A method for optimizing the allocation of cultivated and uncultivated land in salinized irrigation areas

Abstract

The present invention discloses a method for optimizing the allocation of cultivated and uncultivated land in salinized irrigation areas, mainly including two parts: the quantity demand of land use and the spatial layout of cultivated and uncultivated land. By using the method of coupling the water-salt balance model SALTMOD and the land use CLUE-S model, the suitable cultivated and uncultivated ratio based on irrigation and drainage regulation and arable land salt control is combined with the spatial distribution of land use, and a proposal for a suitable spatial dynamic regulation plan for saline wasteland is systematically explored from the aspects of quantity constraint and spatial allocation, considering land use development and regional salt income and expenditure balance, providing certain technical and data support for the treatment of saline-alkali land in arid and semi-arid irrigation areas.

Description

Technical Field

[0001] The present invention belongs to the technical field of optimizing the land use landscape pattern in saline-alkali land treatment, and particularly relates to a method for optimizing the spatial allocation of cultivated and uncultivated land in saline-alkali irrigation areas. Background Technique

[0002] There is a close interaction relationship between the land use spatial pattern and the spatio-temporal evolution of water resources and the redistribution of salt. With the continuous in-depth study of the spatial distribution characteristics, pattern and change process of land use, the importance of land use landscape and spatial optimization allocation has become increasingly prominent. This optimization allocation not only involves the distribution of land use quantity and spatial layout, but also needs to comprehensively consider the combination of ecological processes and land use unit configuration. In arid and semi-arid irrigation areas with alternating cultivated and uncultivated land, saline-alkali wasteland, as a special land use type, the comprehensive consideration of its reasonable spatial layout and ecological salt-accepting function is of great significance for soil salinization control and the sustainable utilization and development of land resources.

[0003] The evolution of the land use spatial pattern of cultivated and uncultivated land is a complex dynamic process, which is affected by a variety of factors. At present, there are few technical methods for optimizing the spatial allocation of cultivated and uncultivated land under the conditions of water and salt migration. How to combine the development of this special land use of "saline-alkali wasteland" with the control of soil salinity in cultivated land and use quantitative methods to optimize the land use structure of saline-alkali wasteland in space, and truly achieve the method and technology of "visible and calculable" are not yet mature. Summary of the Invention

[0004] Aiming at the above deficiencies in the prior art, the method for optimizing the allocation of cultivated and uncultivated land in saline-alkali irrigation areas provided by the present invention overcomes the above deficiencies in the prior art, and can start from the control of soil salinity in irrigation area cultivated land and the salt-accepting dynamics and spatial pattern development of saline-alkali wasteland, and propose regional appropriate suggestions for the spatial dynamic development of saline-alkali wasteland from the aspects of the quantity of saline-alkali wasteland (cultivated and uncultivated ratio), spatial location (spatial pattern), and salt dynamics (water-salt balance), providing certain technical and data support for the treatment of saline-alkali land in arid and semi-arid irrigation areas.

[0005] In order to achieve the above invention purpose, the technical solution adopted by the present invention is: a method for optimizing the allocation of cultivated and uncultivated land in saline-alkali irrigation areas, including the following steps:

[0006] S1. In the non-spatial analysis module of the CLUE-S land use model, construct a water-salt balance SALTMOD model to simulate and calculate the land use quantity demand of the research area, and calculate the area and quantity demand of land use types with an annual step by determining the most suitable cultivated and uncultivated ratio in the research area;

[0007] S2. In the spatial analysis module of the CLUE-S land use model, find the influencing factors of the spatial distribution of different land use types, and combine the calculated area and quantity requirements of land use types. Through running the CLUE-S land use model, simulate and obtain the most suitable spatial layout of different land use grid cells;

[0008] S3. Compare and analyze the simulation results of the CLUE-S land use model with the real results, test the simulation accuracy of the CLUE-S land use model, and optimize the CLUE-S land use model according to the test results;

[0009] S4. Use the calculated most suitable cultivated-saline wasteland ratio as the input data for optimizing the CLUE-S land use model, predict the future long-term land use spatial pattern, analyze the spatial distribution pattern of suitable saline wasteland under the control of arable land salinity, and realize the optimized allocation of cultivated-saline wasteland land use.

[0010] Furthermore, in step S1, in the constructed water-salt balance SALTMOD model, take the cultivated land and saline wasteland in the study area as the research objects, and make the total control area only include the cultivated land and saline wasteland. Its regional division is expressed as:

[0011] A + B = cultivated land in the study area

[0012] U = saline wasteland

[0013] A + B + U = 1

[0014] Among them, A represents the irrigated land with crops or vegetation growing, B represents the fallow or rain-fed irrigated land, and U represents the non-irrigated saline wasteland;

[0015] In the water-salt balance SALTMOD model, the rotation mode of agricultural land Kr = 1. All areas of irrigated land A and B are rotated, and the area of saline wasteland U remains unchanged. And the year is divided into three simulation periods: the growth period, the autumn irrigation period, and the non-growth period.

[0016] Furthermore, in step S1, the specific method for simulating and calculating the land use quantity demand of the research area is as follows:

[0017] S11. Process and analyze the basic data input from the outside through the water-salt balance SALTMOD model, and output the corresponding data;

[0018] Among them, the input basic data includes the agricultural data, hydrological data, meteorological data, and soil data of the study area; the output data includes soil salinity, groundwater depth, drainage volume, and drainage salinity;

[0019] S12. Calibrate and verify the water-salt balance SALTMOD model using the specific parameters of the study area;

[0020] S13. Set different scenario plans based on the current irrigation and drainage conditions and different cultivated land to uncultivated land ratios in the study area, and use the calibrated and verified water-salt balance SALTMOD model to predict and simulate the dynamic changes of cultivated land salinity under different scenario plans;

[0021] S14. Take the predicted cultivated land salinity as a control constraint to determine the ratio of cultivated land to uncultivated land in the study area;

[0022] S15. According to the determined ratio of cultivated land to uncultivated land in the study area, determine the area and quantity requirements of land use types with an annual step.

[0023] Further, the specific steps of step S2 are as follows:

[0024] S21. Based on remote sensing technology, classify the land use types in the study area according to the spectral characteristics and spatial information of different land use types, and uniformly represent the reclassified land use data and driving factor data;

[0025] S22. Conduct a regression analysis of land use driving factors based on the uniformly represented land use data and driving factor data;

[0026] S23. Configure the model files, including:

[0027] Take the area and quantity requirements of land use types as input data, and configure the main parameter files of the CLUE-S land use model based on the raster-based spatial data in the prediction and simulation base period;

[0028] Configure the regression parameter files and driving force files of the CLUE-S land use model according to the results of the regression analysis of land use driving factors;

[0029] Configure the land use type file at the start year of the simulation of the CLUE-S land use model and the land use demand file from the prediction and simulation base period to the end of the simulation;

[0030] S24. Set the land use type distribution rules, including:

[0031] Set the land use transfer rules, the transfer elasticity coefficient of land use types, and the regional constraints for the optimal allocation of land use;

[0032] S25. According to the configured model files and the set land use type distribution rules, run the CLUE-S land use model, continuously iterate and calculate, and simulate the most suitable spatial layout of different land use grid cells.

[0033] Further, in the step S21, the reclassified land use data and driving factor data are represented using a unified coordinate system, projection, spatial extent, and spatial resolution, and all raster data are uniformly converted into floating-point data.

[0034] Further, the step S22 is specifically as follows:

[0035] S22-1. Determine the driving factors based on the uniformly represented land use data and driving factor data, and rasterize each driving factor data;

[0036] The driving factors include elevation, slope, NDVI, annual rainfall, annual average temperature, distance from the center point of the township, distance from major railways, highways, national, provincial, county, and township roads, distance from canal systems, distance from drainage ditches, average groundwater depth over the years, population density, and gross domestic product (GDP);

[0037] S22-2. Perform coding processing on various land use types, and reclassify each land use type map to obtain a binary map of each land type;

[0038] S22-3. Based on the binary map of each land type and the driving factor data for the same period, perform binary Logistic regression analysis on the land use spatial pattern and its driving factors for different years, and analyze to obtain the suitability value for each land use raster cell to be converted into a certain land use type.

[0039] Further, in the step 22-3, the suitability value for a land use raster cell to be converted into a certain land use type is expressed as:

[0040]

[0041] In the formula, P i is the probability of the land use type appearing in each raster cell in the region; x0, x1,..., x n-1 are the respective driving factors; β0, β1, β n-1 are the coefficients of each explanatory variable in the regression equation; n represents the number of driving factors.

[0042] Further, in the step S25, the CLUE-S land use model is run, and the most suitable spatial layout of different land use raster cells is simulated and expressed as:

[0043] TPROP u,i = ELAS i + ITER i + P u,i

[0044] In the formula, TPROP u,iDenote the overall suitability of grid u for land use type i, ELAS i Denote the transfer elasticity coefficient of land type i, representing the conversion cost of land use types, ITER i Denote the competition factor of land type i, which is automatically set during the iterative process of model simulation, P u,i Denote the distribution probability of grid u for i, which is the regression result of the current land use situation on different driving force factors.

[0045] Furthermore, the method for testing the simulation accuracy of the CLUE-S land use model is specifically as follows:

[0046] Perform spatial overlay processing on the true land use type and the simulated land use type. Through the raster calculator, overlay and subtract the remote sensing classification of land use types in different years and the simulated land use spatial classification map, extract the number of raster cells with a value of 0 as the number of correctly simulated raster cells, and analyze the proportion of it in the total number of raster cells;

[0047] Perform spatial comparative analysis on the remote sensing data of land use types in different years and the simulated data of the corresponding simulated land use types, and calculate the Kappa coefficient;

[0048] Based on the proportion of the number of correctly simulated raster cells in the total number of raster cells and the Kappa coefficient, test the simulation accuracy of the CLUE-S land use model.

[0049] The beneficial effects of the present invention are as follows:

[0050] The present invention uses the method of coupling the water-salt balance model SALTMOD and the land use CLUE-S model, combines the suitable cultivated-wasteland ratio based on irrigation and drainage regulation and arable land salinity control with the spatial distribution of land use, and proposes a method for optimizing the spatial allocation of cultivated and wasteland land use in salinized irrigation areas. It systematically explores the recommended dynamic regulation plan for the suitable saline wasteland space considering land use development and regional salt balance from the aspects of quantity constraint and spatial allocation, providing certain technical support for the treatment of saline-alkali land and the sustainable development of maintaining the ecological environment health in arid and semi-arid irrigation areas. Description of the Drawings

[0051] Figure 1 It is the flow chart of the method for optimizing the land use of cultivated and wasteland land in salinized irrigation areas provided by the present invention.

[0052] Figure 2 It is the change trend of the cultivated-wasteland ratio in the research area from 1996 to 2020 provided by the present invention.

[0053] Figure 3 It is the dynamic change of soil water and salt in the root layer under different cultivated-wasteland ratios in the research area provided by the present invention.

[0054] Figure 4 The raster map of land use driving factors provided by the present invention.

[0055] Figure 5 The comparison map of remote sensing interpretation of land use and CLUE-S spatial pattern simulation in 2006 and 2016 provided by the present invention.

[0056] Figure 6 The comparison map of remote sensing interpretation of land use and CLUE-S spatial pattern simulation in 2006 and 2016 provided by the present invention.

[0057] Figure 7 The comparison map of remote sensing interpretation of land use and CLUE-S spatial pattern simulation in 2006 and 2016 provided by the present invention. Detailed implementation manners

[0058] The following describes the detailed implementation manners of the present invention to facilitate those skilled in the art to understand the present invention. However, it should be clear that the present invention is not limited to the scope of the detailed implementation manners. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions created using the concept of the present invention are within the scope of protection.

[0059] The embodiment of the present invention provides a method for optimizing the allocation of cultivated and uncultivated land in salinized irrigation areas, as Figure 1 shown, including the following steps:

[0060] S1. In the non-spatial analysis module of the CLUE-S land use model, construct a water-salt balance SALTMOD model to simulate and calculate the quantity demand of land use in the research area, and calculate the area and quantity demand of land use types in annual steps by determining the most suitable ratio of cultivated and uncultivated land in the research area;

[0061] S2. In the spatial analysis module of the CLUE-S land use model, find the influencing factors of the spatial distribution of different land use types, and combine the calculated area and quantity demand of land use types to simulate and obtain the most suitable spatial layout of different land use grid cells by running the CLUE-S land use model;

[0062] S3. Compare and analyze the simulation results of the CLUE-S land use model with the actual results, test the simulation accuracy of the CLUE-S land use model, and optimize the CLUE-S land use model according to the test results;

[0063] S4. Use the calculated optimal cultivated wasteland ratio as the input data for optimizing the CLUE-S land use model, predict the long-term future land use spatial pattern, analyze the spatial distribution pattern of suitable saline wasteland under the control of cultivated land salinity, and achieve the optimal allocation of cultivated and wasteland land use.

[0064] In step S1 of the embodiment of the present invention, the non-spatial analysis module of the CLUE-S land use model is the predicted value of the land demand in the study area, which needs to be directly input from the outside. Therefore, other models or methods need to be embedded in the non-spatial module for calculation. The non-spatial analysis module mainly predicts the changes in the area and quantity of land use types according to the comprehensive driving factors of land use change in the study area. In this embodiment, the main constraint is the control of the root layer soil salinity of cultivated land. The cultivated-wasteland ratio with better control of cultivated land salinity during the prediction period and the area of various land uses are used as constraint conditions. The water-salt balance model SALTMOD is used to simulate and calculate the land use quantity demand in the study area. The area and quantity sequence data of land use types with an annual time step are determined through the cultivated-wasteland ratio in the study area, and used as the input data for the spatial module of the CLUE-S land use model.

[0065] In step S1 of this embodiment, in the constructed water-salt balance SALTMOD model, the cultivated land and saline wasteland in the study area are used as the research objects, and the total control area only includes cultivated land and saline wasteland. Its regional division is expressed as:

[0066] A + B = cultivated land in the study area

[0067] U = saline wasteland

[0068] A + B + U = 1

[0069] Among them, A represents irrigated land with crops or vegetation growing, B represents fallow or rain-fed irrigated land, and U represents non-irrigated saline wasteland. Specifically, the total control area of the study area is deducted from water areas, construction land, unused land, and non-irrigated forest and grassland, and only includes cultivated land (irrigated farmland grassland and non-irrigated farmland) and saline wasteland.

[0070] In the water-salt balance SALTMOD model, the rotation mode Kr of agricultural land is 1. The A and B areas of irrigated land are all rotated, and the U area of saline wasteland remains unchanged. One year is divided into three simulation periods: the growth period, the autumn irrigation period, and the non-growth period. Exemplarily, the growth period is from April to September, the autumn irrigation period is from October to November, and the non-growth period is from December to March of the following year.

[0071] In step S1 of this embodiment, the method for simulating and calculating the land use quantity demand in the study area is specifically as follows:

[0072] S11. Process and analyze the basic data input from the outside through the water-salt balance SALTMOD model, and output the corresponding data;

[0073] Among them, the input basic data includes agricultural data, hydrological data, meteorological data, and soil data of the study area; the output data includes soil salinity, groundwater depth, drainage volume, and drainage salinity; for example, the input data can include regional soil salinity, groundwater salinity, groundwater depth, surface water hydrological data (rainfall, potential evapotranspiration, irrigation volume, surface runoff), and groundwater hydrological data (groundwater extraction volume, drainage reuse volume), etc.

[0074] S12. Calibrate and verify the water-salt balance SALTMOD model using the specific parameters of the study area;

[0075] Specifically, some input parameters in the model can be measured or estimated, and some parameters such as natural drainage volume Gn, root layer leaching rate Flr, transition layer leaching rate Flx, and natural groundwater drainage rate Gn through the aquifer are difficult to directly obtain through experimental measurement. The appropriate parameter values are determined by fitting the model simulation values with the measured values.

[0076] S13. Based on the current irrigation and drainage conditions of the study area and different cultivated land to uncultivated land ratios, set different scenario schemes, and use the calibrated and verified water-salt balance SALTMOD model to predict and simulate the dynamic changes of cultivated land salinity under different scenario schemes;

[0077] S14. Take the predicted cultivated land salinity as a control constraint to determine the ratio of cultivated land to uncultivated land in the study area;

[0078] S15. According to the determined ratio of cultivated land to uncultivated land in the study area, determine the area and quantity requirements of land use types with an annual step.

[0079] In step S13 of this embodiment, based on the model calibrated and verified with regional specific parameters, carry out scenario scheme simulation and prediction; taking the 2016 current situation as an example, based on the current irrigation and drainage conditions and different cultivated land to uncultivated land ratio scenario schemes, analyze the future long-term change trend of cultivated land soil salinity. As shown in Table 1, carry out scenario scheme setting;

[0080] Table 1: Scenario scheme setting

[0081] Scheme Setting Cultivated Waste Ratio 2016 Status Quo 6.185 Scheme 1 5.500 Scheme 2 6.000 Scheme 3 6.500 Scheme 4 7.000 Scheme 5 7.500 Scheme 6 8.000 Scheme 7 8.500 Scheme 8 9.000 Scheme 9 9.500 Scheme 10 10.000

[0082] As Figure 2 shown, the ratio of cultivated land to uncultivated land in the study area showed a slow upward trend from 1996 to 2020. In 2016, the cultivated land to uncultivated land ratio in the study area was about 6.185. Under the current irrigation and drainage conditions, the salt content in the 1m deep soil layer of cultivated land in the study area showed a slight decreasing trend, and the root layer salinity of cultivated land could be reduced by about 1.03% after 10 years.

[0083] As Figure 3 shown, for the above-mentioned scheme, assuming that the cultivated land area continues to expand in the future, the area of reclaimed saline wasteland increases, the area of saline wasteland continues to decrease, and the cultivated-to-wasteland ratio increases. The simulation results of the SALTMOD water-salt balance model show that as the cultivated-to-wasteland ratio increases, the soil salinity in the root layer of cultivated land shows a gradually increasing trend. When the cultivated-to-wasteland ratio increases to 8, the soil salinity in the regional root layer shows a slight decreasing trend. When the cultivated-to-wasteland ratio increases to 8.5, the soil salinity in the regional root layer shows an increasing trend. Assuming that the irrigation and drainage conditions in the irrigation area remain unchanged in the future, and comprehensively considering the salt control factors in the root layer of cultivated land in the irrigation area, it is proposed that under the background of the development of land use in the irrigation area, the regional cultivated-to-wasteland ratio is more suitable between 8.0 and 8.5. After the cultivated-to-wasteland ratio exceeds 8.5, in the long run, the salt accumulation in the root layer of cultivated land in the irrigation area shows an obvious increasing trend. The continuous and cyclic change of the saline wasteland distribution pattern can extend the dry salt drainage life and make the saline wasteland a permanent salt reservoir. When the irrigation and drainage conditions remain unchanged, when the saline wasteland is reduced to a certain extent, the salt drainage space in the irrigation area decreases, which is not conducive to the salt drainage of cultivated land. As shown in Table 2, it shows the change rate of the soil salinity in the root layer under different cultivated-to-wasteland ratio scenarios in 2016 compared with the current situation.

[0084] Table 2: Change rate of soil salinity in the root layer under different cultivated-to-wasteland ratio scenarios in 2016 compared with the current situation

[0085]

[0086] In step S2 of the embodiment of the present invention, in the spatial analysis module of the CLUE-S land use model, mainly based on the CLUE-S land use model, by finding the main influencing factors of the spatial distribution of different land use types, extracting its distribution rules based on Logistic regression, and using continuous iterative operations to find the most suitable spatial layout of different land use grid cells.

[0087] Specifically, step S2 of the embodiment of the present invention is specifically as follows:

[0088] S21. Based on remote sensing technology, classify the land use types in the study area according to the spectral characteristics and spatial information of different land use types, and uniformly represent the reclassified land use data and driving factor data;

[0089] S22. Conduct a regression analysis of land use driving factors according to the uniformly represented land use data and driving factor data;

[0090] S23. Configure the model file, including:

[0091] Take the land use type area and quantity requirements as input data, and configure the main parameter file of the CLUE-S land use model based on the raster-type spatial data of the prediction simulation base period;

[0092] According to the results of the regression analysis of land use driving factors, configure the regression parameter file and the driving force file of the CLUE-S land use model;

[0093] Configure the land use type file for the starting year of the CLUE-S land use model simulation and the land use demand file from the prediction simulation base period to the end of the simulation;

[0094] S24. Set the land use type distribution rules, including:

[0095] Set the land use transfer rules, the land use type transfer elastic coefficient, and the regional constraints for land use optimization allocation;

[0096] S25. According to the configured model file and the set land use type distribution rules, run the CLUE-S land use model, continuously iterate the operation, and simulate the most suitable spatial layout of different land use grid cells.

[0097] In step S21 of this embodiment, the reclassified land use data and driving factor data are represented using a unified coordinate system, projection, spatial range, and spatial resolution, and all raster data are uniformly converted to floating-point data.

[0098] Specifically, based on remote sensing technology, according to the spectral characteristics and spatial information of different land use types, classify the land use types in the study area. After reclassification, the land use data, driving factor data, etc. all use a unified coordinate system, projection, spatial range, and spatial resolution, and all raster data are uniformly converted to floating-point data.

[0099] Step S22 of this embodiment is specifically as follows:

[0100] S22-1. Determine the driving factors according to the uniformly represented land use data and driving factor data, and rasterize each driving factor data;

[0101] The driving factors include elevation, slope, NDVI, annual rainfall, annual average temperature, distance from the center point of the township, distance from the main railway, highway, national and provincial and county and township roads, distance from the canal system, distance from the drainage ditch, average annual groundwater depth, population density, and gross domestic product GDP;

[0102] S22-2. Perform coding processing on various land use types, and reclassify each land use type map to obtain the binary map of each land type;

[0103] S22-3. Based on the binary maps of each land type and the driving factor data of the same period, perform binary Logistic regression analysis on the land use spatial pattern and its driving factors in different years respectively, and obtain the suitability values for each land use grid cell to transform into a certain land use type.

[0104] In step S22-1 of this embodiment, combining the characteristics of regional land use change and the main driving forces, and considering the scientificity and representativeness of the selected driving factors, 12 driving factors in 4 categories including natural, socio-economic, location and water resources are screened to determine the above-mentioned driving factors. Based on the ArcGIS platform, rasterize each driving factor data, with the raster resampling unit size of 100m×100m, and divide the study area into 799*891 (columns * rows), a total of 246,465 grid cells. All driving factor raster data are converted into ASCII format through the ArcGIS platform and named sc1gr*.fil in sequence, where * represents the coding of the driving factor, and * takes values from 0 to 11. The above-mentioned land use driving factor rasters are as Figure 4 shown.

[0105] In step S22-2 of this embodiment, based on the above-mentioned remote sensing interpretation results of land use types, code each land use type. Cultivated land, forest land, grassland, water area, construction land, unused land, and saline wasteland are coded as 0, 1, 2, 3, 4, 5, and 6 respectively. Then, reclassify the land use type map on the ArcGIS platform to obtain the binary maps of each land type, and set the values of the binary maps to 0 and 1. 1 represents a certain land type (set the ID number of this land type in the attribute table of each vector file to 1), and 0 represents the remaining land use types except this land type (assign the ID of other land types to 0). On the premise of ensuring that all data dimensions are consistent, convert the binary maps of each land type into ASCII format files and name them cov*.asc, where * represents the coding of each land type, and * takes values from 0 to 6. See Table 3 below for the specific descriptions of land use types and driving factor files.

[0106] Table 3: Land Use Type and Driving Factor Files

[0107] File Name Land Use Type / Driving Factor cov0.asc Cultivated Land cov1.asc Forest Land cov2.asc Grassland cov3.asc Water Area cov4.asc Construction Land cov5.asc Unused Land cov6.asc Saline Waste Land Sclgr0.fil Elevation DEM Sclgr1.fil Slope Sclgr2.fil Vegetation Index NDVI Sclgr3.fil Annual Precipitation Sclgr4.fil Annual Average Temperature Sclgr5.fil Distance from the Center Point of the Township Sclgr6.fil Distance from Main Railways, Highways, and Provincial, County, and Township Roads Sclgr7.fil Distance from the Canal System Sclgr8.fil Distance from the Drainage Ditch Sclgr9.fil Average Annual Groundwater Depth (1990 - 2020) Sclgr10.fil Population Density Sclgr11.fil Gross Domestic Product GDP

[0108] In step S22-3 of this embodiment, the suitability value for a land use grid cell to transform into a certain land use type is expressed as:

[0109]

[0110] Where P i is the probability of the land use type appearing in each grid cell in the region; x0, x1,..., x n-1are driving factors; β0, β1, β n-1 are the coefficients of the explanatory variables in the regression equation; n is the number of driving factors.

[0111] Specifically, in this embodiment, based on the above land use types and driving factor data of the same period, Binary Logistic regression analysis was performed on the land use spatial pattern and its driving factors in 1996 and 2016. The ASCII code data files Cov*.0~6.asc and driving factor data files Sclgr0~11.fil of each land use type were converted into Stat.txt files based on the File Convert tool provided by the CLUE-S model. Then Stat.txt was imported into SPSS software, where 12 driving factors were placed in the independent variables and various land types were placed in the dependent variables. Binary Logistic regression analysis was performed in turn, and the analysis results are shown in Tables 4 and 5.

[0112] Table 4: Logistic regression analysis and test results in 1996

[0113]

[0114] Table 5: Logistic regression analysis and test results in 2016

[0115]

[0116]

[0117] In step S23 of this embodiment, during the process of configuring the main parameter file:

[0118] ①Main parameter file configuration

[0119] According to the relevant results of spatial analysis, the main parameters of the simulation operation are set and saved as a "txt" document. The specific parameter meanings and settings are shown in Tables 6 and 7 below.

[0120] Table 6: Input files of CLUE-S model

[0121] File Name Description mian.1 Main Settings for Editing the Model Cov_all.0 Land Use Map in the Starting Year of Simulation demand.in* Land Demand Files under Different Scenarios (* represents different land use types) ELAS Stability Parameter of Land Use Type allow.txt Editing the Land Use Conversion Matrix region*.fil Regional Constraint File (* represents different constraint files) Sclgr*.fil Simulation Driving Force File (Land Use Change Impact Factor) (* represents the serial number of the driving force) alloc.reg Editing the Regression Equation

[0122] Table 7: Main parameter settings of CLUE-S model

[0123]

[0124]

[0125] ②Regression parameter file configuration

[0126] When configuring the regression parameter file, it is calculated from parameters such as the previous driving factors and regression return coefficients. Substitute the regression coefficients of each driving factor into the regression model to obtain the regression equations for each land use type, and rename them as the regression parameter file (allocate.reg), which is saved under the installation directory and used as the input variable. Based on this, the configuration of the regression parameter files for 1996 and 2016 is shown in Tables 8 and 9;

[0127] Table 8: Settings of the regression parameter file in 1996

[0128]

[0129]

[0130]

[0131]

[0132] Table 9: Settings of the regression parameter file in 2016

[0133]

[0134]

[0135]

[0136] ③ Configuration of the driving force file

[0137] Place the ASCII code files converted from each driving factor file under the CLUE-S installation file according to the model requirements.

[0138] ④ Configuration of the land use type file

[0139] Set the attribute values of cultivated land, forest land, grassland, water area, construction land, unused land, and saline wasteland in the land use status maps of the study area in the initial years of 1996 and 2016 to 0, 1, 2, 3, 4, 5, and 6 in sequence, then convert them into 100m×100m raster files, and then reclassify the raster maps according to the codes of each land type. Finally, convert the reclassified results into ASCII format through the Conversion Tools / Raster to ASCII function of ArcGIS 10.5 software, name it cov_all.0, and input it into the CLUE-S model.

[0140] ⑤ Land use demand file

[0141] From the base period of the predictive simulation to the end of the simulation, calculate the quantity of each land use type for each year in years, and input it into the model in the txt format of a notepad as the quantity control end of the optimal allocation of land use.

[0142] In step S25 of this embodiment, the spatial analysis module uses the calculation result of the non-spatial analysis module as input data. Based on the raster-type spatial data in the base period of the predictive simulation, according to the driving factors of land use change, the constraint conditions for obtaining land use change, the land use conversion rules, etc., calculate the spatial distribution probability of each land use type, and use continuous iterative operations to find the most suitable spatial layout for different grid cells, thereby realizing the spatial simulation of landscape change.

[0143] In step S24 of this embodiment, for the setting of land use transfer rules:

[0144] The land use type transfer matrix file is a matrix that sets the parameters for whether different land use types can be converted into each other. The conversion rules between different land use types are named the allow.txt file in the model, and the file is copied to the installation directory of the CLUE-S software. The matrix includes two parameters, 0 and 1. The rows represent the current land use type, and the columns represent the possible future land types. 1 indicates that conversion can occur between the present and the future, and 0 indicates that conversion cannot occur. In this study, it is determined based on the long-time series land use remote sensing results of the region, combined with historical situations and previous knowledge and experience. At the same time, by referring to relevant literature, the initial conversion rule file is determined. All land uses in the study area can undergo transformation, so all the land use type transfer matrix files are 1. Set the allow.txt file in the form of a 7×7 matrix.

[0145] For the setting of the land use type transfer elasticity coefficient:

[0146] When conducting the optimal allocation of land use, in the CLUE-S model, it is necessary to set the conversion rules between different land types in different future scenarios, which are mainly reflected by two parameters: the land use type conversion elasticity coefficient (ELAS) and the land use type conversion file. Among them, ELAS generally reflects the ability or difficulty of converting one land use type into another. The value of the ELAS parameter ranges from 0 to 1. The larger the ELAS value, the higher the difficulty of converting to other land use types, and the higher the stability of this land type. On the contrary, it indicates lower stability and easier conversion.

[0147] The transfer elasticity settings of land types are mainly based on the actual situation of land use type conversion in the region from 1991 to 2020. After repeated simulation, screening, and adjustment, the coefficients with better model effects are selected as the final parameters. The ELAS values of the 7 land types, namely cultivated land, forest land, grassland, water area, construction land, unused land, and saline wasteland, are 0.85, 0.70, 0.55, 0.40, 0.90, 0.45, and 0.10 respectively.

[0148] For the regional constraints setting of land use optimization allocation:

[0149] The regional files include the unrestricted regional file (region_nopark.fil) and the restricted regional file (region_park*.fil). This file is obtained by assigning values to the restricted / unrestricted regions accordingly, and then converting the raster data into an ASCII code file. Assuming no restricted regions, the restricted file for the study area is a raster file with all values being 0, and this file is converted into an ASCII file named region_nopark.fil and copied to the installation directory.

[0150] In this embodiment, based on the above model file configuration and the setting of land use type distribution rules, the spatial analysis process module uses the calculation results of the non-spatial analysis module as input data. Based on the raster-based spatial data of the prediction and simulation base period, according to the driving factors of land use change, the constraint conditions for obtaining land use change, and the land use conversion rules, etc., it calculates the spatial distribution probability of each land use type, and uses continuous iterative operations to find the most suitable spatial layout for different raster cells, thereby realizing the spatial simulation of landscape change.

[0151] Therefore, in step S25 of this embodiment, when running the CLUE-S land use model, the most suitable spatial layout of different land use raster cells obtained by simulation is expressed as:

[0152] TPROP u,i =ELAS i +ITER i +P u,i

[0153] In the formula, TPROP u,i represents the overall suitability of raster u for land use type i, ELAS i represents the transfer elasticity coefficient of land type i, which represents the conversion cost of land use type, ITER i represents the competition factor of land type i, which is automatically set during the iterative process of model simulation, and P u,i represents the distribution probability of raster u for i, which is the regression result of the current land use situation on different driving force factors.

[0154] In step S3 of the embodiment of the present invention, the method for verifying the simulation accuracy of the CLUE-S land use model is specifically as follows:

[0155] Perform spatial overlay processing on the real land use types and the simulated land use types. Use the raster calculator to overlay and subtract the remote sensing classification of land use types in different years and the simulated land use spatial classification map obtained by simulation, extract the number of raster cells with a value of 0 as the number of correctly simulated raster cells, and analyze the proportion of it in the total number of raster cells;

[0156] Perform spatial comparative analysis on the remote sensing data of land use types in different years and the simulated data of the corresponding simulated land use types, and calculate the Kappa coefficient;

[0157] Based on the proportion of the number of correctly simulated raster cells in the total number of raster cells and the Kappa coefficient, verify the simulation accuracy of the CLUE-S land use model.

[0158] Specifically, in this embodiment, in order to verify the accuracy of the CLUE-S model operation results, starting from the land use in 1996 in the study area, simulate the land use in 2006 and 2016. And compare and analyze the simulation results with the real results. Specifically, the Kappa coefficient is used to verify the simulation results of the CLUE-S model, and the Kappa coefficient formula is as follows:

[0159]

[0160] In the formula, P o is the proportion of correctly simulated raster cells.

[0161] In this embodiment, the land use types are divided into 7 categories, so P c = 1 / 7; P p is the proportion of correctly simulated raster cells in the ideal state, that is, 100%. The value range of the Kappa coefficient is [-1, 1]. Generally, it is considered that Kappa > 0.6 has significant consistency, and Kappa > 0.8 indicates a good simulation effect.

[0162] Table 10: Model accuracy test evaluation table

[0163] 1996-2006 1996-2016 Number of Correctly Simulated Rasters 186979 182777 Accuracy Rate % 75.86％ 74.16％ Kappa Coefficient 0.72 0.70

[0164] The model accuracy test results obtained based on the above method are shown in Table 10 and Figure 5As shown in the figure, the land use types interpreted by remote sensing and the land use types simulated by CLUE-S are processed by spatial overlay. By using the raster calculator, the remote sensing classification maps in 2006 and 2016 are overlaid and subtracted from the CLUE-S simulated land use spatial classification maps to extract the number of raster cells with a value of 0, that is, the number of correctly simulated raster cells. The numbers of raster cells with a value of 0 in 2006 and 2016 are 186,979 and 182,777 respectively, accounting for 75.86% and 74.16% of the total raster cells. The spatial contrast analysis is carried out on the land use simulation data and the remote sensing interpreted data in two periods of 2006 and 2016, and the kappa coefficients are 0.72 and 0.70 respectively, indicating that the simulated data is relatively consistent with the current land use data interpreted by remote sensing.

[0165] In step S4 of the embodiment of the present invention, based on the appropriate cultivation and wasteland ratio proposed by the above non-spatial model simulation, it is assumed that the conversion occurs between saline wasteland and cultivated land in the next 10 years, and the areas of other land use types remain unchanged, and the spatial distribution pattern of land use in the next 10 years is simulated. The predicted results of the spatial distribution pattern of land use in the irrigation area in 2026 (as Figure 6 shown), and the spatial pattern of saline wasteland considering the salt control of cultivated land in the irrigation area (as Figure 7 shown) are obtained. From 2016 to 2026 in the next 10 years, the transfer rate of saline wasteland is 23.28%, mainly transferred to cultivated land and forest land, and the transfer rates are 18.68% and 1.56% respectively. The transfer-in rate is 6.65%, mainly transferred from cultivated land and grassland, and the transfer-in rates are 2.42% and 1.34% respectively.

[0166] In the present invention, specific embodiments are used to elaborate the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

[0167] Those of ordinary skill in the art will realize that the embodiments described here are for helping readers understand the principle of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not deviate from the essence of the present invention according to the technical revelations disclosed in the present invention, and these deformations and combinations are still within the protection scope of the present invention.

Claims

1. A method for optimizing the use of cultivated land in salinized irrigation areas, characterized in that: The following steps are involved: S1. In the non-spatial analysis module of the CLUE-S land use model, the water-salt balance SALTMOD model was constructed to simulate and calculate the land use quantity demand in the study area. The area and quantity demand of land use types with an annual step length were calculated by determining the most suitable cultivated land ratio in the study area. S2. In the spatial analysis module of the CLUE-S land use model, find the influencing factors of the spatial distribution of different land use types, and combine the calculated land use type area and quantity requirements to obtain the most suitable spatial layout of different land use grid cells by running the CLUE-S land use model simulation; S3. Compare and analyze the simulation results of the CLUE-S land use model with the actual results to verify the simulation accuracy of the CLUE-S land use model, and optimize the CLUE-S land use model based on the test results; S4. The calculated most suitable cultivated-wasteland ratio is used as input data for optimizing the CLUE-S land use model, predicting the future long-term land use spatial pattern, analyzing the spatial distribution pattern of suitable salt-wasteland under the control of cultivated land salinity, and achieving optimal allocation of cultivated-wasteland land use; The step S2 is specifically as follows: S21. Based on remote sensing technology, the land use types in the study area are classified according to the spectral characteristics and spatial information of different land use types, and the reclassified land use data and driving factor data are uniformly represented; S22. Conduct land use driving factor regression analysis based on the uniformly represented land use data and driving factor data; S23. Configuration model file, including: The land use type area and quantity requirements are used as input data, and the main parameter files of the CLUE-S land use model are configured based on the grid-type spatial data of the prediction simulation base period; According to the results of the regression analysis of land use driving factors, the regression parameter file and driving force file of the CLUE-S land use model are configured; Configure the land use type file for the simulation start year of the CLUE-S land use model and the land use demand file from the forecast simulation base period to the end of the simulation; S24. Set land use type distribution rules, including: Set land use transfer rules, land use type transfer elasticity coefficients, and regional constraints for land use optimization configuration; S25. According to the configured model file and the set land use type distribution rules, the CLUE-S land use model is run, and the operation is continuously iterated to simulate the most suitable spatial layout of different land use grid units.

2. The method for optimizing land use of salinized irrigated land according to claim 1, characterized in that: In step S1, in the constructed water-salt balance SALTMOD model, the cultivated land and salt wasteland in the study area are taken as the research objects, and the total control area only includes the cultivated land and salt wasteland, and its regional division is expressed as: A+B=cultivated land in the study area U = Salt wasteland A+B+U=1 Among them, A represents irrigated land with crops or vegetation growing, B represents fallow or rain-fed irrigated land, and U represents non-irrigated salt wasteland; In the water-salt balance SALTMOD model, the rotation mode of agricultural land Kr=1, the irrigated land areas A and B are all rotated, the salt wasteland area U remains unchanged, and a year is divided into three simulation periods: growth period, autumn irrigation period and non-growth period.

3. The method for optimizing land use of salinized irrigated land according to claim 1, characterized in that: In step S1, the method for simulating and calculating the land use quantity demand in the study area is specifically as follows: S11. Process and analyze the basic data input from the outside through the water-salt balance SALTMOD model and output the corresponding data; The input basic data include agricultural data, hydrological data, meteorological data and soil data of the study area; the output data include soil salinity, groundwater depth, drainage volume and drainage mineralization; S12. Calibrate and verify the water-salt balance SALTMOD model using the specific parameters of the study area; S13. Different scenarios were set based on the current annual irrigation and drainage conditions and different cropland-waste ratios in the study area, and the water-salt balance SALTMOD model after calibration was used to predict and simulate the dynamic changes of salinity in cultivated land under different scenarios. S14, using the predicted salinity of cultivated land as a control constraint, determine the cultivated land ratio in the study area; S15. Based on the determined cultivated land to wasteland ratio of the study area, determine the area and quantity requirements of land use types in annual increments.

4. The method for optimizing land use of salinized irrigated land according to claim 1, characterized in that: In the step S21, the reclassified land use data and driving factor data are represented by a unified coordinate system, projection, spatial range and spatial resolution, and all raster data are uniformly converted into floating point data.

5. The method for optimizing land use of salinized irrigated land according to claim 1, characterized in that: The step S22 is specifically as follows: S22-1, determining the driving factor according to the uniformly represented land use data and driving factor data, and performing rasterization processing on the driving factor data; The driving factors include elevation, slope, NDVI, annual rainfall, annual average temperature, distance from the town center, distance from major railways, highways, national, provincial, county and township roads, distance from canal systems, distance from drainage ditches, multi-year average groundwater depth, population density and gross domestic product (GDP); S22-2, encoding various land use types, and reclassifying each land use type map to obtain a binary map of each land type; S22-3. Based on the binary maps of each land type and the driving factor data of the same period, binary logistic regression analysis was performed on the spatial pattern of land use and its driving factors in different years to obtain the suitability value of each land use grid unit for transformation into a certain land use type.

6. The method for optimizing land use of salinized irrigated land according to claim 5, characterized in that: In step 22-3, the suitability value of converting the land use grid unit into a certain land use type is expressed as: Where P i is the probability of the land use type appearing in each grid cell in the region; x0, x1, …, x n-1 are driving factors; β0, β1, β n-1 are the coefficients of the explanatory variables in the regression equation; n is the number of driving factors.

7. The method for optimizing land use of salinized irrigated land according to claim 1, characterized in that: In step S25, the CLUE-S land use model is run to simulate the most suitable spatial layout of different land use grid cells, which is expressed as: TPROP u，i =THEY i +ITER i +P u，i Where, TPROP u,i represents the overall suitability of grid u for land use type i, ELAS i represents the transfer elasticity coefficient of land type i, representing the conversion cost of land use type, ITER i represents the competition factor of land type i, which is automatically set during the iteration of model simulation, P u,i It represents the distribution probability of grid u to i, which is the regression result of land use status to different driving factors.

8. The method for optimizing land use of salinized irrigated land according to claim 1, characterized in that: The specific method for testing the simulation accuracy of the CLUE-S land use model is as follows: The real land use type and the simulated land use type are spatially superimposed, and the remote sensing classification of land use types in different years and the simulated land use spatial classification map are superimposed and subtracted through the raster calculator, and the number of 0-value grids is extracted as the correct number of grids for simulation, and its proportion to the total number of grids is analyzed; The remote sensing data of land use types in different years were spatially compared with the simulated data of land use types obtained by the corresponding simulation, and the Kappa coefficient was calculated; The simulation accuracy of the CLUE-S land use model was tested based on the proportion of correctly simulated grids to the total number of grids and the Kappa coefficient.

Images