Irrigation decision determination method considering multi-source water conversion process

Abstract

The invention discloses an irrigation decision determination method considering a multi-source water conversion process, and relates to the technical field of agricultural water conservancy irrigation, and the method comprises the steps: obtaining a multi-spectral image and a thermal infrared image of crops in a target irrigation region, calculating a vegetation index and a leaf area index, and combining with the actually measured growth vigor data of the crops, a growth vigor model of crops is constructed by establishing a mapping relation, the rainfall, the irrigation volume, the canal system infiltration replenishment volume and the groundwater capillary rise volume of a target irrigation area are obtained, and a soil moisture income and expenditure model is constructed based on a water circulation process. According to the method, a plurality of paths such as rainfall, irrigation, canal system leakage and groundwater capillary rise are considered through the constructed soil water volume income and expenditure model, crop physiological moisture inflection points are accurately recognized by simulating the root zone volumetric moisture content and growth vigor response curve, threshold control irrigation based on crop requirements is achieved, and the crop yield is improved. And the multi-source input and conversion process of the farmland hydrological system is comprehensively reflected.

Description

Technical Field

[0001] This invention relates to the field of agricultural water conservancy and irrigation technology, specifically a method for determining irrigation decisions that considers the multi-source water conversion process. Background Technology

[0002] In the water cycle of farmland, the soil moisture content in the root zone of crops is not only determined by the amount of irrigation, but is also affected by a variety of factors such as rainfall, evaporation, canal seepage and capillary rise of groundwater. Moreover, these influencing factors do not exist independently, but are interconnected through processes such as infiltration, evaporation, seepage and recharge, and work together to affect the crop root zone. Furthermore, there are time lags and nonlinear transformation relationships between them.

[0003] In current agricultural irrigation management, traditional irrigation strategies often rely solely on soil moisture detection and empirical rules, along with weather forecasts to determine irrigation amounts. This lack of consideration and analysis of the multi-source water input and dynamic transformation in the crop root zone easily leads to either over-irrigation or under-irrigation, consequently affecting crop yield and water resource efficiency. For example, if capillary upwelling is ignored, in shallow-rooted areas, the system may misjudge crop water shortage and irrigate prematurely, increasing irrigation volume, resulting in water waste and increased energy consumption. If canal seepage replenishment is not considered, the degree of crop water shortage may be overestimated, leading to over-irrigation and increased energy consumption. Conversely, after heavy rainfall, if the ratio of infiltration into the soil to deep seepage is not analyzed, the subsequent evapotranspiration deficit may be underestimated, ultimately causing crop water stress. Summary of the Invention

[0004] 1) Technical problems to be solved This invention provides an irrigation decision-making method that considers the multi-source water conversion process, which can quantitatively analyze the contribution of various water sources to the root zone water content of crops at different time periods, thereby formulating corresponding irrigation strategies.

[0005] (ii) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: a method for determining irrigation decisions considering multi-source water conversion processes, comprising the following steps: Simultaneously, multispectral and thermal infrared images of crops within the target irrigation area are acquired and registered to the same coordinate system; wherein, the multispectral images are used to extract the crop vegetation index and leaf area index, and the thermal infrared images are used to extract the crop canopy temperature and soil surface temperature, and calculate the canopy water stress index. By combining the crop's vegetation index and leaf area index with the crop's measured growth data, a crop growth model is constructed by establishing their mapping relationship. This model is used to predict the quantitative evaluation value of crop growth and to obtain meteorological data to calculate the crop's reference evapotranspiration and actual evapotranspiration. The rainfall, irrigation amount, canal infiltration recharge and groundwater capillary rise in the target irrigation area are obtained, and a soil moisture budget model is constructed based on the water cycle process. In the soil moisture budget model, the input and output of water are iteratively simulated on a daily scale to calculate the root zone volumetric water content of crops within the target irrigation area. The root zone volumetric water content obtained from the current simulation is coupled with the crop growth value. By establishing the response function between crop growth and root zone volumetric water content, the response curve of crop growth with water change is obtained. The critical moisture value corresponding to the inflection point of crop growth decline is marked on the response curve, which is the starting threshold of water stress. When the simulated root zone volumetric water content is lower than the starting threshold, an irrigation decision is triggered. The irrigation water volume is set according to the difference between the current moisture content and the target optimal moisture content.

[0006] Furthermore, the acquired multispectral image of the crop includes red, green, blue, and near-infrared light bands; The acquired thermal infrared and multispectral images of the crop are unified into the same geographic coordinate system using spatial registration technology.

[0007] Furthermore, vegetation indices reflecting crop growth status are calculated using each light band in the registered multispectral image; wherein, the chlorophyll content of crops is reflected by selecting the red light band and the near-infrared light band, and the leaf area index is calculated to reflect the size of crop leaf area covering a unit area of ​​ground. The vegetation-covered area and the bare soil area are divided from the registered thermal infrared image. The average temperature of the corresponding area is extracted to represent the canopy temperature of the crop and the soil surface area. The canopy temperature is compared with the theoretical set temperature corresponding to the conditions of sufficient water supply and water shortage to calculate the crop water stress index.

[0008] Furthermore, the obtained measured crop growth data includes plant height and chlorophyll content. A crop growth model is established using regression analysis. The calculated vegetation index and leaf area index are used as inputs to the growth model, and the measured growth data are used as outputs to train the growth model for predicting crop growth. The crop predicted growth value output by the growth model is a normalized value.

[0009] Furthermore, after acquiring meteorological data including thermal radiation, soil heat flux, temperature, humidity, wind speed, and saturated vapor pressure difference, a reference evapotranspiration is calculated using the Peyman formula. The water stress index is then used as a regulating factor and introduced into the crop coefficient corresponding to the planted crop. The corrected crop coefficient is as follows: ; in, The crop coefficient corresponding to the planted crop. The water stress index; The actual evapotranspiration of the crop is calculated based on the corrected crop coefficient and the reference evapotranspiration. ; in, For reference purposes, the evapotranspiration rate is used; when the crop's water stress index is higher, the crop's transpiration activity is restricted, the actual evapotranspiration rate decreases, and the crop's water requirement increases.

[0010] Furthermore, the soil moisture budget model constructed based on the water cycle process is used to track the contribution of input multi-source moisture data to water cycle transformation and to analyze its temporal evolution; wherein, the architecture of the soil moisture budget model includes an input module, an output module, a transformation module and an iteration module. The input module takes the rainfall, irrigation amount, canal infiltration recharge amount and groundwater capillary rise amount as inputs; wherein, the canal infiltration recharge amount is the lateral recharge water infiltrating into the target irrigation area from the canal system, and the groundwater capillary rise amount is the vertical recharge water from shallow groundwater that is replenished to the crop root zone through capillary action. The conversion module uses infiltration function, capillary rise equation and channel infiltration model to convert various input quantities in the input module into usable water in the root zone in a nonlinear manner according to soil moisture content, soil water potential gradient and groundwater depth. The output module simulates the output of water dissipation and loss based on four paths: actual evapotranspiration of the crop, deep seepage, surface runoff, and groundwater evaporation.

[0011] Furthermore, the soil moisture budget model employs a daily-scale iterative simulation mechanism, dynamically iterating the root zone volumetric water content using the water balance equation; specifically: During the dynamic iteration of the water balance equation, the iteration module performs tag tracking for each input and each output; among them, rainfall and irrigation amount first calculate the effective replenishment amount determined by soil infiltration rate, the remainder forms surface runoff, and when the water content exceeds field capacity, the irrigation amount is converted into deep infiltration amount. After each update, the soil moisture budget model outputs the daily root zone volumetric water content and the contribution rate of each water source to that water content, and continues to adjust the input-output ratio of each module in the next step.

[0012] Furthermore, the growth value sequence predicted from the crop growth model is obtained, which is used to quantify the physiological health of the crop at a certain moment. The root zone volumetric water content sequence output from the soil moisture budget model is obtained, which is used to represent the abundance of water in the root zone soil at each moment.

[0013] Furthermore, based on the same time series, the root zone volumetric water content sequence is used as the independent variable, and the growth value sequence is used as the dependent variable. The relationship between the two is fitted using a Logistic curve to establish the physiological response curve of the crop to water status. The bending trend of the fitted curve is used to extract the inflection point of growth decline and define it as the water stress initiation threshold. The water stress initiation threshold indicates the soil moisture content level when the crop begins to be significantly limited by insufficient water.

[0014] (iii) Beneficial effects: Compared with the prior art, this invention has the following beneficial effects: This invention combines vegetation indices, leaf area indices, and water stress indices extracted from multispectral and thermal infrared images to dynamically monitor the true growth and water status of crops, enabling quantifiable perception and diagnosis of water shortage levels.

[0015] The constructed soil water budget model considers multiple water source transformation pathways such as rainfall, irrigation, canal seepage, and capillary upwelling of groundwater. By simulating the root zone volumetric water content and growth response curve, it accurately identifies the inflection point of crop physiological water, realizes threshold-controlled irrigation based on crop needs, and comprehensively reflects the multi-source input and transformation process of the farmland hydrological system. Attached Figure Description

[0016] Figure 1 A flowchart illustrating an irrigation decision-making method considering multi-source water conversion processes, provided in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the acquisition of multispectral and thermal infrared images of crops within an irrigation area in an irrigation decision determination method that considers multi-source water conversion processes, as provided in an embodiment of the present invention. Figure 3 The flowchart illustrates a method for determining irrigation decisions that considers the multi-source water conversion process, provided in an embodiment of the present invention, from acquiring various remote sensing images of crops to coupling various growth and water source data. Detailed Implementation

[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0018] In the description of this invention, it should be understood that the terms "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0019] Furthermore, the terms "first" and "second" are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0020] It should be noted that, where there is no conflict, the features in the embodiments of the present invention can be combined with each other.

[0021] Combination Figures 1 to 3 This paper presents an irrigation decision-making method that considers multiple water conversion processes. This method constructs a crop growth-water stress response mechanism using multispectral and thermal infrared remote sensing data, combines it with a soil water budget model based on the water cycle process, and integrates the dynamic contribution of multiple water conversion processes to soil moisture content. This enables accurate identification of crop water demand and intelligent irrigation triggering, thereby effectively improving irrigation timeliness and water use efficiency, and alleviating the shortcomings of traditional irrigation schemes in terms of accuracy and adaptability.

[0022] For reference here Figure 1 First, in step S10: simultaneously acquire multispectral and thermal infrared images of crops within the target irrigation area and register them to the same coordinate system. The multispectral images are used to extract the crop's vegetation index and leaf area index, while the thermal infrared images are used to extract the crop canopy temperature and soil surface temperature, and to calculate the canopy water stress index. In essence, in this embodiment of the invention, it is necessary to first acquire two types of remote sensing data for the target irrigation area: multispectral images and thermal infrared images. These images can come from drones, satellites, or ground imaging systems, and it is essential to ensure that the weather and crop condition are good during imaging.

[0023] After acquisition, the two types of images are first unified into the same geographic coordinate system using spatial registration techniques to ensure spatial consistency at the pixel level across different data sources. In other embodiments of the invention, registration can be accomplished using methods such as ground control points (GCPs) and affine transformations, cross-correlation matching, etc.

[0024] Next, vegetation parameters reflecting crop growth status are extracted from the registered multispectral images. The most commonly used parameter is NDVI (Normalized Difference Vegetation Index), which reflects chlorophyll content and vegetation density using red and near-infrared bands. Its calculation formula is as follows:

[0025] ; in, It is in the near-infrared band, which strongly reflects crop health; It is the red light band, which reflects the degree to which it is strongly absorbed by chlorophyll.

[0026] NDVI reflects vegetation growth density and vigor, typically ranging from -1 to +1. Additionally, the Leaf Area Index (LAI) can be estimated, reflecting the leaf area of ​​crop cover per unit area of ​​ground; its calculation formula is:

[0027] ; This requires obtaining regression coefficients from multiple collected sample data points beforehand, thereby calculating each coefficient. , and The leaf area index is calculated to further characterize the degree of vegetation growth. Furthermore, these parameters will be used in subsequent crop growth models.

[0028] In summary, it can be understood that NDVI reflects chlorophyll content and photosynthetic capacity, while LAI represents the ratio of total leaf area to ground projection area, thus comprehensively reflecting crop growth density.

[0029] Meanwhile, crop canopy temperature and bare soil surface temperature can be extracted from thermal infrared images. To obtain accurate temperature values, the images need radiometric calibration, atmospheric correction, and consideration of surface emissivity differences. Then, based on the NDVI distribution of the images, they are divided into "vegetated areas" and "bare soil areas," as shown below. Figure 2 As shown, the average temperature of the “vegetation-covered area” and the “bare soil area” is then extracted to obtain the canopy temperature and soil surface temperature.

[0030] Specifically, the original thermal infrared image provides the radiation brightness temperature, which needs to be converted to the actual temperature: ; in, The emissivity is obtained through radiation calibration. and The constants provided by the satellite or thermal imager are not specifically limited here.

[0031] Once the canopy temperature is available, the crop water stress index (CWSI) can be calculated. It's important to understand that the principle behind the CWSI can be understood as follows: under sufficient water conditions, crops cool down through transpiration, and the canopy temperature approaches a "wet limit"; conversely, when crops are short of water, stomata close, transpiration is limited, and the crop canopy temperature rises, approaching a "dry limit."

[0032] Therefore, by comparing the actual canopy temperature with the theoretical temperatures under the two extreme conditions of "ample water supply" and "severe water shortage," the degree of water stress can be determined. The CWSI value ranges from 0 (no stress) to 1 (extreme stress). The theoretical boundary temperature can be estimated from meteorological data or determined through field measurements in wet and dry reference zones. The formula for calculating the water stress index is:

[0033] ; in, The actual measured canopy temperature. This refers to the theoretically assumed crop temperature under full irrigation (sufficient water), i.e., the maximum transpiration rate; while These are theoretically assumed temperatures for severely water-deficient crops, i.e., minimum transpiration. These two reference temperatures can be estimated from meteorological data or determined through field measurements in wet and dry reference zones.

[0034] In summary, it can be understood that the above steps yielded two key parameters: one is a vegetation index that characterizes the crop's growth status, reflecting the crop's growth level and serving as the growth input for subsequent modeling; the other is a water stress index that reflects the crop's current water shortage status, i.e., reflecting the current degree of water shortage, serving as the water input for subsequent modeling.

[0035] The two are combined to construct crop growth models (correspondence between NDVI / LAI and measured growth) and to provide a reference for subsequent evapotranspiration. The correction provides parameters to obtain the actual evapotranspiration, which is then incorporated into the subsequent water budget model.

[0036] After obtaining representative crop growth parameters, step S20 involves combining crop vegetation index and leaf area index with measured crop growth data to establish a mapping relationship and construct a crop growth model. This model is used to predict a quantitatively evaluated crop growth value, and meteorological data is acquired to calculate the crop's reference evapotranspiration and actual evapotranspiration. The task of this step is to use remote sensing indicators to predict crop growth and, combined with meteorological data, calculate the crop's water requirement under that growth condition, providing a basis for subsequent irrigation strategy decisions.

[0037] The core objective of constructing a crop growth model is to establish a mathematical mapping relationship between the two types of remote sensing variables acquired in the previous step: NDVI, which reflects vegetation density and photosynthetic capacity, and LAI (leaf area index), which characterizes the leaf area per unit area of ​​the ground and is a parameter of crop photosynthetic efficiency and yield potential. This mapping relationship yields a growth value that can quantify the crop's growth level.

[0038] In some embodiments of the present invention, the most common approach is to use regression methods, such as NDV and LAI, to perform multiple regressions with the measured aboveground biomass or plant height, respectively, to obtain a linear mapping equation. Furthermore, crop growth curves, such as sigmoid curves or logistic curves, can be fitted using time-series NDVI and LAI to predict growth trends at different stages.

[0039] In summary, it can be understood that the crop growth model ultimately obtained through the above mapping is a function that can transform remote sensing parameters (NDVI, LAI) into quantifiable growth indicators. The specific form of this function includes, but is not limited to, empirical regression equations and time-series growth curve fitting, and is not specifically limited here.

[0040] The quantified growth value predicted by the aforementioned crop growth model can be understood as transforming the complex growth state of the crop into a standardized numerical value to represent the crop's growth level at the current stage. In some embodiments of this invention, the quantification methods for the growth value include, but are not limited to:

[0041] Quantified as biomass / dry matter per unit area ( or This method is the most intuitive quantitative growth indicator; Quantification is expressed as a growth index (i.e., a dimensionless normalized value with a range of 0-1 or 0-100). This method maps measured biomass or plant height to a relative index by standardizing the measured biomass or plant height. For example, NDVI itself is an index in the range of 0-1. Growth vigor can be directly expressed by normalized biomass or leaf area index. The quantification is based on yield-related indicators. This approach takes into account that some growth models will associate growth value with the yield at each stage, using the potential yield percentage as the quantification result. For example, 70% means that the growth has reached 70% of the potential level.

[0042] It is important to note here that, regardless of quantification... The aboveground biomass, or Both of these categories, representing dry matter accumulation, are absolute values; quantifying them as growth indices, however, results in standardized index values, which facilitate comparisons with different crops or regions.

[0043] By using a pre-constructed growth model, remote sensing indicators are input into the model, which outputs predicted crop growth values. This growth model not only reflects the current growth status but also predicts crop growth trends in subsequent simulations under different water conditions, meteorological conditions, or irrigation strategies. The predicted growth values ​​can be coupled with root zone soil moisture content in subsequent processing to establish a water-growth response relationship, thereby determining the crop water shortage inflection point and irrigation trigger threshold.

[0044] In actual irrigation decisions, relying solely on reference evapotranspiration calculated from meteorological data ( ) and crop water requirements ( These models often fail to accurately reflect the actual water consumption of crops because they assume that crops are in an ideal growth state with sufficient water and no stress.

[0045] However, in real-world field environments, crops are often subjected to varying degrees of water stress, and their transpiration rate decreases due to stomatal closure, resulting in actual water consumption being lower than theoretical values. Therefore, to more accurately assess the true water requirements of crops under current conditions, it is necessary to introduce remote sensing methods to obtain information on the crop's "water status."

[0046] Thermal infrared images can be used to monitor crop canopy temperature. Combined with temperature thresholds for theoretical saturated transpiration and drought conditions, the Canopy Water Stress Index (CWSI) can be calculated to quantify the degree of water stress currently experienced by the crop. This index typically ranges from 0 to 1, with 0 indicating no stress and 1 indicating severe stress. The CWSI reflects the crop's ability to "transpire effectively," representing the degree to which actual water conditions inhibit the crop's water consumption process.

[0047] Based on this, in order to make the crop water requirement model more realistic, in some embodiments of the present invention, the water stress index obtained above is introduced as a regulating factor into the crop coefficient ( In the process, the crop coefficient is dynamically adjusted. The adjusted crop coefficient is then used for dynamic correction. It can be calculated in the following way:

[0048] ; in, The crop coefficient corresponding to the planted crop. This represents the water stress index.

[0049] The actual evapotranspiration of the crop is calculated based on the corrected crop coefficient and the reference evapotranspiration. ; in, For reference purposes, the evapotranspiration rate is used; when the crop's water stress index is higher, the crop's transpiration activity is restricted, the actual evapotranspiration rate decreases, and the crop's water requirement increases.

[0050] The core logic of this step can be understood as follows: the higher the CWSI, the more severe the water stress and the weaker the crop transpiration capacity. Therefore, the actual water demand should be reduced accordingly. Through this correction process, the crop water status perceived by remote sensing is effectively fed back into the water demand calculation model, thereby improving the real-time performance and accuracy of water demand prediction and providing a more scientific reference for subsequent irrigation decisions.

[0051] Next, S30 is performed: Rainfall, irrigation volume, canal infiltration recharge, and groundwater capillary upwelling are obtained at the target irrigation area, and a soil moisture budget model is constructed based on the water cycle process. To accurately predict the trend of soil moisture content changes in the crop root zone, a soil water budget model based on the water cycle process needs to be constructed. Then, S40 is performed: In the soil moisture budget model, the input and output of water are iteratively simulated on a daily scale to calculate the volumetric water content of the crop root zone within the target irrigation area.

[0052] The logic of building the model in S30 can be understood as comprehensively calculating all input and output items that affect soil moisture within a certain time step (such as a daily scale) and gradually simulating the cumulative changes in soil moisture.

[0053] Specifically, in order to accurately simulate the spatiotemporal dynamics of soil moisture in the root zone and quantify the transformation paths and contributions of multiple sources of water in the entire water cycle, a soil water balance model that iterates over time is constructed.

[0054] In some embodiments of the present invention, the income and expenditure model is constructed as a four-layer architecture framework, which is represented in the form of modules: input module, transformation module, output module, and iteration module. The specific description of each module is as follows.

[0055] The first module is the input module, which defines the external water supply to the soil moisture system. This can be understood as quantifying the four types of water sources in the target irrigation area into time-scale (e.g., daily) inputs suitable for root zone water calculations. It mainly includes the following four types of input water sources: Rainfall is obtained through daily rainfall measurements from weather stations or remote sensing observations. Irrigation volume is obtained from the irrigation plan or field measurements to determine the actual amount of irrigation water. The infiltration and recharge volume of the canal system, and the amount of water that seeps laterally into the soil from the canal system, can be estimated by the canal system leakage rate and water level. The capillary upwelling of groundwater is estimated based on the groundwater level depth and soil water potential gradient, representing the amount of water that rises to the root zone.

[0056] The four types of water sources were then organized into a unified input on a daily scale to provide a data foundation for the conversion module.

[0057] The conversion module is mainly used for nonlinear equation simulation of multi-source water. This module nonlinearly converts the input water source quantity into usable water in the root zone, while considering the influence of soil properties and water potential. Because each water source has different conversion paths and mechanisms, in some embodiments of this invention, to reduce the computational burden of the model, the conversion of a single type of water source is performed first, and then the conversion results of the four types of water sources are integrated into usable water in the crop's root zone.

[0058] Regarding rainfall The transformation is referenced here using the soil infiltration function: ; When the initial soil moisture content Exceeding the set field water holding capacity At this time, the infiltration rate of water decreases, and for the portion exceeding the soil's water holding capacity, deep seepage or surface runoff occurs. Therefore, rainfall... Transformation root zone water volume The calculation logic can be set as follows: ; in, The thickness of the root zone of the crop.

[0059] Regarding irrigation volume The transformation can also refer to the transformation logic of the soil infiltration function mentioned above. When there is rainfall, the infiltration demand must be met first, and then the excess is converted into deep infiltration, and the final irrigation amount... Transformation root zone water volume The calculation logic can be set as follows: ; It is important to note here that the irrigation amount The transformation needs to be done in terms of rainfall. The calculation is performed after the transformation.

[0060] Regarding the infiltration recharge of the canal system The transformation process mainly follows the logic that infiltration efficiency increases as soil moisture decreases; therefore, the corresponding transformation root zone water volume... The calculation logic can be set as follows: ; in, It is a soil water absorption efficiency function. The specific form of this function is set according to the soil type and its water absorption efficiency, and the specific form is not limited here.

[0061] Regarding the capillary rise of groundwater The transformation logic mainly refers to the corresponding transformation root zone water volume based on the groundwater level depth and soil water potential gradient. The calculation logic can be set as follows: ; in, This represents the soil water potential gradient at the location of the crop root zone. This represents the soil water potential gradient at the groundwater level depth in the crop planting area. It should be noted that both of these values ​​are derived from field surveys in the crop planting area or by referring to historical theoretical values. To correct the coefficient value, no specific limit is made here.

[0062] The conversion results of the four types of water sources were then integrated in a cumulative manner, and the cumulative calculation yielded the usable water volume in the root zone. for: ; It should be noted that although the formula is in the form of addition, the nonlinear coupling relationship between soil saturation, overlapping infiltration, groundwater level and capillary recharge has been taken into account in the transformation process of each term. Therefore, the final summation of the transformation of the four types of water sources is still a nonlinear summation.

[0063] Furthermore, in subsequent iterations, the occupation of the previous type of water source will affect the effective conversion of the next type of water source. For example, if rainfall has already met part of the root zone infiltration capacity, the remaining space will be reduced when the irrigation volume is converted. Canal infiltration and capillary rise will also be regulated by the current soil moisture content.

[0064] Regarding the operation of the output module, it considers the consumption or loss of available water in the root zone through the following four paths.

[0065] The first thing to consider is the water consumption pathway of crop evapotranspiration, and the amount of crop evapotranspiration. The calculation formula can be: ; in, Calculated from meteorological data to reference crop evapotranspiration; This refers to the crop coefficient. It should be noted that the crop coefficient for some crops is related to the growth stage, meaning that the value of the crop coefficient may be different at different growth stages. Therefore, in actual calculations, it is necessary to pay attention to the value of the crop coefficient at different growth stages. This is the soil moisture limiting function, which can be understood as the decrease in evapotranspiration under high water shortage. The specific form can be referenced from historical theoretical values, and no specific limitations are made here.

[0066] When the root zone water content exceeds field capacity, water is consumed through deep seepage. (Regarding deep seepage...) The calculation formula can be: ; in, The thickness of the root zone of the crop. The initial soil moisture content mentioned above, The field water holding capacity is set above.

[0067] When rainfall or irrigation exceeds infiltration capacity, water is consumed as surface runoff. (Regarding surface runoff...) The calculation formula can be: ; in, For rainfall, For irrigation volume, The rainfall infiltration volume is calculated using the soil infiltration function mentioned above.

[0068] The final water consumption pathway comes from the evaporation of surface soil or groundwater, depending on surface moisture content and evaporation potential, with its groundwater evaporation rate... The calculations mainly rely on historical maximum soil moisture content and reference crop evapotranspiration. The calculations of surface moisture and evaporation potential are not specifically limited here.

[0069] Integrating the above four water consumption paths, and assuming that the budget model iterates on a daily scale, the calculation of the root zone residual water volume when integrating the current root zone residual water volume is as follows: ; in, The amount of water available in the root zone after the above calculation is calculated.

[0070] Finally, there's the iteration module, which primarily updates the root zone water content and calculates the contribution rate. Specifically, assuming the budget model iterates on a daily scale, the root zone volumetric water content needs to be updated daily:

[0071] ; Then, label tracking is performed, which involves allocating the contribution ratio for each type of water source. : ; in, For each type of water source, This represents the amount of water transformed in the root zone for each type of water source. For example, the contribution rate of rainfall is... .

[0072] The results were obtained from the iterative calculations of that day. This serves as the input for the next day, and is cyclically updated throughout the entire irrigation cycle.

[0073] In summary, this tagging and tracking mechanism in the model labels the actual infiltration or recharge of each type of input water source at the current time step with its corresponding source tag, and traces it back to the output flow layer by layer through the transformation path. For example, suppose that in today's increase in root zone water content... % comes from rainfall infiltration, % comes from irrigation, The percentage originates from capillary upwelling of groundwater. The evapotranspiration that occurs tomorrow can be used to indicate the source of the water it consumes. It's yesterday's irrigation water. Thus, it's understandable that each output item can be broken down into the cumulative percentage contributed by which water sources, achieving an analysis of water source contribution rates.

[0074] To further understand this, in the previous model simulation, each type of water source does not directly enter the water balance. Instead, its portion entering the root zone is first calculated through a physical process transformation function, and the contribution path from each water source to each output item is recorded at each time step. After multiple days of time-series iterations, not only can the dynamic curve of water content be obtained, but it can also trace back to when and how much each water source contributed to crop water supply through which path in historical stages. This makes the model not only a predictor but also a diagnostic tool for water source structure optimization and irrigation strategy evaluation.

[0075] In summary, we will now integrate and understand these four modules of the model architecture.

[0076] The first is the input module, which receives multi-source water data, mainly from four types of water sources: rainfall, irrigation, canal infiltration recharge, and groundwater capillary rise. At the same time, it acquires the initial soil moisture content and soil physical parameters, such as field capacity, saturated water content, and root zone thickness.

[0077] The conversion module is used to convert each type of water source into usable water in the root zone. The conversion process for each type of water source is non-linear, depending on soil moisture content, soil water potential gradient, and soil layer thickness. The corresponding logic for the conversion process is as follows:

[0078] Rainfall infiltration is limited by the current soil moisture content and field capacity, with the excess forming runoff or deep seepage. Irrigation infiltration prioritizes replenishing excess soil moisture, with any excess water seeping into deeper layers. The infiltration rate of the canal system is adjusted according to soil moisture to improve infiltration efficiency. The capillary ascent of groundwater is affected by the soil water potential gradient in the root zone; the higher the soil moisture content, the lower the capillary ascent.

[0079] The output module is used to simulate water dissipation and loss, specifically by consuming or losing available water in the root zone through four paths, as follows: Crop evapotranspiration ; Deep leakage ; Surface runoff ; Submerged Evaporation Calculations are based on surface moisture and evaporation potential.

[0080] Finally, there is the iteration module, which is used to update the root zone water content according to a unit time scale (such as a daily scale). Specifically, the root zone water content at the end of each day is used as the initial water content for the next day, and the iteration continues while tracking the contribution rate of each water source.

[0081] In summary, the income-expenditure model constructed from these four layers integrates the four types of water sources mentioned above through nonlinear transformation functions, iterative updates, and tracking of the contribution rate of water source transformation, thereby achieving dynamic simulation of root zone water. This model can predict daily trends in root zone water content, quantify the contribution of each type of water source to root zone water volume, and consider the nonlinear coupling and temporal characteristics between water sources. It is not a simple linear superposition as in traditional techniques, but rather takes into account multi-source water transformation, soil saturation constraints, and consumption pathways, providing fundamental data for subsequent crop growth coupling and irrigation strategy formulation.

[0082] In some embodiments of this invention, regarding the final predicted root zone moisture content, to more intuitively reflect the amount of water in the soil that can be absorbed by crop roots, the predicted output is specifically the root zone volumetric moisture content. Root zone volumetric moisture content refers to the average volumetric moisture content within the main distribution layer of the crop roots (soil tillage layer or effective root zone layer). For example, the root zone of wheat is generally 0-60 cm, and the root zone of corn is between 0-100 cm. In calculations, the volumetric moisture content of this soil layer is weighted and averaged to obtain the "root zone volumetric moisture content." The commonly used unit is... or%.

[0083] After obtaining the time series of root zone volumetric water content from the water budget model and the normalized crop growth prediction value series from the crop growth model, step S50 is performed: the currently simulated root zone volumetric water content is coupled with the crop growth value. By establishing a response function between crop growth and root zone volumetric water content, a response curve of crop growth as a function of water change is obtained. This step couples the two sequences obtained above to construct a response function of crop growth as a function of soil moisture change, thereby enabling the prediction of crop growth performance based on soil moisture content and further identifying irrigation critical points.

[0084] Specifically, in the previous step, there were already two time series, one of which was the root zone volumetric water content series ( One is obtained from daily-scale iterations of the soil moisture budget model, which can dynamically reflect the actual soil moisture status; the other is the normalized crop growth prediction value sequence ( The value is output by the crop growth model and ranges from 0 to 1, representing the relative health or growth level of the crop.

[0085] The goal of step S50 is to couple these two sequences using mathematical functions to find the pattern of crop growth changing with water availability, thereby obtaining a response curve. (See reference below.) Figure 3 The curve inflection point and descent point extracted from the response curve can show at what water level crop growth will significantly decline, which represents the following water stress initiation threshold.

[0086] Regarding the method for establishing the response function between crop growth and root zone volumetric water content, in some embodiments of the present invention, the root zone volumetric water content sequence is used as the independent variable, the growth value sequence is used as the dependent variable, and the relationship is fitted by a Logistic curve to establish the physiological response curve of the crop to water status.

[0087] The reason for choosing this function is that the crop's response to soil moisture is not linear. For example, in the low moisture stage, a slight decrease in water content will rapidly impair growth (sensitive to water shortage); within a suitable range (close to field capacity), growth is not sensitive to changes in water content and remains at a high level; however, in the supersaturated stage (close to saturated water content), growth will decline due to factors such as lack of oxygen. Therefore, it can be understood that this relationship between water and crop growth resembles an S-shaped pattern of "stable in the middle and declining at both ends." Thus, in the embodiments of this invention, the logistic function is chosen as the mathematical framework for the response function.

[0088] Specifically, the general form of the Logistic function is: ; in, It is the predicted crop growth, and its normalized value is between 0 and 1; This represents the volumetric water content of the root zone. The water content at the stress critical point, or the critical inflection point, represents the critical point at which crop growth declines. As the response rate factor, it determines the steepness of the curve, indicating the sensitivity of crop growth to changes in water availability.

[0089] Based on the above general expression, and in order to better suit the upper and lower limit considerations in the embodiments of the present invention, its expression is defined as: ; in, and These represent the best and worst crop growth levels, respectively.

[0090] After constructing the response function between crop growth and root zone volumetric water content, the root zone water content sequence ( ) and predicted growth sequence ( Using paired sample points, nonlinear regression methods such as least squares fitting and gradient descent are employed to estimate the parameters of the aforementioned logistic function. , , and The resulting fitted curve is the response curve of crop growth to soil moisture.

[0091] Furthermore, in other embodiments of the present invention, since crops at different growth stages have different sensitivities, the aforementioned logistic function can be fitted separately at each growth stage to obtain different critical moisture values ​​at different stages.

[0092] Finally, S60 is performed: the critical moisture value corresponding to the inflection point of crop growth decline is marked on the response curve, and the initial threshold of water stress is set. When the simulated root zone volumetric water content is lower than the initial threshold, an irrigation decision is triggered; the irrigation water volume is set according to the difference between the current moisture content and the target optimal moisture content.

[0093] In response curve Based on this, the inflection point water content at which crop growth begins to decline significantly is identified and calibrated as the threshold for the onset of water stress, which is the aforementioned... This threshold represents the soil moisture level at which crops begin to be significantly limited by insufficient water; below this level, crops enter a state of water stress.

[0094] In practical applications, a "buffer zone" is usually set up, for example, requiring monitoring of several consecutive days. All are smaller than Only when the irrigation request is triggered will it be activated, thus avoiding false triggering due to instantaneous fluctuations.

[0095] In an embodiment of the present invention, the irrigation water volume is dynamically set based on the difference between the current moisture content and the target optimal moisture content. The target optimal moisture content can be defined as the root zone volumetric moisture content corresponding to the optimal growth or stable high-yield interval of the crop in the response function.

[0096] The calculation of irrigation water demand can be expressed as follows: ; in, The thickness of the root zone of the aforementioned crops can also be understood as depth (mm or cm). The target optimum moisture content obtained above, This indicates the difference between the current moisture state and the target moisture state.

[0097] The irrigation water demand calculated by this formula This can be used as the net water demand, and then the system efficiency can be considered. For example, the water utilization rate of the irrigation system can be further adjusted based on the actual water consumption.

[0098] In summary, this method constructs crop growth and water stress response mechanisms using multispectral and thermal infrared remote sensing data. Combined with a soil water budget model based on the water cycle process, it integrates and analyzes the dynamic contribution of multi-source water transformation to soil moisture content, thereby achieving accurate identification of crop water demand and intelligent irrigation triggering judgment. This effectively improves irrigation timeliness and water use efficiency, alleviating the shortcomings of traditional irrigation schemes in terms of accuracy and adaptability.

[0099] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. The scope of patent protection of the present invention shall be determined by the claims. Similarly, any equivalent structural changes made based on the description and drawings of the present invention shall also be included within the scope of protection of the present invention.

Claims

1. A method for determining irrigation decisions considering multi-source water conversion processes, characterized in that, Includes the following steps: Simultaneously, multispectral and thermal infrared images of crops within the target irrigation area are acquired and registered to the same coordinate system; wherein, the multispectral images are used to extract the crop vegetation index and leaf area index, and the thermal infrared images are used to extract the crop canopy temperature and soil surface temperature, and calculate the canopy water stress index. By combining the crop's vegetation index and leaf area index with the crop's measured growth data, a crop growth model is constructed by establishing their mapping relationship. This model is used to predict the quantitative evaluation value of crop growth and to obtain meteorological data to calculate the crop's reference evapotranspiration and actual evapotranspiration. The rainfall, irrigation amount, canal infiltration recharge and groundwater capillary rise in the target irrigation area are obtained, and a soil moisture budget model is constructed based on the water cycle process. In the soil moisture budget model, the input and output of water are iteratively simulated on a daily scale to calculate the root zone volumetric water content of crops within the target irrigation area. The root zone volumetric water content obtained from the current simulation is coupled with the crop growth value. By establishing the response function between crop growth and root zone volumetric water content, the response curve of crop growth with water change is obtained. The critical moisture value corresponding to the inflection point of crop growth decline is marked on the response curve, which is the starting threshold of water stress. When the simulated root zone volumetric water content is lower than the starting threshold, an irrigation decision is triggered. The irrigation water volume is set according to the difference between the current moisture content and the target optimal moisture content.

2. The irrigation decision-making method considering multi-source water conversion processes according to claim 1, characterized in that, The acquired multispectral images of the crop include red, green, blue, and near-infrared light bands; The acquired thermal infrared and multispectral images of the crop are unified into the same geographic coordinate system using spatial registration technology.

3. The irrigation decision-making method considering multi-source water conversion processes according to claim 2, characterized in that, The vegetation index that reflects the crop growth status is calculated using each light band in the registered multispectral image; wherein, the chlorophyll content of the crop is reflected by selecting the red light band and the near-infrared light band, and the leaf area index is calculated to reflect the size of the crop leaf area covering a unit area of ​​ground. The vegetation-covered area and the bare soil area are divided from the registered thermal infrared image. The average temperature of the corresponding area is extracted to represent the canopy temperature of the crop and the soil surface area. The canopy temperature is compared with the theoretical set temperature corresponding to the conditions of sufficient water supply and water shortage to calculate the crop water stress index.

4. The irrigation decision-making method considering multi-source water conversion processes according to claim 1, characterized in that, The obtained measured crop growth data includes plant height and chlorophyll content. A crop growth model is established using regression analysis. The calculated vegetation index and leaf area index are used as inputs to the growth model, and the measured growth data are used as outputs to train the growth model for predicting crop growth. The predicted crop growth value output by the growth model is a normalized value.

5. The irrigation decision-making method considering multi-source water conversion processes according to claim 3, characterized in that, After acquiring meteorological data including thermal radiation, soil heat flux, temperature, humidity, wind speed, and saturated vapor pressure difference, a reference evapotranspiration was calculated using the Perman formula. The water stress index was then used as a regulating factor and incorporated into the crop coefficient corresponding to the planted crop. The corrected crop coefficient is as follows: ； in, The crop coefficient corresponding to the planted crop. The water stress index; The actual evapotranspiration of the crop is calculated based on the corrected crop coefficient and the reference evapotranspiration. ； in, For reference purposes, the evapotranspiration rate is used; when the crop's water stress index is higher, the crop's transpiration activity is restricted, the actual evapotranspiration rate decreases, and the crop's water requirement increases.

6. The irrigation decision-making method considering multi-source water conversion processes according to claim 1, characterized in that, The soil moisture budget model, constructed based on the water cycle process, is used to track the contribution of input multi-source moisture data to water cycle transformation and to analyze its temporal evolution. The soil moisture budget model architecture includes an input module, an output module, a transformation module, and an iteration module. The input module takes the rainfall, irrigation amount, canal infiltration recharge amount and groundwater capillary rise amount as inputs; wherein, the canal infiltration recharge amount is the lateral recharge water infiltrating into the target irrigation area from the canal system, and the groundwater capillary rise amount is the vertical recharge water from shallow groundwater that is replenished to the crop root zone through capillary action. The conversion module uses infiltration function, capillary rise equation and channel infiltration model to convert various input quantities in the input module into usable water in the root zone in a nonlinear manner according to soil moisture content, soil water potential gradient and groundwater depth. The output module simulates the output of water dissipation and loss based on four paths: actual evapotranspiration of the crop, deep seepage, surface runoff, and groundwater evaporation.

7. The irrigation decision-making method considering multi-source water conversion processes according to claim 6, characterized in that, The soil moisture budget model employs a daily-scale iterative simulation mechanism, dynamically iterating the root zone volumetric water content using the water balance equation; specifically: During the dynamic iteration of the water balance equation, the iteration module performs tag tracking for each input and each output; among them, rainfall and irrigation amount first calculate the effective replenishment amount determined by soil infiltration rate, the remainder forms surface runoff, and when the water content exceeds field capacity, the irrigation amount is converted into deep infiltration amount. After each update, the soil moisture budget model outputs the daily root zone volumetric water content and the contribution rate of each water source to that water content, and continues to adjust the input-output ratio of each module in the next step.

8. The irrigation decision-making method considering multi-source water conversion processes according to claim 4, characterized in that, Obtain the growth value sequence predicted from the crop growth model, which is used to quantify the physiological health of the crop at a certain moment. Obtain the root zone volumetric water content sequence output from the soil moisture budget model, which is used to represent the abundance of water in the root zone soil at each moment.

9. The irrigation decision-making method considering multi-source water conversion processes according to claim 8, characterized in that, Based on the same time series, the root zone volumetric water content sequence is used as the independent variable and the growth value sequence is used as the dependent variable. The relationship between the two is fitted by a Logistic curve to establish the physiological response curve of the crop to water status. The bending trend of the fitted curve is used to extract the inflection point of growth decline and define it as the water stress initiation threshold. The water stress initiation threshold indicates the soil water content level when the crop begins to be significantly limited by insufficient water.

Images