A simulation and prediction method for national land spatial pattern based on coordinated spatial planning and carbon reduction control

Abstract

The present invention discloses a method for simulating and predicting the spatial pattern of land for coordinated spatial planning and carbon reduction control, and belongs to the technical field of land space planning. The present invention includes collecting data, preprocessing the collected data, and establishing a basic database for optimizing low-carbon land use; combining energy carbon emissions with spatial mapping technology to calculate the annual land use carbon emissions, and using the GM (1,1) model to predict the land use carbon emission coefficient of the target year; based on Markov and linear programming models, combined with natural development trends and land space planning requirements, predict and optimize the quantitative structure of low-carbon land use; using the PLUS model, simulate and predict the land use pattern under different scenarios; evaluate the carbon reduction benefits of the low-carbon land use optimization method of coordinated spatial planning, and propose multi-scale low-carbon decision-making and optimization suggestions. The method of the present invention not only reduces policy risks and problems in actual implementation, but also provides a more accurate and dynamic carbon emission optimization solution.

Description

Technical Field

[0001] The present invention belongs to the technical field of national land space planning, and specifically relates to a national land space pattern simulation and prediction method for coordinated spatial planning and carbon reduction control. Background Art

[0002] The current technology for optimizing the low-carbon land spatial pattern has undergone significant changes, from the initial "top-down" structural optimization to the emphasis on "bottom-up" spatial configuration optimization, and then to the current comprehensive method that combines "top-down" quantitative structure optimization and "bottom-up" spatial local optimization. This process not only deepens the understanding of the optimization of low-carbon land use pattern, but also puts forward higher requirements for improving the actual benefits of low-carbon land development. At present, the optimization of low-carbon land use pattern mainly includes the following three methods:

[0003] 1. Structural optimization of low-carbon land use: This method has been widely used to optimize land use structure from a "top-down" perspective, minimizing carbon emissions and maximizing carbon sinks by controlling the amount of land. Based on historical land use change trends and low-carbon emission targets, Markov models, system dynamics models, gray models and linear programming models are used to predict future land use quantities, identify and evaluate the carbon emission effects of different land use options, and provide scientific basis for policy makers.

[0004] 2. Spatial optimization of low-carbon land use: This method emphasizes the local allocation of land resources under the constraints of spatial protection. By analyzing the relationship between policies, economy, population, transportation and natural environment, combined with intelligent algorithms such as cellular automata, a more accurate land use pattern can be simulated. By capturing the spatial heterogeneity and complexity of land use changes, the layout of land use can be optimized.

[0005] 3. Combination of quantitative optimization and spatial optimization: This method combines the "top-down" quantitative optimization method and the "bottom-up" spatial optimization method to explore the optimal solution for the quantity and spatial configuration of land use for low-carbon development. This combined method not only considers the total control of low-carbon land use, but also focuses on the spatial distribution of land use and its impact on carbon effects, providing a more comprehensive and accurate low-carbon land use solution.

[0006] Although existing technologies have made significant progress in the systematic and scientific aspects of low-carbon land use optimization methods, there are still some limitations:

[0007] (1) Existing technologies have the limitations of being out of touch with practice, lacking policy adaptability, and lacking planning coordination. First, although existing technologies perform well in theoretical research and model simulation when formulating land use optimization plans, they often ignore the constraints of policies and regulations in practical applications, resulting in high policy risks in the actual implementation of optimization plans, difficulty in obtaining policy support and legal guarantees, lack of consideration of specific practical conditions, and carbon reduction effects that are often less than expected. Second, many existing technology optimization plans lack coordination with other land use needs and planning, which can easily lead to waste of resources, duplicate construction, and planning conflicts, increasing the cost and time of new construction for land adjustment.

[0008] (2) Existing technologies have limitations in carbon emission assessment and optimization, such as poor parameter applicability, narrow assessment scope, and lack of dynamic prediction. First, some existing technologies directly refer to carbon emission parameters in other regions, failing to accurately reflect the specific economic activities, energy use, and land use characteristics of the region, resulting in the possibility that they may not be fully applicable to the actual situation in a specific region, thereby affecting the scientific nature and applicability of the optimization scheme. In addition, existing technologies often only quantitatively evaluate and determine the carbon emission coefficient of construction land, ignoring the carbon emission characteristics of other land use types such as cultivated land and forest land, resulting in incomplete optimization results. In addition, some studies rely on historical data to calculate the current carbon emission coefficient and optimize it based on this, lacking dynamic prediction of future carbon emission coefficients, which ignores possible changes and uncertainties, making it difficult to adapt to future needs and challenges, and resulting in the inability to provide forward-looking and sustainable optimization schemes.

[0009] (3) The current low-carbon land space optimization technology fails to integrate the advantages of multiple models, resulting in limitations in prediction accuracy, spatial analysis, resource optimization and dynamic simulation. First, the use of a single model can often only solve specific problems and lacks overall coordination and comprehensiveness. Second, the existing technology is insufficient in dealing with complex situations in practical applications, and it is difficult to fully consider and coordinate various policies, regulations, economic, social and environmental factors. In addition, the lack of dynamic prediction capabilities makes it difficult for existing technologies to adapt to future changes and uncertainties, resulting in a lack of foresight in the optimization scheme. Due to these limitations, the existing low-carbon land use optimization scheme has obvious deficiencies in scientificity, applicability and execution, and it is difficult to fully meet the needs of low-carbon land space planning.

[0010] Through the above analysis, the problems and defects of the existing technology are: the current low-carbon land use optimization technology is out of touch with practice, lacks policy adaptability and planning coordination, carbon emission assessment and optimization parameters have poor applicability, narrow assessment scope, and lack of dynamic prediction, which ignores possible changes and uncertainties, making it difficult to provide forward-looking and sustainable optimization solutions, and thus difficult to achieve long-term sustainable development. There is a lack of a national land space pattern simulation and prediction method that coordinates spatial planning and carbon reduction control. Summary of the invention

[0011] In response to the problems existing in the prior art mentioned in the background technology, the present invention proposes a national land space pattern simulation and prediction method that coordinates spatial planning and carbon reduction control. By comprehensively coordinating the constraints on land use indicators and layout of national land space planning, it not only reduces policy risks and problems in actual implementation, but also provides a more accurate and dynamic carbon emission optimization plan, thereby improving the feasibility, flexibility, scientificity and sustainability of low-carbon land use optimization plans.

[0012] Technical solution: In order to solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0013] A method for simulating and predicting the spatial pattern of land for coordinated spatial planning and carbon reduction control, comprising the following steps:

[0014] S1: Collect spatial data and statistical data, pre-process the collected data, and establish a basic database for low-carbon land use optimization in coordination with national land space planning;

[0015] S1-1: Collect spatial vector and raster data on land use, natural ecology, and transportation accessibility, as well as statistical data on social economy, spatial planning, statistical yearbooks, and IPCC reports;

[0016] S1-2: Reclassify the land use data into cultivated land, forest land, grassland, water area, construction land and unused land to generate the reclassified land use data;

[0017] S1-3: Use the Euclidean distance formula to calculate the distance from the regional grid to the driving factors of land use change, and standardize the traffic accessibility data and natural ecological and socio-economic spatial data to obtain standardized spatial data of driving factors;

[0018] S1-4: Assign values ​​to the planning and control space data, set water areas, basic farmland and ecological protection red lines as restricted areas, and generate land use constraint data;

[0019] S1-5: Use GIS software tools to generate spatial raster data with consistent spatial resolution, range, and projection coordinate system;

[0020] S1-6: Establish a raster database and upload all pre-processed raster data to the database for storage;

[0021] S1-7: Organize, summarize and process socio-economic data, statistical yearbooks, spatial planning texts and IPCC scientific and technological reports to form spatial planning land use indicator datasets, energy consumption datasets and socio-economic development datasets;

[0022] S2: Combine energy carbon emissions with spatial mapping technology to calculate annual land use carbon emissions, and use the GM (1,1) model to predict the land use carbon emission coefficient in the target year;

[0023] S2-1: Calculate the energy consumption of production and residential sectors based on the energy balance sheet;

[0024] S2-2: Calculate the carbon emissions of energy consumption in each department based on the type of energy consumption;

[0025] S2-3: Aggregate energy carbon emissions from various sectors to form industry energy carbon emissions;

[0026] S2-4: Establish a mapping relationship between industry carbon emissions and national land space;

[0027] S2-5: Calculate the carbon emissions per unit area of ​​construction land and agricultural land year by year;

[0028] S2-6: Use the output value ratio to refine the carbon emissions per unit area of ​​various agricultural lands year by year;

[0029] S2-7: Use the GM(1,1) model to predict the land use carbon emission coefficient in the target year;

[0030] S3: Based on Markov and linear programming models, combined with natural development trends and national land space planning requirements, predict and optimize the quantitative structure of low-carbon land use;

[0031] S3-1: Set the baseline development scenario and use the Markov model to predict the amount of land demand;

[0032] S3-2: Set low-carbon national land space planning scenarios and construct a low-carbon land use linear planning model;

[0033] The linear programming model consists of three parts: decision variables, objective function, and constraints. It is a decision model used to optimize a linear target to achieve the optimal value under a set of linear constraints. Its expression is as follows:

[0034]

[0035] Among them, f(x) is the objective function, which represents the minimum land use carbon emission; Ci is the carbon emission coefficient of the i-th type of land per unit area; S i is the i-th type of decision variable; n represents the total number of land use types; st represents the constraint condition, a ij is the coefficient corresponding to the i-th type of land use in the j-th constraint; m represents the total number of constraints; Q i is the constraint value;

[0036] S3-3: Determine the objective function for minimizing carbon emissions from land use;

[0037] S3-4: In combination with the land use index requirements of national land space planning, set constraints on the scale of various types of land use;

[0038] S3-5: Combine the objective function of minimizing carbon emissions from land use with the constraints of various land use indicators in national land space planning, use the low-carbon land use linear programming model to optimize the land quantity and structure, and solve and determine the land demand under the low-carbon national land space planning scenario;

[0039] S4: Use the PLUS model to simulate and predict land use patterns under different scenarios;

[0040] S4-1: Convert land use data and land use constraint data into unsigned character format;

[0041] S4-2: Based on the land expansion extraction module, extract the land expansion data and identify the spatial locations of various types of land expansion;

[0042] S4-3: Randomly sample spatial data of land expansion and driving factors, use the random forest algorithm to train the data set, and combine the land expansion analysis strategy module of the PLUS model to calculate the development probability of various types of land use;

[0043] S4-4: Use the expansion ratio of various types of land to set neighborhood weight parameters, and further revise them in combination with the land use indicators of national land space planning;

[0044] S4-5: Set the adaptive inertia coefficient;

[0045] S4-6: Calculate the development probability of the overall land use type;

[0046] S4-7: Differentiated restricted areas are set up based on natural development trends and national land space planning requirements;

[0047] S4-8: Using the PLUS model based on the cellular automaton module of multiple random patch seeds to simulate the land use change of the generated patches;

[0048] S4-9: When the area of ​​all land types reaches the requirement, the land use simulation stops iterating and outputs the simulated land use pattern results;

[0049] S4-10: Compare the simulated land use results with the actual land use pattern, and use Kappa coefficient and numerical merit to verify the accuracy;

[0050] S4-11: Repeat S4-4 to S4-9, and predict the land use pattern under the future baseline development scenario based on the final land use data, the future land demand obtained in S3-1, and the restricted areas;

[0051] S4-12: Repeat S4-4 to S4-9, and predict the land use pattern under the future low-carbon national land space planning scenario based on the final land use data, the future land demand obtained in S3-5, the neighborhood weight factor, the conversion cost matrix, and the restricted area;

[0052] S5: Evaluate the carbon reduction benefits of low-carbon land use optimization methods in collaboration with spatial planning, and propose multi-scale low-carbon decision-making and optimization recommendations;

[0053] S5-1: Compare the differences in carbon emissions from land use between the baseline development scenario and the low-carbon national land space planning scenario, and assess the carbon reduction potential of low-carbon land use optimization;

[0054]

[0055] Among them, ΔE is the difference in carbon emissions between the baseline development scenario and the low-carbon land space planning scenario; E base is the carbon emissions under the baseline development scenario; E plan Carbon emissions under low-carbon national land space planning scenarios; C i is the carbon emission coefficient of the i-th type of land use; is the land area of ​​category i under the baseline development scenario; is the area of ​​land of type i under the low-carbon national land space planning scenario; n is the number of land types; a ΔE value greater than 0 indicates that the low-carbon land use optimization method has carbon reduction potential;

[0056] S5-2: Compare the spatial differences in carbon emissions from land use between the baseline development scenario and the low-carbon national land spatial planning scenario, identify grid cells with carbon reduction potential, and facilitate the implementation of land use conversion systems and land use optimization measures;

[0057] The specific formula is as follows:

[0058]

[0059] Among them, ΔE j is the difference in carbon emissions between the baseline development scenario and the low-carbon national land space planning scenario for grid cell j; is the carbon emission of grid cell j under the baseline development scenario; is the carbon emission of grid cell j under the low-carbon national land space planning scenario; C i is the carbon emission coefficient of the i-th type of land use; is the land area of ​​the i-th category in grid cell j under the baseline development scenario; is the i-th land area of ​​grid unit j under the low-carbon national land space planning scenario; n is the number of land types; ΔE j A value greater than 0 indicates that grid cell j has carbon reduction potential;

[0060] S5-3: Count the carbon reduction potential of each district and county, use hot spot analysis to identify hotspots with the best carbon reduction effects, and select low-carbon demonstration areas at the county level;

[0061] The specific formula for calculating the carbon reduction potential of each district and county is as follows:

[0062]

[0063] Among them, R d represents the total carbon reduction potential of the dth district or county; ΔE d,j represents the carbon reduction potential of the jth grid in the dth district and county; m d represents the total number of grids in the dth district or county;

[0064] The specific hot spot analysis formula is as follows:

[0065]

[0066] Among them, G d * Getis-OrdG d * Hot spot analysis index; m is the total number of evaluated district and county units; d and f are the dth and fth evaluated district and county units, respectively; W df is the spatial weight matrix of unit d and unit f; R d , R f are the observed values ​​of evaluation units d and f respectively; E(G d * ) and Var(G d * ) are G d * The mathematical expectation and variance of Z(G d * ) represents the Z score, when Z(G d * )>+1.96, it indicates that the district or county is a hotspot area with high carbon reduction potential; when Z(G d * )<-1.96, it indicates that the district or county is a cold spot area with low carbon reduction potential;

[0067] S5-4: Calculate the carbon reduction potential of land use optimization in each city as a performance evaluation indicator for municipal land space planning.

[0068] As a preferred method, in S4-4, the neighborhood weight parameters are set using the expansion ratios of various types of land use, and the specific contents of further revisions are combined with the land use indicators of national land space planning:

[0069] The neighborhood weight parameter indicates the degree to which the land use type of an area is affected by the surrounding areas. The specific formula is as follows:

[0070]

[0071] in, It is represented as the neighborhood effect of grid cell j being transformed into land use type i at time node of iteration t; It is represented as the number of spatial grids of the i-th type of land use within the rectangular N×N window at the last iteration t-1 time node, W i represents the neighborhood weight factor of the i-th type of land use;

[0072] Based on the expansion patch data of each land use type, the expansion proportion of each land use type is calculated, and the obtained proportion data is used as the neighborhood weight parameter. The specific formula is as follows:

[0073]

[0074] Among them, W i represents the neighborhood weight factor of the i-th type of land; EXP i represents the expansion area of ​​the i-th type of land use; n is the total number of land use types;

[0075] The parameter setting of the neighborhood factor in the low-carbon national land space planning scenario should not only consider the historical land use expansion trend, but also need to refer to the requirements of the national land space planning for land use indicators for correction. The specific neighborhood weight factor correction formula is as follows:

[0076]

[0077] Among them, W′ i represents the neighborhood weight factor of the i-th type of land in the low-carbon national land space planning scenario; EXP i represents the expansion area of ​​the i-th type of land; n is the total number of land types; LSI i Indicates the planning indicator limit value of the i-th type of land; CSI i Represents the current area of ​​the i-th type of land.

[0078] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0079] (1) Collaborative policy planning improves the feasibility and efficiency of the implementation of optimization plans. National land space planning is the basis of various development and protection work and is considered to be the most effective policy tool for achieving land use optimization goals. First, land use optimization measures coordinated with national land space planning will obtain policy support, meet planning requirements, and be protected by laws and regulations, reducing implementation difficulties and policy risks, and ensuring the feasibility and coordination of various measures in actual operations. Secondly, the optimization plan coordinated with national land space planning alleviates conflicts with other land use needs and plans, avoids waste of resources and duplication of construction, and can make full use of existing infrastructure, reduce the cost and time of new construction of land adjustment, and further improve the implementation efficiency of land optimization measures. Finally, the optimization plan based on national land space planning is closer to reality and more forward-looking, and can dynamically adapt to future development changes, which will more effectively promote the realization of carbon neutrality goals.

[0080] (2) Use local data to ensure the accuracy and comprehensiveness of the carbon emission coefficient. First, the local carbon emission coefficient is determined by quantitative methods instead of directly referring to the parameters of other regions, which ensures the accuracy and reliability of the data. Secondly, the use of local data such as the Jiangsu Provincial Energy Balance Sheet can better reflect the specific economic activities and energy use in the region, and provide more accurate carbon emission estimates, making the optimization plan more practical, more credible and scientific. Finally, not only the carbon emission coefficient of construction land is quantitatively estimated, but also the carbon emission coefficients of various land use types such as cultivated land and forest land, which fully covers the carbon emission characteristics of different land use forms and ensures the comprehensiveness of the optimization results.

[0081] (3) Ensure the foresight and sustainability of the optimization scheme through historical year-by-year data and prediction models. First, the carbon emission coefficients of various types of land use are calculated year by year in the historical period to avoid the errors of average values ​​or single-year data, ensuring the accuracy of the data and the reliability of the trend. Secondly, the optimization scheme not only focuses on the present, but also takes into account future development trends. It uses spatial mapping technology and the GM (1,1) model to dynamically predict the future land use carbon emission coefficient, which not only captures the changing trend of the carbon emission coefficient in the time series, but also reduces the impact of abnormal data through smoothing, thereby improving the accuracy of the carbon emission coefficient prediction. By dynamically predicting the future carbon emission coefficient and taking into account possible changes and uncertainties, it ensures that the optimization scheme can adapt to future changes and development needs, and provides more forward-looking and scientific prediction results and decision-making recommendations.

[0082] (4) The advantages of integrated spatial mapping technology, GM (1,1) model, linear programming model and PLUS model ensure the scientific nature and flexibility of the optimization scheme. Spatial mapping technology can accurately match the carbon emissions of industries to the land space and ensure that the land structure optimization under the carbon reduction target has stronger applicability; GM (1,1) model is suitable for small samples and systems with large uncertainties, and can predict carbon emission data with high precision, enhancing the foresight and accuracy of the scheme; linear programming model is used for resource allocation and optimization problems to determine the optimal low-carbon land use structure and proportion, achieve efficient resource utilization and minimize carbon emissions, and improve the rationality and effectiveness of the scheme; PLUS model can simulate the changes in the spatial pattern of land use under different scenarios, comprehensively consider the influence of multiple driving factors, systematically optimize land resources and make reasonable layout. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] Figure 1 It is a schematic diagram of the overall technical route of the present invention;

[0084] Figure 2 It is a flow chart of the whole of the present invention;

[0085] Figure 3 is a flow chart of S1 of the present invention;

[0086] Figure 4 is a flow chart of S2 of the present invention;

[0087] Figure 5 It is a structural diagram of the mapping relationship between industry carbon emissions and national land space of the present invention;

[0088] Figure 6 is a flow chart of S3 of the present invention;

[0089] Figure 7 is a flow chart of S4 of the present invention;

[0090] Figure 8 is a flow chart of S5 of the present invention;

[0091] Fig. 9 This is a schematic diagram of the spatial distribution of driving factors for the simulation and prediction of the land space pattern of Jiangsu Province;

[0092] Fig.10 This is a schematic diagram of the land use pattern of Jiangsu Province in 2015 and 2020;

[0093] Fig.11 It is a schematic diagram of the spatial pattern of land expansion in Jiangsu Province from 2015 to 2020;

[0094] Fig.12 It is a schematic diagram of the probability distribution pattern of suitability of various types of land use in Jiangsu Province;

[0095] Fig.13 It is a schematic diagram of the simulated land use pattern of Jiangsu Province in 2020, the land use pattern of the baseline development scenario in 2035, and the land use pattern of the low-carbon national land space planning scenario in 2035;

[0096] Fig.14 It is a schematic diagram of the spatial pattern of land use optimization and carbon reduction potential at the city level in Jiangsu Province, the spatial pattern of hotspots of land use optimization and carbon reduction potential at the district and county level, and the spatial pattern of land use optimization and carbon reduction potential at the grid level. DETAILED DESCRIPTION

[0097] The present invention is further illustrated below in conjunction with specific examples. The examples are implemented based on the technical solutions of the present invention. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention.

[0098] like Figure 1-14 As shown, the land space pattern simulation prediction method for coordinated spatial planning and carbon reduction control provided in this embodiment includes the following steps:

[0099] S1: Collect spatial data and statistical data, pre-process the collected data, and establish a basic database for low-carbon land use optimization in coordination with national land space planning;

[0100] like Figure 3 As shown, it includes S1-1 to S1-7.

[0101] S1-1: Collect spatial vector and raster data on land use, natural ecology, transportation accessibility, and other statistical data such as social economy, spatial planning, statistical yearbooks, and IPCC reports;

[0102] S1-2: Reclassify the land use data into six types: cultivated land, forest land, grassland, water area, construction land and unused land, and generate the reclassified land use data;

[0103] S1-3: Use the Euclidean distance formula to calculate the distance from the regional grid to the driving factors of land use change, and standardize the traffic accessibility data and natural ecological and socio-economic spatial data to obtain standardized spatial data of driving factors;

[0104] The Euclidean distance formula is used to calculate the distance between the regional grid and the driving factor of land use change. The calculation formula is:

[0105]

[0106] Among them, d j,p represents the Euclidean distance between the jth grid cell and the pth driving factor point; (x j ,yj ) represents the coordinates of the jth grid cell; (x p ,y p ) are the coordinates of the pth driving factor point.

[0107] The positive and negative indicators are standardized using the range method to obtain standardized driving factor spatial data. The calculation formula for the standardized processing of positive indicators is as follows:

[0108]

[0109] The calculation formula for normalizing negative indicators is as follows:

[0110]

[0111] Among them, X j ' is the normalized index pixel value of grid cell j; X j is the original index pixel value of grid cell j before normalization; max{X1,X2,…,X m} is the maximum value of all pixel values ​​of the indicator; min{X1,X2,…,X m} is the minimum value of all pixel values ​​of the indicator.

[0112] S1-4: Assign values ​​to the planning and control space data, set water areas, basic farmland and ecological protection red lines as restricted areas, and generate land use constraint data;

[0113] S1-5: Use Geography Information System (GIS) software tools to generate spatial raster data with consistent spatial resolution, range and projection coordinate system;

[0114] S1-6: Establish a raster database and upload all pre-processed raster data to the database for storage;

[0115] S1-7: Organize, summarize and process socio-economic data, statistical yearbooks, spatial planning texts and scientific reports of the Intergovernmental Panel on Climate Change (IPCC) to form spatial planning land use indicator datasets, energy consumption datasets such as oil and natural gas, and socio-economic development datasets such as population, economy and energy carbon emission factors.

[0116] S2: Combine energy carbon emissions with spatial mapping technology to calculate annual land use carbon emissions, and use the GM (1,1) model to predict the land use carbon emission coefficient in the target year;

[0117] like Figure 4As shown, it includes S2-1 to S2-7.

[0118] S2-1: Calculate the energy consumption of production and residential sectors based on the energy balance sheet;

[0119] The IPCC energy consumption method is used to calculate the different types of energy consumption of 45 production sectors and 2 residential sectors in Jiangsu Province, that is, the mass balance of fuels produced in Jiangsu Province for energy production, trade, international fuel and inventory changes. The specific formula is as follows:

[0120] FC k =IP k +IM k -EX k +IN k -OUT k ±SC k -LOSS k -NEU,

[0121] Among them, FC k represents the total energy consumption of the kth fossil fuel, where k represents the category of fossil energy; IP k represents the background production of the kth fossil energy; IM k represents the import volume of the kth type of fossil energy; EX k represents the export volume of the kth type of fossil energy; IN k represents the total amount of the kth type of fossil energy imported from other provinces; OUT k represents the total amount of the kth type of fossil energy exported to other provinces; SC k represents the change in the k-th fossil energy inventory; LOSS k represents the loss rate of the kth type of fossil energy; NEU represents non-energy usage.

[0122] S2-2: Calculate the carbon emissions of energy consumption in each department based on the type of energy consumption;

[0123] The specific formula for estimating fossil fuel-related carbon dioxide emissions by energy type is as follows:

[0124]

[0125] Among them, C p represents the total carbon emissions from various fossil energy consumption in the p department of Jiangsu Province, p≤47; FC k represents the total energy consumption of the kth type of fossil fuel; SCE k Indicates the standard coal coefficient of the kth energy type; LHV k Indicates the average lower calorific value of the kth type of energy; CUE k Indicates the carbon content of the k-type energy unit calorific value; COR kIt represents the product of carbon oxidation rates of the kth type of energy; 44 / 12 represents the ratio of the molecular weight of CO2 to carbon.

[0126] In this embodiment, the calculation of each coefficient refers to the calculation methods and coefficients in the "IPCC National Greenhouse Gas List Guidelines", "General Rules for Comprehensive Energy Consumption Calculation (GB / T2589-2020)" and "Provincial Greenhouse Gas Inventory Compilation Guidelines".

[0127] S2-3: Aggregate energy carbon emissions from various sectors to form industry energy carbon emissions;

[0128] Since the carbon emissions of specific sectors cannot be directly matched with land use types, it is necessary to aggregate the carbon emissions of different sectors into industry carbon emissions.

[0129]

[0130] Among them, C q represents the total carbon emissions of industry q, C p It represents the total carbon emissions generated by various fossil energy consumption in p sectors of Jiangsu Province, q≤7. The industry and sector classification is shown in Table 1 below.

[0131] Table 1 Sector-industry aggregation classification mapping ratio table

[0132]

[0133]

[0134] S2-4: Establish a mapping relationship between industry carbon emissions and national land space;

[0135] The specific mapping relationship is as follows Figure 5 As shown in the figure, the mapping relationship between industry carbon emissions and national land space is established to match industry carbon emissions with land use types. Among them, the carbon emission space of seven industries, namely agriculture, mineral mining, industrial energy, transportation, service industry, urban life and rural life, is mapped to two types of land, construction land and agricultural land, forming a data set of industry carbon emissions and national land space mapping.

[0136] S2-5: Calculate the carbon emissions per unit area of ​​construction land and agricultural land year by year;

[0137] By dividing the total carbon emissions of construction land each year by the land area each year, the carbon emissions per unit area are obtained. The specific formula is as follows:

[0138]

[0139] Among them, C urban (t) represents the carbon emission per unit area of ​​construction land in year t; E urban(t) represents the total carbon emissions of construction land in year t; A urban (t) represents the construction land area in year t.

[0140] By dividing the total carbon emissions from agricultural land each year by the land area each year, the carbon emissions per unit area are obtained. The specific formula is as follows:

[0141]

[0142] Among them, C agri (t) represents the carbon emissions per unit area of ​​agricultural land in year t; E agri (t) represents the total carbon emissions from agricultural land in year t; A agri (t) represents the area of ​​agricultural land in year t.

[0143] S2-6: Use the output value ratio to refine the carbon emissions per unit area of ​​various agricultural lands year by year;

[0144] According to the proportion of the output value of agriculture, forestry, animal husbandry and fishery in the total output value each year, the carbon emissions per unit area of ​​agricultural land each year are allocated to various agricultural land spaces such as cultivated land, forest land, grassland and water area. The specific formula is as follows:

[0145]

[0146] Among them, C i (t) represents the carbon emission per unit area of ​​the i-th type of land in year t; C agri (t) represents the carbon emission per unit area of ​​agricultural land in year t; P i (t) represents the output value corresponding to the i-th type of agricultural land in year t; n represents the number of agricultural land types. In this embodiment, the agricultural land type parameter is set to 4, including 4 types of land: cultivated land, forest land, grassland, and water area.

[0147] S2-7: Use the GM(1,1) model to predict the land use carbon emission coefficient in the target year;

[0148] For each land use type, is the original data sequence of carbon emissions per unit area of ​​the i-th type of land use, that is,

[0149] Where n is the number of data points and generates a cumulative sequence That The calculation formula is:

[0150]

[0151] in, represents the kth item in the original data sequence; Represents the kth item in the accumulated data sequence.

[0152] Establish the GM(1,1) model, whose differential equation is:

[0153]

[0154] Among them, a i and b i is the parameter that needs to be estimated; Represents a cumulative sequence The derivative of ; the discrete form is:

[0155]

[0156] in, Generate a sequence for the adjacent values ​​of the cumulative sequence. The calculation formula is:

[0157]

[0158] Among them, a i b is the parameter that needs to be estimated, reflecting the inherent development trend of the system; i are the parameters that need to be estimated, reflecting the external driving factors of the system; Generates the kth item in the sequence for the neighboring values ​​of the accumulated sequence.

[0159] Solving for parameter a i and b i , according to the least squares method, establish the matrix equation:

[0160]

[0161] Among them, B i Y is the constructed matrix, which contains the data of the sequence generated by the cumulative sequence of neighbor values; i The constructed vector contains the data of the original data sequence.

[0162] Calculate the parameter vector and

[0163]

[0164] in, is the parameter a estimated by the least squares method i ; is the parameter b estimated by the least squares method i ; is the matrix B i The transposed matrix of .

[0165] Establish a grey prediction model based on known parameters and The prediction formula of the GM(1,1) model is:

[0166]

[0167] in, is the k+1th item in the predicted cumulative sequence; is an exponential decay term, reflecting the decay trend of the system.

[0168] The predicted value of the original data is obtained by cumulative reduction, specifically:

[0169]

[0170] in, is the k+1th item in the original data sequence to be predicted.

[0171] According to the above calculation steps, the carbon emission coefficients of various land uses in the target year (such as year n+1) are predicted to be:

[0172]

[0173] Among them, C i (n+1) is the carbon emission coefficient per unit area of ​​the i-th type of land in the predicted target year (unit: 10,000 tons / km 2 ); It is the n+1th item in the predicted original data sequence, that is, the carbon emission coefficient per unit area of ​​the i-th type of land over the years.

[0174] S3: Based on Markov and linear programming models, combined with natural development trends and national land space planning requirements, predict and optimize the quantitative structure of low-carbon land use;

[0175] like Figure 6 As shown, it includes S3-1 to S3-5.

[0176] S3-1: Set the baseline development scenario and use the Markov model to predict the amount of land demand;

[0177] Based on the land use data in 2015 and 2020, the land use transition matrix is ​​extracted. A Markov model is constructed and iterative calculations are performed to predict the transition probability and demand quantity of each land use type in the future. The specific formula is as follows:

[0178]

[0179] Among them, S (t) is the state of the random event at time t; S (t+1) is the state of the random event at time t+1; P ijis the transfer matrix, with values ​​ranging from 0 to 1; n is the number of land types.

[0180] Based on the Markov model, the land use transition probability matrix from 2015 to 2020 is obtained, as shown in Table 2:

[0181] Table 2 Land use transition probability matrix from 2015 to 2020

[0182] arable land woodland grassland Waters Construction land Unused land arable land 0.973490 0.001666 0.000011 0.006033 0.018800 0 woodland 0.105357 0.892120 0.000148 0.000148 0.002226 0 grassland 0.395833 0.041667 0.375000 0 0.166667 0.020833 Waters 0.103078 0 0 0.878997 0.017926 0 Construction land 0 0 0 0.003851 0.996090 0 Unused land 0 0 0 0 0.363636 0.454545

[0183] Based on the land use data in 2020 and the land use transition probability matrix from 2015 to 2020, the amount of various types of land under the baseline development scenario in 2035 is predicted, as shown in Table 3 below:

[0184] Table 3 Land quantity structure under the baseline development scenario in 2035

[0185] Land Type arable land woodland grassland Waters Construction land Unused land <![CDATA[Area / km 2 > 68416.55 1489.56 1.79 9070.30 22993.31 0.26

[0186] S3-2: Set low-carbon national land space planning scenarios and construct a low-carbon land use linear planning model;

[0187] The linear programming model is a mathematical optimization model that contains three parts: decision variables, objective functions, and constraints. It is a decision model used to optimize a linear target to achieve the optimal value under a set of linear constraints. Its expression is as follows:

[0188]

[0189] Among them, S i is the i-th decision variable, i.e., the i-th land area, i=1,2,...,n; n represents the total number of land types, n=6; where: S1 is cultivated land, S2 is forest land, S3 is grassland, S4 is water body, S5 is construction land, and S6 is unused land; f(x) is the objective function, which represents the minimum land use carbon emission; C i The carbon emission coefficient of the i-th type of land per unit area; in the constraint condition st, a ij is the coefficient corresponding to the i-th type of land use in the j-th constraint; m represents the total number of constraints; Q i is the constraint value.

[0190] S3-3: Determine the objective function for minimizing carbon emissions from land use;

[0191] In this embodiment, the low-carbon national land space planning scenario aims to minimize carbon emissions from land use, and the objective function of national land space carbon emissions is obtained by cumulatively calculating the area of ​​each type of land use multiplied by the carbon emission coefficient. The specific formula is as follows:

[0192] f(x)=0.0349×S1+0.0016×S2+0.0059×S3+0.0188×S4+7.1613×S5+0.0102×S6→min,

[0193] Among them, S i is the area of ​​land use type i, S1, S2, S3, S4, S5 and S6 are the areas of cultivated land, forest land, grassland, water area, construction land and unused land respectively.

[0194] S3-4: In combination with the land use index requirements of national land space planning, set constraints on the scale of various types of land use;

[0195] The constraints include total area constraints, carbon emission constraints, and constraints on the scale of various land use indicators.

[0196] The total area constraint is set based on the current land area. The study area is Jiangsu Province, with a total area of ​​101,971.77 km 2 .

[0197] The carbon emissions are based on the total amount that can be achieved in the future under the current trend as the maximum value. The cumulative calculation is 167,222.6675 million tons of carbon by multiplying the land area of ​​various types predicted by the baseline development scenario by the land use carbon emission coefficient.

[0198] Spatial planning is the government's arrangement of the future land use structure, and there are corresponding rigid requirements for each land type. Therefore, the area constraints of each type of land are strictly based on the growth and restriction range of various land use indicators in the national land space planning. However, the government planning spatial data year by year is difficult to obtain. This embodiment uses the annual land cover data of China at 30m from 1985 to 2023 released by Professors Yang Jie and Huang Xin of Wuhan University in the prior art, and the data format is raster data (.tif). Data source: Jie Yang, & Xin Huang. The 30-meter annual land cover datasets and its dynamics in China from 1985 to 2023 [J]. Earth System Science Data, 2024, 13(1): 3907–3925. https: / / doi.org / 10.5281 / zenodo.12779975 (Yang Jie, & Huang Xin. The 30-meter annual land cover datasets and its dynamics in China from 1985 to 2023. Published in Earth System Science Data, 2024, 1.0.3, Vol. 13, No. 1, pp. 3907-3925. URL: https: / / zenodo.org / records / 12779975).

[0199] This method can effectively solve the problem of mismatch between the two sets of data. The limit value of the planning indicator constraint is converted into a proportional expression. The specific formula is as follows:

[0200]

[0201] Among them, R i Indicates the conversion of the planning indicator limit value into a proportional value; LSI i Indicates the planning indicator limit value of the i-th type of land in the spatial planning text; CSI i Represents the current area of ​​the i-th type of land in the spatial planning text.

[0202] The names and constraints of various land use planning indicators corresponding to 2021-2035 in the "Jiangsu Provincial Land and Space Planning (2021-2035)" were collected and sorted out, namely, planning indicators such as cultivated land reserves, forest coverage, water space reserves, and urban development boundary expansion multiples, and converted into spatial planning indicator constraints in proportional form, and the "limit value" was converted into the form of "xx% of the current value". The specific information is shown in Table 4:

[0203] Table 4 Constraints on the proportion of various land use indicators in the national land space planning from 2021 to 2035

[0204]

[0205] Although there are no planning indicators matching forest land and unused land, in accordance with the requirements of the national land space planning to attach importance to ecological protection, the scale of grassland (S3) should not be less than the current value in 2020. The change ratio of the unused land area can be obtained by subtracting the planning indicator limit values ​​of cultivated land, forest land, grassland, water area and construction land from the total area of ​​Jiangsu Province, and then dividing it by the current value of unused land.

[0206] According to the planning indicators, the limit value is converted into a proportional value and multiplied by the current data of the land area of ​​each type in this embodiment, and the constraint conditions of various land use indicators under the low-carbon national land space planning scenario can be obtained. The specific information is as follows:

[0207] According to the requirements of the national land space planning for cultivated land and the constraints of the planning index ratio, the cultivated land area (S1) should be no less than 68450.34km 2 According to the forest coverage requirement, the forest area (S2) should be no less than 1636.37km 2 ; Grassland area (S3) not less than 5.64km 2 According to the water space reserve requirements, the water area (S4) shall not be less than 10471.18km 2; The growth trend of construction land is difficult to change, so the minimum area of ​​construction land (S5) should be greater than the total construction land in 2020 (18720.64.44km2) 2 , but according to the urban development boundary expansion multiple requirements, the maximum area (S5) does not exceed 24336.83km 2 ; National land space planning puts forward higher requirements for land use and development efficiency, so the amount of unused land (S6) will gradually decrease and will not exceed 0.71km 2 .

[0208] Based on the total area and carbon emission constraints, combined with the 2035 national land space planning requirements for various land use scales, comprehensive land use scale constraints are determined. The specific information is as follows:

[0209]

[0210] Among them, st represents the constraint condition, S i is the area of ​​land use type i, S1, S2, S3, S4, S5, S6 are cultivated land, forest land, grassland, water area, construction land, and unused land respectively; C i Represents the carbon emission coefficient per unit area of ​​the i-th type of land.

[0211] S3-5: Combining the objective function of minimizing carbon emissions from land use with the constraints of various land use indicators in national land space planning, the linear programming model is used to optimize the quantity and structure of land, and the land demand under the low-carbon national land space planning scenario is determined.

[0212] The amount of land for various types of use in the low-carbon national land space planning scenario in 2035 is shown in Table 5:

[0213] Table 5 Land quantity structure in the low-carbon national land space planning scenario in 2035

[0214]

[0215]

[0216] S4: Use the PLUS model to simulate and predict land use patterns under different scenarios;

[0217] Convert land use data and constraint data into a specific format, analyze land expansion characteristics, and use the random forest algorithm combined with the LEAS module of the PLUS model to calculate the probability of land development; set parameters such as land demand, conversion cost matrix, and neighborhood weight, and use the CARS module of the PLUS model to simulate land use changes; output simulation results and verify accuracy; predict land use patterns under future baseline development scenarios and spatial planning scenarios.

[0218] like Figure 7 As shown, it includes S4-1 to S4-12.

[0219] S4-1: Convert land use data and land use constraint data into unsigned character format;

[0220] S4-2: Extract land expansion data based on the Extract Land Expansion module and identify the spatial locations of various types of land expansion;

[0221] S4-3: Randomly sample spatial data of land expansion and driving factors, use the random forest algorithm to train the data set, and combine the Land Expansion Analysis Strategy (LEAS) module of the PLUS model to calculate the development probability of various types of land use;

[0222] The specific formula for the suitability development probability of each land use type is as follows:

[0223]

[0224] in, represents the growth probability of land use type i output at spatial unit j; I(h c (v) = d) represents the indicator function of the decision tree set; v represents the vector composed of driving factors; h c (v) represents the land prediction type of the cth decision tree when it is a vector v; M represents the number of decision trees; d takes the value of 0 or 1. When d=1, it means that other land types are converted into land type i, and when d=0, it means that no land type is converted into the i-th land type.

[0225] S4-4: Use the expansion ratio of various types of land to set neighborhood weight parameters, and further revise them in combination with the land use indicators of national land space planning;

[0226] The neighborhood weight parameter indicates the degree to which the land use type of an area is affected by the surrounding areas. The specific formula is as follows:

[0227]

[0228] in, It is represented as the neighborhood effect of grid cell j being transformed into land use type i at time node of iteration t; W represents the number of spatial grids of the i-th type of land use within the rectangular N×N window range at the last iteration t-1 time node. In this embodiment, the neighborhood range is set to 3×3. i Represented as weight, value range 0≤W i ≤1, when W iThe closer the value is to 1, the stronger the spatial expansion capacity of the i-th type of land use.

[0229] Based on the expansion patch data of each land use type, the expansion proportion of each land use type is calculated, and the obtained proportion data is used as the neighborhood weight parameter. The specific formula is as follows:

[0230]

[0231] Among them, W i represents the neighborhood weight factor of the i-th type of land; EXP i represents the expansion area of ​​the i-th type of land use; n is the total number of land use types.

[0232] The value range is 0-1. The larger the value, the greater the neighborhood influence. The specific parameters are shown in Table 6:

[0233] Table 6 Neighborhood weight parameters for each land use type in the baseline development scenario

[0234] Land Type arable land woodland grassland Waters Wetlands Construction land Unused land Neighborhood weight 0.4037 0.0325 0.0003 0.1355 0.4280 0.0001 0.4037

[0235] The parameter setting of the neighborhood factor in the low-carbon national land space planning scenario should not only consider the historical land use expansion trend, but also need to refer to the requirements of the national land space planning for land use indicators for correction. The specific neighborhood weight factor correction formula is as follows:

[0236]

[0237] Among them, W i ' represents the neighborhood weight factor of the i-th type of land under the low-carbon national land space planning scenario; EXP i represents the expansion area of ​​the i-th type of land; n is the total number of land types; LSI i Indicates the planning indicator limit value of the i-th type of land; CSI i Represents the current area of ​​the i-th type of land.

[0238] The specific settings of neighborhood weight parameters for each land use type in the low-carbon national spatial planning scenario are shown in Table 7:

[0239] Table 7 Neighborhood weight parameters for each land use type in low-carbon national spatial planning scenarios

[0240] Land Type arable land woodland grassland Waters Wetlands Construction land Unused land Neighborhood weight 0.3938 0.0320 0.0003 0.1243 0.5564 0.0000 0.3938

[0241] S4-5: Set the adaptive inertia coefficient;

[0242] The adaptive inertia coefficient can be appropriately adjusted according to the difference between the land quantity demand and the actual land quantity during the iteration process, so that the land use quantity meets the future land demand quantity, thereby realizing the simulation of land use spatial change. The specific formula is as follows:

[0243]

[0244] in, It is expressed as the inertia coefficient of the i-th land use type at time node t; and They represent the difference between the actual land quantity and the land grid demand quantity at time nodes t-1 and t-2 respectively.

[0245] S4-6: Calculate the development probability of the overall land use type;

[0246] Calculate the overall probability of land use type, the specific formula is as follows:

[0247]

[0248] in, Expressed as the overall probability of land use type i, It is expressed as the growth probability of land type i in grid cell j; It is represented as the neighborhood effect of grid cell j being transformed into land use type i at time node of iteration t; It represents the impact of land demand on the i-th type of land and is the adaptive inertia coefficient.

[0249] A roulette wheel is constructed based on the overall probabilities of all land classes to select the land use state for the next iteration.

[0250] The PLUS model simulates the patch generation in land use change based on the random patch generation mechanism of threshold reduction. This mechanism can better simulate the expansion of different land use types in the process of land change. According to the threshold reduction rule in the competition process, the spontaneous growth process of each land use type is constrained. The specific formula is as follows:

[0251]

[0252] In the formula, Expressed as the overall probability of land use type i in grid cell j; It is expressed as the growth probability of land type i in grid cell j; γ is a random value ranging from 0 to 1; μ i is the threshold for generating new land use patches for land type i; It is represented as the neighborhood effect of grid cell j being transformed into land use type i at time node of iteration t; Expressed as the inertia coefficient of the i-th land use type at time node t.

[0253] A seed may generate a new land use type and grow into a new patch formed by a group of grids with the same land use type. In order to control the generation of multiple land use patches, a decreasing threshold rule based on a competition process is proposed to limit the organic and spontaneous growth of all land use types. If a new land use type wins a round of competition, the candidate land use type i selected by the roulette wheel is evaluated using a decreasing threshold as follows:

[0254]

[0255] in, and represents the difference between the actual amount of land converted to type z land use at time nodes t-1 and t and the amount of land grid demand; step represents the step size of the model simulation approaching the total land use demand, τ represents the land use growth threshold; δ represents the attenuation coefficient of the attenuation threshold τ, ranging from 0 to 1 (set to 0.5 in this embodiment); δ l represents the attenuation coefficient of the attenuation threshold when the attenuation level is l; r1 represents a random value of a normal distribution with a mean value equal to 1; l represents the attenuation level of the threshold; TM i,z It is expressed as a transition transfer matrix, which specifically indicates whether the i-th type of land use can be transformed into the z-th type of land use. The matrix defines the difficulty of conversion between different land use types. 1 means that conversion can occur, and 0 means that conversion cannot occur.

[0256] Considering the actual situation of land use transformation in the study area in the past period, the specific setting of the land conversion cost matrix under the baseline development scenario is shown in Table 8:

[0257] Table 8 Land conversion cost matrix for baseline development scenario

[0258]

[0259]

[0260] Combined with the actual situation of land use transformation in the past period and the requirements of land use conversion in national land space planning, the specific setting of the land conversion cost matrix under the low-carbon national land space planning scenario is shown in Table 9:

[0261] Table 9 Land conversion cost matrix for low-carbon national spatial planning scenarios

[0262] Land Type arable land woodland grassland Waters Construction land Unused land arable land 1 1 1 0 1 0 woodland 1 1 1 0 1 0 grassland 1 1 1 0 1 0 Waters 0 0 0 1 0 0 Construction land 1 1 1 1 1 0 Unused land 1 1 1 1 1 1

[0263] S4-7: Differentiated restricted areas are set up based on natural development trends and national land space planning requirements;

[0264] Under the baseline development scenario, water areas are difficult to convert to other land use types, so water areas are set as restricted areas. According to the ecological protection requirements of national land space planning and the goal of ensuring food security, ecological protection red lines and permanent basic farmland are used as rigid constraints, and land use conversion within this range is prohibited.

[0265] S4-8: Use the PLUS model CA-based on multiple random patch seeds (CARS) module to simulate the land use change of patches;

[0266] S4-9: When the area of ​​all land types reaches the requirement, the land use simulation stops iterating and outputs the simulated land use pattern results;

[0267] S4-10: Compare the simulated land use results with the actual land use pattern, and use Kappa coefficient and Figure of Merit (FoM) to verify the accuracy;

[0268] The Kappa coefficient is a commonly used classification accuracy indicator that reflects the consistency between the simulation results and the real pattern, taking into account the possibility of random matching. The calculation formula is as follows:

[0269]

[0270] The value range of Kappa is between 0 and 1, close to 1 means high consistency, close to 0 means low consistency; P0 represents the correct proportion of simulation results; P i It indicates the correct proportion of predictions under the random state of the model. The Kappa coefficient of this simulation result is 90.25%. The Kappa coefficient is greater than 75%, and the simulation accuracy is relatively high.

[0271] The reliability of the FoM coefficient test accuracy is higher than that of the Kappa coefficient. It is usually used to verify the consistency of the spatial position changes of the simulation results. The calculation formula is as follows:

[0272]

[0273] Among them, FoM value range is between 0-1; FN represents the area where the land use actually changes, but the model simulates that the area has not changed; TP represents the area where the model simulates the same as the actual area; FP represents the error area caused by the model simulating the wrong land use type; TN represents the area where the land use actually has not changed, but the model simulates that the area has changed. In this embodiment, the simulation time interval is 5 years, FoM = 0.08, and the results simulated by the model are of high accuracy.

[0274] S4-11: Repeat S4-4 to S4-9, and predict the land use pattern under the future baseline development scenario based on the land use data at the end of the year (2020), the future land demand obtained in S3-1 (baseline development scenario in 2035), and the restricted areas;

[0275] S4-12: Repeat S4-4 to S4-9, and predict the land use pattern under the future low-carbon national land space planning scenario based on the final land use data (2020), the future land demand obtained in S3-5 (low-carbon national land space planning scenario in 2035), the neighborhood weight factor, the conversion cost matrix and the restricted area.

[0276] S5: Evaluate the carbon reduction benefits of the national spatial pattern of coordinated spatial planning and carbon reduction control, and propose multi-scale low-carbon decision-making and optimization suggestions;

[0277] like Figure 8 As shown, it includes S5-1 to S5-4.

[0278] S5-1: Compare the differences in carbon emissions from land use between the baseline development scenario and the low-carbon national land space planning scenario, and assess the carbon reduction potential of low-carbon land use optimization;

[0279] The specific formula is as follows:

[0280]

[0281] Among them, ΔE is the difference in carbon emissions between the baseline development scenario and the low-carbon land space planning scenario; E base is the carbon emissions under the baseline development scenario; E plan Carbon emissions under low-carbon national land space planning scenarios; C i is the carbon emission coefficient of the i-th type of land use; is the land area of ​​category i under the baseline development scenario; is the area of ​​land of type i under the low-carbon national land space planning scenario; n is the number of land types. A ΔE value greater than 0 indicates that the low-carbon land use optimization method has carbon reduction potential.

[0282] S5-2: Compare the spatial differences in land use carbon emissions between the baseline development scenario and the low-carbon national land spatial planning scenario, identify grid cells with carbon reduction potential, and facilitate the implementation of land use conversion systems and land use optimization measures;

[0283] The specific formula is as follows:

[0284]

[0285] Among them, ΔE j is the difference between the land carbon emissions in the baseline development scenario and the low-carbon national land space planning scenario for grid cell j; is the carbon emission of grid cell j under the baseline development scenario; is the carbon emission of grid cell j under the low-carbon national land space planning scenario; C i is the carbon emission coefficient of the i-th type of land use; is the land area of ​​the i-th category in grid cell j under the baseline development scenario; is the i-th type of land area of ​​grid unit j under the low-carbon national land space planning scenario; n is the number of land types. ΔE j A value greater than 0 indicates that grid cell j has carbon reduction potential.

[0286] S5-3: Count the carbon reduction potential of each district and county, use hot spot analysis to identify hotspots with the best carbon reduction effects, and select low-carbon demonstration areas at the county level;

[0287] The specific formula for calculating the carbon reduction potential of each district and county is as follows:

[0288]

[0289] Among them, R d represents the total carbon reduction potential of the dth district or county; ΔE d,j represents the carbon reduction potential of the jth grid in the dth district and county; m d Represents the total number of grids in the dth district or county.

[0290] The specific hot spot analysis formula is as follows:

[0291]

[0292] Among them, G d * Getis-OrdG d * Hot spot analysis index; m is the total number of evaluated district and county units; d and f are the dth and fth evaluated district and county units respectively; W df is the spatial weight matrix of unit d and unit f; R d , R f are the observed values ​​of evaluation units d and f respectively; E(Gd * ) and Var(G d * ) are G d * The mathematical expectation and variance of Z(G d * ) represents the Z score, which is the multiple of the standard deviation. d * )>+1.96, it indicates that the district or county is a hotspot area with high carbon reduction potential; when Z(G d * )<-1.96, it indicates that the district or county is a cold spot area with low carbon reduction potential.

[0293] S5-4: Count the carbon reduction potential of land use optimization in each city and include it in the performance evaluation indicators of municipal land and space planning.

[0294] The specific formula for calculating the carbon reduction potential of each city is as follows:

[0295]

[0296] Among them, R c represents the total carbon reduction potential of the cth city; ΔE c,j represents the carbon reduction potential of the jth grid in the cth city; g c Represents the total number of grids in the cth city.

[0297] The results show that by 2035, carbon emissions under the low-carbon land space planning scenario will be 136,656,850 tons, a decrease of 305,658,200 tons compared to 167,222,670 tons under the baseline development scenario, indicating that low-carbon land use methods in coordination with land space planning have significant carbon reduction effects. These findings provide important references for policymakers, emphasize the importance of incorporating low-carbon land use optimization methods into land space planning, and provide a scientific basis for the optimization and implementation of future policies.

[0298] The results at the grid scale show that high-potential areas present a distribution characteristic of "overall dispersion and local agglomeration". The areas with high carbon reduction potential are generally scattered and have not formed an obvious continuous block distribution, but the concentrated distribution characteristics of multiple high carbon reduction potential grids in local areas are obvious, which reflects the significant effect of local land use structure and spatial optimization. Taking optimization measures (such as ecological restoration and sustainable agriculture) in these areas can effectively improve the carbon reduction potential.

[0299] The results at the county scale show that hotspots with high carbon reduction potential are mainly concentrated in the north and east, while there are some cold spots and insignificant areas in the central and southern regions, which reflects the differences in land use optimization and carbon reduction potential among different counties. The carbon reduction potential of hotspots is significantly higher than that of other regions, mainly distributed in 12 districts and counties, including Sucheng District, Suqian City, Fengxian County, Xuzhou City, Qidong City, Nantong City, and Guannan County, Lianyungang City. The land use optimization measures implemented in these areas will show significant advantages and have the potential to become low-carbon land use optimization demonstration areas. In the future, these high-potential agglomeration areas can be used as demonstration areas to summarize their successful land use optimization measures and policies and promote them to other similar areas to achieve a wider range of carbon reduction effects.

[0300] The results at the municipal scale show that the carbon reduction potential presents a pattern of "high in the north, low in the south, high in the east, and low in the south" with obvious differentiation characteristics. Regions with high carbon reduction potential may be located in areas with relatively backward economic development, with relatively abundant natural resources and less land development intensity, and with greater space for land use transformation. Regions with low carbon reduction potential are mainly concentrated in economically developed areas, where land resources are highly intensively used and there is less space for land use transformation, resulting in limited carbon reduction potential for land use structure adjustment. For example, Xuzhou City has the highest carbon reduction benefit, which can reach 32.305 million tons, while Zhenjiang City has a negative carbon reduction benefit, increasing carbon emissions by 870,700 tons. In the future, these quantitative carbon reduction potentials can be incorporated into the carbon emission performance assessment indicators of municipal land space planning. By quantifying and specifying the carbon emission reduction targets of each city, it can promote local governments to take more active low-carbon land use measures and improve the overall carbon reduction effect. It is also helpful to scientifically and rationally formulate low-carbon development strategies in land space planning to ensure the realization of carbon reduction targets.

[0301] The present invention proposes a simulation and prediction scheme for the spatial pattern of land and space that coordinates spatial planning and carbon reduction control, aiming to address the deficiencies of existing technologies in terms of policy adaptability and practical conditions. By systematically integrating the latest national spatial planning requirements, the land use indicators and layout constraints in the national spatial planning are embedded in the process of optimizing the land use quantity structure and spatial layout, reducing the policy risks of the scheme in actual implementation, ensuring policy support and legal protection, and improving the carbon reduction effect. By meeting the implementation requirements of the national spatial planning, and then coordinating other land use needs and planning, it is possible to avoid waste of resources and duplicate construction, thereby reducing the new construction cost and time of land optimization layout.

[0302] The present invention constructs an optimization scheme for accurately evaluating regional localized carbon emissions, aiming to solve the problems of poor parameter applicability, narrow evaluation scope, lack of dynamic prediction, etc. in existing technologies in carbon emission assessment and optimization. By adopting carbon emission parameters suitable for local economic activities, energy use and land use characteristics, the carbon emission characteristics of various land use types including construction land, cultivated land, forest land, etc. are comprehensively evaluated. A dynamic prediction method for future carbon emission coefficients that couples spatial mapping technology and GM (1,1) model is established to capture the changing trend of carbon emission coefficients in time series, dynamically adjust future carbon emission coefficients, and adapt to possible changes. At the same time, make full use of existing historical data for reasonable prediction, avoiding errors in average values ​​or single year data. By smoothing to reduce the impact of abnormal data, the accuracy of carbon emission coefficient prediction is improved, providing more forward-looking and scientific prediction results.

[0303] The present invention proposes a new method for simulating and predicting the spatial pattern of low-carbon land that integrates the advantages of spatial mapping technology, GM (1,1) model, linear programming model and PLUS model, aiming to provide a more accurate, comprehensive and dynamic optimization scheme, and improve the scientificity, applicability and foresight of the scheme. Spatial mapping technology can accurately match the land use type by mapping the carbon emissions of industries to the land space, ensuring the applicability of land structure optimization under the carbon reduction target; GM (1,1) model is suitable for systems with small samples and large uncertainties, and can predict carbon emission data with high precision, enhancing the foresight and accuracy of the scheme; linear programming model is used for resource allocation and optimization problems, determining the optimal low-carbon land use structure and proportion, achieving efficient resource utilization and minimization of carbon emissions, and improving the rationality and effectiveness of the scheme; PLUS model can simulate the changes in the spatial pattern of land use under different scenarios, comprehensively consider the influence of multiple driving factors, systematically optimize land resources and reasonably layout.

[0304] The present invention aims to overcome the limitations of the existing technology, and by strictly following the constraints of land use indicators and layout in national land space planning, the coupling of spatial mapping technology, GM (1,1) model, linear programming model and PLUS model can improve the scientificity, applicability and foresight of low-carbon land use optimization schemes. This can not only reduce policy risks and problems in actual implementation, but also provide more accurate and dynamic carbon emission optimization schemes, thereby improving the flexibility, scientificity and sustainability of low-carbon land use optimization decisions, and promoting low-carbon development and efficient use of resources.

[0305] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for simulating and predicting the spatial pattern of land for coordinated spatial planning and carbon reduction control, characterized by: The following steps are involved: S1: Collect spatial data and statistical data, pre-process the collected data, and establish a basic database for low-carbon land use optimization in coordination with national land space planning; S1-1: Collect spatial vector and raster data on land use, natural ecology, and transportation accessibility, as well as statistical data on social economy, spatial planning, statistical yearbooks, and IPCC reports; S1-2: Reclassify the land use data into cultivated land, forest land, grassland, water area, construction land and unused land to generate the reclassified land use data; S1-3: Use the Euclidean distance formula to calculate the distance from the regional grid to the driving factors of land use change, and standardize the traffic accessibility data and natural ecological and socio-economic spatial data to obtain standardized spatial data of driving factors; S1-4: Assign values ​​to the planning and control space data, set water areas, basic farmland and ecological protection red lines as restricted areas, and generate land use constraint data; S1-5: Use GIS software tools to generate spatial raster data with consistent spatial resolution, range, and projection coordinate system; S1-6: Establish a raster database and upload all pre-processed raster data to the database for storage; S1-7: Organize, summarize and process social and economic data, statistical yearbooks, spatial planning texts and IPCC scientific and technological reports to form spatial planning land use indicator data sets, energy consumption data sets and social and economic development data sets; S2: Combine energy carbon emissions with spatial mapping technology to calculate annual land use carbon emissions, and use the GM (1,1) model to predict the land use carbon emission coefficient in the target year; S2-1: Calculate the energy consumption of production and residential sectors based on the energy balance sheet; S2-2: Calculate the carbon emissions of energy consumption in each department based on the type of energy consumption; S2-3: Aggregate energy carbon emissions from various sectors to form industry energy carbon emissions; S2-4: Establish a mapping relationship between industry carbon emissions and national land space; S2-5: Calculate the carbon emissions per unit area of ​​construction land and agricultural land year by year; S2-6: Use the output value ratio to refine the carbon emissions per unit area of ​​various agricultural lands year by year; S2-7: Use the GM(1,1) model to predict the land use carbon emission coefficient in the target year; S3: Based on Markov and linear programming models, combined with natural development trends and national land space planning requirements, predict and optimize the quantitative structure of low-carbon land use; S3-1: Set the baseline development scenario and use the Markov model to predict the amount of land demand; S3-2: Set low-carbon national land space planning scenarios and construct a low-carbon land use linear planning model; The linear programming model consists of three parts: decision variables, objective function, and constraints. It is a decision model used to optimize a linear target to achieve the optimal value while satisfying a set of linear constraints. Its expression is as follows: Among them, f(x) is the objective function, which represents the minimum land use carbon emission; C i is the carbon emission coefficient of the i-th type of land per unit area; S i is the i-th type of decision variable; n represents the total number of land use types; st represents the constraint condition, a ij is the coefficient corresponding to the i-th type of land use in the j-th constraint; m represents the total number of constraints; Q i is the constraint value; S3-3: Determine the objective function for minimizing carbon emissions from land use; S3-4: In combination with the land use index requirements of national land space planning, set constraints on the scale of various types of land use; S3-5: Combine the objective function of minimizing carbon emissions from land use with the constraints of various land use indicators in national land space planning, use the low-carbon land use linear programming model to optimize the land quantity and structure, and solve and determine the land demand under the low-carbon national land space planning scenario; S4: Use the PLUS model to simulate and predict land use patterns under different scenarios; S4-1: Convert land use data and land use constraint data into unsigned character format; S4-2: Based on the land expansion extraction module, extract the land expansion data and identify the spatial locations of various types of land expansion; S4-3: Randomly sample spatial data of land expansion and driving factors, use the random forest algorithm to train the data set, and combine the land expansion analysis strategy module of the PLUS model to calculate the development probability of various types of land use; S4-4: Use the expansion ratio of various types of land to set neighborhood weight parameters, and further revise them in combination with the land use indicators of national land space planning; S4-5: Set the adaptive inertia coefficient; S4-6: Calculate the development probability of the overall land use type; S4-7: Differentiated restricted areas are set up based on natural development trends and national land space planning requirements; S4-8: Using the PLUS model based on the cellular automaton module of multiple random patch seeds to simulate the land use change of the generated patches; S4-9: When the area of ​​all land types reaches the requirement, the land use simulation stops iterating and outputs the simulated land use pattern results; S4-10: Compare the simulated land use results with the actual land use pattern, and use Kappa coefficient and numerical merit to verify the accuracy; S4-11: Repeat S4-4 to S4-9, and predict the land use pattern under the future baseline development scenario based on the final land use data, the future land demand obtained in S3-1, and the restricted areas; S4-12: Repeat S4-4 to S4-9, and predict the land use pattern under the future low-carbon national land space planning scenario based on the final land use data, the future land demand obtained in S3-5, the neighborhood weight factor, the conversion cost matrix, and the restricted area; S5: Evaluate the carbon reduction benefits of low-carbon land use optimization methods in collaboration with spatial planning, and propose multi-scale low-carbon decision-making and optimization recommendations; S5-1: Compare the differences in carbon emissions from land use between the baseline development scenario and the low-carbon national land space planning scenario, and assess the carbon reduction potential of low-carbon land use optimization; Among them, ΔE is the difference in carbon emissions between the baseline development scenario and the low-carbon land space planning scenario; E base is the carbon emissions under the baseline development scenario; E plan Carbon emissions under low-carbon national land space planning scenarios; C i is the carbon emission coefficient of the i-th type of land use; is the land area of ​​category i under the baseline development scenario; is the area of ​​land of type i under the low-carbon national land space planning scenario; n is the number of land types; a ΔE value greater than 0 indicates that the low-carbon land use optimization method has carbon reduction potential; S5-2: Compare the spatial differences in carbon emissions from land use between the baseline development scenario and the low-carbon national land spatial planning scenario, identify grid cells with carbon reduction potential, and facilitate the implementation of land use conversion systems and land use optimization measures; The specific formula is as follows: Among them, ΔE j is the difference in carbon emissions between the baseline development scenario and the low-carbon national land space planning scenario for grid cell j; is the carbon emission of grid cell j under the baseline development scenario; is the carbon emission of grid cell j under the low-carbon national land space planning scenario; C i is the carbon emission coefficient of the i-th type of land use; is the land area of ​​the i-th category in grid cell j under the baseline development scenario; is the i-th land area of ​​grid unit j under the low-carbon national land space planning scenario; n is the number of land types; ΔE j A value greater than 0 indicates that grid cell j has carbon reduction potential; S5-3: Count the carbon reduction potential of each district and county, use hot spot analysis to identify hotspots with the best carbon reduction effects, and select low-carbon demonstration areas at the county level; The specific formula for calculating the carbon reduction potential of each district and county is as follows: Among them, R d represents the total carbon reduction potential of the dth district or county; ΔE d,j represents the carbon reduction potential of the jth grid in the dth district and county; m d represents the total number of grids in the dth district or county; The specific hot spot analysis formula is as follows: Among them, G d * Getis-OrdG d * Hot spot analysis index; m is the total number of evaluated district and county units; d and f are the dth and fth evaluated district and county units, respectively; W df is the spatial weight matrix of unit d and unit f; R d , R f are the observed values ​​of evaluation units d and f respectively; E(G d * ) and Var(G d * ) are G d * The mathematical expectation and variance of Z(G d * ) represents the Z score, when Z(G d * )>+1.96, it indicates that the district or county is a hotspot area with high carbon reduction potential; when Z(G d * )<-1.96, it indicates that the district or county is a cold spot area with low carbon reduction potential; S5-4: Calculate the carbon reduction potential of land use optimization in each city as a performance evaluation indicator for municipal land space planning.

2. The method for simulating and predicting the spatial pattern of land for coordinated spatial planning and carbon reduction control according to claim 1 is characterized by: In S4-4, the neighborhood weight parameters are set using the expansion ratios of various types of land use, and the specific contents of further revisions are combined with the land use indicators of national land space planning: The neighborhood weight parameter indicates the degree to which the land use type of an area is affected by the surrounding areas. The specific formula is as follows: in, It is represented as the neighborhood effect of grid cell j being transformed into land use type i at time node of iteration t; It is represented as the number of spatial grids of the i-th type of land use within the rectangular N×N window at the last iteration t-1 time node, W i represents the neighborhood weight factor of the i-th type of land use; Based on the expansion patch data of each land use type, the expansion proportion of each land use type is calculated, and the obtained proportion data is used as the neighborhood weight parameter. The specific formula is as follows: Among them, W i represents the neighborhood weight factor of the i-th type of land; EXP i represents the expansion area of ​​the i-th type of land use; n is the total number of land use types; The parameter setting of the neighborhood factor in the low-carbon national land space planning scenario should not only consider the historical land use expansion trend, but also need to refer to the requirements of the national land space planning for land use indicators for correction. The specific neighborhood weight factor correction formula is as follows: Among them, W i ' represents the neighborhood weight factor of the i-th type of land under the low-carbon national land space planning scenario; EXP i represents the expansion area of ​​the i-th type of land; n is the total number of land types; LSI i Indicates the planning indicator limit value of the i-th type of land; CSI i Represents the current area of ​​the i-th type of land.

Images