A method for monitoring agricultural drought using surface soil moisture

Abstract

The present invention discloses a method for monitoring agricultural drought using surface soil moisture, comprising the following steps: selecting a soil moisture observation station in a study area as a location for agricultural drought sampling, and collecting soil moisture data at various depths at the sample point locations; calculating the agricultural drought grade on the sample based on a soil relative humidity classification standard for the deep soil moisture data; mining a quantitative relationship between the agricultural drought grade on the sample and surface microwave satellite remote sensing soil moisture data based on a random forest algorithm, calculating the agricultural drought grade of the region through the quantitative relationship, and using a verification index to describe the accuracy of the classification of the agricultural drought grade of the region; the method of the present invention takes into account soil moisture factors from multiple data sources that affect agricultural drought, overcomes the singleness and limitations of previous methods, and realizes the use of multi-source soil moisture data to monitor the spatial distribution of agricultural drought in the region.

Description

Technical Field

[0001] The present invention relates to the fields of water conservancy science, agricultural science and geographic information science and technology, and in particular to a method for monitoring agricultural drought by utilizing surface soil moisture. Background Art

[0002] In my country, losses caused by agricultural drought disasters account for more than 50% of China's agricultural natural disaster losses, and have gradually intensified in most parts of China in the past 20 years, seriously affecting the safe production of grain in my country. Timely and accurate monitoring of regional agricultural drought can provide an accurate scientific theoretical basis for formulating drought resistance and disaster reduction measures. Therefore, studying an effective method to monitor agricultural drought has always been a hot topic in disciplines such as water conservancy science, agricultural science, and geographic information science. However, due to the complex mechanism of drought occurrence, achieving accurate monitoring of drought is currently a scientific problem recognized by the world.

[0003] In recent years, using multi-source soil moisture data as a data source, scholars have explored a method of using machine learning algorithms to monitor agricultural drought, which has practical significance for the study of drought monitoring technology and assessment methods; the degree of water stress suffered by crops is closely related to the soil moisture content. Soil moisture is usually regarded as an important agricultural drought monitoring indicator. Currently, there are two main ways to obtain soil moisture: one is based on site observation, and the other is based on microwave remote sensing observation.

[0004] Soil moisture data based on site observations has high accuracy and can effectively monitor agricultural drought conditions. However, site-based soil moisture monitoring in my country was generally late in development. Many regions lack site-based soil moisture data, and publicly available site-based soil moisture data is also scarce. This presents challenges for many studies, resulting in limited research using site-based soil moisture data for agricultural drought monitoring. Furthermore, regardless of the accuracy and frequency of site-based monitoring, the observation sites are spatially discrete, thus failing to capture continuous surface soil moisture information. Furthermore, the short time series of soil moisture from site-based observations are associated with high costs. These factors restrict the application and promotion of site-based soil moisture methods for agricultural drought monitoring. The development of remote sensing technology has provided an open and efficient channel for soil moisture research. Remotely sensed soil moisture data offers advantages such as all-day, all-weather availability and sensitivity to soil moisture changes, making it highly applicable. However, microwave remote sensing can only retrieve soil moisture in the top 0-5 cm layer, which cannot effectively reflect drought conditions in farmland. Exploring the potential relationship between surface soil moisture and agricultural drought has become a key focus of this study. Summary of the Invention

[0005] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of this application to avoid obscuring the purpose of this section, the abstract and the title of the invention, and such simplifications or omissions should not be used to limit the scope of the present invention.

[0006] In view of the above-mentioned problems, the present invention is proposed.

[0007] A first aspect of an embodiment of the present invention provides a method for monitoring agricultural drought using surface soil moisture, comprising: selecting a soil moisture observation station in a study area as a location for agricultural drought sampling, and collecting soil moisture data at various depths at the sample point locations; calculating the agricultural drought level on the sample based on a soil relative humidity classification standard for the depth soil moisture data; extracting surface microwave satellite remote sensing soil moisture data at the sample point using a microwave satellite remote sensing soil moisture product; mining a quantitative relationship between the agricultural drought level on the sample and the surface microwave satellite remote sensing soil moisture data based on a random forest algorithm, calculating the agricultural drought level of the region using the quantitative relationship, and using a verification indicator to describe the accuracy of the classification of the agricultural drought level in the region.

[0008] As a preferred embodiment of the method for monitoring agricultural drought using surface soil moisture according to the present invention, the calculation of the relative soil humidity includes:

[0009]

[0010] Where RSM represents relative soil moisture, a represents the adjustment coefficient of crop development period, n represents the number of observation layers within the soil layer thickness corresponding to the crop development stage, and w i represents the soil moisture of the i-th layer, f ci represents the field water holding capacity of the i-th soil layer.

[0011] As a preferred embodiment of the method for monitoring agricultural drought using surface soil moisture according to the present invention, the classification criteria of the relative soil moisture include:

[0012] According to the previous work and relevant literature review, the relative soil humidity at the 20cm site is selected as the best reference indicator for the research, that is, the relative soil humidity at the 20cm site is used to classify the agricultural drought level;

[0013] If the relative soil humidity at the 20cm site is greater than 60%, the agricultural drought level is no drought; if the relative soil humidity at the 20cm site is greater than 50% and less than or equal to 60%, the agricultural drought level is light drought; if the relative soil humidity at the 20cm site is greater than 40% and less than or equal to 50%, the agricultural drought level is moderate drought; if the relative soil humidity at the 20cm site is greater than 30% and less than or equal to 40%, the agricultural drought level is severe drought; if the relative soil humidity at the 20cm site is less than or equal to 30%, the agricultural drought level is extreme drought.

[0014] As a preferred embodiment of the method for monitoring agricultural drought using surface soil moisture according to the present invention, the classification of agricultural drought levels on the sample includes:

[0015] The relationship between the relative soil moisture and agricultural drought is expressed by using fuzzy logic, and a membership function of the relative soil moisture to agricultural drought is determined in a form similar to a Gaussian function.

[0016] The calculation of the membership function includes:

[0017]

[0018] Among them, f ijk,v represents the degree of drought of pixel (i, j) at the kth moment, x0 represents the typical value of soil relative humidity when a severe drought occurs, and x ijk,v represents the value of soil relative humidity at the pixel (i, j) k, w represents the difference between the median value of soil relative humidity and x0 when moderate drought occurs, r represents the slope parameter of the control curve, and c represents the membership value at a known point;

[0019] According to the classification principle of agricultural drought levels, a Gaussian membership function of soil relative humidity to agricultural drought is established, which is expressed as:

[0020] f ijk,RSM =1,RSM≤0.3

[0021]

[0022] f ijk,RSM =0,RSM>0.6

[0023] Among them, f ijk,RSM It represents the degree of membership of soil relative moisture to agricultural drought;

[0024] Based on the Gaussian membership function, the minimum operator of the fuzzy set is used to calculate the agricultural drought grade on the final sample. The formula is as follows:

[0025] M i,j,k (RSM)=min(fijk,RSM )

[0026] Among them, M i,j,k represents the drought level on the sample;

[0027] The agricultural drought level on the sample ranges from [0, 1]. When the correlation level is 1, it is considered to be a severe drought, and when the correlation level is 0, it is considered not to be a drought.

[0028] As a preferred embodiment of the method for monitoring agricultural drought using surface soil moisture according to the present invention, the extraction of surface microwave satellite remote sensing soil moisture data at the sample points includes:

[0029] The soil volumetric water content of the SMAP microwave satellite soil moisture product is converted into the relative soil water content. The relative soil water content is the surface microwave satellite remote sensing soil moisture data at the sample point. The formula is as follows:

[0030]

[0031] Among them, Relative SMijk V represents the relative soil water content on the kth day in the pixel of row i and column j. SMk Indicates the soil volumetric water content of the pixel on the kth day.

[0032] As a preferred embodiment of the method for monitoring agricultural drought using surface soil moisture according to the present invention, the agricultural drought level of a region calculated by the quantitative relationship includes:

[0033] Based on the random forest algorithm, the surface microwave satellite remote sensing soil moisture data was used as the independent variable, and the agricultural drought grade based on the soil relative humidity classification sample was used as the dependent variable. The Random Forest data package in Python was used to construct a random forest drought monitoring model.

[0034] During the model training process, 70% to 80% of the sample data are selected to build the random forest drought monitoring model, and 20% to 30% of the samples are used to verify the random forest drought monitoring model. The modeling and verification are repeated multiple times to finally obtain the optimal forest drought monitoring model;

[0035] The optimal forest drought monitoring model is used to explore the quantitative relationship between satellite remote sensing surface soil moisture and deep soil moisture that can best reflect agricultural drought, thereby achieving real-time monitoring of regional farmland drought.

[0036] As a preferred embodiment of the method for monitoring agricultural drought using surface soil moisture according to the present invention, the accuracy of the classification of agricultural drought levels in the region using verification indicators includes:

[0037] By calculating the coefficient of determination R of the observed and simulated values ​​of the test set 2 The accuracy of the optimal forest drought monitoring model was verified by three indicators: root mean square error (RMSE) and mean absolute error (MAE);

[0038] The calculation formulas of the three indicators are as follows:

[0039]

[0040]

[0041]

[0042] Among them, x i represents the simulated value of drought level, The average value of the simulated values ​​representing the drought level, y i represents the observed value of drought level, It represents the average value of the observed values ​​of drought level, m represents the number of samples, and i represents the serial number of the observation month.

[0043] A second aspect of an embodiment of the present invention provides a system for monitoring agricultural drought using surface soil moisture, comprising:

[0044] The data collection unit is used to select soil moisture observation sites in the study area as locations for agricultural drought sampling and to collect soil moisture data at various depths at the sampling sites;

[0045] a sample grade classification unit, configured to calculate the agricultural drought grade of the sample according to a classification standard of soil relative humidity of the deep soil moisture data;

[0046] The regional drought level determination unit is used to use microwave satellite remote sensing soil moisture products to extract surface microwave satellite remote sensing soil moisture data on sample points, and based on the random forest algorithm, to mine the quantitative relationship between the agricultural drought level on the sample and the surface microwave satellite remote sensing soil moisture data, calculate the agricultural drought level of the region through the quantitative relationship, and use verification indicators to describe the accuracy of the classification of the agricultural drought level of the region.

[0047] According to a third aspect of an embodiment of the present invention, a device is provided, comprising:

[0048] processor;

[0049] a memory for storing processor-executable instructions;

[0050] The processor is configured to call the instructions stored in the memory to execute the method described in any embodiment of the present invention.

[0051] According to a fourth aspect of the embodiments of the present invention, a computer-readable storage medium is provided, on which computer program instructions are stored, including:

[0052] When the computer program instructions are executed by a processor, the method according to any embodiment of the present invention is implemented.

[0053] The beneficial effects of the present invention are as follows: the present invention provides a method for monitoring agricultural drought using surface soil moisture. Compared with existing methods, the present invention gives full play to the advantages of remote sensing technology and uses multi-source soil moisture data as a data source. Since remote sensing data has the characteristics of macro, continuous, and large-scale, and the data that can be obtained is surface information, it provides a scientific method for drought monitoring in large areas. Moreover, the random forest algorithm has the advantages of fast learning process, fast calculation speed, good stability, and high prediction accuracy. The study introduces this method, constructs a random forest drought monitoring model, and combines multiple soil moisture data to provide a new method for regional comprehensive drought monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them:

[0055] Figure 1 The overall technical roadmap of the method for monitoring agricultural drought using surface soil moisture provided by the present invention;

[0056] Figure 2 An overview diagram of a method for monitoring agricultural drought using surface soil moisture provided by the present invention;

[0057] Figure 3 A drought distribution map of a method for monitoring agricultural drought using surface soil moisture provided by the present invention. DETAILED DESCRIPTION

[0058] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0059] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0060] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0061] The present invention is described in detail with reference to schematic diagrams. For ease of illustration, cross-sectional views of device structures may be partially enlarged and not to scale when describing embodiments of the present invention. Furthermore, the schematic diagrams are merely illustrative and should not limit the scope of the present invention. Furthermore, in actual production, the three-dimensional dimensions of length, width, and depth should be included.

[0062] In the description of the present invention, it should be noted that the terms "upper, lower, inner, and outer" and other references to orientations or positional relationships are based on the orientations or positional relationships shown in the accompanying drawings and are intended solely to facilitate and simplify the description of the present invention. They are not intended to indicate or imply that the devices or components referred to must have, be constructed, or operate in a specific orientation, and therefore should not be construed as limitations on the present invention. Furthermore, the terms "first, second, or third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0063] In this disclosure, unless otherwise specified or limited, the terms "mounted," "connected," and "connected" should be interpreted broadly. For example, they may refer to fixed, removable, or integral connections. They may also refer to mechanical, electrical, or direct connections, indirect connections through an intermediary, or internal communication between two components. Those skilled in the art will understand the specific meanings of these terms in this disclosure.

[0064] Example 1

[0065] Reference Figure 1 According to one embodiment of the present invention, a method for monitoring agricultural drought using surface soil moisture is provided, comprising:

[0066] S1: Select soil moisture observation sites in the study area as locations for agricultural drought sampling, and collect soil moisture data at various depths at the sample point locations.

[0067] S2: Calculate the agricultural drought level on the sample based on the soil relative humidity classification standard of the deep soil moisture data. It should be noted that:

[0068] The calculation of soil relative moisture includes,

[0069]

[0070] Where RSM represents relative soil moisture, a represents the adjustment coefficient of the crop development period, which is 1.1 in the seedling stage, 0.9 in the critical moisture period, and 1 in the rest of the development period. n represents the number of observation layers (generally divided into 10 cm units) within the soil layer thickness corresponding to the crop development stage (n = 1 in the crop sowing period, n = 2 in the seedling stage, and n = 5 in other growth stages). w i represents the soil moisture of the i-th layer, f ci represents the field water holding capacity of the i-th soil layer;

[0071] It should be noted that the degree of water stress suffered by crops is closely related to soil moisture content. Obtaining effective soil moisture data is of great significance for agricultural drought monitoring. In the monitoring of drought processes, the drought grade has a stronger mechanism for describing the occurrence and development of drought and has been proven to have better applicability. Therefore, it is feasible to introduce the indicator of soil relative humidity to classify drought grades. Soil relative humidity directly reflects the status of available water for dryland crops. It is closely related to environmental meteorological conditions, crop growth and development, and soil physical properties. For different crop varieties, different development stages of the same crop, and different soil textures, there are also certain differences between the indicators of crop available water.

[0072] Furthermore, the classification standards of soil relative moisture include:

[0073] Based on previous work and relevant literature, the relative soil humidity at the 20cm site was selected as the optimal reference indicator for the study. That is, the relative soil humidity at the 20cm site was used to classify agricultural drought levels. The changes in water consumption of each soil layer under different moisture conditions are shown in Table 1, and the classification of soil relative humidity drought levels is shown in Table 2.

[0074] Table 1: Changes in water consumption of each soil layer under different moisture conditions.

[0075] Depth (cm) Percentage of water consumption in soil layers at different depths (%) 20 41.0 40 25.3 60 17.2 80 6.3 100 4.4 120 4.0

[0076] Table 2: Drought classification based on soil relative moisture.

[0077] grade Drought type Soil relative humidity at 20cm depth 1 No drought 60%<RSM 2 Mild drought 50%<RSM<=60% 3 moderate drought 40%<RSM<=50% 4 severe drought 30%<RSM<=40% 5 Extreme drought RSM<=30%

[0078] It should be noted that the water consumption of the 0-20cm soil layer accounts for 40%-70% of the total water consumption of the entire soil layer, and is also the main water-consuming soil layer. Field experiments on winter wheat root systems in the study area showed that the root length density of winter wheat is mainly concentrated in the 0-50cm range, accounting for 57.7% of the total root length. 20cm is in the middle of 0-50cm. Studying the soil moisture at 20cm can intuitively reflect the amount of water absorbed by crops in the soil. Based on previous work and relevant literature, the soil moisture at the 20cm site was selected as the research reference indicator and reflects the optimal agricultural drought.

[0079] Furthermore, the classification of agricultural drought levels in the sample includes:

[0080] Fuzzy logic is used to express the relationship between soil relative humidity and agricultural drought. The membership function of soil relative humidity to agricultural drought is determined in a form similar to Gaussian function. The calculation of membership function includes:

[0081]

[0082] Among them, f ijk,v represents the degree of drought of pixel (i, j) at the kth moment, x0 represents the typical value of soil relative humidity when a severe drought occurs, and x ijk,v represents the value of soil relative humidity at the pixel (i, j) k, w represents the difference between the median value of soil relative humidity during moderate drought and x0, r represents the slope parameter of the control curve, the larger the value, the steeper it is, and the value is usually 2, c represents the membership value at a known point, and is usually 0.5;

[0083] According to the classification principle of agricultural drought levels, a Gaussian membership function of soil relative humidity to agricultural drought is established, which is expressed as:

[0084] f ijk,RSM =1,RSM≤0.3

[0085]

[0086] f ijk,RSM =0,RSM>0.6

[0087] Among them, f ijk,RSM It represents the degree of membership of soil relative moisture to agricultural drought;

[0088] Based on the Gaussian membership function, the minimum operator of the fuzzy set is used to calculate the agricultural drought grade on the final sample. The formula is as follows:

[0089] M i,j,k (RSM)=min(f ijk,RSM )

[0090] Among them, Mi,j,k represents the drought level on the sample;

[0091] Specifically, the agricultural drought level on the sample ranges from [0,1]. When the correlation degree is 1, it is considered a severe drought, and when the correlation degree is 0, it is considered not a drought.

[0092] It should be noted that the calculation of drought level by using the minimum operator operation in fuzzy set operation is mainly based on the following considerations: agricultural drought is affected by relative soil moisture, and the calculation of drought level requires consideration of the degree of membership of relative soil moisture to agricultural drought; therefore, the present invention adopts the principle of limiting factor physics to express the degree of membership of relative soil moisture to agricultural drought.

[0093] S3: Use microwave satellite remote sensing soil moisture products to extract surface microwave satellite remote sensing soil moisture data at the sample point. It should be noted that:

[0094] The extraction of surface microwave satellite remote sensing soil moisture data at the sample point includes:

[0095] The soil volumetric water content of the SMAP microwave satellite soil moisture product is converted into the relative soil water content. The relative soil water content is the surface microwave satellite remote sensing soil moisture data at the sample point. The formula is as follows:

[0096]

[0097] Among them, Relative SMijk V represents the relative soil water content on the kth day in the pixel of row i and column j. SMk Indicates the soil volumetric water content of the pixel on the kth day.

[0098] S4: Based on the random forest algorithm, the quantitative relationship between the agricultural drought level of the sample and the surface microwave satellite remote sensing soil moisture data is mined, the regional agricultural drought level is calculated based on the quantitative relationship, and the accuracy of the regional agricultural drought level classification is described using verification indicators. It should be noted that:

[0099] The agricultural drought level of a region is calculated through quantitative relationships including:

[0100] Based on the random forest algorithm, the surface microwave satellite remote sensing soil moisture data was used as the independent variable, and the agricultural drought grade based on the soil relative humidity classification sample was used as the dependent variable. The Random Forest data package in Python was used to construct a random forest drought monitoring model.

[0101] During the model training process, 70% to 80% of the sample data was selected to build the random forest drought monitoring model, and 20% to 30% of the samples were used to verify the random forest drought monitoring model. The modeling and verification were repeated many times to finally obtain the optimal forest drought monitoring model.

[0102] Using the optimal forest drought monitoring model, we will explore the quantitative relationship between surface soil moisture from satellite remote sensing and deep soil moisture that can best reflect agricultural drought, thus enabling real-time monitoring of regional farmland drought.

[0103] Furthermore, the confusion matrix is ​​introduced to solve the impact of unbalanced data sets on the training of random forest drought monitoring model. The calculation formula of confusion matrix related indicators includes:

[0104]

[0105]

[0106]

[0107]

[0108] Among them, accuracy represents accuracy, precision represents precision, recall represents recall rate, and f1-score represents harmonic mean;

[0109] Specifically, the total number of samples is P+N, and positive and negative are used to represent the two results. The confusion matrix is ​​shown in Table 3;

[0110] Table 3: Confusion matrix.

[0111]

[0112]

[0113] Furthermore, the accuracy of the classification of agricultural drought levels in the region using validation indicators includes:

[0114] By calculating the coefficient of determination R of the observed and simulated values ​​of the test set 2 The accuracy of the optimal forest drought monitoring model was verified by three indicators: root mean square error (RMSE) and mean absolute error (MAE). 2 It is generally used to evaluate the degree of conformity between simulated values ​​and observed values. RMSE is used to measure the deviation between model simulated values ​​and actual values. MAE can reflect the actual situation of simulated value errors. 2 The closer it is to 1, and the closer RMSE and MAE are to 0, the better the model performance;

[0115] Specifically, the calculation formulas for the three indicators are as follows:

[0116]

[0117]

[0118]

[0119] Among them, x i represents the simulated value of drought level, The average value of the simulated values ​​representing the drought level, y i represents the observed value of drought level, It represents the average value of the observed values ​​of drought level, m represents the number of samples, and i represents the serial number of the observation month.

[0120] It should be noted that the present invention provides a method for monitoring agricultural drought using surface soil moisture. Compared with existing methods, the present invention gives full play to the advantages of remote sensing technology and uses multi-source soil moisture data as a data source. Since remote sensing data has the characteristics of macro, continuous, and large-scale, and the available data is surface information, it provides a scientific method for drought monitoring in large areas; and the random forest algorithm has the advantages of fast learning process, fast calculation speed, good stability, and high prediction accuracy. The study introduced this method, constructed a random forest drought monitoring model, and combined with multiple soil moisture data, provided a new method for regional comprehensive drought monitoring methods; in addition, the method of the present invention takes into account soil moisture factors from multiple data sources that affect agricultural drought, overcomes the singleness and limitations of previous methods, and realizes the use of multi-source soil moisture data to monitor the spatial distribution of agricultural drought in the region.

[0121] The second aspect of the present invention is disclosed.

[0122] A system for monitoring agricultural drought using surface soil moisture is provided, comprising:

[0123] The data collection unit is used to select soil moisture observation sites in the study area as locations for agricultural drought sampling and to collect soil moisture data at various depths at the sampling sites;

[0124] A sample classification unit is used to calculate the agricultural drought grade of the sample based on the classification standard of the soil relative moisture of the deep soil moisture data;

[0125] The regional drought level determination unit is used to extract surface microwave satellite remote sensing soil moisture data on the sample points using microwave satellite remote sensing soil moisture products. Based on the random forest algorithm, it mines the quantitative relationship between the agricultural drought level on the sample and the surface microwave satellite remote sensing soil moisture data, calculates the regional agricultural drought level through the quantitative relationship, and uses verification indicators to describe the accuracy of the regional agricultural drought level division.

[0126] The third aspect of the present invention is disclosed.

[0127] Provided is a device comprising:

[0128] processor;

[0129] a memory for storing processor-executable instructions;

[0130] The processor is configured to call instructions stored in the memory to execute any one of the aforementioned methods.

[0131] The fourth aspect of the present invention is disclosed.

[0132] A computer-readable storage medium is provided, on which computer program instructions are stored, including:

[0133] When the computer program instructions are executed by a processor, any of the above methods is implemented.

[0134] The present invention may be a method, an apparatus, a system and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present invention.

[0135] A computer-readable storage medium can be a tangible device that can hold and store instructions for use by an instruction execution device. A computer-readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination thereof. As used herein, a computer-readable storage medium is not to be construed as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through an electrical wire.

[0136] Example 2

[0137] Reference Figures 2-3 This is the second embodiment of the present invention. Different from the first embodiment, this embodiment provides a verification test of a method for monitoring agricultural drought using surface soil moisture, in order to verify and illustrate the technical effects adopted in this method.

[0138] This example selects Hebi City, Henan Province as the research area. Henan Province is located in the North China Plain. The mountains and hills in western Henan are mainly composed of zonal soils such as brown soil and cinnamon soil. The northern and eastern plains are mainly composed of fluvo-aquic soil formed by river alluvial deposits. Sand ginger black soil is widely distributed in the southeast. The southern part is mainly composed of paddy soil, yellow-brown soil, and yellow-cinnamon soil. The main crop studied in this example is winter wheat. The specific implementation method is as follows:

[0139] ① Using ArcGIS software, we can draw a general map of the study area using the DEM elevation and soil moisture monitoring stations of Hebi City, such as Figure 2 As shown in the figure, we can see the distribution of 16 stations in Hebi City;

[0140] ② Use Visio software to draw a specific technical roadmap;

[0141] ③ Use soil relative humidity to classify agricultural drought levels, which is divided into 5 levels;

[0142] ④ Preliminary analysis and literature review revealed that soil moisture at sites at different depths showed that surface soil moisture changes significantly, while deep soil moisture changes more slowly, with relatively higher content, and that deep soil moisture increases with increasing depth. Rainfall can cause changes in soil moisture, and soil moisture at different depths indicates different levels of agricultural drought. The 0-20cm soil layer accounts for 40%-70% of the total water consumption of the entire soil layer, making it the primary water-consuming layer. Ultimately, the 20cm soil moisture level was found to be the optimal indicator of agricultural drought.

[0143] ⑤ Based on the quantitative relationship between relative soil moisture and agricultural drought, a Gaussian membership function was constructed, and the drought level of the samples was quantified based on fuzzy set operations;

[0144] ⑥ The independent variable is the surface 5cm soil moisture data inverted by microwave satellite soil moisture products, and the dependent variable is the drought level of the sample. The random forest model is built by calling the random forest package in Python and scikit-learn. The constructed random forest model is used to explore the quantitative relationship between the agricultural drought level of the sample and the surface microwave satellite soil moisture, and the regional agricultural drought level is obtained. It is displayed in the Acrgis Hebi City grid. The drought level distribution map of Hebi City is shown below. Figure 3 As shown in the figure, it can be seen that Hebi City has experienced droughts of varying degrees;

[0145] ⑦Through confusion matrix and R 2, RMSE, and MAE to verify the accuracy and applicability of the model. At the same time, the drought monitoring model calculation results were verified by the drought level divided by the soil moisture of the site to evaluate whether the results were consistent with the facts. The relevant values ​​are shown in Table 4, and the monthly accuracy of the model is shown in Table 5.

[0146] Table 4: Classification model evaluation metrics.

[0147] precision recall f1-score support 0 1 1 1 45 1 0.85 0.99 0.91 170 2 0 0 0 16 3 0 0 0 5 4 0 0 0 8 5 0 0 0 1 accuracy 0.87 245 macro avg 0.31 0.33 0.32 245 weighted avg 0.77 0.87 0.82 245

[0148] As shown in Table 4, the final test set prediction accuracy is 87%, which shows that the random forest classification accuracy is high and has good superiority;

[0149] Table 5: Model accuracy by month.

[0150] month <![CDATA[R 2 ]]> RMSE MAE 10 0.532 0.563 0.454 11 0.576 0.521 0.489 12 0.578 0.534 0.490 1 0.543 0.581 0.420 2 0.678 0.592 0.468 3 0.684 0.587 0.453 4 0.581 0.504 0.367 5 0.554 0.591 0.581 6 0.586 0.598 0.563

[0151] As can be seen from Table 5, the winter wheat growing season is from October to June of the following year, and the determination coefficient of the model (R 2 ) is between 0.52 and 0.68, maintaining a good correlation, which shows that the drought monitoring model has a high accuracy. Moreover, the change from October to June of the following year is because the snow accumulation on the land in Hebi City when the temperature is low in winter affects the accuracy of multi-source soil moisture data, so the accuracy in winter is slightly lower than that in spring and autumn. The change of the model's monitoring value can also reflect the change of regional soil relative humidity, and soil relative humidity is an important factor affecting agricultural drought. Therefore, for regional agricultural drought, the model also has a certain monitoring capability.

[0152] In summary, the random forest results show that the model is not prone to overfitting, has high prediction accuracy, fast training speed, and can effectively monitor the drought situation in Hebi City.

[0153] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for monitoring agricultural drought using surface soil moisture, characterized in that: include: Soil moisture observation sites in the study area were selected as locations for agricultural drought sampling, and soil moisture data at various depths at the sampling sites were collected; Calculating the agricultural drought grade of the sample according to the classification standard of the soil relative moisture of the deep soil moisture data; Microwave satellite remote sensing soil moisture products are used to extract surface microwave satellite remote sensing soil moisture data at the sample points; Based on the random forest algorithm, the quantitative relationship between the agricultural drought level of the sample and the surface microwave satellite remote sensing soil moisture data is mined, the agricultural drought level of the region is calculated based on the quantitative relationship, and the accuracy of the classification of the agricultural drought level of the region is described using a verification index; The calculation of soil relative humidity includes, in, represents the relative soil moisture, a represents the crop development period regulation coefficient, n Indicates the number of observation layers within the soil thickness corresponding to the crop development stage, Indicates the i Layer soil moisture, Indicates the i Field capacity of soil layer; The classification of agricultural drought levels on the sample includes: The relationship between the relative soil moisture and agricultural drought is expressed by using fuzzy logic, and a membership function of the relative soil moisture to agricultural drought is determined in a form similar to a Gaussian function. The calculation of the membership function includes: in, Represents pixels In the k The membership degree of drought degree at each moment, Indicates the typical value of soil relative moisture during a severe drought. The relative humidity of soil The value of the moment, w The median value of soil relative moisture during moderate drought is The difference, r represents the slope parameter of the control curve, c Represents the membership value at a known point; According to the classification principle of agricultural drought levels, a Gaussian membership function of soil relative humidity to agricultural drought is established, which is expressed as: in, It represents the degree of membership of soil relative moisture to agricultural drought; Based on the Gaussian membership function, the minimum operator of the fuzzy set is used to calculate the agricultural drought grade on the final sample. The formula is as follows: in, represents the drought level on the sample; The agricultural drought level on the sample ranges from [0,1]; The extraction of surface microwave satellite remote sensing soil moisture data at the sample point includes: The soil volumetric water content of the SMAP microwave satellite soil moisture product is converted into the relative soil water content. The relative soil water content is the surface microwave satellite remote sensing soil moisture data at the sample point. The formula is as follows: in, express i OK j The first pixel on the column k The relative soil moisture content at that time, Indicates the pixel k Volumetric soil water content at the time of the day.

2. The method for monitoring agricultural drought using surface soil moisture according to claim 1, wherein: The classification standards of soil relative moisture include: According to the previous work and relevant literature review, the relative soil humidity at the 20cm site is selected as the best reference indicator for the research, that is, the relative soil humidity at the 20cm site is used to classify the agricultural drought level; If the relative soil humidity at the 20cm site is greater than 60%, the agricultural drought level is no drought; if the relative soil humidity at the 20cm site is greater than 50% and less than or equal to 60%, the agricultural drought level is light drought; if the relative soil humidity at the 20cm site is greater than 40% and less than or equal to 50%, the agricultural drought level is moderate drought; if the relative soil humidity at the 20cm site is greater than 30% and less than or equal to 40%, the agricultural drought level is severe drought; if the relative soil humidity at the 20cm site is less than or equal to 30%, the agricultural drought level is extreme drought.

3. The method for monitoring agricultural drought using surface soil moisture according to claim 1, wherein: The agricultural drought level of a region calculated by the quantitative relationship includes: Based on the random forest algorithm, the surface microwave satellite remote sensing soil moisture data was used as the independent variable, and the agricultural drought grade based on the soil relative humidity classification sample was used as the dependent variable. The Random Forest data package in Python was used to construct a random forest drought monitoring model. During the model training process, 70% to 80% of the sample data are selected to build the random forest drought monitoring model, and 20% to 30% of the samples are used to verify the random forest drought monitoring model. The modeling and verification are repeated multiple times to finally obtain the optimal forest drought monitoring model; The optimal forest drought monitoring model is used to explore the quantitative relationship between satellite remote sensing surface soil moisture and deep soil moisture that can best reflect agricultural drought, thereby achieving real-time monitoring of regional farmland drought.

4. The method for monitoring agricultural drought using surface soil moisture according to claim 3, wherein: The accuracy of the classification of agricultural drought levels in the region using validation indicators includes: By calculating the coefficient of determination R of the observed and simulated values ​​of the test set 2 The accuracy of the optimal forest drought monitoring model was verified by three indicators: root mean square error (RMSE) and mean absolute error (MAE); The calculation formulas of the three indicators are as follows: in, represents the simulated value of drought level, represents the average of the simulated values ​​of the drought level, represents the observed value of drought level, represents the mean of the observed values ​​of drought level, m represents the number of samples, i Indicates the serial number of the observation month.

5. A system for monitoring agricultural drought using surface soil moisture, using the method for monitoring agricultural drought using surface soil moisture according to any one of claims 1 to 4, characterized in that: include: The data collection unit is used to select soil moisture observation sites in the study area as locations for agricultural drought sampling and to collect soil moisture data at various depths at the sampling sites; a sample grade classification unit, configured to calculate the agricultural drought grade of the sample according to a classification standard of soil relative humidity of the deep soil moisture data; The regional drought level determination unit is used to use microwave satellite remote sensing soil moisture products to extract surface microwave satellite remote sensing soil moisture data on sample points, and based on the random forest algorithm, to mine the quantitative relationship between the agricultural drought level on the sample and the surface microwave satellite remote sensing soil moisture data, calculate the agricultural drought level of the region through the quantitative relationship, and use verification indicators to describe the accuracy of the classification of the agricultural drought level of the region.

6. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, a method for monitoring agricultural drought using surface soil moisture according to any one of claims 1 to 4 is implemented.

Images