A water-soil-crop multi-objective dynamic coordinated optimization configuration method for drought main grain area
Patent Information
- Application Number
- CN202611203162.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-10
- Publication Date
- 2026-10-27
AI Technical Summary
[0003]但是，现有资源配置技术存在诸多缺陷：一是多采用单一资源静态优化模式，仅针对水资源或土地资源独立调控，未考虑水、土、粮三大要素的互馈耦合关系，优化方案系统性不足；二是未纳入气候变化、冰川消融、跨境径流波动等动态因子，模型预测与实际演化偏差大，长期适应性差；三是生态约束仅作为外部限定条件，未深度融入优化目标，易出现粮食增产与生态破坏冲突；四是忽略农户种植习惯、政策接受度等主体行为因素，理论方案落地难度高；五是研究尺度单一，宏观区域布局无法有效下沉至县域、地块微观单元，调控精度不足
1、水资源利用效益：有效解决干旱区主粮区春灌缺水、汛期弃水的时空错配问题，灌溉水有效利用系数提升至0.65以上，地下水超采现象逐步消除，跨境水资源波动带来的供水风险显著降低。
Smart Images
Figure CN122988321A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of agricultural resource regulation technology, and more specifically to a multi-objective dynamic synergistic optimization allocation method for water, soil, and grain in arid staple food areas. Background Technology
[0002] Arid regions generally suffer from imbalances in the spatial and temporal distribution of water resources, spatial misallocation of water and soil resources, severe degradation of arable land, and fragile ecosystems. In arid staple food areas, grain production is highly dependent on artificial irrigation, and water resources are characterized by spring droughts and summer abundance. There are significant water supply gaps during the critical period of grain greening, and substantial runoff is lost during the flood season. At the same time, the regions face prominent issues of soil salinization and desertification, and the conflict between the development of reserve arable land and ecological protection is acute. Therefore, resource allocation in arid staple food areas is of great significance.
[0003] However, existing resource allocation technologies have many shortcomings: First, they mostly adopt a static optimization model for a single resource, focusing solely on the independent regulation of water or land resources without considering the feedback and coupling relationship among the three major elements of water, soil, and grain, resulting in insufficient systematic optimization schemes; second, they do not incorporate dynamic factors such as climate change, glacial melting, and cross-border runoff fluctuations, leading to large discrepancies between model predictions and actual evolution, and poor long-term adaptability; third, ecological constraints are only used as external limiting conditions and are not deeply integrated into the optimization objectives, easily leading to conflicts between increased grain production and ecological damage; fourth, they ignore the main behavioral factors such as farmers' planting habits and policy acceptance, making it difficult to implement theoretical schemes; and fifth, the research scale is singular, and macro-regional layout cannot be effectively extended to the county and plot micro-units, resulting in insufficient regulation precision.
[0004] Therefore, developing a method for the coordinated optimization of water, soil, and food resources that is adapted to the characteristics of arid regions, integrates multiple models, takes into account economic, food, and ecological goals, and can be dynamically iterated is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention
[0005] In view of this, the present invention provides a multi-objective dynamic collaborative optimization allocation method for water, soil and grain in arid staple food areas. By integrating multi-source monitoring data, climate prediction models, hydrological crop models and system dynamics, a multi-scale, multi-objective, dynamic optimization system is constructed to achieve integrated collaborative optimization of water resource allocation, land consolidation, staple food layout, ecological protection and farmers' behavior, providing technical support for the overall planning of resources in arid staple food areas.
[0006] To achieve the above objectives, the present invention adopts the following technical solution: This invention discloses a multi-objective dynamic collaborative optimization allocation method for water, soil, and food in arid staple food areas, comprising: Step 1: Collect multi-source basic data from arid staple food areas, preprocess the data, and construct a resource database; Step 2: Based on the resource database, evaluate the endowment of water resources, land resources, and staple food production in the study area, calculate the comprehensive matching degree of water-soil-grain, and divide the study area using a clustering algorithm; Step 3: Based on the resource data of each partition in Step 2, analyze the feedback relationship between water resources, land resources, and staple food production, calculate the system coupling coordination degree, and screen the core driving factors based on the coupling coordination degree. Step 4: Based on the driving factors selected in Step 3, construct a multi-objective dynamic optimization model and solve for the basic configuration scheme; Step 5: Analyze farmers' behavioral characteristics based on the discrete choice model, and verify and optimize the basic configuration scheme; Step 6: Set up different simulation scenarios, conduct long-term dynamic simulation and comprehensive benefit evaluation of the optimized basic configuration scheme, and iteratively update the optimization scheme according to a fixed cycle.
[0007] Furthermore, the multi-source basic data includes, within the research scope of arid staple food areas: water resource data, land resource data, staple food production data, environmental and climate data, and social behavior data; The preprocessing includes: using linear interpolation to fill in missing data and removing outlier data based on the 3σ criterion; using Kriging interpolation to convert point data to surface data and unify the coordinate system of all spatial data; and using range standardization for normalization.
[0008] Furthermore, step 2 specifically includes: Data envelopment analysis model was used to evaluate water resource utilization efficiency and total factor productivity of staple food production, and AHP-entropy weight combination weighting method was used to evaluate arable land quality. The formula for calculating the comprehensive matching degree of water, soil, and food is: In the formula, For overall matching accuracy, For water resource matching index, This is a land resource matching index. A matching index for grain production; The study area was divided into high-matching, medium-matching, and low-matching regions using the K-Means clustering algorithm, and further subdivided into functional zones.
[0009] Furthermore, in step 3, a structural equation model is used to identify the feedback paths and correlation strengths among water resources, land resources, and staple food production; a coupling coordination degree model is used to calculate the coupling degree and coordination degree of the system and classify the coordination level; a panel data fixed effects regression model is used, with the coupling coordination degree as the dependent variable, combined with stepwise regression to screen the core driving factors.
[0010] Furthermore, step 4 specifically includes: Based on the aforementioned driving factors, the CMIP6 climate prediction sub-model, SWAT hydrological sub-model, DSSAT crop growth sub-model, and system dynamics SD sub-model are constructed and integrated. A multi-objective function is constructed and constraints are set to obtain the multi-objective dynamic optimization model. The multi-objective function includes objective functions for maximizing economic benefits, stabilizing food yield, and minimizing ecological risks. The constraints include resource constraints, ecological constraints, technological constraints, and policy constraints. The NSGA-II non-dominated sorting genetic algorithm is used to solve the multi-objective dynamic optimization model to generate the optimal solution set; combined with the TOPSIS superior-inferior solution distance method, the basic configuration scheme with the best comprehensive benefits is selected from the optimal solution set.
[0011] Furthermore, in step 5, a discrete choice model is used to quantify the factors influencing farmers' planting decisions, and a farmer acceptance threshold is set. Based on the acceptance threshold, the basic configuration scheme is modified, and differentiated water resource scheduling, farmland improvement, and staple food planting structure schemes are formulated for different zones.
[0012] Furthermore, in step 6, the simulation scenarios include a baseline scenario, a technology optimization scenario, a policy optimization scenario, and a comprehensive optimization scenario; a long-term dynamic simulation with a set number of years is carried out through a system dynamics model, and an economic-food-ecological comprehensive benefit evaluation system is constructed, with an evaluation cycle of 5 years, and the model and configuration scheme are iteratively corrected according to changes in the external environment.
[0013] As can be seen from the above technical solution, compared with the prior art, the present invention provides a multi-objective dynamic collaborative optimization configuration method for water-soil-grain in arid staple food areas, which has the following beneficial effects: 1. Water resource utilization benefits: Effectively solves the problem of water shortage in spring irrigation and water wastage during the flood season in arid and staple food areas. The effective utilization coefficient of irrigation water is increased to over 0.65. The over-extraction of groundwater is gradually eliminated, and the water supply risk caused by cross-border water resource fluctuations is significantly reduced.
[0014] 2. Land resource utilization benefits: Saline-alkali and desertified farmland will be classified and remediated, and the quality of farmland will be steadily improved; the development scope of reserve farmland will be scientifically delineated, ecologically fragile areas will be strictly avoided, and farmland expansion and ecological security will be taken into account.
[0015] 3. Food security benefits: Optimize the regional planting structure, achieve precise matching between crop types and water and soil resources, ensure long-term stability of total staple grain output, control the output fluctuation coefficient within 5%, and continuously consolidate the region's food self-sufficiency.
[0016] 4. Ecological and social benefits: Ensure ecological water use for rivers and wetlands, curb the trend of land desertification and soil degradation; differentiated subsidies and planting programs take into account farmers' income, enhance farmers' enthusiasm for participation, and help stabilize the agricultural industry in border areas and revitalize rural areas.
[0017] 5. Technical promotion value: The method is highly versatile and can be directly promoted to Central Asia, the Middle East, and other arid agricultural areas in Northwest my country, providing a standardized technical solution for the coordinated allocation of water, soil, and food resources in arid regions around the world. Attached Figure Description
[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0019] Figure 1 This is a schematic diagram of the overall process of the present invention. Detailed Implementation
[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0021] This invention discloses a multi-objective dynamic collaborative optimization allocation method for water, soil, and food in arid staple food areas, such as... Figure 1 As shown, it includes: Step 1: Collect multi-source basic data from arid staple food areas, preprocess the data, and construct a resource database; Step 2: Based on the resource database, evaluate the endowment of water resources, land resources, and staple food production in the study area, calculate the comprehensive matching degree of water-soil-grain, and divide the study area into sections using a clustering algorithm; Step 3: Based on the resource data of each matching partition divided in Step 2, analyze the feedback relationship between water resources, land resources, and staple food production, calculate the system coupling coordination degree, and screen the core driving factors based on the coupling coordination degree. Step 4: Based on the driving factors selected in Step 3, construct a multi-objective dynamic optimization model and solve for the basic configuration scheme; Step 5: Analyze farmers' behavioral characteristics based on the discrete choice model, and verify and optimize the basic configuration scheme; Step 6: Set up different simulation scenarios, conduct long-term dynamic simulation and comprehensive benefit evaluation of the optimized basic configuration scheme, and iteratively update the optimization scheme according to a fixed cycle.
[0022] In one specific embodiment, the multi-source basic data includes, within the research scope of arid staple food areas: water resource data, land resource data, staple food production data, environmental and climate data, and social behavior data.
[0023] Specifically, the research scope of the arid staple food area was defined, and the entire area and its subordinate cities, counties, and plots were selected as research objects. Five types of basic data were collected: Water resources data: surface water, groundwater, reservoir capacity, irrigation area water consumption, cross-border river runoff, and ecological water conveyance planning data; Land resource data: cultivated land area, cultivated land quality, distribution of saline-alkali / desertified cultivated land, reserve cultivated land resources, cultivated land red line, and topographic remote sensing data; Staple food production data: sown area, yield, yield per unit area, planting structure, and production cost of wheat, corn, and rice; Environmental and climate data: historical meteorological observation data, CMIP6 multi-scenario climate prediction data, glacier area and meltwater monitoring data; Social behavior data: Survey data on farmers' planting intentions, irrigation methods, policy awareness, agricultural subsidies, and water price policies.
[0024] Preprocessing included: using linear interpolation to fill in missing temporal data and removing outliers based on the 3σ criterion; converting point data to area data using Kriging interpolation and unifying the coordinate system of all spatial data to WGS84; and using range standardization to normalize indices of different dimensions and eliminate differences in data dimensions. After preprocessing, a three-level multi-scale resource database covering the entire research scope (macro-county meso-plot micro) was built to enable data classification, storage, querying, and linked access.
[0025] In one specific embodiment, step 2 specifically includes: A data envelopment analysis (DEA) model was used to evaluate water resource utilization efficiency and total factor productivity (TFP) of staple food production. The AHP-entropy weighted combination method was employed to evaluate arable land quality, thereby quantifying arable land fertility, salinity, and irrigation security levels, and classifying arable land quality into grades. The specific process of arable land quality evaluation using the AHP-entropy weighted combination method is as follows: Constructing an indicator system and data preprocessing: Selecting indicators such as arable land fertility (organic matter, available nutrients, etc.), salinization degree (pH value, salt content, etc.) and irrigation guarantee level (effective irrigation rate, drainage capacity, etc.), and performing range standardization on the raw data to eliminate the influence of dimensions.
[0026] Subjective and objective weighting integration: Using the Analytic Hierarchy Process (AHP), a judgment matrix is constructed and a consistency test is performed (the CR value of the judgment matrix is less than 0.1) to solve for the subjective weights reflecting the experience of business experts. W AHP ； Simultaneously, the entropy weight method is employed to calculate the information entropy and difference coefficient based on the dispersion of each indicator's data, thereby determining the objective weights. W Entropy 。 The combined weights are synthesized using the product normalization formula, taking into account both subjective experience and objective data characteristics.
[0027] Scoring and Quality Grading: The standardized index values and combined weights are weighted and summed to calculate the Integrated Quality Index (IFI) of cultivated land, and the cultivated land quality grade is determined by combining the natural breakpoint method.
[0028] The formula for calculating the comprehensive matching degree of water, soil, and food is: In the formula, For overall matching accuracy, For water resource matching index, This is a land resource matching index. The grain production matching index was used. The study area was divided into high-matching, medium-matching, and low-matching areas using the K-Means clustering algorithm. It was further divided into functional zones, and then further divided into three major functional zones based on the regional resource characteristics: priority guarantee zone, moderate adjustment zone, and restricted development zone. The resource shortcomings and optimization directions of each zone were identified.
[0029] Further subdivision based on regional resource characteristics includes: Ecological / climate geographical zoning: Based on the water resource endowment (e.g., arid and semi-arid regions, humid regions), soil type (e.g., black soil regions, saline-alkali soil regions), and topography (e.g., plains, hills) of the region to be evaluated, the evaluation area is divided into several homogeneous regional resource sub-regions.
[0030] Differentiated Indicator System and Dynamic Weight Correction: The weights and threshold settings of farmland quality evaluation indicators are adjusted for different resource sub-regions. For example, in high-risk drought / salinization areas, the evaluation weights of salinization level (pH value, salinity) and irrigation guarantee level are appropriately increased; in water-rich areas, the weights of soil fertility and organic matter indicators are appropriately increased to achieve refined farmland quality rating tailored to local conditions. Staple Crop Planting Structure Grouping: Before using the DEA model, decision units (DMUs) are grouped or crop coefficients are introduced based on the characteristics of staple crop planting structures in each region (e.g., single-season rice areas, wheat-corn rotation areas, drought-resistant miscellaneous grain areas) to eliminate efficiency measurement biases caused by differences in crop physiological water requirements.
[0031] Construction of the DEA frontier under rigid resource constraints: The upper limit of the total available water resources in each sub-region (water resource red line constraint) and the over-extraction of groundwater are used as limiting conditions or environmental variables (NADP / three-stage DEA adjustment). Environmental noise caused by differences in the natural water resource background in different regions is eliminated, so that the calculated water resource utilization efficiency and total factor productivity of staple food production truly reflect the differences in management and technology levels, rather than simply the differences in natural resource endowment.
[0032] In a specific embodiment, step 3 involves using structural equation modeling to identify the direct and indirect feedback paths between water resources, land resources, and staple food production, and quantifying the strength of factor correlations. A coupling and coordination degree model is used to calculate the coupling and coordination degree of each research unit system, and to classify coordination levels into five categories: extremely coordinated, highly coordinated, moderately coordinated, slightly imbalanced, and severely imbalanced, to analyze the spatiotemporal evolution of the system. A panel data fixed-effects regression model is then used, with coupling and coordination degree as the dependent variable and climate change, water-saving technology, land consolidation, agricultural policies, and food prices as independent variables. Stepwise regression is employed to eliminate insignificant factors, identifying the core driving factors of system evolution. These selected core driving factors are then used as input variables for subsequent optimization models.
[0033] Among these methods, structural equation modeling is used to identify the direct and indirect feedback paths between water resources, land resources, and staple food production, and to quantify the strength of factor correlations. Specifically, this includes: Construct latent and observed variables: Establish three latent variables (comprehensive concepts that cannot be directly observed) of “water resources”, “land resources” and “staple food production”, and map them to specific observed variables (such as water resource utilization efficiency, arable land quality grade / organic matter content, total factor productivity / yield of staple food, etc.).
[0034] Path model and assumptions: Establish a two-way influence path between water resources and land resources, as well as the direct influence paths of water resources and land resources on staple food production, and include the indirect influence path generated by the interaction of "water and soil".
[0035] Model fitting and coefficient calculation: The maximum likelihood estimation method (ML) is used to fit and correct the measurement model and structural model to meet the goodness-of-fit index; the standardized path coefficients of each path are calculated to quantify the direct, indirect and total effects between water resources, land resources and staple food production.
[0036] In one specific embodiment, step 4 specifically includes: Based on driving factors, a dynamic simulation framework was constructed and integrated using the CMIP6 climate prediction sub-model, the SWAT hydrological sub-model, the DSSAT crop growth sub-model, and the SD system dynamics sub-model. This framework predicts the dynamic changes in temperature, precipitation, glacial meltwater, river runoff, and crop yield over multiple years. A multi-objective function was constructed and constraints were set to obtain a multi-objective dynamic optimization model. The multi-objective function includes objectives for maximizing economic benefits, stabilizing food yield, and minimizing ecological risks. The constraints include resource constraints, ecological constraints, technological constraints, and policy constraints. Specifically... The objective function for maximizing economic returns is: In the formula, For planting area, The yield per unit area of staple food, For grain market prices, For water resource costs, For land redevelopment costs, For production and operating costs; The objective function for stabilizing grain yield is: In the formula, To simulate total output, Targeted production levels for regional food security planning; The objective function for minimizing ecological risk is: In the formula, , , These are the risk coefficients for groundwater over-extraction, land desertification, and ecological water shortage, respectively. This refers to the over-extraction of groundwater. The area of desertified land. This represents the ecological water shortage.
[0037] Resource constraints: The total amount of usable water resources and the area of arable land in the region shall not be less than the red line for arable land; Ecological constraints: The amount of ecological water conveyance, the area of wetland protection, and the amount of groundwater extraction shall not exceed the control thresholds; Technical constraints: The effective utilization coefficient of irrigation water and the rate of transformation of saline-alkali farmland must reach the phased targets; Policy constraints: The total sown area of staple grains must meet the requirements of local grain planning.
[0038] The NSGA-II non-dominated sorting genetic algorithm is used to solve the multi-objective dynamic optimization model and generate the Pareto optimal solution set. Combined with the TOPSIS superior-inferior solution distance method, the basic configuration scheme with the best overall benefits is selected from the optimal solution set.
[0039] In one specific embodiment, step 5 employs a discrete choice model to quantify the factors influencing farmers' planting decisions and sets a farmer acceptance threshold. Based on this threshold, the basic configuration plan is revised, and differentiated water resource allocation, farmland improvement, and staple food planting structure plans are developed for different functional zones, along with corresponding tiered water pricing and differentiated agricultural subsidy policies. Specifically: Based on the Discrete Choice Model (DCM), this study quantifies the impact of subsidy standards, planting costs, and technical difficulties on farmers' planting decisions using survey data, and sets a threshold for farmer acceptance. Specifically, it constructs a set of farmers' planting decision choices (e.g., "choosing the optimal mode" and "maintaining the traditional mode"), extracts core influencing factors such as subsidy standards (X1), planting costs (X2), and technical difficulties (X3), and constructs a model integrable utility based on individual farmer characteristics. The random utility function with the random error term is used; based on micro-survey data, a binary Logit model and maximum likelihood estimation (MLE) are employed to solve for the model parameters, quantifying the marginal contribution of each factor to the decision utility; finally, the logistic response function is used. Predict the probability of farmer acceptance under a specific policy and technology combination, set a feasible acceptance threshold, and use the regression coefficient ratio to calculate the subsidy compensation threshold for farmers to overcome technical difficulties or cost increases, providing micro-response constraints for subsequent optimization of the regulation model.
[0040] Develop differentiated configuration strategies for the three functional zones: Priority protection areas: Stabilize the scale of wheat planting, moderately expand corn planting, fully promote drip irrigation water-saving technology, and unleash the production potential of arable land; Moderate adjustment area: Reduce the planting area of water-intensive rice, expand the planting of drought-resistant corn, implement "biological + engineering" combined saline-alkali land transformation, and promote drip irrigation under film; Restricted Development Zones: Strictly control the planting of water-intensive crops, develop dryland agriculture, prioritize ecological water use, and restrict the development of reserve arable land.
[0041] Adjust the allocation plan to match policies such as tiered water pricing, differentiated agricultural subsidies, and compensation for farmland occupation, ensuring that farmers' acceptance meets the threshold requirements.
[0042] In one specific embodiment, step 6 involves simulating scenarios including a baseline scenario, a technology optimization scenario, a policy optimization scenario, and a comprehensive optimization scenario. A long-term dynamic simulation over a set period is conducted using a system dynamics model, and an economic-food-ecological comprehensive benefit evaluation system is constructed. With a 5-year evaluation cycle, the model and configuration scheme are iteratively revised based on changes in the external environment. Specifically: System boundary and causal loop construction: Set the spatial boundary of the system (such as a specific watershed or administrative region) and the temporal boundary (such as setting the simulation period to 2025–2050, with a time step of 1 year); take water resources, land resources, staple food production, socio-economic and ecological environment as subsystems, sort out the causal feedback chains within and between each subsystem, draw the system causal loop diagram (CLD), and clarify the positive and negative feedback nature of the feedback loops.
[0043] Stock-Flow Diagram Construction and State Equation Establishment: The causal loop is transformed into a stock-flow diagram (SFD) containing state variables (Level, such as water resource storage and arable land area), rate variables (Rate, such as annual water consumption growth rate and land degradation rate), and auxiliary variables (Auxiliary, such as technological progress rate and subsidy intensity). Based on the fitting relationship between the system's physical mechanism and historical statistical data, an integral equation and a set of differential equations for the system's state evolution are established. The typical state equation is expressed as follows: Where S(t) is the state variable value at time t, S(t0) is the initial stock, and R... in and R out These are the inflow rate and the outflow rate, respectively.
[0044] Model validation and long-term scenario simulation: Historical observation data (such as data from the past 10–20 years) are used to test the model's historical performance and sensitivity (the historical simulation error is required to be less than 10%). The core driving factors selected earlier (such as climate change, agricultural policies, water-saving technologies, etc.) and the farmer acceptance threshold are introduced as exogenous control variables of the system. Various development scenarios are set (such as baseline continuation scenario, water-saving priority scenario, ecological protection scenario, and high-standard farmland collaborative optimization scenario) to carry out long-term dynamic extrapolation and trend prediction for a set number of years.
[0045] Construction and Calculation Process of Economic-Food-Ecological Integrated Benefit Evaluation System
[0046] Constructing a three-dimensional comprehensive benefit index system: Based on the dynamic simulation time-series data output by the SD model, a benefit evaluation index system covering three core dimensions is established: Economic benefits dimension (B) eco ): This includes indicators such as total agricultural output value, net income of farmers from grain production, and economic output rate of water resource utilization; Food security (B) food ): This includes indicators such as total staple grain output, per capita grain possession, total factor productivity of staple grains, and self-sufficiency rate; Ecological benefits dimension (B) eco-sysThis includes indicators such as groundwater over-extraction / recovery, reduction in fertilizer and pesticide application, non-point source pollution control rate, and green ecological barrier coverage rate.
[0047] Dimensionless data processing and comprehensive benefit index calculation: Data for each dimension is standardized using the efficacy coefficient method or range method to eliminate the influence of dimensions; a combined weighting method (such as AHP-entropy weighting) is used to determine the weight coefficients (W) of each benefit dimension and specific indicator. eco W food W eco-sys Construct the equation for calculating the comprehensive benefit index (EFEI): .
[0048] Multi-scenario benefit comparison and optimal control scheme decision: Run the SD model to compare and analyze the change curve of the comprehensive benefit index EFEI(t) and the cumulative benefit within a set period under different control scenarios. With the goal of maximizing comprehensive benefits and ensuring that the benefits of each subsystem do not decrease, the optimal water and soil resource control and staple food production optimization implementation plan is selected and output.
[0049] Four simulation scenarios are set up: baseline scenario, technology optimization scenario, policy optimization scenario, and comprehensive optimization scenario. Based on a system dynamics model, the evolution trends of the water, soil, and food systems under different scenarios from 2024 to 2053 are simulated, and the economic, food security, and ecological benefits of the four scenarios are compared. An integrated economic-food-ecological benefit evaluation system is established to quantify the effectiveness of the implementation plan. A five-year dynamic evaluation cycle is set, and the optimization model and configuration plan are iteratively revised based on climate fluctuations, runoff changes, policy adjustments, and farmer feedback to achieve long-term dynamic management.
[0050] In a specific embodiment, the present invention will be further described in detail with reference to an application example in a staple grain area.
[0051] 1. Scope of Implementation
[0052] The implementation area is selected from 30 major grain-producing counties and cities, including Changji Prefecture, Shihezi City, Aksu Prefecture, Kashgar Prefecture, Hami City, and Turpan City, spanning from 2018 to 2053.
[0053] 2. Data Acquisition and Preprocessing
[0054] We collected water resources bulletins, land use data, grain production statistics, Landsat and MODIS remote sensing images, meteorological observation data, climate prediction data for five typical SSP paths of CMIP6, cross-border runoff data of the Ili River / Ertis River, and ecological water transfer planning data of the Tarim River from 2018 to 2024 in the above-mentioned areas; we completed data interpolation, outlier removal, coordinate unification, and normalization processing to build a three-level multi-scale database.
[0055] 3. Resource assessment and spatial zoning
[0056] The DEA model was used to evaluate the efficiency of water resources and arable land use, and the AHP-entropy weight method was used to evaluate the quality of arable land. The comprehensive matching degree of water, soil and food was calculated, and the study area was divided into high, medium and low matching areas by K-Means clustering. Three functional areas were delineated: the priority protection area in northern Xinjiang, the moderate adjustment area in southern Xinjiang, and the restricted development area in eastern Xinjiang.
[0057] 4. System Coupling and Driving Factor Analysis
[0058] The SEM model was used to analyze the feedback relationship between water, soil and food, and the coupling coordination degree was calculated from 2018 to 2024. The panel regression model identified glacial meltwater changes, efficient water-saving technologies, saline-alkali land improvement and agricultural subsidies as the core driving factors.
[0059] 5. Construction and solution of multi-objective models
[0060] The CMIP6, SWAT, DSSAT, and SD models are coupled, and three major objectives of economy, food, and ecology are set, along with resource, ecological, technological, and policy constraints. The NSGA-II algorithm is used to solve the problem, and the optimal basic configuration scheme is selected by combining it with the TOPSIS method.
[0061] 6. Farmer behavior verification and differentiated allocation
[0062] A field survey of 500 farming households was conducted to construct a DCM model to analyze farmers' planting preferences and policy acceptance. Based on threshold requirements, the plan was adjusted: in northern Xinjiang, wheat planting will remain stable at 60%, and corn at 30%; in southern Xinjiang, rice planting will be reduced to below 5%, and the saline-alkali land improvement rate will reach 45% by 2025; in eastern Xinjiang, high water-consuming crops will be strictly controlled, and dryland agriculture will be fully developed. A tiered water pricing system and a subsidy policy based on water resource efficiency will be implemented simultaneously.
[0063] 7. Scenario Simulation and Dynamic Iteration
[0064] Four scenarios were set up to conduct long-term simulations from 2024 to 2053. The comprehensive and optimized scenario was determined as the final implementation plan based on the comprehensive comparison results. A benefit assessment was conducted every 5 years, and the model and configuration plan were iteratively optimized based on changes in climate, runoff and policies.
[0065] 8. Implementation Results
[0066] After the implementation of the plan, the total output of staple grains in the region will stabilize at more than 17.5 million tons by 2025, the over-extraction of groundwater will be reduced by 40%, and the effective utilization coefficient of irrigation water will reach 0.6; by 2030, the total output will exceed 18 million tons, the transformation rate of saline-alkali farmland will reach 60%, the ecological water use guarantee rate will be 100%, and the coordinated development of increased grain production, resource conservation and ecological protection will be achieved.
[0067] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.
[0068] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A multi-objective dynamic collaborative optimization allocation method for water, soil, and food in arid staple food areas, characterized in that, include: Step 1: Collect multi-source basic data from arid staple food areas, preprocess the data, and construct a resource database; Step 2: Based on the resource database, evaluate the endowment of water resources, land resources, and staple food production in the study area, calculate the comprehensive matching degree of water-soil-grain, and divide the study area using a clustering algorithm; Step 3: Based on the resource data of each partition in Step 2, analyze the feedback relationship between water resources, land resources, and staple food production, calculate the system coupling coordination degree, and screen the core driving factors based on the coupling coordination degree. Step 4: Based on the driving factors selected in Step 3, construct a multi-objective dynamic optimization model and solve for the basic configuration scheme; Step 5: Analyze farmers' behavioral characteristics based on the discrete choice model, and verify and optimize the basic configuration scheme; Step 6: Set up different simulation scenarios, conduct long-term dynamic simulation and comprehensive benefit evaluation of the optimized basic configuration scheme, and iteratively update the optimization scheme according to a fixed cycle.
2. The method for multi-objective dynamic collaborative optimization of water-soil-grain allocation in arid staple food areas according to claim 1, characterized in that, The multi-source basic data includes: water resources data, land resources data, staple food production data, environmental and climate data, and social behavior data within the research scope of arid staple food areas; The preprocessing includes: using linear interpolation to fill in missing data and removing outlier data based on the 3σ criterion; using Kriging interpolation to convert point data to surface data and unify the coordinate system of all spatial data; and using range standardization for normalization.
3. The method for multi-objective dynamic collaborative optimization of water-soil-grain allocation in arid staple food areas according to claim 1, characterized in that, Step 2 specifically includes: Data envelopment analysis model was used to evaluate water resource utilization efficiency and total factor productivity of staple food production, and AHP-entropy weight combination weighting method was used to evaluate arable land quality. The formula for calculating the comprehensive matching degree of water, soil, and food is: In the formula, For overall matching accuracy, For water resource matching index, This is a land resource matching index. A matching index for grain production; The study area was divided into high-matching, medium-matching, and low-matching regions using the K-Means clustering algorithm, and further subdivided into functional zones.
4. The method for multi-objective dynamic collaborative optimization of water-soil-grain allocation in arid staple food areas according to claim 1, characterized in that, In step 3, structural equation modeling is used to identify the feedback paths and correlation strengths among water resources, land resources, and staple food production; coupling and coordination degree model is used to calculate the coupling degree and coordination degree of the system and classify the coordination level; panel data fixed effects regression model is used, with coupling and coordination degree as the dependent variable, and stepwise regression method is used to screen the core driving factors.
5. The method for multi-objective dynamic collaborative optimization of water-soil-grain allocation in arid staple food areas according to claim 1, characterized in that, Step 4 specifically includes: Based on the aforementioned driving factors, the CMIP6 climate prediction sub-model, SWAT hydrological sub-model, DSSAT crop growth sub-model, and system dynamics SD sub-model are constructed and integrated. A multi-objective function is constructed and constraints are set to obtain the multi-objective dynamic optimization model. The multi-objective function includes objective functions for maximizing economic benefits, stabilizing food yield, and minimizing ecological risks. The constraints include resource constraints, ecological constraints, technological constraints, and policy constraints. The NSGA-II non-dominated sorting genetic algorithm is used to solve the multi-objective dynamic optimization model to generate the optimal solution set; combined with the TOPSIS superior-inferior solution distance method, the basic configuration scheme with the best comprehensive benefits is selected from the optimal solution set.
6. The method for multi-objective dynamic collaborative optimization of water-soil-grain allocation in arid staple food areas according to claim 1, characterized in that, In step 5, a discrete choice model is used to quantify the factors influencing farmers' planting decisions and to set a farmer acceptance threshold. Based on the acceptance threshold, the basic configuration scheme is modified, and differentiated water resource scheduling, farmland improvement, and staple food planting structure schemes are formulated for different zones.
7. The method for multi-objective dynamic collaborative optimization of water-soil-grain allocation in arid staple food areas according to claim 1, characterized in that, In step 6, the simulation scenarios include a baseline scenario, a technology optimization scenario, a policy optimization scenario, and a comprehensive optimization scenario; Long-term dynamic simulations over a set period are conducted using a system dynamics model, and a comprehensive evaluation system for economic, food, and ecological benefits is constructed. The model and configuration scheme are iteratively revised based on changes in the external environment, with an evaluation cycle of 5 years.