A rice field fertilization and pollution reduction synergistic regulation method based on water and fertilizer spatiotemporal coupling

Abstract

The application discloses a rice field fertilization and pollution reduction synergistic regulation method based on water and fertilizer space-time coupling, and relates to the technical field of farmland ecological regulation. A distributed optical fiber temperature measurement system is arranged along a water delivery channel to obtain a soil temperature field, a tracer is put into the upstream of the channel, and the concentration of the tracer is monitored in monitoring wells along the line; the position of a leakage point is identified based on the temperature difference between a real-time temperature field and a background temperature field, and lateral leakage flow velocity and leakage flow of each section of the channel are inverted based on the peak value arrival time of the tracer concentration; the leakage position and the leakage flow are taken as dynamic constraint input of a water and fertilizer space-time coupling model, the water distribution amount of the upstream of the channel is optimized, and the field fertilization time, the fertilization amount and the irrigation system are simultaneously adjusted. Through double coupling inversion of heat tracing and intelligent tracer, the application realizes full-chain collaboration of water delivery and distribution in an irrigation area and field water and fertilizer management, and is suitable for accurate regulation and control of water and fertilizer of various irrigation channels.

Description

Technical Field

[0001] This invention relates to the field of farmland ecological regulation technology, and in particular to a method for coordinated regulation of paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer. Background Technology

[0002] Paddy fields are a significant source of agricultural non-point source pollution and greenhouse gas emissions. Traditional rice cultivation methods, such as excessive application of chemical fertilizers and flood irrigation, not only result in severe nitrogen and phosphorus loss but also lead to excessive methane emissions from paddy fields. Simultaneously, long-term high-intensity farming causes a decline in soil organic matter and soil fertility degradation. Therefore, how to ensure stable and increased rice yields while simultaneously achieving soil fertility improvement, fertilizer reduction, and pollution reduction has become a core issue for the development of green agriculture.

[0003] In recent years, the academic community has proposed the multi-objective irrigation (MOI) strategy, which aims to achieve water conservation, emission reduction, and yield increase by differentially regulating water in paddy fields at different growth stages (such as drying the field during tillering to suppress methane emissions and storing water during fertilization to reduce nitrogen and phosphorus loss). Studies have shown that the MOI strategy can reduce total nitrogen loss by more than 35%, methane emissions by more than 66%, and irrigation water savings by more than 56%.

[0004] However, the engineering implementation of the MOI strategy suffers from a "black box" problem in the water delivery process. Specifically, a significant amount of water is lost through seepage during the water transport from the canal head to the field. Current technologies typically estimate seepage using empirical formulas (such as the Kosgakov empirical formula) or simplify it to a fixed proportionality coefficient, neglecting the dynamic feedback relationship between seepage and factors such as canal water level, soil moisture content, and water delivery duration. This results in the inability to accurately quantify seepage losses, leading to a significant discrepancy between the theoretical water allocation and the actual irrigation volume reaching the field.

[0005] Therefore, there is an urgent need for a collaborative control scheme that can accurately, in real-time, and at low cost quantify the leakage distribution along water conveyance channels and dynamically couple leakage information into water and fertilizer decision-making. To this end, a collaborative control method for paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer is proposed. Summary of the Invention

[0006] The main objective of this invention is to provide a method for coordinated regulation of fertilization and pollution reduction in paddy fields based on spatiotemporal coupling of water and fertilizer, which can effectively solve the problems in the background art.

[0007] To achieve the above objectives, this invention provides a method for synergistic regulation of fertilization improvement and pollution reduction in paddy fields based on spatiotemporal coupling of water and fertilizer, comprising the following steps:

[0008] Step 1, Monitoring of the water transfer process

[0009] A distributed fiber optic temperature measurement system was deployed along the water conveyance channel to obtain the background temperature field and real-time temperature field of the soil along the channel; and a tracer was introduced upstream of the channel, and the concentration of the tracer was monitored downstream of the channel and in soil water monitoring wells along the channel.

[0010] Step 2, Leakage Identification

[0011] Based on the temperature difference between the real-time temperature field and the background temperature field, the location of seepage points along the channel is identified; based on the tracer concentration, the lateral seepage velocity and the seepage flow rate of each section of the channel are calculated.

[0012] Step 3, Water and Fertilizer Decision

[0013] The identified locations of the seepage points, the lateral seepage velocity, and the seepage flow rates of each section of the channel are used as constraints and input into a pre-constructed water-fertilizer spatiotemporal coupling model to dynamically adjust the water distribution upstream of the channel and simultaneously adjust the fertilization time, fertilization amount, and irrigation system of the paddy field.

[0014] Furthermore, in the leakage identification step, identifying the location of the leakage point based on the temperature difference between the real-time temperature field and the background temperature field specifically includes:

[0015] Calculate the spatial gradient ▽(ΔT) of the temperature difference ΔT, where ΔT = ΔT - T b Where T is the real-time temperature field, T b Background temperature field;

[0016] Define the leakage intensity index I leak (x), which is positively correlated with the magnitude of the temperature difference ΔT and the spatial gradient ▽(ΔT), is defined as: I leak (x)=β1×|ΔT max (x)| + β2×‖▽(ΔT)‖, where β1 and β2 are empirical weighting coefficients, and ΔT max (x) represents the maximum temperature difference;

[0017] When the leakage intensity index I leak When (x) exceeds the preset threshold, the location x is determined to be a leak point.

[0018] It should be noted that there is a natural temperature difference between the water in the canal and the surrounding soil / air (e.g., the canal water temperature is lower than the surface soil temperature in summer and higher than the surface soil temperature in winter). When leakage occurs in the canal, the temperature anomaly will be conducted to the surrounding soil, forming a "thermal anomaly patch" in the temperature profile measured by the distributed fiber optic temperature measurement system. By analyzing the amplitude and spatial gradient of this patch, the location and relative intensity of the leakage can be deduced.

[0019] Furthermore, in the leakage identification step, calculating the lateral leakage velocity and leakage flow rate based on the tracer concentration specifically includes:

[0020] Based on the peak arrival time t of the tracer concentration in the monitoring well peak The time difference between the start time t0 of the tracer introduction upstream of the channel and the distance L from the monitoring well to the center of the channel. well Calculate the lateral average velocity v lateral =L well / (t peak -t0);

[0021] Estimate the soil wetting cross-sectional area based on the spatial range of the thermal anomaly at the location of the leakage point. Where Ω represents the leakage-affected area, ε represents the temperature anomaly detection threshold, and 1 {·} It is an indicator function;

[0022] Calculate the leakage flow rate q leak (x)=v lateral (x)×A wet (x)×n, where n is the effective porosity of the soil.

[0023] It should be noted that after the tracer is pulsed and injected upstream of the channel, it migrates downstream with the water flow. The mass of tracer detected downstream of the main channel can reflect the visible loss of the main channel, while the peak arrival time of the tracer captured in the monitoring wells along the channel can be used to calculate the lateral leakage velocity.

[0024] Furthermore, in the water conveyance process monitoring step, an ensemble Kalman filter algorithm is used to assimilate the temperature data acquired by the distributed fiber optic temperature measurement system and the tracer concentration data acquired by the monitoring well, in order to correct the lateral leakage velocity and the leakage flow rate of each section of the channel.

[0025] The specific recursive process is as follows:

[0026] Prediction step: s {k|k-1} = f(s {k-1} ) + w {k-1}

[0027] Update step: s {k|k} = s {k|k-1} + K k · (y k - h(s {k|k-1} ))

[0028] Where s = [v, D, q] leak ] T Let v be the state vector, D be the dispersion coefficient, and q be the flow velocity. leak For leakage flow rate, yk Let f(·) be the observation vector (containing DTS temperature data and tracer concentration data), h(·) be the state transition function, and K be the observation operator. k Let be the Kalman gain matrix, and w be the process noise.

[0029] The EnKF algorithm is used to perform rolling corrections on the model state using real-time observation data, eliminating the systematic errors of a single method (pure thermal tracer or pure tracer).

[0030] Furthermore, in the water and fertilizer decision-making step, the spatiotemporal coupling model of water and fertilizer includes:

[0031] a) Channel water flow model

[0032] Constructed based on the Saint-Venant equations:

[0033] Continuity equation: A c / t + Q / x = -q leak (x)

[0034] Momentum equation: Q / t + / x(Q 2 / A c ) + gA c · h / x = gA c ·(S0 - S f )

[0035] Among them, A c Let Q be the channel surface area, h be the water depth, g be the acceleration due to gravity, S0 be the channel bottom slope, and S be the channel bottom slope. f The friction gradient is calculated using the Manning formula.

[0036] b) Field water and fertilizer response model

[0037] Field water balance equation: dh f / dt = P + I f - ET - D f - R f , where h f P is the depth of the water layer in the field, and I is the amount of precipitation. f ET represents the actual irrigation volume reaching the fields (calculated by deducting leakage losses from the water distribution volume at the canal head), ET represents the evapotranspiration, and D represents the evapotranspiration volume. fR represents the depth of leakage. f This refers to surface runoff.

[0038] Field nitrogen transport equation (one-dimensional convection-diffusion reaction equation): (θ·C N ) / t = / z(θ·D N · C N / z) - (q·C N ) / z - μ·C N + S N , where C N Let θ be the nitrogen concentration in the soil solution, θ be the soil volumetric water content, and D be the nitrogen concentration in the soil solution. N ρ is the hydrodynamic dispersion coefficient, q is the Darcy velocity, μ is the nitrogen conversion / absorption rate, and S is the nitrogen concentration. N This is a source of fertilizer.

[0039] Furthermore, in the water and fertilizer decision-making step, dynamically adjusting the water distribution volume upstream of the channel specifically includes:

[0040] To minimize the total leakage across all channels, a multi-objective optimization function is established: Where u is the opening degree of the upstream gate, L is the channel length, and h is the water depth;

[0041] The downstream water demand Q of the aforementioned channel down and the safe operation of the channel at water depth h max As a constraint, Q(L)≥Q down h min ≤h≤h max h min h max These are the lower and upper limits of the safe operating water depth, respectively.

[0042] Solve the multi-objective optimization function to obtain the optimal gate opening, so as to dynamically adjust the water distribution flow and water level.

[0043] The amount of leakage in a channel is positively correlated with the head pressure (water level). By appropriately lowering the water level in high leakage sections, leakage losses can be reduced at the source while ensuring the water demand downstream.

[0044] Furthermore, in the water and fertilizer decision-making step, the simultaneous adjustment of the fertilization time, fertilization amount, and irrigation system of the paddy field specifically includes:

[0045] When an increase in seepage flow is detected in a certain canal section, the irrigation amount of the corresponding irrigation area of ​​the canal section is reduced in the spatiotemporal coupling model of water and fertilizer, and the concentration of water-soluble fertilizer applied in the area is increased in order to maintain the nutrient concentration of the target root zone.

[0046] The fertilization time is adjusted according to the time lag of the leakage peak (e.g., advancing or delaying the fertilization time so that the fertilizer nutrient peak and the soil effective water peak can achieve spatiotemporal matching in the root zone).

[0047] The wet-dry alternation cycle in the irrigation system is dynamically adjusted according to the rate of change of the seepage flow (when seepage increases, the flooding period is shortened and the drying period is extended to reduce water loss and nutrient loss with water).

[0048] Furthermore, the tracer is an environmentally friendly fluorescent tracer (preferably modified sodium fluorescein, with the dosage controlled below the safe concentration for the water environment), and the tracer is introduced in a pulsed manner. The tracer concentration data in the monitoring well is obtained through an in-situ fluorescence sensor or an automatic water quality sampler.

[0049] Furthermore, the temperature-sensing optical cable in the distributed optical fiber temperature measurement system is buried longitudinally along the channel at a depth of 20 to 50 centimeters below the ground surface to monitor changes in the temperature field of the soil along the channel.

[0050] Furthermore, it also includes a visualization output step, which displays the location of the identified leakage point, the lateral leakage velocity, and the leakage flow rate of each section of the channel in the form of a heat map on a GIS map, and generates an early warning signal based on the leakage intensity level.

[0051] Beneficial effects

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] This invention utilizes the natural temperature difference between the water in the channel and the surrounding soil as a "thermal tracer" signal, eliminating the need for additional heating or cooling. Combined with the pulse injection of an environmentally friendly smart tracer, and through the coupling and inversion of the dual signals, it achieves for the first time the real-time and low-cost location and quantification of the leakage distribution along the entire water conveyance channel. This fundamentally solves the technical problems of traditional methods being unable to accurately locate the leakage location and quantify the leakage amount in real time.

[0054] This invention uses the real-time leakage distribution obtained through inversion as a dynamic constraint, and dynamically adjusts the water flow and water level at the canal head through an optimization algorithm, thereby achieving proactive control of leakage losses while ensuring downstream water demand.

[0055] This invention incorporates the leakage variable of water conveyance channels into a spatiotemporal coupling model of water and fertilizer, so that the deviation between the actual irrigation volume in the field and the decision irrigation volume can be dynamically corrected. With the linkage adjustment of fertilization time, fertilization amount and irrigation system, it can effectively solve the problem of "inaccurate field water volume due to water conveyance leakage, which leads to excessive fertilization and nutrient loss".

[0056] The control method of this invention employs two independent and mutually verifying methods: thermal tracing and tracer. Combined with data assimilation using an ensemble Kalman filter algorithm, it effectively improves the robustness of the system in complex field environments. It is applicable to water conveyance channels with different soil types and lining conditions, and requires minimal modification to existing irrigation infrastructure, thus possessing good prospects for widespread application. Attached Figure Description

[0057] Figure 1 This is a schematic diagram of the overall process of the present invention, which is a synergistic regulation method for improving fertility and reducing pollution in paddy fields based on the spatiotemporal coupling of water and fertilizer.

[0058] Figure 2 This is a schematic diagram of one possible deployment method of the distributed optical fiber temperature measurement system and monitoring well of the present invention.

[0059] Figure 3 This is a schematic diagram of the leakage location principle based on thermal tracing in this invention.

[0060] Figure 4 This is a schematic diagram illustrating the principle of lateral leakage velocity inversion based on tracer concentration-time curve in this invention. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0062] Implementation Method 1: System Deployment and Monitoring System Setup

[0063] Step 11, Obtaining the background temperature field

[0064] Background temperature field measurements were conducted in the target irrigation area when the water conveyance channels were not in operation. Distributed temperature-sensing optical cables were buried along the entire longitudinal direction of the channels, inside the embankment on one (or both) side of the channels, at a depth of 20–50 cm below the ground surface. The optical cables were buried in close contact with the backfill soil to ensure good heat exchange between the optical cables and the surrounding soil.

[0065] The optical cable is connected to a distributed optical fiber temperature measurement system (DTS). The spatial sampling interval is set to 0.5–1.0 meters, with a temperature sampling accuracy of no less than ±0.1℃, and the time sampling interval is 10–30 minutes. Continuous monitoring is conducted for at least 72 hours, collecting time-series temperature data from each measuring point under natural ground temperature conditions. After noise reduction and smoothing, the background temperature field distribution T along the canal is generated. b (x), where x is the longitudinal coordinate along the channel. The background temperature field includes the combined effects of natural factors such as solar radiation and geothermal gradient, and serves as a benchmark for subsequent leakage identification.

[0066] Step 12, Online monitoring of the distributed fiber optic temperature measurement system

[0067] After water is introduced into the canal, the DTS system continues to operate, collecting temperature data T(x,t) from various measuring points along the canal in real time, where t is a time variable. The data is transmitted to the central data processing unit in real time, and after background temperature field correction, a real-time temperature profile is generated.

[0068] In summer irrigation scenarios, the temperature of the canal water is usually lower than the surface soil temperature; in winter irrigation scenarios, the temperature of the canal water is usually higher than the surface soil temperature. This natural temperature difference constitutes the signal source for thermal tracing, eliminating the need for additional heating or cooling devices.

[0069] Step 13, Intelligent tracer dosing and monitoring

[0070] An automatic pulse dispensing device is installed at the channel inlet or a suitable upstream section. This device includes a tracer storage tank, a metering pump, and a spray pipe, and can inject tracer into the channel water in a pulse manner according to preset times and preset dosages.

[0071] The tracer selected is an environmentally friendly fluorescent tracer (such as modified sodium fluorescein), which has identifiable spectral characteristics (specific excitation and emission wavelengths) in water and a controllable half-life in the aquatic environment (from several hours to several days). Its degradation products are ecologically safe for aquatic organisms and farmland soil. The dosage is pre-estimated based on channel flow rate, channel length, and expected dilution factor to ensure detectability at downstream monitoring points and that the concentration is within the instrument's optimal detection range.

[0072] Automatic water quality samplers or in-situ fluorescence sensors are installed at the end section of the channel to capture the tracer's penetration curve (concentration-time process line) in the main channel. Shallow soil water monitoring wells are installed at regular intervals along the longitudinal direction of the channel (e.g., 300–500 meters, with the frequency adjusted based on channel length and expected variability in seepage space). The monitoring wells are located on one (or both) side of the channel, with a horizontal distance of 1–5 meters from the channel centerline (adjusted according to channel cross-sectional dimensions and soil conditions). The well depth should ideally penetrate the topsoil to reach the saturated layer (typically 1.0–2.0 meters). Each monitoring well is equipped with an automatic sampling probe or in-situ fluorescence sensor to record the tracer concentration changes over time in real time.

[0073] Implementation Method 2: Leakage Identification and Quantitative Inversion

[0074] Step 21, Leakage location based on thermal tracing

[0075] The real-time temperature field T(x,t) and the background temperature field T b (x) Taking the difference at each point, we obtain the temperature difference distribution ΔT(x,t)=T(x,t)-T b (x).

[0076] For any location x along the channel, calculate the spatial gradient of temperature difference ▽(ΔT). When there is leakage at a certain point, the leakage water with abnormal temperature enters the soil, forming a temperature anomalous patch at that point, which is characterized by a significant increase in the absolute value of ΔT and a clear spatial gradient change.

[0077] Define the leakage intensity index I leak (x) = β1·|ΔT max (x)| + β2·‖▽(ΔT)‖, where ΔT max (x) represents the peak temperature difference at location x within a set time window, and β1 and β2 are preset weighting coefficients (determined through offline calibration or empirical assignment; the calibration method is as follows: simultaneously measure ΔT and leakage on a known leakage point or a test section with artificially created leakage, and determine the optimal β value through regression analysis). I leak (x) is compared with a preset threshold ε (determined by multiplying the standard deviation of the background temperature field by a set factor, such as 3 times the standard deviation), when I leak When (x)>ε, it is determined that there is leakage at position x.

[0078] Output the spatial coordinates x of the leak point i (i=1,2,…,M, where M is the number of detected leaks), and the relative leakage intensity of each leak point is ranked.

[0079] Step 22, inversion of leakage flow rate based on tracer

[0080] For each monitoring well, extract the peak arrival time t from the tracer concentration-time profile. peak (That is, the time interval from the start time t0 of tracer injection to the concentration reaching its maximum value).

[0081] The lateral average flow velocity corresponding to this monitoring well is calculated using the following formula: v lateral = L well / (t peak -t0), where L well t0 represents the horizontal distance from the monitoring well to the centerline of the channel, and t0 represents the start time of tracer injection.

[0082] For the leak point x located by thermal tracing i Select the monitoring well that is spatially closest to the leakage point. lateral The value is used as an approximate value of the lateral leakage velocity at the leakage point.

[0083] Then, the soil wetting cross-sectional area A is estimated based on the spatial range of temperature anomalies indicated by thermal tracers. wet (x i The specific method is as follows: On the cross-section of the vertical channel, locate the boundary of the spatial region where the temperature anomaly ΔT is lower (in summer) or higher (in winter) than the background value by a certain threshold (e.g., |ΔT|>0.2℃). The cross-sectional area of ​​this region is A. wet This area estimation can be achieved through a pre-established temperature field-moisture content calibration relationship (the relationship between temperature anomaly amplitude and moisture content change is calibrated on the test section using a TDR soil moisture sensor).

[0084] Finally, the leakage flow rate per unit length of channel corresponding to this leakage point is: q leak (x i ) = v lateral (x i )·A wet (x i )·n, where n is the effective porosity of the soil at the location of the monitoring well (obtained through on-site sampling and laboratory measurement or by consulting a regional soil parameter database). For cases where monitoring wells are not deployed near the leakage point, the effective porosity of adjacent monitoring wells is used. lateral Spatial interpolation is performed on the values ​​(such as linear interpolation or inverse distance weighted interpolation).

[0085] Step 23, tracer mass conservation verification

[0086] To verify the reliability of the above quantitative results, a global verification was performed using the tracer mass conservation method.

[0087] The total amount of tracer injected upstream of the channel, M0, is known. The total amount of tracer detected at the end of the channel, M... down =∫C down(t)·Q down (t) dt (where C) down (t) is a function of the downstream cross-sectional tracer concentration as a function of time, Q down (t) represents the downstream cross-sectional flow rate), and the total tracer detected in each monitoring well is M. well = Σ∫C well,j (t)·V sample,j (t) dt (j is the monitoring well number, V) sample,j (The amount of seepage represented by the monitoring well).

[0088] According to the law of conservation of mass, M0 = M down + M wel + M loss (where M) loss This is for tracer attenuation loss; for tracers with a half-life much longer than the water delivery time, this can be ignored. If the above relationship deviation is within the set allowable range (e.g., ±15%), then it indicates that q at each leakage point... leak (x i The quantitative results are consistent; if the deviation exceeds the allowable range, it indicates that there are unidentified leakage points or insufficient monitoring density, and the monitoring wells need to be densified and remeasured.

[0089] Implementation Method 3: Multi-Source Data Assimilation

[0090] Using DTS temperature data and monitoring well tracer concentration data as observation inputs, the Ensemble Kalman Filter (EnKF) algorithm was employed for data assimilation to improve the accuracy and stability of leakage flow field estimation.

[0091] The specific implementation process is as follows:

[0092] Step 31, State Vector Definition: Discretize the channel along its length into N computational cross-sections, where the state vector for each cross-section is s = [v, D, q]. leak ] T Where v is the lateral velocity, D is the hydrodynamic dispersion coefficient, and q leak This represents the leakage flow rate. The overall channel state vector is the set of state vectors for each cross-section.

[0093] Step 32, Initial set generation: Based on the initial estimates obtained from the independent inversion of thermal tracer and tracer in the first step, a random perturbation conforming to a Gaussian distribution is superimposed (the standard deviation of the perturbation is 10% to 20% of the initial estimate) to generate an initial state set containing m members (m is generally 50 to 100).

[0094] Step 33, Prediction Step: For each set member, update the state over time using the state transition equation (based on a convection-diffusion physics model or an empirical recursive model) to obtain the predicted state s. {k|k-1} .

[0095] Step 34, Update Step: When new observation data y k When the temperature and tracer concentration data (including temperature and tracer concentration data) arrive, calculate the Kalman gain K. k The predicted state for each set member is corrected to obtain the analytical state s. {k|k} The corrected state is closer to the true value than the predicted state.

[0096] Step 35, Output Assimilation Results: Take the ensemble average of the analysis states of all ensemble members to obtain the assimilated leakage flow field q. leak (x), and simultaneously calculate the set variance as a measure of the estimated uncertainty.

[0097] Implementation Method 4: Construction and Dynamic Regulation of Spatiotemporal Coupling Model for Water and Fertilizer

[0098] Step 41, Optimization and Regulation of Channel Water Distribution

[0099] With the assimilated leakage flow field q leak (x) serves as the input boundary condition for the channel flow model.

[0100] The channel flow model is constructed based on the one-dimensional Saint-Venant equations. At each decision time t... k (e.g., daily or every few hours) collect real-time boundary conditions such as upstream flow and downstream water level of the channel, and use implicit difference schemes (e.g., Preissmann four-point implicit scheme) to numerically solve the Saint-Venant equations to simulate the distribution of water level h(x) and flow rate Q(x) along the channel.

[0101] Then, with the goal of minimizing the total leakage across all channels, a gate opening optimization problem is established:

[0102] Objective function:

[0103] Where u is the upstream gate opening (decision variable), and q leak It is a function of flow rate and water level.

[0104] Constraints:

[0105] 1) Downstream cross-sectional flow rate Q(L) ≥ Minimum downstream ecological and agricultural water demand Q down ;

[0106] 2) Water depth along the channel h(x) ≤ maximum safe water depth of the channel h max ;

[0107] 3) Water depth along the route h(x) ≥ minimum operating water depth of the channel h min .

[0108] Solve this single-objective constrained optimization problem (using standard optimization algorithms such as particle swarm optimization, genetic algorithm, or gradient descent) to obtain the optimal gate opening u. Send u to the gate control system for execution to achieve dynamic water distribution in the channel based on real-time leakage information.

[0109] Step 42, Field water and fertilizer linkage regulation

[0110] The results of the channel water distribution optimization were converted into the actual amount of water reaching the field in each irrigation unit (I). actual (x unit The specific calculation method is as follows: based on the leakage distribution q along the channel. leak (x) Calculate the leakage loss of each irrigation unit corresponding to the canal section, deduct it from the water distribution at the canal head, and then obtain the actual amount of water reaching the field.

[0111] Field water and fertilizer management should be adjusted according to the following logic:

[0112] 1) Irrigation volume adjustment: If the actual water volume reaching the field of a certain irrigation unit is I actual Below the target irrigation volume for this growth stage I target In this case, irrigation scheduling will be based on the actual amount of water arriving in the field, instead of mechanically implementing the original irrigation quota.

[0113] 2) Fertilizer Concentration Adjustment: For situations where the actual irrigation volume is reduced due to seepage, while maintaining the target nitrogen application rate (pure nitrogen input per unit area), the fertilizer concentration in the irrigation water should be increased accordingly to ensure that the total amount of nutrients entering the root zone with the water is not reduced. Specifically: Adjusted fertilizer concentration = Original fertilizer concentration × (I target / I actual This adjustment applies to water-soluble fertilizers applied with irrigation water.

[0114] 3) Adjustment of fertilization time: Adjust the time window for fertilization operations according to the time lag of the peak leakage during water conveyance. Specifically, for canal sections with large leakage, the peak water volume in the field arrives later than the water release time at the canal head. Correspondingly, the topdressing time in this irrigation area should be appropriately delayed so that the peak release of fertilizer nutrients coincides with the peak of effective soil moisture.

[0115] 4) Dynamic adjustment of irrigation system: based on the rate of change of seepage flow q leak / dt, dynamically adjust the field's wet-dry cycle. When the seepage rate is positive (seepage is increasing), shorten the flooding period and extend the drying period to reduce seepage losses driven by water head pressure; when the seepage rate is negative or stable, restore the regular wet-dry cycle.

[0116] Implementation Method 5: Visualized Output and Early Warning

[0117] The assimilated and corrected leakage flow rate q at each cross-section along the channel. leak (x) Rendered and displayed on a GIS map in the form of a heat map. Based on a preset leakage intensity grading standard, such as:

[0118] Slight leakage: q < 1×10 -5 m 3 / (s·m);

[0119] Moderate leakage: 1×10 -5 ≤ q < 5×10 -5 m 3 / (s·m);

[0120] Severe leakage: q ≥ 5×10 -5 m 3 / (s·m));

[0121] The leakage level of each canal section is marked with different colors.

[0122] When the leakage flow rate of a certain canal section exceeds the preset warning threshold, or the leakage flow rate change rate exceeds the preset change rate threshold, the system automatically generates a warning signal (including leakage location, leakage amount, and suggested treatment measures) and pushes it to the irrigation district management department.

[0123] After the warning is triggered, the system simultaneously sends a "leakage aggravation" event signal to the water and fertilizer decision module, triggering the aforementioned optimization model to be solved again, generating a new water allocation plan and field water and fertilizer adjustment plan.

[0124] Implementation Method 6: Operation Flow in Typical Scenarios

[0125] The following example, using a complete irrigation and fertilization cycle, illustrates the specific operation process of this invention:

[0126] Scenario: A certain irrigation district plans to implement a topdressing irrigation, with a target irrigation volume of 30m³ per mu. 3 Apply 10 kg of urea (46% nitrogen content) per mu as a top dressing.

[0127] Step S61: 12 hours before water is released from the canal head, turn on the DTS system and confirm that the background temperature field is stable (the deviation between the real-time temperature and the background temperature is within ±0.3℃).

[0128] Step S62: Simultaneously with water release from the canal head, the automatic pulse dispenser is activated, injecting fluorescent tracer at a preset amount. The DTS continuously monitors the temperature field along the canal. The automatic sampler in the monitoring well collects water samples at a set frequency (e.g., once every 10 minutes) and records the fluorescence intensity.

[0129] Step S63: The central data processing unit receives DTS temperature and fluorescence intensity data in real time. Two to four hours after water release (depending on channel length and water flow velocity), fluorescence peak signals appear successively in each monitoring well. The system automatically identifies temperature anomaly segments and, combined with the time difference of fluorescence signals, calculates the leakage q in each channel section. leak (x).

[0130] Step S64: q leak (x) Input the Saint-Venant equations model to calculate the actual seepage loss of each channel section under the current water level conditions. If the total seepage loss exceeds a preset threshold (e.g., 15% of the water discharge at the canal head), the optimized model suggests reducing the opening of the upstream gate. After confirmation by the dispatcher, the gate adjustment is executed.

[0131] Step S65: Based on the leakage distribution, estimate the actual water volume reaching the fields in each irrigation district. If the actual water volume in a certain irrigation district is more than 10% lower than the target value due to severe leakage in the corresponding canal section, the system suggests:

[0132] 1) The irrigation quota for this irrigation district has been reduced by 10% this time;

[0133] 2) The concentration of urea for topdressing was increased from the conventional 3.3 kg / mu water (pure) to 3.7 kg / mu water to maintain the same amount of nitrogen applied per unit area;

[0134] 3) The topdressing operation in this irrigation area is postponed to 6 to 8 hours after the canal is opened (i.e., the time window when the local field water volume reaches its peak).

[0135] Step S66: The dispatch instruction is issued to the field management unit for execution. After irrigation and fertilization are completed, the system generates a leakage report for this irrigation cycle, including a heat map of leakage distribution in each canal section, the total leakage amount, and the recommended priority for canal repair.

[0136] The above embodiments describe in detail the entire process of operation, from system deployment, signal acquisition, seepage inversion, data assimilation to water and fertilizer linkage regulation. Those skilled in the art should understand that the specific parameters in each of the above steps (such as monitoring well spacing, sampling frequency, ensemble number, optimization algorithm type, etc.) can be adaptively adjusted according to the actual irrigation area's channel length, soil conditions, hydrological characteristics, and control accuracy requirements, and all of these are within the scope of protection of this invention.

Claims

1. A method for synergistic regulation of fertilization improvement and pollution reduction in paddy fields based on spatiotemporal coupling of water and fertilizer, characterized in that, A distributed fiber optic temperature measurement system is deployed along the water conveyance channel to acquire the background and real-time temperature fields of the soil along the channel. A tracer is introduced upstream of the channel, and the tracer concentration is monitored downstream and in soil water monitoring wells along the channel. Based on the temperature difference between the real-time temperature field and the background temperature field, the location of seepage points along the channel is identified. Based on the tracer concentration, the lateral seepage velocity and the seepage flow rate of each section of the channel are calculated. The identified seepage point locations, lateral seepage velocities, and seepage flow rates of each section of the channel are used as constraints and input into a pre-constructed water-fertilizer spatiotemporal coupling model to dynamically adjust the water distribution upstream of the channel and simultaneously adjust the fertilization time, fertilization amount, and irrigation regime of the paddy fields.

2. The method for coordinated regulation of paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer as described in claim 1, characterized in that, Identifying the location of the leak point based on the temperature difference between the real-time temperature field and the background temperature field specifically includes: Calculate the spatial gradient ▽(ΔT) of the temperature difference ΔT; Define the leakage intensity index I leak (x), which is positively correlated with the magnitude of the temperature difference ΔT and the spatial gradient ▽(ΔT); When the leakage intensity index I leak When (x) exceeds the preset threshold, the location x is determined to be a leak point.

3. The method for coordinated regulation of paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer as described in claim 1, characterized in that, The calculation of the lateral leakage velocity and the leakage flow rate of each section of the channel based on the tracer concentration specifically includes: Based on the peak arrival time t of the tracer concentration in the monitoring well peak The time difference between the start time t0 of the tracer introduction upstream of the channel and the distance L from the monitoring well to the center of the channel. well Calculate the lateral average velocity v lateral =L well / (t peak -t0); Based on the spatial range of the thermal anomaly at the location of the leakage point, estimate the soil wetting cross-sectional area A. wet ; Calculate the leakage flow rate q leak (x)=v lateral (x)×A wet (x)×n, where n is the effective porosity of the soil.

4. The method for coordinated regulation of paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer as described in claim 1, characterized in that, An ensemble Kalman filter algorithm is used to assimilate the temperature data acquired by the distributed fiber optic temperature measurement system and the tracer concentration data acquired by the monitoring well, in order to correct the lateral leakage velocity and the leakage flow rate of each section of the channel.

5. The method for coordinated regulation of paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer as described in claim 1, characterized in that, The spatiotemporal coupling model of water and fertilizer includes a channel flow model based on the Saint-Venant equations, whose continuity equation is: In the formula, A c Let Q be the channel's cross-sectional area and Q be the channel's flow rate.

6. The method for coordinated regulation of paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer as described in claim 5, characterized in that, The dynamic adjustment of water distribution upstream of the channel specifically includes: To minimize the total leakage across all channels, a multi-objective optimization function is established: Where u is the opening degree of the upstream gate, L is the channel length, and h is the water depth; The downstream water demand Q of the aforementioned channel down and the safe operation of the channel at water depth h max As a constraint, Q(L)≥Q down h min ≤h≤h max h min h max These are the lower and upper limits of the safe operating water depth, respectively. Solve the multi-objective optimization function to obtain the optimal gate opening, so as to dynamically adjust the water distribution flow and water level.

7. The method for coordinated regulation of paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer as described in claim 1, characterized in that, The tracer is an environmentally friendly fluorescent tracer; the tracer is introduced in a pulsed manner; the tracer concentration data in the monitoring well is obtained through an in-situ fluorescence sensor or an automatic water quality sampler.

8. The method for coordinated regulation of paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer as described in claim 1, characterized in that, The temperature-sensing optical cable in the distributed optical fiber temperature measurement system is buried longitudinally along the channel at a depth of 20 to 50 centimeters below the ground surface.

9. The method for coordinated regulation of paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer as described in claim 1, characterized in that, The synchronized adjustment of fertilization time, fertilization amount, and irrigation system in paddy fields specifically includes: When an increase in seepage flow is detected in a certain canal section, the irrigation amount of the corresponding irrigation area in the water-fertilizer spatiotemporal coupling model is reduced, and the concentration of water-soluble fertilizer applied in that area is increased to maintain the nutrient concentration in the target root zone. The fertilization time is adjusted according to the time lag of the leakage peak, and the dry-wet alternation cycle in the irrigation system is dynamically adjusted according to the rate of change of the leakage flow.

10. The method for coordinated regulation of paddy field fertilization and pollution reduction based on spatiotemporal coupling of water and fertilizer as described in claim 1, characterized in that, It also includes a visualization output step, which displays the location of the identified leakage point, the lateral leakage velocity, and the leakage flow rate of each section of the channel in the form of a heat map on a GIS map, and generates an early warning signal based on the leakage intensity level.

Images