A farmland water level monitoring and early warning method and system

Abstract

The present application relates to the field of data processing, more particularly, the present application relates to a kind of farmland water level monitoring and early warning method and system, method includes: the spatial mapping relationship of establishment measuring point and field block, rolling calculation early rainfall index and divide runoff inversion state interval;Collect channel water level and rainfall data, carry out scheduling interference determination, calculate water level-rain intensity response coefficient under rainfall dominant working condition to correct baseline inversion coefficient, calculate effective runoff intensity;Effective runoff intensity is deducted after time-varying soil infiltration rate and obtains net water storage rate;According to the real-time water level comparison relationship of field block and channel, in the external discharge scene to drain damping factor reduction, in lateral seepage supply scene increase supply increment, to net water storage rate iterative accumulation and obtain cumulative water storage amount;Conversion into field block water depth after comparison with threshold value, output graded early warning signal.The present application dynamically calculates the water depth of each field block under the condition of no water level sensor in field block, improves early warning accuracy.

Description

Technical Field

[0001] This invention relates to the field of data processing. More specifically, this invention relates to a method and system for monitoring and early warning of farmland water levels. Background Technology

[0002] In modern agricultural production, monitoring farmland waterlogging is crucial for ensuring crop growth and field management. Although some irrigation districts have been equipped with canal water level monitoring and rainfall observation facilities, the interior of fields is usually not used as a regular location for water level sensors due to frequent field disturbances, high equipment maintenance costs, and the economic limitations of large-scale deployment.

[0003] Because there is a lack of direct means of sensing water levels within fields, existing monitoring and early warning methods rely solely on indirect assessments based on rainfall observations. These methods typically set fixed rainfall thresholds, issuing a flood warning once the measured rainfall exceeds this threshold. However, factors such as dynamic changes in soil infiltration capacity with alternating wet and dry periods, fluctuations in surrounding water levels due to canal system scheduling, and differences in local drainage conditions can all lead to significant variations in the actual water depth in fields corresponding to the same rainfall. Therefore, relying solely on rainfall as a single indicator is insufficient to accurately reflect the waterlogging situation in fields, often resulting in false alarms or missed warnings, failing to meet the practical needs of precision agriculture for disaster response. Summary of the Invention

[0004] To address the aforementioned technical problem of the inability to accurately detect the water accumulation status of each field due to the lack of water level sensors inside the fields, the present invention provides solutions in the following aspects.

[0005] In the first aspect, a method for monitoring and early warning of farmland water levels includes: Establish a spatial mapping relationship between measuring points and fields within the target irrigation area, determine the channel water level measuring point to which each field belongs, and calculate the previous rainfall index on a rolling basis; Based on the previous rainfall index, the runoff inversion state intervals are divided, including drought intervals, transition intervals, and wet intervals. The drought interval corresponds to the drought baseline inversion coefficient, the transition interval corresponds to the transition baseline inversion coefficient, and the wet interval corresponds to the wet baseline inversion coefficient. Collect channel water level data and rainfall intensity data, determine the scheduling interference of channel water level data, and when it is determined to be the dominant rainfall condition, calculate the water level-rainfall intensity response coefficient by combining the channel water level rise and fall rate with the synchronously aligned rainfall intensity parameter. The baseline inversion coefficient is corrected based on the water level-rainfall intensity response coefficient to obtain the actual runoff inversion coefficient; the effective runoff intensity is calculated based on the actual runoff inversion coefficient and the synchronously aligned rainfall intensity parameter; the time-varying soil infiltration rate is calculated based on the previous rainfall index, and the effective runoff intensity is subtracted from the time-varying soil infiltration rate and a non-negative value is taken to obtain the net water accumulation rate; Based on the real-time water level comparison between the field and the canal, the water accumulation growth is reduced by the drainage damping factor in the external drainage scenario, and the side seepage replenishment increment is increased in the side seepage replenishment scenario. The net water accumulation rate is iteratively accumulated to obtain the cumulative water accumulation. The accumulated water volume is converted into the water depth of the field, and a graded early warning signal is output based on the comparison result of the water depth of the field with the threshold.

[0006] Optionally, establishing the spatial mapping relationship between measuring points and fields within the target irrigation area includes: The measuring points at the head of the branch canal cover all the fields within the catchment area of ​​the branch canal under its control, while the measuring points at the inlet of the agricultural canal cover the set of fields that have a direct hydraulic connection with the agricultural canal. Using the field canal network and field boundaries as spatial boundaries, and combining field ground elevation data, the catchment areas of each measuring point are divided. Fields whose ground elevation is lower than the design control water level of their respective canals are included in the catchment area of ​​the corresponding measuring point of that canal. The fields located near the boundary of the catchment area of ​​two adjacent measuring points are assigned to the measuring point corresponding to the side with the lower ground elevation. Fields without water level monitoring points are assigned to the nearest monitoring point at the entrance of the irrigation canal that is hydraulically connected. If there is no irrigation canal entrance monitoring point nearby, they are assigned to the monitoring point at the head of the branch canal.

[0007] Optionally, the rolling calculation of the previous rainfall index includes: the remaining value of the previous rainfall index after exponential decay at the previous rainfall sampling time plus the measured rainfall between the previous rainfall sampling time and the current rainfall sampling time to obtain the previous rainfall index at the current rainfall sampling time.

[0008] Optionally, the method for determining channel water level data to prevent scheduling interference includes: comparing real-time channel water level data with historically statistical baseline data of normal channel water level fluctuations; determining that a gate pump scheduling event exists when the absolute value of water level fluctuations in adjacent time periods exceeds a preset threshold; the channel water level rise and fall rate at the current water level sampling time does not participate in runoff generation inversion; and using the value of the runoff generation inversion coefficient from the previous available water level sampling time. When the absolute value of water level fluctuations in adjacent time periods does not exceed the preset threshold, it is determined to be a rainfall-dominated operating condition.

[0009] Optionally, before calculating the water level-rainfall response coefficient using the channel water level rise and fall rate and the synchronously aligned rainfall intensity parameters, a time alignment step is also included: for the water level sampling time of the channel water level measuring point, take the two adjacent rainfall sampling times before and after the water level sampling time and their corresponding measured rainfall intensity values ​​and perform linear interpolation to obtain the synchronously aligned rainfall intensity parameters for the water level sampling time.

[0010] Optionally, correcting the baseline inversion coefficients based on the water level-rainfall intensity response coefficients includes: mapping the water level-rainfall intensity response coefficients to correction weights, using the correction weights as interpolation coefficients, and performing linear interpolation between the baseline inversion coefficients and the baseline inversion coefficients of adjacent humid intervals to obtain the actual runoff inversion coefficients; if the current interval is a humid interval, then the humid baseline inversion coefficients are directly used as the actual runoff inversion coefficients.

[0011] Optionally, the time-varying soil infiltration rate is calculated based on the previous rainfall index, wherein the time-varying soil infiltration rate is negatively correlated with the previous rainfall index, and the lower limit of the time-varying soil infiltration rate is the infiltration rate when the soil is saturated, and the upper limit is the infiltration rate when the soil is dry.

[0012] Optionally, the iterative accumulation of the net water accumulation rate based on the real-time water level comparison between the field and the channel includes: when the water depth in the field is not lower than the water level in the channel, it is determined to be an external discharge scenario, and the net water accumulation rate is accumulated after being reduced by the drainage damping factor; when the water level in the channel is higher than the water depth in the field but not lower than the lowest ground elevation of the field, it is determined to be a lateral seepage recharge scenario, and the lateral seepage recharge increment is added to the net water accumulation rate before accumulation.

[0013] Optionally, converting the cumulative water volume into the water depth of the field includes: converting the cumulative water volume into the water depth of the field based on the field area and the field's elevation-volume curve.

[0014] In a second aspect, a farmland water level monitoring and early warning system includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the farmland water level monitoring and early warning method described in any one of the claims is implemented.

[0015] The present invention has the following effects: 1. This invention addresses the practical constraint of not being able to deploy water level sensors on a large scale within fields. Based on existing water level monitoring and rainfall observation facilities in irrigation districts, it achieves coverage and localization of monitoring signals to all fields through spatial mapping of measuring points to fields. On this basis, it sequentially completes signal separation of scheduling interference and rainfall response, adaptive correction of runoff inversion coefficients with measured channel response, dynamic attenuation of soil infiltration capacity with dry and wet conditions, and scenario-based modeling of bidirectional exchange between drainage and lateral infiltration. This forms a complete calculation link from the original monitoring signal to the output of field water depth. Without adding field hardware, it achieves hourly perception and graded early warning of water accumulation in all fields of the irrigation district, meeting the actual needs of precision agriculture for disaster response with lightweight deployment costs.

[0016] 2. This invention effectively eliminates the disturbance to the water level signal caused by human scheduling factors such as the opening and closing of gate pumps by judging the scheduling interference of channel water level data, so that the runoff inversion coefficient only responds to the actual rainfall process, avoiding false alarms caused by scheduling events being misjudged as strong runoff.

[0017] 3. This invention constructs a time-varying soil infiltration rate that dynamically decreases with the previous rainfall index, so that the net water accumulation process after deducting infiltration from runoff can be adaptively adjusted according to the soil's dryness and wetness. This effectively overcomes the estimation bias of fixed infiltration rate under two extreme scenarios: the first rainfall after drought and continuous rainfall after saturation.

[0018] 4. Based on the real-time water level comparison between fields and channels, this invention models two hydraulic exchange processes with opposite directions: external discharge and lateral seepage replenishment. This makes external discharge suppress dispersion and lateral seepage replenishment compensate for underreporting during the water accumulation process, and the water depth estimation is closer to the real physical process. Attached Figure Description

[0019] Figure 1 This is a flowchart of steps S1-S3 in a farmland water level monitoring and early warning method according to an embodiment of the present invention. Detailed Implementation

[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0021] This invention is applicable to irrigation areas with dense canal networks, where canal level gauges and field rain gauges have been installed, but water level sensors are not installed within the fields. Its application scenario is as follows: in large-scale irrigation areas, field work is frequent, equipment maintenance is difficult, and it is difficult to stably deploy water level sensors within the fields for extended periods. Therefore, it is necessary to rely on existing canal level monitoring and rainfall observation facilities to achieve dynamic calculation and tiered early warning of water depth for all fields without adding additional field sensing equipment.

[0022] Reference Figure 1 A method for monitoring and early warning of farmland water levels includes steps S1-S3, as detailed below: S1: Establish the spatial mapping relationship between measuring points and fields within the target irrigation area and calculate the previous rainfall index on a rolling basis.

[0023] Since no water level sensors are installed inside the fields, the estimation of water depth can only rely on data provided by canal water level measuring points and rainfall measuring points. Therefore, a fundamental problem needs to be addressed first: which fields do the data from the canal water level measuring points correspond to? Without establishing the spatial relationship between measuring points and fields, the estimation results cannot be applied to specific fields. Therefore, the primary task of this step is to clarify the field area served by each canal water level measuring point, and the data allocation method for fields without measuring points.

[0024] To achieve the above objectives, the entire irrigation area covered by this monitoring and early warning system is first defined as the target irrigation district. Let there be a total of [number missing] irrigated areas within this target irrigation district. There are 1 water level monitoring points, among which the monitoring point at the head of the canal is recorded as... The measuring point at the entrance of the irrigation canal is recorded as... The target irrigation area has a total of There are several fields. The spatial mapping relationship between the measuring points and the fields is established through an correlation matrix. Indicates: If the field Belonging to measuring points ,but Otherwise, the value is 0. The sampling period for channel water level gauges is 600 seconds, and the sampling period for rain gauges is 300 seconds. The following basic attributes are recorded for each field: field area, soil texture type, field elevation-volume curve, lowest ground elevation of the field, and crop type and growth stage.

[0025] The field elevation-volume curve and the lowest ground elevation can both be obtained from existing digital elevation model (DEM) data of the target irrigation area. The lowest ground elevation of a field is directly determined by the smallest raster value of the DEM within that field area. The field elevation-volume curve uses water level as a variable, statistically analyzing the water storage corresponding to all elevation grids above that water level within the field at different water levels, thereby establishing the correspondence between water level and volume. This can be directly generated using geographic information system tools such as ArcGIS. For irrigation areas without DEM data, it can be obtained through field surveys or topographic map digitization.

[0026] The establishment of mapping relationships follows these four rules: Rule 1: Hierarchical Attribution of Measuring Points. Measuring points at the head of a branch canal cover all fields within its catchment area. Measuring points at the inlet of a farm canal cover the set of fields directly hydraulically connected to that farm canal. For branch canals without inlet measuring points, all fields controlled by that branch are uniformly assigned to the measuring points at the head of the branch canal.

[0027] Rule 2, Catchment Area Delineation. The catchment area of ​​each measuring point is delineated using the field canal network and field boundaries as spatial boundaries, combined with the field ground elevation data. Any field whose ground elevation is lower than the design control water level of its corresponding canal is considered to have a drainage outlet with that canal and is included in the catchment area of ​​the corresponding measuring point of that canal.

[0028] Rule 3, Assignment of Boundary Fields. When a field is located near the boundary of the catchment area of ​​two adjacent measuring points, it is assigned to the measuring point corresponding to the side with the lower ground elevation, because that side is more prone to water accumulation under the same rainfall conditions and should be given priority in obtaining water level monitoring information.

[0029] Rule 4, Spatial Expansion. For fields within the target irrigation area where no water level monitoring points have been established, they should first be assigned to the nearest monitoring point at the entrance of the irrigation canal with hydraulic connection, according to the rules mentioned above. If there is no irrigation canal entrance monitoring point nearby, they should be assigned to the monitoring point at the head of the branch canal. After the above mapping relationship is established, all fields within the target irrigation area can obtain data input from at least one channel measuring point, achieving full spatial coverage under sparse measuring point conditions. This mapping relationship ultimately forms a table of association between measuring points and fields, serving as a spatial index for estimating the water depth of fields. This table is updated according to the above rules when the channel layout or field boundaries change, and remains stable during daily operation. The channel water level data of the same measuring point will serve as a shared runoff response signal for all fields associated with that measuring point. However, each field independently calculates its own water depth based on its area, soil texture, elevation-volume curve, and crop attributes. Thus, the spatial mapping relationship between measuring point data and fields is established, and the water depth estimation for each field has a clear data source.

[0030] However, spatial mapping alone is insufficient to accurately estimate the depth of water accumulation in fields. Another key issue is that the soil's moisture state directly affects the proportion of rainfall runoff; during droughts, the soil's water absorption capacity is high, resulting in less water accumulation for the same amount of rainfall; conversely, the opposite is true when the soil is saturated. Using a fixed runoff coefficient cannot characterize this time-varying characteristic, inevitably leading to estimation errors. Therefore, it is necessary to construct a state quantity that can represent the change of soil moisture state over time, namely the antecedent rainfall index. The specific operation is as follows; Rainfall data is collected in real time using tipping bucket rain gauges deployed in open areas of the fields. Each tipping bucket rain gauge records a pulse signal with each flip. The average rainfall intensity within the sampling period is calculated by the number of flips per unit time, thus obtaining a rainfall intensity time series. The previous rainfall index is then calculated recursively. The previous rainfall index at the current rainfall sampling time equals the residual value of the previous rainfall index at the previous rainfall sampling time after exponential decay, plus the measured rainfall within the current sampling period. The specific recursive calculation formula is as follows:

[0031] In the formula, Rainfall sampling time The preceding rainfall index represents the accumulated amount of soil moisture status memory; This represents the previous rainfall index at the time of the previous rainfall sampling. The sampling period for the rain gauge is set to 300 seconds. It is an exponential function with the natural constant e as its base; The calibration method for the rainfall decay coefficient is as follows: collect rainfall time series and soil moisture observation data for at least one hydrological year in the irrigation area, take maximizing the correlation coefficient between the API series and the measured soil moisture content series as the optimization objective, and determine the optimal value by one-dimensional search within the preset parameter range; or, in accordance with the principle of minimizing the root mean square error between the estimated water depth and the measured water depth, calibrate it in conjunction with relevant parameters. The current rainfall sampling time The measured rainfall intensity comes from a tipping bucket rain gauge. This indicates the rainfall amount during the current sampling period.

[0032] During the calculation, the rainfall data from the 5-minute sampling cycle needs to be recursively extrapolated periodically according to the timestamp to ensure that the previous rainfall index sequence and the channel water level data have a time alignment relationship. Through the above recursive calculation, the previous rainfall index sequence can be obtained for each time step. This index quantitatively characterizes the dynamic changes in soil moisture state, providing a quantitative basis for adjusting the runoff inversion parameters according to the actual soil conditions.

[0033] In summary, S1 accomplished two fundamental tasks: first, it established a spatial mapping relationship between channel monitoring points and field plots, ensuring a clear data source for each plot; second, it quantitatively characterized the time-varying characteristics of soil moisture state through rolling decay calculations of the previous rainfall index. The completion of these tasks provides a spatial index and state input for estimating field water depth.

[0034] S2: Calculate the effective runoff intensity of each measuring point's catchment area based on the previous rainfall index.

[0035] After obtaining the antecedent rainfall index series, although the index quantitatively characterizes the soil's moisture state, it is only a memory accumulation and cannot be directly used to calculate the amount of rainfall converted into runoff. Therefore, it needs to be transformed into a parameter that can be directly used in runoff estimation. The proportion of rainfall converted into runoff varies significantly depending on the soil's moisture state; during drought, the soil has a strong water absorption capacity and a low runoff proportion; during wet conditions, the soil is close to saturation and the runoff proportion is high.

[0036] To characterize this difference, the runoff generation inversion state is first divided into three intervals—arid, transitional, and moist—based on the magnitude of the previous rainfall index, corresponding to high, medium, and low soil infiltration capacity, respectively. The interval boundaries are defined using the regression relationship between historical soil moisture content and the previous rainfall index: multi-year local soil moisture content observation data and corresponding previous rainfall index sequences are collected in advance, and the regression relationship between the two is fitted. Then, two boundary values ​​are determined according to the quantile of soil moisture content. For example, the API value corresponding to the 25th percentile of soil moisture content is used as the boundary between the arid and transitional intervals, and the API value corresponding to the 75th percentile is used as the boundary between the transitional and moist intervals. Each interval uses its own baseline inversion coefficient: the arid interval corresponds to the arid baseline inversion coefficient, the transitional interval corresponds to the transitional baseline inversion coefficient, and the moist interval corresponds to the moist baseline inversion coefficient. As a preferred example, the arid baseline inversion coefficient, transitional baseline inversion coefficient, and moist baseline inversion coefficient can be set to 0.25, 0.55, and 0.80, respectively. These baseline values ​​are derived from runoff statistical analysis of historical rainfall events and represent the average level of runoff production under various soil conditions. In actual use, they need to be recalibrated based on historical hydrological data of the irrigation area.

[0037] However, the baseline inversion coefficients only represent the average runoff level under various soil conditions, while the measured channel water level data also contains the actual runoff response information of the current catchment area. But the channel water level signal is not a purely natural hydrological signal. In well-irrigated areas, the channel water level is simultaneously affected by rainfall runoff and human-induced scheduling operations, including gate opening and closing, and pumping. If the water level fluctuation rate is directly used for runoff inversion, sudden water level changes caused by gate and pump scheduling will be misjudged as strong runoff events; conversely, if water level information is completely ignored, the true runoff response of the catchment area cannot be reflected. Therefore, the core of this step is to remove scheduling interference from the channel water level signal, extract effective runoff response information, and use this to correct the theoretical baseline coefficients, ultimately calculating the effective runoff intensity.

[0038] To achieve the above objectives, the channel water level data is first assessed for scheduling interference. The real-time channel water level data is then compared with the historical baseline of normal channel water level fluctuations. The standard deviation of channel water levels during periods without rainfall or scheduling is statistically analyzed, and a preset threshold is set at 2 to 3 times this standard deviation. When the absolute value of the water level fluctuation within an adjacent 10-minute interval exceeds this preset threshold, a gate / pump scheduling event is identified. In this case, the channel water level fluctuation rate for the current cycle is not included in the runoff generation inversion, and the runoff generation inversion coefficient uses the value from the previous available cycle. When the absolute value of the water level fluctuation within an adjacent 10-minute interval does not exceed this preset threshold, the event is identified as a rainfall-dominated condition, and subsequent correction calculations are performed. It should be noted that during the cycle identified as a scheduling event, although the channel water level fluctuation rate is not included in the correction of the runoff generation inversion coefficient, the effective runoff generation intensity is still calculated as the product of the runoff generation inversion coefficient from the previous available cycle and the current synchronous rainfall intensity. This allows for continued tracking of the rainfall-runoff generation process without being affected by scheduling disturbances.

[0039] Because the sampling periods of channel water level gauges and tipping bucket rain gauges are different (600 seconds and 300 seconds respectively), and their sampling times are independent and do not overlap, the sampling time of any given channel water level measuring point usually does not correspond to rainfall intensity data collected simultaneously. To obtain synchronous rainfall intensity information at a given water level sampling time, it is necessary to interpolate using two adjacent rainfall sampling times before and after the water level sampling time and their corresponding measured rainfall intensity values. For example, if the water level sampling time of a measuring point is 00:10, and its two adjacent rainfall sampling times are 00:08 and 00:13, both of which have measured rainfall intensity records, then linear interpolation using the measured values ​​from these two points can yield the synchronous rainfall intensity at the measuring point at 00:10.

[0040] Specifically, for the water level sampling time of any channel water level measuring point Take two adjacent rainfall sampling times and their corresponding rainfall intensities, and calculate the water level at that sampling point in a linear manner. The synchronous alignment rainfall intensity parameter, expressed by the following formula:

[0041] In the formula, The time of water level sampling at the measuring point Synchronous alignment of rainfall intensity parameters; , These are the water level sampling times. The rainfall sampling times are adjacent and satisfy the following conditions. ; , These represent the actual rainfall intensity measured at the sampling times of the two rain gauges. This interpolation process is executed periodically on the embedded monitoring terminal according to the timestamp, ensuring that all rainfall data involved in the calculation have a strict time alignment with the channel water level data.

[0042] After time alignment is completed, the channel water level rise and fall rate at the measuring point is calculated under the rainfall-dominant condition. Then, based on the channel water level rise and fall rate and the synchronously aligned rainfall intensity parameters obtained above, the water level sampling time at the measuring point is calculated. The water level-rainfall intensity response coefficient, expressed by the following formula:

[0043] In the formula, The time of water level sampling at this measuring point The water level-rainfall intensity response coefficient is dimensionless and characterizes the degree of water level response caused by a unit rainfall intensity at the measuring point. The larger the value, the more sensitive the current runoff generation in the catchment area is. The rate of rise and fall of the channel water level at this measuring point is obtained by dividing the difference between two consecutive sampling values ​​of the channel water level gauge at this measuring point by the sampling period of the channel water level gauge, which is 600 seconds. The time of water level sampling at the measuring point Synchronous alignment of rainfall intensity parameters, This is a stabilizing term with a value of 0.01. Its purpose is to prevent the denominator from approaching zero when the rainfall intensity is close to zero, which could lead to calculation overflow.

[0044] The water level sampling time at this measuring point was obtained. After determining the water level-rainfall intensity response coefficient, it is further mapped to a corrected weight, which can be expressed as:

[0045] In the formula, The time of water level sampling at this measuring point The correction weight, dimensionless and limited to between 0 and 1, is used to control the gradual position of the runoff inversion coefficient of the measuring point within its state range, reflecting the degree to which the actual response of the channel water level deviates from the baseline level. The time of water level sampling at this measuring point Water level-rainfall intensity response coefficient, To calibrate the lower limit of the response ratio, The upper limit of the response ratio is defined as the lower and upper bounds of the normal fluctuation range of the water level-rainfall intensity response ratio coefficient. and The calibration method is as follows: Extract samples of the ratio of water level rise and fall rates to synchronously aligned rainfall intensity parameters under the dominant rainfall conditions from historical channel water level data, and take the 5th percentile of this sample set as... Take the 95th percentile as .

[0046] Furthermore, based on the runoff inversion state interval to which the measuring point belongs at the water level sampling time, the corresponding baseline inversion coefficient is selected. Then, using the corrected weight as the interpolation coefficient, linear interpolation is performed between this baseline inversion coefficient and the baseline inversion coefficient of the adjacent slightly humid interval to obtain the actual runoff inversion coefficient used at the measuring point at the water level sampling time. If the current interval is a humid interval, the baseline inversion coefficient corresponding to this interval is directly used as the actual runoff inversion coefficient. The physical effect of this approach is to correct the theoretical runoff coefficient with the actual response of the channel water level, so that the runoff inversion coefficient changes continuously and gradually rather than abruptly when switching state intervals, thereby avoiding step-like errors in runoff estimation under alternating soil wet and dry scenarios.

[0047] Finally, the corrected runoff inversion coefficient for this measuring point is multiplied by the synchronously aligned rainfall intensity parameter for that measuring point to obtain the effective runoff intensity of the current catchment area at that measuring point. This effective runoff intensity characterizes the water volume that can actually form runoff in the catchment area of ​​this measuring point per unit time, comprehensively reflecting three types of information: soil moisture status, channel water level response, and real-time rainfall intensity. It serves as the runoff input for subsequent field water accumulation calculations in the catchment area of ​​this measuring point.

[0048] In summary, the above-mentioned S2 method first divides the runoff generation interval based on the previous rainfall index and sets the baseline inversion coefficient. Then, it eliminates disturbances to the water level signal at the measuring point caused by the opening and closing of gate pumps through scheduling interference judgment. Time alignment ensures the synchronization of the water level at the measuring point with the rainfall data. Next, it calculates the water level-rainfall intensity response coefficient of the measuring point and maps it to a correction weight. Adaptive interpolation is then performed on the baseline inversion coefficient of the measuring point to finally deduce the effective runoff intensity of the catchment area at that measuring point. The completion of the above work provides a reliable runoff input for the cumulative calculation of field waterlogging depth.

[0049] S3: Based on the effective runoff intensity, the net water volume of the catchment area of ​​each measuring point is iteratively accumulated, and then converted into the water depth of each field. After threshold comparison, a graded early warning is output.

[0050] After obtaining the effective runoff intensity of each measuring point's catchment area, it needs to be further converted into the actual water depth of each field and an early warning signal needs to be output. However, if the difference between the effective runoff intensity and the infiltration rate is directly integrated over a long period of time, the accumulated water volume will continue to disperse over time. Secondly, after the fields are filled with water, there are two opposing hydraulic exchange processes between the fields and the channel water: outward discharge and lateral seepage replenishment. The physical mechanisms of these two processes are completely different and must be modeled separately.

[0051] Before proceeding, it's important to clarify that of the effective runoff reaching the field surface, a portion infiltrates the soil, while the remainder forms surface water. Existing technologies often employ a fixed infiltration rate, meaning the infiltration capacity is constant regardless of soil moisture. This approach significantly overestimates water accumulation during the first rainfall after a prolonged drought, and significantly underestimates it during continuous rainfall after soil saturation, leading to frequent false alarms and missed alarms. Therefore, this invention does not use a fixed infiltration rate but instead constructs a dynamically changing soil infiltration rate that varies with soil moisture levels, ensuring that the difference between the effective runoff intensity and the infiltration rate always closely approximates the actual physical process.

[0052] Specifically, the aforementioned time-varying infiltration rate can be expressed by the following formula:

[0053] In the formula, The time of water level sampling at this measuring point Soil infiltration rate, , These represent the infiltration rates when the soil is saturated and dry, respectively, and are determined by soil texture. It is an exponential function with the natural constant e as its base. The infiltration attenuation index, Water level sampling time The preceding rainfall index.

[0054] A higher prior rainfall index indicates wetter soil and a closer soil infiltration rate. The lower the current rainfall index, the drier the soil, and the closer the soil infiltration rate is to... Therefore, during the first rainfall after a period of drought, the soil's infiltration capacity is strong, and runoff is significantly reduced, avoiding false alarms; when the soil is saturated and rainfall continues, the infiltration capacity drops to its lowest level, and runoff is fully converted into water accumulation, avoiding underreporting.

[0055] in, and The value of varies depending on the soil texture; as a preferred example, for clay, is . , ; Soil , . The example value is 0.015. All of the above parameters can be optimized during the calibration of historical hydrological data according to the principle of minimizing the error in water depth estimation.

[0056] Furthermore, the effective runoff intensity calculated from S2 above is subtracted from the soil infiltration rate, and a non-negative value is taken. This is equivalent to calculating the difference between the effective runoff intensity and the soil infiltration rate, and taking the maximum value between this difference and 0 to obtain the net water accumulation rate of each measuring point's catchment area. Taking a non-negative value indicates that when the infiltration capacity is greater than the runoff intensity, all rainwater infiltrates, and there is no increase in water accumulation in the field. This achieves real-time balance calculation of the two competing processes of runoff generation and infiltration, and the net water accumulation rate truly reflects the potential for water accumulation growth in the field.

[0057] However, the net water accumulation rate cannot be directly accumulated into water volume because hydraulic exchange occurs between the waterlogged fields and the channels. If this exchange process is ignored, the accumulated water volume in the fields will continue to dissipate over time, failing to reflect the true rise and fall of water levels. After waterlogging, there are two opposing exchange processes between the fields and the channels: outward drainage and lateral seepage replenishment. To characterize the inhibitory effect of outward drainage on water volume growth, a drainage damping factor is introduced, with a value between 0 and 1; a smaller value indicates a stronger inhibitory effect of outward drainage on water volume growth. To characterize the increasing effect of lateral seepage replenishment on water volume, an incremental lateral seepage replenishment is introduced, calculated as a positive value based on the head difference when lateral seepage occurs.

[0058] Furthermore, based on the real-time water level comparison between the fields and the channels, the two exchange effects are modeled separately: when the water depth in the fields is not lower than the channel water level, the fields drain water into the channels, and the increase in water accumulation is reduced by the drainage damping factor; when the channel water level is higher than the water depth in the fields but not lower than the lowest ground elevation of the fields, the channels replenish water to the fields through lateral seepage, directly increasing the water volume. The two scenarios automatically switch based on the real-time water level comparison, without confusion between them. The recursive expression for the water volume increment and cumulative water volume at the corrected water level sampling time is as follows:

[0059]

[0060] In the formula, Water level sampling time The increase in water accumulation refers to the net increase in water volume in a field between the time of the previous water level sampling and the time of the current water level sampling. Water level sampling time The drainage damping factor is dimensionless and ranges from 0 to 1. It is used to characterize the degree to which external drainage inhibits the growth of water accumulation. The smaller the value, the stronger the inhibition. Water level sampling time The net water accumulation rate; The sampling cycle for the channel water level gauge Water level sampling time The increase in lateral seepage recharge, This represents the cumulative water volume at the previous water level sampling time. Water level sampling time The cumulative water volume.

[0061] By modeling in different directions, the growth of water accumulation is suppressed under the external discharge condition, thus avoiding the unlimited accumulation of water volume; under the side seepage recharge condition, the water volume increases directly, thus avoiding underreporting caused by ignoring side seepage recharge in the scenario of continuous high channel water level.

[0062] The cumulative water volume reflects the total water volume of the entire field, but the actual water depth also depends on the field area and topography. The same amount of water will result in shallower water in larger fields and deeper water in smaller fields. This invention uses a field elevation-volume curve to convert the cumulative water volume into the field water depth, which can be expressed by the following formula:

[0063] In the formula, Water level sampling time The depth of water accumulation in the fields, For the area of ​​the field, This is the inverse conversion relationship between elevation and volume. Water level sampling time The cumulative water accumulation. Different fields associated with the same measuring point have different areas, resulting in different calculated water depths, thus realizing differentiated spatial expansion to multiple fields under sparse measuring point conditions.

[0064] Finally, based on the water level sampling time calculated above... The depth of water accumulation in a field is used to determine whether the water level poses a threat to crops. Different crops and different growth stages have different flood tolerance levels; the same water depth may have no effect on flood-tolerant crops, but may cause serious damage to flood-intolerant crops. Therefore, this invention, based on the crop type and current growth stage, exemplarily sets a first threshold, a second threshold, and a third threshold, and outputs a warning signal according to the following rules: When the water level was sampled No warning is issued when the water depth in the field is less than the first threshold, at the time of water level sampling. When the water depth in a field is greater than or equal to the first threshold but less than the second threshold, a "mildly concerned" signal is output. This signal is given at the time of water level sampling. A moderate warning is issued when the water depth in the field is greater than or equal to the second threshold but less than the third threshold. This warning is given at the time of water level sampling. A severe alarm is issued when the water depth in a field exceeds the third threshold. The warning signal includes the field number, water depth, estimated timestamp, and warning level, for irrigation district managers to use as a reference for decision-making.

[0065] In summary, the above-mentioned S3 method first characterizes the decreasing water absorption capacity of soil as it transitions from dry to wet by calculating time-varying infiltration rates. This avoids false alarms from the first rainfall after a drought and underreporting of continuous rainfall after soil saturation. Next, it calculates the net water accumulation rate, achieving a real-time balance between runoff and infiltration. Then, it models the two opposing exchange effects of outflow and lateral seepage, ensuring that outflow inhibits water dispersion and lateral seepage replenishment avoids underreporting. Finally, it obtains the cumulative water accumulation at each water level sampling time through discrete iterative recursion. Finally, it uses elevation-volume curves to convert the cumulative water accumulation into the water depth of each field and compares it with dynamic risk thresholds based on crop type and growth stage, outputting multi-level early warning signals. The completion of this work provides irrigation district managers with field-scale water accumulation status information and tiered early warning decision support.

[0066] The system includes a processor and a memory, the memory storing computer program instructions, which, when executed by the processor, implement the farmland water level monitoring and early warning method according to the first aspect of the present invention.

[0067] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0068] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for monitoring and early warning of farmland water levels, characterized in that, include: Establish a spatial mapping relationship between measuring points and fields within the target irrigation area, determine the channel water level measuring point to which each field belongs, and calculate the previous rainfall index on a rolling basis; Based on the previous rainfall index, the runoff inversion state intervals are divided, including drought intervals, transition intervals, and wet intervals. The drought interval corresponds to the drought baseline inversion coefficient, the transition interval corresponds to the transition baseline inversion coefficient, and the wet interval corresponds to the wet baseline inversion coefficient. Collect channel water level data and rainfall intensity data, determine the scheduling interference of channel water level data, and when it is determined to be the dominant rainfall condition, calculate the water level-rainfall intensity response coefficient by combining the channel water level rise and fall rate with the synchronously aligned rainfall intensity parameter. The baseline inversion coefficients are corrected based on the water level-rainfall intensity response coefficients to obtain the actual runoff inversion coefficients. The effective runoff intensity is calculated based on the actual runoff inversion coefficient and the synchronous aligned rainfall intensity parameter; the time-varying soil infiltration rate is calculated based on the previous rainfall index; the effective runoff intensity is subtracted from the time-varying soil infiltration rate and a non-negative value is taken to obtain the net water accumulation rate. Based on the real-time water level comparison between the field and the canal, the water accumulation growth is reduced by the drainage damping factor in the external drainage scenario, and the side seepage replenishment increment is increased in the side seepage replenishment scenario. The net water accumulation rate is iteratively accumulated to obtain the cumulative water accumulation. The accumulated water volume is converted into the water depth of the field, and a graded early warning signal is output based on the comparison result of the water depth of the field with the threshold.

2. The method for monitoring and early warning of farmland water levels according to claim 1, characterized in that, Establishing the spatial mapping relationship between measuring points and field plots within the target irrigation area includes: The measuring points at the head of the branch canal cover all the fields within the catchment area of ​​the branch canal under its control, while the measuring points at the inlet of the agricultural canal cover the set of fields that have a direct hydraulic connection with the agricultural canal. Using the field canal network and field boundaries as spatial boundaries, and combining field ground elevation data, the catchment areas of each measuring point are divided. Fields whose ground elevation is lower than the design control water level of their respective canals are included in the catchment area of ​​the corresponding measuring point of that canal. The fields located near the boundary of the catchment area of ​​two adjacent measuring points are assigned to the measuring point corresponding to the side with the lower ground elevation. Fields without water level monitoring points are assigned to the nearest monitoring point at the entrance of the irrigation canal that is hydraulically connected. If there is no irrigation canal entrance monitoring point nearby, they are assigned to the monitoring point at the head of the branch canal.

3. The method for monitoring and early warning of farmland water levels according to claim 1, characterized in that, The rolling calculation of the previous rainfall index includes: the remaining value of the previous rainfall index after index decay at the previous rainfall sampling time, plus the measured rainfall between the previous rainfall sampling time and the current rainfall sampling time, to obtain the previous rainfall index at the current rainfall sampling time.

4. The method for monitoring and early warning of farmland water levels according to claim 1, characterized in that, The method for determining channel water level data interference includes: comparing real-time channel water level data with historically statistical baseline data of normal channel water level fluctuations; determining that a gate pump scheduling event exists when the absolute value of water level fluctuations in adjacent time periods exceeds a preset threshold; the channel water level rise and fall rate at the current water level sampling time does not participate in runoff inversion, and the runoff inversion coefficient uses the value from the previous available water level sampling time; determining that rainfall is the dominant condition when the absolute value of water level fluctuations in adjacent time periods does not exceed the preset threshold.

5. The method for monitoring and early warning of farmland water levels according to claim 1, characterized in that, Before calculating the water level-rainfall response coefficient using the channel water level rise and fall rate and the synchronously aligned rainfall intensity parameters, a time alignment step is also included: for the water level sampling time of the channel water level measuring point, take the two adjacent rainfall sampling times before and after the water level sampling time and their corresponding measured rainfall intensity values ​​and perform linear interpolation to obtain the synchronously aligned rainfall intensity parameters for the water level sampling time.

6. The method for monitoring and early warning of farmland water levels according to claim 1, characterized in that, Correcting the baseline inversion coefficients based on the water level-rainfall intensity response coefficients includes: mapping the water level-rainfall intensity response coefficients to correction weights, using the correction weights as interpolation coefficients, and performing linear interpolation between the baseline inversion coefficients and the baseline inversion coefficients of adjacent humid intervals to obtain the actual runoff inversion coefficients; if the current interval is a humid interval, then the humid baseline inversion coefficients are directly used as the actual runoff inversion coefficients.

7. The method for monitoring and early warning of farmland water levels according to claim 1, characterized in that, The time-varying soil infiltration rate is calculated based on the previous rainfall index. The time-varying soil infiltration rate is negatively correlated with the previous rainfall index. The lower limit of the time-varying soil infiltration rate is the infiltration rate when the soil is saturated, and the upper limit is the infiltration rate when the soil is dry.

8. The method for monitoring and early warning of farmland water levels according to claim 1, characterized in that, The net water accumulation rate is iteratively accumulated based on the real-time water level comparison between the field and the channel, including: when the water depth in the field is not lower than the water level in the channel, it is determined to be an external discharge scenario, and the net water accumulation rate is accumulated after being reduced by the drainage damping factor; when the water level in the channel is higher than the water depth in the field but not lower than the lowest ground elevation of the field, it is determined to be a lateral seepage recharge scenario, and the lateral seepage recharge increment is added to the net water accumulation rate before accumulation.

9. The method for monitoring and early warning of farmland water levels according to claim 1, characterized in that, Converting the accumulated water volume into the water depth of the field includes: converting the accumulated water volume into the water depth of the field based on the field area and the field's elevation-volume curve.

10. A farmland water level monitoring and early warning system, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the farmland water level monitoring and early warning method according to any one of claims 1-9.

Images