Saline-alkali soil water-salt motion simulation method based on big data analysis

Abstract

The invention relates to the technical field of data twinning, in particular to a saline-alkali soil water-salt motion simulation method based on big data analysis. The method comprises the following steps: collecting saline-alkali soil multi-source monitoring data; based on the saline-alkali soil multi-source monitoring data, salt soil layer correlation feature analysis and water-salt driving factor coupling feature analysis are carried out, and salt soil layer correlation feature data and water-salt driving factor coupling feature data are obtained; simulation relation model design of saline-alkali soil water-salt movement is carried out through the saline-alkali soil level correlation feature data and the water-salt driving factor coupling feature data, and a saline-alkali soil water-salt movement simulation model is generated; based on the saline-alkali soil water-salt motion simulation model, designing saline-alkali soil farmland irrigation simulation parameters; and carrying out water-salt motion simulation prediction and feedback of the saline-alkali soil farmland irrigation on the saline-alkali soil farmland irrigation simulation parameters. According to the saline-alkali soil water-salt motion simulation analysis method, multi-driving-factor fusion and accurate spatial heterogeneity description are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data twin, and particularly relates to a simulation method for water and salt movement in saline-alkali soil based on big data analysis. BACKGROUND

[0002] Soil salinization is one of the key problems restricting the sustainable development of global agriculture and the stability of the ecological environment. Precise simulation and prediction of the movement law of water and salt in saline-alkali soil is the theoretical basis for formulating scientific management measures and realizing efficient use of water resources and efficient improvement of saline-alkali soil. The research on soil water and salt movement mainly relies on numerical models based on physical mechanisms, such as the HYDRUS model which takes Richards equation and convection-dispersion equation as the core. Such models can simulate the migration process of water and salt in the soil profile to a certain extent by setting soil hydraulic parameters and boundary conditions. However, these traditional models have significant limitations in practical application, especially in the management of saline-alkali soil in large-scale and complex farmland environments. First, the simulation accuracy of these models is severely dependent on accurate soil initial and boundary condition parameters, which are usually obtained through limited point monitoring, making it difficult to effectively characterize the inherent high spatial heterogeneity of saline-alkali soil. Second, traditional models do not adequately consider the multiple sources and dynamic factors that drive water and salt movement. Water and salt movement is a complex process driven by multiple factors such as weather (precipitation, evaporation), hydrology (groundwater level, water quality), farmland irrigation management, and soil physical structure, and existing methods often fail to organically integrate and quantify the coupling effects and synergistic influences between these multiple driving factors. In addition, existing simulation methods cannot quickly and flexibly respond to different farmland irrigation scenarios, and cannot perform forward-looking quantitative assessment of the soil salt flushing effect, water and salt migration path, and potential secondary salinization risk that may be caused by a specific irrigation scheme. SUMMARY

[0003] Therefore, the present application provides a simulation method for water and salt movement in saline-alkali soil based on big data analysis to solve at least one of the above technical problems.

[0004] To achieve the above-mentioned purpose, a simulation method for water and salt movement in saline-alkali soil based on big data analysis comprises the following steps: Step S1: acquiring saline-alkali soil multi-source monitoring data by using a saline-alkali soil multi-source monitoring system to obtain saline-alkali soil multi-source monitoring data, wherein the saline-alkali soil multi-source monitoring data includes saline-alkali soil remote sensing image data, saline-alkali soil soil profile data, saline-alkali soil meteorological data, saline-alkali soil hydrological data, and saline-alkali soil farmland irrigation data; Step S2: analyzing the saline-alkali soil spatial salt soil layer correlation characteristics based on the saline-alkali soil remote sensing image data and the saline-alkali soil soil profile data to generate salt soil layer correlation characteristic data; Step S3: Based on the saline-alkali meteorological data and the saline-alkali hydrological data, the water-salt driving factor coupling characteristic analysis of the saline-alkali soil is performed to generate water-salt driving factor coupling characteristic data; Step S4: The simulation relationship model of the saline-alkali water-salt movement is designed by the salt content soil layer correlation characteristic data and the water-salt driving factor coupling characteristic data to generate the saline-alkali water-salt movement simulation model; Step S5: The saline-alkali farmland irrigation simulation parameter is designed by the saline-alkali farmland irrigation data and the saline-alkali water-salt movement simulation model to generate the saline-alkali farmland irrigation simulation parameter; the saline-alkali farmland irrigation water-salt movement simulation prediction is performed on the saline-alkali farmland irrigation simulation parameter to generate the saline-alkali farmland irrigation water-salt movement simulation prediction data.

[0005] Further, step S2 includes the following steps: Step S21: The remote sensing image multispectral characteristic analysis processing is performed according to the saline-alkali remote sensing image data to generate remote sensing image multispectral characteristic data; Step S22: The multispectral salt content index inversion processing is performed based on the remote sensing image multispectral characteristic data to generate multispectral salt content index inversion data; Step S23: The saline-alkali soil profile index data is designed; Step S24: The soil layer structure characteristic analysis of each spatial node is performed on the saline-alkali soil profile data according to the saline-alkali soil profile index data to generate soil layer structure characteristic data; Step S25: The salt content distribution mapping processing is performed by mapping the multispectral salt content index inversion data to the soil layer structure characteristic data to generate soil salt content distribution-layer structure characteristic data; Step S26: The salt content soil layer correlation characteristic analysis is performed on the soil salt content distribution-layer structure characteristic data to generate salt content soil layer correlation characteristic data.

[0006] Further, step S22 includes the following steps: Step S221: The salt content sensitive spectral feature channel data is obtained by extracting the salt content sensitive spectral feature channel according to the remote sensing image multispectral characteristic data; and the spectral band salt content sensitivity influence weight data is generated by performing the salt content sensitivity influence weight analysis on the salt content sensitive spectral feature channel data; Step S222: The salt content sensitive spectral band feature data is generated by performing the salt content sensitive spectral band feature analysis on the remote sensing image multispectral characteristic data according to the salt content sensitive spectral feature channel data; Step S223: The saline-alkali multispectral salt content evaluation model is established based on the spectral band salt content sensitivity influence weight data and the corresponding pre-collected ground measured salt content spectral data and the salt content sensitive spectral feature channel data; Step S224: The salt-sensitive spectral band feature data is processed by the multi-spectral salt evaluation model to generate multi-spectral salt index inversion data.

[0007] Further, the soil profile index data of the saline-alkali soil in step S23 includes soil bulk density index, soil porosity index, soil moisture content index, and soil salt content index.

[0008] Further, step S26 includes the following steps: Step S261: The time-lag correlation of soil layer salt vertical migration relationship is analyzed according to the soil salt distribution-hierarchical structure feature data to generate soil layer salt vertical migration relationship data; Step S262: The salt vertical migration partial correlation feature analysis is performed on the soil layer salt vertical migration relationship data based on the soil layer structure feature data to generate salt vertical migration partial correlation feature data; Step S263: The salt soil layer correlation feature analysis is performed based on the soil layer salt vertical migration relationship data and the salt vertical migration partial correlation feature data to generate salt soil layer correlation feature data.

[0009] Further, step S261 includes the following steps: The soil layer structure salt cross-time lag correlation data is generated by performing soil layer structure salt cross-time lag correlation analysis on the soil salt distribution-hierarchical structure feature data; The soil layer salt vertical migration relationship data is generated by performing soil layer salt vertical migration relationship analysis according to the soil layer structure salt cross-time lag correlation data.

[0010] Further, step S3 includes the following steps: Step S31: The hydrological derivative correlation data of multiple meteorological factors is analyzed based on the meteorological data of the saline-alkali soil to generate meteorological factor hydrological derivative correlation data; Step S32: The meteorological driving hydrological response feature analysis is performed based on the meteorological factor hydrological derivative correlation data to generate meteorological driving hydrological response feature data; Step S33: The water-salt driving factor coupling feature analysis is performed according to the meteorological driving hydrological response feature data and the saline-alkali soil hydrological data to generate water-salt driving factor coupling feature data.

[0011] Further, step S4 includes the following steps: Step S41: The water-salt driving spatial mapping saline-alkali soil salt movement correlation feature analysis is performed by mapping the water-salt driving factor coupling feature data to the salt soil layer correlation feature data to generate saline-alkali soil salt movement correlation feature data; Step S42: Perform feature fusion splicing processing of water and salt movement time sequence-space-propagation based on the salt and alkali land water and salt movement correlation characteristic data, to generate water and salt movement fusion characteristic data; Step S43: Perform salt and alkali land water and salt movement constraint feature analysis based on the salt and alkali land hydrological data and the salt and alkali land soil profile data, to generate salt and alkali land water and salt movement constraint characteristic data; Step S44: Perform simulation relationship model design of salt and alkali land water and salt movement according to the water and salt movement fusion characteristic data and the salt and alkali land water and salt movement constraint characteristic data, to generate a salt and alkali land water and salt movement simulation model.

[0012] Further, step S42 includes the following steps: Step S421: Perform long and short term time sequence hidden state analysis of water and salt movement based on the salt and alkali land water and salt movement correlation characteristic data, to generate water and salt movement long and short term time sequence hidden state data, and perform water and salt movement time sequence dependent feature analysis on the water and salt movement long and short term time sequence hidden state data, to generate water and salt movement time sequence dependent feature data; Step S422: Perform feature vector multi-layer convolution processing of each local space based on the salt and alkali land water and salt movement correlation characteristic data, to generate water and salt movement space convolution data, and perform water and salt movement space feature analysis on the water and salt movement space convolution data, to generate water and salt movement space feature data; Step S423: Perform water and salt movement adjacency correlation feature analysis based on the salt and alkali land water and salt movement correlation characteristic data, to generate water and salt movement adjacency correlation characteristic data, and perform water and salt movement propagation characteristic aggregation analysis on the water and salt movement adjacency correlation characteristic data, to generate water and salt movement propagation characteristic data; Step S424: Perform water and salt movement feature fusion splicing processing on the water and salt movement time sequence dependent feature data, the water and salt movement space feature data, and the water and salt movement propagation characteristic data, to generate water and salt movement fusion characteristic data.

[0013] Further, step S5 includes the following steps: Step S51: Perform salt and alkali land farmland irrigation simulation parameter design through the salt and alkali land farmland irrigation data and the salt and alkali land water and salt movement simulation model, to generate salt and alkali land farmland irrigation simulation parameters; Step S52: Transmit the salt and alkali land farmland irrigation simulation parameters to the salt and alkali land water and salt movement simulation model for farmland irrigation water and salt movement simulation analysis, to generate farmland irrigation water and salt movement simulation data; Step S53: Perform irrigation water and salt flow direction simulation analysis on the farmland irrigation water and salt movement simulation data, to generate irrigation water and salt flow direction simulation data, and perform irrigation characteristic water and salt movement propagation impact feature analysis through the irrigation water and salt flow direction simulation data, to generate water and salt movement propagation impact feature data; Step S54: According to the water salt movement propagation influence characteristic data, the salt and alkali land salt flushing influence characteristic analysis is carried out, and the salt and alkali land salt flushing influence characteristic data is generated; Step S55: Based on the salt and alkali land water salt movement simulation model, the salt and alkali land outlet flow direction boundary condition analysis is carried out, and the salt and alkali land outlet flow direction boundary condition data is generated; Step S56: According to the water salt movement propagation influence characteristic data and the salt and alkali land outlet flow direction boundary condition data, the salt and alkali land water salt outlet flow direction analysis is carried out, and the salt and alkali land water salt outlet flow direction data is generated; Step S57: The salt and alkali land salt flushing influence characteristic data and the salt and alkali land water salt outlet flow direction data are used for the water salt movement simulation prediction of the salt and alkali land farmland irrigation, and the salt and alkali land farmland irrigation water salt movement simulation prediction data is generated.

[0014] The application has the beneficial effects that the present application constructs a multi-source monitoring system for saline-alkali soil, unifies the collection and integration of multi-dimensional data such as remote sensing images, soil profiles, weather, hydrology and farmland irrigation, so that the simulation of soil water and salt movement in saline-alkali soil is no longer dependent on a single data source, and the dynamic change characteristics of saline-alkali soil in space and time can be fully reflected. This multi-source monitoring method effectively solves the problems of insufficient data coverage and single monitoring in the prior art, while improving the accuracy and diversity of data, providing a rich and reliable data foundation for subsequent feature extraction and simulation modeling. In addition, the collection and integration of multi-source data helps to realize cross-scale and cross-element coupling analysis, so that the simulation model is closer to the actual situation, and the credibility and applicability of the simulation results are improved. By combining remote sensing image data and soil profile data to analyze the salt level correlation characteristics of saline-alkali soil space, the relationship between salt spatial distribution and soil layer structure can be deeply described. The multi-spectral feature analysis of remote sensing images and the salt index inversion can quickly extract the salt distribution characteristics in a wide range, with the advantages of wide coverage and strong real-time performance; while the soil profile index data can provide detailed soil structure information, including bulk density, porosity, water content and salt content. Mapping the remote sensing inversion data to the soil profile structure feature data can realize multi-scale fusion from macro to micro, avoid the limitations of single remote sensing or single profile data, and obtain a more comprehensive coupling relationship between salt spatial distribution and soil layer structure. Not only the accuracy of salt distribution analysis is improved, but also a solid foundation is provided for the subsequent research on the vertical migration law of salt, which helps to establish a more realistic water and salt movement simulation model. Based on the comprehensive analysis of meteorological data and hydrological data, the driving mechanism of water and salt movement in saline-alkali soil can be revealed from the perspective of climate change and hydrological response. Meteorological factors such as precipitation, evaporation, temperature, etc. will directly affect the distribution of surface and underground water, and these changes will indirectly affect the migration of soil salt through hydrological processes. By designing hydrological derivative correlation data analysis, the meteorological factors are converted into parameters closely related to hydrological processes, realizing the organic coupling of meteorology and hydrology. Further through the analysis of meteorological driving hydrological response characteristics, the mapping relationship between climate fluctuations and underground water dynamics, surface runoff, evaporation and transpiration can be established, so as to generate more explanatory water and salt driving factor coupling characteristics. This effectively solves the disadvantages of separate processing of meteorology and hydrology in traditional research, so that the water and salt movement simulation can more accurately reflect the water and salt coupling process under natural conditions, and provides more comprehensive and dynamic driving factors for subsequent model design. By combining the water and salt driving factor coupling characteristic data and the salt soil layer correlation characteristic data, the spatial, temporal and propagation law of water and salt movement can be fully modeled. First, the coupling relationship between water and salt driving and soil layer is established through spatial mapping, so that the action path of driving factors in different soil layers can be quantified.Subsequently, through the fusion analysis of multi-dimensional characteristics such as time, space, and adjacency propagation, not only can the short-term fluctuations and long-term trends of water and salt movement be revealed, but also the diffusion characteristics of salt in local space and the propagation characteristics on the regional scale can be reflected. This multi-dimensional fusion method avoids the limitations of single feature modeling, enhancing the ability of the simulation model to capture the rules of water and salt movement in complex environments. At the same time, through the constraints of hydrological and soil profile data, the stability and prediction accuracy of the model can be improved while maintaining physical rationality. Therefore, the simulation model of water and salt movement in saline-alkali soil not only has strong adaptability, but also can effectively guide land improvement and agricultural production in practical applications. The combination of introducing farmland irrigation data and the simulation model of water and salt movement realizes the accurate prediction of water and salt movement under the irrigation scenario of farmland. Irrigation, as an important means of human regulation, has a significant impact on the leaching, migration, and redistribution of soil salt. Through the design and dynamic transmission of simulation parameters, the process of soil salt transport under different irrigation modes can be simulated. Further analysis of the flow direction and propagation characteristics of irrigation water and salt makes it possible to dynamically depict the flow direction, infiltration, and evaporation of irrigation water in the field, thereby revealing the potential impact of irrigation on water and salt balance. Through the analysis of salt flushing and outlet boundary conditions, the mechanism of irrigation on salt discharge and redistribution can be effectively evaluated, and whether the salt is accumulated or carried out in the region can be predicted. The final simulation and prediction data of water and salt movement under farmland irrigation provide a scientific basis for formulating a reasonable irrigation system, improving saline-alkali soil, and optimizing the water and salt environment of farmland, thereby having significant guiding significance in agricultural production practice.

[0015] Therefore, the simulation method of soil water and salt movement in saline-alkali soil based on big data analysis of the application can achieve high-precision characterization of soil water and salt movement in a large-scale saline-alkali soil environment by introducing multi-source monitoring data and intelligent feature extraction methods, breaking through the bottleneck of insufficient spatial representation caused by the dependence of traditional models on limited point parameters. Through the comprehensive utilization of remote sensing images, soil profile monitoring, groundwater dynamics, meteorological data and farmland irrigation data, a multi-source fusion data system with stronger spatial continuity and better dynamic parameter updating capability is established, thereby effectively overcoming the limitations of traditional methods in obtaining initial condition and boundary condition parameters. Secondly, the water-salt driving factor coupling analysis method is introduced in the modeling process, which can systematically integrate meteorological, hydrological, irrigation management and soil physical structure factors, and comprehensively characterize the dynamic coupling relationship and synergistic mechanism between various driving factors. Avoiding the deviation caused by the fragmented treatment of multiple factors in traditional models, the simulation results of water and salt movement are more consistent with the operation characteristics of the actual complex farmland system, significantly improving the scientificity and reliability of the prediction. By constructing a dynamic simulation framework based on multi-scenario parameterization, the possible salt flushing intensity, water and salt transport path and secondary salinization risk caused by different irrigation scenarios can be quantitatively evaluated in advance. Compared with the prior art, not only the adaptability of the simulation results to the actual irrigation management is improved, but also an operable decision basis for scientifically formulating differentiated and precise saline-alkali soil treatment measures is provided. Not only does the method solve the problems of insufficient spatial heterogeneity characterization, insufficient multi-driving factor fusion and poor irrigation scenario adaptability of traditional saline-alkali soil water and salt movement models, but also builds an intelligent simulation framework with high precision, high adaptability and prediction ability, providing better technical support for saline-alkali soil treatment and agricultural water resource management. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 A step flowchart of the simulation method of soil water and salt movement in saline-alkali soil based on big data analysis of the application is shown in the figure. Figure 2 A detailed implementation step flowchart of step S5 in the method is shown in the figure. Figure 1 A detailed implementation step flowchart of step S5 in the method is shown in the figure. The implementation, functional characteristics and advantages of the application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0017] The technical method of the application will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0018] Further, the accompanying drawings are included to provide a further understanding of the application, and are incorporated in and constitute a part of this specification. The drawings illustrate embodiments of the application and, together with the description, serve to explain the principles of the application. In the drawings:

[0019] To achieve the above object, there is provided Figures 1 to 2 The application provides a simulation method for soil water and salt movement in saline-alkali land based on big data analysis. In the embodiments of the application, please refer to Figure 1 Fig. 1 is a schematic diagram of a step flow of the simulation method for soil water and salt movement in saline-alkali land based on big data analysis, according to the application. The simulation method for soil water and salt movement in saline-alkali land based on big data analysis comprises the following steps: Step S1: collecting multi-source monitoring data of saline-alkali land soil by using a multi-source monitoring system for saline-alkali land to obtain multi-source monitoring data of saline-alkali land, wherein the multi-source monitoring data of saline-alkali land comprises remote sensing image data of saline-alkali land, soil profile data of saline-alkali land, meteorological data of saline-alkali land, hydrological data of saline-alkali land, and farmland irrigation data of saline-alkali land; In the embodiment of the present application, the data acquisition link for coverage is established as the target. First, multi-source sensing devices are arranged in the target saline-alkali land, including ground soil profile sampling point array, several automatic weather observation stations, underground water level observation wells, field irrigation flow and outlet metering devices, and remote sensing carriers (satellite multispectral images and unmanned aerial vehicle multispectral cameras) for regional observation. The ground soil profile is sampled according to the pre-planned spatial sample network, and the profile depth is set as several depth intervals according to soil geomorphology. The density, porosity, water content, and conductivity and salt content are analyzed at each depth interval. The meteorological observation station continuously records rainfall, air temperature, relative humidity, wind speed, solar radiation and other elements, and the zero drift and range check are carried out based on the standardized sensor calibration program. The automatic water level recording device is used in the underground water level observation well, and the manual depth check is combined to ensure the continuity and accuracy of the water level time sequence. The irrigation flow and irrigation time sequence are obtained through the field water meter and time recorder, and the irrigation mode and irrigation duration are recorded to construct the irrigation input function. Remote sensing image acquisition includes multi-temporal multispectral remote sensing data and unmanned aerial vehicle high-resolution images. The images are calibrated, corrected, and screened for cloud artifacts after collection, and the field of view and solar zenith angle information are recorded with the image metadata. All observation data are processed in time and space, and then the redundancy detection, abnormal value identification and missing value filling strategy are executed. Spatial inconsistency of different source data is processed through spatial interpolation and resolution matching strategy. The quality control process includes sensor fault detection, time series stability analysis and sample comparison calibration. Finally, the multi-source monitoring data set including saline-alkali land remote sensing image data, saline-alkali land soil profile data, saline-alkali land meteorological data, saline-alkali land hydrological data and saline-alkali land farmland irrigation data is generated, which provides traceable data basis for subsequent feature extraction and model construction.

[0020] Step S2: Based on the saline-alkali land remote sensing image data and the saline-alkali land soil profile data, the salt soil layer correlation feature analysis of the saline-alkali land space is carried out, and the salt soil layer correlation feature data is generated. In the embodiments of the present application, remote sensing surface domain information and soil profile deep layer properties are systematically fused to form a layered salt space description. First, band preprocessing, principal component analysis and band ratio calculation are performed on the remote sensing image to extract multispectral features, and then salt-sensitive band identification is carried out. The correlation analysis and influence weight calculation are carried out with the ground measured salt spectrum data to obtain the band sensitivity weight distribution. Based on this weight, a multispectral salt inversion model is constructed, and a regression algorithm and an integrated learning algorithm are used to establish an inversion ensemble in parallel. Cross-validation and leave-out validation strategies are used to evaluate the generalization error of the model and form a multi-model ensemble output and uncertainty estimation. In terms of soil profile, according to the designed profile index (bulk density, porosity, water content, salt content), each sampling point is characterized by depth layer, and point profile information is expanded to hierarchical three-dimensional profile attribute field using geostatistical methods (such as variogram fitting and ordinary kriging interpolation). Then, the mapping of remote sensing inversion results to soil hierarchy is implemented, and the mapping relationship between surface inversion salt and each soil layer property is established through a depth infiltration coupling function combining statistical regression and physical inference. The mapping process introduces a vertical attenuation model and soil water-heat state constraint to generate a soil salt distribution-hierarchy feature field. Finally, a hierarchical correlation feature vector is constructed for each spatial unit, and the time lag correlation coefficient and partial correlation index between layers are calculated to describe the coupling mode of salt in different soil layers, and a multiscale clustering or principal component dimensionality reduction method is used to generate salt soil hierarchical correlation feature data and its uncertainty measurement for simulation.

[0021] Step S3: Based on the saline-alkali soil meteorological data and the saline-alkali soil hydrological data, the water-salt driving factor coupling feature analysis of the saline-alkali soil is carried out, and the water-salt driving factor coupling feature data is generated; In the embodiment of the present application, the spatio-temporal coupling representation of water-salt driving factors is formed by constructing the linkage conversion chain of meteorology to hydrology and coupling with the observed hydrological process. The meteorological data is first subjected to element derivation processing. The effective infiltration amount time series is obtained by using the rainfall intensity-sustainability decomposition method for rainfall. The potential evapotranspiration is estimated by using the energy balance or reference evapotranspiration method for radiation and temperature. The dynamic response of actual evapotranspiration loss is calculated in combination with the soil water content field. The hydrological data is estimated by runoff profiling, groundwater level response curve fitting and radial flow model to calculate the aquifer recharge and discharge flow. The meteorologically derived hydrological factors and the original hydrological observation are jointly input into the time series response analysis module. The module uses time lag correlation analysis, Granger causality test and frequency domain response analysis to establish the lag response function of meteorological factors on the surface and underground water level and soil water content. The amplitude, delay and attenuation coefficient of the driving response are further calculated. The coupling characteristics are mapped in space based on the interpolation between observation stations and network flow model, and in time based on the event level statistics of sliding window to form the driving factor time series matrix. Finally, the water-salt driving factor coupling characteristic data is generated by multivariate coupling modeling (including factor analysis and orthogonalization processing). The data is represented in the form of coupling strength matrix, time lag response spectrum and its distribution map on the spatial grid of driving variables, which provides dynamic driving input and coupling constraints for subsequent simulation models.

[0022] Step S4: The simulation relationship model of water-salt movement in saline-alkali soil is designed by associating the salt soil level characteristic data and the water-salt driving factor coupling characteristic data, and a simulation model of water-salt movement in saline-alkali soil is generated. In the embodiment of the present application, a simulation relationship model combining physical constraints and data driving is constructed to represent the transport and retention process of water and salt in space and time. The model architecture consists of two parts: the physical process module describes the basic dynamics of water flow and salt transport in porous media based on the unsteady water flow equation and solute transport equation, adopts spatially discrete grid or cell grid for numerical calculation, and uses explicit-implicit hybrid stepping to ensure convergence and stability; the boundary and initial conditions are based on multi-source monitoring data of saline-alkali soil, and the input items include time-varying rainfall infiltration, irrigation injection and underground recharge. The data-driven residual correction module is built on the physical module to capture the nonlinear residuals that the physical model cannot explain, uses integrated learning and deep learning methods to regress and correct the deviation between the physical model output and the observation data, and feeds the correction term back to the parameter updating process of the physical module. Parameter calibration is achieved through inversion algorithm and data assimilation technology, and gradient-based optimization and sample-based Bayesian update method are used in parallel, and Kalman filter or ensemble Kalman filter is used for time series data assimilation to update the state quantity and parameter estimation in real time. The model output includes grid-level water content, salt concentration and flux spatiotemporal field, accompanied by posterior uncertainty estimation and sensitivity analysis, and the model verification uses the reserved observation data for independent testing and adjusts the error term and model constraints accordingly to ensure the physical consistency of the simulation model while improving the prediction accuracy.

[0023] Step S5: Perform saline-alkali soil farmland irrigation simulation parameter design through saline-alkali soil farmland irrigation data and saline-alkali soil water and salt movement simulation model, generate saline-alkali soil farmland irrigation simulation parameters; perform saline-alkali soil farmland irrigation water and salt movement simulation prediction on the saline-alkali soil farmland irrigation simulation parameters, and generate saline-alkali soil farmland irrigation water and salt movement simulation prediction data.

[0024] In the embodiment of the present application, the irrigation operation parameters are systemized as model recognizable source items and dynamically simulated and predicted for the purpose of irrigation scenario driven prediction. First, the irrigation information is converted into grid level input functions, covering irrigation timing, single water injection volume, injection depth distribution and release rate, which are combined with the hydraulic boundary characteristics of the irrigation mode to act on the source item nodes of the simulation model. During the simulation process, the numerical solver is used to calculate the horizontal flow of irrigation water in the field, vertical infiltration and exchange with groundwater, and the solute conservation equation is introduced to simulate the shear migration, diffusion and adsorption retention process of salt with water flow. The simulation output is subjected to irrigation water salt flow direction simulation, the flow direction tracking algorithm and flux decomposition method are used to generate the irrigation water salt flow field, and the indicators for representing the flushing intensity, such as the salt flux change rate per unit time and the cumulative salt output, are calculated. At the same time, the water and salt propagation influence characteristic analysis is carried out, the spatial adjacency matrix and propagation spectrum analysis are used to quantify the propagation path and speed of salt in the neighborhood caused by irrigation, and the salt flux and solute outflow timing at the regional outlet are estimated based on the outlet boundary condition analysis module. Finally, the farmland irrigation water salt movement simulation prediction data is generated, which includes the temporal and spatial salt concentration field, salt flux distribution, flushing influence characteristics and outlet flow direction information under the irrigation scenario, serving as the decision basis for irrigation regulation and salinization risk assessment.

[0025] Further, step S2 comprises the following steps: Step S21: performing remote sensing image multispectral feature analysis processing according to the saline-alkali soil remote sensing image data to generate remote sensing image multispectral feature data; In the embodiment of the present application, according to the collected remote sensing image data of saline-alkali soil, the image spatial resolution is 10 meters, containing 13 multispectral bands, and the data acquisition time interval is 5 days to capture the dynamic changes of soil salt. First, the image is subjected to radiation calibration processing, and according to the calibration coefficient provided by the satellite launch party, the original DN value of the image is converted into the ground reflectivity, eliminating the influence of sensor response difference on the data; then atmospheric correction is carried out, the dark target method is used to identify the low reflectivity areas such as water body and shadow in the image as dark target, the atmospheric aerosol optical depth is calculated, and then the interference of atmospheric scattering and absorption on reflectivity is removed through the radiation transfer model; subsequently, geometric correction is carried out, taking the 1:10,000 regional topographic map as the reference, selecting not less than 20 ground control points such as road intersections and ridge corners uniformly distributed, using a quadratic polynomial transformation model to perform spatial registration on the image, and ensuring that the planar position error of the corrected image is not more than 1 pixel. After the pretreatment is completed, the ground reflectivity data of each band is extracted, and the normalized difference vegetation index (NDVI, formula: (near-infrared band reflectivity - red band reflectivity) / (near-infrared band reflectivity + red band reflectivity)), normalized difference salt index (NDSI, formula: (short-wave infrared 2 band reflectivity - red band reflectivity) / (short-wave infrared 2 band reflectivity + red band reflectivity)), soil adjusted vegetation index (SAVI) and other band combination indexes are calculated. All data are stored in the form of a grid, and the grid unit size is consistent with the image spatial resolution, and finally a remote sensing image multispectral feature data containing 13 original band reflectivity and 3 derived indexes is formed, providing basic spectral information for subsequent salt inversion.

[0026] Step S22: multispectral salt index inversion processing based on remote sensing image multispectral feature data is performed to generate multispectral salt index inversion data; In the embodiment of the present application, the salt-sensitive spectral feature channels are extracted from the 13 original band reflectances in the multispectral feature data of remote sensing images. The Pearson correlation analysis method is used to calculate the correlation between each band reflectance and the 100 ground soil salt content measured data (measured data obtained by sensors after soil sampling, sampling points are uniformly arranged according to the grid method, and two layers of soil samples (0-20 cm and 20-40 cm) are collected at each sampling point) collected synchronously in the study area. The band with an absolute correlation coefficient greater than 0.6 is selected as the salt-sensitive spectral feature channel. Usually, the short-wave infrared 1 band (center wavelength 1610 nm), the short-wave infrared 2 band (center wavelength 2200 nm) and the red light band (center wavelength 665 nm) will be selected as the sensitive channel. Then, the salt-sensitive spectral feature channel data selected is analyzed for the salt-sensitive influence weight of the spectral band. The random forest algorithm is used to construct a feature importance evaluation model, the sensitive channel reflectance is used as the input variable, and the ground measured salt content is used as the output variable. The weight is calculated by the contribution of each variable to the output result in the model training process. The weight value is between 0 and 1 after normalization. Among them, the weight of the short-wave infrared 2 band is usually the highest, and the weight of the red light band is the second. Then, the salt-sensitive spectral band feature analysis is performed on the multispectral feature data of remote sensing images based on the selected salt-sensitive spectral feature channel data. The reflectance mean, maximum, minimum and standard deviation of each sensitive channel in the study area are extracted. The reflectance difference (such as short-wave infrared 2 band reflectance - red light band reflectance) and ratio (such as short-wave infrared 1 band reflectance / red light band reflectance) between different sensitive channels are calculated to form the salt-sensitive spectral band feature data. Then, the multispectral salt evaluation model of saline-alkali soil is established. The ground measured salt spectral data corresponding to the salt-sensitive spectral feature channel data (i.e. the salt content of the ground sampling point and the reflectance of the sensitive channel at the corresponding position) is used as the sample. The multiple linear regression algorithm is used to construct the model. The model expression is soil salt content = α1 x short-wave infrared 1 band reflectance + α2 x short-wave infrared 2 band reflectance + α3 x red light band reflectance + β (wherein α1, α2, α3 are regression coefficients, and β is a constant term). The least square method is used to solve the coefficients, so that the residual sum of squares of the predicted value and the measured value of the model is minimized. Finally, the salt-sensitive spectral band feature data is substituted into the model, and the soil salt content of each spatial position in the study area is calculated grid by grid to generate spatially continuous multispectral salt index inversion data. The data format is consistent with the multispectral feature data of remote sensing images, and the spatial position correspondence is ensured.

[0027] Step S23: design saline-alkali soil profile index data; In the embodiment of the present application, the designed saline-alkali soil profile index data needs to comprehensively reflect the soil physical properties and salt conditions to support the subsequent soil layer structure analysis and salt transport rule research. Among them, the soil bulk density index is used to characterize the soil compaction degree, directly affects the soil water infiltration rate and salt vertical transport efficiency, and is directly obtained through the bulk density sensor. The sensor is vertically embedded in different depths of the soil profile to ensure that the soil is not disturbed during the monitoring process; the soil porosity index includes total porosity, capillary porosity and non-capillary porosity. The total porosity is calculated by soil bulk density and soil particle density (determined based on regional soil texture type), the formula is total porosity = (1 - soil bulk density / soil particle density) x 100, the capillary porosity is calculated by soil saturated water content and soil bulk density, the formula is capillary porosity = soil saturated water content x soil bulk density x 100, and the non-capillary porosity is the difference between the total porosity and the capillary porosity. This index reflects the soil ventilation and water permeability, and the size of the porosity directly determines the storage space of water and salt; the soil moisture content index includes mass moisture content and volume moisture content. The mass moisture content is obtained by monitoring the soil moisture sensor, the sensor is embedded in different depths of the soil to collect data in real time, and the volume moisture content is obtained by the product of the mass moisture content and the soil bulk density. This index is the carrier of salt transport, and its change directly affects the salt distribution; the soil salt content index includes total salt content and ion composition. The total salt content is directly monitored by the salt sensor, the sensor is inserted into the soil to contact the soil solution to obtain data, and the ion composition is measured by the ion selective electrode sensor to detect the concentration of ion elements in the soil. This index directly reflects the degree of soil salinization and is the core basis for the correlation analysis of salt layers. All index data need to correspond to the specific depth of the soil profile, the interval of the stratification is set to 20 cm, starting from the surface 0 cm to the depth of 100 cm, forming a complete profile index system.

[0028] Step S24: According to the saline-alkali soil profile index data, the soil layer structure characteristics of each spatial node of the saline-alkali soil profile data are analyzed to generate soil layer structure characteristic data; In the embodiment of the present application, the saline-alkali soil profile data is classified by the sampling point position in space, and each sampling point corresponds to a group of soil sample data from five depth levels of 0-20 cm, 20-40 cm, 40-60 cm, 60-80 cm, and 80-100 cm. The data includes index values such as soil bulk density, porosity, water content, and salt content of each level. According to the saline-alkali soil profile index data designed in step S23, the hierarchical structure characteristics of the soil profile data of each spatial node (i.e., sampling point) are analyzed: first, the soil level division standard is determined, and the soil bulk density and total porosity are used as the core division basis. When the soil bulk density is >1.4 and the total porosity is <45, it is determined as plough bottom layer; when the bulk density is 1.2-1.4 and the total porosity is 45-50, it is determined as cultivated layer; and when the bulk density is <1.2 and the total porosity is >50, it is determined as loose soil layer (such as heart soil layer or bottom soil layer, which is further distinguished according to the depth). According to this standard, the five depth level data of each sampling point are classified, for example, the 0-20 cm level of a certain sampling point has a bulk density of 1.3 and a total porosity of 48, which is determined as cultivated layer; the 20-40 cm level has a bulk density of 1.45 and a total porosity of 43, which is determined as plough bottom layer; and the 40-100 cm level has a bulk density of 1.15 and a total porosity of 52, which is determined as heart soil layer. Then, the thickness of each determined level is calculated, such as the thickness of the cultivated layer is 20 cm, the thickness of the plough bottom layer is 20 cm, and the thickness of the heart soil layer is 60 cm; and the mean values of each index in each level are calculated, such as the mean value of the soil water content in the cultivated layer is 18, the mean value of the salt content is 0.3, the mean value of the soil water content in the plough bottom layer is 15, the mean value of the salt content is 0.45, the mean value of the soil water content in the heart soil layer is 12, and the mean value of the salt content is 0.25. Then, a spatial interpolation method (Kriging interpolation method) is used to calculate the hierarchical structure characteristics of the unsampled spatial nodes in the study area based on the level division results and the mean values of the indexes of each level of the sampling points, combined with the latitude and longitude coordinates of the sampling points, to ensure that the spatial resolution of the interpolation results is consistent with the remote sensing image in step S21. The finally generated soil hierarchical structure characteristic data includes the number of soil levels, the names of each level, the thickness, and the mean values of the bulk density, porosity, water content, and salt content of the corresponding level of each spatial node in the study area, and the data is presented in the form of a vector surface layer, and each surface unit corresponds to the soil profile hierarchical structure information of a spatial node.

[0029] Step S25: mapping the multispectral salt index inversion data to the soil hierarchical structure characteristic data for salt distribution mapping processing to generate soil salt distribution-hierarchical structure characteristic data; In the embodiment of the present application, the spatial matching relationship between the multispectral salt index inversion data and the soil hierarchical structure characteristic data is established, both of which use the latitude and longitude coordinate system, to ensure that the multispectral salt index inversion value (raster data) of each spatial node can accurately correspond to the soil hierarchical structure characteristic (vector surface data) of the node, and the matching error is controlled within 1 meter. The multispectral salt index inversion data includes the salt inversion value of each spatial node at different depths (based on the vertical penetration ability of remote sensing image, the inversion value is divided into 0-20cm, 20-40cm, 40-60cm, 60-80cm, and 80-100cm five layers, which are consistent with the soil profile sampling depth), and the soil hierarchical structure characteristic data includes the hierarchical division result (such as plough layer, plough bottom layer, and heart soil layer) of each spatial node and the depth range of each layer. Then, the salt distribution mapping processing is performed, the salt value of the corresponding depth range in the multispectral salt index inversion data is assigned to the corresponding layer in the soil hierarchical structure characteristic data, for example, the depth range of the plough layer of a spatial node is 0-20cm, then the salt value (such as 0.32) of the depth of 0-20cm in the multispectral salt index inversion data is assigned to the plough layer of the node; the depth range of the plough bottom layer is 20-40cm, the salt inversion value (such as 0.48) of the depth of 20-40cm is assigned to the plough bottom layer; the depth range of the heart soil layer is 40-100cm, the salt inversion values (such as 0.26, 0.24, and 0.22) of the depths of 40-60cm, 60-80cm, and 80-100cm are averaged (0.24) and assigned to the heart soil layer. For the case that a layer in the soil hierarchical structure characteristic data spans multiple inversion depths (such as the heart soil layer spanning three inversion depths of 40-100cm), the average value calculation method is used to integrate the salt inversion values of the corresponding depths, to ensure that each soil layer has a unique salt value. Then, the mapping result is subjected to consistency checking, the original measured salt average value of the soil hierarchical structure characteristic data in each spatial node is compared with the salt inversion value after mapping, if the difference between the two is more than 0.05, the depth range division of the layer and the depth matching of the salt inversion value of the node are rechecked for accuracy, and then checked again after correction, until the difference of all nodes is less than 0.05. Finally, the soil salt distribution-hierarchical structure characteristic data is generated, each spatial node includes the name, depth range, bulk density, porosity, water content, and mapped salt value of each soil layer, to realize the depth fusion of salt distribution and soil hierarchical structure.

[0030] Step S26: salt soil layer correlation characteristic analysis is performed on the soil salt distribution-hierarchical structure characteristic data to generate salt soil layer correlation characteristic data.

[0031] In the embodiments of the present application, the time lag correlation of the soil salt distribution-layer structure characteristic data is analyzed, the soil layer salt vertical migration relationship is analyzed, the soil salt distribution-layer structure characteristic data in the research area for 6 consecutive months (1 data is obtained every month) is selected, each spatial node is taken as an analysis unit, for two adjacent soil layers (such as plough layer and plough bottom layer, plough bottom layer and heart soil layer), the cross correlation coefficient of the upper layer soil salt value and the lower layer soil salt value under different time lags (1 month, 2 months, 3 months) is calculated. For example, the correlation coefficient of the first month salt value of the plough layer and the second month salt value of the plough bottom layer (time lag 1 month), the correlation coefficient of the first month salt value of the plough layer and the third month salt value of the plough bottom layer (time lag 2 months) are calculated. By comparing the correlation coefficients of different time lags, the best time lag period of salt migration from the upper layer to the lower layer is determined (usually the correlation coefficient of time lag 1 month is the highest, indicating that there is a time lag of 1 month in the vertical migration of salt), and the migration direction is judged according to the positive and negative of the correlation coefficient (positive correlation indicates that the increase of the upper layer salt will lead to the increase of the lower layer salt, that is, the salt migrates downward; negative correlation is the opposite), and the migration intensity is judged according to the absolute value of the correlation coefficient (the larger the absolute value, the higher the migration intensity), thereby generating the soil layer salt vertical migration relationship data, including the salt migration time lag period, direction and intensity of adjacent layers of each spatial node. Then, the partial correlation characteristic analysis of the vertical migration of salt is carried out, the soil bulk density and capillary porosity in the soil layer structure characteristic data are taken as control variables, and the partial correlation analysis method is used to recalculate the correlation coefficient of the salt values of adjacent layers under the premise of eliminating the influence of bulk density and capillary porosity on salt migration. For example, when analyzing the salt migration of the plough layer and the plough bottom layer, the bulk density values of the two layers are fixed (such as plough layer bulk density 1.3, plough bottom layer bulk density 1.45), the partial correlation coefficient of the salt values of the two layers is calculated, and the interference degree of bulk density and porosity on the vertical migration of salt is judged by comparing the correlation coefficient without control variables (if the difference between the partial correlation coefficient and the original correlation coefficient is large, it indicates that the control variable has a significant influence on the migration of salt), and the critical threshold of bulk density and porosity is determined according to the change trend of the partial correlation coefficient (such as when the bulk density exceeds 1.5, the partial correlation coefficient decreases significantly, indicating that the soil is too compact to inhibit the migration of salt). The partial correlation characteristic data of the vertical migration of salt is generated, including the partial correlation coefficient of the salt of adjacent layers of each spatial node, the main affected soil physical indexes and the critical threshold.Finally, based on the data of vertical salt transport relationship and the data of vertical salt transport partial correlation characteristics of soil layers, the coupling degree model is used to calculate the correlation degree of salt and soil layer structure, the product of the transport intensity (absolute value of correlation coefficient) and the partial correlation coefficient is taken as the basis for correlation degree calculation, and the soil layer thickness is combined for weighted correction (the thicker the layer, the higher the weight), the correlation degree value is between 0 and 1, and the value closer to 1 indicates that the salt is more closely related to the layer structure, thereby generating the salt soil layer correlation characteristic data, including the salt correlation degree of each soil layer of each spatial node, the main influencing factors (such as bulk density, porosity) and the correlation level (0.8-1.0 for strong correlation, 0.5-0.8 for moderate correlation, and 0-0.5 for weak correlation).

[0032] Further, step S22 includes the following steps: Step S221: Extracting salt-sensitive spectral feature channels according to the multispectral feature data of the remote sensing image to obtain salt-sensitive spectral feature channel data; and analyzing the salt sensitivity influence weight of the spectral band to generate spectral band salt sensitivity influence weight data; In the embodiment of the application, 13 original band reflectivity data are extracted from the multispectral feature data of the remote sensing image for salt-sensitive spectral feature channel extraction. 100 ground sampling points are uniformly arranged in the study area by grid method, and soil salt data (directly monitored by a salt sensor) of each sampling point is synchronously obtained, Pearson correlation analysis is performed on the reflectivity data of each band and the salt data of the corresponding sampling point, and the correlation coefficient between each other is calculated. The absolute value of the correlation coefficient greater than 0.6 is set as the screening threshold, and the bands meeting the condition are reserved as salt-sensitive spectral feature channels, and it is necessary to ensure that the reflectivity data of each band and the salt data are strictly matched in space position and time during the screening process. After the sensitive channel extraction is completed, the salt sensitivity influence weight analysis of the spectral band is performed on the screened channel data. A feature importance evaluation model is constructed by using a random forest algorithm, the reflectivity data of the sensitive channel are taken as the input variables of the model, the ground measured salt data are taken as the output variables, the model is trained for multiple iterations, and the contribution degree of each input variable in the model prediction process is recorded. The contribution degrees of the channels are normalized to make the weight value sum of all channels be 1, and the spectral band salt sensitivity influence weight data is generated, wherein the higher the contribution degree of the channel, the greater the weight value corresponding to the channel, so as to quantify the influence degree of different spectral bands on the salt inversion.

[0033] Step S222: performing salt-sensitive spectral band feature analysis on the multispectral feature data of the remote sensing image according to the salt-sensitive spectral feature channel data to generate salt-sensitive spectral band feature data; In the embodiment of the present application, based on the determined salt-sensitive spectral feature channel data, the salt-sensitive spectral band feature analysis is performed on the multispectral feature data of the remote sensing image. For each sensitive channel, an analysis unit of 50m*50m is divided in the study area, and the mean value, maximum value, minimum value and standard deviation of the reflectivity of the channel in each unit are calculated to reflect the distribution and dispersion degree of the reflectivity in space. At the same time, the reflectivity of different sensitive channels is combined to calculate the difference and ratio of the reflectivity of any two sensitive channels, such as the difference feature obtained by subtracting the reflectivity of the red light band from the reflectivity of the short-wave infrared 2 band, and the ratio feature obtained by dividing the reflectivity of the short-wave infrared 1 band by the reflectivity of the red light band. These combined features can enhance the spectral signal difference related to soil salt content and highlight the response of salt content change in the spectrum. All feature data are stored at the same spatial resolution as the remote sensing image to ensure that each spatial position corresponds to a complete set of sensitive band feature values, and finally the salt-sensitive spectral band feature data containing single-band statistical features and multi-band combined features are formed to provide input parameters for the subsequent salt evaluation model.

[0034] Step S223: based on the salt-sensitive spectral feature channel data and the salt-sensitive spectral feature channel data corresponding to the pre-collected ground measured salt spectrum data, a salt-alkali soil multispectral salt evaluation model is established; In the embodiment of the present application, the ground measured salt spectrum data collected in the study area in advance is collected, which contains the salt content (monitored by a salt sensor) of 150 sampling points and the reflectivity of the sensitive spectral feature channel corresponding to the position acquired synchronously with the remote sensing image. The generated spectral band salt sensitivity influence weight data is used as the weight coefficient of the model, and the salt-sensitive spectral feature channel data is combined to construct the salt-alkali soil multispectral salt evaluation model. The multiple linear regression algorithm is used as the basis of the model, the reflectivity data of the sensitive channel is used as the independent variable, and the ground measured salt content is used as the dependent variable. The weight coefficient is used to weight each independent variable. In the model construction process, the 150 sampling point data is divided into a training set and a validation set in a ratio of 7:3, the training set data is used to solve the coefficients of the regression equation by the least square method, so that the residual sum of squares of the calculated value and the measured value of the equation is minimized. The model is tested by using the validation set data, the weight coefficient and the regression coefficient are adjusted to optimize the model, until the prediction error of the model is controlled within the preset range, so that the model can accurately reflect the quantitative relationship between the sensitive spectral feature and the soil salt content, and finally a stable salt-alkali soil multispectral salt evaluation model is formed.

[0035] Step S224: the multispectral salt index inversion data is generated by performing multispectral salt index inversion processing on the salt-sensitive spectral band feature data through the salt-alkali soil multispectral salt evaluation model.

[0036] In the embodiment of the present application, the salt-sensitive spectral band feature data is input into the saline soil multi-spectral salt evaluation model established in step S223 to perform multi-spectral salt index inversion processing. According to the preset regression equation, the model takes the single-band statistical features and multi-band combination features in the feature data as input, combines the spectral band salt sensitivity influence weight data, and calculates the soil salt content of each spatial position in the research area by grid. During the inversion process, the input data needs to be consistent with the feature dimension required by the model to ensure that the feature value of each grid can be correctly analyzed by the model. The inversion result is checked for spatial continuity to check whether the salt value change of adjacent grids conforms to the gradual change rule of soil salt spatial distribution. If a sudden change value appears, the feature data and model calculation process of the position are rechecked, and the result is re-inverted after correction. After the inversion is completed, the generated multi-spectral salt index inversion data and the multi-spectral feature data of the remote sensing image maintain the same spatial coordinate system and resolution, each grid corresponds to a salt inversion value, and the soil salt spatial distribution data covering the entire research area is formed, providing a basis for subsequent salt distribution mapping processing.

[0037] Further, the saline soil soil profile index data in step S23 includes soil bulk density index, soil porosity index, soil moisture content index, and soil salt content index.

[0038] Further, step S26 includes the following steps: Step S261: According to the soil salt distribution-hierarchical structure feature data, the soil layer salt vertical migration relationship analysis of time lag correlation is performed to generate soil layer salt vertical migration relationship data; In the embodiment of the present application, the soil salt distribution-layer structure characteristic data in the research area for 6 consecutive months is selected, and each month of data includes the salt values of different soil layers of each spatial node and layer division information. Taking each spatial node as an independent analysis unit, the time lag correlation analysis is carried out for all adjacent soil layers (such as plough layer and plough bottom layer, plough bottom layer and heart soil layer, etc.) in the unit. The monthly salt value of the upper layer soil is taken as the reference sequence, and the salt values of the lower layer soil in the corresponding month and the subsequent months are taken as the comparison sequence, and the cross correlation coefficients under different time lag intervals (1 month, 2 months, 3 months) are calculated. When calculating, it is necessary to ensure that the upper and lower layer soil data strictly correspond in space position and time node, and the correlation coefficient of each time lag interval is obtained by the ratio of the product of the covariance and the standard deviation of the paired data. By comparing the correlation coefficients of different time lag intervals, the best time lag period of salt migration from the upper layer to the lower layer (i.e. the time lag interval with the highest correlation coefficient) is determined; the direction of migration is judged according to the correlation coefficient (positive correlation for downward salt migration, negative correlation for upward salt migration); and the migration intensity is divided according to the absolute value of the correlation coefficient (the larger the absolute value, the higher the intensity). The time lag period, migration direction and migration intensity of adjacent layers of each spatial node are summarized to generate the soil layer salt vertical migration relationship data, and the data need to maintain the same spatial coordinates as the original soil layer structure characteristic data.

[0039] Step S262: Perform partial correlation characteristic analysis on the soil layer salt vertical migration relationship data by the soil layer structure characteristic data to generate salt vertical migration partial correlation characteristic data; In the embodiment of the present application, the soil bulk density and capillary porosity indexes in the soil layer structure characteristic data are extracted as control variables affecting the salt vertical migration, and the partial correlation characteristic analysis is performed on the soil layer salt vertical migration relationship data. For the adjacent soil layers of each spatial node, the correlation coefficient (i.e. partial correlation coefficient) of the salt values of the upper and lower layers is recalculated under the premise of keeping the values of the bulk density and capillary porosity indexes unchanged. In the calculation process, the index dimension difference is first eliminated by variable standardization, and then the interference of the control variables on the salt migration is removed by inverting the covariance matrix. The difference between the partial correlation coefficient and the original correlation coefficient without the control variables is calculated, and the larger the difference, the more significant the influence of the control variable on the salt migration. By analyzing the trend of the partial correlation coefficient changing with the value of the control variable, the critical threshold of the control variable is determined (i.e. when the control variable exceeds a certain value, the partial correlation coefficient appears a significant mutation). For example, when the bulk density exceeds a certain value, the partial correlation coefficient decreases significantly, indicating that the soil compaction degree has significantly inhibited the salt vertical migration. The partial correlation coefficient, the main influence control variable and the critical threshold of adjacent layers of each spatial node are summarized to generate the salt vertical migration partial correlation characteristic data, and the data need to include the spatial correspondence relationship with the soil layer salt vertical migration relationship data.

[0040] Step S263: Based on the soil layer salt vertical migration relationship data and the salt vertical migration partial correlation characteristic data, salt soil layer correlation characteristic analysis is carried out to generate salt soil layer correlation characteristic data.

[0041] In the embodiment of the application, based on the generated soil layer salt vertical migration relationship data and salt vertical migration partial correlation characteristic data, a coupling degree model is used to carry out salt soil layer correlation characteristic analysis. The model takes the product of the migration intensity (absolute value of correlation coefficient) in the soil layer salt vertical migration relationship data and the partial correlation coefficient in the salt vertical migration partial correlation characteristic data as the basic correlation value, and then combines the layer thickness in the soil layer structure characteristic data for weighted correction (the greater the layer thickness, the higher the weight). The weighted correction is realized by the product of the basic correlation value and the layer thickness ratio. The correlation degree value obtained by calculation needs to be normalized to be between 0 and 1. The value closer to 1 indicates that the salt is more closely related to the soil layer. According to the correlation degree value, the correlation level is divided: 0.8-1.0 for strong correlation, 0.5-0.8 for medium correlation, and 0-0.5 for weak correlation. At the same time, the main influencing factors (such as bulk density, capillary porosity, etc. determined by partial correlation analysis) of each soil layer correlation are recorded. The correlation degree, correlation level and main influencing factors of each soil layer of each spatial node are integrated to generate salt soil layer correlation characteristic data, which needs to maintain the same spatial resolution and coordinate system as the soil layer structure characteristic data, providing a layer correlation basis for subsequent water and salt movement simulation.

[0042] Further, step S261 includes the following steps: The soil salt distribution-layer structure characteristic data is subjected to salt cross-time lag correlation analysis processing of soil layer structure to generate soil layer structure salt cross-time lag correlation data. According to the soil layer structure salt cross-time lag correlation data, soil layer salt vertical migration relationship analysis is carried out to generate soil layer salt vertical migration relationship data.

[0043] In the embodiment of the present application, the soil salt distribution - hierarchical structure characteristic data of the research area for 6 consecutive months is selected, each month of data contains the soil hierarchical division (such as plough layer, plough bottom layer, heart soil layer, etc.) of each spatial node and the salt value of the corresponding layer, and the data needs to maintain the same spatial resolution and coordinate system. Each spatial node is taken as an independent analysis unit, for each soil layer in the unit, the salt value of 6 consecutive months is extracted to form a time series, and the same period salt value of the adjacent upper or lower layer soil in the unit is extracted to form a corresponding time series. Set 1 month, 2 months, 3 months three time lag intervals, carry out cross-time lag correlation analysis on two adjacent layers (such as plough layer and plough bottom layer) in the same unit: take the salt value of the upper layer in the first month as the reference, respectively match with the salt value of the lower layer in the first month (no time lag), the second month (time lag 1 month), the third month (time lag 2 months), and the fourth month (time lag 3 months), and the matching data amount of each time lag interval is not less than 30 groups. The Pearson correlation coefficient calculation formula is used to calculate the correlation coefficient under different time lag intervals through the ratio of the covariance and the standard deviation product of each matching data. Repeat the above operation for all adjacent layers of all spatial nodes, record the correlation coefficient corresponding to each node, each adjacent layer combination and each time lag interval, generate soil hierarchical structure salt cross-time lag correlation data, and the data needs to be clearly marked with spatial node position, layer combination, time lag interval and corresponding correlation coefficient value. Based on the generated soil hierarchical structure salt cross-time lag correlation data, the vertical migration relationship analysis is carried out on the adjacent soil layer combination (such as plough layer-plough bottom layer, plough bottom layer-heart soil layer) of each spatial node. For each layer combination, the correlation coefficients of different time lag intervals are extracted from the cross-time lag correlation data, and the best time lag period is determined by comparing the sizes of these coefficients, that is, the time lag interval with the maximum absolute value of the correlation coefficient, which reflects the time lag feature of the salt migration from the upper layer to the lower layer or from the lower layer to the upper layer. According to the positive and negative of the correlation coefficient corresponding to the best time lag period, the migration direction is judged: positive correlation indicates that the change of salt in the upper layer will cause the same change in the lower layer (salt migration downward), and negative correlation indicates that the change of salt in the upper layer will cause the opposite change in the lower layer (salt migration upward). According to the absolute value of the correlation coefficient of the best time lag period, the migration intensity level is divided, and the larger the absolute value, the higher the migration intensity. Repeat the above analysis for each layer combination of all spatial nodes, and summarize the layer combination name, best time lag period, migration direction, migration intensity level of each node, ensure that the data corresponds to the spatial coordinates of the soil hierarchical structure characteristic data, and finally generate the soil layer salt vertical migration relationship data, which provides a quantitative basis for subsequent salt migration rule analysis.

[0044] Further, step S3 comprises the following steps: Step S31: Perform hydrological derivation correlation data analysis on the saline-alkali soil meteorological data of the multi-source meteorological factors to generate meteorological factor hydrological derivation correlation data; In the embodiment of the present application, the meteorological data of the saline-alkali soil in the research area for 5 consecutive years is collected, and the data is derived from 10 evenly distributed automatic weather stations, including 7 basic meteorological factors of daily precipitation, average temperature, maximum temperature, minimum temperature, wind speed, relative humidity, and sunshine duration. The data collection time interval is 1 hour, ensuring coverage of different seasons and weather conditions. The basic meteorological factors are calculated by hydrological derivation factors: based on temperature and relative humidity, the potential evapotranspiration is calculated by the Penman-Monteith formula, which needs to include sunshine duration and wind speed parameters to reflect the atmospheric evaporation capacity; the daily precipitation is divided into precipitation intensity grades according to the 24-hour cumulative value, and the precipitation intensity is calculated by the ratio of continuous precipitation time and cumulative amount; the air dryness is calculated by combining wind speed and temperature, and the formula is dryness = potential evapotranspiration / precipitation (when the precipitation is zero, the dryness takes a certain maximum value). After the derivation factor calculation is completed, the correlation analysis is carried out: taking precipitation and potential evapotranspiration as the core independent variables, taking the surface runoff simulation value and soil moisture supply as the dependent variables, using the sliding window method (window size is 7 days) to calculate the Pearson correlation coefficient of the independent variables and the dependent variables in different time periods, and analyzing the linear regression relationship between temperature and potential evapotranspiration, the partial correlation relationship between wind speed and precipitation uniformity. All basic meteorological factors, derived factors and their correlation coefficients are arranged in time sequence to form meteorological factor hydrological derivation correlation data containing factor name, value and correlation strength, and the data time resolution remains consistent with the original meteorological data.

[0045] Step S32: Perform meteorological driving hydrological response characteristic analysis based on the meteorological factor hydrological derivation correlation data to generate meteorological driving hydrological response characteristic data; In the embodiment of the present application, based on the generated meteorological factor hydrological derivative correlation data, precipitation, potential evapotranspiration and air dryness are selected as key meteorological driving factors, and combined with the synchronous observation data of 20 hydrological monitoring points in the research area (including surface runoff, 0-100cm soil moisture content and groundwater level depth), the analysis of meteorological driving hydrological response characteristics is carried out. For each hydrological monitoring point, the time sequence corresponding relationship between meteorological factors and hydrological elements is constructed: the daily precipitation and the next day's surface runoff are paired, the runoff change corresponding to the unit precipitation is calculated, which reflects the immediate response of precipitation to surface runoff; the potential evapotranspiration and the soil moisture content reduction amount of the day are paired, the soil water consumption rate caused by evapotranspiration is analyzed, and the proportional relationship between the two is determined by linear fitting; the average of the continuous 7-day air dryness and the groundwater level depth change are paired, and the inhibition effect of long-term dry conditions on groundwater recharge is analyzed. Using the lag analysis method, a lag window of 1-3 days is set, the correlation coefficients of meteorological factors and hydrological elements at different lag periods are calculated, the response time (i.e. the lag period with the highest correlation coefficient) is determined, such as the response lag of precipitation to groundwater level is usually 2 days. The analysis results of all monitoring points are spatially interpolated to form meteorological driving hydrological response characteristic data covering the research area, including the response time, response intensity (such as the change amount of hydrological elements caused by the change of unit meteorological factor) and response type (immediate response or lag response) of each spatial position.

[0046] Step S33: Perform water-salt driving factor coupling characteristic analysis according to the meteorological driving hydrological response characteristic data and the saline-alkali soil hydrological data, and generate water-salt driving factor coupling characteristic data.

[0047] In the embodiment of the present application, hydrological data of saline-alkali soil is collected, including daily variation of underground water level of 30 monitoring points, salt content of surface runoff, soil moisture content profile data (0-20 cm, 20-40 cm, 40-100 cm layering record), and data collection frequency is consistent with meteorological driving hydrological response characteristic data (once a day). The response intensity and response time in the meteorological driving hydrological response characteristic data are coupled and analyzed with the hydrological data: taking the variation of underground water level as a hydrological factor, and taking the variation of salt content of the corresponding soil layer as a salt factor, the synchronous change rate of the two (i.e. the variation of salt content when the underground water level changes per unit value) is calculated, which reflects the driving effect of underground water movement on salt transport; the salt content of surface runoff is multiplied by the surface runoff to obtain the salt leaching flux, and the coupling relationship between the salt leaching flux and the precipitation is analyzed, and the sliding average method (window 15 days) is used to determine the cooperative change rule of precipitation-runoff-salt leaching; for soil moisture content and salt content, the coupling degree model is used to calculate the correlation degree of the two, the model takes the product of the moisture content change rate and the salt change rate as the numerator, and the square root of the sum of the squares of the two change rates as the denominator, and the higher the correlation degree value, the stronger the water-salt coupling. The coupling analysis is carried out for different soil layers (cultivation layer, plow bottom layer, heart soil layer), and the coupling degree, dominant driving factor (such as precipitation dominant or evaporation dominant) and coupling period (such as rainy season or dry season) of each layer are recorded, and all the analysis results are integrated according to the spatial coordinates to generate water-salt driving factor coupling characteristic data, which provides coupling relationship parameters for water-salt movement simulation.

[0048] Further, step S4 comprises the following steps: Step S41: mapping the water-salt driving factor coupling characteristic data to the salt soil layer association characteristic data to perform salt-alkali soil water-salt movement association characteristic analysis of water-salt driving space mapping, and generating salt-alkali soil water-salt movement association characteristic data; In the embodiment of the present application, the spatial coordinate matching relationship of the water-salt driving factor coupling characteristic data and the salt soil layer level correlation characteristic data is established, both of which use the same latitude and longitude coordinate system, so as to ensure that the driving factor data (such as precipitation-salt leaching coupling degree, underground water-salt migration synchronization rate) of each spatial node can be accurately corresponded to the soil layer level correlation data (such as salt correlation degree of each layer, migration intensity) of the node. For each spatial node, the water-salt driving factor is associated and analyzed according to the type (precipitation driving, evaporation driving, underground water driving) and the soil layer structure: the precipitation driving factor is mainly associated with the plough layer and the plough bottom layer, the product of the migration amount of the plough layer salt to the plough bottom layer under the unit precipitation and the correlation degree of the two layers is calculated, so as to reflect the driving effect of the precipitation on the salt migration from the surface layer to the subsurface layer; the evaporation driving factor is mainly associated with the plough layer, the ratio of the potential evaporation amount to the plough layer salt enrichment rate is analyzed, and the driving intensity of the evaporation on the salt accumulation of the surface layer is determined in combination with the plough layer correlation degree; the underground water driving factor is mainly associated with the heart soil layer and the plough bottom layer, the ratio of the underground water level change amount to the salt migration amount of the heart soil layer to the plough bottom layer is calculated, and the driving influence of the underground water movement on the salt migration from the deep layer to the middle layer is quantified in combination with the correlation degree of the two layers. The analysis results of all spatial nodes are subjected to spatial continuity verification, so as to ensure that the driving correlation characteristics of adjacent nodes gradually change reasonably, and finally the salt-alkali soil water-salt movement correlation characteristic data including the correlation strength and direction of each spatial node, each soil layer and each driving type are generated.

[0049] Step S42: performing characteristic fusion splicing processing of water-salt movement time sequence-space-propagation based on the salt-alkali soil water-salt movement correlation characteristic data, to generate water-salt movement fusion characteristic data; In the embodiment of the present application, the correlation characteristic data of water and salt movement of saline-alkali land for 12 consecutive months is collected, and the monthly data includes the correlation intensity of water and salt movement of each spatial node. The sliding time window method is adopted for time sequence characteristic processing, and the window size is set to 3 months. The average value of the correlation intensity of each spatial node in the window is calculated to form the time sequence trend characteristic, reflecting the change rule of water and salt movement with time (such as the difference between rainy season and dry season). The spatial characteristic processing maintains the resolution consistent with the original data, and the data blank area is filled by the Kriging interpolation method to ensure the continuous spatial distribution, and each grid cell corresponds to a group of correlation intensity values. The propagation characteristic processing is based on the data of vertical salt transport relationship of soil layers, and the salt transport path of adjacent layers of each spatial node (such as from plough layer to plough bottom layer to heart soil layer) is extracted. The transport direction and distance are taken as the propagation characteristic parameters. In the fusion and splicing, the time sequence trend characteristic is arranged in time sequence first, and the corresponding spatial distribution characteristic is superimposed on each time node. Then, the propagation characteristic parameters are embedded in the corresponding layer boundary of the spatial distribution (such as marking the transport direction at the junction of plough layer and plough bottom layer). The time sequence, space and propagation characteristics are integrated through the three-dimensional matrix structure, in which the row represents the time sequence, the column represents the spatial coordinate, and the depth dimension represents the propagation path parameter. The fused data needs to be subjected to consistency test to ensure that the propagation characteristics of the same spatial node at different times match the time sequence trend, and finally the water and salt movement fusion characteristic data is generated.

[0050] Step S43: performing saline-alkali land water and salt movement constraint characteristic analysis based on the saline-alkali land hydrological data and the saline-alkali land soil profile data to generate saline-alkali land water and salt movement constraint characteristic data; In the embodiment of the present application, the groundwater level depth, surface runoff, soil saturated hydraulic conductivity in the hydrological data of saline-alkali soil, and the soil bulk density, porosity, clay content in the soil profile data of saline-alkali soil are selected as the basic data for constraint feature analysis. Groundwater level depth constraint analysis: the depth value is divided according to the interval, and the maximum value of the vertical salt transport intensity under different intervals is calculated. When the depth is less than a certain value, the upward salt transport intensity reaches the peak value, and the value is determined as the constraint threshold of groundwater uplift. When the groundwater level exceeds the threshold, the upward salt transport is inhibited, and the constraint relationship is recorded. Soil physical property constraint analysis: the function relationship between water and salt transport rate and the two parameters of bulk density and porosity is established. When the bulk density increases or the porosity decreases, the transport rate decreases. The rate attenuation coefficient is determined by linear fitting as the constraint parameter of soil compactness. The higher the clay content, the stronger the salt adsorption effect. The ratio of clay content to salt transport retardation rate is calculated to determine the adsorption constraint coefficient. Surface runoff constraint analysis: the ratio of runoff to salt leaching flux is calculated. When the runoff exceeds a certain value, the leaching flux no longer increases with the increase of runoff, and the value is determined as the runoff saturation constraint threshold. All constraint parameters are counted according to the soil layers (cultivation layer, plow bottom layer, heart soil layer), and the constraint types (groundwater constraint, physical property constraint, runoff constraint) and corresponding thresholds or coefficients of each layer are determined to generate the salt-alkali soil water-salt movement constraint feature data. The data should correspond to the spatial nodes one by one.

[0051] Step S44: According to the water-salt movement fusion feature data and the salt-alkali soil water-salt movement constraint feature data, the simulation relationship model of salt-alkali soil water-salt movement is designed, and the simulation model of salt-alkali soil water-salt movement is generated.

[0052] In the embodiment of the present application, the water and salt movement fusion feature data is used as the model input variable, including three types of feature parameters of time trend, spatial distribution and propagation path; the water and salt movement constraint feature data of saline-alkali soil is used as the model boundary condition, including constraint thresholds and coefficients at each level. The model adopts a three-dimensional numerical simulation framework, the space dimension is divided into grids according to the soil level (0-20 cm, 20-40 cm, 40-100 cm), the time dimension is taken as a step of each month, and the propagation path is used as the connection parameter between grids. The core equation of the model is constructed based on the water and salt transport conservation law: the water movement equation includes the precipitation infiltration amount, evaporation consumption amount and groundwater level recharge amount, and the water flux at each level is calculated in combination with the soil saturated hydraulic conductivity; the salt movement equation includes the convection term (salt migration with water movement), dispersion term (salt molecular diffusion) and adsorption term (soil particles adsorbing salt), wherein the convection term coefficient comes from the water and salt driving coupling feature, the dispersion term coefficient comes from the salt level correlation strength, and the adsorption term coefficient comes from the soil clay content constraint. The model parameter calibration adopts the historical data (continuous 3-year water and salt measured values) of 20 monitoring points in the research area, and the equation coefficients are adjusted by the least square method, so that the deviation between the model simulation value and the measured value is controlled within the preset range. After calibration, the model can receive the input of different meteorological scenarios (such as different precipitation amounts), output the salt distribution prediction value of each soil level in the future 12 months, and finally generate the water and salt movement simulation model of saline-alkali soil.

[0053] Further, step S42 includes the following steps: Step S421: performing water and salt movement long and short term time series hidden state analysis based on the water and salt movement correlation feature data of saline-alkali soil, generating water and salt movement long and short term time series hidden state data, and performing water and salt movement time series dependence feature analysis on the water and salt movement long and short term time series hidden state data, to generate water and salt movement time series dependence feature data; In the embodiment of the present application, the water and salt movement correlation characteristic data of saline-alkali soil for 18 consecutive months is selected, and is divided into short-term (1-3 months) and long-term (6-12 months) analysis windows according to time granularity. The long and short term time series hidden state analysis adopts time series decomposition method, and the water and salt movement correlation intensity sequence of each spatial node is decomposed into trend item, periodic item and residual item: the trend item is extracted by moving average method (window size is 30 days), and reflects the long-term change rule; the periodic item is extracted by Fourier transform, and captures the seasonal fluctuation characteristics (such as the period caused by the alternation of rainy season and dry season); the residual item is the difference value of the original sequence and the trend item and the periodic item, and reflects the short-term random fluctuation. The trend item and the periodic item are combined as the long-term hidden state, and the residual item is the short-term hidden state, and the long and short term time series hidden state data of water and salt movement is generated. On this basis, the time series dependence characteristic analysis is carried out, the cross correlation coefficient of the short-term hidden state and the long-term hidden state of the previous 1-3 months is calculated, the dependence degree of the short-term fluctuation on the long-term trend is determined; the difference value of the trend item of the two continuous long-term windows is calculated, and the continuity or turning point of the long-term trend is analyzed; the correlation of the same hidden state at different time points is calculated through the autocorrelation function, and the decay rate of time series dependence is determined. The dependence degree, continuity index and decay rate are arranged in time sequence to generate the time series dependence characteristic data of water and salt movement.

[0054] Step S422: performing multi-layer convolution processing of feature vectors of each local space based on the water and salt movement correlation characteristic data of saline-alkali soil, generating water and salt movement spatial convolution data, and performing water and salt movement spatial feature analysis on the water and salt movement spatial convolution data to generate water and salt movement spatial feature data; In the embodiment of the present application, the water-salt movement correlation characteristic data of saline-alkali land is divided into local spatial units according to a 50m*50m grid, each unit contains the water-salt movement correlation intensity, driving factor type and other characteristics in the region, forming a feature vector. The feature vector is processed by multi-layer convolution: the first layer uses a 3*3 grid convolution kernel to calculate the feature mean of each unit and its 8 adjacent units, extracting small-scale local spatial correlation features; the second layer uses a 5*5 grid convolution kernel to calculate the weighted sum of features in a larger range (25 units) (the center unit weight is 0.5, and the total weight of the surrounding units is 0.5), extracting medium-scale spatial correlation features; the third layer uses a 7*7 grid convolution kernel to calculate the maximum value of the features in a larger range, extracting large-scale spatial variation features. The dimension consistency of the feature vector is preserved after each layer of convolution processing, generating water-salt movement spatial convolution data. Based on the data, spatial feature analysis is carried out: calculate the spatial variation coefficient (ratio of standard deviation to mean) of the convolution results at each scale to reflect the spatial heterogeneity at different ranges; determine the spatial autocorrelation range by calculating the rate of change of the eigenvalue with the spatial distance through the semi-variogram function; identify the high-value cluster area (correlation intensity significantly higher than the surrounding) and low-value cluster area in the convolution result, and mark as spatial hotspots and cold spots. Integrate the spatial heterogeneity, autocorrelation range and hotspot distribution at each scale to generate water-salt movement spatial feature data.

[0055] Step S423: based on the water-salt movement correlation characteristic data of saline-alkali land, water-salt movement adjacency correlation feature analysis is carried out, water-salt movement adjacency correlation feature data is generated, and water-salt movement propagation characteristic aggregation analysis is carried out on the water-salt movement adjacency correlation feature data, to generate water-salt movement propagation characteristic data; In the embodiment of the present application, based on the water and salt movement correlation characteristic data of saline-alkali soil, the adjacent nodes of each spatial node in 8 directions (east, south, west, north, southeast, northeast, southwest, northwest) are defined as adjacent units to construct an adjacent relationship network. The water and salt movement correlation intensity difference between the center node and each adjacent unit is calculated, the positive value represents the propagation trend from the center to the adjacent unit, the negative value represents the propagation trend from the adjacent unit to the center, and the absolute value of the difference represents the propagation potential to generate the water and salt movement adjacent correlation characteristic data. On this basis, the propagation characteristics are aggregated and analyzed: according to the soil level (cultivation layer, plow bottom layer, heart soil layer), the propagation potential mean of all adjacent units in the same level is calculated to determine the overall propagation intensity of the level; the adjacent correlation changes in the past 3 months are tracked to record the stability of the propagation direction (such as marking as stable direction if the propagation direction is the same for 3 consecutive months); the ratio of the propagation potential to the water and salt movement correlation intensity is calculated to determine the propagation efficiency (the higher the ratio, the stronger the efficiency). The propagation intensity, stable direction and propagation efficiency of all spatial nodes are regionally aggregated, and the mean value is calculated according to the 100m x 100m block to ensure that the aggregation result retains the spatial distribution characteristics, and finally the water and salt movement propagation characteristic data is generated.

[0056] Step S424: The water and salt movement time sequence dependent characteristic data, the water and salt movement spatial characteristic data and the water and salt movement propagation characteristic data are subjected to water and salt movement feature fusion splicing processing to generate water and salt movement fusion feature data.

[0057] In the embodiment of the present application, the water and salt movement time sequence dependent characteristic data, the water and salt movement spatial characteristic data and the water and salt movement propagation characteristic data are aligned according to the spatial coordinates and time nodes to ensure that the data of the same spatial node and the same time point can be matched. The fusion splicing adopts a three-dimensional matrix structure: the first dimension is the time sequence (divided by month, a total of 12 time points), the second dimension is the spatial coordinates (according to the row and column numbers of the 50m x 50m grid), and the third dimension is the feature type (including 12 core features such as the attenuation rate in the time sequence dependent feature, the hot spot mark in the spatial feature, and the stable direction in the propagation characteristic). During the fusion process, the characteristic values of different magnitudes are normalized (mapped to the 0-1 interval) to avoid feature weight imbalance caused by numerical difference. For the time dimension, the spatial feature and the propagation characteristic data of each month are corresponded to the time sequence dependent feature of the month to form a three-dimensional correlation of "time - space - feature"; for the spatial dimension, the time sequence change and the propagation characteristic of the position are integrated in each grid unit to form a local correlation of "space - time - feature". After fusion, consistency verification is performed, the correlation coefficient of any two features at the same space-time position is calculated, the abnormal values with an absolute correlation coefficient less than 0.3 are removed and corrected by interpolation method, and finally the water and salt movement fusion feature data with complete structure and consistent correlation is generated.

[0058] Further, as an embodiment of the present application, referring to Figure 2 shown, as Figure 1 The detailed step flow diagram of step S5 in the embodiment, step S5 in the embodiment includes the following steps: Step S51: Perform saline-alkali farmland irrigation simulation parameter design through saline-alkali farmland irrigation data and a saline-alkali farmland water-salt movement simulation model, and generate saline-alkali farmland irrigation simulation parameters; In the embodiment of the present application, saline-alkali farmland irrigation data is collected, including irrigation records for 3 consecutive years, and the data covers the time, duration, irrigation method (flood irrigation, drip irrigation), irrigation amount per unit area, and salt content changes of each soil layer (0-20 cm, 20-40 cm, 40-100 cm) before and after irrigation. These data are combined with the saline-alkali farmland water-salt movement simulation model to carry out irrigation simulation parameter design. For the irrigation amount parameter, the ratio of the irrigation amount per unit area to the salt leaching amount of the plough layer in the historical irrigation is calculated to determine the stable range of the ratio under different irrigation methods, which is used as the reference parameter of the correlation between the irrigation amount and the salt leaching in the model; for the irrigation cycle parameter, the relationship between the interval length of consecutive irrigations and the salt accumulation amount of the plough layer is analyzed, and the change rate of the salt accumulation when the interval length increases by a unit value is determined through linear fitting, which is used as the core index of the cycle parameter; for the irrigation intensity parameter, based on the ratio of the duration to the irrigation amount, the ratio of the water infiltration rate to the salt migration rate with water is calculated combined with the soil saturated hydraulic conductivity, to ensure that the ratio matches the soil porosity (the higher the porosity, the higher the ratio). At the same time, the difference in irrigation methods is converted into a parameter coefficient (flood irrigation is set to 1.0, drip irrigation is set to 0.6, reflecting the difference in water distribution uniformity), and all parameters need to be set for each soil layer to ensure that they correspond to the water-salt movement characteristics of each layer in the model, and finally generate saline-alkali farmland irrigation simulation parameters including irrigation amount, cycle, intensity, and method coefficient.

[0059] Step S52: Transmit the saline-alkali farmland irrigation simulation parameters to the saline-alkali farmland water-salt movement simulation model for farmland irrigation water-salt movement simulation analysis, and generate farmland irrigation water-salt movement simulation data; In the embodiment of the present application, the simulated parameters of saline-alkali farmland irrigation are input into the saline-alkali soil water and salt movement simulation model according to the time sequence, the model is run according to the monthly time step, the spatial resolution is maintained at 50 meters x 50 meters grid, and the soil layers are divided into 0-20 cm, 20-40 cm, and 40-100 cm. During the simulation process, the model first calculates the water infiltration amount according to the irrigation amount and irrigation intensity parameters, the water infiltration amount in the plough layer (0-20 cm) is allocated at 70% of the irrigation amount, the plough sole layer (20-40 cm) is allocated at 20%, and the subsoil layer (40-100 cm) is allocated at 10%, and the allocation ratio is determined according to the difference in soil porosity (the allocation ratio of the layer with high porosity is high). Water infiltration drives salt migration, the model calculates the migration amount of salt in each layer with water by the convection-dispersion equation, wherein the convection term coefficient is determined by the ratio of irrigation intensity to soil bulk density, and the dispersion term coefficient is determined by reference to the salt layer correlation intensity data. The change of water content and salt content in each grid unit and each layer is recorded at each step, the salt distribution after irrigation for 24 hours, 72 hours, and 168 hours is calculated synchronously, and the short-term and medium-term change trend is reflected. After the simulation is completed, the results are compared with the measured water and salt data after historical irrigation, the deviation of the simulation value and the measured value of each grid unit is calculated, if the deviation exceeds the preset range, the irrigation parameters are adjusted for re-simulation until the deviation meets the requirements, and finally the farmland irrigation water and salt movement simulation data is generated.

[0060] Step S53: The farmland irrigation water and salt movement simulation data is analyzed for irrigation water and salt flow direction simulation, irrigation water and salt flow direction simulation data is generated, and the water and salt movement propagation influence characteristic data is generated by analyzing the water and salt movement propagation influence characteristics of the irrigation characteristics through the irrigation water and salt flow direction simulation data; In the embodiment of the present application, the flow direction simulation analysis is performed on the simulation data of the salt movement of farmland irrigation water, and the water flux and salt flux of each soil layer at each time node (24 hours, 72 hours, and 168 hours after irrigation) are extracted. The water flux is calculated by the ratio of the change amount of water content of each layer to time, and the salt flux is calculated by the ratio of the change amount of salt content to the water flux. The consistency of the directions of the two (both positive or both negative) indicates that the salt moves in the same direction as the water. The direction of the flux of each grid cell is tracked. When the water flux of a certain layer is positive (downward) and the salt flux is also positive, it is marked as salt downward flow; when the water flux is negative (upward) and the salt flux is negative, it is marked as salt upward flow. In this way, the vertical and horizontal flow paths of the salt of irrigation water are determined, and the salt flow direction simulation data of irrigation water is generated. Based on the data, the propagation influence characteristic analysis is performed: the salt propagation speed (the vertical distance of salt migration per unit time) under different irrigation methods is calculated, the propagation gradient is determined by the difference of the salt fluxes of adjacent layers, the range of the influence of irrigation on each layer of salt (the number of grids whose salt change amount exceeds 10% of the initial value) is counted, the range expansion rate is calculated, the ratio of the propagation speed to the irrigation intensity is analyzed, and the propagation efficiency brought by unit irrigation amount is determined. The propagation speed, gradient, range, and efficiency are classified and arranged according to the soil layer and irrigation method, and the water and salt movement propagation influence characteristic data is generated.

[0061] Step S54: Perform salt division flushing influence characteristic analysis on saline-alkali soil according to the water and salt movement propagation influence characteristic data, and generate saline-alkali soil salt division flushing influence characteristic data; In the embodiment of the present application, based on the water-salt movement propagation influence characteristic data, the flushing effect of salt carried by water after irrigation is focused on, and the salt flushing influence characteristic analysis is carried out. The propagation speed, propagation gradient, propagation range and other parameters of each soil layer (0-20 cm, 20-40 cm, 40-100 cm) in the data are extracted, and the amount of salt flushed in a unit of time is calculated: the product of the propagation speed and the initial salt content of the layer is taken as the basic flushing amount, and then multiplied by the propagation gradient (reflecting the change of flushing intensity with distance) to obtain the actual flushing amount at different depths. For the plough layer, the flushing efficiency within 24 hours after irrigation is analyzed, and the salt flushing capacity per irrigation amount is determined by the ratio of the flushing amount to the irrigation amount; for the plough sole layer, the flushing lag effect within 72 hours is analyzed, and the cumulative rate of the flushing amount with time (i.e. the flushing amount increment every 24 hours) is calculated; for the heart soil layer, the flushing residual amount after 168 hours is analyzed, that is, the difference between the initial salt content and the cumulative flushing amount, which reflects the degree of difficulty of flushing deep salt. At the same time, combined with the soil porosity data, the ratio of porosity to flushing amount is calculated to determine the influence of soil structure on flushing effect (high porosity ratio means more sufficient flushing). The flushing amount, flushing efficiency, lag rate and residual amount of each layer are arranged according to the spatial grid unit to generate the salt flushing influence characteristic data of saline-alkali soil, and the data should correspond to the irrigation method (flood irrigation, drip irrigation) to reflect the flushing difference under different irrigation methods.

[0062] Step S55: analyzing the outlet flow boundary condition of the saline-alkali soil based on the saline-alkali soil water-salt movement simulation model, and generating the outlet flow boundary condition data of the saline-alkali soil; In the embodiments of the present application, based on the simulation model of water and salt movement in saline-alkali soil, the outlet position of the research area is determined (the intersection points of natural drainage channels and artificial drainage ditches in the area are selected, and a total of 3 main outlets are set), and the boundary condition analysis of outlet flow direction is carried out. When the model is running, monitoring nodes are set at the outlet position, and water and salt outflow data for 12 consecutive months are recorded, including daily outflow, salt concentration, and outflow direction (flow direction angle along the drainage channel). The correlation between water and salt outflow at the outlet and the water and salt movement in the region is analyzed: the ratio of daily outflow at the outlet to total irrigation amount in the region on the same day is calculated to determine the proportion of irrigation water converted into outlet drainage; the difference between outlet salt concentration and average salt content in the cultivated layer in the region is calculated to determine the attenuation rate of salt during migration to the outlet; by comparing the outlet flow direction angle with the terrain slope, the constraint relationship of terrain on outlet flow direction is determined (1 degree of slope change corresponds to an adjustment value of flow direction angle). For different seasons (rainy season and dry season), the seasonal differences of boundary conditions are determined: the ratio of outlet outflow to irrigation amount in the rainy season is higher than that in the dry season, and the outlet salt concentration attenuation rate in the dry season is higher than that in the rainy season. The relationship between outlet position coordinates, outflow ratio, salt attenuation rate, flow direction angle, and terrain, and seasonal parameters are integrated to generate saline-alkali soil outlet flow direction boundary condition data, which needs to clearly define the independent boundary parameters of each outlet to ensure accurate docking with the model spatial grid.

[0063] Step S56: Perform salt and water outlet flow direction analysis of saline-alkali soil according to the water and salt movement propagation influence characteristic data and the saline-alkali soil outlet flow direction boundary condition data, and generate saline-alkali soil water and salt outlet flow direction data; In the embodiment of the present application, the water and salt outlet flow direction analysis is carried out in combination with the water and salt movement propagation influence characteristic data and the salt and alkali land outlet flow direction boundary condition data. First, the water and salt flow direction path (vertical and horizontal direction) in the propagation influence characteristic data is superimposed with the terrain elevation data of the research area to determine the potential path (in the direction of decreasing elevation) of water and salt migration from each grid cell to the outlet. For each grid cell, the ratio of the distance to the three main outlets and the propagation speed is calculated to determine the time of water and salt reaching each outlet; by comparing the arrival time, the main outlet direction of the water and salt of the cell (the outlet with the shortest time) is determined. Secondly, in combination with the outflow proportion in the outlet boundary condition data, the contribution amount of each grid cell to the corresponding outlet (the product of the water and salt migration amount in the cell and the outflow proportion) is calculated, and the sum of the contribution amounts should be equal to the total migration amount of the region. The contribution difference of different soil layers to the outlet flow direction is analyzed: the water and salt in the plough layer mainly flow to the nearest outlet (short migration time), and the water and salt in the subsoil layer mainly flow to the outlet at the lowest place in the region due to the slow migration speed. At the same time, the weighted average value of the salt concentration at the outlet and the salt concentration of each contribution cell (the weight is the contribution amount proportion) is calculated to verify the consistency of the flow direction analysis (the weighted average value should be within the preset range of the measured concentration at the outlet). The main outlet direction, contribution amount and migration time of each grid cell are integrated to generate the water and salt outlet flow direction data of the salt and alkali land, and the data is represented by a vector line to represent the flow direction path, and the line thickness represents the contribution amount.

[0064] Step S57: The water and salt movement simulation prediction of the salt and alkali land farmland irrigation is carried out on the salt and alkali land salt flushing influence characteristic data and the salt and alkali land water and salt outlet flow direction data to generate the water and salt movement simulation prediction data of the salt and alkali land farmland irrigation.

[0065] In the embodiment of the present application, the salt flushing influence characteristic data of saline-alkali soil and the water-salt outlet flow direction data of saline-alkali soil are integrated, and based on the spatial coordinate correlation of the two, the simulation prediction of farmland irrigation water-salt movement is carried out. The prediction period is set to 6 months in the future, and the time step is one month. The spatial resolution is maintained at a 50m x 50m grid. In the prediction process, first, according to the salt flushing influence characteristic data, the monthly salt flushing amount of each grid unit and each soil layer is determined (based on the historical flushing efficiency and the predicted irrigation amount); then, combined with the water-salt outlet flow direction data, the proportion of the flushing salt flowing to each outlet in each month is determined (based on the contribution ratio), and the monthly salt discharge amount of the outlet is calculated. For the plough layer, the salt content in the 24 hours after irrigation is predicted to decrease rapidly, and then slowly rise in the next 28 days (affected by the evaporation of underground water carrying salt replenishment); for the plough bottom layer, the salt content is predicted to start to decrease after 72 hours, and the decrease rate is 60% of that of the plough layer (affected by the higher soil bulk density); for the heart soil layer, the salt content is predicted to start to decrease slowly after 30 days, and the decrease amount is 30% of that of the plough layer. At the same time, the monthly total outflow and salt concentration at the outlet are predicted. The outflow needs to maintain water balance with the predicted irrigation amount and the regional evaporation amount (total outflow = total irrigation amount - total evaporation amount - change of soil water storage). The predicted salt content of each layer and the predicted discharge amount of the outlet are integrated according to time and space to generate simulation prediction data of farmland irrigation water-salt movement in saline-alkali soil, which needs to include the spatial distribution map and the outlet monitoring curve of each time step.

[0066] Therefore, from any viewpoint, the embodiments should be considered as exemplary and non-limiting, the scope of the present application being defined by the appended claims and not by the above description, and it is intended to embrace all variations falling within the meaning and range of equivalents of the claims of the application.

[0067] The foregoing is considered as a specific implementation of the present application, enabling those skilled in the art to understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A simulation method for soil water and salt movement in saline-alkali soil based on big data analysis, characterized in that, Comprise the following steps: Step S1: using saline-alkali soil multi-source monitoring system for saline-alkali soil soil multi-source monitoring data acquisition, to obtain saline-alkali soil multi-source monitoring data, wherein the saline-alkali soil multi-source monitoring data includes saline-alkali soil remote sensing image data, saline-alkali soil soil profile data, saline-alkali soil meteorological data, saline-alkali soil hydrological data and saline-alkali soil farmland irrigation data; Step S2: based on saline-alkali soil remote sensing image data and saline-alkali soil soil profile data, the salt soil layer correlation characteristic analysis of saline-alkali soil space is carried out, and the salt soil layer correlation characteristic data is generated; Step S3: based on saline-alkali soil meteorological data and saline-alkali soil hydrological data, the water-salt driving factor coupling characteristic analysis of saline-alkali soil is carried out, and the water-salt driving factor coupling characteristic data is generated; Step S4: through the salt soil layer correlation characteristic data and the water-salt driving factor coupling characteristic data, the simulation relationship model of saline-alkali soil water-salt movement is designed, and the simulation model of saline-alkali soil water-salt movement is generated; Step S5: through the saline-alkali soil farmland irrigation data and the simulation model of saline-alkali soil water-salt movement, the simulation parameter design of saline-alkali soil farmland irrigation is carried out, and the simulation parameter of saline-alkali soil farmland irrigation is generated; the water-salt movement simulation prediction of saline-alkali soil farmland irrigation is carried out on the simulation parameter of saline-alkali soil farmland irrigation, and the water-salt movement simulation prediction data of saline-alkali soil farmland irrigation is generated. 2.The simulation method of soil water and salt movement in saline and alkaline soil based on big data analysis according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: according to the saline-alkali soil remote sensing image data, the remote sensing image multispectral feature analysis processing is carried out, and the remote sensing image multispectral feature data is generated; Step S22: based on the remote sensing image multispectral feature data, the multispectral salt content index inversion processing is carried out, and the multispectral salt content index inversion data is generated; Step S23: design the saline-alkali soil profile index data; Step S24: according to the saline-alkali soil profile index data, the soil layer structure characteristic analysis of each space node of the saline-alkali soil profile data is carried out, and the soil layer structure characteristic data is generated; Step S25: the multispectral salt content index inversion data is mapped to the soil layer structure characteristic data to carry out salt distribution mapping processing, and the soil salt distribution-layer structure characteristic data is generated; Step S26: the salt soil layer correlation characteristic analysis is carried out on the soil salt distribution-layer structure characteristic data, and the salt soil layer correlation characteristic data is generated. 3.The simulation method of soil water and salt movement in saline and alkaline soil based on big data analysis according to claim 2, characterized in that, Step S22 includes the following steps: Step S221: according to the remote sensing image multispectral feature data, the salt sensitive spectral feature channel extraction is carried out, so as to obtain the salt sensitive spectral feature channel data; and the salt sensitive spectral feature channel data is analyzed to generate the spectral band salt sensitive influence weight data; Step S222: according to the salt sensitive spectral feature channel data, the salt sensitive spectral band feature analysis is carried out on the remote sensing image multispectral feature data, and the salt sensitive spectral band feature data is generated; Step S223: based on the spectral band salt sensitive influence weight data and the salt sensitive spectral feature channel data, the saline-alkali soil multispectral salt content evaluation model is established corresponding to the pre-acquired ground measured salt spectrum data; Step S224: The salt-sensitive spectral band feature data is processed by the multi-spectral salt evaluation model to generate multi-spectral salt index inversion data. 4.The simulation method of soil water and salt movement in saline and alkaline soil based on big data analysis according to claim 2, characterized in that, The soil profile index data of the saline-alkali soil in step S23 includes soil bulk density index, soil porosity index, soil moisture content index, and soil salt content index. 5.The simulation method of soil water and salt movement in saline and alkaline soil based on big data analysis according to claim 2, characterized in that, Step S26 includes the following steps: Step S261: Analyzing the time-lag correlation of soil layer salt vertical migration based on the soil salt distribution-hierarchical structure feature data to generate soil layer salt vertical migration relationship data; Step S262: Analyzing the partial correlation characteristics of salt vertical migration based on the soil layer hierarchical structure feature data to generate salt vertical migration partial correlation feature data; Step S263: Analyzing the salt soil layer correlation characteristics based on the soil layer salt vertical migration relationship data and the salt vertical migration partial correlation feature data to generate salt soil layer correlation characteristic data. 6.The simulation method of soil water and salt movement in saline and alkaline soil based on big data analysis according to claim 5, characterized in that, Step S261 includes the following steps: Analyzing the salt cross-time lag correlation of the soil layer hierarchical structure based on the soil salt distribution-hierarchical structure feature data to generate soil layer hierarchical structure salt cross-time lag correlation data; Analyzing the soil layer salt vertical migration relationship based on the soil layer hierarchical structure salt cross-time lag correlation data to generate soil layer salt vertical migration relationship data. 7.The simulation method of soil water and salt movement in saline and alkaline soil based on big data analysis according to claim 1, wherein, Step S3 includes the following steps: Step S31: Analyzing the hydrological derivative correlation data of multi-source meteorological factors based on the meteorological data of the saline-alkali soil to generate meteorological factor hydrological derivative correlation data; Step S32: Analyzing the meteorological driving hydrological response characteristics based on the meteorological factor hydrological derivative correlation data to generate meteorological driving hydrological response characteristic data; Step S33: Analyzing the water-salt driving factor coupling characteristics based on the meteorological driving hydrological response characteristic data and the saline-alkali soil hydrological data to generate water-salt driving factor coupling characteristic data. 8.The simulation method of soil water and salt movement in saline and alkaline soil based on big data analysis according to claim 1, wherein, Step S4 includes the following steps: Step S41: Mapping the water-salt driving factor coupling characteristic data to the salt soil layer correlation characteristic data to analyze the water-salt movement correlation characteristics of the saline-alkali soil based on the water-salt driving space mapping to generate saline-alkali soil water-salt movement correlation characteristic data; Step S42: Processing the feature fusion splicing of water-salt movement time sequence-space-propagation based on the saline-alkali soil water-salt movement correlation characteristic data to generate water-salt movement fusion feature data; Step S43: Analyzing the saline-alkali soil water-salt movement constraint characteristics based on the saline-alkali soil hydrological data and the saline-alkali soil soil profile data to generate saline-alkali soil water-salt movement constraint characteristic data; Step S44: Designing the simulation relationship model of the saline-alkali soil water-salt movement based on the water-salt movement fusion feature data and the saline-alkali soil water-salt movement constraint characteristic data to generate the saline-alkali soil water-salt movement simulation model. 9.The simulation method of soil water and salt movement in saline and alkaline soil based on big data analysis according to claim 8, characterized in that, Step S42 includes the following steps: Step S421: based on the saline-alkali soil water salt movement correlation characteristic data, long and short term time sequence hidden state analysis of water salt movement is performed, water salt movement long and short term time sequence hidden state data is generated, water salt movement time sequence dependence characteristic analysis is performed on the water salt movement long and short term time sequence hidden state data, and water salt movement time sequence dependence characteristic data is generated; Step S422: based on the saline-alkali soil water salt movement correlation characteristic data, feature vector multi-layer convolution processing of each local space is performed, water salt movement space convolution data is generated, and water salt movement space characteristic analysis is performed on the water salt movement space convolution data, and water salt movement space characteristic data is generated; Step S423: based on the saline-alkali soil water salt movement correlation characteristic data, water salt movement adjacency correlation characteristic analysis is performed, water salt movement adjacency correlation characteristic data is generated, and water salt movement propagation characteristic aggregation analysis is performed on the water salt movement adjacency correlation characteristic data, and water salt movement propagation characteristic data is generated; Step S424: water salt movement feature fusion splicing processing is performed on the water salt movement time sequence dependence characteristic data, the water salt movement space characteristic data and the water salt movement propagation characteristic data, and water salt movement fusion feature data is generated. 10.The simulation method of soil water and salt movement in saline and alkaline soil based on big data analysis according to claim 1, wherein, Step S5 includes the following steps: Step S51: saline-alkali soil farmland irrigation simulation parameter design is performed through the saline-alkali soil farmland irrigation data and the saline-alkali soil water salt movement simulation model, and saline-alkali soil farmland irrigation simulation parameters are generated; Step S52: the saline-alkali soil farmland irrigation simulation parameters are transmitted to the saline-alkali soil water salt movement simulation model for farmland irrigation water salt movement simulation analysis, and farmland irrigation water salt movement simulation data are generated; Step S53: irrigation water salt flow direction simulation analysis is performed on the farmland irrigation water salt movement simulation data, irrigation water salt flow direction simulation data are generated, and water salt movement propagation influence characteristic analysis is performed on the irrigation water salt flow direction simulation data, and water salt movement propagation influence characteristic data are generated; Step S54: saline-alkali soil salt flushing influence characteristic analysis is performed according to the water salt movement propagation influence characteristic data, and saline-alkali soil salt flushing influence characteristic data are generated; Step S55: saline-alkali soil outlet flow boundary condition analysis is performed based on the saline-alkali soil water salt movement simulation model, and saline-alkali soil outlet flow boundary condition data are generated; Step S56: saline-alkali soil water salt outlet flow analysis is performed according to the water salt movement propagation influence characteristic data and the saline-alkali soil outlet flow boundary condition data, and saline-alkali soil water salt outlet flow data are generated; Step S57: saline-alkali soil farmland irrigation water salt movement simulation prediction is performed on the saline-alkali soil salt flushing influence characteristic data and the saline-alkali soil water salt outlet flow data, and saline-alkali soil farmland irrigation water salt movement simulation prediction data are generated.

Images