Human factors refer to the abuse of various substances and resources by humans, resulting in resource depletion, ecological damage, and excessive pollutant emissions, which disrupt the ecological balance and pose a huge pressure on the environment.

[0003] Currently, due to different environments in each region, different degrees of factor influence, and different demands due to different goals, the evolution of vulnerability varies in different situations in each region. Moreover, the more complex the environment, the more factors need to be considered, and the more complex the model design. Therefore, in order to improve different environments in each region, different models need to be designed to simulate the evolution of vulnerability in different scenarios, resulting in a large and complicated workload, increased costs, and reduced work efficiency. Therefore, it is desired to design a model that can simulate the evolution of vulnerability in different scenarios in order to develop the best solutions that meet the needs of various stakeholders for different regions and different goals. The environment in sandy areas has relatively weak sensitive response and self-recovery ability, and is greatly affected by both natural and human factors, and its vulnerability evolution is relatively complex. Therefore, considering multiple factors to establish modeling rules, a method for simulating the evolution of vulnerability in sandy areas is designed to achieve the simulation of the evolution of vulnerability in different scenarios. Summary of the Invention

[0004] This application provides a method and system for simulating the evolution of vulnerability in sandy areas. By considering the natural environment and socio-economic characteristics of sandy areas, modeling principles are designed to achieve the optimal allocation of regional land use under the constraints of different development goals to achieve the optimal comprehensive benefits. The method is applicable to simulating the evolution of vulnerability in different scenarios, in order to develop the best solutions that meet the needs of various stakeholders for different regions and different goals, reduce costs, and improve work efficiency.

[0005] In a first aspect, the present application provides a method for simulating the evolution of vulnerability in sandy areas, comprising the following steps: obtaining natural environment data of the research area; wherein, the natural environment data includes climate data, vegetation data, and water resource data; setting constraint conditions based on the social economic and ecological construction requirements of the research area according to the natural environment data; optimizing the land use structure based on the constraint conditions to meet the land use quantity structure; simulating the changes in the development of vegetation, social economy, and water resource data in the research area based on the determination of the land use quantity structure to construct a model; calibrating, validating, and performing sensitivity analysis on key parameters of the model, continuously iteratively optimizing the model, and promoting the change of vulnerability assessment indicators to achieve the vulnerability simulation of the research area.

[0006] In a second aspect, the present application provides a system for simulating the evolution of vulnerability in sandy areas, the system comprising: an acquisition module for obtaining natural environment data of the research area; wherein, the natural environment data includes climate data, vegetation data, and water resource data; a multi-objective constraint module for setting constraint conditions based on the social economic and ecological construction requirements of the research area according to the natural environment data;

[0007] a processing module for optimizing the land use structure based on the constraint conditions to meet the land use quantity structure; a construction module for simulating the changes in the development of vegetation, social economy, and water resource data in the research area based on the determination of the land use quantity structure to construct a model;

[0008] a verification module for calibrating, validating, and performing sensitivity analysis on key parameters of the model, continuously iteratively optimizing the model, and promoting the change of vulnerability assessment indicators to achieve the vulnerability simulation of the research area.

[0009] The present application has at least the following advantages:

[0010] According to the most important environmental characteristics of the sandy area, namely the arid and less rainy climate, poor vegetation conditions, and the widespread water shortage problem in the sandy area, by obtaining the natural data of the research area, namely climate data, vegetation data, and water resource data, and combining the social, economic, and ecological construction needs of the research area, set the constraints; and based on the other important factor affecting the evolution of the vulnerability of the sandy area, land use change, optimize the land use structure based on the constraints, and comprehensively consider the demands of different subjects to optimize the spatial layout of land use, so as to achieve the optimal comprehensive benefit of the regional land use configuration under different development goal constraints. And finally, further simulate the changes in the development of the vegetation, social economy, and water resource data of the research area according to the determination of the land use quantity structure, construct a simulation model of the vulnerability evolution of the sandy area, and verify the model to achieve accurate simulation of the vulnerability of the research area, and realize the prediction of the vulnerability evolution of the EES composite system in the sandy area under different scenarios, such as climate change, social and economic development, and ecological governance, in order to formulate the best solutions that meet the needs of various stakeholders for different regions and different goals, reduce costs, and improve work efficiency. Description of the Drawings

[0011] Figure 1 It is an application environment diagram showing the simulation method of the vulnerability evolution of the sandy area in an embodiment;

[0012] Figure 2 It is a schematic diagram of the step flow showing the simulation method of the vulnerability evolution of the sandy area in an embodiment;

[0013] Figure 3 It is a schematic diagram of the process showing the simulation method of the vulnerability evolution of the sandy area in an embodiment;

[0014] Figure 4 It is a flowchart showing the simulation method of the vulnerability evolution of the sandy area in an embodiment;

[0015] Figure 5 It is a schematic diagram of the process showing the land use quantity structure in an embodiment;

[0016] Figure 6 It is a flowchart showing the land use quantity structure in an embodiment;

[0017] Figure 7 It is a bar chart showing the dynamic evolution of the regional desertification simulated based on NPP in an embodiment;

[0018] Figure 8 It is a schematic diagram of the structure module showing the social and economic characteristics and land use in an embodiment;

[0019] Figure 9 It is a schematic diagram of the structure module showing the water resource system and land use in an embodiment;

[0020] Figure 10 Schematic diagram showing the detection results of the ROC curve of the land suitability probability model in an embodiment;

[0021] Figure 11 Schematic diagram showing the comparison between the simulated value and the actual value of NPP in an embodiment;

[0022] Figure 12 Schematic diagram showing the comparison between the simulated value and the actual value of NDVI in an embodiment;

[0023] Figure 13 Schematic diagram showing the comparison between the simulated value and the actual value of GDP in an embodiment;

[0024] Figure 14 Schematic diagram showing the comparison between the simulated value and the actual value of the total water consumption in an embodiment;

[0025] Figure 15 Structural block diagram of the simulation system for the evolution of vulnerability in sandy areas in an embodiment;

[0026] Figure 16 Schematic structural diagram of a computer device in an embodiment. Detailed implementation manners

[0027] The following further elaborates on the present application in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely used to explain the present application and are not used to limit the present application.

[0028] It should be noted that the terms used herein are only for describing the specific implementation manners and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should also be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the embodiments of the present application will be elaborated in detail below in conjunction with the accompanying drawings. However, those of ordinary skill in the art can understand that in the embodiments of the present application, many technical details are provided to help readers better understand the present application. However, even without these technical details and various changes and modifications based on the following embodiments, the technical solutions required to be protected by the present application can still be achieved. The labels of the following embodiments are for convenience of description and should not constitute any limitation to the specific implementation manners of the present application. The various embodiments can be combined with each other and cross-referenced on the premise of not conflicting with each other.

[0030] Figure 2The flowchart shows a method for simulating the evolution of vulnerability in sandy areas provided by an embodiment of this application. This method can be executed by the user space file server in a system as shown in Figure 1 and executed by the user space file server in the system shown in Figures 2 - 4 . As shown, this method may include the following steps:

[0031] S201. Obtain the natural environment data of the research area;

[0032] Among them, the natural environment data includes climate data, vegetation data, and water resource data;

[0033] S202. Set constraint conditions based on the natural environment data according to the social, economic, and ecological construction needs of the research area;

[0034] S203. Optimize the land use structure based on the constraint conditions to obtain the target quantity of various types of land to meet the land use quantity structure;

[0035] S204. Based on the determination of the land use quantity structure, simulate the changes in the vegetation, social economy, and water resource data development in the research area, and build a model;

[0036] S205. Calibrate, verify, and perform sensitivity analysis on key parameters of the model, continuously iterate and optimize the model, and promote the change of vulnerability assessment indicators to achieve the vulnerability simulation of the research area.

[0037] By obtaining the natural environment data of the research area, including climate data, vegetation data, and water resource data, simulating climate change based on the climate data, simulating vegetation change based on the vegetation data to measure the degree of land desertification, and simulating water resource change based on the water resource data to reflect the constraints of water resources on social and economic development. Since land use change is affected by both natural and human factors, land use can not only reflect the guiding role of the overall regional development goal in land resource allocation, but also reflect the demands and initiatives of micro-subjects. In addition, land use will also have a feedback effect on the simulation change of vulnerability in sandy areas, affecting vegetation conditions and economic development. Therefore, based on the natural environment data and combined with future development goals, set constraint conditions, optimize the quantity structure of land use, and comprehensively consider the demands of different subjects to achieve the optimization of spatial layout. Finally, based on the determination of the land use quantity structure, further simulate the changes in the vegetation, social economy, and water resource data development in the research area, build a model, and verify the model to achieve the accurate simulation of vulnerability in the research area.

[0038] The following will specifically elaborate on each step in detail:

[0039] Please refer to Figure 2 、 Figure 3As shown, step S201: Obtain the natural environment data of the research area; among them, the natural environment data includes climate data, vegetation data, and water resource data;

[0040] In this embodiment, it should be noted that to obtain the natural environment data of the research area, climate change scenarios are simulated based on the climate data to determine future climate parameters such as temperature, precipitation, and solar radiation. The vegetation data is used to simulate land changes under the combined influence of climate change and human activities to measure the degree of land desertification. Water resources are an important issue for promoting the sustainable development of desert areas. Obtaining water resource data and simulating water resource changes can reflect the constraints of water resources on social and economic development, so as to provide a data basis for subsequent simulations.

[0041] Please continue to refer to Figures 2 - 4 As shown, step S202: Set constraint conditions based on the natural environment data according to the social and economic and ecological construction needs of the research area;

[0042] In this embodiment, it should be noted that according to the natural environment data, total area constraints, economic development constraints, food production constraints, ecological protection constraints, development trend constraints, reasonable regulation constraints, and non - negative constraints are set according to the social and economic and ecological construction needs of the research area; among them, the total area constraint means that the sum of the areas of various types of land is equal to the total land area of the research area; the economic development constraint means that the GDP growth rate reaches the expected target; the food production constraint means that the comprehensive food production capacity reaches the expected target; the ecological protection constraint means that the forest land area meets the forest coverage rate requirements; the development trend constraint means that the water area remains unchanged, the urban land and other construction land areas are greater than the current values, and the unused land area is less than the current value; the reasonable regulation constraint means that the upper and lower regulation thresholds are set according to the historical change trends of the areas of various types of land; the non - negative constraint means that the areas of various types of land meet the non - negative conditions. The constraint conditions are shown in Table 1:

[0043] Table 1: Optimization Constraint Conditions for the Quantity Structure of Land Use

[0044]

[0045] Refer to Figure 5 、 Figure 6 As shown, step S203: Optimize the land use structure based on the constraint conditions to meet the quantity structure of land use;

[0046] In this embodiment, it should be noted that in step S2031, first, divide the land use type X. Specifically, the land classification is mainly based on the Chinese land use / land cover remote sensing monitoring data classification system, and the land use type is divided into 8 categories as decision variables, including cultivated land x1, forest land x2, grassland x3, water area x4, urban land x5, rural residential areas x6, other construction land x7, and unused land x8. As shown in Table 2:

[0047] Table 2: Optimization Decision Variables of Land Use Quantity Structure

[0048]

[0049] Step S2032: Establish an objective function around the three aspects of economy, ecology, and water resources, which is described by formulas (1), (2), and (3):

[0050]

[0051] In the formula, Z1, Z2, and Z3 are the economic benefit, ecological benefit, and total water consumption respectively. EO k is the economic output per unit area of land use of type k, which is determined by dividing the total output value of the industry corresponding to land use of type k by the land area. ESV k is the ecosystem service value per unit area of land use of type k. WC k is the water consumption per unit area of land use of type k. WC else is the water consumption that is not affected by the land use structure. X k is the land area of type k.

[0052] In an example, according to the Inner Mongolia Autonomous Region's plan and target outline, in addition to pursuing high-quality economic development, the regional development goal should also be guided by ecological priority and green development. Considering the actual situation of water shortage in the Inner Mongolia Autonomous Region, taking the Inner Mongolia Autonomous Region as an example for the study area, aiming to promote economic growth, improve the ecological environment, and alleviate water resource contradictions in the Inner Mongolia Autonomous Region, an objective function Z1, Z2, and Z3 is established around the three aspects of economy, ecology, and water resources to maximize economic benefits, maximize ecological benefits, and minimize total water consumption.

[0053] Step S2033: Then, based on the multi-objective optimization of economic benefit, ecological benefit, and total water consumption, the quantity of various types of land use is obtained. Specifically, combined with the natural environmental conditions of the Inner Mongolia Autonomous Region, the development goals and the constraint conditions of the future social and economic development plan and ecological construction needs are determined. The parameters in the constraint equations are determined according to different scenarios, and finally the quantity of various types of land use is obtained. In an example, the specific parameter settings are a population size of 100 and a maximum number of evolutionary generations of 200. The simulated binary crossover (SBX) operator is used. During the solution process, for the three objective functions of economy, ecology, and water resources, the solution is carried out under 7 constraint conditions. Since the various objectives often conflict with each other, there is generally no unique global optimal solution, but a set of multiple optimal solutions, that is, the Pareto optimal solution set. In actual planning decisions, an optimal solution is selected from all the solutions according to the preference for different objectives, and finally the target quantity of various types of land use is obtained.

[0054] Step S2034. Finally, based on the quantity of various land use targets, calculate the probability of a plot being converted into land use type K according to land use conversion, and conduct spatial allocation to meet the land use quantity structure. Among them, land use conversion includes urban expansion conversion, land consolidation conversion, and artificial greening conversion. Urban expansion converts into urban land, rural settlements, and other construction land occupying cultivated land, forest land, grassland, and unused land. Land consolidation converts rural settlements vacated into cultivated land, forest land, and grassland. Artificial greening converts cultivated land, forest land, grassland, and unused land according to the adaptive inertia competition mechanism, and after multiple iterations until the expected value is reached. In this embodiment, three types of land use conversion methods are specifically set and executed in sequence, that is, first urban expansion, then land consolidation, and finally artificial greening. After multiple iterative processes until the expected value is reached to meet the land use quantity structure.

[0055] Specifically, as Figure 6 shown, the land use spatial change simulation presets two types of entities, including the government entity and the department entity, and the two entities jointly decide on land use conversion. The department entity is used to submit a land use application to the government entity to guide the layout behavior of different land uses, calculate the probability of a plot being selected by the department entity to initially select the location of a certain type of land use. There are three types of departments preset in the department entity, namely the economic department, the optimization department, and the ecological department. The economic department is responsible for the urban expansion conversion of land use conversion, the optimization department is responsible for the land consolidation conversion of land use conversion, and the ecological department is responsible for the artificial greening conversion. The government entity is used to determine the final land use plan, calculate the probability that the government finally accepts the conversion of a plot into land use type k according to the Monte Carlo method, and conduct conversion from high to low according to the final probability until the land use demand is met. After selecting the land use type with the highest probability and allocating it to the plot, if the quantity of various land uses is inconsistent with the target quantity, the department entity introduces the adaptive inertia competition mechanism for spatial allocation to resolve the competition relationship of different land uses, constructs a land use module, and realizes the optimal land use quantity structure.

[0056] In this embodiment, it should be noted that in the land use change spatial simulation, two types of entities, the government and the department entity, are preset, and the two entities jointly decide on urban expansion, land consolidation, and artificial greening. Specifically, the government entity is an abstraction of the local government, which plays a macro-allocation role in land resources and is responsible for approving land expropriation and land transfer. The department entity is an abstraction of the departments under the local government, which is responsible for initially selecting the location of a certain type of land use, putting forward a land use plan and submitting a land use application to the government entity, and then the government entity conducts land use review and determines the final land use plan. There are three types of departments set in the department entity, namely the economic department, the optimization department, and the ecological department, and different departments are responsible for guiding the layout behavior of different land uses.

[0057] In one example, the department entity is used to submit a land use application to the government entity to guide the layout behavior of different land uses, and calculate the probability that a plot is selected by the department entity to preliminarily select the location of a certain type of land use. Specifically: the economic department is responsible for the urban expansion conversion of land use conversion, and the optimization department is responsible for the land consolidation conversion of land use conversion. The two types of department entities calculate the probability that a plot unit is selected using a discrete choice model considering the comprehensive suitability probability, neighborhood influence, and conversion cost, which is described by formulas (4) and (5):

[0058] P k (i, j) = exp(TP k (i, j)) / exp(TP k (i, j)) (4),

[0059] TP k (i, j) = S k (i, j) × Ω k (i, j) × (1 - C k (i, j)) (5),

[0060] In the formula, Pk(i, j) is the probability that the department entity selects the plot unit, TPk(i, j) is the comprehensive probability that the plot unit is converted into k-type land use, Sk(i, j) is the suitability probability that the plot unit is converted into k-type land use, Ωk(i, j) is the neighborhood influence; Ck(i, j) is the conversion cost for the plot unit to be converted into k-type land use.

[0061] For the government entity to determine the final land use plan, calculate the probability that the government finally accepts the conversion of the plot into k-type land use according to the Monte Carlo method, and perform the conversion from high to low according to the final probability until the land use demand is met. Specifically: after determining the probability that the plot is selected by the department entity, use the Monte Carlo method for random sampling to simulate the department entity submitting the land use plan; the government entity accepts the land use application of the department entity and conducts a review. The more times the plot is applied for by the department entity, the higher the probability of its acceptance, and at the same time, the probability of acceptance of its neighboring plots will also increase. After reaching the sampling times, calculate the probability that the government entity finally accepts the conversion of the plot into k-type land use, and perform the conversion from high to low according to the final probability until the land use demand is met, which is described by the formula: P gov_k (i, j) = P k (i, j) + g × ΔP1 + h × ΔP2 (6),

[0062] P gov_k (i, j) ≤ 1 (7),

[0063] Wherein, Pk(i, j) is the probability that the department entity selects the plot unit, g is the number of times the plot unit is applied for, ΔP1 is the increased acceptance probability each time the plot is applied for, h is the number of times the plots within the neighborhood range are applied for, and ΔP2 is the increased acceptance probability each time the plots within the neighborhood range are applied for.

[0064] In addition, after allocating the land use type with the highest selection probability to the plot, if the quantity of each type of land use is inconsistent with the target quantity, the department entity introduces an adaptive inertia competition mechanism for spatial allocation to resolve the competition relationship among different land uses and meet the land use quantity structure. Specifically: in each update iteration when calculating the probability that the government finally accepts the conversion of the plot to the k-th type of land use, the ecological department selects the land use type with the highest probability based on the comprehensive probability of the plot unit converting to cultivated land, forest land, grassland, and unused land and allocates it to the plot; if the quantity of each type of land use is inconsistent with the target quantity, the adaptive inertia coefficient is adjusted to increase or decrease the comprehensive probability of the plot converting to a certain type of land use, and the next iteration is carried out to achieve the optimal land use quantity structure; among them, the adaptive inertia coefficient is determined by the difference between the actual quantity and the target quantity of each current type of land use and is described by formula (8):

[0065]

[0066] Wherein, Intertia t,k is the adaptive inertia coefficient of the k-th type of land use in the t-th iteration, Num k is the target quantity of the k-th type of land use, Num t-1,k is the allocated quantity of the k-th type of land use in the t-th iteration; after adding the adaptive inertia mechanism, the comprehensive probability that the plot unit converts to the k-th type of land use in the t-th iteration is obtained and is described by formula (9):

[0067] TP t，k (i, j) = S k (i, j) × Ω k (i, j) × (1 - C k (i, j)) × Intertia t，k (9).

[0068] Referring to Figures 2 - 4 shown, in step S204, based on the determination of the land use quantity structure, change simulation is carried out on the vegetation, social economy, and water resource data development of the research area, and a model is constructed;

[0069] First, the net primary productivity (NPP) of vegetation is selected as an index to reflect the vegetation condition and used to measure the degree of land desertification. The effects of climate change, grazing, land use, and natural grassland restoration factors on the vegetation NPP are selected and quantified, and a vegetation - soil module is constructed to simulate the dynamic evolution of desertification in the research area;

[0070] As shown Figure 7 in this embodiment, it should be noted that the vegetation-soil module is used to simulate the changes in the desert ecosystem under the combined influence of climate change and human activities. Here, NPP is selected as an indicator to reflect the vegetation status and used to measure the degree of land desertification. Since the driving forces of desertification include natural factors such as temperature, precipitation, soil texture, etc., and human factors such as grazing, land use, ecological protection policies, etc., considering the desertification evolution and driving process comprehensively, factors such as climate change, grazing, land use, and natural grassland restoration are selected. On the basis of quantifying the effects of these factors on vegetation NPP, the dynamic evolution of regional desertification is simulated. Specifically, on the basis of the actual NPP of the previous year, the changes in vegetation NPP caused by each driving factor are summarized to obtain the actual NPP of the current year, which is described by formula (10):

[0071] NPP t = NPP t-1 + ΔNPP t，c + ΔNPP t，l + NPP t，nr - NPP t，gr (10)

[0072] In the formula, NPP t is the actual NPP in the t-th year; NPP t-1 is the actual NPP in the (t - 1)-th year; ΔNPP t,c is the change in NPP caused by climate change, that is, the difference in the potential NPP (PNPP) of vegetation between the t-th year and the (t - 1)-th year considering only natural factors such as climate and soil water; ΔNPP t,l is the change in NPP caused by land use; NPP t,nr is the change in NPP caused by natural grassland restoration; NPP t,gr is the change in grassland NPP caused by grazing.

[0073] The following explains the specific calculation processes of ΔNPP t,c , ΔNPP t,l , NPP t,nr and NPP t,gr respectively:

[0074] For ΔNPP t,c, since the potential vegetation net primary productivity (PNPP) refers to the net primary productivity of vegetation under the influence of only natural factors such as climate and soil without human interference. Specifically, vegetation PNPP is a relative concept. That is, in the land use state of the previous year, from the dynamic processes of climate - soil water - vegetation PNPP, the changes in temperature, precipitation, solar radiation and the resulting changes in evapotranspiration will affect the increase or decrease of soil moisture, and both together lead to the change of vegetation PNPP in the natural state. Here, the CASA model is used to estimate PNPP, and the calculation formula is as follows:

[0075] PNPP = FRAR × ε max × sol(t) × T ε1 (t) × T ε2 (t) × W ε (t) × 0.5 (11)

[0076] In the formula, PNPP is the potential vegetation productivity; FRAR is the percentage of incident photosynthetically active radiation absorbed by the vegetation canopy; ε max is the maximum light use efficiency; sol(t) represents the total solar radiation in month t (MJ / m 2 / month); T ε1 (t), T ε2 (t) is the monthly temperature stress influence coefficient; W ε (t) is the soil moisture stress influence coefficient. The monthly FPAR and the maximum light use efficiency ε max are determined by the land use type.

[0077] For ΔNPP t,l , since land use change will have an important impact on the vegetation NPP and desertification evolution in the study area, there are obvious differences in the NPP of different land use types. Based on the average NPP values of various land uses, the difference between the average NPP values of the land use type in the previous year and the current year is calculated, which is the change in NPP caused by land use change.

[0078] For NPP t,nr , the change in NPP caused by the natural restoration of grassland is calculated using the classical forage consumption - growth model. The specific formula is as follows:

[0079]

[0080] In the formula, NPP t,nr is the change in NPP caused by the natural restoration of grassland; r is the intrinsic growth rate of the grassland. Here, this parameter specifically refers to the natural growth rate of the grassland under certain climate conditions, which is determined by the ratio of the potential NPP of the grassland in the current year and the previous year; NPP t-1 is the actual NPP in the (t - 1)th year; NPP maxIt is the maximum value of grassland NPP in the study area.

[0081] For NPP t,gr , the change in grassland NPP caused by grazing is affected by the grazing ban policy. In the designated grazing ban areas, grazing activities are prohibited on the grassland. In the areas with grassland-livestock balance, referring to previous literature, without damaging the natural restoration ability of the grassland, the maximum NPP value that can be consumed by grazing is set to 50% of the grassland NPP. The specific formula is as follows:

[0082] NPP t，gr = 50% × (NPP t-1 + ΔNPP t，c + NPP t，nr ) (13)

[0083] In the formula, NPP t,gr is the change in grassland NPP caused by grazing; NPP t-1 is the actual NPP in the (t - 1)th year; ΔNPP t,c is the change in NPP caused by climate change; NPP t,nr is the change in NPP caused by the natural restoration of the grassland.

[0084] According to the above description, after obtaining the actual NPP value of the current year, according to the NPP threshold system of different desertification levels, the degree of land desertification is divided into 5 levels, namely non-desertification (> 150 gC / m 2 / a), mild desertification (115 - 150 gC / m 2 / a), moderate desertification (70 - 115 C / m 2 / a), severe desertification (25 - 70 gC / m 2 / a), and extremely severe desertification (< 25 gC / m 2 / a). And compare the obtained actual NPP value of the current year with the NPP threshold system of desertification levels to determine the current desertification level.

[0085] Secondly, social economy is closely related to land use. As Figure 8 shown, according to the basic characteristics of the social and economic system in the study area and its correlation with the evolution of vulnerability, the concept of social economy is divided into four parts: population change, economic development, technological progress, and income growth. By simulating the key variables of the total population, total GDP, and per capita income of farmers and herdsmen, a social and economic module is constructed to simulate the dynamic evolution of desertification in the study area;

[0086] In one example, taking the 12 leagues and cities under the jurisdiction of Inner Mongolia Autonomous Region as an example, its social and economic module is divided into four parts: population change, economic development, technological progress, and income growth. By means of this module, key variables of the system such as the total population, the total GDP, and the per capita income of farmers and herdsmen are simulated, and the dynamic evolution of desertification in Inner Mongolia Autonomous Region is studied.

[0087] Among them, the population change simulation is based on the future provincial population and high-resolution (1km) gridded population database of China under five shared socioeconomic pathways from 2010 to 2100. A multi-dimensional recursive model is used for this data. By setting parameters such as the fertility rate, mortality rate, migration rate, and education level, and comprehensively considering the impact of national policy adjustments, the natural hierarchical movement of the newly born population and the population with different age structures is simulated, and the mutual conversion between multiple education levels is realized. Specifically, the number of births in a certain year is determined by the corresponding population divided by age and education level in the previous year, the fertility rate of women of childbearing age, and the sex ratio at birth. The calculation formula is as follows:

[0088]

[0089] In the formula, Pm and Pf are the male and female populations respectively; FER is the fertility rate; Bm and Bf are the proportions of male and female births respectively; t is a certain year between 2011 and 2100; a is a certain age group from 0 to 99 years old, and also includes the age group of 100 years old and above; edu is the seven education stages in the Chinese population census. The formula for recursively calculating the population with different ages and education levels is as follows:

[0090] Pm t，a+1，edu =(Pm t-1，a，edu ×(1 - MORm t，a，edu )×(1 + NetPIM t，a，edu )×(1 - Gm t，a，edu ) + Pm t-1，a，edu-1 ×(1 - MORm t，a，edu-1 )×(1 + NetPIM t，a，edu-1 )×Gm t，a，edu-1 )×(1 + NetGIM t ) (16)

[0091] Pf t，a+1，edu =(Pf t-1，a，edu ×(1 - MORf t，a，edu )×(1 + NetPIM t，a，edu )×(1 - Gf t，a，edu ) + Pf t-1，a，edu-1 ×(1 - MORf t，a，edu-1 )×(1 + NetPIM t，a，edu-1 )×Gf t，a，edu-1 )×(1 + NetGIMt ) (17)

[0092] Where MORm and MORf are the mortality rates of males and females respectively; NetPIM is the provincial net immigration rate; Gm and Gf are the enrollment rates between different education levels of males and females respectively; NetGIM is China's global net immigration rate. The total regional population is further divided into urban population and rural population through the urbanization rate. Since the growth of the urbanization rate conforms to the S-shaped curve, that is, the urbanization rate is relatively slow in the initial stage, accelerates in the middle stage, and approaches saturation in the later stage, a sigmoid function is used to predict the change of the urbanization rate over time. The specific formula is as follows:

[0093]

[0094] Where PU is the population urbanization rate; T is the predicted time span (the predicted year minus the base year); b is the upper limit value of the urbanization rate; c is the rate of change of the urbanization rate over time; d is the time when the urbanization rate reaches the inflection point of half of the curve. Finally, the predicted urban population and rural population are downscaled to a 1km population grid according to the proportion data of the rural population and urban population.

[0095] The economic development situation is mainly reflected by the total regional production (GDP) and the added value of each industry. The added value of each industry is determined by the corresponding land area and the economic output per unit area. Among them, the added value of the primary industry is the sum of the economic outputs of cultivated land, forest land, grassland and water areas. The added value of the secondary industry includes the added value of industry and the added value of construction. The added value of industry is the industrial economic output of other construction land, and the added value of construction is the construction economic output of all construction land. The added value of the tertiary industry is the sum of the tertiary industry economic outputs of urban land and rural settlements. In addition, considering that in addition to the increase in output caused by the input of land factors, there are other factors that contribute to the growth of output. Here, the Cobb-Douglas production function is used to fit the added value of each industry. According to the needs of this study, capital, labor, land resources and scientific and technological input are mainly considered. The formula is as follows:

[0096] Y = AK a L b S γ e λt (19),

[0097] Where Y is the added value of each industry, K is the capital input, represented by the capital stock of each industry per year; L is the labor input, represented by the number of people engaged in each industry per year; S is the land resource input, represented by the corresponding land area; λ is the scientific and technological contribution rate, which changes over time and affects the factor output level.

[0098] Regarding technological progress, since regional economic growth will lead to an increase in R & D investment, resulting in technological progress. Due to economic agglomeration, there will be economies of scale effects. Therefore, there is a positive correlation between total factor productivity growth and output growth, and economic development will improve the utilization efficiency of land factors. At the same time, the nature of capital's pursuit of profit and the role of the market mechanism will prompt resources to flow and transfer to industries with high economic efficiency. The continuous innovation and upgrading of industries, combined with the accumulation of fixed asset investment, will have an impact on land output rate. Here, a logarithmic function is used to quantify the feedback effect of regional economic growth on the economic output per unit area of land. The formula is as follows:

[0099] ln EO t，k =b1 + k1×ln GDP t-1 (20)

[0100] In the formula, EO t，k is the economic output per unit area of the k - type land in the t - th year; GDP t-1 is the total GDP in the (t - 1) - th year. On the other hand, the improvement of technological level can improve water use efficiency by improving production equipment, using water - saving irrigation equipment, popularizing water - saving appliances, etc., so as to reduce the water consumption per unit area of farmland irrigation and the water consumption per unit industrial added value, and achieve the goal of water - saving production and green development. Research shows that the relationship between industrial water intensity and economic growth may show various forms such as an inverted "N" shape, a "U" shape, and a monotonically decreasing shape, and the change direction is uncertain. Therefore, this application also sets the feedback relationship of regional economic growth on the water consumption per unit area of farmland irrigation and the water consumption per unit industrial added value as follows:

[0101] WC t，p =k1×exp(b1 + k2×GDP t-1 ) (21)

[0102] WC t，p =b1 + k1×ln GDP t-1 (22)

[0103] In the formula, WC t,p is the water consumption per unit area of farmland irrigation or the water consumption per unit industrial added value in the t - th year; GDP t-1 is the total GDP of the previous year.

[0104] Regarding income growth, since economic development has a profound impact on agricultural product demand and prices, farmers' employment and wages, etc., the prospects for farmers' income increase are highly correlated with the fundamentals of the macroeconomic trend. The development of modern agriculture increases the labor productivity through specialized and large-scale operations, intensifies capital and integrates technologies to improve the land output rate, thereby increasing the agricultural income of farmers and herdsmen. At the same time, along with the process of urbanization, the rural labor force continues to transfer to cities and towns, and farmers achieve high-quality and full employment in the non-agricultural industries. Wage income has already occupied a dominant position in farmers' income, and the rapid development of the secondary and tertiary industries has become the key driving force for promoting farmers' income increase. This application assumes that macroeconomic growth has a positive effect on farmers' income growth, and the equation is set as follows:

[0105] ln Income=b1+k1×ln GDP (23)

[0106] In the formula, Income is the per capita net income of farmers and herdsmen; GDP is the gross national economic product of the corresponding year.

[0107] Income growth will drive the improvement of the social investment capacity and consumption capacity. Farmers and herdsmen will increase the input of agricultural factors such as the amount of chemical fertilizer applied and the total power of agricultural machinery. The effective use of science and technology in the agricultural production process helps the rational allocation of various agricultural production factors and the restoration and improvement of soil fertility, laying a foundation for the improvement of agricultural production conditions and the efficient output of agricultural products, thereby comprehensively improving agricultural productivity and agricultural economic efficiency. The main formulas are as follows:

[0108] ln FC=b1+k1×ln GDP+k2×ln x1 (24),

[0109] ln AM=b1+k1×ln GDP+k2×ln x1 (25),

[0110] ln IA=b1+k1×ln GDP+k2×ln x1 (26),

[0111] In the formula, FC, AM, and IA are the amount of chemical fertilizer applied, the total power of agricultural machinery, and the effective irrigation area respectively, GDP is the gross national economic product of the corresponding year, and x1 is the cultivated land area.

[0112] Then, as Figure 9 shown, according to the entire water resource system of the research area, it is divided into four parts: domestic water use, production water use, ecological water use, and available water resources. The water resource allocation is correlated with land use. Taking the water resource supply-demand ratio as the core variable of the module, a water resource module is constructed to simulate the dynamic evolution of desertification in the research area;

[0113] In this embodiment, it should be noted that according to the characteristics of the water use structure in Inner Mongolia Autonomous Region, the entire water resources system is divided into four parts: domestic water use, production water use, ecological water use, and available water resources, and the water resources supply-demand ratio is used as the core variable of the module.

[0114] For domestic water use, domestic water use is the sum of urban domestic water use and rural domestic water use. Population change is the key factor determining urban and rural domestic water use, and the influence of urban and rural populations on domestic water use is further distinguished through the urbanization rate. Urban domestic water use is determined by the urban population and the per capita domestic water use of urban residents, and rural domestic water use is determined by the rural population and the per capita domestic water use of rural residents. There are many factors affecting domestic water use. With the acceleration of the urbanization process, the per capita housing area increases, the living standards of the people improve, and high-water-consuming equipment such as showers, sanitation, and kitchens is generally configured, resulting in an increase in the demand for residential water. On the other hand, the popularization of water-saving appliances, the promotion of water-saving policies, and the enhancement of water-saving awareness and environmental protection awareness will all reduce the per capita domestic water use. Considering that the domestic water use structure and influencing factors of residents are relatively complex and show the characteristics of non-linear changes, the following formula is used to establish the feedback relationship between regional economic growth and the per capita social water use of residents:

[0115] WC t，d = k1×exp(b1 + k2×GDP t-1 ) (27)

[0116] WC t，p = b1 + k1×ln GDP t-1 (28)

[0117] In the formula, WC t，d is the per capita water use of urban or rural residents in the t-th year; GDP t-1 is the total GDP of the previous year.

[0118] For production water use, production water use refers to the amount of water resources required to meet the development of regional industries, which is the sum of the water use of the primary industry, secondary industry, and tertiary industry. The water use of the primary industry includes farmland irrigation water use and forestry, animal husbandry, and fishery water use. Farmland irrigation water use is determined by the irrigation area and the water use per unit area of farmland irrigation. The water use of forestry, animal husbandry, and fishery includes fruit and forest irrigation water use, grassland irrigation water use, fish pond make-up water, and livestock water use, which is determined by the sum of the added value of the forestry, animal husbandry, and fishery industries and the water use of the added value of the forestry, animal husbandry, and fishery industries. Industrial water use and urban public water use are closely related to industrial economic development and the water use per unit industrial added value, and are determined by the corresponding industrial added value and the water use per unit industrial added value. The water use of each industry is related to the local technical conditions. The improvement of irrigation and production technology levels can effectively reduce production water use. By establishing the correlation with the total GDP, the change of the water use per unit area of farmland irrigation and the water use per unit added value of each industry in the future can be predicted.

[0119] For ecological water use, ecological water use is the water resources that must be consumed through artificial control and management to maintain the ecological environment from deteriorating further and gradually improve. According to the actual water use situation in Inner Mongolia Autonomous Region, ecological water use only includes the urban environmental water supplied by artificial measures. The urban environmental water consumption is affected by the urban ecological construction level and is determined by the green area and the water consumption for landscaping. Urban construction land is the basis for urban green space construction. The expansion of the scale of urban construction land determines the potential for urban green space construction. In this study, the green space rate is used as an indicator to measure the relationship between the area of garden green space and the area of urban construction land.

[0120] For available water resources, available water resources are the sum of surface water resources, groundwater resources, and the amount of recycled sewage. Among them, surface water resources and groundwater resources mainly come from precipitation and are also affected by the amount of surface evapotranspiration. In this study, two main climate elements, temperature and precipitation, which affect the amount of natural water resources, are selected, and an artificial neural network algorithm is used to establish a non-linear relationship between them for simulating the changes in future surface water resources and groundwater resources. In addition to natural water resources, some domestic sewage and industrial sewage can be recycled through treatment and used as part of the water resources supply. A linear regression equation is established to describe the relationship between the amount of recycled sewage and time:

[0121] SR t = b1 + k1×t (29),

[0122] In the formula, SR t is the amount of recycled sewage in the t-th year.

[0123] Finally, as Figure 4 shown, a vulnerability evolution model for sandy areas is constructed based on the land use module, vegetation soil module, social economy module, and water resources module to simulate the vulnerability evolution of sandy areas.

[0124] In addition, the simulation method for the evolution of vulnerability in sandy areas also includes optimizing each module of the model after it is constructed. Specifically: for the land use module, optimize its quantity structure and the spatial layout of land use respectively. The optimization of the land quantity structure is to use a multi-objective optimization algorithm to solve the optimal land use quantity structure under different scenarios on the basis of setting the objective function and constraints. In multi-objective linear programming, the various objectives often conflict with each other, and generally there is no unique global optimal solution, but there is a set of multiple optimal solutions, that is, the Pareto optimal solution set. Specifically, in an example, the non-dominated sorting genetic algorithm NSGA-Ⅲ is used to solve the Pareto optimal solution set for the optimization of the land use quantity structure. The non-dominated sorting genetic algorithm (Non-dom nated Soring Genetic Agonthms, NSGA) is one of the most classic multi-objective optimization algorithms. On the basis of the NSGA algorithm, two algorithms, NSGA-Ⅱ and NSGA-Ⅲ, have been updated. Here, NSGA-Ⅲ is used to solve the optimal solution. Its core idea is to search for evenly distributed Pareto optimal solutions in the solution space, which ensures the diversity of the population and reduces the risk of falling into local optima; in addition, NSGA-Ⅲ introduces a reference point in the objective space, and on the basis of reducing the sorting complexity of NSGA-Ⅱ, improves the optimization effect for multi-objective problems with the number of objectives greater than or equal to 3. In actual planning decisions, it is necessary to select an optimal solution from all solutions according to the preference for different objectives. Therefore, it is necessary to set the weights of the objective function for different scenarios and calculate the comprehensive benefits. To avoid the influence of the dimension inconsistency of the objective function and the absolute value size on the estimation of the comprehensive benefits, first normalize the values, then sum them with weights and select the land use optimization plan with the largest comprehensive benefits under different scenarios.

[0125] Regarding the optimization of land use spatial layout, in the spatial simulation of land use change, departmental entities need to calculate the probability of a plot unit being selected using a discrete choice model based on the comprehensive probability of the plot unit being converted into a certain land use type. Specifically, cellular automata (CA) is used to calculate the probability of a plot unit converting its land use type. Since cellular automata is a grid dynamics model with discrete time, space, and state, and local spatial interaction and time causality, it has the ability to simulate the spatio-temporal evolution process of complex systems. The FLUS model is a land use change simulation model based on cellular automata (CA). By combining the artificial neural network (ANN) algorithm and the adaptive inertia and competition mechanism, the traditional CA model is improved to better simulate various land use scenarios. Its basic principle is to first calculate the suitability probability of various land uses based on driving factors using the neural network algorithm, then combine neighborhood influence and conversion costs, and introduce an adaptive inertia competition mechanism to solve the competition relationship during different land use conversions, obtaining the overall conversion probability of plot units. Finally, the simulation results are obtained through the roulette wheel mechanism during CA iteration.

[0126] In this study, a BP (back propagation) neural network is used to establish a model for generating suitability probabilities. The BP neural network is a multi-layer feedforward neural network trained according to the error backpropagation algorithm. Its basic components include an input layer, a hidden layer, and an output layer. Specifically, the input layer is the evaluation indicators for land use suitability, and the output layer is the suitability probabilities of various land uses. In the input layer, 12 evaluation indicators are selected, namely elevation, slope, surface soil texture, organic matter content, distance to water area, distance to the center of the league, distance to the center of villages and towns, distance to the railway station, distance to the road, distance to hospitals and clinics, distance to parks, and distance to schools. In the hidden layer, the number of neurons is required to be at least 2 / 3 of the number of neurons in the input layer. Therefore, the number of neurons in the hidden layer is set to 8, and the connection between the hidden layer and the output layer is established by the Sigmoid function. In an example, 10,000 plot units of each land use type are randomly selected as sample points, that is, a total of 80,000 sample points, and they are divided into a training set and a validation set according to a ratio of 7:3. Neighborhood influence is used to reflect the influence of the interaction between plot units within the neighborhood range on the conversion of land use types, and the calculation formula is as follows:

[0127]

[0128] In the formula, Ω k (i, j) is the neighborhood influence, Σ N×Ncon(c(i, j) = k) represents the total number of plots of k types of land within the N×N neighborhood range. Considering that the expansion ability of construction land is relatively strong while that of other types of land is relatively weak, a 25×25 Moore neighborhood is selected to calculate the influence of the construction land neighborhood, and a 7×7 Moore neighborhood is selected to calculate the influence of other types of land neighborhoods. The conversion cost represents the difficulty of converting the current land use type to other land use types. It is preset that there is no mutual conversion between various types of land and water areas, and urban land and other construction land cannot be converted to other types of land. According to the land use conversion matrix of the study area and referring to relevant research, the conversion cost matrix is set as shown in Table 3.

[0129] Table 3: Land Use Type Conversion Cost Matrix

[0130] Land use type Cultivated land Forest land Grassland Urban land Rural residential areas Other construction land Unused land Cultivated land 0 0.95 0.9 0.4 0.5 0.6 0.4 Forest land 0.8 0 0.9 0.6 0.7 0.8 0.8 Grassland 0.6 0.8 0 0.4 0.4 0.6 0.1 Urban land 1 1 1 0 1 1 1 Rural residential areas 0.95 0.99 0.95 0.99 0 0.99 0.95 Other construction land 1 1 1 1 1 0 1 Unused land 0.9 0.99 0.9 0.3 0.3 0.6 0

[0131] Note: The rows represent the future land use types, and the columns represent the current land use types.

[0132] For the optimization of the vegetation-soil module, specifically, calculating the potential NPP using the CASA model is a key process of the vegetation-soil module. The CASA model is a process-based remote sensing model that couples ecosystem productivity and soil carbon and nitrogen fluxes and is driven by gridded global climate, radiation, soil, and remote sensing vegetation index datasets. Since climate directly affects the vegetation growth process, on the basis of quantitatively evaluating the driving effect of climate change on vegetation growth without human interference, the impacts of grazing and land use on vegetation are superimposed to comprehensively reflect the driving process of natural and human factors on desertification evolution. Therefore, by setting the relative concept of potential NPP, which refers to an ideal state NPP that vegetation reaches under the influence of only climate in the land use state of the previous year, it is a relative reference reflecting the impact of climate change on vegetation growth. The calculation method of potential NPP is similar to that of the CASA model, except that the absorption percentage of incident photosynthetically active radiation by the vegetation canopy FRAR is not calculated by the normalized difference vegetation index (NDVI), but is set according to empirical values related to the land use type, and the calculation of other parts such as the light use efficiency is consistent with the CASA model. Table 4 is as follows:

[0133] Table 4: FPAR and Maximum Light Use Efficiency εmax for Different Months Corresponding to Different Land Use Types

[0134] Farmland Forest land Grassland Water area Construction land Unused land FPAR 1 0.023 0.026 0.022 0 0.022 0.021 FPAR 2 0.023 0.025 0.022 0 0.022 0.021 FPAR 3 0.024 0.026 0.023 0 0.022 0.021 FPAR 4 0.025 0.028 0.024 0 0.024 0.021 FPAR 5 0.037 0.047 0.032 0 0.030 0.023 FPAR 6 0.045 0.094 0.041 0 0.034 0.023 FPAR 7 0.054 0.144 0.06 0 0.036 0.024 FPAR 8 0.056 0.153 0.061 0 0.036 0.024 FPAR 9 0.048 0.090 0.049 0 0.034 0.024 FPAR 10 0.030 0.039 0.028 0 0.026 0.021 FPAR 11 0.025 0.028 0.024 0 0.023 0.021 FPAR 12 0.024 0.026 0.023 0 0.023 0.021 <![CDATA[ε max > 0.542 0.526 0.536 0 0.529 0.537

[0135] In addition, this application has also optimized the soil water module. The input parameters of soil water content include soil texture, field water holding capacity, precipitation, potential evapotranspiration, etc., and the simulation is carried out on a monthly time step. On this basis, the actual evapotranspiration is calculated more accurately according to the soil water content, rainfall, potential evapotranspiration, soil relative drying rate and soil wilting water content (i.e., the lower limit value of soil water content) of the previous month, and is further used to calculate the water stress coefficient. On the basis of quantifying the impact of climate change on vegetation NPP, the changes in NPP caused by natural restoration, grazing and land use are further calculated. After summarizing the driving effects of all factors, the dynamic simulation of vegetation NPP under the influence of climate change and human activities can be carried out. Specifically, the vegetation soil module is reconstructed through land stability assessment (ArcPy) to realize the development of the vegetation soil module. The livestock density is closely related to the desertification process, and considering the natural conditions and social and economic impacts comprehensively, the livestock density LD is simulated based on NPP and GDP. The specific formula is as follows:

[0136] ln LD = b1 + k1×ln NPP + k2×n NPP (31),

[0137] There is a correlation between NDVI and NPP, as well as between temperature, rainfall and radiation. On the basis of NPP calculation, an empirical equation established by referring to relevant research and combining the basic principles of grassland genesis and measured data is used to estimate the annual maximum NDVI of the study area. The formula is as follows:

[0138]

[0139] In the formula, T is the annual average temperature (°C) of the corresponding year, W is the annual precipitation (mm), and R is the annual radiation (MJ / m2).

[0140] There is a highly significant linear correlation between vegetation coverage and NDVI. Usually, by establishing the conversion relationship between the two, the vegetation coverage information is directly extracted. Therefore, the vegetation coverage is calculated according to the principle of the dichotomy of pixels. The formula is as follows:

[0141]

[0142] In the formula, NDVI veg is the NDVI value of the pure vegetation coverage unit, NDVI soil is the NDVI value of the bare soil covered pixel. Due to the influence of conditions such as the atmosphere, surface conditions, year, season and region, the NDVI veg , NDVI soil values vary with time and space. Here, on the basis of statistically analyzing the cumulative probability distribution histogram of NDVI in the study area, with the cumulative percentages of 5% and 95% as the confidence intervals, the corresponding pixel values are read, so as to respectively determine the effective NDVI vegand NDVI soil value

[0143] Vegetation water consumption can reflect the water consumption status of vegetation in a region, and can be calculated using potential evapotranspiration PE, vegetation water consumption coefficient KC, and soil water coefficient KS. The water consumption coefficient KC of specific vegetation is calculated according to the land use type and empirical parameters provided by the Food and Agriculture Organization of the United Nations (FAO). According to the local actual soil water content S, field capacity S * and wilting water content S w calculate the soil water coefficient KS. The specific formula is as follows:

[0144] ET = PE × KC × KS (34)

[0145] KS = ln[(S - S w ) / (S * - S w ) × 100 + 1] / ln 101 (35).

[0146] For the optimization of the social - economic module, since establishing the feedback relationship between regional economic growth and land economic output is the core content of the social - economic module, it is not only the key to linking the social - economic module and the land - use module, but also provides support for setting land economic output parameters in multi - objective land - use optimization. According to the GDP data of each league city in Inner Mongolia Autonomous Region from 2010 to 2020 and the calculated economic output data per unit area of various land uses, the least - squares method is used to fit the equation parameters using the set equation form. Taking cultivated land as an example, the feedback relationships of capital, labor, land, and science and technology inputs in each league city on the economic output per unit area of cultivated land obtained by fitting are shown in Table 5:

[0147] Table 5: Feedback relationships of regional factor inputs on land economic output - taking cultivated land as an example

[0148] League cities Equation Hohhot <![CDATA lnY = -6.63 + 1.19E-17t + 0.96lnK + 1.03lnL + 0.01lnS]]> Baotou <![CDATA lnY = -12.56 + 1.19E-18t + 0.29lnK + 0.01lnL + 1.70lnS]]> Hulunbuir <![CDATA lnY = -4.62 - 3.5E-18t + 0.49lnK + 1.40lnL + 0.10lnS]]> Xing'an League <![CDATA lnY = 1.56 + 0.08t + 0.05lnK + 0.05lnL + 0.09lnS]]> Tongliao <![CDATA lnY = 0.80 + 0.06t + 0.01lnK + 0.13lnL + 0.26lnS]]> Chifeng <![CDATA lnY = -15.56 + 0.05t + 0.01lnK + 0.01lnL + 1.98lnS]]> Xilingol League <![CDATA lnY = 1.45 + 0.03t + 0.18lnK + 0.01lnL + 0.01lnS]]> Ulanqab <![CDATA lnY = -8.89 + 0.20t + 0.01lnK + 1.90lnL + 0.09lnS]]> Ordos <![CDATA lnY = -3.36 + 0.09t + 0.17lnK + 0.84lnL + 0.19ln]]> Bayannur <![CDATA lnY = -5.71 - 5.7E-19t + 0.84lnK + 1.03lnL + 0.13lnS <!-- 14 -->]]> Wuhai <![CDATA lnY = -2.46 + 0.06t + 0.08lnK + 1.81lnL + 0.01lnS]]> Alxa League <![CDATA lnY = -3.54 + 0.08t + 0.01lnK + 1.84lnL + 0.15lnS]]>

[0149] Based on the quantitative feedback relationships of each element in the fitted social - economic module, the social - economic module is reconstructed using the Pandas package to realize the development of the social - economic module.

[0150] Regarding the optimization of the water resources module, establishing the interaction relationship between key variables in the water resources module and the social and economic module is the basis for simulating the change of water use for three types of production and the total water use. Establishing the feedback relationship between regional economic growth and water use is the core content. On the one hand, it realizes the coupling of the water resources module and the social and economic module. On the other hand, it supports the setting of water use parameters for various types of land use in the multi-objective optimization of land use. According to the GDP data of each league city in Inner Mongolia Autonomous Region from 2010 to 2020, and the calculated per capita water use of residents, water use per unit area of various types of land, and water use per unit added value of various industries, the least squares method is used to fit the equation parameters by using the set equation form. Taking the irrigation water use per unit area of farmland as an example, the feedback relationship between economic growth and water use in each league city obtained by fitting is shown in Table VI:

[0151] Table VI: Feedback Relationship between Regional Economic Growth and Water Use - Taking Irrigation Water Use per Unit Area of Farmland as an Example

[0152] League cities Equation Hohhot <![CDATA WC t ，1 = 2.87E-07 + 0.0011exp(-4.4E05GDP t-1 )]]> Baotou <![CDATA WC t ，1 = 0.0012 + 0.75exp(-0.0069GDPt-1)]]> Hulunbuir <![CDATA WC t ，1 = 1.19 + ln (2.41E-08GDP t-1 - 0.17)]]> Xing'an League <![CDATA WC t ，1 = 1.22 + ln (2.68E-07GDPt-1 - 0.20)]]> Tongliao <![CDATA WC t ，1 = 0.0012 + 0.0011 exp(-4.4E05 GDP t-1 )]]> Chifeng <![CDATA WC t ，1 = 1.13 + ln (6.78E-08GDPt-1 - 0.13)]]> Xilingol League <![CDATA WC t ，1 = 1.16 + ln (8.18E-09GDPt-1 - 0.15)]]> Ulanqab <![CDATA WC t ，1 = 3.46E-05 + 0.00033 exp(-0.00072 GDP t-1 )]]> Ordos <![CDATA WC t ，1 = 0.0021 + 0.0048exp(-0.0017GDP t-1 )]]> Bayannur <![CDATA WC t ，1 = 0.97 + ln (1.12E-06GDPt-1 + 0.035)]]> Wuhai <![CDATA WC t ，1 = 0.0039 + 0.0080exp(-0.0078GDP t-1 )]]> Alxa League <![CDATA WC t ，1 = 0.00050 + 0.55exp(-0.035GDP t-1 )]]>

[0153] Based on the quantitative relationship between the elements of the water resources module and the social and economic module obtained by fitting, the water resources module is reconstructed using the Pandas package. Among them, Pandas is a tool based on NumPy, which is created to solve data analysis tasks. Pandas incorporates a large number of libraries and some standard data models, providing the tools required to efficiently operate large datasets. Here, the Pandas package is used to reconstruct the water resources module to realize the development of the water resources module.

[0154] As Figures 2 - 4 shown, in step S205, the model is calibrated, verified, and the sensitivity analysis of key parameters is carried out, and the model is continuously iteratively optimized to promote the change of vulnerability assessment indicators to realize the vulnerability simulation of the research area;

[0155] First, based on the land use module, the receiver operating characteristic curve (ROC) and the area under the receiver operating characteristic curve (AUC value) are used to test the land use module to evaluate the simulation accuracy. In this embodiment, it should be noted that after the neural network model is established, the ROC curve and the AUC value are used to test the performance of the neural network model. The ROC curve detection is a detection method for quantitatively evaluating the performance of a classification model, which is suitable for the accuracy test of generating a land use suitability probability model. The AUC value is the area under the ROC curve. Among them, the closer the AUC value is to 1.0, the better the classification effect of the model. When the AUC value is greater than 0.7, the classification result has a certain accuracy and can pass the test; As Figure 10, The ROC curve detection results show that the AUC values of various land uses are all greater than 0.8, and the prediction effect of the neural network model is relatively ideal. Generally speaking, the simulation accuracy of various land uses can reach a relatively high level, and the spatial distribution pattern of land use in the study area has been relatively stable. Therefore, it is considered that the model and corresponding model parameters can be used to simulate and optimize the future spatial layout of land use in the study area.

[0156] Secondly, based on the vegetation-soil module, the effectiveness of the simulated NPP in the study area is verified. The simulated NPP results and actual NPP data at specific times are respectively extracted, and the correlation coefficient is used for comparison and verification to evaluate the simulation accuracy; based on the social-economic module, the simulated values and real values of the added value of each industry and GDP are respectively compared to evaluate the simulation accuracy.

[0157] In this embodiment, it should be noted that taking 2010 as the benchmark, the NPP of Inner Mongolia Autonomous Region from 2011 to 2020 is simulated, and the effectiveness verification is carried out in combination with the simulation results in 2020. Specifically, 1000 verification points are randomly selected, and the simulated NPP results and actual NPP data in 2020 are respectively extracted, and the correlation coefficient is used for comparison and verification. As Figure 11 shown, from the results, the correlation coefficient between the two is 0.976, indicating that the simulation accuracy is relatively high, reflecting the good simulation effect of the vegetation-soil module. As Figure 12 shown, the NDVI data estimated by NPP also achieves good accuracy, and the correlation coefficient with the actual NDVI data is 0.908, which can be used for further calculation of subsequent model indicators.

[0158] Then, based on the social-economic module, the simulated values and real values of the added value of each industry and GDP are respectively compared to evaluate the simulation accuracy.

[0159] In this embodiment, it should be noted that taking 2010 as the benchmark, using the GDP data in 2010 and the interpolated land use quantity structure data from 2011 to 2020, the added value of each industry and GDP from 2011 to 2020 are simulated. As Figure 13 shown, taking GDP as the verification variable, the simulated value is compared with the actual value to evaluate the simulation accuracy. After testing, it is found that the simulation results are in good agreement with the historical data, and the errors in each year do not exceed 10%, and the average error is 2.14%. In addition, the average errors of the simulated added value of the primary, secondary, and tertiary industries, other key variables, are 5.97%, 3.24%, and 2.92% respectively. Generally speaking, the effectiveness of the feedback relationship established in the social-economic module is good and can be used to simulate the future changes of various elements in this module.

[0160] Finally, based on the water resources module, the per capita water consumption of residents, the water consumption per unit area of various land uses, and the water consumption of added value of each industry corresponding to the GDP data within a specific time period are simulated, and the simulated values are compared with the actual values to evaluate the simulation accuracy.

[0161] In this embodiment, it should be noted that taking 2010 as the benchmark, the per capita water consumption of residents, the water consumption per unit area of various land uses, and the water consumption of added value of each industry from 2011 to 2020 are simulated using the GDP data from 2010 to 2019. Combining the population data, the added value data of each industry, and the land use quantity structure data obtained by interpolation, the water consumption for three types of livelihoods and the total water consumption from 2011 to 2020 are simulated. As Figure 14 shown, taking the total water consumption as the verification variable, the simulated values are compared with the actual values to evaluate the simulation accuracy. The test results show that the simulated values are basically in line with the actual values, with an average error of 3.99%. However, there are relatively large errors in individual years in some leagues and cities, ranging from 10% to 15%. In addition, the average errors of other key variables such as domestic water, production water, and ecological water simulation are 7.21%, 5.94%, and 15.48% respectively. Generally speaking, the simulation accuracy of each variable in the water resources module is slightly lower, mainly because the annual fluctuation range of the water consumption for three types of livelihoods in Inner Mongolia is relatively large, especially the ecological water consumption, and the gap between adjacent years can even reach dozens of times, resulting in a serious increase in the fitting difficulty. Considering that there is no obvious change rule in the water consumption of some leagues and cities in Inner Mongolia, it is considered that the simulation error is acceptable, the simulation results basically conform to the actual situation, and the quantitative relationship between the elements of the water resources module and the social and economic module can be used to simulate the future changes in water consumption.

[0162] Implementation principle of this embodiment: The above steps mainly rely on the most important environmental characteristics of the sandy area, namely the arid and less rainy climate. The climate affects the growth of vegetation, directly impacts the regional dry-wet conditions and wind erosion intensity. Human activities are also one of the driving forces of desertification. In addition, the vegetation condition is an important indicator reflecting the degree of land desertification, and the sandy area generally faces water shortage problems. How to utilize limited water resources to promote the sustainable development of the sandy area is an important issue. Therefore, by obtaining the natural data of the research area, namely climate data, vegetation data, and water resource data, simulating climate change based on the climate data, simulating vegetation changes based on the vegetation data to measure the degree of land desertification, and simulating water resource changes based on the water resource data to reflect the constraints of water resources on social and economic development. Land use change is also an important factor affecting the evolution of the vulnerability of the sandy area. It is affected by both natural and human factors, which can not only reflect the guiding role of the overall regional development goal in land resource allocation but also reflect the demands and initiatives of micro-subjects. In addition, land use will also act on the simulated changes in the vulnerability of the sandy area, affecting the vegetation condition and economic development. Therefore, based on the natural environment data and combined with the future development goals, constraint conditions are set, the quantity structure of land use is optimized, and then based on the determined optimized quantity of land use, the spatial layout is optimized by considering the demands of different subjects. Finally, based on the determination of the land use quantity structure, the development of vegetation, social economy, and water resource data in the research area is further simulated, a model is constructed, and the model is verified to achieve accurate simulation of the vulnerability of the research area. According to the simulation model, the vulnerability evolution of the EES composite system in the sandy area under different scenarios such as climate change, social and economic development, and ecological governance can be predicted, so as to formulate the best solutions that meet the needs of various stakeholders for different regions and different goals, reduce costs, and improve work efficiency.

[0163] Refer to Figure 15 As shown in

[0164] The acquisition module 301 is used to acquire the natural environment data of the research area, where the natural environment data includes climate data, vegetation data, and water resource data;

[0165] The multi-objective constraint module 302 is used to set constraint conditions based on the natural environment data according to the social and economic and ecological construction needs of the research area;

[0166] The processing module 303 is used to optimize the land use structure based on the constraint conditions to meet the land use quantity structure;

[0167] The construction module 304 is used to perform change simulation and construct a model for the development of vegetation, socio - economic, and water resource data in the research area based on the determination of the land - use quantity structure.

[0168] The verification module 305 is used to calibrate, verify the model, and perform sensitivity analysis on key parameters, continuously iterating and optimizing the model to promote changes in vulnerability assessment indicators so as to achieve vulnerability simulation in the research area.

[0169] As Figure 16 shown, it is a block diagram of a computer device according to an embodiment of the present application. The computer device is intended to represent various forms of digital computers or mobile devices. Among them, the digital computer may include a desktop computer, a portable computer, a workbench, a personal digital assistant, a server, a mainframe computer, and other suitable computers. The mobile device may include a tablet computer, a smart phone, a wearable device, etc.

[0170] As Figure 16 shown, the device 600 includes a computing unit 601, a ROM 602, a RAM 603, a bus 604, and an input / output (I / O) interface 605. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other through the bus 604. The input / output (I / O) interface 605 is also connected to the bus 604.

[0171] The computing unit 601 can execute various processes in the method embodiments of the present application according to computer instructions stored in the read - only memory (ROM) 602 or computer instructions loaded from the storage unit 608 into the random - access memory (RAM) 603. The computing unit 601 can be various general - purpose and / or special - purpose processing components with processing and computing capabilities. The computing unit 601 may include, but is not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine - learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. In some embodiments, the method provided by the embodiments of the present application can be implemented as a computer software program, which is tangibly contained in a computer - readable storage medium, such as the storage unit 608.

[0172] The RAM 603 can also store various programs and data required for the operation of the device 600. Part or all of the computer programs can be loaded and / or installed onto the device 600 via the ROM 602 and / or the communication unit 609.

[0173] The input unit 606, output unit 607, storage unit 608, and communication unit 609 in the device 600 can be connected to the I / O interface 605. Among them, the input unit 606 can be, for example, a keyboard, mouse, touch screen, microphone, etc.; the output unit 607 can be, for example, a display, speaker, indicator light, etc. The device 600 can exchange information, data, etc. with other devices through the communication unit 609.

[0174] It should be noted that the device may also include other components necessary for normal operation. It may also only include the components necessary to implement the solution of this application, and does not necessarily include all the components shown in the figure.

[0175] The various embodiments of the systems and techniques described herein can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof.

[0176] The computer instructions for implementing the methods of this application can be written in any combination of one or more programming languages. These computer instructions can be provided to the computing unit 601, such that when the computer instructions are executed by a computing unit 601 such as a processor, the steps involved in the method embodiments of this application are executed.

[0177] The computer-readable storage medium provided by this application can be a tangible medium that can contain or store computer instructions for executing the steps involved in the method embodiments of this application. The computer-readable storage medium can include, but is not limited to, storage media in the forms of electronic, magnetic, optical, electromagnetic, etc.

Claims

1. A method for simulating the evolution of vulnerability in sandy areas, characterized in that, It includes the following steps: Obtain the natural environment data of the research area; Among them, the natural environment data includes climate data, vegetation data, and water resource data; Set constraint conditions based on the social and economic and ecological construction needs of the research area according to the natural environment data, including: setting total area constraints, economic development constraints, food production constraints, ecological protection constraints, development trend constraints, reasonable regulation constraints, and non - negative constraints based on the natural environment data according to the social and economic and ecological construction needs of the research area; Optimize the land use structure based on the constraint conditions to obtain the target quantity of various types of land to meet the land use quantity structure, including: dividing land use types X, including cultivated land x1, forest land x2, grassland x3, water area x4, urban land x5, rural residential areas x6, other construction land x7, and unused land x8; establish an objective function around the three aspects of economy, ecology, and water resources, and describe it through formulas (1), (2), and (3): (1)， (2)， (3)， Wherein, Z1, Z2, and Z3 are economic benefits, ecological benefits, and total water consumption respectively, and EO k is the economic output per unit area of land use type k, which is determined by dividing the total output value of the industry corresponding to land use type k by the land area, and ESV k is the ecosystem service value per unit area of land use type k, and WC k is the water consumption per unit area of land use type k, and WC else is the water consumption not affected by the land use structure, and X k is the area of land use type k; the target quantities of various land use types are obtained by optimizing and solving according to the multi-objectives of economic benefits, ecological benefits, and total water consumption based on the above constraints; based on the target quantities of various land use types, the probability of a plot being converted into land use type K is calculated according to land use conversion, and spatial allocation is carried out to meet the land use quantity structure; Based on the determination of the land use quantity structure, conduct change simulations on the development of vegetation, social economy, and water resource data in the research area, and build models, including: selecting net primary productivity of vegetation NPP as an indicator to reflect the vegetation status and using it to measure the degree of land desertification, selecting and quantifying the effects of climate change, grazing, land use, and natural grassland restoration factors on vegetation NPP, building a vegetation - soil module to simulate the dynamic evolution of desertification in the research area; based on the close relationship between social economy and land use, divide the social economy concept into four parts: population change, economic development, technological progress, and income growth according to the basic characteristics of the social - economic system in the research area and its correlation with the evolution of vulnerability. By simulating the key variables of the total population, total GDP, and per capita income of farmers and herdsmen, build a social - economic module to simulate the dynamic evolution of desertification in the research area; divide the entire water resource system in the research area into four parts: domestic water use, production water use, ecological water use, and available water resources. Since water resource allocation is correlated with land use, use the water resource supply - demand ratio as the core variable of the module to build a water resource module to simulate the dynamic evolution of desertification in the research area; build a vulnerability evolution model of the sandy area based on the land use module, the vegetation - soil module, the social - economic module, and the water resource module to conduct vulnerability evolution simulation of the sandy area; Calibrate, verify, and conduct sensitivity analysis of key parameters on the model, continuously iterate and optimize the model, and promote the change of vulnerability assessment indicators to achieve the vulnerability simulation of the research area.

2. The method for simulating the evolution of vulnerability in the sandy area according to claim 1, wherein The total area constraint means that the sum of the areas of various types of land is equal to the total area of the land in the research area; The economic development constraint means that the GDP growth rate reaches the expected target; The food production constraint means that the comprehensive food production capacity reaches the expected target; The ecological protection constraint means that the forest land area meets the requirements of the forest coverage rate; The described development trend constraint indicates that the water area remains unchanged, the areas of urban land and other construction land are greater than the current values, and the area of unused land is less than the current value; The described reasonable regulation constraint means setting upper and lower regulation thresholds according to the historical change trends of the areas of various types of land; The described non - negative constraint means that the areas of various types of land meet the non - negative conditions.

3. The method for simulating the evolution of vulnerability in sandy areas according to claim 2, characterized in that The land use conversion includes urban expansion conversion, land consolidation conversion, and artificial greening conversion. The urban expansion conversion means that urban land, rural settlements, and other construction land occupy cultivated land, forest land, grassland, and unused land. The land consolidation conversion means that rural settlements are vacated into cultivated land, forest land, and grassland. The artificial greening conversion means that cultivated land, forest land, grassland, and unused land, according to the self - adaptive inertia competition mechanism, are iterated multiple times until the expected value is reached.

4. The simulation method for the evolution of vulnerability in sandy areas according to claim 3, wherein Based on the target quantities of various types of land, according to the land use conversion, calculating the probability that a plot is converted into the K - type land and performing spatial allocation to meet the land use quantity structure, including: The simulation of land use spatial change presets two types of entities, including the government entity and the department entity, and the two types of entities jointly make decisions on land use conversion; The department entity is used to submit land use applications to the government entity to guide the layout behaviors of different land uses, calculate the probability that a plot is selected by the department entity, and initially select the location of a certain type of land; The government entity is used to determine the final land use plan, calculate the probability that the government finally accepts the conversion of a plot into the k - type land according to the Monte Carlo method, and perform conversions from high to low according to the final probability until the land use demand is met; After allocating the land use type with the highest probability to the plot, if the department entity finds that the quantity of various types of land is inconsistent with the target quantity of various types of land, an adaptive inertia competition mechanism is introduced for spatial allocation to resolve the competition relationship between different land uses, and a land use module is constructed to meet the land use quantity structure; Among them, three types of departments are preset in the department entity, namely the economic department, the optimization department, and the ecological department. The economic department is responsible for the urban expansion conversion of the land use conversion, the optimization department is responsible for the land consolidation conversion of the land use conversion, and the ecological department is responsible for the artificial greening conversion.

5. The method for simulating the evolution of the vulnerability in sandy areas according to claim 4, wherein The department entity is used to submit land use applications to the government entity to guide the layout behaviors of different land uses, calculate the probability that a plot is selected by the department entity, and initially select the location of a certain type of land, including: The economic department is responsible for the urban expansion conversion of the land use conversion, and the optimization department is responsible for the land consolidation conversion of the land use conversion. The two department entities calculate the probability that a plot unit is selected by using the discrete choice model by integrating the comprehensive suitability probability, neighborhood influence, and conversion cost, and are described by formulas (4) and (5); (4)， (5)， Where, P k (i, j) is the probability that the department entity selects the plot unit, TP k (i, j) is the comprehensive probability that the plot unit is converted into land use of type k, S k (i, j) is the suitability probability that the plot unit is converted into land use of type k, Ω k (i, j) is the neighborhood influence; C k (i, j) is the conversion cost for the plot unit to be converted into land use of type k.

6. The simulation method for the evolution of the vulnerability in sandy areas according to claim 5, wherein The government entity is used to determine the final land use plan, calculate the probability that the government finally accepts the conversion of a plot into the k - type land according to the Monte Carlo method, and perform conversions from high to low according to the final probability until the land use demand is met, including: After determining the probability that a plot is selected by the department entity, the Monte Carlo method is used for random sampling to simulate the submission of land use plans by the department entity; The government entity accepts and reviews the land use applications of the department entity. The more times a plot is applied for by the department entity, the higher the probability of its acceptance. At the same time, the probability of acceptance of its adjacent plots also increases. After reaching the sampling times, calculate the probability that the government entity finally accepts the conversion of the plot into land use type k, and perform the conversion from high to low according to the final probability until the land use demand is met, which is described by the formula: (6)， (7)， Where P k (i, j) is the probability that the department entity selects the plot unit, g is the number of times the plot unit has been applied for, ΔP1 is the increased acceptance probability for each time the plot is applied for, h is the number of times the plots within the neighborhood range have been applied for, and ΔP2 is the increased acceptance probability for each time the plots within the neighborhood range are applied for.

7. The method for simulating the evolution of vulnerability in sandy areas according to claim 6, wherein After allocating the land use type with the highest selection probability to the plot, if the quantity of each type of land use is inconsistent with the target quantity of each type of land, an adaptive inertia competition mechanism is introduced for spatial allocation to resolve the competition relationship of different land uses and meet the land use quantity structure, including: In each update iteration when calculating the probability that the government finally accepts the conversion of the plot into land use type k, the ecological department selects the land use type with the highest probability according to the comprehensive probability of the plot unit being converted into cultivated land, forest land, grassland, and unused land and allocates it to the plot; If the quantity of each type of land use is inconsistent with the target quantity, adjust the adaptive inertia coefficient to increase or decrease the comprehensive probability of the plot being converted into a certain type of land use, and perform the next iteration to meet the land use quantity structure; Among them, the adaptive inertia coefficient is determined by the difference between the actual quantity of each current type of land use and the target quantity, and is described by formula (8): (8)， In the formula, Intertiat,k is the adaptive inertia coefficient of land use type k in the t-th iteration, Numk is the target quantity of land use type k, and Numt-1,k is the allocated quantity of land use type k in the t-th iteration; After adding the adaptive inertia mechanism, the comprehensive probability that the plot unit is converted into land use type k in the t-th iteration is obtained, which is described by formula (9): (9)。 8. The method for simulating the evolution of vulnerability in sandy areas according to any one of claims 1-7, characterized in that, Calibrate, verify, and perform sensitivity analysis of key parameters on the model to continuously iterate and optimize the model, promote changes in vulnerability assessment indicators, and achieve vulnerability simulation in the research area, including: Based on the land use module, use the receiver operating characteristic curve ROC curve and the area under the receiver operating characteristic curve AUC value to test the land use module and evaluate the simulation accuracy; Based on the vegetation soil module, perform validity verification according to the NPP simulated in the research area. Extract the simulated NPP results and real NPP data at specific times respectively, and compare and verify them through the correlation coefficient to evaluate the simulation accuracy; Based on the social economy module, compare the simulated values and real values of the added value of each industry and GDP respectively to evaluate the simulation accuracy; Based on the water resources module, simulate the per capita water consumption of residents, the water consumption per unit area of each type of land use, and the water consumption of the added value of each industry corresponding to the GDP data within a specific time, and compare the simulated values with the actual values to evaluate the simulation accuracy.

9. A simulation system for the evolution of vulnerability in sandy areas, characterized in that, Including: An acquisition module for acquiring the natural environment data of the research area; among them, the natural environment data includes climate data, vegetation data, and water resources data; The multi-objective constraint module is used to set constraint conditions based on the social-economic and ecological construction requirements of the research area according to the natural environment data, including: setting the total area constraint, economic development constraint, food production constraint, ecological protection constraint, development trend constraint, reasonable regulation constraint, and non-negative constraint based on the natural environment data according to the social-economic and ecological construction requirements of the research area; The processing module is used to optimize the land use structure based on the constraint conditions to meet the land use quantity structure, including: dividing the land use types X, including cultivated land x1, forest land x2, grassland x3, water area x4, urban land x5, rural residential areas x6, other construction land x7, and unused land x8; Establish an objective function around the three aspects of economy, ecology, and water resources, which is described by formulas (1), (2), and (3): (1)， (2)， (3)， Where, Z1, Z2, and Z3 are economic benefits, ecological benefits, and total water consumption respectively, EO k is the economic output per unit area of land use type k, which is determined by dividing the total output value of the industry corresponding to land use type k by the land area, ESV k is the ecosystem service value per unit area of land use type k, WC k is the water consumption per unit area of land use type k, WC else is the water consumption not affected by the land use structure, X k is the land area of land use type k; Based on multiple objectives of economic benefits, ecological benefits, and total water consumption, optimize and solve according to the constraint conditions to obtain the target quantity of various types of land use; Based on the target quantity of each type of land use, calculate the probability of the plot being converted into the K type of land use according to the land use conversion, and perform spatial allocation to meet the land use quantity structure; The construction module is used to simulate the changes in the development of vegetation, social economy, and water resources data in the research area based on the determination of the land use quantity structure, and construct a model, including: selecting the net primary productivity NPP of vegetation as an indicator to reflect the vegetation status and using this to measure the degree of land desertification, selecting and quantifying the effects of climate change, grazing, land use, and natural grassland restoration factors on vegetation NPP, constructing a vegetation-soil module to simulate the dynamic evolution of desertification in the research area; based on the close relationship between social economy and land use, according to the basic characteristics of the social-economic system in the research area and its correlation with the evolution of vulnerability, dividing the social-economic concept into four parts: population change, economic development, technological progress, and income growth, simulating the key variables of the total population, total GDP, and per capita income of farmers and herdsmen, constructing a social-economic module to simulate the dynamic evolution of desertification in the research area; dividing the entire water resources system in the research area into four parts: domestic water, production water, ecological water, and available water resources, and since water resources allocation is related to land use, using the water resources supply-demand ratio as the core variable of the module, constructing a water resources module to simulate the dynamic evolution of desertification in the research area; constructing a vulnerability evolution model of the sandy area based on the land use module, the vegetation-soil module, the social-economic module, and the water resources module to conduct vulnerability evolution simulation of the sandy area; The verification module is used to calibrate, verify, and conduct sensitivity analysis of key parameters for the model, continuously iterate and optimize the model, and promote the change of vulnerability assessment indicators to achieve the vulnerability simulation of the research area.

Images