Method and system for predicting response of water-saving and salt discharge and evaluating risk of salt accumulation in irrigation district based on strategy classification

Abstract

The application discloses a kind of based on strategy typing's irrigation area water-saving salt discharge response prediction and salt accumulation risk assessment method and system, including based on irrigation area basic data construction and rate verification distributed hydrology model, based on the correction coefficient of distributed hydrology model simulation result calculation, and according to the correction coefficient each water-saving scenario predicted monthly drainage capacity is obtained;Mineralization model is constructed and the annual average salt discharge of each water-saving scenario prediction is calculated;Salt discharge ratio and salt retention rate are calculated to carry out double-index joint risk grade assessment;Sensitivity equation of each water-saving strategy is fitted to invert water diversion reduction safety red line, and monthly drainage capacity and salt discharge regression relationship are established to determine monthly high-risk drainage capacity threshold value.The application discloses the different strategy under the differentiation sensitivity of salt discharge capacity to water diversion reduction, realizes from salt accumulation risk assessment to the quantitative conversion of operable water-saving safety boundary, provides technical support for water-saving policy making and monthly dynamic early warning.

Description

Technical Field

[0001] This invention belongs to the field of agricultural water and soil engineering and irrigation area water resources management technology, and specifically relates to a method and system for predicting irrigation area water-saving and salt discharge response and assessing salt accumulation risk based on strategy classification. Background Technology

[0002] Arid and semi-arid irrigation areas widely exhibit unique water-salt cycling mechanisms involving both salt inflow and outflow. Large-scale water-saving measures can drastically disrupt solute transport and discharge fluxes, breaking the long-established regional water-salt balance. Against this backdrop, predicting the dynamic water-salt response of irrigation areas under different water control strategies and scientifically assessing the systemic salt accumulation risk caused by leaching dynamic decline have become critical technical issues that urgently need to be addressed in modern irrigation areas to formulate scientific water-saving policies, implement precise water rights allocation, and ensure the quality and safety of arable land.

[0003] Currently, existing technical approaches for water and salinity prediction and risk assessment in irrigation areas under the background of water conservation mainly fall into two categories: The first category is mechanism-driven distributed hydrological modeling, which constructs a distributed hydrological model of solute transport processes by dividing the irrigation area into hydrological response units, simulating the dynamics of regional water volume and soil salinity under uniform water restriction conditions. The second category is data-driven empirical statistical models, which construct a macroscopic mapping relationship between water diversion, drainage, and salinity based on long-term historical monitoring section data of the irrigation area, using statistical regression methods to assess the long-term salinity balance and evolution trend of the irrigation area.

[0004] However, both of these technologies have significant limitations in practical applications: For distributed hydrological models, the complex control process of artificial irrigation and drainage systems in irrigation districts, coupled with the strong spatial heterogeneity of water-salt exchange processes in soil and aquifers, leads to difficulties in model parameter calibration and high computational costs. This results in systematic biases in predicting indicators such as the absolute amount of solute at irrigation district outlets, making it difficult to support rapid simulation and batch evaluation across multiple strategies and scenarios. For empirical statistical models, their predictive ability relies on the statistical regularities of historical monitoring data, lacking a clear physical mechanism. They cannot reasonably extrapolate nonlinearly for extreme water-saving scenarios or unconventional water control strategies that have not occurred historically, resulting in poor applicability for scenario expansion. Furthermore, existing technologies generally lack strategy classification research, making it impossible to quantify the differentiated response patterns of drainage and salt removal under different irrigation time windows, and also failing to characterize the spatiotemporal lag effects of water-salt transport processes. Simultaneously, they fail to reveal the game-theoretic conflict mechanism between the decline in external salt intake and the deterioration of drainage leaching dynamics leading to deterioration in salt removal efficiency during water diversion reduction.

[0005] In summary, existing technologies cannot simultaneously achieve prediction accuracy, extrapolation rationality, and strategy adaptability. Summary of the Invention

[0006] Purpose of the invention: One objective of this invention is to provide a method for predicting water-saving and salt-discharge response and assessing salt accumulation risk in irrigation districts based on strategy classification. This method can comprehensively consider the differentiated responses to drainage and salt discharge under different irrigation time windows in the irrigation district, and quantitatively provide the safety red line for water diversion reduction and the monthly high-risk drainage threshold for each strategy type.

[0007] Another objective of this invention is to provide a system for predicting water-saving and salt-discharge response and assessing salt accumulation risk in irrigation areas based on strategy classification.

[0008] Technical solution: The method for predicting water-saving and salt-discharge response and assessing salt accumulation risk in irrigation areas based on strategy classification as described in this invention includes:

[0009] A distributed hydrological model was constructed and calibrated using basic data from the irrigation district. Based on the calibrated distributed hydrological model, the simulated monthly drainage volume for each water-saving scenario and the baseline scenario was obtained. Correction coefficients were calculated based on the simulated monthly drainage volume for each water-saving scenario and the baseline scenario, and the predicted monthly drainage volume for each water-saving scenario was obtained based on the correction coefficients. Each water-saving scenario was generated by combining water-saving strategies and water-saving intensities, and each type of water-saving strategy had multiple levels of water-saving intensities.

[0010] Using measured monthly drainage volume, current and previous month water diversion volume, previous accumulated water deficit, and monthly fixed effects as independent variables, and drainage mineralization as the dependent variable, a mineralization model is constructed. The mineralization model is validated using basic irrigation district data. The predicted monthly drainage volume for each water-saving scenario is input into the validated mineralization model to calculate the predicted monthly drainage mineralization. Based on the predicted monthly drainage mineralization and the predicted monthly drainage volume, the predicted annual average salt discharge for each water-saving scenario is calculated.

[0011] Based on the predicted annual average salt discharge under each water-saving scenario, the salt discharge ratio and salt retention rate under each water-saving scenario are calculated. The salt discharge ratio is used to assess the salt accumulation risk level under each water-saving scenario, and the salt retention rate is used to assess the degree of salt accumulation.

[0012] Sensitivity equations for salt discharge ratio and water diversion reduction ratio are fitted for each water-saving strategy. Based on these sensitivity equations, the safety red line for water diversion reduction is derived by using the critical threshold of salt discharge ratio corresponding to each salt accumulation risk level. Linear regression equations are fitted for the predicted monthly drainage volume and annual salt discharge ratio under each water-saving scenario. Based on these linear regression equations, the corresponding monthly high-risk drainage volume threshold is derived by using the salt discharge ratio equal to the high-risk lower limit threshold in the salt accumulation risk level. Finally, the safety red line for water diversion reduction and the monthly high-risk drainage threshold for each water-saving strategy are output.

[0013] Furthermore, the construction and calibration of the distributed hydrological model using basic irrigation district data includes:

[0014] The distributed hydrological model includes a runoff generation module, a runoff collection module, an evapotranspiration module, and an irrigation management module. Each module uses a Hydrological Response Unit (HRU) as the basic computational unit and the watershed system as the river confluence network. For each HRU, the evapotranspiration module calculates the actual evapotranspiration based on irrigation district baseline data. The irrigation management module determines the irrigation window and irrigation depth for that HRU based on the irrigation district baseline data and the input water diversion plan. The runoff generation module integrates the irrigation district baseline data and the actual evapotranspiration to perform water balance calculations and obtain the runoff volume. After being aggregated within sub-watersheds, the runoff volume is collected step-by-step along the watershed system to the irrigation district drainage outlet through the runoff collection module, forming a monthly drainage volume simulation result. A parameter sensitivity analysis algorithm is used to identify key parameters sensitive to the drainage volume simulation, and automatic calibration is performed using the measured monthly drainage volume as the calibration target, with the Nash efficiency coefficient as the calibration target. Coefficient of determination and percentage deviation The model has passed the validation criteria;

[0015] The correction factor is the ratio of the simulated monthly drainage volume of each water-saving scenario to that of the baseline scenario. Multiplying the correction factor by the measured monthly drainage volume of the baseline scenario yields the predicted monthly drainage volume for each water-saving scenario.

[0016] Furthermore, the water-saving strategies are divided into year-round water saving, growing season water saving, and autumn irrigation season water saving according to the irrigation window, and the water-saving intensity is the reduction ratio of the water diversion volume under each water-saving strategy relative to the water diversion volume under the baseline scenario.

[0017] Furthermore, the mineralization model is a weighted sum of the natural logarithm of monthly drainage, the natural logarithm of the current month's water diversion, the natural logarithm of the previous month's water diversion, the previous accumulated water deficit, and the monthly fixed effect. The previous accumulated water deficit is calculated based on the previous potential evapotranspiration, effective precipitation, and effective irrigation water.

[0018] Based on the continuous monthly observation sequence of drainage mineralization, the mineralization model was validated using model evaluation indices, including the coefficient of determination. Nash efficiency coefficient Klin-Gupta efficiency coefficient and percentage deviation The simulation accuracy of drainage salinity was evaluated in logarithmic space and the simulation accuracy of coupled salt discharge was evaluated in the original physical space, with the validation period determination coefficient as the basis for the evaluation. And Nash efficiency coefficient As the criterion for judgment, and As an auxiliary reference indicator, the measured monthly drainage mineralization continuous observation sequence is divided into a training period and a validation period. The training period is used to estimate the parameters of the regression coefficients of the mineralization model, and the validation period is used to independently evaluate the generalization ability of the mineralization model.

[0019] After verification, the predicted monthly drainage volume for each water-saving scenario is input into the verified mineralization model, and the predicted drainage mineralization is calculated monthly. The predicted monthly drainage volume is multiplied by the predicted drainage mineralization for the corresponding month to obtain the predicted monthly coupled salt discharge sequence for each scenario. The predicted monthly coupled salt discharge for each scenario is accumulated and averaged to obtain the predicted annual average salt discharge for each scenario.

[0020] Furthermore, based on the predicted annual average salt discharge under each water-saving scenario, the salt discharge ratio and salt retention rate for each water-saving scenario are calculated; including: the salt discharge ratio To predict the percentage of annual average salt discharge under the water-saving scenario relative to the baseline scenario, when This indicates that the salt excretion capacity is not impaired; This indicates a decline in salt removal capacity and a risk of salt retention; the salt accumulation risk level includes: basically balanced. Low risk High risk ;in, To achieve a basic balance between the lower threshold and the upper threshold for low risk, These are the lower limit threshold for low risk and the upper limit threshold for high risk. This is the lower limit threshold for high risk.

[0021] The amount of salt introduced by external irrigation under each water-saving scenario is calculated based on the annual external water diversion volume and the multi-year average water diversion salinity. The salt retention rate is equal to the difference between the amount of salt introduced by external irrigation and the predicted annual average salt discharge, divided by the amount of salt introduced by external irrigation. When the salt retention rate is greater than zero, it indicates that the soil is in a state of net salt accumulation. The higher the salt retention rate, the more serious the salt accumulation.

[0022] Furthermore, the process of fitting a sensitivity equation between the salt discharge ratio and the water diversion reduction ratio for each water-saving strategy, and then using the critical threshold of the salt discharge ratio corresponding to each salt accumulation risk level to invert the water diversion reduction safety line based on the sensitivity equation, includes:

[0023] Using the water diversion reduction ratio as the independent variable and the salt discharge ratio as the dependent variable for each water-saving scenario, a sensitivity equation for each water-saving strategy is fitted. The water diversion reduction ratio is the percentage reduction in water diversion volume relative to the baseline scenario, which is numerically equal to the water-saving intensity of each water-saving scenario. The sensitivity equation adopts a linear regression with an intercept of 100%. The regression slope reflects the sensitivity of the salt discharge capacity of the strategy to water diversion reduction. The larger the absolute value of the slope, the more sensitive it is.

[0024] Substitute the critical threshold of salt discharge ratio corresponding to each salt accumulation risk level into the sensitivity equation of each fitted water-saving strategy, and solve the corresponding water diversion reduction safety red line in reverse to obtain the water diversion reduction safety red line for each strategy type under the three salt accumulation risk levels of basic balance, low risk and high risk.

[0025] Using the predicted monthly drainage volume under the water-saving scenario as the independent variable and the annual salt discharge ratio as the dependent variable, a linear regression equation is fitted. Based on the linear regression equation and the empirical regression parameters for each month, the annual salt discharge ratio is set to equal the high-risk lower limit threshold, and the high-risk drainage volume threshold for that month is calculated in reverse.

[0026] Based on the water diversion reduction safety red line and the monthly high-risk drainage threshold, a two-layer decision support output is formed: strategy layer and monthly layer. The strategy layer summarizes the water diversion reduction safety red line of each water-saving strategy under each risk level using a decision matrix. In the decision matrix, rows correspond to water-saving strategies, columns correspond to salt accumulation risk levels, and the elements of the decision matrix are the water diversion reduction safety red line of each strategy under the corresponding salt accumulation risk level. The monthly layer outputs the monthly high-risk drainage threshold and the ranking of critical months for salt discharge safety.

[0027] Based on the same inventive concept, another aspect of the present invention provides a system for predicting water-saving and salt-discharge response and assessing salt accumulation risk in irrigation districts based on strategy classification, comprising:

[0028] The model building, calibration and verification, and prediction unit is used to build and calibrate a distributed hydrological model using basic irrigation district data; based on the calibrated and verified distributed hydrological model, it obtains the simulated monthly drainage volume of each water-saving scenario and the baseline scenario; it calculates correction coefficients based on the simulated monthly drainage volume of each water-saving scenario and the baseline scenario; and it obtains the predicted monthly drainage volume of each water-saving scenario based on the correction coefficients; each water-saving scenario is generated by combining water-saving strategies and water-saving intensities, and each type of water-saving strategy has multiple levels of water-saving intensities.

[0029] The salt discharge calculation unit is used to construct a mineralization model with measured monthly drainage volume, current and previous month water diversion volume, previous accumulated water deficit and monthly fixed effect as independent variables and drainage mineralization as dependent variable. The mineralization model is verified using irrigation district basic data. The predicted monthly drainage volume of each water-saving scenario is input into the verified mineralization model to calculate the predicted drainage mineralization. Based on the predicted drainage mineralization and the predicted monthly drainage volume, the predicted annual average salt discharge volume of each water-saving scenario is calculated.

[0030] The salt accumulation risk level assessment unit is used to predict the average annual salt discharge under each water-saving scenario, calculate the salt discharge ratio and salt retention rate under each water-saving scenario, assess the salt accumulation risk level under each water-saving scenario based on the salt discharge ratio, and assess the degree of salt accumulation based on the salt retention rate.

[0031] The early warning unit is used to fit a sensitivity equation between the salt discharge ratio and the water diversion reduction ratio for each water-saving strategy. Based on the sensitivity equation, the water diversion reduction safety red line is inverted using the critical threshold of the salt discharge ratio corresponding to each salt accumulation risk level. The unit also fits a linear regression equation between the predicted monthly drainage volume and the annual salt discharge ratio under each water-saving scenario. Based on the linear regression equation, the corresponding monthly high-risk drainage volume threshold is inverted using the salt discharge ratio equal to the high-risk lower limit threshold in the salt risk level. Finally, the water diversion reduction safety red line and the monthly high-risk drainage threshold for each water-saving strategy are output.

[0032] Based on the same inventive concept, another aspect of the present invention provides an electronic device, comprising:

[0033] Memory, used to store computer programs;

[0034] A processor is used to execute the computer program to implement the method for predicting water-saving and salt-discharge response and assessing salt accumulation risk in irrigation areas based on strategy classification.

[0035] Based on the same inventive concept, another aspect of the present invention provides a non-volatile storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the aforementioned method for predicting irrigation district water-saving and salt discharge response and assessing salt accumulation risk based on strategy classification.

[0036] Based on the same inventive concept, another aspect of the present invention provides a computer program product, comprising a computer program / instruction that, when executed by a processor, implements the aforementioned method for predicting irrigation district water-saving and salt discharge response and assessing salt accumulation risk based on strategy classification.

[0037] Beneficial effects: Compared with the prior art, the significant technical effects of the present invention are as follows: (1) By adopting irrigation time window strategy classification, salt discharge ratio (SDR) and salt retention rate (SRR) dual indicators for joint evaluation, it can quantitatively characterize the differential response sensitivity of irrigation area salt discharge capacity to water diversion reduction under three water control strategies: annual water saving, growing season water saving and autumn irrigation season water saving; through dual indicator synergistic evaluation, it can effectively avoid the evaluation bias caused by the single absolute salt accumulation indicator under the two-way effect of water saving and salt reduction, improve the accuracy and objectivity of risk assessment results, and provide a quantitative decision basis for upgrading irrigation area water saving policy from empirical total reduction to strategy classification and precise control; (2) By constructing a sensitivity regression equation of SDR and water diversion reduction ratio, the salt accumulation risk is transformed into a quantifiable water diversion reduction safety red line; by the regression analysis of monthly drainage volume and SDR, the high-risk drainage volume threshold of key control months is identified; at the same time, the ratio-driven method is used to realize the engineering coupling of hydrological model and statistical model, and a complete decision support technology chain is constructed. Compared to the fully coupled distributed model, this invention significantly reduces computational complexity and operating costs, making it more suitable for the routine business application needs of irrigation district management departments. Attached Figure Description

[0038] Figure 1 This is a flowchart of the method of the present invention;

[0039] Figure 2 This is a schematic diagram of the coupling of drainage volume and salt discharge volume in this invention;

[0040] Figure 3 These are the calibration and verification results of the distributed hydrological model of the present invention, wherein (a) is the result during the calibration period and (b) is the result during the verification period;

[0041] Figure 4 The present invention shows the changes in drainage volume and salt discharge volume under multiple scenarios, where (a) is the monthly drainage volume, (b) is the monthly salt discharge volume, (c) is the rate of change of drainage volume, mineralization and salt discharge volume under each scenario, and (d) is the curve of water diversion-salt discharge rate.

[0042] Figure 5 This invention presents a monthly SDR heatmap and a monthly salt accumulation change line, wherein (a) is a monthly SDR heatmap and (b) is a monthly salt accumulation change line.

[0043] Figure 6 This is a monthly distribution chart of high-risk drainage thresholds according to the present invention. Detailed Implementation

[0044] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0045] Example 1:

[0046] like Figure 1 As shown, the method for predicting water-saving and salt-discharge response and assessing salt accumulation risk in irrigation districts based on strategy classification includes: constructing and calibrating a distributed hydrological model using basic data of the irrigation district; obtaining the simulated monthly drainage volume of each water-saving scenario and the baseline scenario based on the calibrated and validated distributed hydrological model; calculating correction coefficients based on the simulated monthly drainage volume of each water-saving scenario and the baseline scenario; and obtaining the predicted monthly drainage volume of each water-saving scenario based on the correction coefficients; wherein each water-saving scenario is generated by combining water-saving strategies and water-saving intensities, and each type of water-saving strategy has multiple levels of water-saving intensities.

[0047] Using measured monthly drainage volume, current and previous month water diversion volume, previous accumulated water deficit, and monthly fixed effects as independent variables, and drainage mineralization as the dependent variable, a mineralization model is constructed. The mineralization model is validated using basic irrigation district data. The predicted monthly drainage volume for each water-saving scenario is input into the validated mineralization model to calculate the predicted monthly drainage mineralization. Based on the predicted monthly drainage mineralization and the predicted monthly drainage volume, the predicted annual average salt discharge for each water-saving scenario is calculated.

[0048] Based on the predicted annual average salt discharge under each water-saving scenario, the salt discharge ratio and salt retention rate under each water-saving scenario are calculated. The salt discharge ratio is used to assess the salt accumulation risk level under each water-saving scenario, and the salt retention rate is used to assess the degree of salt accumulation.

[0049] Sensitivity equations for the ratio of salt discharge to water diversion reduction are fitted for each water-saving strategy. Based on these sensitivity equations, the safety threshold for water diversion reduction is derived using the critical threshold of the salt discharge ratio corresponding to each salt accumulation risk level. Linear regression equations are fitted for the predicted monthly drainage volume and the annual salt discharge ratio under each water-saving scenario. Based on these linear regression equations, the corresponding monthly high-risk drainage volume threshold is derived by using the salt discharge ratio equal to the high-risk lower limit threshold of the salt accumulation risk level. Finally, the safety threshold for water diversion reduction and the monthly high-risk drainage threshold for each water-saving strategy are output. The specific steps include:

[0050] S1. Collect and organize basic data of the irrigation area: including spatial data of the irrigation area, meteorological driving data, hydrological monitoring data and irrigation management data; among which, spatial data includes digital elevation model (DEM) data, land use data, soil type data and irrigation and drainage canal distribution data; meteorological driving data includes daily precipitation, temperature, wind speed, solar radiation and relative humidity; hydrological monitoring data includes monthly drainage volume, water diversion volume, salt discharge and drainage mineralization of drainage outlets; irrigation management data includes crop planting date, crop planting area, irrigation window and irrigation water depth.

[0051] S2. A distributed hydrological model is constructed using the spatial data, meteorological driving data, and irrigation management data obtained in S1. The distributed hydrological model is then calibrated and verified using hydrological monitoring data. The calibrated and verified distributed hydrological model is used to simulate the monthly drainage volume of the baseline scenario and each water-saving scenario. A ratio-driven method is employed, using the measured monthly drainage volume of the baseline scenario as the benchmark point. The ratio of the simulated monthly drainage volume of each water-saving scenario to the baseline scenario's monthly drainage volume is used to correct the system error of the corresponding water-saving scenario, thus obtaining the predicted monthly drainage volume for each water-saving scenario. In this embodiment, the historical actual water diversion status is used as the baseline scenario. The water-saving scenario is generated by combining water-saving strategies and water-saving intensity. Water-saving strategies are divided into three types according to the irrigation time window: S-type (year-round water saving), G-type (growing season water saving), and A-type (autumn irrigation season water saving). The water-saving intensity is the reduction ratio relative to the baseline scenario's water diversion volume. Each type of water-saving strategy has multiple levels of water-saving intensity. All types of water-saving strategies and multiple levels of water-saving intensity constitute a strategy × intensity scenario matrix. Specifically, the following steps are included:

[0052] S2.1 The distributed hydrological model includes a runoff generation module, a runoff collection module, an evapotranspiration module, and an irrigation management module. Each of these modules uses a Hydrological Response Unit (HRU) as its basic computational unit and the watershed system as its river confluence network. For each HRU, the evapotranspiration module calculates the actual evapotranspiration based on the meteorological driving data obtained in S1 and the land use data of that HRU. The irrigation management module determines the irrigation window and irrigation depth for that HRU based on irrigation management data and the input water diversion scheme. The runoff generation module integrates the meteorological driving data, irrigation management data, and actual evapotranspiration to perform water balance calculations and obtain the runoff volume. After being aggregated within the sub-watershed, the runoff volume flows through the runoff collection module along the watershed system to the irrigation district drainage outlet, forming a monthly drainage volume simulation result.

[0053] The distributed hydrological model receives different water diversion schemes through the irrigation management module, and can support batch simulation of baseline scenarios and various water-saving scenarios.

[0054] Based on Digital Elevation Model (DEM) data, the D8 single-flow direction algorithm was used to calculate flow direction and runoff accumulation. A runoff accumulation threshold was set to extract the watershed system, and sub-watershed boundaries were initially delineated according to the same threshold. The watershed system defined the topological relationship of water flow transmission between sub-watersheds, with the runoff from each sub-watershed flowing step by step along the system to the irrigation district's drainage outlet. The initially delineated sub-watershed boundaries were spatially overlaid with the irrigation and drainage canal distribution data obtained in S1. The sub-watershed boundaries were then manually adjusted using the irrigation canal system as the physical boundary to form a sub-watershed delineation scheme that conforms to the actual runoff characteristics of the irrigation district. The land use data and soil type data obtained in S1 were reclassified. The original land use data classification system was a secondary land use category (such as dry land, grassland, etc.), and the classification was compared with the actual crop planting area ratio of the irrigation district. The dominant cultivated land type is further subdivided into the main crop types in the irrigation area so that the model can distinguish the evapotranspiration characteristics and irrigation differences of different crops. For soil type data, it is reclassified according to the soil type level of the Chinese Soil System Classification. Soil types with smaller distribution areas in the irrigation area are merged into similar main soil types, and several main soil types with significant area proportions in the irrigation area are retained. In this embodiment, the soil physical parameters required by each HRU of the distributed hydrological model are supplemented by the Chinese Soil Attribute Database, including soil bulk density, field capacity, saturated hydraulic conductivity, effective water holding capacity, and soil organic carbon content, in order to calculate the soil moisture transport process. By superimposing the reclassified land use data and soil type data, several hydrological response units (HRUs) are generated by discretizing the entire watershed.

[0055] S2.2. The distributed hydrological model is calibrated and validated using hydrological monitoring data. The simulation period for the distributed hydrological model is defined, including a warm-up period, a calibration period, and a validation period. The warm-up period is used to eliminate uncertainties in the initial state variables of the model; the calibration period is used for parameter sensitivity analysis and optimization calibration; and the validation period is used to independently evaluate the model's generalization ability. The SUFI-2 parameter sensitivity analysis algorithm is used to identify key parameters sensitive to the discharge volume simulation for subsequent automatic calibration processes to improve calibration efficiency. Automatic calibration is performed using monthly measured discharge volume data from the discharge outlet as the calibration target, and the evaluation index includes the Nash efficiency coefficient. Coefficient of determination and percentage deviation During the verification period , and The evaluation metric formula is as follows to ensure the model passes validation:

[0056] (1)

[0057] (2)

[0058] (3)

[0059] in, The Nash efficiency coefficient. As the coefficient of determination, This is a percentage deviation. For the first Monthly measured drainage volume (10,000 m³) For the first Simulated monthly drainage volume (10,000 m³) This represents the average value (in ten thousand m³) of the measured drainage volume series. To simulate the average value of the drainage volume series (in ten thousand m³), The total number of months in the verification period. For continuous time series indexes, .

[0060] S2.3 Water-saving scenarios are generated by combining water-saving strategies and water-saving intensities. Water-saving strategies are categorized into three types based on irrigation time windows: Type S (year-round water saving), Type G (growing season water saving), and Type A (autumn irrigation water saving). Each type of water-saving strategy reduces the water diversion volume within its corresponding irrigation time window, while maintaining the baseline scenario level for water diversion in non-target time windows. Specifically: Type S (year-round water saving) reduces the water diversion volume proportionally and evenly across all months of the year; Type G (growing season water saving) reduces the irrigation water diversion volume only during the crop growing season (April-September), while maintaining the baseline scenario level for water diversion during the non-growing season (October-March of the following year); Type A (autumn irrigation water saving) reduces the irrigation water diversion volume only during the autumn irrigation window (October-November), while maintaining the baseline scenario level for water diversion in other months. Water-saving intensity is defined as the reduction ratio of water diversion volume under each water-saving strategy relative to the baseline scenario water diversion volume. Each type of water-saving strategy has at least three levels of water-saving intensity. The three types of water-saving strategies combined with multiple levels of water-saving intensity constitute a scenario matrix. The specific form is as follows:

[0061] (4)

[0062] in, , , , These are water-saving scenarios for the S-shaped (year-round water-saving) strategy under water-saving intensities of level 1, 2 to n. , , , These are water-saving scenarios for the G-type (growing season water-saving) strategy under water-saving intensities of level 1, 2 to n. , , , These represent water-saving scenarios for the Type A (autumn irrigation water-saving) strategy at water-saving intensities of levels 1, 2, to n. In the scenario matrix, each row represents the same water-saving strategy, each column represents the same water-saving intensity level, and each element corresponds to a water-saving scenario. Historical actual water diversion conditions are used as the baseline scenario, and its water diversion scheme is denoted as the baseline scenario water diversion scheme. The water diversion schemes for each water-saving scenario are denoted as the water-saving scenario water diversion schemes.

[0063] S2.4 Using the distributed hydrological model constructed and validated in S2.1-S2.2, the validated distributed hydrological model is driven by the water diversion scheme of the baseline scenario and the water diversion schemes of each water-saving scenario, respectively, to obtain the simulated monthly drainage volume of the baseline scenario and each water-saving scenario. The ratio of the simulated monthly drainage volume of each water-saving scenario to that of the baseline scenario is calculated as the correction coefficient for the corresponding water-saving scenario. The correction coefficient is multiplied by the measured monthly drainage volume of the baseline scenario to obtain the predicted monthly drainage volume of each water-saving scenario. Specifically, the predicted monthly drainage volume of each water-saving scenario is calculated using a ratio-driven formula:

[0064] (5)

[0065] in, This indicates the predicted monthly wastewater discharge (in ten thousand m³) under water-saving scenarios. 3 ); This represents the measured drainage volume (in ten thousand m³) at the drainage outlet for the month corresponding to the baseline scenario. 3 (i.e., the measured monthly drainage volume under the baseline scenario); and The simulated monthly wastewater discharge (in ten thousand m³) represents the baseline scenario and the water-saving scenario, respectively. 3 ). This is the dimensionless variation coefficient of the water quantity response in the distributed hydrological model, i.e., the correction coefficient.

[0066] S3, such as Figure 2 As shown, a log-linear regression model for drainage mineralization (denoted as the mineralization model) is constructed using measured monthly drainage volume, current and previous month water diversion volume, previous accumulated water deficit, and monthly fixed effects as independent variables, and drainage mineralization as the dependent variable. The model is validated using the continuous observation sequence of measured monthly drainage mineralization at the drainage outlet obtained in S1. After validation, the predicted monthly drainage volume for each water-saving scenario output from S2 is input into the validated mineralization model to calculate the predicted monthly drainage mineralization. Then, the predicted drainage mineralization for month t of each water-saving scenario is multiplied by the predicted drainage volume for the same month under the corresponding water-saving scenario to obtain the predicted coupled salt discharge for month t of that water-saving scenario. The predicted monthly coupled salt discharge is accumulated and averaged to obtain the predicted annual average salt discharge for each scenario. The specific steps include:

[0067] S3.1 Construct a logarithmic linear regression model for drainage salinity (denoted as the salinity model). The salinity model is a weighted sum of the natural logarithm of monthly drainage, the natural logarithm of monthly water diversion, the natural logarithm of the previous month's water diversion, the accumulated water deficit from the previous period, and the monthly fixed effect. The accumulated water deficit from the previous period is calculated based on the previous potential evapotranspiration, effective precipitation, and effective irrigation water. The model form is as follows:

[0068] (6)

[0069] (7)

[0070] in, The value is the salinity of the drainage water (g / L). This represents the monthly drainage volume (in ten thousand m³). This represents the monthly water diversion volume (in ten thousand m³). This represents the water diversion volume for the previous month, used to account for the lag effect of irrigation on drainage; The cumulative water deficit in the previous period (100 mm) reflects the impact of the previous soil moisture status on the current salt discharge. In this example, the cumulative water deficit in the previous 3 months is taken. To fix the monthly effect, seasonal background differences in drainage mineralization were controlled; The empirically estimated coefficients of the model are to be determined. For the first Monthly potential evapotranspiration (mm), For the first Monthly effective precipitation (mm), For the first Monthly effective irrigation water volume (mm), To trace back to the month sequence number, During the model building phase, all variables in equation (6) are based on the measured values ​​of hydrological monitoring data obtained by S1, and the monthly drainage volume is calculated accordingly. Monthly drainage volume and mineralization were measured at the drainage outlet. Monthly drainage salinity and water diversion volume were measured at the drainage outlet. The measured monthly water diversion volume and the accumulated water deficit in the early stage were used as the basis for the calculation. The irrigation water volume is calculated using equation (7) based on meteorological data and measured data. During the water-saving scenario analysis phase, the monthly predicted drainage volume for each water-saving scenario output by S2 is substituted into equation (6). Substitute the water diversion scheme for the corresponding scenario into and The predicted monthly drainage mineralization for each scenario was calculated.

[0071] S3.2. Based on the continuous monthly observation sequence of measured drainage mineralization at the irrigation district's drainage outlet, the mineralization model was validated using model evaluation indices. The simulation accuracy of drainage mineralization was evaluated in logarithmic space and the simulation accuracy of coupled salt discharge was evaluated in the original physical space. The validation period was used as the basis for further verification. and The criteria for judgment are as follows: Specifically, the measured monthly drainage mineralization continuous observation sequence is divided into a training period and a validation period. The training period is used to estimate the parameters of the regression coefficients of the mineralization model. The measured monthly drainage mineralization of the training period is used as the dependent variable, and the measured monthly drainage volume, monthly water diversion volume, the cumulative water deficit of the previous three months calculated by formula (7) from the measured data of the same period and the month fixed effect are used as independent variables. The parameters of each regression coefficient are estimated by substituting them into formula (6). The estimated coefficients are substituted back into formula (6) to obtain the monthly drainage mineralization of the training period. The validation period is used to independently evaluate the generalization ability of the mineralization model. The measured monthly drainage volume, measured monthly water diversion volume, the cumulative water deficit of the previous three months calculated by formula (7) and the month fixed effect are used as independent variables. The estimated coefficients are input into formula (6) to obtain the monthly drainage mineralization of the validation period. The accuracy is verified by calculating the evaluation index and judging whether it meets the preset standard. The model evaluation index includes the coefficient of determination. Nash efficiency coefficient Klin-Gupta efficiency coefficient and percentage deviation The simulation accuracy of drainage mineralization was evaluated in both logarithmic and physical spaces. In logarithmic space, the accuracy was assessed by comparing the monthly drainage mineralization during the validation period with the measured monthly drainage mineralization. In physical space, the accuracy was assessed by comparing the simulation accuracy of coupled salt discharge. During the training period, coupled salt discharge was obtained by multiplying the measured monthly drainage volume during the training period with the drainage mineralization during the training period. During the validation period, coupled salt discharge was obtained by multiplying the measured monthly drainage volume during the validation period with the measured monthly drainage mineralization during the validation period. The coupled salt discharge during the validation period was compared with the measured monthly salt discharge from the drainage outlet obtained in S1. The same evaluation index system was used in both spaces, with the validation period... and As the criterion for judgment, and As a supplementary reference indicator The calculation formula is as follows:

[0072] (8)

[0073] (9)

[0074] (10)

[0075] in, For the first Monthly measured salt discharge (10,000 tons); This represents the average value (in ten thousand tons) of the measured salt discharge series. For the first Monthly coupled salt discharge (10,000 tons), the coupled salt discharge during the training period is taken, and the coupled salt discharge during the verification period is taken; This represents the average value (in ten thousand tons) of the coupled salt discharge sequence; Pearson correlation coefficient; The variation ratio is the ratio of the coupled salt excretion to the standard deviation of the measured salt excretion. The mean ratio is the ratio of the coupled salt excretion to the average measured salt excretion. , These represent the measured salt discharge volume and the standard deviation of the coupled salt discharge volume series (in ten thousand tons), respectively. , This represents the measured salt discharge volume and the average value of the coupled salt discharge volume sequence (in ten thousand tons).

[0076] S3.3 After verification, input the predicted monthly drainage volume for each water-saving scenario obtained in step S2 into the verified salinity model, and calculate the predicted drainage salinity monthly. Multiply the predicted monthly drainage volume by the predicted drainage salinity for the corresponding month to obtain the predicted monthly coupled salt discharge sequence for each water-saving scenario:

[0077] (11)

[0078] in, For water conservation scenario number Monthly forecast of coupled salt discharge (10,000 tons); For water conservation scenario number Monthly forecasted drainage (10,000 m³) 3 ); For water conservation scenario number Monthly predicted drainage mineralization (g / L); 10 −3 The unit conversion factor is used; the monthly salt discharge predictions for each water-saving scenario are coupled and averaged to obtain the predicted annual average salt discharge for each water-saving scenario, which serves as the basic input for the salt accumulation risk assessment in step S4.

[0079] S4. Based on the predicted annual average salt discharge under each water-saving scenario, calculate the Salt Drainage Ratio (SDR) and Salt Retention Rate (SRR) for each scenario. Assess the salt accumulation risk level for each scenario based on the SDR, and assess the degree of salt accumulation based on the SRR. Classify the salt accumulation risk into three levels: basically balanced, low risk, and high risk, according to the critical threshold of the SDR. The critical thresholds for the SDR corresponding to each salt accumulation risk level are denoted as follows: ,in , To achieve a basic balance between the lower threshold and the upper threshold for low risk, and to cover the uncertainty tolerance of model predictions, These are the lower and upper risk thresholds, corresponding to the equivalent salt removal efficiency constraints for a 10% salt tolerance loss in crops. The lower limit threshold for high risk represents the critical inflection point where salt removal efficiency will collapse. Below this value, soil salt storage will increase positively and cause structural salt accumulation. Used for cross-sectional comparison of salt removal capacity among water-saving scenarios. The steps used to assess the degree of salt accumulation are as follows:

[0080] S4.1 For each water-saving scenario, calculate the percentage of its predicted annual average salt discharge relative to the baseline scenario's annual average salt discharge, i.e., the salt discharge ratio. As the primary evaluation indicator, the calculation formula is as follows:

[0081] (12)

[0082] in, Predicted annual average salt discharge (10,000 tons / year) under water-saving scenario; This represents the average annual salt discharge (10,000 tons / year) corresponding to the baseline scenario. This indicates that the salt excretion capacity is not impaired; This indicates a decline in salt removal capacity and a risk of salt retention. Specific salt accumulation risk levels include: basically balanced (…). Low risk High risk ).in, To achieve a basic balance between the lower threshold and the upper threshold for low risk, and to cover the uncertainty tolerance of model predictions, These are the lower and upper risk thresholds, corresponding to the equivalent salt removal efficiency constraints for a 10% salt tolerance loss in crops. The lower limit threshold for high risk represents the critical inflection point where salt removal efficiency will collapse. Below this value, soil salt storage will increase positively and cause structural salt accumulation.

[0083] S4.2 Calculate the amount of salt introduced from external irrigation under each water-saving scenario based on the annual external water diversion volume and the multi-year average water diversion salinity, and introduce the salt retention rate ( As the second evaluation indicator, the salt retention rate is defined as the proportion of exogenous salt entering the irrigation area that is retained in the soil system. The salt retention rate is equal to the difference between the amount of salt introduced by exogenous irrigation and the predicted average annual salt discharge, divided by the amount of salt introduced by exogenous irrigation. The formula is as follows:

[0084] (13)

[0085] (14)

[0086] in, The salt retention rate (%) of the irrigation area under water-saving scenarios. The amount of salt introduced by external irrigation under this scenario (ten thousand tons / year); Annual external water diversion volume (10,000 m³) under the given scenario 3 / a); This represents the multi-year average salinity of the diverted water. When... A value greater than zero indicates that the soil is in a state of net salt accumulation. The higher the value, the more severe the salt accumulation; Reflecting the relative decline in salt excretion capacity, The absolute retention ratio, which reflects soil salinity, and the two together constitute a dual-indicator assessment.

[0087] S5. Fit each water-saving strategy separately. A sensitivity equation for the water diversion reduction ratio is used to invert the safety red line for water diversion reduction based on the critical threshold of salt discharge ratio corresponding to each salt accumulation risk level; the predicted monthly drainage volume under each water-saving scenario is fitted to the annual data. The linear regression equation, based on the linear regression equation, with The corresponding monthly high-risk drainage threshold is calculated by reverse calculation. The final output includes the water diversion reduction safety line and the monthly high-risk drainage threshold for each water-saving strategy. The specific steps include:

[0088] S5.1, Using the water diversion reduction ratio for each water-saving scenario as the independent variable, As the dependent variable, sensitivity equations for each water-saving strategy are fitted. The water diversion reduction ratio is the percentage reduction in water diversion volume relative to the baseline scenario, which is numerically equal to the water-saving intensity of each water-saving scenario in S2.3. A linear regression with an intercept of 100% (i.e., when the water diversion reduction ratio is 0, the salt discharge ratio is 100%) is used, and the sensitivity equation formula is as follows:

[0089] (15)

[0090] in, The salt discharge ratio calculated for S4.1; The percentage reduction in water diversion relative to the baseline scenario (%). The response sensitivity coefficient (slope) obtained from regression fitting has the physical meaning of the response rate when the water pressure is reduced by 1%. Marginal changes; The higher the value, the more sensitive the salt removal capacity of this strategy type is to water diversion reduction, and the more cautious one should be when implementing water conservation measures.

[0091] S5.2, The critical threshold for salt discharge ratio corresponding to each salt accumulation risk level defined in step S4.1 ( , , Substitute the values ​​into the sensitivity equations of each water-saving strategy fitted in step S5.1, and solve in reverse for the corresponding water diversion reduction safety threshold. The safety red lines for water diversion reduction were obtained for each strategy type under three salt accumulation risk levels: basic balance, low risk, and high risk. The calculation formula is:

[0092] (16)

[0093] in, For each risk level, the critical threshold for salt excretion ratio is taken sequentially. , , This leads to the conclusion that each water-saving strategy is in basic equilibrium ( Low risk ), and high risk ( The corresponding safety red lines for water diversion reduction under the three risk levels.

[0094] S5.3, Regarding the first Monthly predicted drainage volume under water-saving scenario (monthly) 10,000 m 3 (month) is the independent variable, and the whole year Using [variable name] as the dependent variable, fit a linear regression equation:

[0095] (17)

[0096] in, and For the first Empirical fitting parameters for the month;

[0097] Based on the aforementioned linear regression equation and the empirical regression parameters for each month, let the annual salt excretion ratio equal to the high-risk lower limit threshold (i.e., let...). (High-risk lower limit threshold), and then reverse the process to deduce the high-risk drainage threshold for that month. :

[0098] (18)

[0099] The predicted drainage volume for the month is lower than hour Will fall below This leads to structural salt accumulation.

[0100] S5.4, combining the water diversion reduction safety red line of S5.2 and the high-risk drainage threshold of S5.3, forms a two-layer decision support output: a strategy layer and a monthly layer. The strategy layer uses a decision matrix to summarize the water diversion reduction safety red lines of the three water-saving strategies under three risk levels. In summary, the decision matrix consists of rows corresponding to water-saving strategies and columns corresponding to salt accumulation risk levels. Each element represents the water diversion reduction safety limit for each strategy at its corresponding salt accumulation risk level. Represented as:

[0101] (19)

[0102] in, , , The S-shaped (year-round water conservation) strategy is basically balanced ( Low risk ) and high risk ( The water diversion reduction safety red line below; , , The G-type (water-saving during the growing season) strategy is in basic balance ( Low risk ) and high risk ( The water diversion reduction safety red line below; , , The two strategies are: Type A (water-saving during autumn irrigation) and Basic Balance. Low risk ) and high risk ( The water diversion reduction safety red line under ( ); monthly layer output of high-risk drainage threshold for each month. Ranking of key months for salt removal safety.

[0103] The steps of the present invention will be described in detail below with specific implementation examples.

[0104] This embodiment takes a large irrigation area as the object and collects spatial data, including the digital elevation model (DEM) data of the irrigation area, the distribution data of the irrigation and drainage canal system, land use data, and soil type data; it also collects daily precipitation, temperature, wind speed, solar radiation, and relative humidity data during the study period; and continuous observation sequences of monthly drainage volume, water diversion volume, salt discharge, and drainage mineralization at the irrigation area's drainage outlets, with an observation period of no less than 100 months, as benchmark data for the calibration and verification of the distributed hydrological model and the estimation of parameters for the log-linear regression model of drainage mineralization; and obtains the planting date, planting area, irrigation window, and irrigation water depth of the main crops (spring wheat, corn, and sunflower) in the irrigation area. Irrigation management data are shown in Table 1.

[0105] Table 1 Crop Irrigation Management Scheme for the Study Area

[0106]

[0107] Based on DEM data, sub-basin drainage systems were extracted and boundaries were delineated to form a sub-basin delineation scheme that conforms to the actual runoff characteristics of the basin. By overlaying reclassified land use maps, soil type maps, and slope data, the entire basin was discretized to generate hydrological response units (HRUs). The model simulation period was set from 2013 to 2023, with 2013-2014 as the warm-up period, 2015-2020 as the calibration period, and 2021-2023 as the validation period.

[0108] Figure 3 Figures (a) and (b) show the model evaluation results during the rate-up and validation periods, where all indicators meet the requirements. , , The validation criteria were met, indicating that the model has good reproducibility of monthly variations in drainage volume and can be used for comparative analysis of drainage responses under different irrigation reduction scenarios.

[0109] The log-linear regression model of drainage salinity constructed in step S3 was used, and the model evaluation index results are shown in Table 2. Three strategy types were set according to irrigation time windows: Type S (year-round water saving), Type G (growing season water saving), and Type A (autumn irrigation season water saving), with three levels of water saving intensity (5%, 10%, and 15%) for each type, resulting in a total of nine water-saving scenarios plus one baseline scenario (historical actual water diversion status). The ratio-driven method was used to jointly solve for the predicted drainage volume of each scenario, and the salinity discharge was obtained through coupling.

[0110] Table 2. Evaluation Indicators of the Accuracy of Mineralization Model and Coupled Salt Excretion Simulation

[0111]

[0112] Figure 4 Figures (a) to (d) show the monthly changes in drainage and salt discharge under nine water-saving scenarios. The three strategies exhibit clear seasonal response differences: year-round water saving shows a simultaneous decrease in drainage and salt discharge in most months of the year; the impact of growing season water saving is mainly concentrated in April to September, with a weaker response in autumn and winter; autumn irrigation water saving mainly alters the drainage and salt discharge process from October to December, indicating that autumn irrigation is a key window for water-saving policies to affect salt discharge.

[0113] Calculate according to step S4 and , Figure 5 (a) and (b) combined month by month The heat map and monthly salt accumulation curve show that winter (December-March) and the later stages of autumn irrigation are the main contributing periods to salt accumulation risk. Low drainage, evaporation, and groundwater salinity are prominent in winter, easily creating natural salt-dissipation blind spots even without direct irrigation reduction. Water conservation during the growing season mainly reduces salt-dissipation capacity from April to September, but triggers a weak salt-dissipation compensation effect in autumn and winter; however, water conservation during autumn irrigation results in a sudden and concentrated release of risk in November-December (the later stages of autumn irrigation), with a steep and precipitous collapse in salt discharge. This dramatic difference in seasonal response indicates that the risks of autumn irrigation regulation have a strong lag and cascading effects.

[0114] Using the sensitivity equation from step S5, a linear regression with an intercept of 100% is fitted to each water-saving strategy. (All > 0.98), and based on the sensitivity equation, a preliminary water-saving safety boundary can be obtained, as shown in Table 3. Under basic equilibrium ( Under the requirements, the safety red lines for water diversion reduction for types S, G, and A are ≤1.4%, ≤1.5%, and ≤0.8%, respectively; under low risk ( Under the standard, the safe red lines for water diversion reduction are ≤5.0%, ≤5.4%, and ≤2.9%, respectively; while when the water diversion reduction exceeds 10.7%, 11.6%, and 6.2%, respectively, the system will directly fall below the high-risk lower limit. ).

[0115] Table 3 Safety Red Lines for Water Diversion Reduction under Each Strategy

[0116]

[0117] Using the established monthly drainage volume and annual By using the statistical regression relationship, the monthly drainage threshold for each month that falls below the high-risk lower limit can be calculated. Figure 6 The monthly distribution of high-risk drainage thresholds shows that different months saw a decrease in the high-risk threshold (…). The threshold values ​​for drainage volume vary significantly: approximately 0.12 billion m³ in January-February, approximately 0.43-0.61 billion m³ in May-October, and the highest threshold in November, approximately 0.86 billion m³.

[0118] As can be seen from the above embodiments, the irrigation district water-saving and salt discharge response prediction and salt accumulation risk assessment method based on strategy classification described in this embodiment can clearly characterize the differentiated response sensitivity of irrigation district salt discharge capacity to water diversion reduction under different irrigation time window water control strategies, quantitatively determine the water diversion reduction safety red line and monthly high-risk drainage threshold corresponding to each strategy type; and can also be used to determine the salt discharge ratio ( ) and salt retention rate ( The dual-indicator joint evaluation framework simultaneously meets the dual evaluation needs of the dynamic evolution of salt discharge capacity in irrigation areas and the maintenance of soil salinity stability under water-saving scenarios, providing a systematic decision support chain for the refined management of water resources in irrigation areas.

[0119] Example 2:

[0120] A method and system for predicting water-saving and salt-discharge response and assessing salt accumulation risk in irrigation districts based on strategy classification, including:

[0121] The model building, calibration and verification, and prediction unit is used to build and calibrate a distributed hydrological model using basic irrigation district data; based on the calibrated and verified distributed hydrological model, it obtains the simulated monthly drainage volume of each water-saving scenario and the baseline scenario; it calculates correction coefficients based on the simulated monthly drainage volume of each water-saving scenario and the baseline scenario; and it obtains the predicted monthly drainage volume of each water-saving scenario based on the correction coefficients; each water-saving scenario is generated by combining water-saving strategies and water-saving intensities, and each type of water-saving strategy has multiple levels of water-saving intensities.

[0122] The salt discharge calculation unit is used to construct a mineralization model with measured monthly drainage volume, current and previous month water diversion volume, previous accumulated water deficit and monthly fixed effect as independent variables and drainage mineralization as dependent variable. The mineralization model is verified using irrigation district basic data. The predicted monthly drainage volume of each water-saving scenario is input into the verified mineralization model to calculate the predicted drainage mineralization. Based on the predicted drainage mineralization and the predicted monthly drainage volume, the predicted annual average salt discharge volume of each water-saving scenario is calculated.

[0123] The salt accumulation risk level assessment unit is used to predict the average annual salt discharge under each water-saving scenario, calculate the salt discharge ratio and salt retention rate under each water-saving scenario, assess the salt accumulation risk level under each water-saving scenario based on the salt discharge ratio, and assess the degree of salt accumulation based on the salt retention rate.

[0124] The early warning unit is used to fit a sensitivity equation between the salt discharge ratio and the water diversion reduction ratio for each water-saving strategy. Based on the sensitivity equation, the water diversion reduction safety red line is inverted using the critical threshold of the salt discharge ratio corresponding to each salt accumulation risk level. The unit also fits a linear regression equation between the predicted monthly drainage volume and the annual salt discharge ratio under each water-saving scenario. Based on the linear regression equation, the corresponding monthly high-risk drainage volume threshold is inverted using the salt discharge ratio equal to the high-risk lower limit threshold in the salt risk level. Finally, the water diversion reduction safety red line and the monthly high-risk drainage threshold for each water-saving strategy are output.

[0125] Example 3:

[0126] An electronic device, comprising:

[0127] Memory, used to store computer programs;

[0128] A processor is used to execute the computer program to implement the strategy-based method for predicting irrigation district water-saving and salt discharge response and assessing salt accumulation risk.

[0129] Example 4:

[0130] A non-volatile storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the strategy-based water-saving and salt-discharge response prediction and salt accumulation risk assessment method for irrigation areas.

[0131] Example 5:

[0132] A computer program product includes a computer program / instructions that, when executed by a processor, implement the aforementioned strategy-based irrigation district water-saving and salt discharge response prediction and salt accumulation risk assessment method.

Claims

1. A method for predicting water-saving and salt-discharge response and assessing salt accumulation risk in irrigation districts based on strategy classification, characterized in that, include: A distributed hydrological model was constructed and calibrated using basic data from the irrigation district. The monthly drainage volume of each water-saving scenario and the baseline scenario is obtained based on the distributed hydrological model. The correction coefficient is calculated based on the monthly drainage volume of each water-saving scenario and the baseline scenario, and the predicted monthly drainage volume of each water-saving scenario is obtained based on the correction coefficient. Each water-saving scenario is generated by a combination of water-saving strategies and water-saving intensities, and each type of water-saving strategy has multiple levels of water-saving intensities. Using measured monthly drainage volume, current and previous month water diversion volume, previous accumulated water deficit, and monthly fixed effects as independent variables, and drainage mineralization as the dependent variable, a mineralization model was constructed. The mineralization model was validated using basic irrigation district data. The predicted monthly drainage volume for each water-saving scenario was input into the validated mineralization model to calculate the predicted monthly drainage mineralization. Based on the predicted monthly drainage mineralization and the predicted monthly drainage volume, the predicted annual average salt discharge for each water-saving scenario was calculated. Based on the predicted annual average salt discharge under each water-saving scenario, the salt discharge ratio and salt retention rate under each water-saving scenario are calculated. The salt discharge ratio is used to assess the salt accumulation risk level under each water-saving scenario, and the salt retention rate is used to assess the degree of salt accumulation. Sensitivity equations for salt discharge ratio and water diversion reduction ratio are fitted for each water-saving strategy. Based on the sensitivity equations, the safety red line for water diversion reduction is inverted using the critical threshold of salt discharge ratio corresponding to each salt accumulation risk level. A linear regression equation is fitted to the predicted monthly drainage volume and annual salt discharge ratio under each water-saving scenario. Based on the linear regression equation, the corresponding monthly high-risk drainage volume threshold is calculated by back-calculating the salt discharge ratio equal to the high-risk lower limit threshold in the salt accumulation risk level. Finally, the water diversion reduction safety red line and monthly high-risk drainage threshold for each water-saving strategy are output.

2. The method according to claim 1, characterized in that, The construction and calibration of the distributed hydrological model using basic irrigation district data includes: The distributed hydrological model includes a runoff generation module, a runoff collection module, an evapotranspiration module, and an irrigation management module. Each module uses a Hydrological Response Unit (HRU) as the basic computational unit and the watershed system as the river confluence network. For each HRU, the evapotranspiration module calculates the actual evapotranspiration based on irrigation district baseline data. The irrigation management module determines the irrigation window and irrigation depth for that HRU based on the irrigation district baseline data and the input water diversion plan. The runoff generation module integrates the irrigation district baseline data and the actual evapotranspiration to perform water balance calculations to obtain the runoff volume. After being aggregated within sub-watersheds, the runoff volume is collected step-by-step along the watershed system to the irrigation district drainage outlet via the runoff collection module, forming a monthly drainage volume simulation result. A parameter sensitivity analysis algorithm is used to identify key parameters sensitive to the drainage volume simulation, and automatic calibration is performed using the measured monthly drainage volume as the calibration target, with the Nash efficiency coefficient as the calibration target. Coefficient of determination and percentage deviation The model has passed the validation criteria; The correction factor is the ratio of the simulated monthly drainage volume of each water-saving scenario to that of the baseline scenario. Multiplying the correction factor by the measured monthly drainage volume of the baseline scenario yields the predicted monthly drainage volume for each water-saving scenario.

3. The method according to claim 1, characterized in that, The water-saving strategies are divided into year-round water saving, growing season water saving, and autumn irrigation season water saving according to the irrigation window. The water-saving intensity is the reduction ratio of water diversion under each water-saving strategy relative to the baseline scenario water diversion.

4. The method according to claim 1, characterized in that, The mineralization model is a weighted sum of the natural logarithm of monthly drainage, the natural logarithm of monthly water diversion, the natural logarithm of previous month's water diversion, the accumulated water deficit in the previous period, and the monthly fixed effect. The accumulated water deficit in the previous period is calculated based on the potential evapotranspiration, effective precipitation, and effective irrigation water in the previous period. Based on the continuous monthly observation sequence of drainage mineralization, the mineralization model was validated using model evaluation indices, including the coefficient of determination. Nash efficiency coefficient Klin-Gupta efficiency coefficient and percentage deviation The simulation accuracy of drainage salinity was evaluated in logarithmic space and the simulation accuracy of coupled salt discharge was evaluated in the original physical space, with the validation period determination coefficient as the basis for the evaluation. And Nash efficiency coefficient As the criterion for judgment, and As a supplementary reference indicator; The measured monthly drainage mineralization continuous observation sequence is divided into a training period and a validation period; the training period is used to estimate the parameters of the regression coefficients of the mineralization model, and the validation period is used to independently evaluate the generalization ability of the mineralization model. After verification, the predicted monthly drainage volume for each water-saving scenario is input into the verified mineralization model, and the predicted drainage mineralization is calculated monthly. The predicted monthly drainage volume is multiplied by the predicted drainage mineralization for the corresponding month to obtain the predicted monthly coupled salt discharge sequence for each scenario. The predicted monthly coupled salt discharge for each scenario is accumulated and averaged to obtain the predicted annual average salt discharge for each scenario.

5. The method according to claim 1, characterized in that, Based on the predicted annual average salt discharge under each water-saving scenario, the salt discharge ratio and salt retention rate under each scenario are calculated; including: The salt excretion ratio To predict the percentage of annual average salt discharge under the water-saving scenario relative to the baseline scenario, when This indicates that the salt excretion capacity is not impaired; This indicates a decline in salt removal capacity and a risk of salt retention; the salt accumulation risk level includes: basically balanced. Low risk High risk ;in, To achieve a basic balance between the lower threshold and the upper threshold for low risk, These are the lower limit threshold for low risk and the upper limit threshold for high risk. This is the lower limit threshold for high risk. The amount of salt introduced by external irrigation under each water-saving scenario is calculated based on the annual external water diversion volume and the multi-year average water diversion salinity. The salt retention rate is equal to the difference between the amount of salt introduced by external irrigation and the predicted annual average salt discharge, divided by the amount of salt introduced by external irrigation. When the salt retention rate is greater than zero, it indicates that the soil is in a state of net salt accumulation. The higher the salt retention rate, the more serious the salt accumulation.

6. The method according to claim 1, characterized in that, The process involves fitting a sensitivity equation between the salt discharge ratio and the water diversion reduction ratio for each water-saving strategy, and then, based on this sensitivity equation, using the critical threshold of the salt discharge ratio corresponding to each salt accumulation risk level to invert the safety red line for water diversion reduction; including: Using the water diversion reduction ratio of each water-saving scenario as the independent variable and the salt discharge ratio as the dependent variable, a sensitivity equation for each water-saving strategy is fitted. The water diversion reduction ratio is the percentage reduction of water diversion volume relative to the baseline scenario, which is numerically equal to the water-saving intensity of each water-saving scenario. The sensitivity equation adopts a linear regression with an intercept of 100% and the slope is the response sensitivity coefficient obtained by regression fitting. Substitute the critical threshold of salt discharge ratio corresponding to each salt accumulation risk level into the sensitivity equation of each fitted water-saving strategy, and solve the corresponding water diversion reduction safety red line in reverse to obtain the water diversion reduction safety red line for each strategy type under the three salt accumulation risk levels of basic balance, low risk and high risk. Using the predicted monthly drainage volume under the water-saving scenario as the independent variable and the annual salt discharge ratio as the dependent variable, a linear regression equation is fitted. Based on the linear regression equation and the empirical regression parameters for each month, the annual salt discharge ratio is set to equal the high-risk lower limit threshold, and the high-risk drainage volume threshold for that month is calculated in reverse. Based on the water diversion reduction safety red line and the monthly high-risk drainage threshold, a two-layer decision support output is formed: strategy layer and monthly layer. The strategy layer summarizes the water diversion reduction safety red line of each water-saving strategy under each risk level using a decision matrix. In the decision matrix, rows correspond to water-saving strategies, columns correspond to salt accumulation risk levels, and the elements of the decision matrix are the water diversion reduction safety red line of each strategy under the corresponding salt accumulation risk level. The monthly layer outputs the monthly high-risk drainage threshold and the ranking of critical months for salt discharge safety.

7. A system for predicting water-saving and salt-discharge response and assessing salt accumulation risk in irrigation districts based on strategy classification, characterized in that, include: The model building, calibration and verification, and prediction unit is used to build and calibrate a distributed hydrological model using basic data from the irrigation district; based on the calibrated and verified distributed hydrological model, the simulated monthly drainage volume of each water-saving scenario and the baseline scenario is obtained; the correction coefficient is calculated based on the simulated monthly drainage volume of each water-saving scenario and the baseline scenario; and the predicted monthly drainage volume of each water-saving scenario is obtained based on the correction coefficient. Each water-saving scenario is generated by a combination of water-saving strategies and water-saving intensities, and each type of water-saving strategy has multiple levels of water-saving intensities. The salt discharge calculation unit is used to construct a mineralization model with measured monthly drainage volume, current and previous month water diversion volume, previous accumulated water deficit and monthly fixed effect as independent variables and drainage mineralization as dependent variable. The mineralization model is verified using irrigation district basic data. The predicted monthly drainage volume of each water-saving scenario is input into the verified mineralization model to calculate the predicted drainage mineralization. Based on the predicted drainage mineralization and the predicted monthly drainage volume, the predicted annual average salt discharge volume of each water-saving scenario is calculated. The salt accumulation risk level assessment unit is used to predict the average annual salt discharge under each water-saving scenario, calculate the salt discharge ratio and salt retention rate under each water-saving scenario, assess the salt accumulation risk level under each water-saving scenario based on the salt discharge ratio, and assess the degree of salt accumulation based on the salt retention rate. The early warning unit is used to fit a sensitivity equation between the salt discharge ratio and the water diversion reduction ratio for each water-saving strategy. Based on the sensitivity equation, the water diversion reduction safety red line is inverted using the critical threshold of the salt discharge ratio corresponding to each salt accumulation risk level. A linear regression equation is fitted to the predicted monthly drainage volume and annual salt discharge ratio under each water-saving scenario. Based on the linear regression equation, the corresponding monthly high-risk drainage volume threshold is deduced by using the salt discharge ratio equal to the high-risk lower limit threshold in the salt risk level. Finally, the water diversion reduction safety red line and monthly high-risk drainage threshold for each water-saving strategy are output.

8. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the method for predicting irrigation district water-saving and salt discharge response and assessing salt accumulation risk based on strategy classification as described in any one of claims 1-6.

9. A non-volatile storage medium, characterized in that, Used to store a computer program, wherein the computer program, when executed by a processor, implements the method for predicting water-saving and salt-discharging response and assessing salt accumulation risk in irrigation areas based on strategy classification as described in any one of claims 1-6.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the method for predicting water-saving and salt-discharge response and assessing salt accumulation risk in irrigation areas based on strategy classification as described in any one of claims 1-6.

Images