A crop drought monitoring method based on multi-source data and application

Abstract

The application belongs to the technical field of data processing calculation, and provides a crop drought monitoring method based on multi-source data and application, which comprises collecting ground data; crop drought category labeling is performed on the ground data; an inversion model of meteorological data, remote sensing data and ground crop drought category is established; and the inversion model is applied to a target area for crop drought; the application provides a more comprehensive and more accurate crop drought monitoring and evaluation method by combining meteorological data, remote sensing data and ground data; and the model is based on the collection of ground samples and pays attention to the importance of ground data used in the model, because this provides a key link between satellite observation and actual field conditions, the integration of such data sources strengthens a stronger and more accurate understanding of crop drought, provides valuable insights into current crop conditions and potential future trends, and helps to improve crop management practices and reduce the impact of drought on agricultural production.

Description

Technical Field

[0001] This invention relates to the fields of remote sensing and drought monitoring technology, specifically to a crop drought monitoring method and application based on multi-source data. Background Technology

[0002] Central Asia's arid regions are highly vulnerable to drought, and poor water resource management and continued reliance on irrigated agriculture further exacerbate its impact. Identifying and monitoring the occurrence and development of drought in Central Asia is of great significance for formulating regional policies and disaster prevention and mitigation. Currently, drought monitoring mainly relies on meteorological data and remote sensing, but current monitoring technologies still have some limitations:

[0003] 1. The drought index calculated from the meteorological data reanalysis dataset has low accuracy and cannot achieve high-precision drought detection. Ground-based measured data has higher accuracy, but its acquisition and laboratory processing are difficult; remote sensing data has higher accuracy, but it requires verification by ground-based measured data. Existing drought indices are not suitable for drought monitoring in arid areas.

[0004] 2. Existing methods use remote sensing data to invert meteorological indices, but the results are not very accurate and cannot accurately monitor the drought conditions of specific farmland.

[0005] 3. Existing drought level assessment methods are mostly based on subjective human evaluation and lack objective standards. Summary of the Invention

[0006] (a) Technical problems to be solved

[0007] To address the shortcomings of existing technologies, this invention provides a crop drought monitoring method and application based on multi-source data, in order to solve the problems mentioned in the background section.

[0008] (II) Technical Solution

[0009] To achieve the above objectives, the present invention provides a method for monitoring crop drought based on multi-source data, characterized by comprising the following steps:

[0010] S1: Collect ground data;

[0011] S2: Label crop drought categories in ground data;

[0012] S3: Establish an inversion model of meteorological data, remote sensing data, and drought categories of ground crops;

[0013] S4: Apply the inversion model to the target area for crop drought monitoring.

[0014] As a further preferred option, the ground data in S1 includes relevant data on farmland physiological conditions and habitat information. Specifically, the relevant data on farmland physiological conditions and habitat information includes soil moisture at 0-10cm depth, soil moisture at 10-20cm depth, soil temperature at 0-10cm depth, soil temperature at 10-20cm depth, canopy temperature, air temperature at 2m depth, crop wet weight, dry weight after drying, and relative moisture content of the sample, while also recording the latitude and longitude information of the area where the sample point is located.

[0015] As a further preferred option, the crop drought category labeling of the ground data in S2 specifically involves labeling the collected ground data using data mining methods, and the labeling results can be used as the dependent variable dataset for establishing the inversion model.

[0016] As a further preferred embodiment, the annotation of the collected ground data using data mining methods specifically includes processing the ground data collected by S1 using the K-means clustering algorithm and dividing it into four categories: no drought, mild drought, moderate drought, and severe drought. The specific classification process is as follows:

[0017] By analyzing the Spearman rank correlation of the sample data, the data types that are strongly correlated with temperature were identified. Then, the data of the strongly correlated data types in the sample data were compared with the measured WDI value calculated at 2m. After processing by the K-means clustering algorithm, the samples were divided into four categories: no drought, mild drought, moderate drought, and severe drought.

[0018] As a further preferred embodiment, the process of establishing the inversion model in S3 includes:

[0019] The inverse feature elimination method is used to reduce the dimensionality of the collected data indicators. The features are ranked according to their importance in the random forest model. While ensuring that the model results are not reduced, the indicators with low feature importance are gradually eliminated. The selected indicators are used as inversion factors.

[0020] Based on measured data and remote sensing data from sample points, a crop drought index (CDI) was constructed. The independent variable is the inversion factor corresponding to the coordinates of the sample points, and the dependent variable is the drought level label of the sample points, thus completing the establishment of the inversion model.

[0021] As a further preferred embodiment, the process of applying the inversion model to the target area for crop drought in S4 includes first preprocessing the farmland data of the target area, using the established inversion model to convert multi-source data into a quantitative assessment of crop drought, calculating the current crop drought status, and completing the quantitative analysis of the current crop drought status.

[0022] The farmland data for the target area includes meteorological data, remote sensing data, and ground data.

[0023] As a further preferred option, the above-mentioned inversion model based on the multi-source data crop drought monitoring method is applied to agricultural drought monitoring.

[0024] (III) Beneficial Effects

[0025] This invention provides a crop drought monitoring method and application based on multi-source data, which has the following beneficial effects:

[0026] 1. This invention provides a more comprehensive and accurate method for monitoring and assessing crop drought by combining meteorological data, remote sensing data, and ground data. The integration of these different data sources highlights the combined advantages of satellite observation and ground measurement in crop drought research. Furthermore, this model is based on ground sample collection and emphasizes the importance of ground data usage in the model, as it provides a crucial link between satellite observations and actual field conditions. This integration of data sources enhances a more robust and accurate understanding of crop drought, providing valuable insights into current crop conditions and potential future trends, and contributing to improved crop management practices and mitigation of the impact of drought on agricultural production.

[0027] 2. The inversion model of this invention is based on measured samples and remote sensing data, and has high accuracy and reliability. Compared with traditional methods that identify regional crop drought conditions based on literature or recorded drought events, this model can identify drought conditions more accurately. This model can achieve real-time monitoring of crop drought conditions over a large area. With the continuous development of satellite remote sensing technology and meteorological observation technology, the cost of acquiring remote sensing data and meteorological data is constantly decreasing, while the spatiotemporal resolution of the data is also increasing, which can meet the needs of large-scale crop drought monitoring. This model can also provide decision support for agricultural production. Through real-time monitoring and prediction of crop drought conditions, corresponding measures can be taken in a timely manner, such as irrigation and adjustment of planting structure, to ensure the normal operation of agricultural production and reduce economic losses and social impacts caused by drought. Attached Figure Description

[0028] Figure 1 This is a flowchart of the crop drought monitoring method based on multi-source data of the present invention;

[0029] Figure 2 This is a graph showing the correlation of key physiological parameters of crops in Embodiment 2 of the present invention;

[0030] Figure 3 Principal component analysis and drought classification diagram of key crop parameters in Example 2 of this invention;

[0031] Figure 4 This is a correlation analysis diagram of variables after dimensionality reduction of crop data in Embodiment 2 of the present invention;

[0032] Figure 5This is a graph showing the importance of variables after dimensionality reduction in Embodiment 2 of the present invention.

[0033] Figure 6 This is a statistical chart showing the number of model iterations and accuracy in Embodiment 2 of the present invention;

[0034] Figure 7 This is a graph showing the actual drought category and predicted category analysis in Embodiment 2 of the present invention;

[0035] Figure 8 This is a TVDI map showing the spatial distribution of crop drought in the sampling area and the Central Asian study area in Embodiment 2 of the present invention. Detailed Implementation

[0036] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0037] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the 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, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0038] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection, an electrical connection, or a connection that allows for communication; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0039] In this invention, unless otherwise explicitly specified and limited, "above" or "below" the second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of the second feature includes the first feature being directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature includes the first feature being directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature.

[0040] The following disclosure provides many different embodiments or examples for implementing various structures of the invention. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the invention. Furthermore, reference numerals and / or letters may be repeated in different examples; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed. In addition, examples of various specific processes and materials are provided in this invention, but those skilled in the art will recognize the application of other processes and / or the use of other materials.

[0041] Crop drought monitoring is a research hotspot in agricultural remote sensing monitoring. The impact of drought is widespread, spatially and temporally variable, and influenced by both natural and anthropogenic factors. Therefore, before conducting drought monitoring, it is necessary to establish an index that can describe drought phenomena. This index is formed by merging one or more meteorological or hydrological elements into a specific value to characterize and describe the dryness and wetness of a region and the level of drought in the area. Compared with the original meteorological or hydrological elements, this index is more reliable, representative, and easy to read. However, the drought index needs to be combined with expert experience to conduct quantitative and qualitative analysis of drought conditions. Therefore, machine learning methods will be used to automatically learn the drought characteristics of the samples, greatly reducing the error of human identification, and showing significant effects in fields such as crop drought.

[0042] In the past, meteorological data was the primary source of information for crop drought monitoring. This method was limited by geographical location, weather conditions, and instrument accuracy, resulting in lower accuracy. With the development of remote sensing technology, existing techniques have adopted a combined approach using remote sensing, meteorological, and ground-based data, improving the overall data quality. Remote sensing data, with its wide global coverage and real-time performance, comprehensively reflects the state of the atmosphere and surface environment, providing more comprehensive and accurate information for forecasting. By combining the advantages of remote sensing and meteorological data with ground-based data, the precision and accuracy of drought monitoring have been significantly enhanced.

[0043] This invention provides a crop drought monitoring method based on multi-source data, characterized by the following steps:

[0044] S1: Collect ground data;

[0045] S2: Label crop drought categories in ground data;

[0046] S3: Establish an inversion model of meteorological data, remote sensing data, and drought categories of ground crops;

[0047] S4: Apply the inversion model to the target area for crop drought monitoring.

[0048] Specifically, by collecting and analyzing ground data, the degree of drought affecting crops can be determined, laying the foundation for the subsequent establishment of an inversion model. Through the combined analysis of meteorological, remote sensing, and ground data, an inversion model can be established to identify crop drought conditions from remote sensing data. By applying this inversion model, the drought status of crops in arid areas can be monitored precisely, providing technical support for agricultural production. The model design of this invention fully utilizes the advantages of meteorological, remote sensing, and ground data, enabling comprehensive monitoring of drought conditions in arid areas.

[0049] Furthermore, the ground data in S1 includes relevant data on farmland physiological conditions and habitat information. Specifically, the relevant data on farmland physiological conditions and habitat information includes soil moisture at 0-10cm depth, soil moisture at 10-20cm depth, soil temperature at 0-10cm depth, soil temperature at 10-20cm depth, canopy temperature, air temperature at 2m depth, crop wet weight, dry weight after drying, and relative moisture content of the sample, while also recording the latitude and longitude information of the area where the sample point is located.

[0050] Furthermore, in S2, the labeling of crop drought categories on the ground data specifically involves labeling the collected ground data using data mining methods. The labeling results can be used as the dependent variable dataset for establishing the inversion model.

[0051] Furthermore, the annotation of the collected ground data using data mining methods specifically includes processing the ground data collected by S1 using the K-means clustering algorithm to divide it into four categories: no drought, mild drought, moderate drought, and severe drought. The specific classification process is as follows:

[0052] By analyzing the Spearman rank correlation of the sample data, the data types that are strongly correlated with temperature were identified. Then, the data of the strongly correlated data types in the sample data were compared with the measured WDI value calculated at 2m. After processing by the K-means clustering algorithm, the samples were divided into four categories: no drought, mild drought, moderate drought, and severe drought.

[0053] Furthermore, the process of establishing the inversion model in S3 includes:

[0054] The inverse feature elimination method is used to reduce the dimensionality of the collected data indicators. The features are ranked according to their importance in the random forest model. While ensuring that the model results are not reduced, the indicators with low feature importance are gradually eliminated. The selected indicators are used as inversion factors.

[0055] Based on measured data and remote sensing data from sample points, a crop drought index (CDI) was constructed. The independent variable is the inversion factor corresponding to the coordinates of the sample points, and the dependent variable is the drought level label of the sample points, thus completing the establishment of the inversion model.

[0056] Furthermore, the application process of the inversion model to the target area for crop drought in S4 includes first preprocessing the farmland data of the target area, using the established inversion model to convert multi-source data into a quantitative assessment of crop drought, calculating the current crop drought status, and completing the quantitative analysis of the current crop drought status.

[0057] The farmland data for the target area includes meteorological data, remote sensing data, and ground data.

[0058] Example 2

[0059] This embodiment discloses a crop drought monitoring method based on multi-source data, including:

[0060] Step 1: Collect relevant data on farmland physiological conditions and habitat information, specifically including soil moisture at 0-10cm depth, soil moisture at 10-20cm depth, soil temperature at 0-10cm depth, soil temperature at 10-20cm depth, canopy temperature, air temperature at 2m depth, wet weight of wheat, dry weight after drying, and relative moisture content of the samples. Simultaneously, record the latitude and longitude information of the sampling points. This data can be acquired using various sensors and devices, such as weather stations, soil moisture sensors, and handheld GPS recorders.

[0061] Step 2: Label the ground data with crop drought categories.

[0062] It should be noted that this step requires labeling the collected ground data using data mining methods. Data mining methods such as correlation analysis and K-means clustering analysis can be used in this step, and the labeling results can be used as the dependent variable dataset for model building.

[0063] Specifically, please refer to Figure 2 and 3K-means clustering algorithm was used to classify key physiological parameters of wheat samples to achieve a reasonable classification of crop drought conditions. Key physiological parameters included soil moisture at 0-10 cm depth, soil moisture at 10-20 cm depth, soil temperature at 0-10 cm depth, soil temperature at 10-20 cm depth, canopy temperature, air temperature at 2 m depth, wet weight of wheat, dry weight after drying, and relative moisture content of the sample.

[0064] Spearman's rank correlation analysis of the sample data revealed that temperature and humidity were the main classification factors. Within the temperature category, canopy temperature, canopy-air temperature difference, and soil temperature showed a strong correlation. Regarding humidity, soil moisture and plant water content also showed a certain correlation. To ensure the classification results accurately reflected reality, this section used the measured WDI values ​​calculated from the canopy temperature and atmospheric temperature at 2m from the sample data as a reference. After processing with the K-means clustering algorithm, the samples were divided into four categories: no drought, mild drought, moderate drought, and severe drought.

[0065] Step 3: Establish an inversion model of meteorological data, remote sensing data, and drought categories of ground crops.

[0066] Specifically, this step requires using the random forest method to establish the relationship between multi-source data and crop drought categories, thereby obtaining an inversion model.

[0067] Please see Figure 4 In this step, the data dimensionality reduction method employs a correlation matrix and the feature importance method from a random forest. Specifically, one or two highly correlated indicators are retained from the correlation matrix, while indicators with low or zero feature importance from the random forest are removed. To further integrate the spectral characteristics of ground crop physiological conditions, this experiment collected spectra of wheat under different dry and wet conditions and plotted spectral characteristic curves. To reduce the dimensionality of the indicators, this section uses a reverse feature elimination method. Features are ranked according to their importance in the random forest model, and indicators with low feature importance are gradually eliminated without compromising the model results. The selected indicators are S2_REP, S2_MNDVI, S2_PSRI, MOD_MSAVI, MOD_TCI, and ERA_t2m.

[0068] It should be noted that the criteria for selecting appropriate inversion factors from multi-source data include two aspects: first, the correlation between indicators should be as weak as possible to avoid multicollinearity problems between indicators; second, the indicators should be able to summarize the physiological characteristics of crop drought as comprehensively as possible.

[0069] Furthermore, vegetation indices were calculated using multi-source images from Sentinel-2 remote sensing data (Sentinel 2), the MODIS medium-resolution imaging spectrometer, and ERA-5 meteorological data. Inversion factors were selected for each sample point based on correlation analysis and reverse feature elimination. The final inversion factors include S2_REP, S2_MNDVI, S2_PSRI, MOD_MSAVI, MOD_TCI, and ERA_t2m, which reflect the crop drought status at each sample point.

[0070] For further details, please refer to Figure 5 Based on the measured data and remote sensing data of the sample points, the crop drought index (CDI) was constructed. The independent variable is the inversion factor corresponding to the coordinates of the sample points, and the dependent variable is the drought level label of the sample points.

[0071] For example, in order to obtain a more accurate inversion model, we chose the linear model, support vector machine, backpropagation neural network and random forest method, and divided the samples into training set and test set in a 6:4 ratio to build the model.

[0072]

[0073] Inversion accuracy statistics of different inversion models

[0074] The inversion accuracy statistics show that all four methods deliver excellent prediction results. Linear models and random forests exhibit relatively small fitting and prediction errors, while support vector machines and backpropagation neural networks show slightly higher prediction errors on the test set, but still meet the needs of practical applications. Considering the inversion accuracy, computational efficiency, and feasibility of practical applications, we can select the appropriate model for crop drought prediction and monitoring based on the specific application scenario. Under the current data conditions, random forests show the highest fitting accuracy. Furthermore, to better adjust the model parameters, the sample classification and iteration count were adjusted to achieve the most suitable parameters. The iteration count and accuracy score are shown below. Figure 6 and Figure 7 .

[0075] Step 4: Apply the model established in Step 3 to farmland in the arid Central Asian region to monitor crop drought conditions.

[0076] Specifically, in practical applications, the fourth step requires preprocessing farmland data in the arid Central Asian region, including meteorological data, remote sensing data, and ground data. Using the established inversion model, the multi-source data is converted into a quantitative assessment of crop drought. The current crop drought status is calculated, thereby enabling a quantitative analysis of the current crop drought situation.

[0077] Comparative Example 1

[0078] The Temperature-Vegetation Drought Index (TVDI) is calculated using meteorological or remote sensing data.

[0079] The crop drought monitoring method obtained in Example 2 and Comparative Example 1 were applied to spatial monitoring of crop drought in the Central Asian study area, and the monitoring results are as follows: Figure 8 .

[0080] See Figure 8 It is known that the Temperature Vegetation Drought Index (TVDI) can be calculated using meteorological or remote sensing data. However, TVDI is not very effective in arid regions due to resolution limitations; it can only characterize the presence and severity of drought, but cannot distinguish between different degrees of drought. This invention patent uses a method combining ground data, remote sensing data, and meteorological data to overcome the shortcomings of single data sources, providing a more comprehensive and accurate method for monitoring and assessing crop drought.

[0081] This data integration method highlights the advantages of combining satellite observation and ground measurements, particularly emphasizing the use of ground data in the model, thus providing a crucial link between satellite observations and actual field conditions. This method can enhance our understanding of crop drought, help improve crop management practices, and mitigate the impact of drought on agricultural production. By applying this method to monitor crop drought in farmland, we can study the growth patterns of crops in arid regions of Central Asia, further optimize planting methods, promote crop growth, and increase yields. This invention patent, through multi-source data joint analysis, quantitatively assesses crop drought conditions. It features simple research methods, abundant data sources, accurate results, and wide applicability, making it an important tool for studying crop drought in arid regions and of great significance for promoting crop growth and improving agricultural production efficiency.

[0082] In summary, the multi-source data-based crop drought monitoring method of this invention combines meteorological data, remote sensing data, and ground data, providing a more comprehensive and accurate method for monitoring and assessing crop drought. The integration of these different data sources highlights the combined advantages of satellite observation and ground measurement in crop drought research. Furthermore, this model is based on ground sample collection and emphasizes the importance of ground data usage in the model, as this provides a crucial link between satellite observations and actual field conditions. This integration of data sources enhances a more robust and accurate understanding of crop drought, providing valuable insights into current crop conditions and potential future trends, and contributing to improved crop management practices and mitigation of the impact of drought on agricultural production.

[0083] Furthermore, by establishing an inversion model, this invention's inversion model, based on measured samples and remote sensing data, possesses high accuracy and reliability. Compared to traditional methods that identify regional crop drought conditions based on literature or recorded drought events, this model can more accurately identify drought conditions. This model can achieve real-time monitoring of crop drought conditions over a large area. With the continuous development of satellite remote sensing and meteorological observation technologies, the cost of acquiring remote sensing and meteorological data is constantly decreasing, while the spatiotemporal resolution of the data is also increasing, meeting the needs of large-scale crop drought monitoring. This model can also provide decision support for agricultural production. Through real-time monitoring and prediction of crop drought conditions, corresponding measures can be taken in a timely manner, such as irrigation and adjusting planting structures, to ensure the normal operation of agricultural production while reducing economic losses and social impacts caused by drought.

[0084] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A crop drought monitoring method based on multi-source data, characterized in that: Includes the following steps: S1: Collect ground data; S2: Label crop drought categories in ground data; S3: Establish an inversion model of meteorological data, remote sensing data, and drought categories of ground crops; S4: Apply the inversion model to the target area for crop drought monitoring; In S2, the labeling of crop drought categories on ground data specifically involves labeling the collected ground data using data mining methods. The labeling results can be used as the dependent variable dataset for establishing the inversion model. The annotation of the collected ground data using data mining methods specifically includes processing the ground data collected by S1 using the K-means clustering algorithm and dividing it into four categories: no drought, mild drought, moderate drought, and severe drought. The specific classification process is as follows: By analyzing the Spearman rank correlation of the sample data, the data types that are strongly correlated with temperature were identified. Then, the data of the strongly correlated data types in the sample data and the measured WDI values ​​calculated at 2m were used as references. After processing by the K-means clustering algorithm, the samples were divided into four categories: no drought, mild drought, moderate drought, and severe drought. The process of establishing the inversion model in S3 includes: The inverse feature elimination method is used to reduce the dimensionality of the collected data indicators. The features are ranked according to their importance in the random forest model. While ensuring that the model results are not reduced, the indicators with low feature importance are gradually eliminated. The selected indicators are used as inversion factors. Based on measured data and remote sensing data from sample points, a crop drought index (CDI) was constructed. The independent variable is the inversion factor corresponding to the coordinates of the sample points, and the dependent variable is the drought level label of the sample points, thus completing the establishment of the inversion model.

2. The crop drought monitoring method based on multi-source data according to claim 1, characterized in that: The ground data in S1 includes relevant data on farmland physiological conditions and habitat information. Specifically, the relevant data on farmland physiological conditions and habitat information includes soil moisture at 0-10cm depth, soil moisture at 10-20cm depth, soil temperature at 0-10cm depth, soil temperature at 10-20cm depth, canopy temperature, air temperature at 2m depth, crop wet weight, dry weight after drying, and relative moisture content of the sample. At the same time, the latitude and longitude information of the area where the sample point is located is recorded.

3. The crop drought monitoring method based on multi-source data according to claim 1, characterized in that: The application process of applying the inversion model to the target area for crop drought monitoring in S4 includes first preprocessing the farmland data of the target area, then using the established inversion model to convert multi-source data into a quantitative assessment of crop drought, and finally calculating the current crop drought status to complete the quantitative analysis of the current crop drought status. The farmland data for the target area includes meteorological data, remote sensing data, and ground data.

4. The application of the inversion model established by the multi-source data crop drought monitoring method as described in any one of claims 1-3 in agricultural drought monitoring.

Images