A method for monitoring and regulating the relationship between water and salt migration driving carbonate migration time at field scale

Abstract

The application discloses a kind of field scale water salt migration drive carbonate migration time relationship monitoring and control method, comprising: based on different irrigation system setting field test processing, using profile water temperature salt monitoring system arranged in each test plot automatically continuous acquisition high-frequency data;Based on different irrigation mode stage acquisition low-frequency data;Using analytical detection technology to process low-frequency data, obtain low-frequency carbonate index data;Based on low-frequency data and high-frequency data, establish deviation function and carry out regression correction to high-frequency parameter to obtain calibrated high-frequency data;Calibrated high-frequency data and low-frequency carbonate index data are processed, to obtain synchronized water salt parameter sequence and carbonate index sequence;Using time lag regression modeling method to process synchronized water salt parameter sequence and carbonate index sequence, to obtain prediction model;Based on prediction model, identify key water salt driving factor and corresponding lag time, and generate irrigation control decision.

Description

Technical Field

[0001] This invention belongs to the field of soil science and agricultural water conservancy engineering technology, and in particular relates to a method for monitoring and regulating the temporal relationship between water and salt transport and carbonate migration at the field scale. Background Technology

[0002] Saline soils are widely distributed in arid and semi-arid regions and are one of my country's important agricultural resource types. Due to high evaporation, limited rainfall, and high salinity of irrigation water, these soils are prone to salt accumulation and uneven water-salt distribution, thus affecting soil structure, nutrient supply capacity, and crop productivity. Simultaneously, the migration and transformation of carbonates in saline soils are directly regulated by water-salt transport, thereby influencing soil carbon cycling, soil carbon fixation capacity, and soil improvement effects. Therefore, in-depth research into the spatiotemporal regulation mechanism of water-salt transport on carbonate migration is of great significance for the scientific management and sustainable utilization of saline soils.

[0003] Currently, research on the relationship between water and salt transport and carbonate migration in saline soils mainly focuses on indoor soil column simulations or sporadic manual sampling experiments. While these methods can provide basic insights into the underlying patterns, they have significant limitations in field-scale applications: 1) Insufficient continuous dynamic monitoring: Indoor simulations or intermittent sampling cannot continuously monitor changes in soil water, salt, and carbonate levels under real field conditions, making it difficult to reflect the complex seasonal and diurnal variations in the natural environment. 2) Limited temporal resolution: Traditional manual sampling has long cycles, typically on a daily or monthly scale, making it difficult to capture the short-term response characteristics of water and salt transport to carbonate migration, such as the instantaneous dissolution, migration, and deposition processes after irrigation events. 3) Limited spatial coverage: Existing methods typically monitor a single profile or a small number of soil layers, making it difficult to comprehensively reflect the spatial heterogeneity of water and salt states and carbonate migration at different depths of the soil profile. 4) Lack of comprehensive control methods: There is a lack of systematic technical means to integrate different irrigation patterns, meteorological conditions, and long-term monitoring data, making it difficult to establish quantitative and predictable time-driven models between water and salt transport and carbonate migration, and also making it difficult to provide an operational decision-making basis for field irrigation management and carbonate control.

[0004] Therefore, there is an urgent need for a technical means applicable to field conditions that can monitor soil water and salt status and carbonate migration in a long-term, continuous, and systematic manner, and establish a time-driven relationship model. This method can combine high-frequency sensor data and periodic soil sample analysis to achieve quantitative analysis and dynamic prediction of water-salt-carbonate processes, providing a scientific basis for optimizing irrigation regimes in saline soils, salt management, and soil carbon cycle regulation, while also overcoming the shortcomings of existing technologies such as short research cycles, isolated parameters, and lack of dynamic response. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a method for monitoring and regulating the temporal relationship between water-salt transport and carbonate migration at the field scale, thereby resolving the issues present in the prior art.

[0006] To achieve the above objectives, this invention provides a method for monitoring and regulating the temporal relationship between water-salt transport and carbonate migration at the field scale, comprising: Field trials were conducted based on different irrigation regimes. A profile water, temperature and salinity monitoring system deployed in each experimental plot was used to automatically and continuously collect high-frequency data, including soil moisture, salinity and temperature data. Based on preset time intervals and spatial depths, low-frequency data are manually collected in stages, including soil solution, groundwater samples, and profile soil samples. The low-frequency data is processed using analytical detection techniques to obtain low-frequency carbonate index data; Using the processing results of the soil samples from the profile as a reference value, a deviation function is established based on the low-frequency data and the high-frequency data, and regression correction is performed on the high-frequency parameters to obtain calibrated high-frequency data; The calibrated high-frequency data and the low-frequency carbonate index data are sequentially subjected to time scale unification, numerical standardization and time series matching to obtain synchronized water and salt parameter sequences and carbonate index sequences. The synchronized water-salt parameter sequence and carbonate index sequence were processed using the time-delay regression modeling method to obtain a predictive model describing the time relationship between water-salt transport and carbonate migration. Based on the prediction model, key water and salt driving factors and their corresponding lag times are identified, and irrigation regulation decisions are generated.

[0007] Optionally, the preset time interval is any time interval within the range of 5-30 minutes; The preset depth is the depth of multiple soil layers covering the cultivated layer and root zone, and the multiple soil layers include several layers within the range of 0-120cm.

[0008] Optionally, the process of manually collecting soil solution, groundwater, and profile soil samples in stages based on preset time intervals and depths includes: During the irrigation period, soil solution and groundwater samples were collected at a sampling frequency ranging from minutes to hours. During non-irrigation periods, soil solution and groundwater samples were collected at a sampling frequency of weekly to monthly. During the crop growing season, profile soil samples are collected at several soil depths on a monthly timescale.

[0009] Optionally, the process of sequentially performing time-scale unification, numerical standardization, and time-series matching on the calibrated high-frequency data and the low-frequency carbonate index data to obtain synchronized water-salt parameter sequences and carbonate index sequences includes: The calibrated high-frequency data were processed using a time aggregation method to obtain a water-salt parameter sequence with a uniform time scale; The low-frequency carbonate index data is processed using a smoothing method to obtain a carbonate index sequence that is consistent with the time scale of the water-salt parameter sequence.

[0010] Optionally, the process of using time-delay regression modeling to process the synchronized water-salt parameter sequence and carbonate index sequence to obtain a predictive model describing the time relationship between water-salt transport and carbonate migration includes: Based on time series data of soil moisture content, salt concentration and temperature, the salt diffusion-convection equation was used to calculate the salt flux time series. The salt flux time series, soil moisture content series, evaporation force series, groundwater depth series and temperature series are used as candidate driving variables. Based on the candidate driving variables and the carbonate index change series, time lag correlation analysis is performed to determine the potential lag time of each candidate driving variable's influence on carbonate migration. Based on the potential lag time, a predictive model is constructed to characterize the time relationship between water-salt transport and carbonate migration.

[0011] Optionally, the calculation expression for the evaporation force sequence is: ; In the formula, Δ is the slope of the curve relating saturated water vapor pressure and air temperature, with units of kPa °C. -1 ; The net radiation input to the canopy is expressed in MJ / m². -2 d -1 ; G Soil heat flux, in MJ / m³ -2 d -1 ; c This is the hygrometer constant, in kPa °C. -1 ; T The average daily temperature at a height of 2 m above the ground is expressed in °C. Wind speed at a height of 2 m above the ground, in milliseconds (ms). -1 ; This represents the pressure difference between saturated and actual water vapor, expressed in kPa. C n and C d The first and second constants are determined by the reference crop type and the calculation time step, respectively.C n The unit is K mm s 3 Mg -1 h -1 The unit of Cd is sm. -1 ; Evaporation force, unit: mm h -1 .

[0012] Optionally, the process of constructing the prediction model includes: constructing a multiple lag regression model based on the carbonate change rate and water-salt transport driving factors, wherein the multiple lag regression model is: ; In the formula, In order to be in t Rate of change of soil carbonate content per unit time at any given moment ； These are the lagged and standardized driving variables; t Indicates the monitoring time or monitoring moment; For the first i The time lag of the effect of each water-salt transport driver variable on carbonate changes; For the first m Indicator variables for the irrigation event; This is the constant term in the regression model; The first regression coefficient, The second regression coefficient; This is the random error term.

[0013] Optionally, the process of identifying key water-salt driving factors and their corresponding lag times based on the prediction model includes: Based on the regression coefficients of the prediction model, time series analysis is performed using a sliding window of a preset duration; When the regression coefficient of the key water-salt driving factor is significantly higher than the historical benchmark level within a certain number of consecutive time windows, the corresponding time period is determined to be the critical driving period of the current key water-salt driving factor.

[0014] Optionally, the process of generating irrigation regulation decisions includes: predicting the risk of carbonate migration over a future set time period based on the prediction model; and triggering an irrigation system adjustment instruction when the predicted value exceeds a preset risk threshold, wherein the adjustment includes at least one of adjusting irrigation time, single irrigation volume, or number of irrigations.

[0015] Compared with the prior art, the present invention has the following advantages and technical effects: This invention, by setting up different irrigation patterns in the field and combining continuous monitoring of water and salt profiles with soil-water sample collection and analysis, can systematically reveal the driving effect and temporal relationship between water and salt transport and carbonate migration. This method provides strong data continuity and reliable results, offering technical support for saline soil improvement and carbon cycle research, and has significant application and promotion value. Attached Figure Description

[0016] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a profile monitoring system according to an embodiment of the present invention; Figure 2 This is a flowchart illustrating the monitoring, sampling, and calculation processes in an embodiment of the present invention. Detailed Implementation

[0017] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0018] It should be noted that the steps shown in the flowcharts in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flowcharts, in some cases, the steps shown or described may be performed in a different order than that shown here. Without conflict, the embodiments and features in the embodiments of this application can be combined with each other. The following embodiments are only used to illustrate the technical solutions of this application and are not intended to limit the scope of protection of this application.

[0019] Example 1 like Figure 1 and Figure 2 As shown, this embodiment provides a method for monitoring and regulating the temporal relationship between water-salt transport and carbonate migration at the field scale, including the following steps: S1. Experimental Site Setup: Three irrigation modes were set up within the experimental station: Summer irrigation: 400 mm of water, divided into four irrigations, implemented in May-June; Summer irrigation + autumn irrigation: Summer irrigation as above, plus an additional 200 mm autumn irrigation from July to September; Summer irrigation + autumn irrigation + post-autumn irrigation: In addition to the first two, a post-autumn irrigation of 100 mm was added from September-October. Each treatment was replicated in three places, with a plot area of ​​50 m². 2(5 m × 10 m), with plots spaced 1 m apart, and impermeable barriers are installed to prevent lateral seepage. Irrigation is achieved using surface canals, equipped with flow meters (accuracy ±2%) to control the irrigation volume. Sunflowers are the crop planted. Plot management (crop planting, tillage, fertilization, and canal irrigation) is consistent with local field practices. The above irrigation system and crop type are specific settings in this embodiment and can be adjusted according to the research object and regional conditions.

[0020] Furthermore, the irrigation system for the three irrigation models in 2025 is shown in Table 1: Table 1 In practice, the irrigation time can be adjusted appropriately according to meteorological conditions and crop growth period, while maintaining consistency between treatments, but the total water volume should remain the same, so as to compare the water and salt transport and carbonate migration characteristics under different irrigation modes.

[0021] S2. Deployment of the profile monitoring system: Deploy profile water, temperature and salinity monitoring devices in each test plot to continuously collect high-frequency data. The high-frequency data includes soil moisture, salinity and temperature data. The sampling interval is on the minute time scale, which is 10 minutes in this embodiment.

[0022] In other implementations, any time between 5 and 30 minutes may be used.

[0023] Soil moisture, salinity, and temperature profile monitoring systems were deployed at the center of each experimental plot to continuously observe soil water and salinity dynamics and temperature changes. The main components of the monitoring system include an MTD15 three-parameter sensor (volume moisture content sensor). i Soil electrical conductivity EC and soil temperature T ), wireless data acquisition module and groundwater level monitoring equipment.

[0024] (1) Monitoring depth and deployment method: The soil profile monitoring depth is set to multiple depth layers covering the topsoil, root zone and subsoil. In this embodiment, these depths are 0, 15, 30, 45, 60, 90 and 120 cm to cover the water and salt characteristics of the surface, root zone and subsoil. Sensors are vertically deployed at the center point of each plot to ensure that the profile data can represent the overall water and salt status of the plot.

[0025] (2) Sensor acquisition and recording: MTD15 three-parameter sensor real-time recording: soil volumetric water content i ( z,t ), unit m 3 m -3 Soil salinity (expressed as electrical conductivity EC), unit dS m -1 This can be converted into salt concentration. C ( z,tSoil temperature T ( z, t (Unit: °C). Data sampling interval is 10 minutes to ensure high-frequency monitoring captures irrigation events and rapid water-salt response.

[0026] (3) Data transmission and storage: The sensor transmits data to the field data acquisition unit via a wireless module; the data acquisition unit automatically uploads the data to the remote server via the GPRS network, realizing real-time remote monitoring and long-term continuous recording. All data is stored according to the original timestamp and backed up locally and on the server to ensure data integrity and traceability.

[0027] (4) Groundwater level monitoring: Install perforated water level gauges in the underground area of ​​the community to monitor the groundwater depth. H ( t Groundwater depth data is synchronized with soil water and salt data in time, and is used for subsequent construction of water and salt flux and carbonate migration models. The water level gauge has an accuracy of ±0.5 cm and can capture short-term groundwater fluctuations after irrigation.

[0028] (5) System calibration and maintenance: All sensors are calibrated in the laboratory before installation to ensure the accuracy of volumetric moisture content and conductivity measurements; the status of wireless transmission and data acquisition devices is checked regularly to ensure long-term stable operation; rainproof and animal-proof protective covers are installed to ensure sensor safety and measurement reliability.

[0029] S3. Sample Collection: Under different irrigation modes, a phased sampling system is established for the irrigation and non-irrigation periods to obtain low-frequency data, including spatiotemporal variation data of soil solution, groundwater, and soil profile samples. During the irrigation period, soil solution and groundwater samples are collected once every minute to hour to capture the rapid water-salt response process under short-term irrigation events. During the non-irrigation period, soil solution and groundwater samples are collected once every week to month to monitor water-salt migration under long-term equilibrium conditions. During the crop growing season, soil profile samples are collected once every month. The sampling depth interval is consistent with the soil water, temperature, and salt profile monitoring depth, which in this embodiment is 0, 15, 30, 45, 60, 90, 120, and 150 cm, until the groundwater aquifer (generally 120-150 cm) is reached. Soil solution was collected by pre-burying clay tubes at different depths (15, 30, 45, 60, 90, 120, 150 cm). During sampling, 30-50 mL of soil solution was extracted by hand pump or vacuum pump. Before each sampling, 5-10 mL of residual liquid from the previous sampling was discarded to ensure the representativeness of the sample. Groundwater samples were collected by monitoring well sampler or submersible pump.

[0030] S4. Sample testing: Analyze the pH, salinity, and content of major ions (major cations and anions) and dissolved carbonate content of soil solution and groundwater samples; analyze the pH, salinity, basic cation content and different forms of carbonate content of soil samples.

[0031] Soil solution and groundwater samples were filtered through a 0.45 μm filter membrane, and pH, conductivity, and major cations (Na+, Na ... + K + Ca 2+ Mg 2+ ) and anions (HCO3) - CO3 2- Cl - SO4 2- The total inorganic carbon (TIC) content was determined using a total inorganic carbon analyzer (TIC Analyzer). Soil sample measurements included: total salinity, pH, electrical conductivity, cation exchange capacity (CEC), and basic ion composition (Na₂O₃). + K + Ca 2+ Mg 2+ The results were used to construct water-salt parameter sequences and carbonate index sequences at different depths and time scales.

[0032] Using the analysis results of manually collected profile soil samples as a reference value, synchronous sampling is performed at the corresponding depth of the sensor. The high-frequency data obtained by the water temperature and salinity monitoring system is compared with the reference value to identify systematic deviations in the high-frequency data. Corresponding correction relationships are established for soil moisture content, salinity and temperature parameters respectively. The high-frequency data is calibrated and corrected to obtain calibrated high-frequency data, thereby improving the reliability and accuracy of continuous monitoring data.

[0033] The calibrated high-frequency data and the low-frequency carbonate index data are sequentially subjected to time scale unification, numerical standardization and time series matching to obtain synchronized water and salt parameter sequences and carbonate index sequences.

[0034] Specifically, calibration parameters are generated based on the correspondence, and the data output by the monitoring system is modified by means of proportional correction, offset correction, or a combination thereof. The correction relationship is a linear or nonlinear function relationship, and the calibration parameters include proportional correction coefficients and offset correction amounts, which are used to modify the data output by the monitoring system by means of the parameters.

[0035] When the modified form is a linear relationship, its modified form can be expressed as: X c = aXm + b in, X m The raw data collected by the monitoring system X c For the calibrated data, a This is the proportional correction factor. b This is the offset correction amount. The above method is used to calibrate and correct the moisture and salinity monitoring data.

[0036] In other implementations, the calibration correction may also be achieved using nonlinear regression, piecewise functions, or machine learning regression models.

[0037] S5. Evaporation Force Calculation and Data Standardization: Using continuous meteorological data from the experimental station's weather station, calculate the atmospheric evaporation force during the experiment. (mm h) –1 The FAO Penman–Monteith formula is used as an example method for calculating evaporative force, to illustrate the changes. Where Δ represents the slope of the curve relating saturated vapor pressure and air temperature (kPa °C). -1 ); : Net radiation input to the canopy (MJ m -2 d -1 ); G Soil heat flux (MJ m -2 d -1 ); c Hygrometer constant (kPa °C) -1 T: Daily average temperature (°C) at a height of 2 m above the ground; Wind speed at 2 m above ground (m / s) -1 ); : The difference between saturated and actual water vapor pressure (kPa); C n (K mms 3 Mg -1 h -1 )and C d (sm -1 The constant 0.408 is determined based on the crop type and the calculation time step. When the time step is in hours, it is 37 and 0.34; when the time step is in days, it is 900 and 0.34. The unit of the constant 0.408 is meters. 2 mm MJ -1 .

[0038] In other implementations, simplified empirical formulas or other standard evaporation force models may be used for calculation.

[0039] The obtained evaporation force was time-registered with groundwater depth, water content, and salinity (time step 10 min) and uniformly converted into daily average values.

[0040] S6. Construction of the Time-Driven Relationship Model. The goal of this section is to couple the high-frequency (sensor) water-salt time series with low-frequency (laboratory) carbonate speciation indicators, and to determine "which water-salt factor lags at what time" by establishing a time-driven relationship model. t The study aims to "drive the dissolution, migration, or deposition of carbonates" and translate this result into actionable regulatory rules. After completing preliminary continuous water and salt monitoring and carbonate speciation analysis, data on soil moisture content, conductivity, temperature, and groundwater depth continuously collected by sensors are compared with data on total carbonate content and calcium in periodically collected soil samples. 2+ HCO3 - CO3 2- The content is matched over time. The goal of this section is to couple high-frequency (sensor) water-salt time series with low-frequency (laboratory) carbonate speciation indicators to determine "which water-salt factor lags at what time." t This study investigates the driving forces behind carbonate dissolution, migration, and deposition, and translates this knowledge into actionable regulatory rules. To ensure repeatability, detailed operational procedures, formulas, parameter selection, verification thresholds, and output formats are provided below.

[0041] S6.1 Data organization and synchronization.

[0042] (1) Time reference: The minimum time step is 10 min from the sensor, and all high-frequency data (including moisture content) are used. i ( z, t ), soil electrical conductivity EC ( z, t → Salt concentration C ( z, t ),temperature T ( z, t ), groundwater depth H ( t Meteorological variables and calculated evaporative force E 0 ( t Saved according to the original timestamp. z, t These represent soil profile depth and time, respectively. Periodic samples (solution DIC, profile CaCO3-C) are entered into the database according to the sampling time.

[0043] (2) Registration / interpolation: Map the low-frequency carbonate samples to the daily scale using linear interpolation (or nearest neighbor); when summarizing the 10-minute data into daily values ​​(or hourly values, depending on the modeling resolution), calculate the daily average, daily maximum, daily minimum and cumulative amount (e.g., the cumulative evaporation force of the day); if the modeling is done on a daily scale, replace the 10-minute data with the daily average.

[0044] (3) Variable transformation: Perform natural logarithmic transformation on concentration variables (salt concentration, DIC, CaCO3-C). ln ( x+c )( c (Using small constants to avoid zero), z-score standardization is performed on all explanatory variables to compare coefficient sizes.

[0045] (4) Missing data handling: Short missing data (<72 h) is filled with linear interpolation; long missing data is recorded and treated as missing indicator variables in the model and is not used for coefficient estimation.

[0046] S6.2 Candidate Driver Set and Lag Detection.

[0047] (1) Set of candidate driving variables: salt flux J ( z, t (See S6.4 for calculation), soil moisture content i ( z, t ), soil salt concentration C ( z, t ), Evaporation power E 0 ( t ), groundwater depth H ( t Precipitation / irrigation events I irrigation ( t (0 / 1 or irrigation volume), temperature T ( t ). Variables are retained for different depths and depth aggregation can be performed (upper layer 0–30 cm, root region 30–60 cm, lower layer 60–120 cm).

[0048] (2) Lag determination ( t For each candidate driving variable and the target time series ΔC c(t) (The rate of change in carbonate form or content is defined as the difference between samples and normalized.) Calculate the cross-correlation function (CCF) and record the lag corresponding to significant peaks. t For each pair of variables, Granger causality tests were performed to confirm predictive power (lag order 1-30 days or 1-72 hours, depending on sampling frequency). The results were compiled into a variable-lag pair list, for example... J root (t -6h) E 0 ( t -48h) H ( t -14d).

[0049] S6.3 Physical Quantification: Salt Flux and Seepage Calculation (Generation) J ).

[0050] (1) Definition of salt flux (depth) z (location) in The effective salt diffusion coefficient is taken in the form: parameter , α , β Initial values ​​are given in the laboratory or literature and calibrated using nonlinear least squares on field data; spatial gradient Estimation is performed using the difference between adjacent monitoring depth centers.

[0051] seepage velocity Estimated by Richards equations or empirical seepage approximation: If conditions are limited, empirical approximations can be used for calculation: in The average hydraulic conductivity of the layer, Dh For head difference. Hydraulic conductivity function. The water conservation characteristics can be determined using the van Genuchten or Mualem models and calibrated with on-site hydraulic parameters.

[0052] (2) The above calculations yielded The time series was used to obtain the total salt flux of the profile by depth or profile area. J total(t) This is used for subsequent regression modeling.

[0053] The parameters are shown in Table 2.

[0054] Table 2 To verify the feasibility of the model calculation, a typical profile of the test area (0–120 cm) was used as an example to demonstrate the calculation process, rather than actual experimental data.

[0055] (1) Setting: D 0 = 1.0×10-7 m 2 s -1 ; i = 0.25 (dimensionless); T = 20 °C; α =2.1; β =0.04.

[0056] Step-by-step calculation: 1) Calculate the exponential term: a·i - β·T = 2.1×0.25 − 0.04×20 = 0.525 − 0.8 = −0.275.

[0057] 2) Take the index: exp (−0.275) ≈ 0.759.

[0058] 3) Calculation D s (unit: m) 2 s -1 ): D s = D0 × 0.759 = 1.0 × 10 -7 × 0.759 = 7.59 × 10 -8 m 2 s -1 .

[0059] 4) Convert to m 2 d -1 (multiplied by 86400 sd) -1 ): D s ( d ) = 7.59 × 10 -8 × 86400 ≈ 0.00656m 2 d -1 .

[0060] (2) Concentration gradient: Assume that the solution concentrations measured at the 0.15 m and 0.30 m layers are respectively C (0.15) = 20 kgm -3 (equal to 20 g L) -1 ), C (0.30) = 15 kg m -3 Then the spatial gradient is: −50.15 = −33.333 kg m −4 (3) Convection term: Assuming seepage velocity q = 0.01 md -1(Typical slow infiltration), and with C (0.15) = 20 kgm -3 express.

[0061] (4) Calculate the diffusion term: =−0.00656×(−33.333)=0.2187 kg m −2 d −1 (5) Calculate the convection term: q⋅C =0.01×20=0.2 kg m -2 d −1 (6) Daily-scale composite flux: J =0.2187 + 0.2 = 0.4187 kg m −2 d −1 (approximately 418.7 gm) -2 d -1 ).

[0062] This example demonstrates that meaningful salt flux values ​​can be obtained using the parameters and methods described above. Further details will follow... J ( t )and ΔCc ( t Coupled modeling is performed.

[0063] The example values ​​are for demonstration purposes only; actual parameters should be corrected based on the soil properties and calibration experiments.

[0064] S6.4 Baseline time-lag regression model (multivariate lag regression).

[0065] Time series modeling and relationship identification: The following baseline regression model is established. In this embodiment, a multiple linear regression form is used, but it can also be extended to a nonlinear model or other statistical learning models: The carbonate change rate Δ is established using multiple regression or path analysis. C c Time-driven relationship with multidimensional water-salt parameters: in, In order to be in t The rate of change of soil carbonate content per unit time is used to characterize the intensity of carbonate migration or transformation. These are the lagged and standardized driving variables; t This indicates the monitoring time or monitoring moment, corresponding to the sampling time point in the continuous field observation sequence; To indicate the first iThe time lag of the influence of each water-salt transport driver variable on carbonate changes is used to characterize the time-delay response of the water-salt transport process to carbonate migration. For the first m Indicator variables for the irrigation event; This is the constant term in the regression model, used to characterize the baseline level of the rate of change of carbonates when each driving variable is zero or at the baseline state; The first regression coefficient, The second regression coefficient The random error term characterizes the impact of environmental disturbances, measurement errors, or other random factors not explained by the model on carbonate variations; coefficients , m , The solution is obtained using weighted least squares estimation (WLS) or regression with regularization (LASSO / Ridge) to control for multicollinearity. The model is evaluated using AIC, BIC, and adjusted values ​​within the sample. R ², RMSE comparison to select the optimal set of variables and lag combination.

[0066] S6.5 Direct / Indirect Impacts and Mechanism Identification (Path Analysis).

[0067] (1) Constructing SEM: Organize the observed variables into a path structure, example of the assumed relationship: E 0 → i → q → J → Δ C c Irrigation → q → J ; H → i & H → J .

[0068] (2) Estimate the path coefficients using lavaan (R) or semiopy (Python) and test the model fit (CFI>0.90, RMSEA<0.08, SRMR<0.08). SEM can distinguish the different driving pairs Δ. C c The direct and indirect contributions were analyzed to obtain the significance test and confidence interval of the path coefficient.

[0069] S6.6 Time-varying coefficients and critical period identification (sliding window).

[0070] (1) Sliding window regression: Select the window length (e.g., 30 days or 4320 10-min time steps) with a step size of 7 days; repeat the regression estimation in S6.4 within each window to obtain the time series coefficients. ai ( t w ).

[0071] (2) Critical Period Judgment Rule: When the absolute value of a factor coefficient is in a continuous period of time... k window( k ≥ 2) Significantly higher than the long-term median (e.g., higher than the historical 75th percentile and p If the value of a factor is less than 0.05 and its contribution to the model is greater than 30%, then the time interval is determined to be the "driving critical period" of that factor. Save the start and end dates of the critical period and the dominant factor.

[0072] S6.7 Model Validation and Robustness Testing.

[0073] (1) Validation scheme: The generalization ability of the model is evaluated by time cross-validation (leave-one-year-out) and spatial cross-validation (leave-one-plot-out).

[0074] (2) Robustness test: Bootstrap resampling of model parameters ( N =1000) to obtain the parameter confidence interval; calculate VIF Indicators for detecting collinearity ( VIF (When the value is greater than 10, dimensionality reduction or regularization is required).

[0075] (3) Residual diagnosis: If the residuals are autocorrelation, use AR ( p If heteroscedasticity exists, use weighted regression or GLS for correction; if heteroscedasticity exists, use weighted regression.

[0076] (4) Uncertainty assessment: Monte Carlo simulation or Bayesian method (PyMC3 / Stan) can be used to assess the prediction interval (95% CI). S6.8 Output Results and Documentation.

[0077] (1) Output file: Coefficient table (including lag) t Standard error p Data such as values, path coefficient tables, critical period lists, driving contribution rate charts, model fitting and residual diagnosis charts are exported in CSV / Excel and PDF report formats to quantitatively characterize the time response of water-salt transport to carbonate migration.

[0078] (2) Decision output: Based on the identification of critical periods, a "critical period calendar" is generated, and in conjunction with sensitivity analysis, an "irrigation response curve" is obtained (i.e., for every increase in critical period). X mm irrigation for Δ C c (Expected impact).

[0079] (3) Software / Tool Recommendations: Data processing and modeling can be done using Python (pandas, numpy, statsmodels, scikit-learn, semopy, matplotlib) or R Languages ​​(tidyverse, lm / glmnet, lavaan, ggplot2).

[0080] From Modeling to Irrigation Regulation Decision-Making (Examples of Implementation Rules) (1) Threshold and triggering rules (example): based on historical 4-year daily Δ C c The distribution is set with a high-risk threshold at the historical 90th percentile. If the model predicts Δ over a 7-day rolling period... C c ( t If +7d) exceeds the threshold, "mitigation measures" are triggered.

[0081] (2) Examples of mitigation measures: If the risk of surface carbonate migration is high (model indication) E 0 If the water level is shallow and the main factor is low, it is recommended to increase or advance the autumn irrigation (e.g., add 100–150 mm, the specific amount is given by the sensitivity curve); if the model indicates that a large amount of water causes a strong downward shift (leading to Ca... 2+ / HCO 3- If there is a loss, it is recommended to irrigate in multiple sessions or reduce the amount of water irrigated at one time to reduce the loss.

[0082] (3) Feedback loop: The observation results after implementation are incorporated into the training data of the next cycle to continuously update the model parameters (annual update). The monitoring and modeling process of this invention has been continuously implemented for many years in the same experimental area, covering different crop growing seasons and climatic conditions, ensuring that the model can reflect the dynamic changes of water, salt, and carbonate on both seasonal and interannual scales. The constructed method can not only be used to study the carbonate migration mechanism in saline soil areas, but also is applicable to the optimization of irrigation regimes and carbonate accumulation risk assessment in different irrigation areas and soil types; salinity regulation and groundwater utilization management; and the simulation of soil carbon cycle processes and the development of carbon sequestration regulation technologies. This method has advantages such as automation, long-term continuous monitoring, quantitative modeling, and decision-making applications, overcoming the limitations of previous studies with short experimental cycles, isolated parameters, and a lack of dynamic response. The above model construction, parameter selection, and threshold setting are all exemplary implementation methods and do not constitute a limitation on the scope of protection of this invention.

[0083] Example 2 Other conditions were the same as in Example 1 (including: S2 profile monitoring system deployment, S3 sample collection, S4 sample detection, etc.), except that the sensor preset time interval was set to 5 minutes, and the following key results were obtained: S6.3 Salt flux calculation: (1) Setting: D 0 = 1.0×10 -7 m 2 s -1 ; i = 0.25 (dimensionless); T = 20 ℃; α =2.1; β =0.04.

[0084] Step-by-step calculation: 1) Calculate the exponential term: a·i - β·T = 2.1×0.25 − 0.04×20 = 0.525 − 0.8 = −0.275.

[0085] 2) Take the index: exp (−0.275) ≈ 0.759.

[0086] 3) Calculation D s (unit: m) 2 s -1 ): D s = D0 × 0.759 = 1.0 × 10 -7 × 0.759 = 7.59 × 10 -8 m 2 s -1 .

[0087] 4) Convert to m 2 d -1 (multiplied by 86400 sd) -1 ): D s ( d ) = 7.59 × 10 -8 × 86400 ≈ 0.00656m 2 d -1 .

[0088] (2) Concentration gradient: Assume that the solution concentrations measured at the 0.15 m and 0.30 m layers are respectively C (0.15) = 20 kgm -3 (equal to 20 g L) -1 ), C(0.30) = 15 kg m -3 Then the spatial gradient is: −50.15 = −33.333 kg m −4 (3) Convection term: Assuming seepage velocity q = 0.01 md -1 (Typical slow infiltration), and with C (0.15) = 20 kgm -3 express.

[0089] (4) Calculate the diffusion term: =−0.00656×(−33.333)=0.2187 kg m −2 d −1 (5) Calculate the convection term: q⋅C =0.01×20=0.2 kg m -2 d −1 (6) Daily-scale composite flux: J =0.2187 + 0.2 = 0.4187 kg m −2 d −1 (approximately 418.7 gm) -2 d -1 ).

[0090] (7) 5-minute time scale conversion: Δ t =5 / 1440≈0.003472 d J 5min = J ×Δ t ≈0.00145 kg / m -2 5min -1 ≈1.45 g / m -2 5min -1 S6.6 Sliding Window Regression The sliding window length (e.g., 30 days or 8640 5-min time steps) is used as the baseline, with a step size of 7 days. The regression calculation within the window uses the actual number of time steps, adjusted according to different sampling intervals: (1) Sliding window regression: Select the window length, with a step size of 7 days; repeat the regression estimation in S6.4 within each window to obtain the time series coefficients. a i ( t w ).

[0091] Example 3 In this embodiment, all other conditions are the same as in Embodiment 1, except that the sensor preset time interval is set to 30 minutes.

[0092] Key results example (0-120 cm profile, schematic values): 1. The effective diffusion coefficient, concentration gradient, convection term, diffusion term, and daily total salt flux are all the same as in Example 1; 2. Daily-scale composite flux: J =0.2187 + 0.2 = 0.4187 kg m −2 d −1 (approximately 418.7 gm) -2 d -1 ).

[0093] 3. 30-minute time scale conversion: Δ t =30 / 1440=0.020833d J 30min = J ×Δ t ≈0.4187×0.020833≈0.00872 kg m -2 30min -1 ≈8.72 gm -2 30min -1 S6.6 Sliding Window Regression The sliding window length (e.g., 30 days or 1440 30-min time steps) is used as the baseline, with a step size of 7 days. The regression calculation within the window uses the actual number of time steps, adjusted according to different sampling intervals: (1) Sliding window regression: Select the window length, with a step size of 7 days; repeat the regression estimation in S6.4 within each window to obtain the time series coefficients. a i ( t w ).

[0094] This embodiment shows that low-frequency sampling (30 minutes) can still obtain reliable daily-scale cumulative salt flux values, but it has a lower temporal resolution for capturing rapid irrigation responses. Other steps (evaporation force calculation, low-frequency carbonate calibration, regression modeling, etc.) are the same as in Embodiment 1.

[0095] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for monitoring and regulating the temporal relationship between water-salt transport and carbonate migration at the field scale, characterized in that, Includes the following steps: Field trials were conducted based on different irrigation regimes. A profile water, temperature and salinity monitoring system was deployed in each experimental plot to automatically and continuously collect high-frequency data based on preset time intervals and spatial depths. The high-frequency data included soil moisture, salinity and temperature data. Low-frequency data were manually collected in stages based on different irrigation modes. The low-frequency data included soil solution, groundwater samples, and profile soil samples. The low-frequency data is processed using analytical detection techniques to obtain low-frequency carbonate index data; Using the processing results of the soil samples from the profile as a reference value, a deviation function is established based on the low-frequency data and the high-frequency data, and regression correction is performed on the high-frequency parameters to obtain calibrated high-frequency data; The calibrated high-frequency data and the low-frequency carbonate index data are sequentially subjected to time scale unification, numerical standardization and time series matching to obtain synchronized water and salt parameter sequences and carbonate index sequences. The synchronized water-salt parameter sequence and carbonate index sequence were processed using the time-delay regression modeling method to obtain a predictive model describing the time relationship between water-salt transport and carbonate migration. Based on the prediction model, key water and salt driving factors and their corresponding lag times are identified, and irrigation regulation decisions are generated.

2. The method for monitoring and regulating the temporal relationship between water-salt transport and carbonate migration at the field scale according to claim 1, characterized in that, The preset time interval is any time interval within the range of 5-30 minutes; The preset depth is the depth of multiple soil layers covering the topsoil and root zone, and the multiple soil layers include several layers in the range of 0-120 cm.

3. The method for monitoring and controlling the temporal relationship between water-salt transport and carbonate migration at the field scale according to claim 2, characterized in that, The process of manually collecting soil solution, groundwater, and profile soil samples in stages based on different irrigation patterns includes: During the irrigation period, soil solution and groundwater samples were collected at a sampling frequency ranging from minutes to hours. During non-irrigation periods, soil solution and groundwater samples were collected at a sampling frequency of weekly to monthly. During the crop growing season, profile soil samples are collected at several soil depths on a monthly timescale.

4. The method for monitoring and controlling the temporal relationship between water-salt transport and carbonate migration at the field scale according to claim 1, characterized in that, The process of sequentially performing time-scale unification, numerical standardization, and time-series matching on the calibrated high-frequency data and the low-frequency carbonate index data to obtain synchronized water-salt parameter sequences and carbonate index sequences includes: The calibrated high-frequency data were processed using a time aggregation method to obtain a water-salt parameter sequence with a uniform time scale; The low-frequency carbonate index data is processed using a smoothing method to obtain a carbonate index sequence that is consistent with the time scale of the water-salt parameter sequence.

5. The method for monitoring and controlling the temporal relationship between water-salt transport and carbonate migration at the field scale according to claim 1, characterized in that, The process of using time-delay regression modeling to process the synchronized water-salt parameter sequence and carbonate index sequence to obtain a predictive model describing the time relationship between water-salt transport and carbonate migration includes: Based on time series data of soil moisture content, salt concentration and temperature, the salt diffusion-convection equation was used to calculate the salt flux time series. The salt flux time series, soil moisture content series, evaporation force series, groundwater depth series and temperature series are used as candidate driving variables. Based on the candidate driving variables and the carbonate index change series, time lag correlation analysis is performed to determine the potential lag time of each candidate driving variable's influence on carbonate migration. Based on the potential lag time, a predictive model is constructed to characterize the time relationship between water-salt transport and carbonate migration.

6. The method for monitoring and controlling the temporal relationship between water-salt transport and carbonate migration at the field scale according to claim 5, characterized in that, The calculation expression for the evaporation force sequence is: ； In the formula, Δ is the slope of the curve relating saturated water vapor pressure and air temperature, with units of kPa °C. -1 ; The net radiation input to the canopy is expressed in MJ / m². -2 d -1 ; G Soil heat flux, in MJ / m³ -2 d -1 ; γ This is the hygrometer constant, in kPa °C. -1 ; T The average daily temperature at a height of 2 m above the ground is expressed in °C. Wind speed at a height of 2 m above the ground, in milliseconds (ms). -1 ; This represents the pressure difference between saturated and actual water vapor, expressed in kPa. C n and C d The first and second constants are determined by the reference crop type and the calculation time step, respectively. C n The unit is K mm s 3 Mg -1 h -1 , C d The unit is sm -1 ; Evaporation force, unit: mmh –1 .

7. The method for monitoring and controlling the temporal relationship between water-salt transport and carbonate migration at the field scale according to claim 5, characterized in that, The process of constructing the prediction model includes: constructing a multiple lag regression model based on the carbonate change rate and water-salt transport driving factors, wherein the multiple lag regression model is as follows: ； In the formula, In order to be in t The rate of change of soil carbonate content per unit time at any given moment; These are the lagged and standardized driving variables; t Indicates the monitoring time or monitoring moment; For the first i The time lag of the effect of each water-salt transport driver variable on carbonate changes; For the first m Indicator variables for the irrigation event; This is the constant term in the regression model; The first regression coefficient, The second regression coefficient; This is the random error term.

8. The method for monitoring and regulating the temporal relationship between water-salt transport and carbonate migration at the field scale according to claim 1, characterized in that, The process of identifying key water-salt driving factors and their corresponding lag times based on the prediction model includes: Based on the regression coefficients of the prediction model, time series analysis is performed using a sliding window of a preset duration; When the regression coefficient of the key water-salt driving factor is significantly higher than the historical benchmark level within a certain number of consecutive time windows, the corresponding time period is determined to be the critical driving period of the current key water-salt driving factor.

9. The method for monitoring and regulating the temporal relationship between water-salt transport and carbonate migration at the field scale according to claim 1, characterized in that, The process of generating irrigation regulation decisions includes: predicting the risk of carbonate migration in a future set time period based on the prediction model; when the predicted value exceeds the preset risk threshold, triggering an irrigation system adjustment instruction, wherein the adjustment includes adjusting at least one of the following: irrigation time, single irrigation volume, or number of irrigations.

Images