A method and device for dynamically regulating nitrogen in crops at a regional scale

Abstract

The application discloses a regional scale crop nitrogen dynamic regulation method and equipment, comprising the following steps: acquiring multi-source data and constructing a time-space database; selecting a crop growth model, calibrating and verifying the crop growth model based on the time-space database, obtaining the crop growth model and generating a localized model parameter set; dividing a target region into spatial units, inputting soil data, meteorological data and the localized model parameter set of each spatial unit into the crop growth model, simulating the crop growth process under different nitrogen gradients, extracting optimal biomass and critical nitrogen concentration data point pairs of each spatial unit, fitting and generating a spatialized pCNDC parameter layer; based on remote sensing data, inversing actual biomass and nitrogen concentration spatial distribution maps, combining the pCNDC parameter layer to calculate the critical nitrogen concentration and nitrogen nutrition index of each spatial unit and generate spatial distribution maps; when the nitrogen nutrition index is lower than a preset threshold, calculating nitrogen deficiency and generating a variable fertilization prescription map.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of agricultural technology, more particularly, to a regional-scale crop nitrogen dynamic regulation method and device. BACKGROUND

[0002] Nitrogen is one of the essential macronutrients for crop growth and development, and plays a decisive role in photosynthesis, protein synthesis, and the formation of final yield and quality. In global agricultural production, the application of nitrogen fertilizer is a key measure to ensure food supply and improve yield per unit area. However, for a long time, the traditional nitrogen management mode is often based on experience or regional average recommendation, without fully considering the huge variability of crop actual nitrogen demand in time process and spatial distribution.

[0003] This extensive management mode often leads to insufficient nitrogen application in some areas, limiting the realization of crop yield potential; while in other areas, excessive application not only increases the cost of agricultural production, but also wastes valuable nitrogen resources, and even causes serious environmental problems, such as soil acidification, groundwater nitrate pollution, water eutrophication, and increased greenhouse gas (such as nitrous oxide N2O) emissions. Therefore, developing and applying crop nitrogen precision management technology to realize intelligent fertilization according to demand, quantity, time and location has become an inevitable requirement and important direction for the sustainable development of modern agriculture.

[0004] The theory of critical nitrogen dilution curve (CNDC) provides an important physiological basis for crop nitrogen diagnosis. The theory points out that, during the growth of crops, in order to maintain the maximum growth rate, the critical nitrogen concentration (N c , i.e. the minimum nitrogen concentration required to ensure the maximum growth rate) in the aboveground part of the plant will show a power function dilution law with the increase of aboveground biomass (W), usually expressed as N c =a×W b , where a and b are parameters related to crop type and environmental conditions.

[0005] Based on the CNDC theory, by comparing the actual measured nitrogen concentration (N actual ) of the crop with the corresponding critical nitrogen concentration (N c ) under the current biomass, the nitrogen nutrition index (Nitrogen Nutrition Index, NNI) can be calculated, i.e. NNI=N actual / N cNNI is widely accepted as a reliable indicator for assessing the nitrogen status of crops in real time, whether the crop is in a state of deficiency (NNI < 1), adequacy (NNI ≈ 1), or surplus (NNI > 1).

[0006] Although the CNDC theory and its derived NNI indicator have shown great potential for application in scientific research, their promotion and application in actual agricultural production, especially at the regional scale, face serious challenges. Most existing CNDC research is based on point experiments or conducted on relatively homogeneous small plots. Obtaining CNDC through experimental research not only faces high time and economic costs, but also the CNDC parameters (a, b) obtained are often only applicable to specific test sites, crop types, and environmental conditions.

[0007] However, real agricultural regions (such as a county, an irrigation district, or a large farm) often exhibit significant spatial heterogeneity, which includes differences in climate conditions (such as uneven spatial distribution of precipitation, temperature, and light), diversity of soil properties (such as variations in soil type, organic matter content, texture, and fertility level in space), regionalized layout of crop types, and differences in field management measures (such as sowing time and irrigation method). This high spatial heterogeneity leads to significant differences in crop growth processes and nitrogen requirements at different locations within the region, making it difficult to directly and accurately apply CNDC parameters obtained from point studies to the entire region.

[0008] Currently, research is actively being conducted on combining CNDC theory, crop growth models, and remote sensing technology for dynamic and spatial evaluation of crop nitrogen status at the regional scale or for variable fertilization decision-making. Related Chinese patent documents also reflect research and development activities in the application of CNDC and NNI diagnosis, remote sensing nitrogen monitoring, and variable fertilization technology. For example, CN106841051A proposes using greenness and coverage information from crop canopy images in combination with the critical nitrogen dilution model to calculate NNI. CN107505271A focuses on plant nitrogen estimation based on a nitrogen component radiation transfer model. CN102487644A and CN103959973A involve variable fertilization devices and precise application methods.

[0009] However, related technologies do not provide any technical inspiration on how to overcome the spatial heterogeneity barrier, accurately and dynamically generate and apply nitrogen diagnostic indicators, achieve variable fertilization based on real-time crop demand, and adapt to long-term changing environments through feedback for self-optimization in a new type of nitrogen fertilizer management method. SUMMARY

[0010] In view of the above problems existing in the prior art, the application provides a regional-scale crop nitrogen dynamic regulation method and equipment, which can realize precise variable fertilization based on crop physiological needs, environmental potential and real-time growth potential through real-time monitoring and a closed-loop learning mechanism.

[0011] The object of the application is achieved by the following technical solutions.

[0012] As a first aspect of the application, some embodiments of the application provide a regional-scale crop nitrogen dynamic regulation method, comprising the following steps:

[0013] Obtain multi-source data of the target region to construct a spatio-temporal database, the multi-source data including meteorological data, soil data, crop growth dynamic data, crop management measure data and remote sensing data;

[0014] Select a crop growth model with a nitrogen response module, calibrate and verify the crop growth model based on the spatio-temporal database, obtain a crop growth model reflecting the dynamic response of crop nitrogen in the target region, and generate a localized model parameter set;

[0015] Divide the target region into spatial units, input the soil data, meteorological data and localized model parameter set of each spatial unit into the crop growth model, simulate the crop growth process under the nitrogen gradient without water stress, extract the optimal biomass and critical nitrogen concentration data point pair of each spatial unit, and fit the spatialized pCNDC parameter layer of the target region;

[0016] Based on the remote sensing data, the actual biomass and nitrogen concentration spatial distribution map are inversed, the critical nitrogen concentration and nitrogen nutrition index of each spatial unit are dynamically calculated in combination with the pCNDC parameter layer, and the spatial distribution maps are respectively generated; the nitrogen deficiency amount is calculated based on the generated spatial distribution maps, the variable fertilization prescription map is generated in combination with the soil nitrogen supply potential and the nitrogen fertilizer utilization efficiency to guide fertilization, and regional-scale crop nitrogen dynamic regulation is realized.

[0017] Further, after implementing fertilization according to the variable fertilization prescription map, feedback data are collected and supplemented to the spatio-temporal database, the feedback data including crop yield, crop nitrogen absorption data at the harvest period, fertilization cost data and control data;

[0018] Based on the feedback data, the crop growth model is optimized, the remote sensing data inversion is optimized, and the threshold is adjusted, so as to form a closed-loop regulation mechanism and realize continuous optimization of the regional-scale crop nitrogen dynamic regulation method.

[0019] Further, the crop growth model is selected to be screened by a quantitative evaluation system, the crop growth model has a nitrogen response module and is suitable for the main variety of the target region, and the quantitative evaluation system includes:

[0020] Data fitting dimension, used to evaluate the matching degree of model input requirements and spatio-temporal database;

[0021] Availability and scalability dimension, used to evaluate the openness of model code and the convenience of parameter adjustment;

[0022] Case and literature support dimension, used to count the number of application cases of the model in the target crop type.

[0023] Further, the step of calibrating the crop growth model comprises: setting parameters of the crop growth model;

[0024] The global sensitivity analysis method is used to analyze the sensitivity of the crop growth model parameters to the output variables, and to identify the key model parameters that have the greatest impact on nitrogen response simulation;

[0025] The crop growth dynamic data in the spatio-temporal database is used, and a multi-objective optimization algorithm is used to calibrate the key model parameters, to ensure that the comprehensive performance of the crop growth model in the entire nitrogen response range from nitrogen deficiency to sufficient nitrogen is optimal.

[0026] Further, in the simulation of crop growth under different nitrogen gradients, variance analysis is performed and different nitrogen gradients are divided into: the low biomass group with the smallest numerical value, the high biomass group or plateau group with the largest numerical value, and the remaining medium biomass group according to the biomass numerical value;

[0027] In the high biomass group or plateau group, the treatment with the lowest nitrogen application rate is selected, which represents the minimum nitrogen supply required for the crop to achieve the maximum growth rate in the current growth period. The corresponding aboveground biomass is recorded as the optimal biomass, and the plant nitrogen concentration is recorded as the critical nitrogen concentration, to obtain the optimal biomass and critical nitrogen concentration data point pair of the spatial unit.

[0028] Further, based on the data point pair of the optimal biomass and the critical nitrogen concentration of each spatial unit, the power function equation of the critical nitrogen dilution curve is fitted to obtain the pCNDC parameters a and b of each spatial unit, and the power function equation of the critical nitrogen dilution curve is expressed as:

[0029] N c =a×W optimal b ;

[0030] In the formula, W optimal represents biomass, and N c represents critical nitrogen concentration;

[0031] The parameters a and b of all the spatial units are collected and stored as raster layers according to spatial positions, forming a spatialized pCNDC parameter a layer and a spatialized pCNDC parameter b layer, which together constitute a spatialized pCNDC parameter layer covering the target region.

[0032] Further, the step of obtaining the actual biomass and nitrogen concentration spatial distribution map based on remote sensing data comprises: obtaining remote sensing data covering the key growth period of crops in the spatiotemporal database, calculating a vegetation index sensitive to biomass;

[0033] Based on the vegetation index and the measured biomass on the ground, an actual biomass inversion model is established to perform inversion, generating an actual biomass spatial distribution map covering the target region and corresponding to the time phase of the remote sensing data;

[0034] Based on the spectral characteristics sensitive to plant nitrogen in the remote sensing data, including red edge parameters, chlorophyll index, nitrogen reflection index or specific band combination spectral information, an actual nitrogen concentration inversion model is established in combination with the measured plant nitrogen concentration on the ground to perform inversion, generating an actual nitrogen concentration spatial distribution map corresponding to the time phase of the remote sensing data.

[0035] Further, the actual biomass spatial distribution map and the spatialized pCNDC parameter layer are spatially superimposed and matched in spatial units to obtain the actual biomass and parameters a and b of each spatial unit, and the critical nitrogen concentration of the spatial unit is calculated and expressed as:

[0036] N c =a×(W actual ) b ;

[0037] In the formula, N c represents the critical nitrogen concentration, and W actual represents the actual biomass.

[0038] The critical nitrogen concentrations of all the spatial units are collected to generate a critical nitrogen concentration spatial distribution map corresponding to the time phase covering the target region;

[0039] The actual nitrogen concentration spatial distribution map and the critical nitrogen concentration spatial distribution map are spatially superimposed and matched in spatial units to obtain the actual plant nitrogen concentration value and the critical nitrogen concentration value of each spatial unit, and the nitrogen nutrition index of the spatial unit is calculated and expressed as:

[0040] NNI=N actual / N c ;

[0041] In the formula, NNI represents the nitrogen nutrition index, N actual represents the actual plant nitrogen concentration value, and N c represents the critical nitrogen concentration.

[0042] The nitrogen nutrient index of all spatial units is collected to generate a spatial distribution map of the nitrogen nutrient index of the corresponding time phase covering the target area;

[0043] Based on the spatial distribution map of nitrogen nutrient index, the crop nitrogen nutrient status of each spatial unit is assessed by dynamically preset thresholds. When the nitrogen nutrient index is lower than the threshold, the nitrogen deficit is calculated, and a variable fertilization prescription map is generated by combining soil nitrogen supply potential and nitrogen fertilizer utilization efficiency.

[0044] Furthermore, the steps for generating a variable fertilization prescription map include: reading the critical nitrogen concentration, actual nitrogen concentration, and actual biomass of each spatial unit, calculating the nitrogen deficit, and expressing it as:

[0045] N deficit =(N c N actual )×W actual ;

[0046] In the formula, N deficit Nitrogen deficit, N c The critical nitrogen concentration for the current biomass, N actual W represents the real-time actual nitrogen concentration of the crop. actual This refers to the actual biomass of the crop in real time.

[0047] Subtracting the soil's nitrogen supply potential from the nitrogen deficit yields the net nitrogen demand (N). net_demand , is represented as:

[0048] N net_demand =N deficit SoilN supply ;

[0049] Predicting nitrogen fertilizer use efficiency and calculating recommended nitrogen application rate ΔN recommend , is represented as:

[0050] ΔN recommend =N net_demand / NUE;

[0051] The recommended nitrogen application rates for all spatial units are collected and stored as raster layers according to spatial location to obtain a variable fertilization prescription map.

[0052] As a second aspect of this application, some embodiments of this application provide a regional-scale crop nitrogen dynamic regulation device, including a memory storing a computer program; a processor executing the computer program to implement the steps of the above-described method; and a communication module for receiving input from an external meteorological data server and a remote sensing data platform.

[0053] Compared with the prior art, the advantages of this invention are:

[0054] (1) Propose a potential critical nitrogen dilution curve (pCNDC) spatialization method based on crop growth model. By simulating the nitrogen gradient under non-water stress, the parameters for each spatial unit are fitted to accurately quantify the influence of meteorological conditions, soil characteristics, and variety differences on crop nitrogen demand. This method replaces the traditional point-based repeated test determination mode, significantly reducing data collection costs and improving regional applicability.

[0055] (2) Establish a dynamic diagnosis and real-time decision-making mechanism to improve management timeliness. By combining the model-simulated growth "potential" with the remote sensing-monitored crop "actual" growth, a nitrogen nutrition index (NNI) spatial distribution map is dynamically generated, enabling immediate diagnosis of crop nitrogen balance. The decision-making process also incorporates weather forecasts to adjust the fertilization plan, significantly outperforming the traditional static management mode and ensuring the real-time and efficiency of the fertilization decision.

[0056] (3) Introduce a yield-nitrogen absorption-economic benefit closed-loop learning mechanism that can use end-of-season yield maps and other actual data to automatically correct and optimize crop growth models, remote sensing algorithms, and fertilization decision logic. This self-learning closed-loop system enables the method to continuously adapt to changes in climate, variety, and management measures, ensuring the robustness and advancement of the technical solution in long-term applications. BRIEF DESCRIPTION OF DRAWINGS

[0057] Fig. 1 is a step flow chart of the regional-scale crop nitrogen dynamic regulation method in an embodiment of the present application.

[0058] Fig. 2 is a critical nitrogen dilution curve for a spatial unit in an embodiment of the present application. DETAILED DESCRIPTION

[0059] The present application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0060] In combination Figs. 1-2 , a regional-scale crop nitrogen dynamic regulation method of the present application includes the following steps:

[0061] S1, constructing a spatio-temporal database

[0062] Obtain multi-source data of the target region to construct a spatio-temporal database, which is used to comprehensively characterize the meteorological conditions, soil characteristics, crop growth dynamics, and crop management measures of the target region, thereby serving as the data basis for crop growth model calibration, nitrogen dynamic diagnosis, and variable fertilization decision.

[0063] Specifically, the multi-source data includes meteorological data, soil data, crop growth dynamic data, crop management measure data, and remote sensing data, and the acquisition process is as follows:

[0064] (1) Meteorological data acquisition:

[0065] Obtain historical daily weather data (at least the past 10 years) and real-time daily weather data for the target area or nearby weather stations. Weather data includes maximum temperature, minimum temperature, average temperature, precipitation, solar radiation, wind speed, and relative humidity.

[0066] Weather data sources include national weather networks, local weather stations, and reanalysis datasets. Spatial interpolation methods such as inverse distance weighting or Kriging interpolation are used to process the weather data observed at the stations, and finally generate a gridded weather data layer covering the target area, reflecting the spatial and temporal variation of climate conditions.

[0067] (2) Soil data acquisition:

[0068] Extract soil texture (percentage of sand, silt, and clay), bulk density, pH value, organic matter content, total nitrogen, available nitrogen, phosphorus, and potassium content, field moisture capacity, and wilting coefficient from soil survey maps or digital soil maps. Combine with geostatistical methods to interpolate and generate gridded soil property layers, obtaining soil data for quantifying soil fertility and nitrogen supply potential.

[0069] (3) Crop growth dynamic data acquisition:

[0070] Crop growth dynamic data aims to quantitatively describe the morphology, phenology, and nutrient absorption process of different crops in the target area during the entire growth period, which is used for subsequent calibration of crop growth models. Crop growth dynamic data includes growth indicators at key growth stages, soil moisture, and soil mineral nitrogen dynamics.

[0071] By designing and implementing field trials covering major varieties, main soil types, and nitrogen supply level gradients in the target area, crop growth dynamic data is obtained. Monitor the growth indicators of crops at key growth stages, soil moisture, and soil mineral nitrogen dynamics during the field trial process as crop growth dynamic data. Among them, the main varieties refer to the main varieties planted on the cultivated land in the target area, such as crop varieties with planting area accounting for more than 10% of the cultivated land area in the region. The main soil type refers to the soil type whose cultivated land area reaches more than 10% of the cultivated land area of the target area.

[0072] Key growth period refers to a period of time that has a significant impact on crop nitrogen response, nutrient absorption, and final yield formation. For example, for rice, it usually includes tillering stage, jointing stage, booting stage, full heading stage, filling stage, and maturity stage.

[0073] The growth indicators include aboveground biomass, biomass of different organs (e.g. stem, leaf and spike or grain of rice), plant nitrogen concentration or nitrogen content, leaf area index (LAI), final yield and yield component, and quality indicators. The quality indicators can be protein content, amylose content, chalkiness, etc.

[0074] In a specific embodiment, when the field test conditions are lacking, the historical data of the field tests carried out in the target region can be obtained by consulting literature as the crop growth dynamic data.

[0075] (4) Crop management measure data acquisition:

[0076] The crop management measure data are obtained through agricultural statistical yearbook, local agricultural department records or farmer survey.

[0077] The crop management measure data include main cultivars and their growth period characteristics, sowing date, transplanting date, planting density, irrigation system and historical fertilization data, the historical fertilization data including fertilization type, fertilization period and fertilization amount, which together constitute the crop management measure data depicting the basic situation of agricultural production activities in the region.

[0078] (5) Remote sensing data acquisition:

[0079] The RGB (red, green, blue three-band visible light image), multispectral or hyperspectral remote sensing images covering the entire growth season of the crop are obtained and preprocessed to obtain the remote sensing data.

[0080] Specifically, the data source of the remote sensing data is medium-high resolution satellite (e.g. Sentinel-2 or Landsat 8 / 9), unmanned aerial vehicle platform or aerial remote sensing. The images without cloud coverage or with cloud coverage percentage less than 10% in the target region are screened, and the preprocessing of radiometric calibration, atmospheric correction, geometric precise correction and cloud mask is performed to obtain the remote sensing data. The preprocessing can ensure that the image quality meets the accuracy requirements of subsequent biomass and nitrogen concentration inversion.

[0081] The preprocessed remote sensing data are standardized and registered for spatial and temporal information, and are integrated into the time-space database as the key dynamic data layer, which specifically includes:

[0082] Time series construction: associate accurate imaging date and time for each remote sensing image, and organize and index according to time sequence to form an image time series covering the entire growth season and reflecting the dynamic process of crop growth.

[0083] Spatial alignment and fusion: Ensure that all remote sensing image layers use the same coordinate system and spatial resolution as the static data layers such as meteorological and soil data in the database. Through spatial alignment, any location can be accurately matched and linked with soil properties and meteorological data for analysis.

[0084] Through the above storage process, discrete and static remote sensing images are converted into dynamic monitoring layers that can be queried, analyzed, spatially explicit, and time-continuous in the spatio-temporal database. This not only realizes effective management of earth observation data, but also makes it a bridge connecting crop growth model simulation and actual growth in the field.

[0085] S2, crop growth model selection and calibration

[0086] Select a crop growth model and calibrate and verify the model based on the spatio-temporal database to obtain a set of localized model parameters reflecting the dynamic response of nitrogen in the target area, including:

[0087] (1) Model selection:

[0088] Select a crop growth model that can simulate the mechanism of nitrogen absorption, distribution, and utilization of crops and has a good application basis, such as WOFOST, DSSAT series, APSIM, ORYZA, and STICS, etc.

[0089] First, select a candidate crop growth model with a perfect nitrogen response module and suitable for the main crop in the target area (referring to the crop type that occupies a dominant position in the planting area, total yield, or economic value in the region). Quantitative evaluation is used to select from the candidate crop growth models based on data fit, availability and scalability, and case and literature support.

[0090] Data fit is used to consider the matching degree and availability of the input data required by the crop growth model and the spatio-temporal database. Availability and scalability are used to consider the cost of obtaining the crop growth model (preferably open source), the openness of the code, the convenience of parameter adjustment and function expansion. Case and literature support are used to consider the breadth and depth of published verification studies and application cases of the crop growth model in similar agricultural ecological regions or for the target crop type.

[0091] In a specific embodiment, the target crop is rice, and the candidate crop growth models can include the ORYZA (v3) model and the CERES-Rice model in DSSAT v4.8. Among them, the ORYZA (v3) model is designed specifically for rice and has high data fit; the CERES-Rice model is a module of DSSAT and is widely used but may require data format conversion. Finally, the most suitable crop growth model is selected based on the comprehensive evaluation of the above dimensions.

[0092] In one specific embodiment, the candidate crop growth model can be quantitatively scored according to the evaluation dimensions, for example, 1 point represents poor performance or does not meet the requirements, 2 points represents average performance or basically meets the requirements, and 3 points represents excellent performance or completely meets the requirements.

[0093] In one specific embodiment, the target crop type is rice, and the candidate crop growth model is the ORYZA (v3) model with a perfect nitrogen response module and the CERES-Rice model in DSSAT v4.8.

[0094] The ORYZA (v3) model is designed for rice and has high input data compatibility; the CERES-Rice model needs to be adapted to the physiological and ecological characteristics of rice, that is, it needs to be processed through data conversion or parameter estimation. Therefore, in this embodiment, the ORYZA (v3) model scores 3 points in data compatibility, and the CERES-Rice model scores 2 points in data compatibility.

[0095] The source code of some versions of the ORYZA (v3) model can be obtained under the framework of scientific research cooperation; the CERES-Rice model is highly available but needs to be deeply modified and developed at the source code level. Therefore, in this embodiment, the ORYZA (v3) model scores 3 points in availability and scalability, and the CERES-Rice model scores 2 points in availability and scalability.

[0096] The ORYZA (v3) model, as one of the mainstream rice growth models, has a wide range of verification and application, and its simulation accuracy and reliability have been fully verified; the CERES-Rice model also has a wide range of verification and application. Therefore, in this embodiment, the ORYZA (v3) model scores 3 points in case and literature support, and the CERES-Rice model scores 3 points in case and literature support.

[0097] As shown in Table 1, the ORYZA (v3) model scores 9 points in the three evaluation dimensions, and the CERES-Rice model in DSSAT v4.8 scores 7 points in the three evaluation dimensions. Finally, the ORYZA (v3) model is selected as the crop growth model in this embodiment.

[0098] Table 1 Comparison and evaluation of ORYZA (v3) model and CERES-Rice model

[0099]

[0100] (2) Model calibration:

[0101] According to the genetic characteristics of the crops in the target area, the soil data and the meteorological data in the spatio-temporal database, the selected crop growth model is parameterized, and the crop growth model parameters include light energy utilization rate, maximum photosynthetic rate, nitrogen absorption efficiency, root growth parameters, nitrogen distribution coefficient between organs, and nitrogen stress effect on photosynthesis or growth, etc.

[0102] The global sensitivity analysis method is used to analyze the sensitivity of the crop growth model parameters to the output variables, and the key model parameters that have the greatest impact on the nitrogen response simulation are identified.

[0103] Using the crop growth dynamic data in the spatio-temporal database, a multi-objective optimization algorithm (such as NSGA-II or MOPSO) is used to calibrate the key model parameters determined by the sensitivity analysis, and the optimization objectives are to minimize the comprehensive error (combined use of root mean square error RMSE, normalized root mean square error nRMSE, coefficient of determination R², and consistency index d, etc.) between the simulated biomass, plant nitrogen concentration (or nitrogen content), key growth period, final yield, etc. of the crop growth model under all nitrogen treatment levels and the measured data. Ensure that the comprehensive performance of the crop growth model in the entire nitrogen response range from nitrogen deficiency to sufficient nitrogen is optimal, and avoid the deviation caused by single nitrogen level calibration.

[0104] Through model calibration, a set of localized crop growth model parameters that can reflect the dynamic response of crops in the target area to nitrogen are obtained.

[0105] (3) Model verification:

[0106] An independent data set from the model calibration is used to evaluate the calibrated crop growth model. Quantitative evaluation indicators such as root mean square error (RMSE), normalized root mean square error (nRMSE), coefficient of determination (R 2 ), and consistency index (d-index) are used to evaluate the consistency between the simulation results of the crop growth model and the measured data.

[0107] Specifically, it is verified whether the prediction accuracy of the crop growth model meets the preset acceptance requirements, for example, the normalized root mean square error (nRMSE) of the key growth indicators (such as biomass, plant nitrogen concentration) is less than 15%, the coefficient of determination (R²) is greater than 0.80, and the consistency index (d-index) is greater than 0.85; at the same time, the simulation error of the key growth period is within ±7 days.

[0108] Specifically, if the prediction accuracy of the crop growth model does not meet the aforementioned acceptance requirements, an iterative feedback adjustment mechanism is started, and the model calibration stage is returned to. According to the bias revealed by the model verification result, the key model parameters identified in the sensitivity analysis stage, which have a significant impact on the output of the crop growth model, are optimized and adjusted. This iterative verification and optimization process will continue until all indicators of the crop growth model in the verification meet the preset acceptance requirements, thereby ensuring that the model has reliable prediction ability and good generalization for different environmental conditions and management measures in the target region.

[0109] Thus, a crop growth model calibrated and verified locally is obtained, which can accurately simulate the growth response dynamics of crops to nitrogen in the target region. The crop growth model and its parameter set are the core tools for generating the spatialized potential critical nitrogen dilution curve (pCNDC) and parameter layer at the regional scale in step S3, and by simulating the nitrogen gradient test under potential growth conditions, the critical nitrogen concentration data points are extracted for each spatial unit. In addition, the predicted values of biomass and nitrogen concentration output by the crop growth model are combined with the actual data retrieved by remote sensing to form a closed-loop optimized data chain, ensuring the continuous improvement of nitrogen management strategies.

[0110] This step solves the problem of insufficient simulation accuracy of general crop growth model parameters in heterogeneous environments by model selection, model calibration, and model verification, and builds a crop growth model that adapts to regional characteristics. Moreover, the model system established by this technical solution has the ability to continuously evolve, and can dynamically adapt to variety updates and climate change by regularly incorporating new crop growth dynamic data, solving the defect of insufficient regional adaptability of traditional empirical models.

[0111] S3, construction of spatialized potential critical nitrogen dilution curve (pCNDC) parameter layer

[0112] The target region is divided into spatial units, and the crop growth process under different nitrogen supply gradients under potential growth conditions without water stress is simulated based on the crop growth model, to generate potential critical nitrogen dilution curve (pCNDC) parameters for each spatial unit in the target region, and to generate a spatialized pCNDC parameter layer covering the target region, specifically including:

[0113] (1) Regional grid division:

[0114] According to the size of the target region, the resolution of the expected variable rate technology map (VRT map) to be generated, and the spatial variation scales of factors affecting crop growth such as soil types and climate factors in the target region, the resolution of the spatial unit is determined, and the target region is divided into square spatial units of the same size.

[0115] Specifically, for small target areas such as farms, high resolution (tens to hundreds of meters) is adopted, and the target area is divided into spatial units such as 30 m x 30 m or 300 m x 300 m; for large target areas such as counties, provinces or larger areas, lower resolution (several to tens of kilometers) is adopted, and the target area is divided into spatial units such as 1 km x 1 km or 10 km x 10 km.

[0116] (2) Crop growth model input acquisition and simulation setting:

[0117] The soil data and meteorological data corresponding to each spatial unit are extracted from the spatio-temporal database, and the extracted unit soil data and meteorological data and the localized model parameter set obtained in step S2 are input into the calibrated crop growth model. In the crop growth model, the nitrogen supply gradient is set to simulate the growth process of the crop from sowing to maturity.

[0118] Specifically, in the crop growth model, a nitrogen supply gradient from low to high is set to cover the range from nitrogen deficiency to sufficient nitrogen supply for the crop, and the nitrogen gradient is set to 0 N / ha - 400 kg N / ha; the gradient interval needs to be small enough to capture the response change, such as 20 kg N / ha or 30 kg N / ha (for example: 0 kg N / ha, 20 kg N / ha, 40 kg N / ha... 400 kg N / ha).

[0119] For each nitrogen gradient, the entire growth process of the crop from sowing to maturity is simulated. By setting a nitrogen supply gradient from low to high to cover the range from nitrogen deficiency to sufficient nitrogen supply for the crop, the biomass and plant nitrogen concentration response curves of the crop under different nitrogen supply conditions can be captured.

[0120] More specifically, the crop growth model needs to run under ideal growth conditions without water stress.

[0121] (3) Key growth period data extraction and critical point determination:

[0122] When the nitrogen supply is below the critical level, increasing nitrogen fertilizer can significantly increase biomass; when the nitrogen supply reaches or exceeds the critical level, the biomass tends to be stable or increase slowly, while the plant nitrogen concentration may continue to rise. By statistical analysis, the inflection point corresponding to the minimum nitrogen application rate at which the biomass reaches the platform period is found, and the plant nitrogen concentration at this biomass level is the critical nitrogen concentration.

[0123] Therefore, for the key growth period in the simulation process, the aboveground biomass (W) and plant nitrogen concentration (N) under each nitrogen gradient are extracted. For the same growth period, the biomass under different nitrogen gradient treatments is subjected to variance analysis, and according to the significance difference result (p<0.1) of the variance analysis, all the nitrogen gradient treatments are grouped.

[0124] Specifically, the key growth stages include tillering, jointing, booting, heading, grain-filling, and maturity. Based on ANOVA results and by comparing the significant differences in biomass among different nitrogen level treatments, the treatments were divided into several groups.

[0125] In one specific embodiment, different nitrogen gradient treatments were divided into low biomass, medium biomass, and high biomass or plateau phase groups. First, analysis of variance was used to divide the treatments of different nitrogen gradients into several groups with significant differences based on biomass values. The group with the smallest value was defined as the low biomass group, the group with the largest value was defined as the high biomass or plateau phase group, and the rest were defined as the medium biomass group.

[0126] Within the high biomass or plateau phase groups, select the treatment with the lowest nitrogen application rate. This treatment represents the minimum nitrogen supply required for the crop to reach its maximum growth rate during this growth stage. Record the aboveground biomass corresponding to this treatment as the optimal biomass (W). optimal The unit is t / ha, and the plant nitrogen concentration is used as the critical nitrogen concentration (N). c (The unit is %).

[0127] Repeat the above process during the key growth stages of the crop to obtain a series of (W) for this spatial unit. optimal N c Data point pairs.

[0128] (4) Spatialized pCNDC parameter layer generation:

[0129] Based on a series of (W) for each spatial unit optimal N c Data point pairs, with optimal biomass W optimal The critical nitrogen concentration N is the independent variable. c As the dependent variable, fit the power function equation of the critical nitrogen dilution curve: N c =a×W optimal b The pCNDC parameters a and b of the unit are obtained by fitting.

[0130] like Fig. 2 As shown, in a specific embodiment, with a spatial unit (W) optimal N c By fitting a power function equation to the data points, the critical nitrogen dilution curve for this spatial cell was obtained: N c =3.44×W optimal 0.437 .

[0131] The parameters a fitted by all the spatial units are collected and stored as a raster layer according to the spatial position to form a spatialized pCNDC parameter a layer. The parameters b fitted by all the spatial units are collected and stored as a raster layer according to the spatial position to form a spatialized pCNDC parameter b layer. The two layers together constitute a spatialized pCNDC parameter layer covering the entire target area.

[0132] This step extends the concept of point-scale critical nitrogen dilution curve to the regional scale, generating pCNDC parameters a and b specific to each spatial unit in the region based on potential growth conditions. The generated spatialized pCNDC parameter layer accurately quantifies the impact of regional environmental heterogeneity on the physiological nitrogen demand benchmark of crops, providing core dynamic data for subsequent real-time nitrogen nutrition index spatialization diagnosis and variable fertilization decision-making based on remote sensing.

[0133] S4, dynamic monitoring of crop growth and nitrogen concentration based on remote sensing

[0134] Based on the remote sensing data in the spatiotemporal database, the actual biomass spatial distribution map and the actual nitrogen concentration spatial distribution map of crops in the target area are obtained through inversion algorithms, specifically including:

[0135] Specifically, remote sensing data covering the key growth period of crops are obtained from the spatiotemporal database. The vegetation index sensitive to biomass is calculated using remote sensing data. Based on the vegetation index and the actual measured biomass on the ground, an actual biomass inversion model is established for inversion, or a radiation transfer model (such as PROSAIL) is used for inversion and applied to remote sensing data to generate a real-time actual biomass (W actual ) spatial distribution map covering the target area corresponding to the time phase of the remote sensing data.

[0136] Among them, the vegetation index includes normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), soil-adjusted vegetation index (SAVI), and wide dynamic range vegetation index (WDRVI).

[0137] The spectral characteristics sensitive to plant nitrogen in remote sensing data, such as red edge parameter (Red Edge Position, REP), chlorophyll index (SPAD), nitrogen reflectance index (Nitrogen Reflectance Index, NRI), or specific band combination spectral information, are used to establish an actual nitrogen concentration inversion model for inversion, or a radiation transfer model is used for inversion. Apply to remote sensing data and generate a real-time actual nitrogen concentration (N actual ) spatial distribution map corresponding to the time phase of the remote sensing data.

[0138] Specifically, the actual biomass inversion model and the actual nitrogen concentration inversion model are verified in accuracy using ground measured data. Quantitative indicators such as root mean square error (RMSE) or coefficient of determination (R²) are used to evaluate the consistency of the inversion results and the measured values, to ensure that the inversion results have physiological rationality and spatial and temporal continuity.

[0139] The actual biomass (W actual ) spatial distribution map and the actual nitrogen concentration (N actual ) spatial distribution map generated in this step are necessary inputs for calculating the spatialized critical nitrogen concentration and nitrogen nutrition index, and the real-time crop growth state data obtained in this step can be fed back to the crop growth model in step S2 for dynamic optimization of model parameters. Compared with the traditional sampling detection method, this step realizes dynamic regional monitoring, which can overcome the defects of time-consuming, labor-intensive and limited spatial coverage of traditional field sampling, and realize high-frequency, large-scale and non-destructive dynamic monitoring of regional crop growth state (biomass and nitrogen concentration). The standardized inversion process established can ensure the comparability, reliability and traceability of the monitoring results.

[0140] S5, dynamic evaluation of nitrogen balance

[0141] Combined with the spatialized pCNDC parameter layer, the actual biomass spatial distribution map and the actual nitrogen concentration spatial distribution map, the critical nitrogen concentration and nitrogen nutrition index of each spatial unit in the target region are dynamically calculated and spatial distribution maps are generated respectively, and the nitrogen balance state of regional crops is dynamically evaluated. Specifically, it includes:

[0142] The actual biomass (W actual ) spatial distribution map generated in step S4 is spatially superimposed and spatially matched with the spatialized pCNDC parameter layer (including parameter a layer and parameter b layer) generated in step S3. The specific method is as follows:

[0143] Ensure that all grid layers involved in the calculation have a unified coordinate reference system. If the resolutions of the layers are not consistent, resample all layers to the same grid system using the resampling algorithm based on the preset calculation resolution, so that each spatial unit can be one-to-one corresponding.

[0144] Under the unified grid system, traverse each spatial unit, and at the same geographic location, extract the actual biomass W actual and parameters a and b corresponding to the spatial unit from the actual biomass spatial distribution map, the pCNDC parameter a layer and the parameter b layer respectively. Based on the values of each spatial unit extracted above, the critical nitrogen concentration of the spatial unit is calculated using the critical nitrogen dilution curve equation:

[0145] N c =a×(W actual ) b ;

[0146] The calculation results of all spatial units are collected to generate a critical nitrogen concentration (N c ) spatial distribution map covering the target area at the corresponding phase, which represents the minimum plant nitrogen concentration required for crops to reach the maximum growth potential or optimal economic benefit at the current actual biomass level.

[0147] The actual nitrogen concentration (N actual ) spatial distribution map generated in step S4 is spatially superimposed and spatial unit matched with the critical nitrogen concentration (N c ) spatial distribution map generated above.

[0148] The actual plant nitrogen concentration value (N actual ) of each spatial unit is obtained from the actual nitrogen concentration (N actual ) spatial distribution map, and the critical nitrogen concentration value (N c ) of the spatial unit is obtained from the critical nitrogen concentration (N c ) spatial distribution map. The nitrogen nutrition index NNI of the spatial unit is calculated:

[0149] NNI = N actual / N c ;

[0150] The calculation results of all spatial units are collected to generate a nitrogen nutrition index (NNI) spatial distribution map covering the target area at the corresponding phase.

[0151] Specifically, based on the nitrogen nutrition index (NNI) spatial distribution map, the nitrogen nutrition status of crops in each spatial unit is evaluated by dynamically presetting thresholds. The thresholds include a nitrogen deficiency threshold and a nitrogen surplus threshold, which are not fixed but dynamically set according to crop type, growth stage, and management objectives. The evaluation process is as follows:

[0152] When the NNI value is lower than the preset nitrogen deficiency threshold (e.g., 0.95), it indicates that the crops are in a nitrogen stress state, and the actual nitrogen concentration is lower than the critical level. The smaller the value, the more severe the nitrogen deficiency.

[0153] When the NNI value is between the preset nitrogen deficiency threshold (e.g., 0.95) and the nitrogen surplus threshold (e.g., 1.05), it indicates that the nitrogen nutrition of crops is in an optimal or near-optimal state, and the recommended nitrogen application amount is determined as zero.

[0154] When the NNI value is higher than the preset nitrogen surplus threshold (e.g., 1.05), it indicates that the nitrogen supply is sufficient or excessive, and there may be luxury absorption, which does not require additional nitrogen fertilizer.

[0155] Luxury absorption refers to the excessive absorption phenomenon when nitrogen supply exceeds crop demand.

[0156] The critical nitrogen concentration (N c ) spatial distribution map and nitrogen nutrition index (NNI) spatial distribution map provide regional scale crop nitrogen physiological demand benchmarks and crop nitrogen surplus or deficit state quantification indicators. Combined with remote sensing data and spatialized pCNDC parameters, high frequency, large scale and dynamic diagnosis of crop nitrogen nutrition status is achieved. Through spatial unit level calculation, it can accurately identify which locations in the target region have nitrogen deficiency or nitrogen excess, overcoming the limitations of traditional average or one-size-fits-all diagnosis methods based on limited point locations. The calculation framework based on the critical nitrogen dilution curve theory ensures that the diagnosis results conform to the basic physiological laws of crop nitrogen uptake and biomass accumulation.

[0157] S6, variable fertilization prescription map generation

[0158] Based on the nitrogen nutrition index spatial distribution map, when the spatial unit NNI value is lower than the preset NNI threshold, the fertilization decision is triggered, the nitrogen deficiency amount is calculated, and the variable fertilization prescription map is generated after deducting the soil nitrogen supply potential. Specifically, it includes:

[0159] (1) Fertilization decision trigger:

[0160] According to the crop type and the current growth period, set the NNI threshold, read the real-time NNI value of each spatial unit in the target region based on the NNI spatial distribution map, and when the real-time NNI value of the spatial unit is lower than the NNI threshold, trigger the fertilization decision of the spatial unit. The NNI threshold in the tillering stage is lower than that in the jointing stage.

[0161] In a specific embodiment, the NNI threshold is set to 0.90 or 0.95.

[0162] (2) Nitrogen deficiency amount calculation:

[0163] For the spatial unit that triggers the fertilization recommendation, read its critical nitrogen concentration, actual nitrogen concentration and actual biomass and calculate the nitrogen deficiency amount. The nitrogen deficiency amount is the amount of nitrogen required for the spatial unit to recover to the optimal nitrogen nutrition state (NNI≈1), and the calculation process is represented as:

[0164] N deficit =(N c N actual )×W actual ;

[0165] In the formula, N deficit is the nitrogen deficiency amount, N c is the critical nitrogen concentration under the current biomass, N actual is the real-time actual nitrogen concentration of the crop, and W actual is the real-time actual biomass of the crop.

[0166] (3) Recommended nitrogen application rate calculation:

[0167] For the spatial unit that needs fertilization, the soil nitrogen supply potential (SoilN supply ) that is the effective nitrogen amount that the soil itself can provide in the future key nitrogen uptake period can be estimated by a simplified empirical model. A specific estimation method is:

[0168] SoilN supply = SOM x k min

[0169] In the formula, SOM is the soil organic matter content of the spatial unit (which can be directly obtained from the spatio-temporal database constructed in step S1); k min is the soil nitrogen mineralization coefficient, which is an empirical coefficient, and its value can be determined by consulting local agricultural research reports or setting a regional general value according to the soil type and climate conditions of the target area (for example, for paddy soil, it is set to 2%).

[0170] Subtract the estimated soil nitrogen supply potential from the nitrogen deficiency amount to obtain the net nitrogen demand N net_demand :

[0171] N net_demand = N deficit SoilN supply ;

[0172] Specifically, when N net_demand ≤ 0, it means that the soil supply has met or exceeded the deficiency, and the recommended nitrogen application rate is 0 kg N / ha.

[0173] The nitrogen fertilizer applied to the soil is not all absorbed by the crops, and the recommended application amount needs to be calculated according to the estimated nitrogen use efficiency (NUE). In the present application, the estimated nitrogen use efficiency (NUE) can be determined by consulting the pre-set "NUE reference lookup table". The "NUE reference lookup table" is pre-established according to a large amount of localized test data or published literature. The lookup table provides an empirical benchmark NUE value according to the combination of several key influencing factors (such as crop growth period, fertilizer type, application method, soil texture, etc.). For example, the lookup table can specify that the NUE value under the condition of "rice, jointing stage, urea, deep application" is 40%-50%.

[0174] According to the NUE value obtained by looking up the table, the recommended nitrogen application amount ΔN recommend is calculated:

[0175] ΔN recommend = N net_demand / NUE;

[0176] Finally, adjust the recommended nitrogen application rate according to the future short-term weather forecast (e.g., the next 3 days). If the forecast predicts heavy rainfall, which may lead to nitrogen leaching, reduce ΔN (e.g., by 10-30%) or recommend split applications or post-rain fertilization. If the future short-term weather forecast predicts sustained high temperatures and drought without irrigation, adjust the fertilization timing (e.g., in the evening) or apply fertilizer after a small amount of irrigation to reduce volatilization losses and improve absorption efficiency. recommend

[0177] (4) Generation of variable rate fertilization prescription map:

[0178] Collect the recommended nitrogen application rates for all spatial units and store them as a raster layer based on spatial location to obtain the variable rate fertilization prescription map (VRT map).

[0179] Ensure that the VRT map uses a format compatible with mainstream precision agriculture equipment (e.g., Shapefile or ISOXML) to be directly imported into variable rate fertilization equipment for field fertilization operations.

[0180] The variable rate fertilization prescription map output in this step provides direct location-based nitrogen application instructions for variable rate fertilization equipment. Based on real-time nitrogen balance assessment, fertilization is recommended only in areas where it is needed, and the fertilization rate is accurately matched to the degree of deficiency, soil supply potential, and crop absorption capacity. By considering soil nitrogen supply and expected nitrogen use efficiency, we can reduce ineffective investment and waste. This avoids excessive fertilization and reduces the risk of non-point source pollution caused by nitrogen leaching, volatilization, etc. (groundwater nitrate pollution, water eutrophication, and greenhouse gas emissions). Dynamic adjustment in combination with growth stages and weather forecasts can make the fertilization strategy more consistent with crop physiological needs and actual field conditions.

[0181] S7, closed-loop learning and iterative optimization

[0182] Introduce a feedback mechanism to optimize crop growth model parameters, remote sensing inversion algorithms, or fertilization decision logic based on feedback data after applying the variable rate fertilization prescription map. Specifically, this includes:

[0183] Obtain actual feedback data at the end of the season or at harvest in fields where variable rate fertilization was applied using the variable rate fertilization prescription map generated in step S6, including crop yield, crop nitrogen uptake data at harvest, fertilization cost data, and control data.

[0184] ​Specifically, the actual yield of the crop is obtained by the yield monitoring device of the combine harvester or field trials, and a yield spatial distribution map reflecting the actual impact of variable fertilization on yield is generated. The crop nitrogen uptake data at harvest includes the nitrogen concentration of the plant (or grain) and the total nitrogen accumulation at harvest, which is used to calculate the actual nitrogen use efficiency to evaluate the resource efficiency of the fertilization strategy. The fertilization cost data is calculated by the cost of fertilizer and labor, and the output of yield and quality. The yield and nitrogen uptake data of the traditional fertilization treatment or zero nitrogen treatment set in the plot are collected as control data for comparison with the efficiency of variable fertilization.

[0185] The expected management targets (such as target yield and expected NNI) are compared with the actual feedback results. The difference is located by statistical analysis or model simulation diagnosis method. The difference reasons include crop growth model parameter deviation, remote sensing inversion algorithm deviation and fertilization decision logic deviation.

[0186] Specifically, the crop growth model parameter deviation will cause the generated pCNDC parameters or the biomass or nitrogen concentration simulated by the crop growth model to deviate significantly from the measured values in some areas.

[0187] The remote sensing inversion algorithm deviation will cause the W actual or N actual There is a systematic deviation between the ground measured data (such as vegetation index saturation leading to underestimation of biomass).

[0188] The fertilization decision logic deviation will cause the recommended fertilization amount to be generally excessive in certain soil types (such as sandy soil) or the actual nitrogen use efficiency to deviate significantly from the estimated nitrogen use efficiency.

[0189] When the difference is mainly caused by extreme environmental events (such as unforecasted long-term drought, flood) or large-scale disease and pest outbreaks that are not fully characterized in the model, the system will record the event and its impact range as a "disturbance event" and associate it with the corresponding yield loss data. When performing subsequent optimization, the data in this area can be treated as an outlier to avoid the system incorrectly attributing yield reduction caused by non-fertilization factors to the failure of fertilization decision. At the same time, these accumulated disturbance event data can be used as a basis for selecting or improving future crop model versions.

[0190] Based on the difference analysis results and accumulated feedback data, the regional scale crop nitrogen dynamic regulation method is improved, including optimizing crop growth model parameters, optimizing remote sensing inversion algorithm and optimizing fertilization decision logic.

[0191] Specifically, the optimization of crop growth model parameters is to supplement the feedback data to the spatio-temporal database, and to prioritize recalibration of key parameters that have a significant impact on nitrogen response identified in the sensitivity analysis of step S2, to improve the adaptability of the crop growth model to local environment and management practices.

[0192] Optimization of remote sensing inversion uses feedback of ground measured data and inversion results for comparison to optimize inversion model (such as by adjusting parameters, replacing algorithm and fusing multi-source remote sensing data, etc.), and improve inversion accuracy.

[0193] Optimization of fertilization decision is to adjust NNI threshold value according to feedback data, dynamically calibrate NUE lookup table (such as finding that NUE in sandy soil area is generally high, and the corresponding item in the table can be lowered from 40% to 35%), so that the recommended fertilization amount is more in line with the actual demand and optimal benefit.

[0194] This step periodically incorporates new data to dynamically adapt to climate change, main planting variety update and evolution of field management measures. By timely correcting the deviation of crop growth model parameters, the deviation of remote sensing inversion algorithm and the deviation of fertilization decision logic, the difference accumulation is suppressed to improve the reliability and robustness of the method in long-term application. Based on real production feedback continuous optimization, a self-improving closed-loop mechanism is formed to ensure continuous improvement of nitrogen fertilizer management accuracy and benefit, and the fertilization recommendation is close to the actual optimum.

[0195] Example 1

[0196] This embodiment takes the subtropical monsoon climate zone in the lower reaches of the Yangtze River as an example, and the main rice variety in the target area is japonica rice or hybrid indica rice. The specific process of applying the regional scale crop nitrogen dynamic regulation method of the application is as follows:

[0197] S1, constructing a space-time database:

[0198] Meteorological data acquisition:

[0199] Integrate daily meteorological data of national and provincial meteorological stations from 1991 to now, including maximum temperature, minimum temperature, average temperature, precipitation, sunshine duration, solar radiation, wind speed and relative humidity. Generate 1km×1km resolution daily meteorological grid data by inverse distance weighting method, and update the daily data of the current year through real-time meteorological monitoring network.

[0200] Soil data acquisition:

[0201] Extract the soil type, texture (percentage of sand, silt and clay), bulk density, organic matter content, pH value, total nitrogen, alkali-hydrolyzable nitrogen, available phosphorus and available potassium of the region from the soil survey map or digital soil map, and supplement the characteristics of paddy soil, including plough bottom depth and oxidation-reduction potential. Generate 250m×250m resolution soil raster layer by ordinary kriging method.

[0202] Crop growth dynamic data acquisition:

[0203] Field experiments are carried out in representative locations such as Nanjing, Jiaxing and Hefei in the target region for 3-5 years. The field experiments are aimed at main rice varieties such as Nanjing series and Yongyou series, and the nitrogen level gradient is 0 kg / ha, 90 kg / ha, 180 kg / ha, 270 kg / ha and 360 kg / ha (total amount of nitrogen used during the whole growth period).

[0204] The above-ground biomass, stem leaf spike organ biomass, plant total nitrogen concentration, LAI and final yield and yield component factors at the tillering stage, jointing stage, booting stage, full heading stage, grain filling stage and mature stage are obtained. In addition, the same region crop growth dynamic data in historical literature are supplemented.

[0205] Crop management measure data acquisition:

[0206] The main rice varieties, sowing period, planting density and irrigation method in each county and city in the target region are collected as crop management measure data.

[0207] Remote sensing data acquisition:

[0208] Sentinel-2 L2A image data (10m resolution, 5-day revisit period) during the rice growing season (May-October) is obtained, and image data with less than 10% cloud coverage in the target region is selected. After preprocessing such as radiation calibration, atmospheric correction, geometric precision correction and cloud mask, the remote sensing data is obtained.

[0209] Construction of spatio-temporal database:

[0210] The above multi-source data is standardized, including unifying the coordinate system, resolution (250m) and time format, integrated into the geographic information system (GIS) platform and established spatio-temporal index and association rules, to obtain the constructed spatio-temporal database.

[0211] S2, crop growth model selection and calibration:

[0212] Model selection:

[0213] According to the quantitative evaluation results in Table 1, the ORYZA (v3) model is selected as the crop growth model for rice planting in the lower reaches of the Yangtze River in this embodiment.

[0214] Model calibration:

[0215] Global sensitivity analysis was conducted using the 3-year crop growth dynamic data of representative sites to identify the key parameters that had the greatest impact on the simulation of nitrogen response, including parameters related to rice development rate (DRDV, DRDF, DRDR), relative growth rate (RGRLF), light use efficiency (RUE), specific leaf area (SLA), maximum leaf nitrogen concentration (NMAXLV), and parameters related to nitrogen uptake and distribution such as maximum root nitrogen uptake rate (NUPRMX), leaf / stem / spike nitrogen distribution coefficients (FLN, FST, FSO). The PEST tool was used in combination with the NSGA-II multi-objective optimization algorithm to calibrate the key parameters. The objectives of the model calibration were to minimize the average absolute error (MAE) of the simulated key growth stages under multiple nitrogen levels;

[0216]

[0217] The root mean square error (RMSE), normalized root mean square error (nRMSE), coefficient of determination R², and consistency index (d-index) between the simulated aboveground biomass (dry weight), plant nitrogen concentration, leaf area index (LAI), and the measured values;

[0218] The RMSE, nRMSE, R², and d-index between the simulated final yield and the measured values.

[0219] The results of the model calibration were as follows: the MAE of the simulated mature stage was less than 4 days; the nRMSE of the simulated aboveground biomass was 9.5%, the R² was 0.91, and the d-index was 0.94; the nRMSE of the simulated plant nitrogen concentration was 11.2%, the R² was 0.88, and the d-index was 0.92; the nRMSE of the simulated LAI was 13.0%, the R² was 0.85, and the d-index was 0.90; the nRMSE of the simulated final yield was 7.8%, the R² was 0.90, and the d-index was 0.95.

[0220] Model verification:

[0221] The crop growth dynamic data obtained in Jiaxing from 2018 to 2020 was used to evaluate the prediction accuracy of the crop growth model. The evaluation indicators included the simulated aboveground biomass, plant nitrogen concentration, final yield, and key growth stages.

[0222] The preset acceptance requirements were as follows: the prediction error MAE of the key growth stages was within ±7 days; the nRMSE of the predicted aboveground biomass and LAI was less than 15%; the nRMSE of the predicted plant nitrogen concentration was less than 20%; the nRMSE of the predicted final yield was less than 15%; and the d-index of each indicator was greater than 0.85.

[0223] ​In this embodiment, the verification results show that the MAE of the mature period prediction is 5.2 days (within the ±7-day threshold); the nRMSE of the aboveground biomass prediction is 12.8% (less than 15%), R² is 0.85, and d-index is 0.91; the nRMSE of the plant nitrogen concentration prediction is 15.5% (less than 20%), R² is 0.80, and d-index is 0.88; the nRMSE of the LAI prediction is 14.5% (less than 15%), R² is 0.82, and d-index is 0.89; and the nRMSE of the final yield prediction is 10.5% (less than 15%), R² is 0.86, and d-index is 0.92. The ORYZA (v3) model meets the preset acceptance requirements.

[0224] S3, spatial potential critical nitrogen dilution curve (pCNDC) parameter layer construction:

[0225] Based on the resolution of meteorological data and soil data, the target area is divided into 250m×250m spatial units, and the soil property data, meteorological data and localized parameter set of each spatial unit are input into the ORYZA (v3) model to simulate the growth process of rice with a nitrogen supply gradient of 0 kg / ha-400 kg / ha (interval 20 kg / ha).

[0226] For the simulation data at the tillering peak period, the jointing period, the booting period, the heading period and the mid-grain filling period, the lowest nitrogen application treatment of the biomass plateau period is determined by ANOVA analysis, and the biomass (W optimal ) and plant nitrogen concentration (N c ) of the treatment are recorded.

[0227] Based on the (W optimal , N c ) data points (W optimal in t / ha, N c in %) of each spatial unit, the equation N c =a×W optimal is fitted. b The pCNDC parameter a layer and the pCNDC parameter b layer of the target area with a resolution of 1 km are generated.

[0228] S4, remote sensing-based crop growth and nitrogen concentration dynamic monitoring:

[0229] Based on the vegetation index such as NDVI, EVI and WDRVI and the ground measured biomass data, a segmented regression model or a machine learning model (such as random forest regression) based on the growth period is constructed. The W actual spatial distribution map (unit: t / ha) at multiple key growth stages in the rice growing season is generated.

[0230] The red edge band of Sentinel-2 is used to calculate the NDRE and CI rededge index, and machine learning methods such as partial least squares regression (PLSR) or gradient boosting decision tree (GBDT) are used to construct N actual estimation model combined with ground measured plant nitrogen concentration data, to generate N actual spatial distribution map (unit: %).

[0231] S5, nitrogen balance dynamic evaluation:

[0232] The obtained W actual spatial distribution map is spatially overlaid and matched with the obtained pCNDC parameter a layer and parameter b layer, and calculation is performed at a resolution of 250 m in this embodiment.

[0233] Critical nitrogen concentration (N c ) is calculated per unit: N c = a x W actual -b , to generate N c spatial distribution map. For example, W actual = 5 t / ha, a = 5.0, and b = 0.4, then N c = 5.0 x 5 -0.4 ≈ 2.63%.

[0234] The obtained N actual spatial distribution map is spatially overlaid and matched with the N c spatial distribution map. Nitrogen nutrition index (NNI) is calculated per unit: NNI = N actual / N c , to generate NNI spatial distribution map. The NNI value range is classified and displayed: NNI < 0.90 is nitrogen deficiency, NNI = 0.90-1.05 is suitable, and NNI > 1.05 is excessive.

[0235] S6, variable fertilization prescription map generation:

[0236] Based on the NNI spatial distribution map, the actual biomass of crops, the growth period of crops, the estimated soil nitrogen supply potential, the expected nitrogen fertilizer use efficiency, and the future short-term weather forecast are comprehensively considered to generate a variable fertilization prescription map. Taking the early jointing stage of nitrogen fertilizer as an example, the NNI threshold is set to 0.90:

[0237] When the NNI of a spatial unit is less than 0.90, the nitrogen deficiency amount is calculated: N deficit = (N c N actual ) x W actual . In this embodiment, Nc = 2.63%, N actual= 2.2%, W actual = 5 t / ha, then N deficit = (0.0263 - 0.0220) x 5000 kg / ha = 0.0043 x 5000 kg / ha = 21.5 kg N / ha.

[0238] According to the soil data, the soil nitrogen supply potential is estimated to be 15 kg N / ha. Subtracting the soil nitrogen supply potential, the net nitrogen requirement N net_demand : N net_demand = N deficit Soil N supply In this embodiment, the spatial unit N net_demand = 21.5 kg / ha - 15 kg / ha = 6.5 kg / ha.

[0239] The current rice jointing stage has a high absorption efficiency, and the urea is combined with the scattering method and the field management level, and the estimated nitrogen use efficiency (NUE) is 40%, ΔN recommend = N net_demand / NUE, in this embodiment, the spatial unit ΔN recommend = 6.5 kg N / ha / 0.40 = 16.25 kg N / ha.

[0240] According to the future short-time weather forecast, if there is a lot of rainfall in the next 3-7 days, reduce the amount of this time fertilization (adjust 10%-30% as follows) or suggest to apply after rain. If there is no extreme weather in the next 3-7 days, maintain the calculated value.

[0241] According to the recommended nitrogen application amount of each spatial unit, a Shapefile or ISOXML format raster map, i.e. a variable fertilization prescription map (VRT map), is generated.

[0242] The recommended nitrogen application amount is limited to 15 kg N / ha-60 kg N / ha. If the recommended nitrogen application amount is 16.25 kg N / ha, adjust it to an integer according to the agricultural requirements, and set the upper and lower limits of application, the minimum is 15 kg N / ha, and the maximum is 60 kg N / ha.

[0243] S7, closed loop learning and iterative optimization:

[0244] Feedback data collection:

[0245] In the field applying VRT, yield spatial distribution map is obtained using combine harvester with yield monitor. Plant samples are collected from representative sampling points (selected from different fertilization management levels based on NNI map and VRT map) to measure plant nitrogen concentration and total nitrogen accumulation at harvest stage, and actual nitrogen use efficiency (NUE) is calculated. Meanwhile, fertilizer and operation cost are collected, and economic benefit data are calculated combined with yield and quality (such as protein content).

[0246] Effect evaluation and difference analysis:

[0247] Compare the expected target (such as target yield, NNI improvement) with the actual feedback result. Analyze the difference reasons, which may include: model parameter deviation (such as inaccurate ORYZA model parameters), remote sensing inversion error (large deviation between Wactual or Nactual and actual measurement), decision logic defect (such as unreasonable NUE estimation leading to systematic high or low recommended fertilizer amount, for example, recommended fertilizer amount generally exceeds or is 20 kg N / ha lower than the actual optimal fertilizer amount), or external interference factors such as extreme weather, pests and diseases that are not predicted.

[0248] Iteration optimization:

[0249] Optimize crop growth model parameters: supplement the collected yield, biomass, nitrogen content and other data to the spatio-temporal database of S1, and recalibrate or fine-tune the parameters in ORYZA (v3) model that have significant impact on nitrogen response.

[0250] Optimize remote sensing inversion algorithm: use ground measured data to improve the inversion model of Wactual and Nactual.

[0251] Optimize fertilization recommendation algorithm: according to the actual fertilization effect and economic benefit, adjust the NNI triggering threshold in S6, the SoilNsupply estimation model, the NUE preset value (for example, if it is found that the NUE of a sandy soil area is generally lower than 40%, it can be lowered to 35%) or the coefficient in the recommendation formula.

[0252] This embodiment fully demonstrates the technical process of regional rice nitrogen precision management through the construction of spatio-temporal database, localized model calibration, spatialized pCNDC generation, remote sensing dynamic monitoring, variable fertilization prescription generation and closed-loop optimization. Regional heterogeneity is accurately quantified through multi-scale data processing of 250m-1km; timeliness of topdressing decision is realized based on real-time remote sensing data and model diagnosis; nitrogen fertilizer application amount is reduced based on real-time NNI diagnosis; crop growth model parameters are continuously optimized through yield feedback, improving the robustness of long-term application of technical solutions.

[0253] Example 2

[0254] The embodiment is directed to the temperate monsoon climate zone of North China Plain, where drought occurs frequently in winter and spring, and winter wheat-summer corn rotation is adopted. Steps S1-S5 are consistent with embodiment 1, and the difference points include: soil data supplement soil moisture characteristic parameters; crop growth model selection CERES-Wheat model, strengthening water and nitrogen simulation); the inversion resolution is 20m, and the adaptive variable fertilizer machine is adapted. The implementation process of the remaining steps is as follows:

[0255] S6, variable fertilization prescription map generation:

[0256] At the early jointing stage, when the spatial unit NNI is less than 0.90, the fertilization decision is triggered. The nitrogen deficit amount Ndeficit_total is calculated, and the soil seasonal nitrogen supply potential and the residual nitrogen amount of the previous season (corn) estimated according to the soil data are deducted, combined with jointing water and jointing fertilizer, good water condition is beneficial to improve NUE, according to the statistical data of recent years, it is 42%, the recommended fertilization amount is adjusted N = (Ndeficit_total + jointing to maturity nitrogen demand increment-soil supply) / NUE. Set the fertilization range to 30 N / ha-90 kg N / ha).

[0257] Based on the recommended nitrogen application amount of all spatial units, a 20m resolution ISOXML format variable fertilization prescription map is generated.

[0258] S7, closed loop learning and iterative optimization:

[0259] The yield map, plant / grain nitrogen content at harvest, actual nitrogen fertilizer utilization efficiency and grain protein content are calculated; the influence of variable fertilization on yield, protein content and nitrogen fertilizer utilization efficiency is analyzed, and the CERES-Wheat model parameters, remote sensing inversion model and fertilization recommendation algorithm (based on economic benefit feedback, adjusting NUE preset value and irrigation correlation coefficient) are optimized.

[0260] The embodiment proves that the application solves the problems of winter and spring drought and high efficiency of nitrogen fertilizer in wheat production in North China Plain, nitrogen supply heterogeneity caused by straw returning and quality and yield synergistic regulation through water and nitrogen simulation, stress precise diagnosis and dynamic decision optimization. Compared with the prior art, the method realizes the significant benefits of reducing irrigation water and nitrogen fertilizer by specific values under the premise of ensuring food security.

[0261] The above description of the application and its embodiments is illustrative and not restrictive, and the application can be practiced in other specific forms without departing from the spirit or essential character thereof. The drawings are intended to be illustrative, and not limiting, and the appended claims should not be limited to the drawings. Thus, if a person of ordinary skill in the art is inspired to design a similar structure and embodiment to the technical solution without departing from the spirit of the invention, it should be within the scope of protection of the patent. In addition, the word "comprising" does not exclude other elements or steps, and the word "one" before an element does not exclude the inclusion of "multiple" such elements. Multiple elements stated in a product claim can also be implemented by one element through software or hardware. The words "first", "second", etc. are used to indicate names, not any specific order.

Claims

1. A method for dynamic regulation of crop nitrogen at a regional scale, comprising the following steps: Multi-source data of the target area are acquired to construct a spatiotemporal database. The multi-source data includes meteorological data, soil data, crop growth dynamic data, crop management measures data, and remote sensing data. Select a crop growth model with a nitrogen response module, calibrate and validate the crop growth model based on a spatiotemporal database, obtain a crop growth model that reflects the dynamic response of crop nitrogen in the target area, and generate a localized model parameter set. The selection of crop growth models employs a quantitative evaluation system to screen candidate crop growth models. These models include a nitrogen response module and are applicable to the main cultivated varieties in the target region. The quantitative evaluation system includes: The data fit dimension is used to evaluate the degree of matching between the model's input requirements and the spatiotemporal database; Accessibility and scalability are dimensions used to evaluate the openness of the model code and the ease of parameter adjustment; Case studies and literature support dimension, used to count the number of application cases of the model in the target crop type; The target area is divided into spatial units. Soil data, meteorological data and localized model parameter sets of each spatial unit are input into the crop growth model. The crop growth process under nitrogen gradient is simulated under no water stress conditions. The optimal biomass and critical nitrogen concentration data point pairs of each spatial unit are extracted and the spatialized pCNDC parameter layer of the target area is fitted. Based on remote sensing data, the actual biomass and nitrogen concentration spatial distribution map are inverted. The critical nitrogen concentration and nitrogen nutrient index of each spatial unit are dynamically calculated by combining the pCNDC parameter layer and spatial distribution map is generated respectively. The nitrogen deficit is calculated based on the generated spatial map. The variable fertilizer prescription map is generated by combining the soil nitrogen supply potential and nitrogen fertilizer use efficiency to guide fertilization and realize the dynamic regulation of crop nitrogen at the regional scale. The steps for retrieving the actual biomass and nitrogen concentration spatial distribution map based on remote sensing data include: acquiring remote sensing data covering the key growth stages of crops in a spatiotemporal database, and calculating vegetation indices sensitive to biomass. A biomass inversion model is established based on vegetation index and ground-measured biomass to generate a spatial distribution map of actual biomass covering the target area and corresponding to the time phase of remote sensing data. Based on the spectral features of plants sensitive to nitrogen in remote sensing data, including red edge parameters, chlorophyll index, nitrogen reflectance index or specific band combination spectral information, an actual nitrogen concentration inversion model is established by combining the measured nitrogen concentration of plants on the ground to generate a spatial distribution map of the actual nitrogen concentration corresponding to the time phase of the remote sensing data.

2. The regional-scale crop nitrogen dynamic regulation method according to claim 1, characterized in that: After fertilization is carried out according to the variable fertilization prescription map, feedback data is collected and added to the spatiotemporal database. The feedback data includes crop yield, crop nitrogen uptake data at harvest, fertilization cost data, and control data. Based on the feedback data, the crop growth model is optimized, the remote sensing data inversion is optimized, and the threshold is adjusted to form a closed-loop control mechanism, thereby achieving continuous optimization of the regional-scale crop nitrogen dynamic control method.

3. The regional-scale crop nitrogen dynamic regulation method according to claim 1, characterized in that: The steps for calibrating the crop growth model include: setting parameters for the crop growth model; The sensitivity of crop growth model parameters to output variables was analyzed using a global sensitivity analysis method, and the key model parameters with the greatest impact on nitrogen response simulation were identified. By utilizing crop growth dynamic data from a spatiotemporal database and employing a multi-objective optimization algorithm to calibrate key model parameters, the overall performance of the crop growth model is optimized across the entire nitrogen response range, from nitrogen deficiency to nitrogen abundance.

4. The regional-scale crop nitrogen dynamic regulation method according to claim 3, characterized in that: In the simulated crop growth process with different nitrogen gradients, the treatments with different nitrogen gradients were divided into the following groups based on the value of biomass: the low biomass group with the smallest value, the high biomass group or plateau group with the largest value, and the rest were medium biomass groups. Within the high biomass group or plateau period group, the treatment with the lowest nitrogen application rate is selected, representing the minimum nitrogen supply required for the crop to reach its maximum growth rate during the current growth stage. The corresponding aboveground biomass is recorded as the optimal biomass, and the plant nitrogen concentration is recorded as the critical nitrogen concentration, thus obtaining the optimal biomass and critical nitrogen concentration data point pairs for the spatial unit.

5. The regional-scale crop nitrogen dynamic regulation method according to claim 4, characterized in that: Based on the data point pairs of optimal biomass and critical nitrogen concentration for each spatial cell, a power function equation for the critical nitrogen dilution curve is fitted to obtain the pCNDC parameters a and b for each spatial cell. The power function equation for the critical nitrogen dilution curve is expressed as: N c =a×W optimal b ; In the formula, W optimal N represents biomass. c Indicates the critical nitrogen concentration; The parameters a and b of all spatial units are collected and stored as raster layers according to their spatial location, forming a spatialized pCNDC parameter a layer and a spatialized pCNDC parameter b layer, which together constitute a spatialized pCNDC parameter layer covering the target area.

6. The regional-scale crop nitrogen dynamic regulation method according to claim 1, characterized in that: The actual biomass spatial distribution map is spatially overlaid with the spatialized pCNDC parameter layer, and spatial cell matching is performed to obtain the actual biomass and parameters a and b for each spatial cell. The critical nitrogen concentration of the spatial cell is then calculated and expressed as: N c =a×(W actual ) b ; In the formula, N c W represents the critical nitrogen concentration. actual Indicates actual biomass; The critical nitrogen concentrations of all spatial units are collected and a spatial distribution map of the critical nitrogen concentrations of the corresponding time phases covering the target area is generated. The actual nitrogen concentration spatial distribution map is spatially overlaid with the critical nitrogen concentration spatial distribution map, and spatial cell matching is performed to obtain the actual plant nitrogen concentration value and critical nitrogen concentration value of each spatial cell. The nitrogen nutrient index of the spatial cell is then calculated and expressed as: NNI=N actual / N c ; In the formula, NNI represents the nitrogen nutritional index, N actual This represents the actual nitrogen concentration value of the plant, N. c Indicates the critical nitrogen concentration; The nitrogen nutrient index of all spatial units is collected to generate a spatial distribution map of the nitrogen nutrient index of the corresponding time phase covering the target area; Based on the spatial distribution map of nitrogen nutrient index, the crop nitrogen nutrient status of each spatial unit is assessed by dynamically preset thresholds. When the nitrogen nutrient index is lower than the threshold, the nitrogen deficit is calculated, and a variable fertilization prescription map is generated by combining soil nitrogen supply potential and nitrogen fertilizer utilization efficiency.

7. The regional-scale crop nitrogen dynamic regulation method according to claim 6, characterized in that: The steps for generating the variable fertilization prescription map include: reading the critical nitrogen concentration, actual nitrogen concentration, and actual biomass of each spatial unit, and calculating the nitrogen deficit, expressed as: N deficit =(N c N actual )×W actual ; In the formula, N deficit Nitrogen deficit, N c The critical nitrogen concentration for the current biomass, N actual W represents the real-time actual nitrogen concentration of the crop. actual This refers to the actual biomass of the crop in real time. Subtracting the soil's nitrogen supply potential from the nitrogen deficit yields the net nitrogen demand (N). net_demand , represented as: A net_demand =N deficit SoilN supply ; Calculate the recommended nitrogen application rate ΔN to estimate nitrogen fertilizer use efficiency. recommend , represented as: ΔN recommend =N net_demand / NUE; Among them: SoilN supply NUE represents the soil's nitrogen supply potential, and NUE is the estimated nitrogen fertilizer use efficiency. The recommended nitrogen application rates for all spatial units are collected and stored as raster layers according to spatial location to obtain a variable fertilization prescription map.

8. A regional-scale crop nitrogen dynamic regulation device, characterized in that, include: Memory, which stores computer programs; The processor, when executing the computer program, implements the steps of the method as described in any one of claims 1-7; The communication module is used to receive input from external meteorological data servers and remote sensing data platforms.

Images