A basin seasonal drought analysis method and system based on MCDI index

Abstract

The present invention relates to the technical field of electronic digital data processing, and in particular to a method and system for analyzing watershed quarterly drought based on the MCDI index. The method comprises the following steps: performing dry spectrum and hydrological classification analysis on watershed information to obtain watershed dry spectrum and watershed classification hydrological data; constructing a correlation matrix on watershed classification hydrological data based on dry spectrum to obtain dryness-related hydrological matrix; performing drought parameter analysis on watershed classification hydrological data based on dryness-related hydrological matrix to obtain relevant drought parameters; constructing a neural network model for relevant drought parameters and performing MCDI drought index analysis to obtain MCDI drought index; performing drought analysis on watershed according to MCDI drought index to obtain watershed drought analysis data. The present invention can determine multiple drought parameters through dry spectrum to ensure the accuracy of MCDI drought index.

Description

Technical Field

[0001] The invention relates to the technical field of electronic digital data processing, and in particular to a basin seasonal drought analysis method and system based on the MCDI index. Background Art

[0002] The analysis of seasonal drought in a watershed plays an important role in regional water supply. Previous studies on seasonal drought in a watershed mainly used single indices such as the Standardized Precipitation Index (SPI) to analyze seasonal drought conditions. Most of them only considered a single factor and did not fully consider the impact of meteorological, hydrological, and agricultural droughts on seasonal droughts. The multivariate comprehensive drought index MCDI was proposed by combining multiple indices with a function to comprehensively analyze seasonal drought conditions. However, to analyze seasonal drought in a watershed based on the MCDI drought index, it is necessary to obtain a variety of drought parameters and process the drought parameters to output an index. Using a single drought parameter cannot accurately analyze seasonal drought in a watershed. It is necessary to select the correlation closeness of drought parameters based on a drought correlation factor, otherwise the practicality and convenience of the MCDI drought index will be limited. Summary of the invention

[0003] Based on this, it is necessary for the present invention to provide a basin quarterly drought analysis system and method based on the MCDI index to solve at least one of the above technical problems.

[0004] To achieve the above purpose, a basin quarterly drought analysis method based on MCDI index includes the following steps:

[0005] Step S1: Acquire the watershed information of the watershed to obtain watershed information data; perform dry spectrum analysis on the watershed information data to obtain the watershed dry spectrum; perform hydrological classification analysis on the watershed to obtain watershed classification hydrological data;

[0006] Step S2: constructing a correlation matrix for the classified hydrological data of the watershed based on the dry spectrum to obtain a dry correlation hydrological matrix; performing drought parameter analysis on the classified hydrological data of the watershed based on the dry correlation hydrological matrix to obtain meteorological drought parameters, hydrological drought parameters and agricultural drought parameters;

[0007] Step S3: constructing a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters to obtain a drought neural network model; performing MCDI drought index analysis on the drought neural network model based on the dry spectrum to obtain the MCDI drought index;

[0008] Step S4: divide the watershed into intervals according to the MCDI drought index to obtain watershed drought assessment intervals; perform drought analysis on the watershed based on the watershed drought assessment intervals to obtain watershed drought analysis data.

[0009] The present invention obtains basin information data, including dry spectrum and hydrological classification data, and has a clear and comprehensive understanding of the basin, and can understand the characteristics and classification of the basin, providing basic data support for subsequent drought analysis. By constructing a correlation matrix and analyzing drought parameters, the hydrological conditions of the basin can be quantified and understood. By combining dry spectrum with hydrological data, the drought trend of the basin is determined, the drought situation of the basin is studied in depth, and the drought risk is comprehensively assessed from the perspectives of meteorology, hydrology and agriculture. A comprehensive assessment method is provided, which simultaneously considers multiple parameters to have a more comprehensive analysis perspective on the severity and impact of drought, and constructs a neural network model based on multiple analysis perspectives to calculate the drought index. Using dry spectrum analysis, the model can be improved and the MCDI drought index can be obtained. By using dry spectrum analysis, this step makes the analysis of the MCDI drought index more convenient and accurate. According to the MCDI drought index, the basin drought assessment interval is divided, and drought analysis is performed, so that the degree of drought in each interval of the basin can be intuitively understood, providing an important reference for disaster risk assessment and response planning under different degrees of drought.

[0010] Preferably, step S1 comprises the following steps:

[0011] Step S11: Acquire watershed information of the watershed to obtain watershed information data;

[0012] Step S12: performing remote sensing processing on the watershed to obtain watershed remote sensing data;

[0013] Step S13: performing dry spectrum analysis on the watershed information data based on the watershed remote sensing data to obtain a watershed dry spectrum;

[0014] Step S14: performing basin hydrological monitoring processing on the basin information data based on the basin remote sensing data to obtain basin hydrological monitoring data;

[0015] Step S15: Perform data classification processing on the watershed hydrological monitoring data to obtain watershed classified hydrological data.

[0016] The present invention acquires basin information from the basin, collects and analyzes basin-related data, including geographic, climate, and soil information, and can establish a comprehensive database of the basin. It provides an information basis for subsequent analysis and modeling, and provides basic data support for subsequent basin analysis. The application of remote sensing technology provides data for basin research. Through the processing and analysis of remote sensing images, the remote sensing-based data acquisition method has a wide coverage and high efficiency, and can accurately process the acquired data to provide remote sensing images of the basin, help identify and analyze basin data, and provide more comprehensive data support for subsequent analysis. Based on the basin remote sensing data, dry spectral analysis is performed on the basin information data to reveal the water dynamics of the basin. Spectral analysis is performed on remote sensing data to determine the content and distribution of water in the basin. A new analysis angle is provided for the MCDI drought index, and it can also help to obtain and define subsequent drought parameters. Through comprehensive analysis of remote sensing data and basin information, the hydrological dynamic data of the basin can be monitored, including changes in rainfall, water flow, and water level. By classifying the basin hydrological monitoring data to identify different hydrological characteristics, it is helpful to better understand the hydrological changes in the basin and provide more accurate data support for basin drought analysis.

[0017] Preferably, step S13 comprises the following steps:

[0018] Step S131: performing dust data and dryness analysis on the watershed information data based on the watershed remote sensing data to obtain the watershed dust data and watershed dryness;

[0019] Step S132: performing spectral analysis on the watershed dust data based on the watershed remote sensing data to obtain the watershed dust spectrum;

[0020] Step S133: performing spatial ratio analysis on the watershed dust data to obtain watershed dust spatial ratio data;

[0021] Step S134: mapping the horizontal and vertical coordinates of the coordinate axis of the watershed dryness based on the watershed dust space ratio data to obtain dust dryness related coordinate data;

[0022] Step S135: constructing a correlation curve for the dust drying correlation coordinate data to obtain a dust drying correlation curve;

[0023] Step S136: performing association rule analysis on the dust drying association curve to obtain dust drying association rules;

[0024] Step S137: Perform data replacement on the watershed dust spectrum based on the dust drying association rule to obtain the watershed drying spectrum.

[0025] The present invention obtains the dust data and dryness of the watershed by analyzing the dust data and dryness of the watershed information data based on the watershed remote sensing data. By analyzing the dust data and dryness, the dust distribution and dryness in the watershed can be more comprehensively understood. By performing spectral analysis on the watershed dust data, different types of dust components and distribution can be identified by spectral analysis of the dust data. It is helpful to provide basic spectral information for subsequent dry spectrum construction. Spatial ratio analysis can reveal the law of dust distribution in the watershed. Through spatial ratio analysis, the distribution ratio and density of dust in the watershed can be understood, which is helpful for subsequent correlation analysis based on dust ratio and dryness, and provides data support for analysis. Coordinate axis mapping processing helps to visualize the relationship between dust and dryness. By mapping the watershed dryness data and the dust space ratio data into a coordinate system, the relationship between the two can be intuitively expressed. Coordinate axis mapping processing helps to visualize the relationship between dust and dryness. By mapping the watershed dryness data and the dust space ratio data into a coordinate system, the relationship between the two can be intuitively expressed. The construction of the correlation curve enables the quantification of the relationship between dust and dryness. By drawing the dust-dryness correlation curve, the corresponding dryness changes under different dust levels can be determined, which is conducive to the subsequent conversion between spectral data. The dust-dryness correlation curve is analyzed by association rules to determine the rules between dust concentration changes and dryness changes, providing a reference for dust and dryness changes. By applying the dust-dryness association rules to the dust spectrum of the basin, the spectral data is modified and improved to better reflect the drought characteristics of the basin.

[0026] Preferably, step S132 includes the following steps:

[0027] Step 1321: performing dust particle data acquisition processing on the watershed dust data based on the watershed remote sensing data to obtain the watershed dust particle data;

[0028] Step 1322: performing wavelength analysis on the dust particle data in the watershed to obtain wavelength data of dust particles in the watershed;

[0029] Step S1323: Perform spatial concentration analysis on the dust particle data in the watershed to obtain spatial concentration data of dust in the watershed;

[0030] Step S1324: extracting the scattering characteristic peaks of the watershed remote sensing data according to the wavelength data of the watershed dust particles to obtain the watershed dust peak characteristic data;

[0031] Step S1325: Perform spectrum construction processing on the watershed dust spatial concentration data based on the watershed dust peak characteristic data to obtain the watershed dust spectrum.

[0032] The present invention processes remote sensing data to extract information about dust particles in the watershed, including size, shape and concentration data. It provides basic data for subsequent dust analysis and modeling, and helps to more comprehensively understand the dynamic state of dust in the watershed. By analyzing the data of dust particles in the watershed, the reflection, absorption and scattering behavior of dust at different wavelengths can be determined. It helps to distinguish dust particles in space from other particles in space. Spatial concentration analysis is performed on the data of dust particles in the watershed to obtain spatial concentration data of dust in the watershed. By analyzing the spatial variation of dust particle data in the watershed, the hot spot area of ​​dust concentration can be determined. It helps to acquire data for the impact analysis of dust concentration in the watershed. Peak extraction processing can identify the unique characteristics of dust in the spectrum. By analyzing the wavelength data of dust particles in the watershed, the obvious peaks generated by dust scattering and absorption can be detected. It helps to distinguish the difference between the dust spectrum and the spectrum of other substances, and can improve the accuracy of subsequent spectral analysis and modeling. By combining the dust peak characteristic data in the watershed with the dust spatial concentration data to construct spectral information, the unique optical properties of the dust in the watershed are represented, providing a tool for dust monitoring and analysis of dryness.

[0033] Preferably, step S2 comprises the following steps:

[0034] Step S21: performing correlation scatter plot processing on the watershed classification hydrological data according to the dry spectrum to obtain a dry hydrological correlation scatter plot;

[0035] Step S22: performing correlation coefficient analysis on the dry spectrum and the watershed classification hydrological data based on the dry hydrological correlation scatter plot to obtain a dry hydrological correlation coefficient;

[0036] Step S23: constructing a matrix of the dry hydrological correlation coefficient, the dry spectrum and the watershed classification hydrological data to obtain a dry correlation hydrological matrix;

[0037] Step S24: constructing an influence chain for the basin classification hydrological data based on the dry-related hydrological matrix to obtain a dry spectral hydrological influence chain;

[0038] Step S25: extracting hydrological data from the classified hydrological data of the watershed to obtain the watershed precipitation, watershed runoff and watershed evaporation;

[0039] Step S26: Based on the dry spectral hydrological impact chain, drought parameter analysis is performed on the basin precipitation, basin runoff and basin evaporation to obtain meteorological drought parameters, hydrological drought parameters and agricultural drought parameters.

[0040] The present invention draws a scatter plot to visually observe the relationship and distribution between two sets of data. It is helpful to evaluate the strength and direction of the association between the dry spectrum and the hydrological characteristics of the basin, and provides a basis for subsequent quantitative analysis. The relationship number analysis quantifies the relationship between the dry spectrum and the basin classification hydrological data. The statistical correlation between the two variables is determined by calculating the correlation coefficient. It is helpful to understand how strong the linear relationship is between the dry spectrum and the basin hydrological characteristics. By creating a dry-related hydrological matrix, the dry spectrum, the dry hydrological correlation coefficient and the basin classification hydrological data are simultaneously visualized and analyzed, providing a data structure basis for subsequent comprehensive analysis. By analyzing the dry-related hydrological matrix, it is possible to identify how spectral data affects hydrological parameters and determine the cascade effect between hydrological parameters. This influence chain is conducive to understanding the propagation and impact of drought conditions in the basin. Precipitation, runoff and evaporation are extracted from the basin classification hydrological data to provide basic data support for subsequent drought parameter analysis. The dry spectrum hydrological impact chain is applied to analyze precipitation, runoff and evaporation data to obtain meteorological, hydrological and agricultural drought parameters. These parameters can quantify the severity and impact of drought and facilitate the impact analysis of drought on different aspects of a particular watershed.

[0041] Preferably, step S3 comprises the following steps:

[0042] Step S31: constructing a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters to obtain a drought neural network model;

[0043] Step S32: performing data injection processing on the drought neural network model based on the dry spectrum to obtain dry spectrum injection data;

[0044] Step S33: performing drought parameterization processing on the dryness spectrum injection data to obtain dryness-related drought data;

[0045] Step S34: performing MCDI drought index analysis on drought-related drought data based on the drought neural network model to obtain the MCDI drought index.

[0046] The present invention obtains a drought neural network model by constructing a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters. By constructing a neural network model, the influence of multiple drought parameters can be comprehensively considered, and the drought parameters can be preprocessed, which is helpful for the subsequent calculation of the drought index. Through data injection processing, the dry spectrum data is introduced into the neural network model to consider the influence of dry factors on drought and improve the accuracy and reliability of the model. By analyzing the dry spectrum injection data, key features related to the severity and impact of drought are extracted. This parameterization process quantifies the information of the dry spectrum and converts it into drought parameters that can be directly interpreted. By applying the drought neural network model, we can analyze dryness-related drought data and obtain the MCDI drought index. This index combines multiple drought parameters and provides a unified scale to evaluate the severity of drought. The MCDI drought index can quantify the quarterly drought analysis of the basin, which is helpful for the analysis of drought and flood changes in the basin and provides data support.

[0047] Preferably, step S31 includes the following steps:

[0048] Step S311: performing dryness analysis on the drying spectrum to obtain the watershed dryness;

[0049] Step S312: constructing a model input and output for the meteorological drought parameters, the hydrological drought parameters, and the agricultural drought parameters to obtain a model input layer and a model output layer;

[0050] Step S313: constructing an influence degree sequence of meteorological drought parameters, hydrological drought parameters and agricultural drought parameters based on the watershed dryness to obtain a drought parameter influence sequence;

[0051] Step S314: dividing the drought parameter impact sequence by impact ratio to obtain drought parameter impact ratio data;

[0052] Step S315: Based on the drought parameter impact ratio data, the meteorological drought parameters, the hydrological drought parameters and the agricultural drought parameters are processed by connection weight division to obtain the connection weight data between models.

[0053] Step S316: constructing a drought function for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters based on the connection weight data between models to obtain the MCDI drought index function;

[0054] Step S317: constructing a hidden layer for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters based on the MCDI drought index function to obtain a hidden layer of the MCDI drought index function;

[0055] Step S318: Based on the connection weight data between models, the model input layer, the MCDI drought index function hidden layer and the model output layer are merged and constructed to obtain a drought neural network model.

[0056] The present invention can accurately evaluate the dryness of the basin through dryness analysis, and provide an important basis for subsequent drought parameter analysis and prediction. By defining the model input layer and output layer, the model is specified to process specific drought parameters and expect to obtain output. It is ensured that the model focuses on the drought analysis task and processes these specific parameters. By analyzing the influence sequence of basin dryness on different drought parameters, the dependence between parameters is determined. It is helpful to understand how to analyze drought under different drought parameters. By quantifying the influence of dryness on each drought parameter, it helps to identify key drought parameters and improve the accuracy of drought prediction. By dividing the influence ratio, the relative importance of each drought parameter in drought prediction can be clarified, providing a reference basis for the subsequent establishment of a drought neural network model. By dividing the connection weight, the correlation between each drought parameter can be determined, providing a basis for the connection weight of the drought neural network model. By constructing a drought function, the influence of each drought parameter can be comprehensively considered, and the MCDI drought index function can be established to provide a method for calculating and evaluating the drought index. Through the exponential function hidden layer, the processing capacity of the neural network model is increased. The hidden layer identifies complex data relationships, which is also conducive to enhancing the ability of the model to analyze the dynamic changes of drought. By constructing a drought neural network model, we can comprehensively consider the influence and correlation of various drought parameters, achieve a comprehensive assessment and prediction of drought conditions, and provide scientific support for drought monitoring.

[0057] Preferably, step S4 comprises the following steps:

[0058] Step S41: constructing a drought data distribution lattice for the MCDI drought index to obtain a MCDI drought index distribution lattice;

[0059] Step S42: performing normal analysis of drought data on the basin based on the MCDI drought index distribution matrix to obtain a normal distribution curve of the MCDI drought index;

[0060] Step S43: performing drought interval division processing on the MCDI drought index based on the MCDI drought index normal distribution curve to obtain the watershed drought interval;

[0061] Step S44: performing drought assessment processing on the drought interval of the watershed to obtain watershed drought assessment data;

[0062] Step S45: Perform drought analysis on the watershed according to the watershed drought assessment data to obtain watershed drought analysis data.

[0063] The present invention can intuitively display the spatial distribution of the MCDI drought index in the basin by constructing a drought index distribution dot matrix, help identify the spatial differences in the degree of drought, and provide a basis for subsequent analysis and interpretation. Through normal analysis, it is possible to understand whether the distribution characteristics of the MCDI drought index conform to the normal distribution, and provide a statistical basis for data processing and analysis. By analyzing the normal distribution curve of the MCDY drought index, different drought level intervals are defined. It is helpful to standardize drought assessment and provide a unified scale for quantifying the severity of drought. Through drought assessment processing, the drought situation in the basin can be assessed according to the division results of the drought intervals, and the assessment is conducive to more quickly defining the drought and flood situation in the basin. By analyzing the basin drought assessment data, the drought situation in the basin can be deeply understood, which is conducive to the prevention of drought and flood disasters in the basin area.

[0064] Preferably, step S43 includes the following steps:

[0065] Step S431: constructing a characteristic curve for the MCDI drought index to obtain a drought characteristic curve;

[0066] Step S432: performing quantile threshold division processing on the drought characteristic curve based on the MCDI drought index to obtain the curve threshold quantile;

[0067] Step S433: performing curve fitting processing on the MCDI drought index normal distribution curve according to the curve threshold quantile points to obtain the normal threshold quantile points;

[0068] Step S434: sorting the normal threshold points in ascending order to obtain an ascending sequence of normal threshold points;

[0069] Step S435: construct drought intervals for the MCDI drought index based on the normal threshold point ascending sequence to obtain the watershed drought intervals.

[0070] By constructing a characteristic curve, the present invention can intuitively display the changing trend and characteristics of the MCDI drought index, help identify the laws and characteristics of drought development, and provide an important reference for drought characteristic analysis. By dividing the exponential distribution into different quantile thresholds, the drought grade interval is defined. It is helpful to standardize drought assessment and provide a unified scale for quantifying the severity of drought. By performing curve fitting on the exponential distribution and considering the quantile threshold, the quantile of the normal distribution curve is obtained. It helps to understand the regularity and statistical characteristics of the drought index distribution and provides a basis for subsequent statistical inference intervals. Through ascending sorting processing, the normal threshold points can be arranged in order of size to help identify the severity and distribution law of drought. Through the construction of drought intervals, the MCDI drought index can be divided into different drought intervals according to the normal threshold points, providing a data basis for determining the drought grade.

[0071] Preferably, the present invention further provides a basin seasonal drought analysis system based on the MCDI index, which is used to execute the basin seasonal drought analysis method based on the MCDI index as described above. The basin seasonal drought analysis system based on the MCDI index comprises:

[0072] The watershed information acquisition module is used to acquire watershed information in the watershed to obtain watershed information data; perform dry spectrum analysis on the watershed information data to obtain watershed dry spectrum; perform hydrological classification analysis on the watershed to obtain watershed classification hydrological data;

[0073] The watershed drought parameter analysis module is used to construct a correlation matrix of the watershed classification hydrological data based on the dry spectrum to obtain the dryness-related hydrological matrix; based on the dryness-related hydrological matrix, the watershed classification hydrological data is analyzed for drought parameters to obtain meteorological drought parameters, hydrological drought parameters and agricultural drought parameters;

[0074] The neural network model MCDI drought index analysis module is used to construct a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters to obtain a drought neural network model; the MCDI drought index analysis is performed on the drought neural network model based on the dry spectrum to obtain the MCDI drought index;

[0075] The watershed drought analysis module is used to divide the watershed into intervals based on the MCDI drought index to obtain the watershed drought assessment interval; based on the watershed drought assessment interval, drought analysis is performed on the watershed to obtain watershed drought analysis data.

[0076] In summary, the present invention provides a watershed quarterly drought analysis system based on the MCDI index. The watershed quarterly drought analysis system based on the MCDI index consists of a watershed information acquisition module, a watershed drought parameter analysis module, a neural network model MCDI drought index analysis module, and a watershed drought analysis module. It can realize any watershed quarterly drought analysis method based on the MCDI index described in the present invention, and is used to combine the operations between computer programs running on each module to realize any watershed quarterly drought analysis method based on the MCDI index. The internal structures of the system cooperate with each other, which can greatly reduce duplication of work and manpower investment, and can quickly and effectively provide a more accurate and efficient drought analysis process, thereby simplifying the operational process of watershed quarterly drought analysis based on the MCDI index. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings:

[0078] Figure 1It is a schematic diagram of the steps of the method for analyzing seasonal drought in a watershed based on the MCDI index of the present invention;

[0079] Figure 2 for Figure 1 Detailed step flow diagram of step S1;

[0080] Figure 3 for Figure 2 Detailed step flow diagram of step S13;

[0081] Figure 4 for Figure 3 Detailed step flow chart of step S132 in FIG. DETAILED DESCRIPTION

[0082] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0083] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0084] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0085] To achieve this, please refer to Figures 1 to 4 The present invention provides a basin quarterly drought analysis method based on the MCDI index, comprising the following steps:

[0086] Step S1: Acquire the watershed information of the watershed to obtain watershed information data; perform dry spectrum analysis on the watershed information data to obtain the watershed dry spectrum; perform hydrological classification analysis on the watershed to obtain watershed classification hydrological data;

[0087] Step S2: constructing a correlation matrix for the classified hydrological data of the watershed based on the dry spectrum to obtain a dry correlation hydrological matrix; performing drought parameter analysis on the classified hydrological data of the watershed based on the dry correlation hydrological matrix to obtain meteorological drought parameters, hydrological drought parameters and agricultural drought parameters;

[0088] Step S3: constructing a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters to obtain a drought neural network model; performing MCDI drought index analysis on the drought neural network model based on the dry spectrum to obtain the MCDI drought index;

[0089] Step S4: divide the watershed into intervals according to the MCDI drought index to obtain watershed drought assessment intervals; perform drought analysis on the watershed based on the watershed drought assessment intervals to obtain watershed drought analysis data.

[0090] In the embodiment of the present invention, please refer to Figure 1 FIG. 1 is a schematic diagram of the steps of the method for analyzing seasonal drought in a watershed based on the MCDI index of the present invention. In this example, the method for analyzing seasonal drought in a watershed based on the MCDI index includes the following steps:

[0091] Step S1: Acquire the watershed information of the watershed to obtain watershed information data; perform dry spectrum analysis on the watershed information data to obtain the watershed dry spectrum; perform hydrological classification analysis on the watershed to obtain watershed classification hydrological data;

[0092] The embodiment of the present invention obtains watershed information, including geographical, climate, and soil information, to establish watershed information data, uses remote sensing technology to obtain dry spectral data of the watershed, analyzes and processes the obtained information data, performs hydrological classification analysis, and classifies the watershed according to hydrological characteristics to obtain watershed classification hydrological data.

[0093] Step S2: constructing a correlation matrix for the classified hydrological data of the watershed based on the dry spectrum to obtain a dry correlation hydrological matrix; performing drought parameter analysis on the classified hydrological data of the watershed based on the dry correlation hydrological matrix to obtain meteorological drought parameters, hydrological drought parameters and agricultural drought parameters;

[0094] The embodiment of the present invention establishes a correlation matrix by mining the intrinsic relationship between the dry spectrum and the hydrological data of the basin. The drought characteristics are extracted by analyzing the spectral data. The historical hydrological data of the basin, including rainfall, evapotranspiration, and runoff, are combined to establish a correlation relationship. The hydrological data is analyzed by establishing a relationship chain, and the drought factors are determined by multiple aspects such as meteorology, hydrology, and agriculture.

[0095] Step S3: constructing a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters to obtain a drought neural network model; performing MCDI drought index analysis on the drought neural network model based on the dry spectrum to obtain the MCDI drought index;

[0096] The embodiment of the present invention pre-processes the parameter data and selects a suitable network architecture, such as a feed-forward neural network, and defines the number of model layers and the connection mode between layers. The network weight is adjusted by a back propagation algorithm and an error minimization technique. Based on the well-trained neural network model, the MCDI drought index analysis is performed. The MCDI drought index is calculated for the above drought parameter data to obtain the MCDI drought index.

[0097] Step S4: divide the watershed into intervals according to the MCDI drought index to obtain watershed drought assessment intervals; perform drought analysis on the watershed based on the watershed drought assessment intervals to obtain watershed drought analysis data.

[0098] The embodiment of the present invention divides the drought degree of the basin into intervals according to the MCDI drought index, and divides the intervals by establishing the intermediate threshold quantile. The MCDI drought index is divided by percentile thresholds, and the intermediate threshold is determined. The determined thresholds are ranked in ascending order. The drought characteristic data between the thresholds are associated, and then the intervals are divided into degrees. By comparing the MCDI drought index of the basin with the thresholds of each drought level, the drought interval of the basin is determined.

[0099] The present invention obtains basin information data, including dry spectrum and hydrological classification data, and has a clear and comprehensive understanding of the basin, and can understand the characteristics and classification of the basin, providing basic data support for subsequent drought analysis. By constructing a correlation matrix and analyzing drought parameters, the hydrological conditions of the basin can be quantified and understood. By combining dry spectrum with hydrological data, the drought trend of the basin is determined, the drought situation of the basin is studied in depth, and the drought risk is comprehensively assessed from the perspectives of meteorology, hydrology and agriculture. A comprehensive assessment method is provided, which simultaneously considers multiple parameters to have a more comprehensive analysis perspective on the severity and impact of drought, and constructs a neural network model based on multiple analysis perspectives to calculate the drought index. Using dry spectrum analysis, the model can be improved and the MCDI drought index can be obtained. By using dry spectrum analysis, this step makes the analysis of the MCDI drought index more convenient and accurate. According to the MCDI drought index, the basin drought assessment interval is divided, and drought analysis is performed, so that the degree of drought in each interval of the basin can be intuitively understood, providing an important reference for disaster risk assessment and response planning under different degrees of drought.

[0100] Preferably, step S1 comprises the following steps:

[0101] Step S11: Acquire watershed information of the watershed to obtain watershed information data;

[0102] Step S12: performing remote sensing processing on the watershed to obtain watershed remote sensing data;

[0103] Step S13: performing dry spectrum analysis on the watershed information data based on the watershed remote sensing data to obtain a watershed dry spectrum;

[0104] Step S14: performing basin hydrological monitoring processing on the basin information data based on the basin remote sensing data to obtain basin hydrological monitoring data;

[0105] Step S15: Perform data classification processing on the watershed hydrological monitoring data to obtain watershed classified hydrological data.

[0106] As an embodiment of the present invention, refer to Figure 2 As shown, Figure 1 Detailed step flow diagram of step S1 in FIG. 1 , in this embodiment, step S1 includes the following steps:

[0107] Step S11: Acquire watershed information of the watershed to obtain watershed information data;

[0108] The embodiment of the present invention collects various data and information related to the watershed. It is necessary to obtain the physical geographic information of the watershed, including the location, area, topography, and hydrological elements of the watershed. The above information data can be obtained through on-site investigation, remote sensing image analysis, existing data collection, etc.

[0109] The embodiments of the present invention utilize remote sensing data, such as satellite images and aerial photographs, to perform remote sensing processing on the watershed, extract watershed characteristics, identify the land use type, water body distribution, vegetation conditions, etc. of the watershed through classification technology, and extract terrain elements such as the surface height and slope of the watershed in combination with a digital elevation model.

[0110] Step S12: performing remote sensing processing on the watershed to obtain watershed remote sensing data;

[0111] The embodiment of the present invention performs remote sensing processing on the watershed and uses remote sensing technology to obtain image data of the watershed, including satellite images, aerial images, etc., to obtain spatial information of the watershed. Through remote sensing processing, large-scale, high-resolution watershed remote sensing data can be obtained.

[0112] Step S13: performing dry spectrum analysis on the watershed information data based on the watershed remote sensing data to obtain a watershed dry spectrum;

[0113] The embodiment of the present invention obtains spectral data of a watershed through remote sensing technology. The spectrum contains information on the response of an object to electromagnetic waves and can reflect the drought condition of the watershed. The band or spectral index for drought analysis is selected, and the spectral data is analyzed. For example, the spectral band is related to the spatial moisture content, and the spatial drought condition is obtained. The spectral characteristics of soil drought are analyzed to evaluate the dust content of the soil.

[0114] Step S14: performing basin hydrological monitoring processing on the basin information data based on the basin remote sensing data to obtain basin hydrological monitoring data;

[0115] The embodiment of the present invention monitors the hydrological conditions of a river basin through remote sensing data. The water body characteristics in the remote sensing data are combined with other data sources, such as meteorological data and water level monitoring data, to extract the area and shape information of the water body, and the trend of hydrological changes is calculated through the water level monitoring data.

[0116] Step S15: Perform data classification processing on the watershed hydrological monitoring data to obtain watershed classified hydrological data.

[0117] The embodiment of the present invention obtains basin classification hydrological data by classifying and sorting the monitoring data. This step can divide the hydrological monitoring data according to different characteristics, and can be divided into rainfall, evapotranspiration, and runoff data categories according to the data type.

[0118] The present invention acquires basin information from the basin, and can establish a comprehensive database of the basin by collecting and analyzing basin-related data, including geographical, climate, and soil information. It provides an information basis for subsequent analysis and modeling, and provides basic data support for subsequent basin analysis. The application of remote sensing technology provides data for basin research. Through the processing and analysis of remote sensing images, the remote sensing-based data acquisition method has a wide coverage and high efficiency, and can accurately process the acquired data to provide remote sensing images of the basin, help identify and analyze basin data, and provide more comprehensive data support for subsequent analysis. Based on the basin remote sensing data, dry spectral analysis is performed on the basin information data to reveal the water dynamics of the basin. Spectral analysis of remote sensing data is performed to determine the water content and distribution in the basin. It provides a new analysis angle for the MCDI drought index, and can also help with the subsequent acquisition and definition of drought parameters. Through the comprehensive analysis of remote sensing data and watershed information, the hydrological dynamic data of the basin is monitored, including changes in rainfall, water flow, and water level. By classifying the basin hydrological monitoring data to identify different hydrological characteristics, it helps to better understand the hydrological changes in the basin and provide more accurate data support for basin drought analysis.

[0119] Preferably, step S13 comprises the following steps:

[0120] Step S131: performing dust data and dryness analysis on the watershed information data based on the watershed remote sensing data to obtain the watershed dust data and watershed dryness;

[0121] Step S132: performing spectral analysis on the watershed dust data based on the watershed remote sensing data to obtain the watershed dust spectrum;

[0122] Step S133: performing spatial ratio analysis on the watershed dust data to obtain watershed dust spatial ratio data;

[0123] Step S134: mapping the horizontal and vertical coordinates of the coordinate axis of the watershed dryness based on the watershed dust space ratio data to obtain dust dryness related coordinate data;

[0124] Step S135: constructing a correlation curve for the dust drying correlation coordinate data to obtain a dust drying correlation curve;

[0125] Step S136: performing association rule analysis on the dust drying association curve to obtain dust drying association rules;

[0126] Step S137: Perform data replacement on the watershed dust spectrum based on the dust drying association rule to obtain the watershed drying spectrum.

[0127] As an embodiment of the present invention, refer to Figure 3 As shown, Figure 2 Detailed step flow diagram of step S13 in the embodiment, step S13 includes the following steps:

[0128] Step S131: performing dust data and dryness analysis on the watershed information data based on the watershed remote sensing data to obtain the watershed dust data and watershed dryness;

[0129] The embodiment of the present invention obtains dust data and drought information by analyzing the remote sensing data of the watershed. Remote sensing data can reveal the surface characteristics, and through processing and analysis, dust content and its distribution information can be extracted. Spectral information in remote sensing data, such as vegetation index and water body characteristics, can be used to assess the dryness of the watershed.

[0130] Step S132: performing spectral analysis on the watershed dust data based on the watershed remote sensing data to obtain the watershed dust spectrum;

[0131] The embodiment of the present invention performs spectral analysis on the dust data of the watershed. The size, composition and concentration of dust particles will affect the absorption and reflection of electromagnetic waves, making them present unique spectral characteristics in remote sensing images. By analyzing the spectral characteristics of dust in remote sensing data, the dust information in the watershed can be identified and quantified.

[0132] Step S133: performing spatial ratio analysis on the watershed dust data to obtain watershed dust spatial ratio data;

[0133] The embodiment of the present invention calculates the proportion of the area covered by dust in the entire watershed to understand the distribution range and density of dust. The area covered by dust is extracted and its area is calculated through spatial analysis techniques such as image overlay and pixel statistics. This step provides quantitative information on the spatial distribution of dust in the watershed.

[0134] Step S134: mapping the horizontal and vertical coordinates of the coordinate axis of the watershed dryness based on the watershed dust space ratio data to obtain dust dryness related coordinate data;

[0135] The embodiment of the present invention adopts the coordinate axis horizontal and vertical mapping technology, takes the dryness data as the horizontal coordinate and the dust space ratio data as the vertical coordinate, so as to obtain data points on a two-dimensional coordinate plane. The data points reflect the relationship between the dryness and the dust ratio.

[0136] Step S135: constructing a correlation curve for the dust drying correlation coordinate data to obtain a dust drying correlation curve;

[0137] The embodiment of the present invention reflects the dust space ratio under different dry and wet conditions through data points. The data points are fitted using a linear function to obtain one or more correlation curves. This curve reveals the changing trend of the dust space ratio when the basin is in different degrees of drought.

[0138] Step S136: performing association rule analysis on the dust drying association curve to obtain dust drying association rules;

[0139] The embodiment of the present invention analyzes the slope, shape and data distribution of the curve to infer the changing pattern of the dust space ratio when the basin is in different drought levels. For example, the curve shows an upward trend, indicating that the dust space ratio increases as the drought level increases.

[0140] Step S137: Perform data replacement on the watershed dust spectrum based on the dust drying association rule to obtain the watershed drying spectrum.

[0141] The embodiment of the present invention replaces the original dust spectrum data by analyzing the obtained association rules, so that the dust spectrum data better reflects the dryness. For example, the association rules show that when the degree of drought is high, the dust space ratio is high, and the dust spectrum data is adjusted to increase the weight of the high-spectral drought feature.

[0142] The present invention obtains the dust data and dryness of the watershed by analyzing the dust data and dryness of the watershed information data based on the watershed remote sensing data. By analyzing the dust data and dryness, the dust distribution and dryness in the watershed can be more comprehensively understood. By performing spectral analysis on the watershed dust data, different types of dust components and distribution can be identified by spectral analysis of the dust data. It is helpful to provide basic spectral information for subsequent dry spectrum construction. Spatial ratio analysis can reveal the law of dust distribution in the watershed. Through spatial ratio analysis, the distribution ratio and density of dust in the watershed can be understood, which is helpful for subsequent correlation analysis based on dust ratio and dryness, and provides data support for analysis. Coordinate axis mapping processing helps to visualize the relationship between dust and dryness. By mapping the watershed dryness data and the dust space ratio data into a coordinate system, the relationship between the two can be intuitively expressed. Coordinate axis mapping processing helps to visualize the relationship between dust and dryness. By mapping the watershed dryness data and the dust space ratio data into a coordinate system, the relationship between the two can be intuitively expressed. The construction of the correlation curve enables the quantification of the relationship between dust and dryness. By drawing the dust-dryness correlation curve, the corresponding dryness changes under different dust levels can be determined, which is conducive to the subsequent conversion between spectral data. The dust-dryness correlation curve is analyzed by association rules to determine the rules between dust concentration changes and dryness changes, providing a reference for dust and dryness changes. By applying the dust-dryness association rules to the dust spectrum of the basin, the spectral data is modified and improved to better reflect the drought characteristics of the basin.

[0143] Preferably, step S132 includes the following steps:

[0144] Step 1321: performing dust particle data acquisition processing on the watershed dust data based on the watershed remote sensing data to obtain the watershed dust particle data;

[0145] Step 1322: performing wavelength analysis on the dust particle data in the watershed to obtain wavelength data of dust particles in the watershed;

[0146] Step S1323: Perform spatial concentration analysis on the dust particle data in the watershed to obtain spatial concentration data of dust in the watershed;

[0147] Step S1324: extracting the scattering characteristic peaks of the watershed remote sensing data according to the wavelength data of the watershed dust particles to obtain the watershed dust peak characteristic data;

[0148] Step S1325: Perform spectrum construction processing on the watershed dust spatial concentration data based on the watershed dust peak characteristic data to obtain the watershed dust spectrum.

[0149] As an embodiment of the present invention, refer to Figure 4 As shown, Figure 3 FIG. 1 is a schematic diagram of a detailed step flow chart of step S132 in FIG. 1 . In this embodiment, step S132 includes the following steps:

[0150] Step 1321: performing dust particle data acquisition processing on the watershed dust data based on the watershed remote sensing data to obtain the watershed dust particle data;

[0151] The embodiment of the present invention obtains watershed dust data by using remote sensing technology, including using optical remote sensing instruments. Extract dust particle information. The processing process includes correction, enhancement, filtering, etc., using algorithms and models to identify and separate dust particles from remote sensing images, and obtain information such as the distribution, size and concentration of dust particles.

[0152] Step 1322: performing wavelength analysis on the dust particle data in the watershed to obtain wavelength data of dust particles in the watershed;

[0153] The embodiment of the present invention analyzes dust samples through a spectrometer and other equipment to obtain information about the wavelengths of light reflected, absorbed and scattered by the dust particles. The behavior of dust particles at different wavelengths, including ultraviolet, visible light and infrared bands, is analyzed. Through wavelength analysis, the composition and structure of dust particles and the characteristics of their interaction with light can be determined.

[0154] Step S1323: Perform spatial concentration analysis on the dust particle data in the watershed to obtain spatial concentration data of dust in the watershed;

[0155] The embodiment of the present invention creates a dust particle concentration map to display the dust concentration in different areas of the watershed. The distribution pattern, concentration gradient and possible accumulation area of ​​dust particles are analyzed through spatial statistical technology and modeling. The migration and deposition pattern of dust in the watershed is understood, and the dust impact area is determined.

[0156] Step S1324: extracting the scattering characteristic peaks of the watershed remote sensing data according to the wavelength data of the watershed dust particles to obtain the watershed dust peak characteristic data;

[0157] The embodiments of the present invention use the unique scattering characteristics of dust particles to produce obvious peaks or valleys at specific wavelengths. These scattering feature peaks are identified and extracted through advanced image processing technology and algorithms. The peak extraction process includes background removal, enhancement, and edge detection to accurately obtain the scattering characteristics.

[0158] Step S1325: Perform spectrum construction processing on the watershed dust spatial concentration data based on the watershed dust peak characteristic data to obtain the watershed dust spectrum.

[0159] The embodiment of the present invention uses dust peak characteristic data and combines the results of spatial concentration analysis to construct a dust spectrum in the watershed. Spectral construction includes quantifying and modeling the optical properties of reflectivity, absorptivity and scattering rate of dust particles at different wavelengths. Through spectral modeling technology, the behavior of dust particles in the atmosphere can be simulated and a unique dust spectrum can be constructed.

[0160] The present invention processes remote sensing data to extract information about dust particles in the watershed, including size, shape and concentration data. It provides basic data for subsequent dust analysis and modeling, and helps to more comprehensively understand the dynamic state of dust in the watershed. By analyzing the data of dust particles in the watershed, the reflection, absorption and scattering behavior of dust at different wavelengths can be determined. It helps to distinguish dust particles in space from other particles in space. Spatial concentration analysis is performed on the data of dust particles in the watershed to obtain spatial concentration data of dust in the watershed. By analyzing the spatial variation of dust particle data in the watershed, the hot spot area of ​​dust concentration can be determined. It helps to acquire data for the impact analysis of dust concentration in the watershed. Peak extraction processing can identify the unique characteristics of dust in the spectrum. By analyzing the wavelength data of dust particles in the watershed, the obvious peaks generated by dust scattering and absorption can be detected. It helps to distinguish the difference between the dust spectrum and the spectrum of other substances, and can improve the accuracy of subsequent spectral analysis and modeling. By combining the dust peak characteristic data in the watershed with the dust spatial concentration data to construct spectral information, the unique optical properties of the dust in the watershed are represented, providing a tool for dust monitoring and analysis of dryness.

[0161] Preferably, step S2 comprises the following steps:

[0162] Step S21: performing correlation scatter plot processing on the watershed classification hydrological data according to the dry spectrum to obtain a dry hydrological correlation scatter plot;

[0163] The embodiment of the present invention analyzes the two variables by preprocessing the watershed classification hydrological data and the dry spectral data, and draws a scatter plot, in which each point represents a specific time period or sample, the dry spectral value is on the x-axis, and the corresponding hydrological data value is on the y-axis.

[0164] Step S22: performing correlation coefficient analysis on the dry spectrum and the watershed classification hydrological data based on the dry hydrological correlation scatter plot to obtain a dry hydrological correlation coefficient;

[0165] The embodiment of the present invention can quantify the linear relationship between different variables by calculating the data points in the scatter plot. A correlation coefficient close to 1 indicates a strong positive correlation, close to -1 indicates a strong negative correlation, and close to 0 indicates a weak or non-existent correlation, thereby obtaining a dry hydrological correlation coefficient.

[0166] Step S23: constructing a matrix of the dry hydrological correlation coefficient, the dry spectrum and the watershed classification hydrological data to obtain a dry correlation hydrological matrix;

[0167] The embodiment of the present invention combines the dry hydrological correlation coefficient, the dry spectral value and the watershed classification hydrological data. The rows of the matrix may represent different time periods or samples, and the columns contain the dry spectral values, the hydrological data values ​​and the correlation coefficients between them. This matrix is ​​called the dry correlation hydrological matrix.

[0168] Step S24: constructing an influence chain for the basin classification hydrological data based on the dry-related hydrological matrix to obtain a dry spectral hydrological influence chain;

[0169] The embodiment of the present invention establishes a dry spectrum hydrological impact chain by analyzing the association relationship in the matrix. This can be accomplished by using path analysis and structural equation modeling causal inference techniques. It establishes how dry spectrum affects hydrological conditions and how these impacts propagate to downstream variables along the causal chain.

[0170] Step S25: extracting hydrological data from the classified hydrological data of the watershed to obtain the watershed precipitation, watershed runoff and watershed evaporation;

[0171] The embodiment of the present invention collects basin precipitation data, including precipitation measurement, through meteorological stations, satellite remote sensing or meteorological models. To calculate the runoff of the basin, it is necessary to measure the flow of rivers flowing into and out of the basin. The evaporation of the basin is analyzed, including soil evaporation and plant transpiration.

[0172] Step S26: Based on the dry spectral hydrological impact chain, drought parameter analysis is performed on the basin precipitation, basin runoff and basin evaporation to obtain meteorological drought parameters, hydrological drought parameters and agricultural drought parameters.

[0173] The present invention uses the previously established dryness spectrum hydrological impact chain as the analysis framework. For meteorological drought parameters, precipitation data is analyzed, the dry day sequence is identified, and the correlation between precipitation and dryness spectrum is evaluated. For hydrological drought parameters, the basin runoff data and their association with the dryness spectrum are examined. For agricultural drought parameters, basin evaporation indicators are considered.

[0174] The present invention draws a scatter plot to visually observe the relationship and distribution between two sets of data. It is helpful to evaluate the strength and direction of the association between the dry spectrum and the hydrological characteristics of the basin, and provides a basis for subsequent quantitative analysis. The relationship number analysis quantifies the relationship between the dry spectrum and the basin classification hydrological data. The statistical correlation between the two variables is determined by calculating the correlation coefficient. It is helpful to understand how strong the linear relationship is between the dry spectrum and the basin hydrological characteristics. By creating a dry-related hydrological matrix, the dry spectrum, the dry hydrological correlation coefficient and the basin classification hydrological data are simultaneously visualized and analyzed, providing a data structure basis for subsequent comprehensive analysis. By analyzing the dry-related hydrological matrix, it is possible to identify how spectral data affects hydrological parameters and determine the cascade effect between hydrological parameters. This influence chain is conducive to understanding the propagation and impact of drought conditions in the basin. Precipitation, runoff and evaporation are extracted from the basin classification hydrological data to provide basic data support for subsequent drought parameter analysis. The dry spectrum hydrological impact chain is applied to analyze precipitation, runoff and evaporation data to obtain meteorological, hydrological and agricultural drought parameters. These parameters can quantify the severity and impact of drought and facilitate the impact analysis of drought on different aspects of a particular watershed.

[0175] Preferably, step S3 comprises the following steps:

[0176] Step S31: constructing a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters to obtain a drought neural network model;

[0177] The embodiment of the present invention integrates meteorological drought parameters, hydrological drought parameters and agricultural drought parameters into a neural network architecture. Provide historical drought parameter data and corresponding drought conditions to train the model. The network may include multiple hidden layers and nodes to capture the complex dynamics of drought. By repeatedly adjusting the model weights and biases.

[0178] Step S32: performing data injection processing on the drought neural network model based on the dry spectrum to obtain dry spectrum injection data;

[0179] The embodiment of the present invention enhances the understanding of drought process by using dry spectral data. In the injection stage, the dry spectral data is combined with the constructed drought neural network model and introduced into the neural network as the input layer. These spectral values ​​represent the dryness of the land surface.

[0180] Step S33: performing drought parameterization processing on the dryness spectrum injection data to obtain dryness-related drought data;

[0181] The embodiment of the present invention obtains the dryness of the watershed by analyzing the dryness of the dry spectrum injected into the data. Based on the dryness of the watershed, the watershed classification hydrological data is advanced to propose relevant precipitation, runoff and evaporation, and the meteorological, hydrological and agricultural parameters that have an impact on the drought index are converted to obtain meteorological, hydrological and agricultural parameters.

[0182] Step S34: performing MCDI drought index analysis on drought-related drought data based on the drought neural network model to obtain the MCDI drought index.

[0183] The embodiment of the present invention calculates the MCDI index by using a drought neural network model and drought-related drought data. MCDI takes into account multiple variables, including precipitation, evapotranspiration, soil moisture, and remote sensing indicators. By weighting and combining these variables, MCDI provides an indicator for indicating the severity and impact of drought.

[0184] The present invention obtains a drought neural network model by constructing a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters. By constructing a neural network model, the influence of multiple drought parameters can be comprehensively considered, and the drought parameters can be preprocessed, which is helpful for the subsequent calculation of the drought index. Through data injection processing, dry spectral data is introduced into the neural network model to consider the influence of dry factors on drought and improve the accuracy and reliability of the model. By analyzing the dry spectral injection data, key features related to the severity and impact of drought are extracted. This parameterization process quantifies the information of the dry spectrum and converts it into drought parameters that can be directly interpreted. By applying the drought neural network model, we can analyze dryness-related drought data and obtain the MCDI drought index. This index combines multiple drought parameters and provides a unified scale to evaluate the severity of drought. The MCDI drought index can quantify the quarterly drought analysis of the basin, which is helpful for the analysis of drought and flood changes in the basin and provides data support.

[0185] Preferably, step S31 includes the following steps:

[0186] Step S311: performing dryness analysis on the drying spectrum to obtain the watershed dryness;

[0187] The embodiment of the present invention obtains drying spectrum data, including reflectance data of different wavelengths, removes noise from the drying spectrum data, and identifies and selects bands sensitive to drying based on the pre-processed spectrum data. This usually includes near infrared and short-wave infrared bands, and performs remote sensing monitoring on the drying sensitive bands to obtain dryness data of the corresponding bands.

[0188] Step S312: constructing a model input and output for the meteorological drought parameters, the hydrological drought parameters, and the agricultural drought parameters to obtain a model input layer and a model output layer;

[0189] The embodiment of the present invention provides information about the drought conditions in the study area by specifying meteorological drought parameters, hydrological drought parameters and agricultural drought parameters as input. The input layer consists of these parameters, each of which represents a feature. For the output, the expected response of the neural network is defined, such as the MCDI drought index.

[0190] Step S313: constructing an influence degree sequence of meteorological drought parameters, hydrological drought parameters and agricultural drought parameters based on the watershed dryness to obtain a drought parameter influence sequence;

[0191] The embodiment of the present invention determines the degree to which each drought parameter is affected by dryness based on the dryness analysis results of the watershed, establishes an influence sequence of drought parameters, checks the flow volume during the drought period in a specific watershed for hydrological drought parameters, performs characteristic analysis during the drought period in the watershed, constructs a drought characteristic and hydrological drought parameter model, analyzes the model, and establishes the degree to which the hydrological drought parameters affect drought.

[0192] Step S314: dividing the drought parameter impact sequence by impact ratio to obtain drought parameter impact ratio data;

[0193] The embodiment of the present invention analyzes the data of the impact sequence and allocates the impact ratio. For example, it is found that in a specific drought event, the reduction of precipitation leads to the impact of the meteorological drought parameter, while the reduction of river flow leads to the impact of the hydrological drought parameter. The contribution of each parameter to the drought condition is determined, and the impact ratio division is obtained to provide quantitative information.

[0194] Step S315: Based on the drought parameter impact ratio data, the meteorological drought parameters, the hydrological drought parameters and the agricultural drought parameters are processed by connection weight division to obtain the connection weight data between models.

[0195] The embodiment of the present invention assigns weights according to the strength of the correlation and causal relationship between parameters. A higher weight indicates a stronger connection between parameters. For example, precipitation is a strongly correlated factor affecting evaporation and soil moisture, so the precipitation-evaporation connection is assigned a higher weight. Basin runoff has only a minor impact on hydrological conditions, and the connection of basin runoff parameters is assigned a lower weight.

[0196] Step S316: constructing a drought function for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters based on the connection weight data between models to obtain the MCDI drought index function;

[0197] The embodiment of the present invention creates a mathematical formula by using the weight data and drought parameters of the model connection. The Copula function combines multiple parameters to propose a new drought index. The function may involve a weighted combination of multiple parameters, threshold application and logical operations. For example, the function includes precipitation, flow and evaporation.

[0198] Step S317: constructing a hidden layer for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters based on the MCDI drought index function to obtain a hidden layer of the MCDI drought index function;

[0199] The embodiment of the present invention identifies drought patterns by constructing hidden layers for meteorological drought parameters, hydrological drought parameters, and agricultural drought parameters. Each neuron performs mathematical operations based on the MCDI function and other inputs. Through these calculations, the hidden layer can learn and explain the potential relationships between different drought parameters.

[0200] Step S318: Based on the connection weight data between models, the model input layer, the MCDI drought index function hidden layer and the model output layer are merged and constructed to obtain a drought neural network model.

[0201] The embodiment of the present invention combines and constructs the model input layer, the MCDI drought index function hidden layer and the model output layer through the inter-model connection weight data, aggregates the components to form a neural network model, and defines the relationship and strength between parameters by applying the inter-model connection weight data.

[0202] The present invention can accurately evaluate the dryness of the basin through dryness analysis, and provide an important basis for subsequent drought parameter analysis and prediction. By defining the model input layer and output layer, the model is specified to process specific drought parameters and expect to obtain output. It is ensured that the model focuses on the drought analysis task and processes these specific parameters. By analyzing the influence sequence of basin dryness on different drought parameters, the dependence between parameters is determined. It is helpful to understand how to analyze drought under different drought parameters. By quantifying the influence of dryness on each drought parameter, it helps to identify key drought parameters and improve the accuracy of drought prediction. By dividing the influence ratio, the relative importance of each drought parameter in drought prediction can be clarified, providing a reference basis for the subsequent establishment of a drought neural network model. By dividing the connection weight, the correlation between each drought parameter can be determined, providing a basis for the connection weight of the drought neural network model. By constructing a drought function, the influence of each drought parameter can be comprehensively considered, and the MCDI drought index function can be established to provide a method for calculating and evaluating the drought index. Through the exponential function hidden layer, the processing capacity of the neural network model is increased. The hidden layer identifies complex data relationships, which is also conducive to enhancing the ability of the model to analyze the dynamic changes of drought. By constructing a drought neural network model, we can comprehensively consider the influence and correlation of various drought parameters, achieve a comprehensive assessment and prediction of drought conditions, and provide scientific support for drought monitoring.

[0203] Preferably, step S4 comprises the following steps:

[0204] Step S41: constructing a drought data distribution lattice for the MCDI drought index to obtain a MCDI drought index distribution lattice;

[0205] The embodiment of the present invention obtains the MCDI drought index from a data source to preprocess the data, converts the data into a dot matrix format according to the spatial resolution and range of the drought index preprocessing, divides the matrix into grid units, maps the data to the corresponding grid units, and processes the data using a spatial interpolation method to obtain a drought index distribution dot matrix.

[0206] Step S42: performing normal analysis of drought data on the basin based on the MCDI drought index distribution matrix to obtain a normal distribution curve of the MCDI drought index;

[0207] The embodiment of the present invention extracts the MCDI drought index of the grid cells in the watershed, calculates the mean and standard deviation of the MCDI drought index, describes the central position and dispersion of the data, performs a normality test, determines whether the data conforms to the normal distribution, and then fits the MCDI drought index to a normal distribution curve.

[0208] Step S43: performing drought interval division processing on the MCDI drought index based on the MCDI drought index normal distribution curve to obtain the watershed drought interval;

[0209] The embodiment of the present invention determines the threshold of the drought index according to the mean and standard deviation of the normal distribution curve, and the MCDI drought index is classified according to the threshold, and is defined according to a specific percentile of the classified data. The data is divided into different drought intervals based on the percentile method.

[0210] Step S44: performing drought assessment processing on the drought interval of the watershed to obtain watershed drought assessment data;

[0211] The embodiment of the present invention requires drought indicators for drought assessment of drought intervals, and these indicators can reflect the impact of drought, conduct a comprehensive analysis of different indicators, unify the drought indicators to convert influencing parameters, classify based on the unified influencing parameters, and standardize and assess drought index data based on the unified influencing parameters.

[0212] Step S45: Perform drought analysis on the watershed according to the watershed drought assessment data to obtain watershed drought analysis data.

[0213] The embodiment of the present invention uses basin drought assessment data to analyze the drought conditions in different areas or time periods within the basin, analyzes the relationship between the drought index and meteorological elements and hydrological variables, obtains the changing trend and periodicity of the drought index and the law of drought events, and performs drought analysis on the entire basin based on the changing trend and periodicity of the drought index and the law of drought events to obtain basin drought analysis data.

[0214] The present invention can intuitively display the spatial distribution of the MCDI drought index in the basin by constructing a drought index distribution dot matrix, help identify the spatial differences in the degree of drought, and provide a basis for subsequent analysis and interpretation. Through normal analysis, it is possible to understand whether the distribution characteristics of the MCDI drought index conform to the normal distribution, and provide a statistical basis for data processing and analysis. By analyzing the normal distribution curve of the MCDY drought index, different drought level intervals are defined. It is helpful to standardize drought assessment and provide a unified scale for quantifying the severity of drought. Through drought assessment processing, the drought situation in the basin can be assessed according to the division results of the drought intervals, and the assessment is conducive to more quickly defining the drought and flood situation in the basin. By analyzing the basin drought assessment data, the drought situation in the basin can be deeply understood, which is conducive to the prevention of drought and flood disasters in the basin area.

[0215] Preferably, step S43 includes the following steps:

[0216] Step S431: constructing a characteristic curve for the MCDI drought index to obtain a drought characteristic curve;

[0217] The embodiment of the present invention extracts the time series data of the MCDI drought index, constructs a curve with the time series data of the MCDI drought index as the horizontal axis and the MCDI drought index as the vertical axis, performs trend and feature analysis on the curve, and obtains the development and periodicity characteristics of the MCDI drought index.

[0218] Step S432: performing quantile threshold division processing on the drought characteristic curve based on the MCDI drought index to obtain the curve threshold quantile;

[0219] The embodiment of the present invention analyzes the curve distribution and defines quantiles representing different drought levels. By selecting quantiles, such as the 25th percentile, the 50th percentile, and the 75th percentile, a drought severity level is established. The quantile thresholds serve as reference points to classify and quantify drought conditions.

[0220] Step S433: performing curve fitting processing on the MCDI drought index normal distribution curve according to the curve threshold quantile points to obtain the normal threshold quantile points;

[0221] The embodiment of the present invention analyzes by identifying the threshold quantiles of the curve and uses the normal distribution to fit these data points. The flexibility of the normal distribution is used to describe the changes in the MCDI drought index. By using a statistical algorithm to determine the best fitting normal curve, and mapping the threshold quantiles of the curve, the threshold quantiles of the curve on the normal curve are obtained.

[0222] Step S434: sorting the normal threshold points in ascending order to obtain an ascending sequence of normal threshold points;

[0223] The embodiment of the present invention creates a continuous sequence by sorting these threshold quantiles in ascending order. The sequence represents the order of drought levels from mild to severe. By sorting the thresholds, a unified scale is established, allowing a consistent interpretation of the MCDI drought index, and the sorted sequence provides a clear reference point.

[0224] Step S435: construct drought intervals for the MCDI drought index based on the normal threshold point ascending sequence to obtain the watershed drought intervals.

[0225] The embodiment of the present invention sorts the MCDI drought index data to construct an ascending sequence, determines the threshold points on the normal distribution curve according to the characteristics of the normal distribution, and uses them to divide the drought intervals. The drought index data is classified according to the threshold points to form different drought intervals. The threshold points are adjusted in combination with the actual watershed drought data to obtain the drought intervals for dividing the watershed.

[0226] By constructing a characteristic curve, the present invention can intuitively display the changing trend and characteristics of the MCDI drought index, help identify the laws and characteristics of drought development, and provide an important reference for drought characteristic analysis. By dividing the exponential distribution into different quantile thresholds, the drought grade interval is defined. It is helpful to standardize drought assessment and provide a unified scale for quantifying the severity of drought. By performing curve fitting on the exponential distribution and considering the quantile threshold, the quantile of the normal distribution curve is obtained. It helps to understand the regularity and statistical characteristics of the drought index distribution and provides a basis for subsequent statistical inference intervals. Through ascending sorting processing, the normal threshold points can be arranged in order of size to help identify the severity and distribution law of drought. Through the construction of drought intervals, the MCDI drought index can be divided into different drought intervals according to the normal threshold points, providing a data basis for determining the drought grade.

[0227] Preferably, the present invention further provides a basin seasonal drought analysis system based on the MCDI index, which is used to execute the basin seasonal drought analysis method based on the MCDI index as described above. The basin seasonal drought analysis system based on the MCDI index comprises:

[0228] The watershed information acquisition module is used to acquire watershed information in the watershed to obtain watershed information data; perform dry spectrum analysis on the watershed information data to obtain watershed dry spectrum; perform hydrological classification analysis on the watershed to obtain watershed classification hydrological data;

[0229] The watershed drought parameter analysis module is used to construct a correlation matrix of the watershed classification hydrological data based on the dry spectrum to obtain the dryness-related hydrological matrix; based on the dryness-related hydrological matrix, the watershed classification hydrological data is analyzed for drought parameters to obtain meteorological drought parameters, hydrological drought parameters and agricultural drought parameters;

[0230] The neural network model MCDI drought index analysis module is used to construct a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters to obtain a drought neural network model; the MCDI drought index analysis is performed on the drought neural network model based on the dry spectrum to obtain the MCDI drought index;

[0231] The watershed drought analysis module is used to divide the watershed into intervals based on the MCDI drought index to obtain the watershed drought assessment interval; based on the watershed drought assessment interval, drought analysis is performed on the watershed to obtain watershed drought analysis data.

[0232] In summary, the present invention provides a watershed quarterly drought analysis system based on the MCDI index. The watershed quarterly drought analysis system based on the MCDI index consists of a watershed information acquisition module, a watershed drought parameter analysis module, a neural network model MCDI drought index analysis module, and a watershed drought analysis module. It can realize any watershed quarterly drought analysis method based on the MCDI index described in the present invention, and is used to combine the operations between computer programs running on each module to realize any watershed quarterly drought analysis method based on the MCDI index. The internal structures of the system cooperate with each other, which can greatly reduce duplication of work and manpower investment, and can quickly and effectively provide a more accurate and efficient drought analysis process, thereby simplifying the operational process of watershed quarterly drought analysis based on the MCDI index.

[0233] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A basin seasonal drought analysis method based on MCDI index, characterized in that: The following steps are involved: Step S1: Acquire watershed information of the watershed to obtain watershed information data; Perform dry spectrum analysis on the watershed information data to obtain the watershed dry spectrum; Carry out hydrological classification analysis on the watershed to obtain the watershed classification hydrological data; Step S2: constructing a correlation matrix for the classified hydrological data of the watershed based on the dry spectrum to obtain a dry correlation hydrological matrix; performing drought parameter analysis on the classified hydrological data of the watershed based on the dry correlation hydrological matrix to obtain meteorological drought parameters, hydrological drought parameters and agricultural drought parameters, wherein step S2 is specifically as follows: Step S21: performing correlation scatter plot processing on the watershed classification hydrological data according to the dry spectrum to obtain a dry hydrological correlation scatter plot; Step S22: performing correlation coefficient analysis on the dry spectrum and the watershed classification hydrological data based on the dry hydrological correlation scatter plot to obtain a dry hydrological correlation coefficient; Step S23: constructing a matrix of the dry hydrological correlation coefficient, the dry spectrum and the watershed classification hydrological data to obtain a dry correlation hydrological matrix; Step S24: constructing an influence chain for the basin classification hydrological data based on the dry-related hydrological matrix to obtain a dry spectral hydrological influence chain; Step S25: extracting hydrological data from the classified hydrological data of the watershed to obtain the watershed precipitation, watershed runoff and watershed evaporation; Step S26: performing drought parameter analysis on the basin precipitation, basin runoff and basin evaporation based on the dry spectrum hydrological impact chain to obtain meteorological drought parameters, hydrological drought parameters and agricultural drought parameters; Step S3: constructing a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters to obtain a drought neural network model; performing MCDI drought index analysis on the drought neural network model based on the dry spectrum to obtain the MCDI drought index, wherein step S3 is specifically as follows: Step S31: constructing a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters to obtain a drought neural network model, wherein step S31 is specifically as follows: Step S311: performing dryness analysis on the drying spectrum to obtain the watershed dryness; Step S312: constructing a model input and output for the meteorological drought parameters, the hydrological drought parameters, and the agricultural drought parameters to obtain a model input layer and a model output layer; Step S313: constructing an influence degree sequence of meteorological drought parameters, hydrological drought parameters and agricultural drought parameters based on the watershed dryness to obtain a drought parameter influence sequence; Step S314: dividing the drought parameter impact sequence by impact ratio to obtain drought parameter impact ratio data; Step S315: performing connection weight division processing on meteorological drought parameters, hydrological drought parameters and agricultural drought parameters based on drought parameter impact ratio data to obtain connection weight data between models; Step S316: constructing a drought function for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters based on the connection weight data between models to obtain the MCDI drought index function; Step S317: constructing a hidden layer for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters based on the MCDI drought index function to obtain a hidden layer of the MCDI drought index function; Step S318: Based on the connection weight data between models, the model input layer, the MCDI drought index function hidden layer and the model output layer are merged and constructed to obtain a drought neural network model; Step S32: performing data injection processing on the drought neural network model based on the dry spectrum to obtain dry spectrum injection data; Step S33: performing drought parameterization processing on the dryness spectrum injection data to obtain dryness-related drought data; Step S34: performing MCDI drought index analysis on drought-related drought data based on the drought neural network model to obtain the MCDI drought index; Step S4: divide the watershed into intervals according to the MCDI drought index to obtain watershed drought assessment intervals; perform drought analysis on the watershed based on the watershed drought assessment intervals to obtain watershed drought analysis data.

2. The method for analyzing seasonal drought in a watershed based on the MCDI index according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: Acquire watershed information of the watershed to obtain watershed information data; Step S12: performing remote sensing processing on the watershed to obtain watershed remote sensing data; Step S13: performing dry spectrum analysis on the watershed information data based on the watershed remote sensing data to obtain a watershed dry spectrum; Step S14: performing basin hydrological monitoring processing on the basin information data based on the basin remote sensing data to obtain basin hydrological monitoring data; Step S15: Perform data classification processing on the watershed hydrological monitoring data to obtain watershed classified hydrological data.

3. The basin seasonal drought analysis method based on MCDI index according to claim 2 is characterized in that: Step S13 includes the following steps: Step S131: performing dust data and dryness analysis on the watershed information data based on the watershed remote sensing data to obtain the watershed dust data and watershed dryness; Step S132: performing spectral analysis on the dust data in the watershed based on the watershed remote sensing data to obtain a dust spectrum in the watershed; Step S133: performing spatial ratio analysis on the watershed dust data to obtain watershed dust spatial ratio data; Step S134: mapping the horizontal and vertical coordinates of the coordinate axis of the watershed dryness based on the watershed dust space ratio data to obtain dust dryness related coordinate data; Step S135: constructing a correlation curve for the dust drying correlation coordinate data to obtain a dust drying correlation curve; Step S136: performing association rule analysis on the dust drying association curve to obtain dust drying association rules; Step S137: Perform data replacement on the watershed dust spectrum based on the dust drying association rule to obtain the watershed drying spectrum.

4. The method for analyzing seasonal drought in a watershed based on the MCDI index according to claim 3, characterized in that: Step S132 includes the following steps: Step 1321: performing dust particle data acquisition processing on the watershed dust data based on the watershed remote sensing data to obtain the watershed dust particle data; Step 1322: performing wavelength analysis on the dust particle data in the watershed to obtain wavelength data of dust particles in the watershed; Step S1323: Perform spatial concentration analysis on the dust particle data in the watershed to obtain spatial concentration data of dust in the watershed; Step S1324: extracting the scattering characteristic peaks of the watershed remote sensing data according to the wavelength data of the watershed dust particles to obtain the watershed dust peak characteristic data; Step S1325: Perform spectrum construction processing on the watershed dust spatial concentration data based on the watershed dust peak characteristic data to obtain the watershed dust spectrum.

5. The method for analyzing seasonal drought in a watershed based on the MCDI index according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: constructing a drought data distribution matrix for the MCDI drought index to obtain a MCDI drought index distribution matrix; Step S42: performing normal analysis of drought data on the basin based on the MCDI drought index distribution matrix to obtain a normal distribution curve of the MCDI drought index; Step S43: performing drought interval division processing on the MCDI drought index based on the MCDI drought index normal distribution curve to obtain the watershed drought interval; Step S44: performing drought assessment processing on the drought interval of the watershed to obtain watershed drought assessment data; Step S45: Perform drought analysis on the watershed according to the watershed drought assessment data to obtain watershed drought analysis data.

6. The method for analyzing seasonal drought in a watershed based on the MCDI index according to claim 5, characterized in that: Step S43 includes the following steps: Step S431: constructing a characteristic curve for the MCDI drought index to obtain a drought characteristic curve; Step S432: performing quantile threshold division processing on the drought characteristic curve based on the MCDI drought index to obtain the curve threshold quantile; Step S433: performing curve fitting processing on the MCDI drought index normal distribution curve according to the curve threshold quantile points to obtain the normal threshold quantile points; Step S434: sorting the normal threshold points in ascending order to obtain an ascending sequence of normal threshold points; Step S435: construct drought intervals for the MCDI drought index based on the normal threshold point ascending sequence to obtain the watershed drought intervals.

7. A basin seasonal drought analysis system based on MCDI index, characterized in that: Used to execute the basin seasonal drought analysis method based on the MCDI index as claimed in claim 1, the basin seasonal drought analysis system based on the MCDI index comprises: The watershed information acquisition module is used to acquire watershed information in the watershed to obtain watershed information data; perform dry spectrum analysis on the watershed information data to obtain watershed dry spectrum; perform hydrological classification analysis on the watershed to obtain watershed classification hydrological data; The watershed drought parameter analysis module is used to construct a correlation matrix of the watershed classification hydrological data based on the dry spectrum to obtain the dryness-related hydrological matrix; based on the dryness-related hydrological matrix, the watershed classification hydrological data is analyzed for drought parameters to obtain meteorological drought parameters, hydrological drought parameters and agricultural drought parameters; The neural network model MCDI drought index analysis module is used to construct a neural network model for meteorological drought parameters, hydrological drought parameters and agricultural drought parameters to obtain a drought neural network model; the MCDI drought index analysis is performed on the drought neural network model based on the dry spectrum to obtain the MCDI drought index; The watershed drought analysis module is used to divide the watershed into intervals based on the MCDI drought index to obtain the watershed drought assessment interval; based on the watershed drought assessment interval, drought analysis is performed on the watershed to obtain watershed drought analysis data.

Images