A drought early warning method and system based on soil moisture detection

Abstract

The present invention provides a drought warning method and system based on soil moisture detection, which relates to the technical field of drought warning. The present invention first divides the soil evenly into a plurality of grid areas, collects data in the central area of ​​each grid, and integrates them into a sample data set, and uses an expert scoring method to assign a drought risk index to each timestamp. Then, a drought risk model is constructed based on a random forest algorithm, and the model is trained with a training set. The environmental data collected in real time is then input into the trained model to obtain the drought risk index and determine the number of grid areas with drought risk. If there are many grid areas with drought risk, the drought risk index and related data of each grid are further analyzed and compared with the drought risk threshold, thereby evaluating the current state of the soil and providing a scientific basis for drought monitoring and management.

Description

Technical Field

[0001] The present invention relates to the technical field of drought early warning, and in particular to a drought early warning method and system based on soil moisture detection. Background Art

[0002] The background of drought early warning technology is mainly due to the increasingly serious global climate change and water shortage problems. Traditional meteorological monitoring methods can no longer meet the needs of modern agriculture and ecological protection. With the development of remote sensing technology and the Internet of Things, researchers have begun to explore the use of soil moisture monitoring, plant physiological parameter collection and other methods, combined with data analysis and machine learning and other advanced algorithms to build a more accurate drought early warning system. The application of these technologies not only improves the real-time and accuracy of drought monitoring, but also provides a scientific basis for crop management, irrigation decision-making and ecological restoration, helping agricultural producers cope with uncertain climate change and ensure food security.

[0003] The prior art has the following deficiencies:

[0004] Although drought early warning methods based on soil moisture detection have significant advantages in monitoring accuracy and real-time performance, there are still some technical deficiencies that affect their effectiveness. For example, the layout of sensors may fail to cover all soil types and terrain features, resulting in missing or unrepresentative soil moisture data in some areas, which may lead to deviations in drought risk assessment. In addition, plant physiological parameters such as leaf temperature, chlorophyll content, and transpiration rate have large variability and may be interfered with by environmental factors, which limits the model's predictive ability under extreme climate conditions, thereby affecting the accuracy of real-time decision-making. Therefore, how to optimize the sensor layout and improve the model's adaptability to a changing environment remains a key challenge to improving the effectiveness of this technology.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a drought early warning method and system based on soil moisture detection to solve the problems raised in the above background technology.

[0007] To achieve the above object, the present invention provides a drought early warning method based on soil moisture detection, the specific steps of which include:

[0008] Step 1: Divide the soil area into multiple 5m×5m grids, and arrange sensors in the soil in the central area of ​​each grid to obtain soil moisture content data in each grid area;

[0009] Step 2: Place sensors on a plant in the central area of ​​each grid to collect data on plant leaf temperature, chlorophyll content, and transpiration rate. In addition, integrate soil moisture content, leaf temperature, chlorophyll content, and transpiration rate into a sample data set based on timestamps, and use an expert scoring method to assign a drought risk index corresponding to each timestamp data.

[0010] Step 3: Construct a drought risk model based on the random forest algorithm, input soil moisture content, leaf temperature, chlorophyll content and transpiration rate as training sets, and the corresponding drought risk index as labels to train the drought risk model;

[0011] Step 4: The soil moisture content, leaf temperature, chlorophyll content and transpiration rate of each grid area are collected in real time and input into the trained drought risk model to obtain the corresponding drought risk index and determine the number of grid areas with drought risk;

[0012] Step 5: Analyze the drought risk index, soil moisture content, leaf temperature, chlorophyll content and transpiration rate data of each grid area to obtain the drought risk probability of the soil area, and compare it with the drought risk threshold to determine the current soil status.

[0013] Furthermore, the specific logic for collecting soil moisture content is as follows:

[0014] After the soil area is evenly divided into 5m×5m grids, a humidity sensor is installed in the center of each grid. The soil humidity of each grid is collected every five minutes. The collected soil humidity data is used to represent the soil moisture content. The soil moisture content collected in the i-th grid area at time t is calibrated as θ i (t), i is the index of the grid.

[0015] Furthermore, the specific logic for obtaining the sample data set is:

[0016] An infrared thermometer, a SPAD meter, and a gas exchange system are placed on a plant in the central area of ​​each grid to obtain the leaf temperature, chlorophyll content, and transpiration rate data of the plant. The data are collected every five minutes and stored in the sample data set in sequence according to the timestamp and soil moisture content. The leaf temperature, chlorophyll content, and transpiration rate of the i-th grid area collected at time t are calibrated as T i (t), S i (t), E i (t).

[0017] Furthermore, the specific logic for obtaining the drought risk index is as follows:

[0018] Soil moisture content, leaf temperature, chlorophyll content and transpiration rate are input into the drought risk model, and the drought risk index of the corresponding timestamp is used as a label to train the drought risk model. The formula for obtaining the drought risk index at time t is:

[0019] R i (t) = α·θ i (t)+β·T i (t)+γ·S i (t)+δ·E i (t)

[0020] Among them, R i (t) is the drought risk index of the ith grid area at time t, and α, β, γ, and δ are the weight coefficients of soil moisture content, leaf temperature, chlorophyll content, and transpiration rate after the model training is completed.

[0021] Furthermore, the specific logic for determining whether the soil is at risk of drought is as follows:

[0022] The soil moisture content, leaf temperature, chlorophyll content and transpiration rate data of each grid area at the current moment are obtained and input into the trained drought risk model to obtain the drought risk index R of the soil in the i-th grid area at the current moment. i , if R i ≥T, determine that the soil in the current area is at risk of drought and mark the area. i <T, the soil in the current area is judged to be at a normal moisture level, and the number of grid areas marked as having drought risk is calibrated as M. If M>2, the drought risk index of the soil area is obtained for further judgment. If M<2, the drought risk index of each grid area continues to be monitored, and T is the drought risk index threshold.

[0023] Furthermore, the specific logic for judging the current soil state is as follows:

[0024] The formula for obtaining the mean value of the drought risk index for this soil area is:

[0025]

[0026] in, is the mean value of the drought risk index of the soil region at the current moment, and i = 1, 2, ..., N, where N is the number of divided grid areas;

[0027] Similarly, the mean soil moisture content at the current moment is obtained Mean leaf temperature Average chlorophyll content and the mean transpiration rate

[0028] The formula for obtaining the drought risk probability of the soil area at the current moment is:

[0029]

[0030] Where P is the drought risk probability of the soil area at the current moment, ω1, ω2, ω3, ω4, and ω5 are the weight coefficients of the corresponding parameters, Z is the normalization factor, ω1=ω2>ω3>ω4>ω5, the weight coefficients of ω2 and ω4 are negative, the weight coefficients of ω1, ω3, and ω5 are positive, and ω1+|ω2|+ω3+|ω4|+ω5=1;

[0031] The calculated drought risk index of the soil area is compared with the preset drought risk threshold. If the drought risk index of the soil area is greater than the preset drought risk threshold, it is judged that the soil is currently at drought risk and an early warning is issued. If the drought risk index of the soil area is lower than the preset drought risk threshold, it is judged that the soil area is currently in a normal state and the soil state continues to be monitored.

[0032] The present invention further provides a drought early warning system based on soil moisture detection, the drought early warning system based on soil moisture detection is used to implement the above-mentioned drought early warning method based on soil moisture detection, comprising:

[0033] The soil moisture collection module is used to divide the soil area into multiple 5|×5m grids, and arrange sensors in the soil in the central area of ​​each grid to obtain soil moisture content data of each grid area;

[0034] The plant status monitoring module is used to place sensors on a plant in the central area of ​​each grid to collect plant leaf temperature, chlorophyll content and transpiration rate data, and integrate soil moisture content, leaf temperature, chlorophyll content and transpiration rate into a sample data set based on the timestamp, and use the expert scoring method to assign a drought risk index corresponding to each timestamp data;

[0035] The drought risk model building module is used to build a drought risk model based on the random forest algorithm. The soil moisture content, leaf temperature, chlorophyll content and transpiration rate are input as training sets, and the corresponding drought risk index is used as a label to train the drought risk model.

[0036] The real-time risk monitoring module is used to collect the soil moisture content, leaf temperature, chlorophyll content and transpiration rate of each grid area in real time and input them into the trained drought risk model to obtain the corresponding drought risk index and determine the number of grid areas with drought risk;

[0037] The risk analysis and status judgment module is used to analyze the drought risk index, soil moisture content, leaf temperature, chlorophyll content and transpiration rate data of each grid area, obtain the drought risk probability of the soil area, and compare it with the drought risk threshold to judge the current soil status.

[0038] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0039] The drought early warning method and system based on soil moisture detection have significant technical effects and advantages. First, by arranging sensors in a 5m×5m grid area, the system can efficiently and in real time monitor multiple environmental factors such as soil moisture content and plant status to ensure the accuracy and reliability of the data. This high-density monitoring method can capture subtle changes in soil and plants in a timely manner and provide a more detailed drought risk assessment. Secondly, the drought risk model constructed using the random forest algorithm can effectively process complex nonlinear data and improve the accuracy of early warning. The drought risk index is assigned by the expert scoring method, combined with multi-dimensional data analysis, so that the model has strong adaptability and interpretability, and can provide a scientific basis for agricultural management. In addition, the system's real-time data collection and analysis capabilities enable it to respond quickly to environmental changes, issue drought warnings in a timely manner, provide decision support for farmland managers, and reduce losses caused by drought. By comprehensively analyzing the mean and risk probability of the drought risk index, the soil state can be more comprehensively evaluated, so as to implement targeted irrigation and water resource management. Finally, the multi-parameter comprehensive evaluation mechanism of this method adjusts the importance of each parameter through the weight coefficient, realizes flexible risk assessment and status monitoring, adapts to the needs of different environments and crops, and enhances the practicality and scope of application of the system. These advantages make this drought early warning method an efficient and accurate agricultural management tool, which helps to improve the water resource utilization efficiency and crop yield of farmland. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0041] Figure 2 Schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0042] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0043] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0044] Example:

[0045] See also Figure 1 The present invention provides a drought early warning method based on soil moisture detection, and the specific steps include:

[0046] Step 1: Divide the soil area into multiple 5m×5m grids, and arrange sensors in the soil in the central area of ​​each grid to obtain soil moisture content data in each grid area;

[0047] In this embodiment, the specific logic for collecting soil moisture content is:

[0048] After the soil area is evenly divided into 5m×5m grids, a humidity sensor is installed in the center of each grid. The soil humidity of each grid is collected every five minutes. The collected soil humidity data is used to represent the soil moisture content. The soil moisture content collected in the i-th grid area at time t is calibrated as θ i (t), i is the index of the grid.

[0049] This step can achieve high-precision, real-time monitoring of soil moisture content by evenly dividing the soil area into 5m×5m grids and arranging humidity sensors in the center of each grid. This detailed spatial division makes the soil moisture data of each grid representative, which helps to fully understand the soil moisture conditions in different regions, thereby providing a reliable data basis for drought warning. By regularly collecting soil moisture content data, moisture change trends can be identified in a timely manner, especially under drought conditions, and rapid response can help to combine with plant growth conditions for drought assessment. In addition, accurate moisture content data provides important input for subsequent risk assessment models, helps to improve the overall accuracy and effectiveness of the drought warning system, and has important guiding significance for agricultural management and decision-making.

[0050] Step 2: Place sensors on a plant in the central area of ​​each grid to collect data on plant leaf temperature, chlorophyll content, and transpiration rate. In addition, integrate soil moisture content, leaf temperature, chlorophyll content, and transpiration rate into a sample data set based on timestamps, and use an expert scoring method to assign a drought risk index corresponding to each timestamp data.

[0051] In this embodiment, the specific logic for obtaining the sample data set is:

[0052] An infrared thermometer, a SPAD meter, and a gas exchange system are placed on a plant in the central area of ​​each grid to obtain the leaf temperature, chlorophyll content, and transpiration rate data of the plant. The data are collected every five minutes and stored in the sample data set in sequence according to the timestamp and soil moisture content. The leaf temperature, chlorophyll content, and transpiration rate of the i-th grid area collected at time t are calibrated as T i (t), S i (t), E i (t).

[0053] This step arranges sensors on the central plants of each grid to obtain plant physiological parameters such as leaf temperature, chlorophyll content and transpiration rate, which can deeply analyze the growth status and water use efficiency of plants under specific soil moisture conditions. Integrating these data with soil moisture content into a sample data set can help establish a more comprehensive drought risk assessment model. By assigning a drought risk index to each timestamp data through an expert scoring method, real-time monitoring is not limited to soil moisture, but also includes plant physiological responses, thereby more accurately reflecting the impact of drought on plants. This multi-dimensional data integration enhances the system's predictive capabilities, allowing farmers and managers to take timely countermeasures, optimize irrigation management, and improve the stability and sustainability of agricultural production.

[0054] Step 3: Construct a drought risk model based on the random forest algorithm, input soil moisture content, leaf temperature, chlorophyll content and transpiration rate as training sets, and the corresponding drought risk index as labels to train the drought risk model;

[0055] In this embodiment, the specific logic for obtaining the drought risk index is:

[0056] Soil moisture content, leaf temperature, chlorophyll content and transpiration rate are input into the drought risk model, and the drought risk index of the corresponding timestamp is used as a label to train the drought risk model. The formula for obtaining the drought risk index at time t is:

[0057] R i (t) = α·θ i (t)+β·T i (t)+γ·Si (t)+δ·E i (t)

[0058] Among them, R i (t) is the drought risk index of the ith grid area at time t, and α, β, γ, and δ are the weight coefficients of soil moisture content, leaf temperature, chlorophyll content, and transpiration rate after the model training is completed.

[0059] This step aims to establish a comprehensive and accurate drought risk assessment tool by constructing a drought risk model based on the random forest algorithm, using soil moisture content, leaf temperature, chlorophyll content and transpiration rate as input features. The random forest algorithm has strong nonlinear modeling capabilities and high anti-overfitting capabilities, and can effectively handle the complex relationship between plant physiology and environmental factors. And by collecting data regularly, the model can be updated at any time. This dynamic characteristic enables the model to reflect the changes in the state of plants and soil in a timely manner and better adapt to changes in environmental conditions.

[0060] During the model training process, the input features are combined with the corresponding drought risk index, so that the model can automatically learn the impact of each factor on drought risk (represented by weight coefficients α, β, γ, δ). This automatic adjustment of weights makes the model adaptive and can update the drought risk assessment results in real time according to changes in different time and environmental conditions.

[0061] Using this method, the drought risk index R i (t) not only reflects the state of soil moisture, but also takes into account the physiological response of plants, providing a more comprehensive drought risk assessment. This is of great significance for agricultural managers to take effective irrigation measures in a timely manner, optimize resource allocation, and reduce the impact of drought on crop yields. Therefore, this step is a key link in the drought early warning method and provides a basis for the realization of scientific and intelligent agricultural management.

[0062] Step 4: The soil moisture content, leaf temperature, chlorophyll content and transpiration rate of each grid area are collected in real time and input into the trained drought risk model to obtain the corresponding drought risk index and determine the number of grid areas with drought risk;

[0063] In this embodiment, the specific logic for determining whether the soil is at risk of drought is as follows:

[0064] In this embodiment, the soil moisture content, leaf temperature, chlorophyll content and transpiration rate data of each grid area at the current moment are obtained and input into the trained drought risk model to obtain the drought risk index R of the soil in the i-th grid area at the current moment. i , if R i≥T, determine that the soil in the current area is at risk of drought and mark the area. i <T, the soil in the current area is judged to be at a normal moisture level, and the number of grid areas marked as having drought risk is calibrated as M. If M>2, the drought risk index of the soil area is obtained for further judgment. If M<2, the drought risk index of each grid area continues to be monitored, and T is the drought risk index threshold.

[0065] By obtaining the soil moisture content, leaf temperature, chlorophyll content and transpiration rate data of each grid at the current moment and inputting them into the drought risk model, real-time monitoring of the crop growth environment can be achieved. This timeliness enables farmland managers to respond quickly to potential drought risks and take appropriate measures. Set the drought risk index R i The comparison with the threshold T provides a clear standard for judging whether the soil has drought risk. Here, the drought risk index threshold is taken as the mean of the drought risk index of the soil with drought warning in the historical data. This quantitative method can effectively avoid the error caused by subjective judgment and improve the scientific nature of decision-making. By calibrating the number of regions with drought risk M, the drought risk of the entire soil area can be macro-evaluated. If M>2, it means that there are risks in multiple regions and further drought risk index analysis is required; if M<2, it can be focused on monitoring to reduce resource waste. This regional assessment helps to optimize resource allocation and improve management efficiency. With the results of this step, preventive measures can be taken in advance. If there is drought risk in multiple places, further evaluation will be carried out. If there is less possibility of local data abnormalities, the data of the soil will continue to be monitored.

[0066] Step 5: Analyze the drought risk index, soil moisture content, leaf temperature, chlorophyll content and transpiration rate data of each grid area to obtain the drought risk probability of the soil area, and compare it with the drought risk threshold to determine the current soil status;

[0067] In this embodiment, the specific logic for determining the current soil state is:

[0068] The formula for obtaining the mean value of the drought risk index for this soil area is:

[0069]

[0070] in, is the mean value of the drought risk index of the soil region at the current moment, and i = 1, 2, ..., N, where N is the number of divided grid areas;

[0071] Similarly, the mean soil moisture content at the current moment is obtained Mean leaf temperature Average chlorophyll content and the mean transpiration rate

[0072] The formula for obtaining the drought risk probability of the soil area at the current moment is:

[0073]

[0074] Where P is the drought risk probability of the soil area at the current moment, ω1, ω2, ω3, ω4, and ω5 are the weight coefficients of the corresponding parameters, Z is the normalization factor, ω1=ω2>ω3>ω4>ω5, the weight coefficients of ω2 and ω4 are negative, the weight coefficients of ω1, ω3, and ω5 are positive, and ω1+|ω2|+ω3+|ω4|+ω5=1;

[0075] The calculated drought risk probability of the soil area is compared with the preset drought risk threshold. If the drought risk probability of the soil area is greater than the preset drought risk threshold, it is judged that the soil is currently at drought risk and an early warning is issued. If the drought risk probability of the soil area is lower than the preset drought risk threshold, it is judged that the soil area is currently in a normal state and the soil state continues to be monitored.

[0076] This formula combines a variety of factors that affect drought risk and generates a risk probability P. The weight coefficient ω here is used to adjust the influence of different parameters on the final risk probability. Finally, the result is normalized to ensure that the probability value is reasonable. ) indicates that they are positively correlated with the probability of drought risk, that is, the higher the values ​​of these parameters, the higher the probability of drought risk. A higher drought risk index means a higher risk, and higher leaf temperature and transpiration rate may indicate an increase in plant water consumption, thereby increasing drought risk. Negative weights of ω2 and ω4 (such as ) show that they are negatively correlated with the probability of drought risk. That is, the higher the value of these parameters, the lower the probability of drought risk. Higher soil moisture content indicates that the soil is moist, thereby reducing the risk of drought, and higher chlorophyll content indicates that the plant is in good health, which also reduces the risk of drought. The normalization factor Z is to ensure that the final output drought risk probability P is within a reasonable range (usually 0 to 1). This makes comparison and interpretation more intuitive. Z can be dynamically valued, taking the sum of the maximum values ​​of each parameter in the data collected in all grid areas, for example Z = R max +θ max +T max +S max +E max , where R max is the maximum value of the drought risk index, θ max is the maximum soil moisture content, Tmax is the maximum leaf temperature, S max is the maximum chlorophyll content and E max The maximum transpiration rate can better adapt to environmental changes. If environmental conditions change significantly, regular updates of Z will help maintain the accuracy of the prediction. The drought risk threshold here is the mean of the drought risk probability generated by the soil data with drought conditions in the historical data. When the drought risk probability exceeds the threshold, the system can issue an early warning, allowing managers to quickly take necessary measures to reduce the impact of drought on crops. Setting thresholds and making comparisons can make management decisions more dependent on data rather than intuition, thereby improving the accuracy and reliability of decisions.

[0077] See also Figure 2 The present invention further provides a drought early warning system based on soil moisture detection, the drought early warning system based on soil moisture detection is used to implement the above-mentioned drought early warning method based on soil moisture detection, including:

[0078] The soil moisture collection module is used to divide the soil area into multiple 5m×5m grids, and arrange sensors in the soil in the central area of ​​each grid to obtain soil moisture content data of each grid area;

[0079] The plant status monitoring module is used to place sensors on a plant in the central area of ​​each grid to collect plant leaf temperature, chlorophyll content and transpiration rate data, and integrate soil moisture content, leaf temperature, chlorophyll content and transpiration rate into a sample data set based on the timestamp, and use the expert scoring method to assign a drought risk index corresponding to each timestamp data;

[0080] The drought risk model building module is used to build a drought risk model based on the random forest algorithm. The soil moisture content, leaf temperature, chlorophyll content and transpiration rate are input as training sets, and the corresponding drought risk index is used as a label to train the drought risk model.

[0081] The real-time risk monitoring module is used to collect the soil moisture content, leaf temperature, chlorophyll content and transpiration rate of each grid area in real time and input them into the trained drought risk model to obtain the corresponding drought risk index and determine the number of grid areas with drought risk;

[0082] The risk analysis and status judgment module is used to analyze the drought risk index, soil moisture content, leaf temperature, chlorophyll content and transpiration rate data of each grid area, obtain the drought risk probability of the soil area, and compare it with the drought risk threshold to judge the current soil status.

[0083] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0084] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0085] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0086] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A drought early warning method based on soil moisture detection, characterized in that: The specific steps include: Step 1: Divide the soil area into multiple 5m×5m grids, and arrange sensors in the soil in the central area of ​​each grid to obtain soil moisture content data in each grid area; Step 2: Place sensors on a plant in the central area of ​​each grid to collect data on plant leaf temperature, chlorophyll content, and transpiration rate. In addition, integrate soil moisture content, leaf temperature, chlorophyll content, and transpiration rate into a sample data set based on timestamps, and use an expert scoring method to assign a drought risk index corresponding to each timestamp data. Step 3: Construct a drought risk model based on the random forest algorithm, input the soil moisture content, leaf temperature, chlorophyll content and transpiration rate of each grid area as the training set, and the corresponding drought risk index as the label to train the drought risk model; Step 4: The soil moisture content, leaf temperature, chlorophyll content and transpiration rate of each grid area are collected in real time and input into the trained drought risk model to obtain the corresponding drought risk index and determine the number of grid areas with drought risk; Step 5: Analyze the drought risk index, soil moisture content, leaf temperature, chlorophyll content and transpiration rate data of each grid area to obtain the drought risk probability of the soil area, and compare it with the drought risk threshold to determine the current soil status; After the soil area is evenly divided into 5m×5m grids, a humidity sensor is installed in the center of each grid. The soil humidity of each grid is collected every five minutes. The collected soil humidity data is used to represent the soil moisture content. The soil moisture content collected in the i-th grid area at time t is calibrated as θ i (t), i is the index of the grid; An infrared thermometer, a SPAD meter, and a gas exchange system are placed on a plant in the central area of ​​each grid to obtain the leaf temperature, chlorophyll content, and transpiration rate data of the plant. The data are collected every five minutes and stored in the sample data set in sequence according to the timestamp and soil moisture content. The leaf temperature, chlorophyll content, and transpiration rate of the i-th grid area collected at time t are calibrated as T i (θ), S i (t), E i (t); Soil moisture content, leaf temperature, chlorophyll content and transpiration rate are input into the drought risk model, and the drought risk index of the corresponding timestamp is used as a label to train the drought risk model. The formula for obtaining the drought risk index at time t is: R i (t)=α·θ i (t)+β·T i (t)+γ·S i (t)+δ·E i (t) Among them, R i (t) is the drought risk index of the ith grid area at time t, α, β, γ, and δ are the weight coefficients of soil moisture content, leaf temperature, chlorophyll content, and transpiration rate after the model training is completed; Obtain the soil moisture content, leaf temperature, chlorophyll content, and transpiration rate data of each grid area at the current moment, and input them into the trained drought risk model to obtain the drought risk index R of the soil in the i-th grid area at the current moment i , if R i ≥T, it is determined that the soil in the current area has a drought risk, and the area is marked. If R i <T, it is determined that the soil in the current area is at a normal moisture level. The number of grid areas marked as having a drought risk is designated as M. If M>2, obtain the drought risk index of this soil area for further judgment. If M<2, continue to monitor the drought risk index of each grid area. T is the drought risk index threshold; The formula for obtaining the mean value of the drought risk index for this soil area is: in, is the mean value of the drought risk index of the soil region at the current moment, and i = 1, 2, ..., N, where N is the number of divided grid areas; Similarly, the mean soil moisture content at the current moment is obtained Mean leaf temperature Mean chlorophyll content and mean transpiration rate The formula for obtaining the drought risk probability of the soil area at the current moment is: Where P is the drought risk probability of the soil area at the current moment, ω1, ω2, ω3, ω4, and ω5 are the weight coefficients of the corresponding parameters, Z is the normalization factor, ω1=ω2>ω3>ω4>ω5, the weight coefficients of ω2 and ω4 are negative, the weight coefficients of ω1, ω3, and ω5 are positive, and ω1+|ω2|+ω3+|ω4|+ω5=1; The calculated drought risk probability of the soil area is compared with the preset drought risk threshold. If the drought risk probability of the soil area is greater than the preset drought risk threshold, it is judged that the soil is currently at drought risk and an early warning is issued. If the drought risk probability of the soil area is lower than the preset drought risk threshold, it is judged that the soil area is currently in a normal state and the soil state continues to be monitored.

2. A drought early warning system based on soil moisture detection, characterized in that: The drought early warning system based on soil moisture detection is used to implement the drought early warning method based on soil moisture detection according to claim 1, comprising: The soil moisture collection module is used to divide the soil area into multiple 5m×5m grids, and arrange sensors in the soil in the central area of ​​each grid to obtain soil moisture content data of each grid area; The plant status monitoring module is used to place sensors on a plant in the central area of ​​each grid to collect plant leaf temperature, chlorophyll content and transpiration rate data, and integrate soil moisture content, leaf temperature, chlorophyll content and transpiration rate into a sample data set based on the timestamp, and use the expert scoring method to assign a drought risk index corresponding to each timestamp data; The drought risk model building module is used to build a drought risk model based on the random forest algorithm. The soil moisture content, leaf temperature, chlorophyll content and transpiration rate are input as training sets, and the corresponding drought risk index is used as a label to train the drought risk model. The real-time risk monitoring module is used to collect the soil moisture content, leaf temperature, chlorophyll content and transpiration rate of each grid area in real time and input them into the trained drought risk model to obtain the corresponding drought risk index and determine the number of grid areas with drought risk; The risk analysis and status judgment module is used to analyze the drought risk index, soil moisture content, leaf temperature, chlorophyll content and transpiration rate data of each grid area, obtain the drought risk probability of the soil area, and compare it with the drought risk threshold to judge the current soil status.

Images