Agricultural water consumption verification method based on two-dimensional response relationship between precipitation and inflow

By using a method based on the two-dimensional response relationship between precipitation and inflow, meteorological and hydrological characteristics of agricultural irrigation areas are extracted, and total regulation coefficients and temporal distribution correction functions are generated. This solves the problem of the lack of graded evaluation of water source security capacity in existing technologies, achieves more accurate agricultural water consumption accounting, and reduces the impact of climate fluctuations.

CN123118293APending Publication Date: 2026-11-17MWR GENERAL INSTITUTE OF WATER RESOURCES & HYDROPOWER PLANNING & DESIGN +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611467994.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-21
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies fail to effectively classify and evaluate the water supply security capabilities of different physical storage characteristics when calculating agricultural water consumption. This results in numerical discrepancies and systematic deviations between the output water consumption calculation data and the actual water consumption level under complex operating conditions involving long-term climate fluctuations and multiple water supply sources.

Method used

Based on the two-dimensional response relationship between precipitation and inflow, by acquiring meteorological and hydrological data, agricultural irrigation area attributes and crop planting structure of the target area, precipitation climate response characteristics and equivalent inflow guarantee characteristics are extracted, total regulation coefficient and time distribution correction function are generated, and orthogonal decoupling correction is performed to obtain the verified agricultural water consumption.

Benefits of technology

This reduces the interference of hydrological fluctuations on agricultural water consumption calculation results, making the verified data closer to the actual level of water conservation and intensive use in the region.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123118293A_ABST
    Figure CN123118293A_ABST
Patent Text Reader

Abstract

This invention discloses a method for verifying agricultural water consumption based on the two-dimensional response relationship between precipitation and inflow. The method includes: acquiring basic data such as original agricultural water consumption and meteorological and hydrological data for the target area; extracting precipitation climate response features characterizing precipitation fluctuations, and extracting equivalent inflow guarantee features that integrate water supply differences based on irrigation district attributes; extracting a comprehensive water demand time window based on crop planting structure; generating a total water resource regulation coefficient describing the degree of deviation in annual total water resources based on precipitation and equivalent inflow guarantee features; generating a time distribution correction function describing the deviation in time distribution within the year, combining the hydrological distribution status and storage capacity within the window; and using the coefficient and function to perform orthogonal decoupling correction on the original data. This invention reduces the interference of hydrological abundance and scarcity, restoring the true water use efficiency.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of water resources management and agricultural water conservancy information processing technology, and in particular to a method for verifying agricultural water consumption based on the two-dimensional response relationship between precipitation and inflow. Background Technology

[0002] Agricultural irrigation water consumption is an important component of the total regional water resource consumption. Accurately calculating its actual water use efficiency is the basis for conducting watershed water cycle evolution analysis and formulating water rights allocation strategies.

[0003] Current accounting mechanisms for regional agricultural water use data primarily rely on direct statistical aggregation through year-end meter readings and ledger reporting. Some modern research attempts to introduce empirical compensation algorithms based on total annual precipitation to correct statistical errors arising from direct aggregation. However, these methods typically treat historical water use statistics as a stationary time series, using only total rainfall as the sole driving variable. They fail to classify and evaluate the water source security capabilities of different physical storage characteristics. When faced with complex conditions involving long-term climate fluctuations and multiple water supply sources, the output water consumption accounting data exhibits numerical dispersion and systematic bias compared to the actual intensive water use levels in the region. Summary of the Invention

[0004] The purpose of this invention is to provide an agricultural water consumption verification method based on the two-dimensional response relationship between precipitation and inflow, in order to solve the above-mentioned problems in the prior art.

[0005] Technical solution: A method for verifying agricultural water consumption based on the two-dimensional response relationship between precipitation and inflow, comprising:

[0006] Obtain the raw agricultural water consumption of the target area, as well as basic data including meteorological and hydrological data, agricultural irrigation area attributes, and crop planting structure;

[0007] Based on meteorological and hydrological data, precipitation climate response characteristics that characterize precipitation fluctuations are extracted, and equivalent water supply guarantee characteristics that integrate the differences in multi-source water supply are extracted by combining agricultural irrigation area attributes.

[0008] Based on crop planting structure, a comprehensive water demand time window is extracted to characterize the regional water-sensitive cycle.

[0009] Based on precipitation climate response characteristics and equivalent water inflow security characteristics, a total amount adjustment coefficient is generated to describe the degree of deviation of annual total water resources.

[0010] Based on the distribution of meteorological and hydrological data within the comprehensive water demand time window, and combined with the engineering storage capacity in the attributes of agricultural irrigation districts, a time distribution correction function is generated to describe the degree of deviation of the time distribution of water resources within the year.

[0011] The original agricultural water consumption was orthogonally decoupled and corrected using the total amount adjustment coefficient and the time distribution correction function to obtain the verified agricultural water consumption.

[0012] Beneficial effects: This invention can reduce the interference of hydrological fluctuations on agricultural water consumption calculation results, making the verified data closer to the actual level of water conservation and intensive use in the region. Attached Figure Description

[0013] Figure 1 This is a flowchart of the agricultural water consumption verification method based on the two-dimensional response relationship between precipitation and inflow of water according to the present invention.

[0014] Figure 2 This is a flowchart of the process for extracting precipitation climate response characteristics according to the present invention.

[0015] Figure 3 This is a flowchart of the construction of precipitation climate response characteristics according to the present invention.

[0016] Figure 4 This is a flowchart illustrating the extraction of equivalent incoming water protection features according to the present invention.

[0017] Figure 5 This is a flowchart of the extraction of the comprehensive water demand time window of the present invention. Detailed Implementation

[0018] To enable those skilled in the art to better understand the present invention, 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 only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0019] In order to solve the problems mentioned in the background art, combined with Figures 1 to 5 The present invention will be specifically described through the following embodiments.

[0020] The precipitation-inflow two-dimensional response relationship of this invention refers to the response influence of agricultural water consumption on mutually independent physical driving dimensions, specifically including: precipitation climate dimension, used to characterize the driving effect of natural precipitation abundance and scarcity on irrigation demand; inflow security dimension, used to characterize the constraint effect of multi-source engineering water supply reliability on actual water intake capacity; water demand time sequence dimension, used to characterize the matching status between crop water sensitivity period and water resource allocation timing; and regulation and buffer dimension, used to characterize the ability of engineering facilities to physically shift and resolve the temporal and spatial misalignment of supply and demand.

[0021] Example 1

[0022] According to one aspect of this application, this embodiment provides an agricultural water consumption verification method based on four-dimensional response relationships, the method comprising the following steps:

[0023] Step 101: Obtain the original agricultural water consumption of the target area, as well as basic data including meteorological and hydrological data, agricultural irrigation area attributes, and crop planting structure.

[0024] When executing the data acquisition process, it is necessary to establish the spatial boundaries of data analysis. Specifically, county-level administrative divisions are used as the conventional spatial accounting units, and a multi-scale raster system is constructed on top of these. For example, a 1km-scale raster is used for spatial gridding of regional climate and hydrological process data, a 30m-scale raster carries crop planting structure information, and a 10m-scale raster is used to express crop growth stage characteristics. Through this multi-scale nesting mechanism, the spatial alignment and integration of macro-hydrological characteristics and micro-agricultural information are achieved.

[0025] The acquired basic data are presented in a pre-defined format within the system. For example, meteorological and hydrological data are represented as a two-dimensional numerical matrix containing long-term historical daily rainfall, evapotranspiration, and river runoff. Crop planting structure is represented as a vector raster layer carrying geographic coordinates and crop type identification codes. Agricultural irrigation district attributes include the design capacity of reservoirs within the region and the water distribution flow thresholds for each level of canal system.

[0026] Accordingly, in order to address the differences in data abundance across different regions, the process of acquiring basic meteorological and hydrological data includes the following two logical branches.

[0027] When the data is complete, if the target area has a continuous hydrological station observation sequence for more than a preset number of years, the data acquisition module can directly retrieve the two-dimensional numerical matrix of the local station for cleaning and storage.

[0028] In cases of data gaps, if there are no long-sequence hydrological stations in the target area, a spatial translation substitution mechanism is triggered. The computation module, based on topographic elevation and watershed topology, finds the nearest neighboring accounting units with similar underlying surfaces and uses the data sequences of these neighboring units as the local meteorological and hydrological baseline data, thereby improving the coherence of global network computation.

[0029] After completing the basic data collection, based on the regional irrigation water source structure characteristics and the completeness of supporting engineering facilities, the conventional accounting units were clustered into large and medium-sized irrigation areas, small and rain-fed agricultural areas, and inter-regional water transfer irrigation areas. For different irrigation area types after classification, corresponding differentiated operation parameters were assigned to the dynamic blocking rule in the total regulation coefficient and the feature fusion calculation in the time distribution correction function to match the actual drought resistance buffering capacity of each type of irrigation area.

[0030] Step 102: Extract precipitation climate response features that characterize precipitation fluctuations based on meteorological and hydrological data, and extract equivalent water supply guarantee features that integrate the differences in multi-source water supply based on agricultural irrigation area attributes.

[0031] To reflect the dual-driven underlying physical characteristics of natural climate disturbances, this embodiment de-stationes the long-sequence historical precipitation matrix, removes interference from human-induced water-saving technologies, and extracts the deviations from the baseline in precipitation abundance and scarcity.

[0032] Accordingly, considering the differences in the reliability of water from different sources, this embodiment, following the laws of natural variation, uses a calculation module to map river runoff, reservoir storage, and inter-regional water transfer to a unified equivalent supply level. This reduces the complexity of multi-source water supply data to a standard equivalent inflow guarantee characteristic, which is then used to establish a joint correction model. In this embodiment, the equivalent inflow includes at least irrigation water.

[0033] Step 103: Extract the comprehensive water demand time window based on the crop planting structure to characterize the regional water-sensitive cycle.

[0034] When dealing with accounting units involving multiple crop intercropping, the information processing system extracts peak water demand periods at the regional scale by reading area proportion weights from vector raster layers. This defines the start and end dates of the calendar period, providing a temporal reference boundary for evaluating the timing of precipitation and water allocation, and reducing the dilution interference of irrelevant hydrological data throughout the year on the assessment results.

[0035] Step 104: Based on the precipitation climate response characteristics and equivalent water inflow guarantee characteristics, generate a total amount adjustment coefficient to describe the degree of deviation of the annual total water resources.

[0036] After extracting the core climate and water inflow characteristics, a parameter construction module is used to nonlinearly orthogonally couple the two feature components in a dimensionality-reduced feature space. This describes the surge in theoretical irrigation demand caused by annual precipitation shortages, as well as the physical constraints on actual water intake capacity imposed by bottlenecks in the supply of key water sources. The coupled superposition of these two factors jointly characterizes the level of deviation between the current year's macro-level water abundance and scarcity. For passive water use fluctuations caused by changes in the absolute abundance of interannual water resources, a total water supply adjustment coefficient is used to correct and calibrate the water use to the baseline state of a standard normal water year.

[0037] Step 105: Based on the distribution of meteorological and hydrological data within the comprehensive water demand time window, and combined with the engineering storage capacity in the agricultural irrigation area attributes, generate a time distribution correction function to describe the degree of deviation of the time distribution of water resources within the year.

[0038] Besides the annual absolute amount of water resources, the staggered rainfall events within a year can also mask the true water-saving performance. In this embodiment, the computing unit extracts time-series values ​​within the comprehensive water demand time window to assess the degree of overlap between the time axis of concentrated water supply and the actual drought period of crops. Combining the physical buffering capacity index of reservoirs and canal systems, a time distribution correction function is fitted to correct the spatiotemporal distribution misalignment error. This function is independent of the total amount adjustment coefficient in terms of the physical target dimension of the correction; the former focuses on correcting the deviation of the time distribution within the year, while the latter focuses on correcting the deviation of the annual total amount.

[0039] Step 106: The original agricultural water consumption is orthogonally decoupled and corrected using the total amount adjustment coefficient and the time distribution correction function to obtain the verified agricultural water consumption.

[0040] To mitigate the risk of overcompensation or repeated penalties caused by interference between multiple hydrological driving factors during joint correction, this embodiment employs an orthogonal decoupled multiplicative correction mechanism. This mechanism sequentially multiplies the collected raw agricultural water consumption with the calculated total volume regulation coefficient and time distribution correction function. The corresponding calculation formula is as follows:

[0041] W _corrected =W _original ×K _total ×f _t ;

[0042] Among them, W _corrected To verify the agricultural water consumption, W _original K represents the original agricultural water consumption. _total f is the total adjustment coefficient. _t This is a time distribution correction function.

[0043] Based on this, the total amount adjustment coefficient takes a value less than 1 in dry years and a value greater than 1 in wet years. The time distribution correction function takes a value less than 1 in years with severe spatiotemporal misalignment and a value close to 1 in years with ideal distribution. This multiplication operation logic enables the independent occurrence of total amount deviation correction and time domain misalignment correction. After verification, the output agricultural water consumption is free from redundant interference caused by uncontrollable fluctuations in natural gas meteorology and water inflow, and can objectively reflect the true level of water conservation and intensive use in the target area.

[0044] Example 2

[0045] Based on Example 1 above, this example further explains the extraction of precipitation climate response characteristics that characterize precipitation fluctuations. In one possible implementation, the following steps are included:

[0046] Step 201: Extract historical precipitation sequences from meteorological and hydrological data, and obtain corresponding historical agricultural water use sequences from historical statistical records of original agricultural water use.

[0047] Obtain the annual rainfall data set and the corresponding annual average irrigation water consumption set within a preset time span. Construct a data combination with equal time dimensions and time-series correspondence as the basic data source for subsequent feature separation.

[0048] Step 202: Perform a non-stationarity test on the historical agricultural water use series to identify the technological progress trends implied in the series.

[0049] With the advancement of water-saving renovation projects in irrigation districts and the popularization of efficient irrigation technologies, under the same rainfall level, the actual water consumption of farmland shows a downward trend over a long period. If this gradual trend introduced by the systematic changes in engineering facilities is not removed, the fitted response function will be influenced by non-natural meteorological factors. Therefore, the computational module employs a non-parametric trend test algorithm to determine the stationarity of historical water use data, calculate relevant detection statistics, and assess the temporal distribution of the time series.

[0050] Accordingly, in this time series analysis stage, based on the results of the nonparametric test, the computational processing logic will enter the corresponding branch path. When the calculated output test statistic exceeds the set confidence level, it is determined that the current historical agricultural water use series is a non-stationary series, indicating a trend of technological progress, and the detrending feature extraction process is activated.

[0051] If the calculated test statistic does not reach the decision boundary, it is determined that the historical water consumption of the current accounting unit remains relatively stable and there is no efficiency improvement process. In this case, the original historical agricultural water consumption sequence is retained for subsequent mapping steps, and the data smoothing and difference calculation modules are skipped.

[0052] Step 203: When it is determined that there is a trend of technological progress, the historical agricultural water use sequence is stripped to remove the benchmark offset interference caused by human water-saving technology, and precipitation climate response features are constructed based on the historical precipitation sequence; when it is determined that there is no trend of technological progress, precipitation climate response features are constructed directly based on the statistical correlation between the historical precipitation sequence and the historical agricultural water use sequence.

[0053] In this embodiment, the macro-level goals of trend stripping and feature construction are achieved through the following decomposition execution instructions.

[0054] Step 2031: The historical agricultural water use sequence is decomposed into a technology progress trend component that characterizes the systematic changes in water use efficiency, and a climate response residual component that characterizes the impact of interannual meteorological fluctuations, using a data smoothing algorithm.

[0055] Specifically, the system invokes a locally weighted scatter smoothing algorithm to filter long-sequence water consumption data. Through dynamic fitting, it tracks low-frequency, slowly varying signals within the sequence, generating and outputting a smoothed fitted value sequence, which is then extracted as a technological progress trend component. This component characterizes the expected routine water consumption level of the farmland irrigation system after excluding hydrological disturbances of the year. Correspondingly, the algebraic difference set between the original sequence observations and the corresponding year's technological progress trend component is extracted to construct a climate response residual component. This component reflects the amplitude of water consumption fluctuations caused by interannual meteorological variations.

[0056] Step 2032: Normalize the residual component of the climate response using the technological progress trend component to obtain the climate response deviation rate.

[0057] Due to differences in water efficiency baselines, the absolute values ​​of residuals from earlier historical years and more recent years are not meaningful for direct horizontal comparison.

[0058] In this embodiment, a proportional calculation rule based on the technical level benchmark of each year is introduced to eliminate this dimensional difference. The corresponding calculation relationship is as follows:

[0059] D(t) = W _C (t) / W _T (t);

[0060] Where D(t) is the climate response deviation rate in year t on the time axis, and W _C (t) represents the residual components of the climate response output by the smoothing decomposition operation in year t, W _T (t) represents the technological progress trend component corresponding to year t.

[0061] The calculated climate response deviation rate is used to characterize the rate at which water use deviates from the standard expectation due to natural climate fluctuations under preset facility and equipment operating conditions; positive values ​​represent irrigation supplementation caused by drought, while negative values ​​represent natural substitution caused by humid climate.

[0062] Step 2033: Extract the annual precipitation frequency based on the historical precipitation sequence.

[0063] In this embodiment, the analysis unit performs a descending sorting operation on the samples of each year in the historical precipitation sequence and calculates the empirical frequency value corresponding to each sorting node. The higher the frequency value, the higher the degree of drought in that year.

[0064] Step 2034: Establish a monotonic mapping relationship between precipitation frequency and climate response deviation rate, and use this monotonic mapping relationship as a precipitation climate response characteristic.

[0065] The precipitation frequency array for each year is extracted and used as the horizontal axis variable, while the corresponding year's climate response deviation rate is used as the vertical axis variable, constructing a two-dimensional coordinate mapping relationship. Based on the fundamental farmland hydrological mechanism that the scarcer the precipitation resources, the higher the demand for supplementary irrigation in agriculture, an ordinal-preserving regression tool is used to perform constrained model fitting on discrete feature coordinate points. This process generates a feature function curve in which the output variable increases monotonically with the input variable in a non-negative manner, and this function is encapsulated as a localized precipitation climate response feature unit.

[0066] Example 3

[0067] Because different water supply projects have physical boundaries that separate their regulation characteristics, simply summing the water volume at the source can easily obscure the spatiotemporal buffering capacity of water resources. Therefore, a computer mechanism for accurate conversion of water volumes from multiple sources is introduced.

[0068] Before initiating feature extraction, a sub-process for testing the independence of regional water resources is pre-imported. The computational unit retrieves the annual series of natural precipitation and the annual series of total inflow including all forms of water supply for the region, performs Pearson product-moment correlation coefficient calculation, and assesses the coupling correlation between climate fluctuations and hydrological water supply fluctuations.

[0069] If the calculated correlation coefficient exceeds a preset threshold, it is determined that precipitation and inflow are positively correlated, meaning that inflow mainly relies on immediate runoff replenishment, and the engineering system has weak anti-interference capabilities. Under this condition, the two-factor independent decoupling model is prone to logical failure. The algorithm platform automatically disables the equivalent inflow frequency domain evaluation program, degenerating into a single-factor verification mode, relying on precipitation climate characteristics for subsequent calculations.

[0070] If the correlation coefficient does not meet the conditions for triggering the threshold, it is determined that the regional water inflow is dominated by water storage projects or external replenishment, thus decoupling from local meteorological conditions. The computational network is then activated normally, and the process proceeds to the subsequent multi-source protection feature conversion execution steps.

[0071] In one possible implementation, the equivalent water supply guarantee characteristics, which combine the attributes of agricultural irrigation districts with the differences in multi-source water supply, are extracted and integrated, including the following steps:

[0072] Step 301: Extract multi-source water inflow data from meteorological and hydrological data, specifically including natural river inflow components, reservoir regulation inflow components, and inter-regional water transfer components.

[0073] Specifically, based on meteorological and hydrological datasets, three types of flow time series with different fluctuation characteristics were separated according to the data source channels. These include the flow series of rivers at various levels representing natural runoff patterns, the available replenishment water series of reservoirs representing the buffering capacity of water storage projects, and the water series of water introduced from external watersheds reflecting the characteristics of long-distance intervention.

[0074] Step 302: Based on the characteristics of the agricultural irrigation area, assess the physical supply stability of various water components and determine the corresponding independent guarantee coefficients.

[0075] In this implementation, for various independent and separate sequence subsets, the calculation unit extracts ratio indicators that reflect the degree of water supply jump over time and the reservoir's interception capacity.

[0076] When calculating the independent guarantee coefficients for natural river components, the dispersion evaluation formula in mathematical statistics is used to obtain the coefficient of variation values ​​of the series. The reciprocal of these values ​​is then taken to obtain the final coefficient parameters. This relationship describes a state mapping where the greater the natural fluctuations in runoff, the lower the reliability of water supply from direct water diversion operations.

[0077] When calculating the independent guarantee coefficient corresponding to the reservoir regulation component, the regulation volume of large and medium-sized flood control and water conservancy reservoirs recorded in the agricultural irrigation district attributes is read. This constant volume parameter is divided by the long-term average annual flow at the reservoir cross-section to construct the reservoir regulation ratio. A constant offset is added to this regulation ratio, and this is used as the independent guarantee coefficient parameter for this component. This reflects the physical fact that the higher the proportion of flood control storage volume to annual runoff, the more ample the space for water allocation and storage in the face of drought.

[0078] For the water diversion component, since the amount of water resources in this component is mostly constrained by rigid dispatch orders and has a low degree of local climate intervention, its independent guarantee coefficient is set to a preset constant value that is greater than the average level of the first two components.

[0079] Step 303: Use each independent guarantee coefficient to perform a weighted transformation on the original water volume of the corresponding water components to obtain the equivalent water guarantee characteristics that reflect the actual supply reliability.

[0080] The computational core performs global peak finding and normalization scaling operations based on the collected independent guarantee coefficients. The specific normalization and feature transformation are completed through the following steps:

[0081] Step 3031: Determine the independent protection coefficient of the natural river inflow components based on the degree of variation of the natural river inflow components over time.

[0082] Step 3032: Based on the beneficial storage capacity of the reservoir where the water inflow component is located and the regulation ratio of the average annual inflow, determine the independent guarantee coefficient of the water inflow component; set the independent guarantee coefficient of the inter-regional water transfer component as a preset rigid guarantee constant.

[0083] Step 3033: Based on the highest level of protection parameter among various independent protection coefficients, normalize and transform all independent protection coefficients.

[0084] Specifically, the largest value is detected from three independent guarantee coefficients corresponding to natural river channels, reservoir regulation, and external water transfer. The value of the largest value is then used as the divisor of the global reference base. Each of the three coefficients is then placed under this reference base divisor, generating a set of normalized weighted vectors that fall within the specified interval. This process ensures that the water supply channel with the highest reliability receives the full evaluation weight.

[0085] Step 3034: Calculate the comprehensive equivalent guarantee volume by aggregating the original water volumes of each water component using the normalized independent guarantee coefficients.

[0086] The computational unit multiplies the original physical water volume readings recorded from each channel that year with the calculated normalized weighted weight vector, performs scalar summation, and obtains the dimensionality-reduced equivalent total through a linear polynomial superposition relationship. The relevant linear calculation formula is as follows:

[0087] W _eq =C _r ×W _riv +C _s ×dW _res +C _t ×W _trans ;

[0088] Among them, W _eq To characterize the equivalent guaranteed water volume of the region's adjustable comprehensive water supply capacity, C _r W represents the weighted parameter of natural river inflow after normalization. _riv C represents the total amount of natural river water recorded in the current period. _s The normalized inflow weight parameter for reservoir regulation, dW _res To calculate the change in net water release or storage in reservoirs over a given period, C _t W represents the weighted parameters for inter-regional water transfer after normalization. _trans The total amount of water transferred from other sources recorded in the current year is allocated.

[0089] Step 3035: Perform frequency characteristic analysis on the comprehensive equivalent guaranteed water volume, and use the obtained equivalent water inflow frequency as the equivalent water inflow guarantee characteristic.

[0090] After water volume conversion, the generated equivalent water volume time series set is processed using an empirical frequency statistical algorithm to generate the corresponding equivalent inflow frequency. This feature data retains the original multi-source heterogeneous information and reduces the impact of reliability differences. It is updated, recorded, and stored in the equivalent inflow guarantee feature output interface for downstream calibration algorithms.

[0091] Example 4

[0092] Because different crops have different growth and development rhythms, the irrigation experience cycle of a single crop is difficult to represent the actual concentrated period of water resource consumption in the entire region. Therefore, a spatially weighted time series fusion mechanism is introduced to delineate the macroscopic observation window.

[0093] In one possible implementation, extracting a comprehensive water demand time window characterizing the water-sensitive cycle of a region based on crop planting structure includes the following steps:

[0094] Step 401: Analyze the crop planting structure to determine the various crop types covered in the region and the planting area ratio of each crop.

[0095] For example, land cover classification codes stored in a 30m spatial resolution raster layer are extracted, and all crop types present in the calculation unit are summarized through classification counting operations. Next, the raster cell areas of the same type of crop are accumulated, and the total area is divided by the total effective irrigated cultivated land area in the entire calculation unit to obtain a numerical set of planting area proportions for each type of crop. For scattered crop categories with an area proportion lower than a preset filtering threshold, a removal operation is performed, retaining the main crops to form the calculation set. In this embodiment, for example, the preset filtering threshold is set to 0.05, that is, scattered crops with a planting area proportion of less than 5% are not included in the calculation of water demand periods. Those skilled in the art can appropriately adjust this threshold according to the complexity of the regional planting structure.

[0096] Step 402: Based on the pre-stored crop water requirement parameters, identify the single crop water requirement period when the average daily water requirement of each crop type reaches its peak during its entire growth period.

[0097] For each crop category retained within the dataset, the data retrieval module retrieves the daily average irrigation intensity variation curves for the corresponding crop throughout its entire growth period from the built-in farmland water conservancy basic parameter database, identifying the intervals where the intensity values ​​reach their peak. During this process, a boundary convergence constraint judgment rule is executed: provided that the cumulative irrigation intensity values ​​within the selected time period are not less than a preset proportional limit of the total irrigation demand for the entire growth period, the set of consecutive dates with the narrowest coverage of days is found and recorded as the single crop water requirement period for the corresponding crop category.

[0098] Step 403: Using the planting area ratio of each crop as a weighting coefficient, the water demand periods of all single crops are weighted and integrated to construct a weighted water demand intensity distribution that represents the comprehensive irrigation demand of the entire region.

[0099] Specifically, the calculation unit uses the natural calendar days as the time axis and maps the acquired daily average irrigation intensity time series curves of each crop onto a unified time metric scale. The planting area ratio corresponding to each crop is extracted and used as a constant coefficient factor. Daily multiplication and horizontal summation operations are performed on the daily intensity values ​​of all crops to construct a weighted daily water demand weight function at the whole region scale.

[0100] Step 404: Extract the comprehensive water demand time window based on the weighted water demand intensity distribution.

[0101] Furthermore, after establishing the weight function distribution shape, the main high-frequency sensitive periods are extracted through cumulative calculation. This process specifically includes:

[0102] Step 4041: Calculate the cumulative distribution characteristics of the weighted water demand intensity distribution over the entire year.

[0103] Specifically, the daily water demand weight values ​​are arranged in ascending order of time to generate a monotonically increasing cumulative distribution function characteristic curve. Its value range is constrained by the overall normalization operation and is limited to a preset constant interval.

[0104] Step 4042: In the cumulative distribution characteristics of the time series, search for the continuous period that meets the preset cumulative water volume percentage threshold and has the shortest time span, and use it as the comprehensive water demand time window to lock the macro water sensitive period of the whole region.

[0105] The search program initiates a sliding window optimization algorithm, traversing the cumulative distribution curve. A dynamically moving time scanning window is set, filtering the cumulative numerical differences within this window to determine if they equal or exceed a preset threshold for the cumulative water volume percentage. Among all candidate time intervals that meet this threshold difference requirement, the segment with the smallest start and end date difference is selected. The start and end dates of this shortest intercepted segment are extracted and established as the comprehensive water demand time window parameter applicable to the entire region.

[0106] Step 4043: Map the natural precipitation process and regional water inflow process recorded in the meteorological and hydrological data to the comprehensive water demand time window, and compare the supply and demand of the natural precipitation process with the weighted water demand intensity distribution on a time-by-time basis to calculate the effective precipitation matching degree within the comprehensive water demand time window.

[0107] After defining the core time observation section, time-domain slicing is performed on the annual natural meteorological fluctuation data to extract the daily natural rainfall values ​​and equivalent river inflow data that occur within the locked window period.

[0108] To avoid including excess runoff and ineffective evaporation, a daily logical comparison truncation is performed for this window period. Specifically, the total daily natural water supply input (sum of precipitation and inflow) is compared with the county's weighted water demand intensity distribution value for that day, and the smaller value is extracted for effective hydrological calculation. All truncated smaller values ​​within the entire interception window period are summed up, and based on this, the effective precipitation matching degree characteristic parameters that can be truly converted into usable water resources for farmland during this period are calculated.

[0109] Example 5

[0110] Traditional univariate fitting is prone to extrapolating limit values ​​at both ends of hydrological data that do not conform to the laws of physics. Therefore, a dual-constraint formula model based on the physical mechanism of water resource supply and demand is introduced.

[0111] In one possible implementation, based on precipitation climate response characteristics and equivalent water inflow security characteristics, a total quantity adjustment coefficient is generated to describe the degree of deviation of the annual total water resources, including the following steps:

[0112] Step 501: Map the precipitation climate response characteristics into demand-driven response units that characterize the surge in irrigation demand under conditions of sufficient water resources.

[0113] The precipitation climate response characteristics are obtained, and the equivalent precipitation frequency parameter is extracted. As an example, assuming an ideal boundary condition with no external water supply interception or restriction, the theoretical irrigation compensation is calculated based on meteorological drought performance. The parameter calculation engine uses a logarithmic linear model for parameterization, meaning the logarithm of water demand is linearly related to the precipitation frequency. The corresponding calculation formula is:

[0114] F _req =W _0 ×exp(α×(p _r -0.5);

[0115] Among them, F _req W is the output value of the demand-driven response unit. _0 Here, exp is the calibration constant for baseline water consumption in a normal year, α is the exponential function operator with the natural constant e as the base, p is the pre-optimized precipitation elasticity parameter offline, and α is the parameter p. _r This represents the annual precipitation frequency. The formula is essentially a log-linear model, i.e., F0. _req The natural logarithm of precipitation maintains a linear relationship with the precipitation frequency, and in the original space, it exhibits an exponentially nonlinearly amplified response behavior with the degree of drought.

[0116] In this embodiment, the difference between the precipitation frequency and the median value of a normal year is used as the input variable of the exponential function. As the climate shifts towards drier conditions, the theoretical water demand exhibits a non-linear amplification curve to match the uneven increase in transpiration consumption by crops under drought conditions.

[0117] Step 502: Based on the equivalent water supply guarantee characteristics, construct a supply constraint response unit that characterizes the passive decline state of actual water consumption under water source constraints.

[0118] Water consumption depends not only on plant needs but also on the rigid constraints of the actual allocation capacity of key water conservancy facilities. In this embodiment, the calculation module extracts the engineering storage capacity index, which characterizes the physical buffer level of the irrigation district. A local truncation protection mechanism is used when calculating the engineering storage capacity index, and the corresponding calculation formula is as follows:

[0119] I _cap =w _1 ×min(V _r / V _ref,1.0 )+w _2 ×min(Q _r / Q _ref,1.0 );

[0120] Among them, I _cap w is the engineering storage capacity index. _1 with w _2 These are the weighting coefficients for storage capacity and canal system, respectively, w _1 with w _2 The sum is 1, V _r V represents the physical observation value of the reservoir's beneficial storage capacity. _ref Q is the high percentile reference reservoir capacity for the secondary zone of this watershed. _r Q represents the physical observation value of the water conveyance and distribution flow rate in the canal system. _ref Here, represents the high percentile reference flow rate corresponding to the secondary zone of the watershed, and 'min' is the boundary truncation operator that takes the smaller value. This mechanism limits the over-configured local extreme volume truncation to within 1.0, preventing singular overflow in subsequent multiplication and division operations.

[0121] Accordingly, based on the calculated I _cap A dynamic blocking threshold is established that fluctuates with engineering conditions, and the corresponding calculation formula is as follows:

[0122] p _w_th =0.5+c _0 ×I _cap ;

[0123] Where, p _w_th For dynamic blocking threshold, c _0 I is the preset tolerance elastic constant. _cap This refers to the engineering water storage capacity index obtained from the previous step. In other words, the more complete the supporting facilities, the higher the upper limit of the frequency of water shortages that can be tolerated.

[0124] Next, the current equivalent inflow frequency is compared with the dynamic blocking threshold, specifically including:

[0125] When the equivalent water inflow frequency is not greater than the dynamic interruption threshold, the current supply capacity is determined to be able to cope with water demand fluctuations. No constraints are imposed on demand; the value F of the supply constraint response unit is... _sup The value is assigned to be the constant 1.

[0126] When the equivalent inflow frequency exceeds the dynamic blocking threshold, the system determines that a water supply bottleneck has occurred. The calculation engine performs a secondary attenuation pressure reduction operation, and the corresponding calculation formula is as follows:

[0127] F _sup =max(C _min ,1-β×(p _w -p _w_th ) 2 );

[0128] Among them, F _sup To provide the value of the constraint response unit, max is the minimum operator that takes the larger value, C _min To maintain the minimum guarantee constant for crop survival baseline, β is the supply constraint strength parameter, p _w p is the equivalent water inflow frequency for that year. _w_th This is the calculated dynamic blocking threshold. The formula outputs a penalty coefficient less than 1, used to simulate the phenomenon where water source depletion forces a reduction in actual water diversion.

[0129] In this embodiment, β, c _0 C _min Supply constraint parameters can be determined by those skilled in the art based on regional irrigation practices: β can be obtained by fitting the relationship between actual water consumption and theoretical water demand in historical dry years; c _0 It can be determined by back-calculation based on the historical maximum drought resistance frequency of irrigation areas with different water storage capacities; C _min It can be determined by referring to the ratio of the minimum irrigation quota for maintaining the survival of basic staple crops in the local area to the normal water consumption per mu; w _1 with w _2 The allocation can be based on the ratio of engineering investment or water supply volume between reservoir water supply and canal water supply in the actual water source structure of the region.

[0130] Step 503: The demand-driven response unit and the supply-constrained response unit are coupled and synthesized to generate a total quantity adjustment coefficient that is decoupled from the spatiotemporal misalignment factors, which is specifically used to correct the water use deviation caused by the deviation of the annual total water resources.

[0131] Specifically, after completing the parameterized calculations for both demand and supply, the aggregation module normalizes the water consumption by dividing the baseline water consumption constant by the product of the output values ​​of the demand-driven response unit and the supply-constrained response unit, and outputs a dimensionless check multiplier reflecting the intensity of the total annual fluctuation, i.e., the total adjustment coefficient. The expression for this calculation relationship is as follows:

[0132] K _total =W _0 / (F _req ×F _sup );

[0133] Among them, K _total W is the total adjustment coefficient. _0 F is the calibration constant for the baseline water consumption in a normal water year. _req F is the output value of the demand-driven response unit. _sup To provide the value of the supply constraint response unit; F _req W is already included in the middle. _0 K _total Essentially, it simplifies to 1 / (exp(α×(p)). _r -0.5))×F _sup ), with W _0 The absolute value is irrelevant.

[0134] According to this calculation formula, in drier years K _total K less than 1, wet year _total If the value is greater than 1, the original statistical data will be aligned to the standard average water year baseline in subsequent multiplication corrections. This quotient structure isolates the influence of the distribution pattern of hydrological events on the micro-timeline within the year, and constrains the extraction of features to the macro-interannual span level.

[0135] Furthermore, this embodiment provides a further explanation of the process for generating the total adjustment coefficient, specifically including:

[0136] Step 5031: The demand-driven response unit is constructed through the following mechanism: an elastic response boundary is set that amplifies non-linearly with the degree of precipitation drought to ensure the theoretical surge in irrigation demand under extreme drought conditions, thus obtaining the demand-driven response unit; the supply-constrained response unit is constructed through the following mechanism: a corresponding dynamic blocking threshold is set based on the engineering storage capacity in the agricultural irrigation area attributes; when the equivalent water inflow guarantee characteristics indicate that the water inflow is sufficient and does not exceed the dynamic blocking threshold, no water quantity constraint is imposed on the demand-driven response unit, and the supply-constrained response unit takes the value of a constant 1; when the dynamic blocking threshold is exceeded, a forced attenuation constraint is imposed according to the degree of water inflow drought, thus obtaining a supply-constrained response unit with a value less than 1.

[0137] In another alternative embodiment, when there is a large area of ​​missing basic historical monitoring data, making it difficult to calibrate the elastic response boundary and attenuation parameters based on the algorithm, this embodiment provides a static conversion alternative rule based on classification and degradation.

[0138] Specifically, the classification identifiers of irrigation district attributes are extracted. For large and medium-sized irrigation districts, a static attenuation multiplier is set using the equivalent water shortage magnitude; for small irrigation districts without regulation and storage, a direct linear summation conversion ratio of the difference in natural rainfall and the difference in river inflow is used; for irrigation districts receiving water from outside the region, direct coefficient replacement is implemented based on a preset constant proportion of the externally allocated water volume to the total planned water volume. This replacement rule directly reduces dimensionality, enabling the calculation module to operate resiliently even in data-scarce environments.

[0139] Example 6

[0140] Even if the total interannual water resources are in line with the expected annual levels, deviations in water availability from the actual agricultural demand period will still trigger passive supplemental irrigation. Building upon Example 5 above, this example further explains the generation of a time distribution correction function to describe the degree of deviation in the temporal distribution of water resources within the year.

[0141] In one possible implementation, the following steps are included:

[0142] Step 601: Based on the degree of deviation in the temporal distribution of meteorological and hydrological data within and outside the comprehensive water demand time window, extract the concentrated characteristics of precipitation periods.

[0143] Specifically, the process involves accessing a pre-defined comprehensive water demand time window, calculating the actual cumulative precipitation within that time boundary, and determining its current percentage of the total annual precipitation. Based on historical meteorological records, the average percentage within the same time window over many years is calculated. The current percentage is then divided by the average percentage using a computational unit; the corresponding linear calculation formula is as follows:

[0144] I _p =R _cur / R _avg ;

[0145] Among them, I _p R represents the concentrated characteristic value of precipitation periods. _cur R represents the current percentage of precipitation within the current year's comprehensive water demand window relative to the total annual precipitation. _avg This is the average percentage over many years, obtained based on historical data. This quotient represents the timing of rainfall distribution. A value below a constant of 1 indicates that rainfall deviates from the water-sensitive period; even with abundant rainfall throughout the year, crops may still face drought pressure during critical periods.

[0146] In one alternative implementation, the precipitation period is concentrated in one characteristic. _p The calculation can be further combined with the effective precipitation matching degree for supply and demand truncation correction. Specifically, the ratio of the effective precipitation matching degree to the proportion of precipitation within the current year's window is used as I. _pAlternative or weighted inputs can be used to enhance the accuracy of the representation of precipitation time series validity. Those skilled in the art can choose either the simple ratio method or the supply-demand truncation method to calculate the concentrated characteristics of precipitation periods, depending on the regional data conditions.

[0147] Step 602: Extract water inflow process guarantee features based on the fluctuation of equivalent water inflow guarantee characteristics during the key irrigation period of the year, or in other words, based on the deviation of multi-source water inflow process data in meteorological and hydrological data from the same period of the year during the key irrigation period.

[0148] Specifically, the equivalent cumulative inflow within the comprehensive water demand time window of the year is statistically analyzed. This includes the weighted sum of external inflows, reservoir discharges, and river diversions after equivalent conversion during the window period. The current percentage of this cumulative inflow within the window is then calculated as the total equivalent inflow for the year. Based on historical records over many years, an average percentage benchmark for inflows within the same time window is statistically analyzed. The current percentage is then divided by this average benchmark using a calculation unit. The corresponding formula is as follows:

[0149] I _w =S _cur / S _avg ;

[0150] Among them, I _w为 Incoming water process protection characteristic value, S _cur S represents the current proportion of the equivalent inflow within the current year's comprehensive water demand time window to the total equivalent inflow for the entire year. _avg This serves as a benchmark for the average proportion of water inflow within a given annual window, derived from historical sequence statistics. This characteristic value is used to assess whether the supply nodes of key water sources maintain temporal consistency with peak water demand periods for farmland.

[0151] Step 603: Based on the engineering storage capacity in the attributes of agricultural irrigation areas and its buffering capacity for supply and demand peak shifting, extract the spatiotemporal buffering characteristics of storage.

[0152] Obtain the engineering regulation and storage capacity index calculated based on the reservoir's beneficial storage capacity and the canal system's design flow. This index, as a fixed parameter, represents the static transfer potential of the irrigation area to intercept non-flood season precipitation through physical containers and shift it to the dry season for crops. Areas with larger facility parameters exhibit greater resistance to the misalignment between precipitation and water demand timing.

[0153] Step 604: After excluding the influence of the total annual water resources, the time distribution correction function is generated orthogonally by integrating the concentrated characteristics of precipitation periods, the characteristics of water inflow process guarantee, and the characteristics of regulation and storage spatiotemporal buffering. This function is specifically used to correct the misalignment deviation of the time distribution within the year.

[0154] Specifically, the information integration module employs a multiplicative weighted superposition structure to aggregate temporal feature parameters. Since the total amount adjustment coefficient already handles the correction of the absolute value of interannual precipitation, the aggregation function in this step is used to capture the pure spatiotemporal offset vector, and the corresponding calculation formula is:

[0155] f _t =1-λ _p ×(1-I _p )-λ _w ×(1-I _w) +λ _c ×I _cap ;

[0156] Among them, f _t The output value of the time distribution correction function is generated, where constant 1 is the baseline constant under the condition of a normal water year and uniform distribution, and λ is... _p To control the precipitation displacement elastic parameter of the precipitation time series fluctuation volume, I _p λ represents the concentrated characteristic value of the extracted precipitation period. _w为 Controlling the water supply deviation from the volume of the water supply's inflow displacement elasticity parameter, I _w为 The characteristic value of the incoming water process is λ. _c To adjust the buffer elasticity parameter, I _cap This is a characteristic value of the spatiotemporal buffer used to characterize the physical damping strength. For f... _t Perform boundary truncation to limit it to a preset reasonable range.

[0157] It is understandable that when the timing misalignment of precipitation is more severe (the smaller I_p) or the water supply guarantee is worse (the smaller I_w), f_t is less than 1. In the multiplicative correction, the original water consumption is reduced downward to remove the irrigation volume passively increased due to the temporal and spatial mismatch. When the engineering regulation and storage capacity is stronger, that is, when I_cap is larger, the positive correction term partially offsets the misalignment penalty, reflecting the role of physical buffer in mitigating the interference of timing misalignment.

[0158] Furthermore, the time distribution correction function includes preset elastic parameters to control the weights of each extracted feature. To improve the alignment of various elastic parameters with the actual drought resistance of specific regions and reduce regional distortions caused by artificially set uniform empirical values, an offline optimization program for closed-loop calibration of historical data is run independently before the system is officially put into verification calculation.

[0159] Step 6041: Obtain the historical verification residual dataset remaining in the historical statistical sample after only the annual water resource total amount correction.

[0160] Specifically, historical measured agricultural water consumption samples from each accounting unit are read over many years, and a division operation is performed using the total amount adjustment coefficient under known conditions for that unit. The resulting divisor represents the transient water consumption that has eliminated annual fluctuations in water availability but not the spatiotemporal distribution disorder within the year. Correspondingly, the difference between this transient water consumption and the calibrated baseline target water consumption is calculated, and the differences from all years are collected to construct a historical verification residual dataset. This residual set reflects the abnormal water consumption induced by unfavorable rainfall timing within the year.

[0161] Step 6042: Construct a regularization penalty term by combining the physical direction criterion of the spatiotemporal buffer characteristics of the storage capacity, so as to limit the abnormal weight inversion phenomenon caused by the difference in storage capacity.

[0162] To avoid the fitting algorithm outputting weight combinations that violate objective common sense due to insufficient samples in some short sequence segments, the modeling unit introduces inequality constraint boundaries to construct a penalty term matrix. In specific mapping relationships, the regulation and buffer elasticity parameter corresponding to large and medium-sized irrigation districts must be numerically greater than that of similar parameters for small irrigation districts. This constraint follows the objective mechanism of water conservancy facilities: the larger the volume of the engineering container, the higher its absorption and buffering efficiency for temporal misalignment should be. The system uses the coefficient modulus sum-of-squares normal form inherent in ridge regression to construct the basic normalized spatial constraint boundary.

[0163] Step 6043: With minimizing the historical verification residual dataset and satisfying the regularization penalty term as the joint optimization objective, the weight coefficients corresponding to each extracted feature are solved offline to obtain the preset elastic parameters adapted to the current region features.

[0164] Specifically, the computing host invokes the restricted nonlinear least squares algorithm module to load a sample matrix containing eigenvectors. Through the optimization engine, it searches for the optimal solution space that minimizes the sum of squared residuals, simultaneously calculating the algebraic sum of the regularization penalty terms. Iterative steps are allowed only if the weight evolution direction satisfies the preset physical direction boundary. The multivariate parameter vector output after convergence is the locked preset elasticity parameter.

[0165] Based on this, the parameter combination output by this offline calibration mechanism can maintain consistency with the inherent facility attributes of each agricultural irrigation area while filtering sample noise.

[0166] Example 7

[0167] When there are abnormal errors in the basic data, the verification coefficients output by the statistical model may deviate beyond the limits, causing the generated verification quantities to fail to meet the farmland water consumption baseline or exceed the physical evapotranspiration limit. Before decoupling correction, a baseline verification and protection procedure based on the farmland thermodynamic evapotranspiration dissipation mechanism is introduced.

[0168] In one possible implementation, before orthogonally decoupling the original agricultural water consumption using the total amount adjustment coefficient and the time distribution correction function, the following steps are also included:

[0169] Step 701: Based on meteorological and hydrological data and crop planting structure, simulate the theoretical net water demand boundary of the region using the physical mechanism of crop evapotranspiration.

[0170] Specifically, the system extracts multi-source environmental meteorological parameter sequences, including net radiation flux, average temperature, water vapor pressure, and wind speed, from the basic meteorological observation station matrix. The computational unit calls the FAO Penman-Monteith physical model to obtain potential evapotranspiration values ​​at the grid scale. The model's internal calculations involve conversion of wet and dry meter constants, derivation of saturated water vapor pressure difference, and estimation of aerodynamic impedance. This process belongs to the classic environmental meteorological energy balance calculation mechanism in agricultural irrigation, used to output the benchmark for the maximum water loss rate of crops under ideal growing conditions.

[0171] Accordingly, after obtaining the potential evaporation loss benchmark based on millimeter-thickness measurements, it is converted into an absolute volumetric water consumption limit at a macroscopic spatial scale to reduce systematic divergence errors caused by dimensional misalignment. The relevant volumetric and non-negative truncation calculation expressions are as follows:

[0172] E _net =max(0,K _c ×E _t -P _eff )×A _c / N _eff ;

[0173] Among them, E _net This outputs the calculated theoretical net water demand boundary, with values ​​in volume units. `max` is the non-negative zero-truncation forced screening operator. K _c E is the pre-entered crop variety moisture coefficient constant. _t P represents the potential evaporation thickness at the grid scale. _eff This refers to the effective precipitation retention thickness within the root zone depth. (A) _c N represents the total effective sown area corresponding to this crop. _eff The water utilization efficiency constant coefficient is calculated by comprehensively considering the facilities, irrigation systems, and field transportation links.

[0174] In this embodiment, the net shortage water depth is obtained by subtracting the effective precipitation from the potential loss, and then multiplied by the water-receiving area to achieve dimensional alignment. If the added thickness of natural precipitation can compensate for evapotranspiration loss, the calculated shortage value will be negative. The truncation operator is applied to set it to zero, thereby reducing the calculation deviation of falsely reported supplementary water volume when there is no urgent need for irrigation. Based on this, the theoretical absolute lower limit or upper limit of the actual farmland consumption in the region is defined.

[0175] Step 702: Compare the original agricultural water consumption with the theoretical net water demand boundary to assess the actual water demand satisfaction of the region.

[0176] The total agricultural water consumption statistics originally reported and registered for the current year are compiled and then divided by the theoretical net water demand boundary volume obtained earlier. The quotient of these two values ​​constitutes the actual water demand satisfaction parameter for the region. This parameter is used for dimensionless assessment of the physical match between the actual water supply volume and the thermodynamic limit of water consumption.

[0177] Step 703: Limit the reasonable range of values ​​for the total regulation coefficient by using the actual water demand satisfaction to obtain the limited total regulation coefficient, so as to cut off the coefficient distortion caused by statistical overestimation or severe irrigation shortage; use the limited total regulation coefficient to replace the original total regulation coefficient to participate in orthogonal decoupling correction.

[0178] Specifically, a preset empirical limit range is used as the basis for judging the reported water consumption data and identifying the decision boundary for whether there is systematic distortion. Corresponding limiting intervention actions are executed based on the deviation of this parameter from either side.

[0179] In the anti-inflated water demand processing branch, when the actual water demand satisfaction significantly exceeds the upper limit range, it identifies statistical errors in the original reported water intake of the target area, indicating inflated levels or large-scale irrational flooding. Even if the dual-drive physical statistical model calculates a high total volume adjustment coefficient, the system will perform a downgrade intervention, reducing it to the equivalent multiplier corresponding to meeting the upper limit threshold. This avoids providing a reasonable basis for wasteful water use and prevents weakening the fairness of the water-saving assessment after dilution and verification.

[0180] In the anti-false low water demand processing branch, when the actual water demand satisfaction falls below the lower limit range, the system identifies that the region faces physical water restriction and may experience large-scale abandonment of farmland, reduced yields, or statistical underreporting. Since the predictive response curve of the underlying demand-supply model cannot cover this type of agricultural degradation, the system cuts off the downward movement of the total adjustment coefficient calculated by the original model, implementing a bottom-line limit adjustment and raising it to the corresponding lower limit benchmark threshold. Through this type of forced bidirectional truncation operation, the corrected water consumption output falls within the boundary that conforms to the thermodynamic evaporation mechanism.

[0181] Example 8

[0182] Building upon Example 7 above, this example further explains the application of the verification results and the generation of evaluation indicators for the actual water-saving effectiveness in the region. The pure values ​​output by orthogonal decoupling and multiple boundary checks are fed back into the existing macro-level water resources ledger management system, completing the closed loop of the methodology in the scenario of evaluating the effectiveness of regional water conservation and intensive use.

[0183] In one possible implementation, after obtaining the verified agricultural water consumption, the following steps are also included:

[0184] Step 801: Obtain the industrial water consumption, domestic water consumption, ecological water consumption, and GDP of the target area during the same period.

[0185] Specifically, the external business interface module retrieves statistical yearbooks or water resources bulletins for the preset accounting year, extracting water usage records from other industry categories with the same time-sensitivity as the verification year. This includes the summary value of water usage ledgers for industrial enterprises above a certain size, the total value of water consumption for urban and rural residential use, and data on non-agricultural water use such as supplementary ecological discharge. Correspondingly, it obtains the financial accounting parameters of GDP under the same accounting caliber to construct a reference calculation and comparison base parallel to the current verification algorithm.

[0186] Accordingly, after completing the main algorithm's extension and acquisition of data from the external data environment, the system's built-in post-processing unit initiates a subroutine for evaluating the logical validity of the verification curve. It calculates the coefficient of determination and root mean square error across the entire spatial timescale. The simulated total water output sequence, generated based on offline calibration parameters, is compared with historically observed available water retention records. When this set of statistical evaluation quantities meets the predetermined empirical fitting threshold, it indicates that the physical parameter surface extracted for the current application year possesses computational coherence. The system is then allowed to continue issuing subsequent data integration and fusion operation commands.

[0187] Step 802: Summarize the verified agricultural water consumption, industrial water consumption, domestic water consumption, and ecological water consumption to obtain the verified annual total water consumption, which can eliminate the interference of natural hydrological abundance and scarcity and objectively represent the true level of water conservation and utilization in the region.

[0188] Specifically, through the computational accumulation module, the independent agricultural water consumption, after orthogonal verification and double-sided amplitude normalization due to physical evapotranspiration, is used as a replacement primitive. This primitive replaces the original agricultural water consumption values ​​in the bulletin system that are heavily influenced by meteorological factors, and is then algebraically accumulated with industrial, domestic, and ecological water consumption. The newly synthesized dataset isolates the disturbances caused by rainfall and water inflow fluctuations, and the dynamic fluctuations of its change curves can map the effectiveness of human intervention actions, including water-saving engineering renovations in agricultural irrigation areas.

[0189] Step 803: Using the verified annual total water consumption and GDP, calculate and obtain the verified water consumption per 10,000 yuan of GDP, and use this indicator as the standardized output result for judging the actual water-saving effect of the region.

[0190] By using a parameter conversion network, the total water volume after consolidation and normalization is divided by the corresponding economic accounting parameters to output a composite evaluation dependent variable reflecting production efficiency and resource matching degree. This indicator is reported to the external control system through a terminal display module or an electronic ledger upload module. Based on the final value obtained by this algorithm pipeline, the systematic error phenomenon caused by extreme climate years can be reduced, providing a stable benchmark for horizontal and vertical cross-referencing between different administrative assessment units and within the same management unit over a long historical span. Among them, the water consumption per 10,000 yuan of GDP after verification can be output to the water resources management information system in structured data format through the data interface module, or presented to management decision-makers in chart form through a visualization terminal. Those skilled in the art can choose an appropriate data delivery method according to the actual information construction needs.

[0191] Based on the same inventive concept as the above-described method embodiments, this embodiment also provides an electronic device and a computer-readable storage medium. This device and medium are used to support and execute the methods provided in the above embodiments. Their underlying implementation logic and technical effects have been detailed in the preceding chapters, and the computational details will not be repeated here. The following focuses on describing the hardware architecture and data carrier format.

[0192] The electronic equipment includes a memory and a processor. The memory is used to store computer programs, and the computer programs are executed by the processor to implement an agricultural water consumption verification method based on four-dimensional response relationships.

[0193] At the physical hardware level, the computing entity is equipped with a general-purpose processor capable of data processing and instruction scheduling. This type of processor manifests as a central processing unit (CPU), digital signal processor (DSP), or application-specific integrated circuit (ASIC). The memory, physically connected to this processor via a communication bus, consists of high-speed random access memory (RAM) and non-volatile memory modules. A pre-compiled machine-readable instruction set is embedded within the physical space of the memory.

[0194] When basic data inputs such as meteorological and hydrological data and agricultural irrigation area attributes are received, the processor extracts and loads the aforementioned instruction set from memory via the system bus. On the silicon-based hardware platform, it sequentially instantiates and executes operations such as precipitation climate response feature extraction, dual-constraint parameterized model construction, offline regularization calibration, and amplitude limiting constraints and verification. Various communication interface modules, in coordination with the input / output bus and the processor's timing, complete the reading of raw water usage data and the external transmission of verified water consumption indicators.

[0195] A computer-readable storage medium stores a computer program that, when executed by a processor, implements an agricultural water consumption verification method based on a four-dimensional response relationship. Specifically, the computer-readable storage medium is a physical medium used to carry logical instructions and multi-dimensional operational functions, possessing the characteristics of long-term information retention and protection against power loss. In terms of selection, it covers magnetic storage media, optical storage substrates, and solid-state storage arrays based on flash memory technology. The binary code fragments stored within this physical carrier can be recognized and parsed by the read drives in the aforementioned electronic devices or distributed server clusters. By executing the program sequence loaded on this computer-readable storage medium, the computing unit can map the hydrological and physical mechanisms of farmland into electronic signal processing actions, thereby supporting the automated operation of the entire four-dimensional response verification system for precipitation, water inflow, crops, and irrigation water demand.

[0196] It should be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the present invention will not describe the various possible combinations separately.

Claims

1. A method for verifying agricultural water consumption based on the two-dimensional response relationship between precipitation and inflow, characterized in that, include: Obtain the raw agricultural water consumption of the target area, as well as basic data including meteorological and hydrological data, agricultural irrigation area attributes, and crop planting structure; Based on meteorological and hydrological data, precipitation climate response characteristics that characterize precipitation fluctuations are extracted, and equivalent water supply guarantee characteristics that integrate the differences in multi-source water supply are extracted by combining agricultural irrigation area attributes. Based on crop planting structure, a comprehensive water demand time window is extracted to characterize the regional water-sensitive cycle. Based on precipitation climate response characteristics and equivalent water inflow security characteristics, a total amount adjustment coefficient is generated to describe the degree of deviation of annual total water resources. Based on the distribution of meteorological and hydrological data within the comprehensive water demand time window, and combined with the engineering storage capacity in the attributes of agricultural irrigation districts, a time distribution correction function is generated to describe the degree of deviation of the time distribution of water resources within the year. Based on this, the original agricultural water consumption was orthogonally decoupled and corrected to obtain the verified agricultural water consumption.

2. The method according to claim 1, characterized in that, Extracting precipitation climate response characteristics, including: Historical precipitation sequences were extracted from meteorological and hydrological data, and historical agricultural water use sequences were obtained from raw agricultural water consumption data. Perform nonstationarity tests on historical agricultural water use sequences to identify the technological progress trends implied in the sequences; When a technological progress trend is identified, the historical agricultural water use sequence is stripped to remove the baseline offset interference caused by human water-saving technologies, and precipitation climate response characteristics are constructed based on the historical precipitation sequence. Otherwise, precipitation climate response characteristics are constructed based on the statistical correlation between historical precipitation sequences and historical agricultural water use sequences.

3. The method according to claim 2, characterized in that, Constructing precipitation climate response characteristics, including: A data smoothing algorithm was used to decompose the historical agricultural water use series into a technological progress trend component that characterizes the systematic changes in water use efficiency, and a climate response residual component that characterizes the impact of interannual meteorological fluctuations. The climate response residual component is normalized using the technological progress trend component to obtain the climate response deviation rate. Extracting annual precipitation frequencies based on historical precipitation sequences; A monotonic mapping relationship between precipitation frequency and climate response deviation rate is established, and this monotonic mapping relationship is used as a feature of precipitation climate response.

4. The method according to claim 1, characterized in that, Extracting equivalent water supply security features, including: Multi-source water inflow data were extracted from meteorological and hydrological data, including components of water inflow from natural rivers, components of water inflow from reservoirs, and components of water transferred from other regions. Based on the characteristics of agricultural irrigation areas, the physical supply stability of various water components is assessed, and the corresponding independent guarantee coefficients are determined. By using each independent guarantee coefficient to perform a weighted transformation on the original water volume of the corresponding water components, an equivalent water supply guarantee characteristic reflecting the actual supply reliability is obtained.

5. The method according to claim 1, characterized in that, Extract the comprehensive water demand time window, including: Analyze the crop planting structure to determine the various crop types covered in the region and the planting area ratio of each crop; Based on crop water requirement parameters, the single crop water requirement period when the average daily water requirement of each crop type reaches its peak during the entire growth period is identified; Using the planting area ratio of each crop as a weighting coefficient, the water demand periods of all single crops are weighted and integrated to construct a weighted water demand intensity distribution that represents the comprehensive irrigation demand of the entire region. Extracting the comprehensive water demand time window based on the weighted water demand intensity distribution.

6. The method according to claim 1, characterized in that, Generate the total adjustment coefficient, including: The precipitation climate response characteristics are mapped into demand-driven response units that characterize the surge in irrigation demand under conditions of sufficient water resources. Based on the equivalent water supply guarantee characteristics, a supply constraint response unit is constructed to characterize the passive decline state of actual water consumption under water source constraints. By coupling and synthesizing the demand-driven response unit and the supply-constrained response unit, a total adjustment coefficient that decouples the spatiotemporal misalignment factors is generated.

7. The method according to claim 1, characterized in that, Generate a time distribution correction function, including: Based on the degree of deviation in the temporal distribution of meteorological and hydrological data within and outside the comprehensive water demand time window, the concentrated characteristics of precipitation periods are extracted. Based on the deviation of multi-source water inflow process data from the annual key irrigation period and the same period of the year in meteorological and hydrological data, the water inflow process guarantee characteristics are extracted. Based on the engineering storage capacity of agricultural irrigation districts and its buffering capacity against supply and demand peak shifts, the spatiotemporal buffering characteristics of storage are extracted. After excluding the influence of the total annual water resources, the time distribution correction function is generated orthogonally by integrating the characteristics of concentrated precipitation periods, the characteristics of water inflow process guarantee, and the characteristics of spatiotemporal buffering of water storage.

8. The method according to claim 1, characterized in that, Before performing orthogonal decoupling correction, the following is also included: Based on meteorological and hydrological data and crop planting structure, the theoretical net water demand boundary of the region is simulated using the physical mechanism of crop evapotranspiration. The actual water demand satisfaction of the region is assessed by comparing the original agricultural water consumption with the theoretical net water demand boundary. By using the actual water demand satisfaction rate to limit the reasonable range of values ​​for the total quantity regulation coefficient, the limited total quantity regulation coefficient is obtained; the limited total quantity regulation coefficient is then used to replace the original total quantity regulation coefficient in orthogonal decoupling correction.

9. The method according to claim 1, characterized in that, After obtaining the verified agricultural water usage, the following is also included: Obtain the industrial water consumption, domestic water consumption, ecological water consumption, and GDP of the target area during the same period; The verified agricultural water consumption, industrial water consumption, domestic water consumption, and ecological water consumption are summarized to obtain the verified annual total water consumption, which can eliminate the interference of natural hydrological abundance and scarcity and objectively represent the true level of water conservation and utilization in the region. By using the verified annual total water consumption and GDP, the verified water consumption per 10,000 yuan of GDP is calculated and used as a standardized output result for evaluating the actual water-saving effect of the region.

10. The method according to claim 1, characterized in that, Acquire basic data for the target area and delineate accounting units, including: County-level administrative divisions are used as the conventional spatial accounting unit; Based on the characteristics of the regional irrigation water source structure and the degree of engineering support, the conventional accounting units are clustered into large and medium-sized irrigation areas, small and rain-fed agricultural areas, and inter-regional water transfer irrigation areas. For different irrigation district types after classification, corresponding differentiated operation parameters are assigned to the dynamic blocking rule in the total regulation coefficient and the feature fusion calculation in the time distribution correction function.