A unit area irrigation water consumption estimation method and system based on multi-source data fusion

Abstract

The application discloses a unit area irrigation water consumption estimation method and system based on multi-source data fusion, relates to the field of irrigation water consumption estimation, and through fusing multi-source heterogeneous data such as remote sensing images, meteorological data, soil humidity and crop growth information, uses data assimilation to reduce observation and model errors, estimates crop evapotranspiration under the principle of energy balance; adopts a water balance method to estimate daily scale irrigation water consumption; introduces a crop actual irrigation area identification model based on deep learning, effectively distinguishes actual irrigation areas of different crop types, and dynamically reflects the influence of spatial and temporal changes on irrigation efficiency. The application can realize high-precision and rapid estimation of unit area irrigation water consumption of single crops and multiple crops in different regional scales, significantly improves the timeliness and accuracy of irrigation water quota evaluation, has strong adaptability to climate change, soil differences and crop diversity, and provides strong technical support for precise irrigation and intelligent water conservancy management.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of irrigation water quantity estimation, and more particularly to a unit area irrigation water quantity estimation method and system based on multi-source data fusion. BACKGROUND

[0002] At present, irrigation water quantity is an important technical basis for realizing farmland water saving, improving water resource utilization efficiency and guaranteeing sustainable development of agriculture. However, there are still obvious deficiencies in the timeliness of data acquisition, estimation accuracy and adaptability to complex environmental conditions in the current irrigation water monitoring and irrigation water quantity formulation. Especially in the region where multiple climate types, multiple soil conditions and multiple crop planting structures coexist, the comprehensive influence of meteorological, soil moisture and crop growth and other factors makes it difficult for the traditional irrigation water quantity formulation method to balance the precision and timeliness. Therefore, developing a new method that can fuse multi-source heterogeneous data, improve the estimation accuracy of unit area irrigation water quantity and adapt to complex changing environment has important significance for promoting precise irrigation management, improving water resource allocation efficiency and guaranteeing food security.

[0003] Guizhou Water Conservancy Scientific Research Institute discloses a method for calculating water quota in karst mountainous areas (CN2024108089318). The method divides the region into zones based on natural geography, hydrology, meteorology, and irrigation conditions, and selects typical zone sample plots. According to the precipitation conditions, hydrological year types are set and rainfall frequency analysis is performed, and the most unfavorable year is selected. Combined with the full growth period water requirement law of the main characteristic crops in the region and the representative annual rainfall process, the water balance analysis method is used to calculate the irrigation quota for the full growth period, and the basic quota is calculated combined with the water use coefficient of the mountainous water distribution system. This method is suitable for specific complex topography, but the data source mainly depends on field monitoring and statistical data, lacks multi-source remote sensing, meteorological and soil information fusion, and has insufficient dynamic updating capability; and it is not flexible enough to respond to rapid changes in meteorological and crop conditions, and its promotion range is limited. Pearl River Water Conservancy Scientific Research Institute discloses a method for estimating total agricultural irrigation water use based on remote sensing (CN2019111267540), which identifies irrigation areas and crop distribution by obtaining remote sensing images, and calculates total water use by combining planting structure, crop water quota and effective precipitation. This method improves spatial coverage and accuracy, but the water quota is used as a static input parameter, and the dynamic optimization of irrigation water quota itself is not addressed; the precipitation utilization rate and other coefficients are empirical values, and cannot be adjusted in real time based on soil moisture, crop growth stage and meteorological changes; remote sensing is only used for area and crop type identification, and lacks multi-dimensional data fusion such as evapotranspiration and soil moisture, making it difficult to accurately reflect the actual water consumption process. In general, existing technologies have certain practical value in specific application scenarios, but generally have the following shortcomings: single data dimension, insufficient integration of multi-source heterogeneous information, lack of dynamic updating mechanism, insufficient response to changes in weather and crops, and limited adaptability in complex multi-factor environments. These problems limit the widespread promotion and efficient application of unit area irrigation water consumption estimation in modern precision irrigation and intelligent water management.

[0004] Therefore, how to achieve high-precision and rapid estimation of unit area irrigation water consumption for single crops and multiple crops in different scale regions, and thus significantly improve the timeliness and accuracy of irrigation water quota evaluation, is a problem that needs to be solved by those skilled in the art. SUMMARY

[0005] Therefore, the application provides a unit area irrigation water consumption estimation method and system based on multi-source data fusion, which fuses multi-source heterogeneous data such as remote sensing images, meteorological data, soil moisture and crop growth information, reduces observation and model errors by using data assimilation technology, and estimates crop evapotranspiration with high precision under the principle of energy balance; the water balance method is used to estimate the daily scale irrigation water consumption; a crop actual irrigation area identification model based on deep learning is further introduced to effectively distinguish the actual irrigation area of different crops and dynamically reflect the influence of spatial and temporal changes on irrigation efficiency. Through the above steps, the application can realize high-precision and rapid estimation of unit area irrigation water consumption of single crops and multiple crops in different scale regions, significantly improve the timeliness and accuracy of irrigation water quota evaluation, and have strong adaptability to climate change, soil difference and crop diversity, thereby providing strong technical support for precise irrigation and intelligent water conservancy management.

[0006] In order to achieve the above object, the application adopts the following technical solutions:

[0007] A unit area irrigation water consumption estimation method based on multi-source data fusion comprises the following steps:

[0008] Vegetation index, meteorological data and soil moisture data in an irrigation area are collected, and the collected vegetation index, meteorological data and soil moisture data are processed to obtain daily scale vegetation index, daily scale meteorological data and daily scale soil moisture;

[0009] The daily scale vegetation index, meteorological data and soil moisture are processed by using statistical downscaling technology, so that all the data have a unified spatial resolution;

[0010] On the basis of the data with a unified spatial resolution, daily scale evapotranspiration is estimated based on a SEBAL (Surface Energy Balance Algorithm for Land) model;

[0011] On the basis of the obtained daily scale evapotranspiration, daily scale unit area irrigation water consumption is estimated by using a water balance method with the aid of data such as runoff, leakage and soil water storage change;

[0012] A deep learning and ensemble model technology is used to identify the actual irrigation area;

[0013] Based on the calculated data, single crop unit area irrigation water consumption and crop comprehensive unit area irrigation water consumption are calculated.

[0014] Optionally, the collection of vegetation index, meteorological data and soil moisture data in an irrigation area and the processing of the collected vegetation index, meteorological data and soil moisture data to obtain daily scale vegetation index, daily scale meteorological data and daily scale soil moisture are specifically as follows:

[0015] Vegetation index: Extract regional-scale normalized difference vegetation index, enhanced vegetation index, greenness index, normalized water index, soil-adjusted vegetation index using MODIS satellite sensors to achieve large-scale dynamic monitoring of crop growth; At the same time, multi-spectral unmanned aerial vehicles are used for aerial survey at the key growth period of crops to obtain high-spectral images with a ground resolution of <5cm, and simultaneously collect canopy spectral reflectance, soil moisture spatial distribution and visible irrigation marks. These high-precision data (collected data) are corrected by a scale conversion model to obtain daily-scale vegetation index;

[0016] Weather data: Fusion of three types of data sources, first, integrate the daily-scale continuous observation data of meteorological stations in the irrigation area, including rainfall, surface temperature, air temperature, wind speed and relative humidity parameters; Second, introduce the ERA5 reanalysis data set of the European Center for Medium-Range Weather Forecasts to supplement global meteorological elements; Based on the thermal infrared sensor of Landsat-8 satellite, the split window algorithm is used to retrieve the land surface temperature, and finally the ensemble Kalman filter method is used to integrate multi-source daily-scale meteorological data;

[0017] Soil moisture: In order to obtain the dynamic of root layer water, a data assimilation fusion method of multi-source satellite data is used, combined with Noah-MP land surface process model and ensemble Kalman filter data assimilation framework; Under this framework, first, soil moisture remote sensing data from different satellite sensors, Sentinel-1, medium resolution imaging spectrometer, soil moisture and salinity satellite and interferometric synthetic aperture radar are obtained, which cover soil moisture information with different spatial and temporal resolutions; Enhanced spatio-temporal adaptive reflectance fusion model is used for data fusion; Enhanced spatio-temporal adaptive reflectance fusion model combines different remote sensing data with the simulation results of Noah-MP land surface process model to dynamically adjust the state variables of the model and optimize the estimation of soil moisture in real time, and then obtain daily-scale soil moisture.

[0018] Optionally, the statistical downscaling technique is used to process the daily-scale vegetation index, daily-scale meteorological data and daily-scale soil moisture respectively, so that all data have a unified spatial resolution, specifically including:

[0019] The multi-source data downscaling processing framework is adopted to realize spatial unification of different resolution data; the input data includes 30-meter resolution vegetation index generated by Landsat-8, 30-meter daily meteorological data fused by an enhanced spatio-temporal adaptive reflectance fusion model, and soil moisture fusion data; the processing procedure firstly performs feature engineering to extract the time variation gradient of the vegetation index and the terrain adjustment factor; then, interpolation and resampling methods are adopted to unify all the input data to the spatial resolution of 30 meters; finally, the soil humidity is optimized by using physical constraints to generate soil humidity data with unified spatial resolution, so that the spatial consistency of the data from different sources is ensured.

[0020] Optionally, the daily evapotranspiration is estimated based on the SEBAL model on the basis of the data with unified spatial resolution, so that the water requirement of the crops is calculated, and specifically, the method comprises the following steps:

[0021] All the input data are unified to the spatial resolution of 30 meters through multi-source data fusion and downscaling processing; in the calculation process of the SEBAL model, firstly, the net radiation is calculated by using the remote sensing data and the meteorological data through the principle of the surface energy balance; secondly, the surface heat flux and the sensible heat flux are calculated by combining the surface reflectivity and the solar radiation information; finally, the SEBAL model calculates the daily evapotranspiration by using the energy balance equation on the basis of the net radiation, the surface heat flux and the sensible heat flux.

[0022] Optionally, the specific calculation formula of the SEBAL model is as follows:

[0023] ;

[0024] ;

[0025] ;

[0026] ;

[0027] Wherein, is the net radiation, G is the surface heat flux, H is the sensible heat flux, is the evapotranspiration, is the difference between the aerodynamic surface temperature and the reference height temperature, is the surface temperature, a and b are empirical coefficients, is the air density, is the air specific heat, is the aerodynamic resistance of the heat transfer between the aerodynamic surface and the reference height, is the surface albedo, and are the incident and outgoing long-wave radiation, is the surface thermal emissivity, is the actual direct and diffuse solar radiation flux reaching the earth's surface;

[0028] Through the calculation of the above formula, the evapotranspiration of crops is obtained.

[0029] Optionally, the calculation formula for estimating the daily irrigation water consumption on the basis of the obtained daily evapotranspiration is:

[0030] I t = ET t +R t +D t -P t + ΔS t ;

[0031] In the formula, P t is the precipitation, I t is the daily irrigation amount, ET t is the evapotranspiration, R t is the runoff, D t is the deep seepage, Δ S t is the change of soil water storage.

[0032] Optionally, the deep learning and integrated model technology for identifying the actual irrigation area specifically includes:

[0033] In the model training stage, first, the input data is divided into different crop types, and a deep learning model is trained for each crop type; a U-Net model is used for fine segmentation of the irrigation area;

[0034] In the integrated model part, the outputs of multiple models are fused using an integrated learning method, and the advantages of each model are combined using weighted averaging or voting mechanism;

[0035] After the training is completed, feature variable importance analysis is performed, and a tool is used to analyze the contribution of each feature variable in the model to the identification of the irrigation area;

[0036] The result output: the model will output the monthly actual irrigation area data, and all results will be spatially aggregated to the field scale.

[0037] Optionally, the calculation formula of the single crop unit area irrigation amount and the unit area irrigation amount of multiple crops is:

[0038] Single crop irrigation amount calculation: according to crop type, growth stage and climate division, the basic irrigation amount W is formulated base , dynamic correction factor:

[0039] W adj =W base × ) ) ;

[0040] Among them, ET actual is the actual ET reversed by remote sensing, ET ref is the reference ET, S soil is the real-time soil moisture, S threshold is the irrigation trigger threshold;

[0041] Using GIS spatial overlay analysis, the unit area irrigation water consumption of different crops is weighted and averaged according to the planting area to generate county or irrigation district level unit area irrigation water consumption of multiple crops:

[0042] W comp = ;

[0043] In the formula, W i is the correction unit area irrigation water consumption of crops i , A i is the planting area.

[0044] A unit area irrigation water consumption estimation system based on multi-source data fusion, comprising:

[0045] A data acquisition module acquires vegetation index, meteorological data and soil moisture data in the irrigation area, and processes the acquired vegetation index, meteorological data and soil moisture data respectively to obtain daily scale vegetation index, daily scale meteorological data and daily scale soil moisture;

[0046] A data downscaling module uses statistical downscaling technology to process the daily scale vegetation index, daily scale meteorological data and daily scale soil moisture, so that all data have a unified spatial resolution;

[0047] an evapotranspiration estimation module, based on the obtained data with uniform spatial resolution, estimates daily evapotranspiration based on SEBA model; L a daily evapotranspiration estimation module;

[0048] an irrigation water volume estimation module, based on the obtained daily evapotranspiration, with the aid of runoff, leakage and soil water storage change data, estimates daily irrigation water volume per unit area by using water balance method;

[0049] an irrigation area output module, which identifies actual irrigation area by using deep learning and ensemble model technology;

[0050] an irrigation volume output module, which calculates irrigation volume per unit area of single crop and comprehensive irrigation volume per unit area of multiple crops based on the obtained data.

[0051] Compared with the prior art, the technical scheme of the present application provides a unit area irrigation water volume estimation method and system based on multi-source data fusion, which fuses multi-source heterogeneous data such as remote sensing images, meteorological data, soil moisture and crop growth information, reduces observation and model errors by using data assimilation technology, and estimates crop evapotranspiration with high precision under the principle of energy balance; estimates daily irrigation water volume by using water balance method; further introduces a crop irrigation area identification model based on deep learning, which effectively distinguishes irrigation areas of different crops and dynamically reflects the influence of spatial and temporal changes on irrigation efficiency. Through the above steps, the present application can realize high-precision and rapid estimation of unit area irrigation water volume of single crop and multiple crops, significantly improve the timeliness and accuracy of irrigation water quota evaluation, and has strong adaptability to climate change, soil difference and crop diversity, providing strong technical support for precise irrigation and intelligent water conservancy management. BRIEF DESCRIPTION OF DRAWINGS

[0052] In order to more clearly illustrate the technical scheme in the embodiments of the present application or the prior art, the drawings needed in the embodiments or prior art description will be briefly introduced as follows. Obviously, the drawings in the following description are only embodiments of the present application, and those skilled in the art can obtain other drawings according to the provided drawings without creative labor.

[0053] Figure 1 The equation flowchart provided by the present application. DETAILED DESCRIPTION

[0054] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be described clearly and completely below. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all the other embodiments obtained by those of ordinary skill in the art without creative efforts should fall into the scope of the present application.

[0055] The embodiment of the present application discloses a unit area irrigation water consumption estimation method based on multi-source data fusion, as shown in the formula (1), comprising: Figure 1

[0056] Collecting vegetation index, meteorological data and soil moisture data in the irrigation area, and processing the collected vegetation index, meteorological data and soil moisture data to obtain daily scale vegetation index, daily scale meteorological data and daily scale soil moisture;

[0057] Processing the daily scale vegetation index, daily scale meteorological data and daily scale soil moisture by using statistical downscaling technology, so that all the data have a unified spatial resolution;

[0058] Based on the data with unified spatial resolution, estimating daily scale evapotranspiration based on the SEBAL model;

[0059] Based on the obtained daily scale evapotranspiration, auxiliary runoff, leakage and soil water storage change data, and using water balance method to estimate daily scale unit area irrigation water consumption;

[0060] Using deep learning and integrated model technology to identify the actual irrigation area;

[0061] Based on the calculated data, calculating single crop unit area irrigation water consumption and multiple crop comprehensive unit area irrigation water consumption.

[0062] In a specific embodiment, the multi-source data acquisition and fusion specifically comprises:

[0063] ​Vegetation Index: Extract regional-scale vegetation parameters such as normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), greenness index (GI), normalized water index (NDWI), and soil-adjusted vegetation index (SAVI) using MODIS satellite sensors (spatial resolution 250-1000 m, revisit period 1 day) to achieve large-scale dynamic monitoring of crop growth. To improve data accuracy, multi-spectral unmanned aerial vehicles are used for aerial survey during the key growth period of crops, with flight height controlled at 50 m to obtain high-spectral images with ground resolution <5 cm. Simultaneously, canopy spectral reflectance, soil moisture spatial distribution, and visible irrigation marks are collected. These high-precision data are corrected using satellite vegetation index through scale conversion models, and ground GPS positioning sampling points (horizontal accuracy ±3 cm) are used for irrigation event identification accuracy verification (producer's accuracy and user's accuracy can be calculated using the confusion matrix).

[0064] Weather Data: The construction of the weather database integrates three types of data sources. First, daily-scale continuous observation data from weather stations within the irrigation area are integrated, including rainfall (pre), land surface temperature (lst), air temperature (temp), wind speed, and relative humidity. Second, the ERA5 reanalysis dataset (spatial resolution 0.25°x0.25°, time resolution 1 hour) from the European Centre for Medium-Range Weather Forecasts is introduced to supplement global meteorological elements such as solar radiation and atmospheric pressure. Third, the split window algorithm is applied to the thermal infrared sensor (TIRSBand 10-11) of the Landsat-8 satellite to retrieve land surface temperature (LST). Finally, the ensemble Kalman filter (EnKF) method is used to integrate multi-source daily meteorological data: based on the observation data from weather stations within the irrigation area, the daily products (spatial resolution 0.25°) from the ERA5 reanalysis dataset of the European Centre for Medium-Range Weather Forecasts are integrated, and the daily land surface temperature products (30 m resolution) retrieved from the Landsat-8 satellite are used.

[0065] Soil Moisture: To obtain root layer water dynamics, a data assimilation method is used to integrate multi-source satellite data, combined with the Noah-MP land surface process model and the ensemble Kalman filter (EnKF) data assimilation framework. Under this framework, soil moisture remote sensing data from different satellite sensors (such as the Soil Moisture Active Passive satellite (SMAP), Sentinel-1, Moderate Resolution Imaging Spectrometer (MODIS), Soil Moisture and Ocean Salinity satellite (SMOS), etc.) and interferometric synthetic aperture radar (InSAR) are preprocessed, covering different spatial resolutions (e.g., 10 km, 1 km) and time resolutions (e.g., 1 day or less). These data contain soil moisture information at different levels, and each data source has different spatial and temporal resolutions, so standardization and spatial interpolation processing are needed to ensure the uniformity of different data.

[0066] Next, the Enhanced Spatio-Temporal Adaptive Reflectance Fusion Model (ESTARFM) is used for data fusion. ESTARFM combines different remote sensing data with the simulation results of the Noah-MP land surface process model, dynamically adjusts the state variables of the model (such as the depth distribution of soil moisture), and optimizes the estimation of soil moisture in real time. The specific steps include: first, using the Noah-MP model to make a preliminary prediction of soil moisture; then, ESTARFM updates the model according to real-time satellite remote sensing data, corrects the prediction results, so that the final output of soil moisture data is more accurate. ESTARFM combines low spatial resolution remote sensing images with high spatial resolution data, effectively reduces model errors, and improves estimation accuracy.

[0067] To further improve the accuracy and reliability of the model, historical meteorological data and soil type information will also be used to assist the data assimilation process. Specifically, the precipitation, temperature, humidity data of the weather station are used for model input to better reflect the changes of soil moisture. In addition, the spatial distribution of soil types will also be included in the assimilation framework to consider the effects of different soil types on water retention and conduction.

[0068] Spatial statistical downscaling method: The invention adopts a multi-source data downscaling processing framework to realize the spatial unification of different resolution data. The input data includes 30-meter resolution vegetation index generated by Landsat-8, 30-meter daily meteorological data (temperature, precipitation, etc.) fused by ESTARFM, and soil moisture fusion data. The processing flow first performs feature engineering to extract the time change gradient of vegetation index and terrain adjustment factors (such as elevation, slope). Then, interpolation and resampling methods are used to unify all data to 30-meter spatial resolution. Finally, the soil humidity is optimized using physical constraints to generate soil humidity data with unified spatial resolution. This method ensures the spatial consistency of data from different sources and provides support for accurate irrigation management.

[0069] The role of this part of the content is:

[0070] Vegetation index, meteorology data, soil moisture, and environmental information (e.g., elevation) are based on multiple optical remote sensing satellites (Landsat / MODIS / Sentinel-2 MSI), microwave remote sensing satellites (Sentinel-1A), meteorology data, statistical data, and sensor data. The different sources and different temporal and spatial resolution data are converted into daily scale, unified spatial resolution vegetation index, meteorology data, and soil moisture data through data fusion (ESTARFM), data assimilation (Noah-MP+EnKF), and spatial statistical downscaling methods. This part solves the problems of inconsistent scales, non-uniform precision, and large differences in temporal and spatial resolution between multi-source heterogeneous data, providing a unified and reliable data basis for subsequent high-precision irrigation water estimation.

[0071] In one specific embodiment, the energy balance principle method for crop water requirement specifically includes:

[0072] The SEBAL model is used to estimate crop water requirement, and the goal is to generate daily scale evapotranspiration (ET) data. The SEBAL model accurately estimates evapotranspiration based on the physical principle of surface energy balance, and thus calculates the crop water requirement. In the previous step, all input data (including vegetation index, meteorology data, soil moisture, etc.) are unified to 30-meter spatial resolution through multi-source data fusion and downscaling processing, in order to fully utilize high-resolution remote sensing images. Through these data assimilation processes, the consistency and high precision of the input data are ensured, making the final estimated crop water requirement result more reliable and accurate. In the SEBAL model calculation process, first, the remote sensing data (such as Landsat-8 reflectivity and surface temperature) and meteorology data (such as precipitation, air temperature, etc.) are used to calculate the net radiation (Rn) ) through the surface energy balance principle. Net radiation is the total radiant energy received by the surface, which is partially converted into surface heat after reflection, absorption, and transmission. Then, combined with surface reflectivity, solar radiation, etc., the surface heat flux (G) and sensible heat flux (H) are calculated. Sensible heat flux is the exchange process between surface heat and air, while surface heat flux is the main source of surface temperature change. Next, the SEBAL model uses these data to accurately calculate the evapotranspiration (ET) ), i.e., the crop water requirement, through the energy balance equation. The specific calculation formula is:

[0073] ;

[0074] ;

[0075] ;

[0076] ;

[0077] where, is the net radiation, G is the ground heat flux, H is the sensible heat flux, is the evapotranspiration, is the difference between the aerodynamic surface temperature and the reference height temperature, is the surface temperature, a and b are empirical coefficients, is the air density, is the specific heat of air, is the aerodynamic resistance of heat transfer between the aerodynamic surface and the reference height, is the surface albedo, and are the incoming and outgoing longwave radiation, is the surface emissivity of heat, is the actual direct and diffuse solar radiation flux reaching the Earth's surface;

[0078] Through the calculation of the above formula, the evapotranspiration of crops is obtained.

[0079] Finally, the evapotranspiration data output by the SEBAL model has a spatial resolution of 30 meters and is generated according to a daily time scale. The evapotranspiration calculation results of each pixel are accurately assigned and analyzed according to the climate and crop type in the specific area. This method can provide high-precision estimation of crop water requirements by accurately simulating the surface energy balance, providing strong data support for precise irrigation decision-making, water resource scheduling, and agricultural water management. This result not only effectively improves the use efficiency of water resources, but also provides a scientific basis for optimizing agricultural irrigation strategies, promoting the implementation of precision agriculture and sustainable water resource management.

[0080] The role of this part of the content is:

[0081] Using the high-precision multi-source fusion input dataset generated in the first step, the SEBAL (Surface Energy Balance Algorithm for Land) method is used to accurately estimate the crop evapotranspiration through surface energy balance calculation, using remote sensing data (such as surface temperature, reflectivity, etc.) and meteorological data. This method solves the problem of insufficient accuracy of traditional empirical formula estimation and the inability to dynamically reflect changes in weather and surface conditions, achieving accurate calculation of daily crop water requirements.

[0082] In a specific embodiment, the irrigation water estimation based on the water balance method specifically includes:

[0083] The daily irrigation water consumption is estimated using the water balance method, and the input data includes the 30-meter resolution daily actual evapotranspiration (AET) data estimated based on the energy balance principle in the second step, combined with multi-source auxiliary data such as local meteorological station interpolation data, runoff data, leakage data, and soil moisture data (30-meter resolution) obtained in the foregoing downscaling step. The above data is used to calculate the daily irrigation water consumption through the water balance equation, as follows:

[0084] I t = ET t +R t +D t -P t + ΔS t ;

[0085] In the formula, P t P is the precipitation, I t I is the daily irrigation, ET t ET is the evapotranspiration, R t R is the runoff, D t L is the deep percolation, Δ S t ΔSW is the change of soil water storage.

[0086] To enhance the physical rationality of the calculation results, the present application introduces double constraints: one is to use evapotranspiration (ET) data as an energy balance constraint to ensure the crop water requirement mechanism; the other is to use soil moisture data as a mass conservation constraint, combined with threshold filtering of abnormal values (for example, when the soil moisture reaches saturation, limit the overestimation of percolation). These two constraints help to ensure that the irrigation water consumption estimation conforms to the actual law of hydrological cycle, and further improves the accuracy of the results.

[0087] The specific implementation steps include three parts: first, in the preprocessing stage, all data are aligned to the 30-meter grid, and the cloud coverage affected area is removed to ensure the spatial consistency of the data; second, in the core calculation stage, an iterative optimization algorithm (based on the gradient descent method) is used to solve the water balance equation, and the output is corrected after each iteration by applying the constraint term; finally, in the post-processing stage, the results are calculated to the field scale through spatial aggregation, and the time series are smoothed by 7-day moving average filtering to remove noise, ensuring the stability and reliability of the results.

[0088] The verification process of the present application will be carried out in Inner Mongolia Lengkou Irrigation Area and Henan Yuzhuang Irrigation Area, and the focus is to use the measured irrigation amount data of intelligent water meters for comparison. Through spatial alignment and time consistency verification, the error between the irrigation water amount output by the model and the measured data is evaluated, and indicators such as RMSE, MAE and R2 are used for error analysis. In addition, the ten-fold cross-validation method is used to evaluate the stability and generalization ability of the model. Further, by comparing with the measured soil moisture and meteorological data, the estimation accuracy of the change of soil water storage in the water balance equation is verified. Finally, spatial aggregation and field scale application are carried out, and 7-day moving average filtering is used to smooth the time series, ensuring the accuracy and practicality of the model.

[0089] This part is based on the second step of daily evapotranspiration data, combined with runoff data, leakage data and precipitation data, water balance analysis is carried out, and daily irrigation water amount is accurately calculated by simulating soil moisture change and hydrological process. This method solves the problem that soil water storage change and water budget balance are not fully considered in existing irrigation water estimation, and improves the scientificity and accuracy of water consumption estimation.

[0090] In a specific embodiment, the actual irrigation area monitoring specifically includes:

[0091] The present application uses deep learning and integrated model technology to identify the actual irrigation area, and the goal is to use multi-source data fusion method to generate high-precision irrigation area monitoring data. The input data includes the high-resolution multi-source data set generated in the early stage, which includes vegetation index (NDVI), evapotranspiration (ET), land surface temperature, meteorological data (such as air temperature, precipitation, etc.), soil moisture, spectrum, drought index, SAR data, elevation, slope, land use information and phenological characteristics, etc. All data have been processed by the previous step to reduce the scale, and unified to 30 meters spatial resolution to ensure data consistency and high precision.

[0092] In the model training stage, first, the input data is divided into different crop types (such as paddy field, corn, etc.), and a deep learning model is trained for each crop type. U-Net model is used for fine segmentation of irrigation area, and U-Net has excellent image segmentation ability, which can identify subtle irrigation area boundaries. In order to further improve the precision and robustness of the model, traditional machine learning methods such as XGBoost and random forest are combined, which can effectively process the complex relationship and nonlinear characteristics in the input data and provide strong feature fusion ability.

[0093] In the integrated model part, the outputs of multiple models such as U-Net and XGBoost / random forest are fused using the integrated learning method. The advantages of each model are combined through weighted average or voting mechanism to improve the accuracy and stability of the overall model. In addition, the deep learning model combined with traditional machine learning method can better extract feature information in complex irrigation areas.

[0094] After the training is completed, feature variable importance analysis is performed using tools such as SHAP value (Shapley Additive Explanations) to deeply analyze the contribution of each feature variable in the model to the identification of irrigation area, thereby improving the interpretability and transparency of the model.

[0095] Model verification will be carried out in Inner Mongolia Lengkou Irrigation District and Henan Yuzhuang Irrigation District, which have different climate conditions, soil types and irrigation management modes, providing wide applicability for model verification. The verification data will include high-resolution actual irrigation area data obtained through unmanned aerial vehicle data, which will be compared with the model output results. In addition, statistical data such as irrigation management records will also be used as verification data sources. Through ten-fold cross-validation method, the performance of the model on different data sets is evaluated to ensure its stability and generalization ability. Unmanned aerial vehicle data will be used as residual input to further optimize model training and enhance model accuracy.

[0096] Result output: Finally, the model will output monthly-scale actual irrigation area data, all results will be spatially aggregated to field-scale, and 7-day moving average filtering will be used to smooth the time series and reduce noise.

[0097] This part uses the high-precision multi-source fusion input dataset generated in the first step, selects vegetation index (NDVI), evapotranspiration, land surface temperature, meteorological data, soil moisture, spectrum, drought index, SAR data, elevation, slope, land use information and phenological characteristics as feature variables, and trains U-Net, XGBoost, random forest and other deep learning and integrated models for different crops (such as paddy fields), combined with irrigation farmland edge monitoring and segmentation model, to realize monthly-scale actual irrigation area identification. This method solves the problem of relying on manual investigation, long update cycle and insufficient precision in existing irrigation area statistics, and realizes large-scale, high-frequency and high-precision dynamic monitoring of irrigation area.

[0098] In one specific embodiment, the irrigation water consumption monitoring specifically includes:

[0099] Single crop irrigation water consumption calculation: according to crop type, growth stage and climate zoning, the basic irrigation water consumption W base is determined, and the dynamic correction factor is:

[0100] W adj =W base × ) ) ;

[0101] wherein, ET actual is the actual ET , ET ref is the reference ET, S soil is the real-time soil moisture, S threshold is the irrigation triggering threshold;

[0102] By using GIS spatial overlay analysis, the irrigation water consumption of different crops is weighted and averaged according to the planting area to generate county-level or irrigation district-level comprehensive irrigation water consumption of multiple crops:

[0103] W comp = ;

[0104] wherein, W i is the corrected irrigation water consumption of the crop, i A i is the planting area thereof.

[0105] This part accurately estimates the irrigation water consumption of a single crop according to the obtained irrigation water consumption of different crops and the obtained actual irrigation area data, combines the crop water requirement characteristics, calculates the comprehensive irrigation water consumption of the entire irrigation district by comprehensively considering the irrigation requirements of multiple crops and the distribution of regional water resources, realizes the accurate assessment of the water demand of crops at different growth stages, and provides a decision basis for irrigation scheduling. This method solves the problem of lack of real-time update and difficulty in dynamically adapting to climate and crop changes in the existing irrigation water consumption formulation, and improves the scientificity and rationality of water resource allocation.

[0106] An irrigation water consumption estimation system based on multi-source data fusion, comprising:

[0107] A data acquisition module acquires vegetation index, meteorological data and soil moisture data in an irrigation area, and processes the acquired vegetation index, meteorological data and soil moisture data respectively to obtain daily-scale vegetation index, daily-scale meteorological data and daily-scale soil moisture;

[0108] ​The data downscaling module processes the daily vegetation index, daily meteorological data and daily soil moisture respectively by using statistical downscaling technology, so that all the data have a unified spatial resolution.

[0109] The evapotranspiration estimation module estimates the daily evapotranspiration based on the SEBAL model on the basis of the data with the unified spatial resolution.

[0110] The irrigation water quantity estimation module estimates the daily irrigation water quantity per unit area by using the water balance method on the basis of the obtained daily evapotranspiration, the runoff, the leakage and the change in soil water storage.

[0111] The irrigation area output module identifies the actual irrigation area by using deep learning and ensemble model technology.

[0112] The irrigation water quantity output module calculates the single-crop irrigation water quantity and the comprehensive crop irrigation water quantity based on the obtained data.

[0113] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments, and the same or similar parts between the embodiments can be referred to each other. For the device disclosed by the embodiments, since it corresponds to the method disclosed by the embodiments, the description is relatively simple, and the related parts can be referred to the method part.

[0114] The above description of the disclosed embodiments enables a person skilled in the art to implement or use the present application. Various modifications to the embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for estimating irrigation water consumption per unit area based on multi-source data fusion, characterized in that, include: Collect vegetation index, meteorological data, and soil moisture data within the irrigated area; The collected vegetation index, meteorological data, and soil moisture data were processed to obtain daily-scale vegetation index, daily-scale meteorological data, and daily-scale soil moisture. Statistical downscaling techniques were used to process daily vegetation index, daily meteorological data, and daily soil moisture to give all data a uniform spatial resolution. Based on data with uniform spatial resolution, diurnal evapotranspiration is estimated using the SEBAL model. Based on the obtained daily evapotranspiration, supplemented by data on changes in runoff, infiltration, and soil water storage, the water balance method is used to estimate the daily irrigation water consumption per unit area. The actual irrigated area was identified using deep learning and ensemble modeling techniques. Based on the calculated data, the irrigation water volume per unit area for a single crop and the comprehensive irrigation water volume per unit area for multiple crops are calculated. The formulas for calculating the irrigation water volume per unit area for a single crop and the combined irrigation water volume per unit area for multiple crops are as follows: Calculation of irrigation water per unit area for a single crop: Based on crop type, growth stage, and climate zone, determine the basic irrigation water per unit area W. base Dynamic correction factor: W adj =W base × （ ）； in, ET actual For the actual remote sensing inversion ET , ET ref For reference to ET, S soil For real-time soil moisture, S threshold This is the irrigation trigger threshold; Using GIS spatial overlay analysis, the irrigation amount per unit area for different crops is weighted and averaged according to the planting area to generate the comprehensive irrigation amount per unit area for multiple crops at the county or irrigation district level: W comp = ； In the formula, W i For crops i Corrected irrigation water consumption per unit area A i The planting area.

2. The method for estimating irrigation water consumption per unit area based on multi-source data fusion according to claim 1, characterized in that, The collection of vegetation index, meteorological data, and soil moisture data within the irrigation area; and the processing of the collected vegetation index, meteorological data, and soil moisture data to obtain daily-scale vegetation index, daily-scale meteorological data, and daily-scale soil moisture are specifically as follows: Vegetation indices: Using MODIS satellite sensors, regional-scale normalized vegetation index, enhanced vegetation index, green index, normalized water index, and soil-adjusted vegetation index are extracted to achieve large-scale dynamic monitoring of crop growth. Simultaneously, during the key growth stages of crops, multispectral UAVs are used for aerial surveys to acquire hyperspectral images with a ground resolution of <5cm. Canopy spectral reflectance, spatial distribution of soil moisture, and visible irrigation traces are collected simultaneously. The collected data are then corrected using a scale transformation model to obtain the daily-scale vegetation index. Meteorological data: The data sources are integrated. First, the daily continuous observation data of meteorological stations in the irrigation area are integrated, including rainfall, surface temperature, air temperature, wind speed and relative humidity parameters. Second, the reanalysis dataset of the weather forecast center is introduced to supplement global meteorological elements. The surface temperature is retrieved by applying the split window algorithm based on the thermal infrared sensor of the Landsat-8 satellite. Finally, the multi-source daily meteorological data is integrated using the ensemble Kalman filter method. Soil Moisture: To obtain the dynamics of root zone moisture, a multi-source satellite data assimilation and fusion method was adopted, combining the Noah-MP land surface process model with an ensemble Kalman filter data assimilation framework. Within this framework, soil moisture remote sensing data were first acquired from different satellite sensors, Sentinel-1, medium resolution imaging spectrometer, soil moisture and ocean salinity satellites, and interferometric synthetic aperture radar, covering soil moisture information at different spatial and temporal resolutions. The enhanced spatiotemporal adaptive reflectivity fusion model was used for data fusion. This model dynamically adjusts the model's state variables by combining different remote sensing data with the simulation results of the Noah-MP land surface process model, optimizing the estimation of soil moisture in real time, and thus obtaining the daily-scale soil moisture.

3. The method for estimating irrigation water consumption per unit area based on multi-source data fusion according to claim 2, characterized in that, The statistical downscaling technique is used to process the daily-scale vegetation index, daily-scale meteorological data, and daily-scale soil moisture to give all data a uniform spatial resolution. Specifically, this includes: A multi-source data downscaling framework was adopted to achieve spatial uniformity of data with different resolutions. The input data included 30-meter resolution vegetation indices generated by Landsat-8, 30-meter daily meteorological data fused by an enhanced spatiotemporal adaptive reflectance fusion model, and fused soil moisture data. The processing flow first performed feature engineering to extract the temporal gradient of vegetation indices and topographic adjustment factors. Then, interpolation and resampling methods were used to unify all input data to a spatial resolution of 30 meters. Finally, physical constraints were used to optimize soil moisture, generating soil moisture data with uniform spatial resolution, thus ensuring spatial consistency of data from different sources.

4. The method for estimating irrigation water consumption per unit area based on multi-source data fusion according to claim 1, characterized in that, The estimation of diurnal evapotranspiration based on the SEBAL model, using data with uniform spatial resolution, specifically includes: All input data were fused from multiple sources and downscaled to a spatial resolution of 30 meters. In the SEBAL model calculation process, firstly, remote sensing data and meteorological data were used to calculate net radiation using the principle of surface energy balance. Secondly, surface heat flux and sensible heat flux were calculated by combining surface reflectance and solar radiation information. Finally, based on the net radiation, surface heat flux, and sensible heat flux obtained, the SEBAL model calculated diurnal evapotranspiration using the energy balance equation.

5. The method for estimating irrigation water consumption per unit area based on multi-source data fusion according to claim 1, characterized in that, The specific calculation formula for the SEBAL model is as follows: ； ； ； ； in, Where G is net radiation and G is surface heat flux. H For sensible heat flux, Evaporation rate It is the difference between the aerodynamic surface temperature and the reference altitude temperature. Where is the surface temperature, and a and b are empirical coefficients. It is air density. It is the specific heat of air. It is the aerodynamic drag that accounts for heat transfer between the aerodynamic surface and the reference altitude. For surface albedo, and For incident and outgoing longwave radiation, It is the surface thermal emissivity. It is the actual direct and diffuse solar radiation flux reaching the Earth's surface; The evapotranspiration of crops can be calculated using the formula above.

6. The method for estimating irrigation water consumption per unit area based on multi-source data fusion according to claim 1, characterized in that, The specific formula for estimating daily irrigation water consumption using the water balance method based on the obtained daily evapotranspiration is as follows: I t =ET t +R t +D t -P t +ΔS t ； In the formula, P t For precipitation, I t This refers to the daily irrigation volume. ET t Evaporation rate R t For runoff, D t This refers to the depth of leakage. ΔS t This refers to changes in soil water storage.

7. The method for estimating irrigation water consumption per unit area based on multi-source data fusion according to claim 1, characterized in that, The deep learning and ensemble model technology for identifying the actual irrigated area specifically includes: During the model training phase, the input data is first divided into different crop types, and a deep learning model is trained separately for each crop type; the U-Net model is used for fine segmentation of the irrigation area. In the ensemble model section, an ensemble learning method is used to fuse the outputs of multiple models, and the advantages of each model are combined by using a weighted average or voting mechanism. After training, feature variable importance analysis is performed, and tools are used to analyze in depth the contribution of each feature variable in the model to the identification of irrigation area. Output results: The model will output actual irrigated area data at the monthly scale, and all results will be spatially aggregated to the field scale.

8. A system for estimating irrigation water consumption per unit area based on multi-source data fusion, characterized in that, The method for estimating irrigation water consumption per unit area based on multi-source data fusion according to any one of claims 1-7 includes: The data acquisition module collects vegetation index, meteorological data, and soil moisture data within the irrigation area; and processes the collected vegetation index, meteorological data, and soil moisture data to obtain daily-scale vegetation index, daily-scale meteorological data, and daily-scale soil moisture. The data downscaling module uses statistical downscaling techniques to process daily vegetation index, daily meteorological data, and daily soil moisture, so that all data have a uniform spatial resolution. The evapotranspiration estimation module estimates daily evapotranspiration based on the SEBAL model, using data with uniform spatial resolution. The irrigation water consumption estimation module, based on the obtained daily evapotranspiration, supplements runoff, infiltration and soil water storage change data, and uses the water balance method to estimate the daily irrigation water consumption per unit area. The irrigation area output module uses deep learning and ensemble model technology to identify the actual irrigation area. The irrigation output module calculates the irrigation volume per unit area for a single crop and the comprehensive irrigation volume per unit area for multiple crops in different scale regions based on the calculated data.

Images