A precipitation phase identification and multi-source data storage and retrieval system for alpine regions

Abstract

The present invention discloses a precipitation phase identification and multi-source data storage and retrieval system for high-altitude cold regions, which relates to the technical field of multi-source precipitation data storage and retrieval, and comprises a multi-source precipitation data acquisition module, a remote sensing positioning module, a precipitation type identification module, a precipitation intensity assessment module and a storage and retrieval module. The multi-source precipitation data acquisition module collects satellite remote sensing data and ground-observed precipitation data in high-altitude cold regions, pre-processes and standardizes the collected data, and performs feature extraction and analysis on the pre-processed precipitation data, thereby obtaining precipitation particle characteristics and precipitation amount characteristic parameters. The data acquisition technology, feature analysis technology, remote sensing positioning technology and model construction technology in the system of the present invention are closely integrated with modern information technology, thereby solving the problems that the existing technology is difficult to effectively integrate and analyze multi-source precipitation data, difficult to accurately distinguish different types of precipitation data, and difficult to accurately assess precipitation intensity and spatial distribution.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-source precipitation data storage and retrieval, and in particular to a precipitation phase recognition and multi-source data storage and retrieval system in alpine regions. Background Art

[0002] In alpine regions, precipitation conditions are complex and changeable. Accurately acquiring multi-source precipitation data is crucial for climate research, disaster warning, and water resources management. However, current technologies have many limitations. On the one hand, precipitation in alpine regions takes various forms, including rain and snow, and is unevenly distributed in time and space. Existing technologies have difficulty effectively analyzing multi-source precipitation data and cannot accurately distinguish different types of precipitation data, which poses a challenge to the accurate acquisition and management of precipitation data. On the other hand, the harsh environment in alpine regions makes data collection difficult. Existing collection and processing methods cannot obtain data comprehensively and accurately in complex environments. These problems seriously restrict in-depth research and effective response to precipitation in alpine regions, resulting in low precision of stored multi-source precipitation data, which cannot truly reflect the actual precipitation situation in alpine regions, leading to inaccurate subsequent retrieval results. Therefore, a new multi-source precipitation data storage and retrieval system for alpine regions is urgently needed to address the shortcomings of existing technologies, improve data quality and retrieval accuracy, and thus provide more powerful support for climate research, disaster warning, and water resources management in alpine regions.

[0003] While existing technologies have made significant progress in the storage and retrieval of multi-source precipitation data, some challenges remain. When analyzing and processing multi-source precipitation data, existing technologies struggle to accurately distinguish between different types of precipitation and adequately assess precipitation intensity and spatial location. This leads to errors in the stored data, which in turn affects the accuracy of retrieval results for multi-source precipitation data in high-altitude and cold regions. Therefore, there is an urgent need to innovate and optimize existing precipitation data storage and retrieval methods to address these technical bottlenecks and improve the reliability and retrieval accuracy of multi-source precipitation data. Summary of the Invention

[0004] To achieve the above objectives, the present invention is implemented through the following technical solutions: a precipitation phase identification and multi-source data storage and retrieval system in alpine regions, comprising a multi-source precipitation data acquisition module, a remote sensing positioning module, a precipitation type identification module, a precipitation intensity assessment module, and a storage and retrieval module, wherein each module is communicatively connected;

[0005] The multi-source precipitation data acquisition module collects satellite remote sensing data and ground observation precipitation data in alpine areas, preprocesses and standardizes the collected data, and performs feature extraction and analysis on the preprocessed precipitation data to obtain precipitation particle characteristics and precipitation amount characteristic parameters, providing data preparation for the construction of models in subsequent modules;

[0006] The remote sensing positioning module divides the alpine region into several sub-regions using pre-processed satellite remote sensing data and obtains positioning data for each sub-region;

[0007] The precipitation type identification module classifies the precipitation type corresponding to the precipitation particle data based on the characteristics of the precipitation particles, thereby obtaining the precipitation phase change index and updating the precipitation type corresponding to the precipitation particles. This solves the problem that the existing technology has difficulty in distinguishing the types of precipitation data from multiple sources.

[0008] The precipitation intensity assessment module obtains the solid precipitation intensity index and the liquid precipitation intensity index based on the precipitation characteristics, and then assesses the solid precipitation intensity and the liquid precipitation intensity, solving the problem that the existing technology is difficult to assess precipitation intensity;

[0009] The storage and retrieval module integrates the positioning data and precipitation intensity assessment results of each sub-area, and stores and retrieves them.

[0010] A further improvement of the technical solution of the present invention is that the multi-source precipitation data acquisition module collects satellite remote sensing data and ground observation precipitation data in alpine areas and obtains precipitation particle characteristics and precipitation amount characteristic parameters in the following process:

[0011] The target detection area in the alpine region is divided into several sub-areas of equal area based on the terrain, climate characteristics and precipitation distribution characteristics. Different types of collection equipment are deployed in each sub-area to collect high-precision ground precipitation data and satellite remote sensing data for each sub-area. The collection equipment includes microwave remote sensing equipment, total precipitation sensor, laser raindrop spectrometer and laser precipitation particle spectrometer with the same area, which are used to measure the precipitation intensity, precipitation phase, precipitation particle size distribution and precipitation particle motion characteristics in each sub-area respectively;

[0012] The satellite remote sensing data includes multispectral, microwave or radar remote sensing precipitation information, which is satellite remote sensing data for each sub-region and is used to invert precipitation type and intensity; the ground observation precipitation data includes precipitation particle data and hourly precipitation data in each sub-region;

[0013] Data cleaning and normalization are performed on precipitation data, outliers are removed, and missing data are filled. Geometric correction, atmospheric correction, and noise reduction are performed on satellite remote sensing data. Timestamps are added to satellite remote sensing data and precipitation data. Time interpolation and synchronization algorithms are used to synchronize the timestamps of satellite remote sensing data and precipitation data to keep the time series consistent.

[0014] Based on the hourly precipitation data in each sub-region, the total amount of precipitation in each sub-region within 24 hours and the number of precipitation events in each sub-region within 24 hours are counted, and the precipitation frequency and precipitation volume of each sub-region are further extracted. The precipitation particle data in each sub-region are combined with a laser precipitation particle spectrometer based on the principle of laser scattering to obtain the velocity and particle size of precipitation particles in each sub-region.

[0015] The precipitation frequency and precipitation volume of each sub-region are used as precipitation characteristics to form a precipitation feature vector. The velocity and particle size of precipitation particles in each sub-region are used as precipitation particle characteristics to form a precipitation particle feature vector. The precipitation feature vector and precipitation particle feature vector are integrated to construct a precipitation feature dataset, and finally the precipitation feature dataset is divided into a training set and a test set.

[0016] A further improvement of the technical solution of the present invention is that: the remote sensing positioning module, the process of acquiring positioning data of each sub-area includes:

[0017] Import the pre-processed satellite remote sensing data into GIS software and analyze the satellite remote sensing data of each sub-region through GIS software;

[0018] Select the geographic coordinate system of satellite remote sensing data and adjust the GIS software coordinate system to make it consistent with the geographic coordinate system of satellite remote sensing data;

[0019] Number each sub-region and arrange them in numerical order. Then, use the feature extraction tool of GIS software to divide the boundaries of each sub-region and read the location coordinates of each sub-region.

[0020] The positioning coordinates of each sub-region are integrated and matched with the corresponding labels to obtain a standardized positioning data table of the target detection area.

[0021] A further improvement of the technical solution of the present invention is that the process of the precipitation type identification module classifying the precipitation types corresponding to the precipitation particles based on the characteristics of the precipitation particles includes:

[0022] Establish a rectangular coordinate system, use the particle size of precipitation particles in each sub-area as the horizontal coordinate, and the velocity of precipitation particles in each sub-area as the vertical coordinate, and draw a particle size-velocity curve;

[0023] Select two representative points in the particle size-velocity curve and calculate the local slope value of the curve between the two selected points according to the differential slope calculation formula. The slope calculation formula of the curve between the two points is as follows:

[0024]

[0025] Where k is the slope of the curve between the two points, and are the precipitation particle velocities at two points in the particle size-velocity curve, and are the precipitation particle sizes of two points selected from the particle size-velocity curve;

[0026] According to the slope characteristics of the particle size-velocity relationship curve, when the slope of the curve between two points is less than 1, the precipitation type corresponding to the precipitation particles is classified as solid precipitation; when the slope of the curve between two points is greater than 1, the precipitation type corresponding to the precipitation particles is classified as liquid precipitation.

[0027] The precipitation type corresponding to the precipitation particles is encoded, the precipitation type code is converted into a precipitation type feature vector, and the vector is incorporated into the precipitation feature dataset to update the precipitation feature dataset.

[0028] A further improvement of the technical solution of the present invention is that: the precipitation type identification module, the process of obtaining the precipitation phase change index includes:

[0029] Weights are assigned to the velocity and size of precipitation particles in each sub-region, as well as the meteorological factors (temperature, humidity, and 0° layer height) of each sub-region. The process of calculating the precipitation phase change index using the assigned weights includes:

[0030]

[0031] in, is the precipitation phase change index, , , , and are the speed of precipitation particles in each sub-area, the particle size of precipitation particles in each sub-area, the temperature, humidity and the weight of the 0° layer height in each sub-area, , , 、 and are the velocity of precipitation particles in each sub-area, the particle size of precipitation particles in each sub-area, the temperature, humidity and 0° layer height of each sub-area;

[0032] Extract the precipitation particle feature vector and precipitation type feature vector from the precipitation feature dataset respectively;

[0033] Using the precipitation particle feature vectors and precipitation type feature vectors extracted from the training set, combined with a neural network algorithm, the precipitation particle feature vectors and precipitation type feature vectors extracted from the training set are used as input, and the precipitation phase change index is used as output. The nonlinear relationship between the precipitation particle feature vectors, precipitation type feature vectors and precipitation phase change index is learned to train the phase change model.

[0034] The precipitation particle feature vectors and precipitation type feature vectors extracted from the test set are input into the phase change model to evaluate the performance of the phase change model. The phase change model parameters are adjusted based on error feedback to optimize the model structure and obtain the final phase change model.

[0035] The precipitation particle feature vectors and precipitation type feature vectors in the precipitation feature dataset are input into the phase change model. The precipitation phase change index is output through the phase change model. The precipitation type feature vectors and their corresponding precipitation phase change indices are analyzed to determine the precipitation phase change index threshold in alpine regions.

[0036] A further improvement of the technical solution of the present invention is that: in the precipitation type identification module, the updating process of the precipitation type corresponding to the precipitation particles includes:

[0037] Analyze the numerical characteristics of the precipitation phase change index. When the index is lower than the threshold of the precipitation phase change index in high-altitude cold regions, the precipitation type is determined to be solid precipitation. When the index is higher than the threshold of the precipitation phase change index in high-altitude cold regions, the precipitation type is determined to be liquid precipitation.

[0038] Combined with the original precipitation type corresponding to the precipitation particles, when the precipitation type corresponding to the precipitation phase change index is the same as the precipitation type corresponding to the precipitation particles, the precipitation type corresponding to the precipitation particles is not updated; when the precipitation type corresponding to the precipitation phase change index is different from the precipitation type corresponding to the precipitation particles, the precipitation type corresponding to the precipitation particles is updated to the precipitation type corresponding to the precipitation phase change index;

[0039] Based on the precipitation type corresponding to the updated precipitation particles, the precipitation feature dataset is subdivided, and the precipitation feature vector is subdivided into solid precipitation feature vectors and liquid precipitation feature vectors. The solid precipitation feature vector includes the solid precipitation frequency and solid precipitation volume of each subregion, and the liquid precipitation feature vector includes the liquid precipitation frequency and liquid precipitation volume of each subregion.

[0040] A further improvement of the technical solution of the present invention is that: the precipitation intensity assessment module, the process of obtaining the solid precipitation intensity index includes:

[0041] Solid precipitation feature vectors are extracted from the precipitation feature dataset. The solid precipitation feature vectors in the training set are used in combination with a multivariate linear regression algorithm. The solid precipitation feature vectors extracted from the training set are used as input, and the solid precipitation intensity index is used as output. The linear relationship between the solid precipitation feature vectors and the solid precipitation intensity index is learned to train a solid precipitation intensity model.

[0042] The solid precipitation feature vectors in the test set are input into the solid precipitation intensity model to evaluate the performance of the solid precipitation intensity model. The intercept term and regression coefficient of the solid precipitation intensity model are adjusted to optimize the solid precipitation intensity model and obtain the final solid precipitation intensity model. The expression of the solid precipitation intensity model is:

[0043]

[0044] in, is the solid precipitation intensity index, is the intercept term, and are the regression coefficients of the solid precipitation frequency and the solid precipitation volume in each sub-region, respectively. is the solid precipitation frequency in each sub-region, is the volume of solid precipitation in each sub-region, is the error term;

[0045] The solid precipitation characteristic vector is input into the solid precipitation intensity model, and the solid precipitation intensity index is output through the solid precipitation intensity model.

[0046] A further improvement of the technical solution of the present invention is that: the precipitation intensity assessment module, the process of obtaining the liquid precipitation intensity index includes:

[0047] Liquid precipitation feature vectors are extracted from the precipitation feature dataset. By combining the liquid precipitation feature vectors in the training set with a multivariate linear regression algorithm, the liquid precipitation feature vectors extracted from the training set are used as input and the liquid precipitation intensity index is used as output. The linear relationship between the liquid precipitation feature vectors and the liquid precipitation intensity index is learned to train a liquid precipitation intensity model.

[0048] The liquid precipitation feature vectors in the test set are input into the liquid precipitation intensity model to evaluate the performance of the liquid precipitation intensity model, adjust the parameters of the liquid precipitation intensity model, optimize the liquid precipitation intensity model, and obtain the final liquid precipitation intensity model. The expression of the liquid precipitation intensity model is:

[0049]

[0050] Where Y is the liquid precipitation intensity index, is the intercept term, and are the regression coefficients of the liquid precipitation frequency and the liquid precipitation volume in each sub-region, respectively. is the liquid precipitation frequency of each sub-region, is the volume of liquid precipitation in each sub-region, is the error term;

[0051] The liquid precipitation characteristic vector is input into the liquid precipitation intensity model, and the liquid precipitation intensity index is output through the liquid precipitation intensity model. A further improvement of the technical solution of the present invention is that the precipitation intensity evaluation module, the solid precipitation intensity and liquid precipitation intensity evaluation process includes:

[0052] Based on the historical process of the alpine region, the assessment levels of the solid precipitation intensity index and the liquid precipitation intensity index were calculated according to each historical sequence using the percentile calculation method. The percentiles corresponding to the 50th, 80th, 95th and 98th percentiles were calculated respectively. These were used as thresholds to divide the solid precipitation intensity level and the liquid precipitation intensity level into five levels. The corresponding assessment results are: general (level 5), moderate (level 4), strong (level 3), strong (level 2), and very strong (level 1).

[0053] Combining the output results of the solid precipitation intensity model and the liquid precipitation intensity model, the solid precipitation intensity and liquid precipitation intensity are evaluated respectively, and finally the solid precipitation intensity evaluation results and liquid precipitation intensity evaluation results are obtained.

[0054] A further improvement of the technical solution of the present invention is that the storage and retrieval module integrates the positioning data and precipitation intensity assessment results of each sub-region and stores and retrieves them, including:

[0055] Integrate the liquid precipitation intensity assessment results and the solid precipitation intensity assessment results to generate a multi-source precipitation data retrieval report for high-altitude and cold regions;

[0056] Based on the MySQL database architecture, a precipitation data storage system is constructed, and the positioning data table of the target detection area and the multi-source precipitation data retrieval report of the high-altitude and cold regions are stored in the constructed precipitation data storage system;

[0057] The precipitation data storage system includes positioning data, precipitation type and precipitation intensity of the target detection area in the alpine region. By constructing the precipitation data storage system, accurate retrieval of multi-source precipitation data in the alpine region can be achieved.

[0058] The beneficial effects of the present invention are as follows: a precipitation phase identification and multi-source data storage and retrieval system in a high-altitude cold region in the present invention, compared with a traditional precipitation phase identification and multi-source data storage and retrieval system in a high-altitude cold region, the data acquisition technology, feature analysis technology, remote sensing positioning technology and model construction technology in the system of the present invention are closely integrated with modern information technology, accurately capturing precipitation data and satellite remote sensing data of each sub-region in the high-altitude cold region, and then obtaining satellite remote sensing data, precipitation particles and hourly precipitation data of each sub-region in the high-altitude cold region. Through the slope analysis of the particle size-velocity curve graph, the present invention can effectively distinguish different precipitation types, combine with the neural network algorithm, construct a phase change model, and update the precipitation type of each sub-region on this basis, thereby significantly improving the accuracy of precipitation type monitoring. Through the multivariate linear regression algorithm, real-time and comprehensive monitoring of precipitation intensity is achieved, which solves the problem that the existing technology is difficult to analyze multi-source precipitation data, distinguish the types of multi-source precipitation data, and evaluate their corresponding precipitation intensity and positioning, resulting in inaccurate stored multi-source precipitation data, and further leading to errors in the retrieval results of multi-source precipitation data in high-altitude cold areas. This ensures that the present invention can refine the dynamic monitoring standards of precipitation phase identification and multi-source data storage and retrieval systems in high-altitude cold areas within a more precise range, making the monitored data a more accurate indicator under the same conditions. The development and application of this method have significantly enhanced the level of intelligence in the process of precipitation phase identification and multi-source data storage and retrieval in high-altitude cold areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0060] Figure 1 This is a block diagram of a precipitation phase identification and multi-source data storage and retrieval system for high-altitude cold regions according to the present invention. DETAILED DESCRIPTION

[0061] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0062] like Figure 1As shown, the present invention provides a precipitation phase identification and multi-source data storage and retrieval system in alpine regions, including a multi-source precipitation data acquisition module, a remote sensing positioning module, a precipitation type identification module, a precipitation intensity assessment module and a storage and retrieval module, wherein each module is communicatively connected;

[0063] The multi-source precipitation data acquisition module collects satellite remote sensing data and ground-based precipitation data from high-altitude cold regions, preprocesses and standardizes the collected data, and performs feature extraction and analysis on the preprocessed precipitation data to obtain precipitation particle characteristics and precipitation amount characteristics, providing data preparation for the construction of models in subsequent modules;

[0064] The remote sensing positioning module uses pre-processed satellite remote sensing data to divide the alpine region into several sub-regions and obtain the positioning data of each sub-region;

[0065] The precipitation type recognition module classifies the precipitation type corresponding to the precipitation particle data based on the characteristics of the precipitation particles, and then obtains the precipitation phase change index and updates the precipitation type corresponding to the precipitation particles. This solves the problem that existing technologies have difficulty in distinguishing the types of precipitation data from multiple sources.

[0066] The precipitation intensity assessment module obtains the solid precipitation intensity index and the liquid precipitation intensity index based on precipitation characteristics, and then evaluates the solid precipitation intensity and the liquid precipitation intensity, solving the problem that existing technologies are difficult to evaluate precipitation intensity;

[0067] The storage and retrieval module integrates the positioning data and precipitation intensity assessment results of each sub-area, and stores and retrieves them.

[0068] The multi-source precipitation data acquisition module collects satellite remote sensing data and precipitation data in alpine areas, and obtains precipitation particle characteristics and precipitation amount characteristics. The process includes:

[0069] The target detection area in the alpine region is divided into several sub-areas of equal size based on topography, climate characteristics, and precipitation distribution characteristics. Different types of collection equipment are deployed in each sub-area to collect high-precision ground precipitation data and satellite remote sensing data for each sub-area. The collection equipment includes microwave remote sensing equipment, total precipitation sensors, laser raindrop spectrometers, and laser precipitation particle spectrometers to measure the precipitation intensity, precipitation phase, precipitation particle size distribution, and precipitation particle motion characteristics in each sub-area.

[0070] Among them, satellite remote sensing data covers multispectral, microwave or radar remote sensing precipitation information, which is satellite remote sensing data for each sub-region and is used to invert precipitation type and intensity. Ground observation precipitation data includes precipitation particle data and hourly precipitation data in each sub-region;

[0071] Data cleaning and normalization are performed on precipitation data, outliers are removed, and missing data are filled. Geometric correction, atmospheric correction, and noise reduction are performed on satellite remote sensing data. Timestamps are added to satellite remote sensing data and precipitation data. Time interpolation and synchronization algorithms are used to synchronize the timestamps of satellite remote sensing data and precipitation data to keep the time series consistent.

[0072] Based on the hourly precipitation data in each sub-region, the total precipitation amount and the number of precipitation events in each sub-region within 24 hours are counted to obtain the precipitation frequency and precipitation volume of each sub-region. The precipitation particle data in each sub-region are combined with a laser precipitation particle spectrometer based on the principle of laser scattering to obtain the velocity and particle size of precipitation particles in each sub-region.

[0073] The precipitation frequency and precipitation volume of each sub-region are used as precipitation characteristics to form a precipitation feature vector. The velocity and particle size of precipitation particles in each sub-region are used as precipitation particle characteristics to form a precipitation particle feature vector. The precipitation feature vector and precipitation particle feature vector are integrated to construct a precipitation feature dataset, and finally the precipitation feature dataset is divided into a training set and a test set.

[0074] Remote sensing positioning module, the process of obtaining positioning data for each sub-area includes:

[0075] Import the pre-processed satellite remote sensing data into GIS software and analyze the satellite remote sensing data of each sub-region through GIS software;

[0076] Select the geographic coordinate system of satellite remote sensing data and adjust the GIS software coordinate system to make it consistent with the geographic coordinate system of satellite remote sensing data;

[0077] Number each sub-region and arrange them in numerical order. Then, use the feature extraction tool of GIS software to divide the boundaries of each sub-region and read the location coordinates of each sub-region.

[0078] The positioning coordinates of each sub-area are integrated and matched with the corresponding labels to obtain the positioning data table of the target detection area.

[0079] The precipitation type recognition module classifies precipitation particles into their corresponding precipitation types based on their characteristics. The process includes:

[0080] Establish a rectangular coordinate system, use the particle size of precipitation particles in each sub-area as the horizontal coordinate, and the velocity of precipitation particles in each sub-area as the vertical coordinate, and draw a particle size-velocity curve;

[0081] Select two representative points in the particle size-velocity curve and calculate the local slope value of the curve in the selected interval according to the differential slope calculation formula. The slope calculation formula of the curve between the two points is as follows:

[0082]

[0083] Where k is the slope of the curve between the two points, and are the precipitation particle velocities of two points in the particle size-velocity curve, and are the precipitation particle sizes of two points selected from the particle size-velocity curve;

[0084] According to the slope characteristics of the particle size-velocity relationship curve, when the slope of the curve between two points is less than 1, the precipitation type corresponding to the precipitation particles is classified as solid precipitation; when the slope of the curve between two points is greater than 1, the precipitation type corresponding to the precipitation particles is classified as liquid precipitation.

[0085] The precipitation type corresponding to the precipitation particles is encoded, the precipitation type code is converted into a precipitation type feature vector, and the vector is incorporated into the precipitation feature dataset to update the precipitation feature dataset.

[0086] In the precipitation type identification module, the process of obtaining the precipitation phase change index includes:

[0087] Weights are assigned to the velocity and size of precipitation particles in each sub-region, as well as the meteorological factors (temperature, humidity, and 0° layer height) of each sub-region. The process of calculating the precipitation phase change index using the assigned weights includes:

[0088]

[0089] in, is the precipitation phase change index, , , and are the speed of precipitation particles in each sub-area, the particle size of precipitation particles in each sub-area, the temperature, humidity and the weight of the 0° layer height in each sub-area, , , 、 and are the velocity of precipitation particles in each sub-area, the particle size of precipitation particles in each sub-area, the temperature, humidity and 0° layer height of each sub-area;

[0090] Extract precipitation particle feature vectors and precipitation type feature vectors from the precipitation feature dataset respectively;

[0091] Using the precipitation particle feature vectors and precipitation type feature vectors extracted from the training set, combined with a neural network algorithm, the precipitation particle feature vectors and precipitation type feature vectors extracted from the training set are used as input, and the precipitation phase change index is used as output. The nonlinear relationship between the precipitation particle feature vectors, precipitation type feature vectors and precipitation phase change index is learned to train the phase change model.

[0092] Input the precipitation particle feature vectors and precipitation type feature vectors extracted from the test set into the phase change model, evaluate the performance of the phase change model, adjust the parameters of the phase change model, optimize the phase change model, and obtain the final phase change model;

[0093] The precipitation particle feature vectors and precipitation type feature vectors in the precipitation feature dataset are input into the phase change model. The precipitation phase change index is output through the phase change model. The precipitation type feature vectors and their corresponding precipitation phase change indices are analyzed to determine the precipitation phase change index threshold in alpine regions.

[0094] In the precipitation type recognition module, the updating process of the precipitation type corresponding to the precipitation particles includes:

[0095] Analyze the numerical characteristics of the precipitation phase change index. When the index is lower than the threshold of the precipitation phase change index in high-altitude cold regions, the precipitation type is determined to be solid precipitation. When the index is higher than the threshold of the precipitation phase change index in high-altitude cold regions, the precipitation type is determined to be liquid precipitation.

[0096] Combined with the original precipitation type corresponding to the precipitation particles, when the precipitation type corresponding to the precipitation phase change index is the same as the precipitation type corresponding to the precipitation particles, the precipitation type corresponding to the precipitation particles is not updated; when the precipitation type corresponding to the precipitation phase change index is different from the precipitation type corresponding to the precipitation particles, the precipitation type corresponding to the precipitation particles is updated to the precipitation type corresponding to the precipitation phase change index;

[0097] Based on the precipitation type corresponding to the updated precipitation particles, the precipitation feature dataset is subdivided, and the precipitation feature vector is divided into a solid precipitation feature vector and a liquid precipitation feature vector. The solid precipitation feature vector includes the solid precipitation frequency and solid precipitation volume of each subregion, and the liquid precipitation feature vector includes the liquid precipitation frequency and liquid precipitation volume of each subregion.

[0098] Precipitation intensity assessment module, the process of obtaining the solid precipitation intensity index includes:

[0099] Solid precipitation feature vectors are extracted from the precipitation feature dataset. The solid precipitation feature vectors in the training set are used in combination with a multivariate linear regression algorithm. The solid precipitation feature vectors extracted from the training set are used as input, and the solid precipitation intensity index is used as output. The linear relationship between the solid precipitation feature vectors and the solid precipitation intensity index is learned to train a solid precipitation intensity model.

[0100] The solid precipitation feature vectors in the test set are input into the solid precipitation intensity model to evaluate the performance of the solid precipitation intensity model. The intercept term and regression coefficient of the solid precipitation intensity model are adjusted to optimize the solid precipitation intensity model and obtain the final solid precipitation intensity model. The expression of the solid precipitation intensity model is:

[0101] in, is the solid precipitation intensity index, is the intercept term, and are the regression coefficients of the solid precipitation frequency and the solid precipitation volume in each sub-region, respectively. is the solid precipitation frequency in each sub-region, is the volume of solid precipitation in each sub-region, is the error term;

[0102] The solid precipitation characteristic vector is input into the solid precipitation intensity model, and the solid precipitation intensity index is output through the solid precipitation intensity model.

[0103] In the precipitation intensity assessment module, the process of obtaining the liquid precipitation intensity index includes:

[0104] Liquid precipitation feature vectors are extracted from the precipitation feature dataset. By combining the liquid precipitation feature vectors in the training set with a multivariate linear regression algorithm, the liquid precipitation feature vectors extracted from the training set are used as input and the liquid precipitation intensity index is used as output. The linear relationship between the liquid precipitation feature vectors and the liquid precipitation intensity index is learned to train a liquid precipitation intensity model.

[0105] The liquid precipitation feature vectors in the test set are input into the liquid precipitation intensity model to evaluate the performance of the liquid precipitation intensity model, adjust the parameters of the liquid precipitation intensity model, optimize the liquid precipitation intensity model, and obtain the final liquid precipitation intensity model. The expression of the liquid precipitation intensity model is:

[0106]

[0107] Where Y is the liquid precipitation intensity index, is the intercept term, and are the regression coefficients of the liquid precipitation frequency and the liquid precipitation volume in each sub-region, respectively. is the liquid precipitation frequency of each sub-region, is the volume of liquid precipitation in each sub-region, is the error term;

[0108] The liquid precipitation characteristic vector is input into the liquid precipitation intensity model, and the liquid precipitation intensity index is output through the liquid precipitation intensity model.

[0109] The precipitation intensity assessment module, the assessment process of solid precipitation intensity and liquid precipitation intensity includes:

[0110] Based on the historical process of the alpine region, the assessment levels of the solid precipitation intensity index and the liquid precipitation intensity index were calculated according to each historical sequence using the percentile calculation method. The percentiles corresponding to the 50th, 80th, 95th, and 98th percentiles were calculated respectively. These percentiles were used as thresholds to divide the solid precipitation intensity level and the liquid precipitation intensity level into five levels (Table 1). The corresponding assessment results are: general (level 5), moderate (level 4), strong (level 3), strong (level 2), and very strong (level 1).

[0111] Table 1 Different percentile ranges (P) and evaluation index levels and intensity evaluation results

[0112]

[0113] Combining the output results of the solid precipitation intensity model and the liquid precipitation intensity model, the solid precipitation intensity and liquid precipitation intensity are evaluated respectively, and finally the solid precipitation intensity evaluation results and liquid precipitation intensity evaluation results are obtained.

[0114] The storage and retrieval module integrates the positioning data and precipitation intensity assessment results of each sub-region, and the storage and retrieval process includes:

[0115] Integrate the liquid precipitation intensity assessment results and the solid precipitation intensity assessment results to generate a multi-source precipitation data retrieval report for high-altitude and cold regions;

[0116] Based on the MySQL database architecture, a precipitation data storage system is constructed, and the positioning data table of the target detection area and the multi-source precipitation data retrieval report of the high-altitude and cold regions are stored in the constructed precipitation data storage system;

[0117] Among them, the construction of the precipitation data storage system includes the positioning data, precipitation type and precipitation intensity of the target detection area in the high-altitude and cold regions. By constructing the precipitation data storage system, accurate retrieval of multi-source precipitation data in the high-altitude and cold regions can be achieved.

[0118] First, the target alpine region is divided into several sub-regions of equal area according to the terrain, climate characteristics and precipitation distribution characteristics. Different types of collection equipment are deployed in each sub-region to collect high-precision ground precipitation data and satellite remote sensing data of each sub-region. The collected data are preprocessed and feature analyzed to extract the precipitation frequency and precipitation volume in each sub-region, obtain the velocity and particle size of precipitation particles in each sub-region, integrate the precipitation amount feature vector and the precipitation particle feature vector to construct a precipitation feature dataset, and finally divide the dataset into a training set and a test set; secondly, the preprocessed satellite remote sensing data is used in combination with GIS software to divide the boundaries of each sub-region, read the positioning coordinates of each sub-region, and then obtain the positioning data of the alpine region; then, based on the characteristics of precipitation particles, a particle size-velocity curve is drawn, and the slope of the curve between the two points is calculated according to the slope calculation formula. , the precipitation types corresponding to the precipitation particles are divided into liquid precipitation and solid precipitation, and the precipitation feature data set is updated; then, through the neural network algorithm, the precipitation particle feature vector and the precipitation type feature vector in the precipitation feature data set are combined to construct a phase change model to output the precipitation phase change index, and the precipitation phase change index is used to update the precipitation type corresponding to the precipitation particles; then, the multivariate linear regression algorithm is used to match the solid precipitation feature vector and the liquid precipitation feature vector in the precipitation feature data set to obtain the solid precipitation intensity index and the liquid precipitation intensity index respectively, and the solid precipitation intensity and liquid precipitation intensity are evaluated according to the solid precipitation intensity index and the liquid precipitation intensity index respectively; finally, a precipitation data storage system is established, and the positioning data table of the target detection area and the multi-source precipitation data retrieval report in the high-altitude cold area are stored in the precipitation database to realize the accurate retrieval of multi-source precipitation data in the high-altitude cold area.

[0119] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A precipitation phase identification and multi-source data storage and retrieval system for high-altitude cold regions, comprising a multi-source precipitation data acquisition module, a remote sensing positioning module, a precipitation type identification module, a precipitation intensity assessment module, and a storage and retrieval module, wherein: The modules are connected in communication, characterized by: The multi-source precipitation data acquisition module collects satellite remote sensing data and ground observation precipitation data in alpine areas, preprocesses and standardizes the collected data, and performs feature extraction and analysis on the preprocessed precipitation data to obtain precipitation particle characteristics and precipitation amount characteristic parameters; The remote sensing positioning module divides the alpine region into several sub-regions using pre-processed satellite remote sensing data and obtains positioning data for each sub-region; The precipitation type identification module classifies the precipitation type corresponding to the precipitation particle data based on the characteristics of the precipitation particles, thereby obtaining a precipitation phase change index and updating the precipitation type corresponding to the precipitation particles. The precipitation type identification module obtains the precipitation phase change index in the following steps: Extract the precipitation particle feature vector and precipitation type feature vector from the precipitation feature dataset respectively; Using the precipitation particle feature vectors and precipitation type feature vectors extracted from the training set, combined with a neural network algorithm, the precipitation particle feature vectors and precipitation type feature vectors extracted from the training set are used as input, and the precipitation phase change index is used as output. The nonlinear relationship between the precipitation particle feature vectors, precipitation type feature vectors and precipitation phase change index is learned to train the phase change model. The precipitation particle feature vectors and precipitation type feature vectors extracted from the test set are input into the phase change model to evaluate the performance of the phase change model. The phase change model parameters are adjusted based on error feedback to optimize the model structure and obtain the final phase change model. The precipitation particle feature vectors and precipitation type feature vectors in the precipitation feature dataset are input into the phase change model. The precipitation phase change index is output through the phase change model. The precipitation type feature vectors and their corresponding precipitation phase change index are analyzed to determine the precipitation phase change index threshold in the alpine region. The precipitation intensity assessment module obtains a solid precipitation intensity index and a liquid precipitation intensity index based on precipitation characteristics, and then assesses the solid precipitation intensity and the liquid precipitation intensity. The solid precipitation intensity index acquisition process includes: Solid precipitation feature vectors are extracted from the precipitation feature dataset. The solid precipitation feature vectors in the training set are used in combination with a multivariate linear regression algorithm. The solid precipitation feature vectors extracted from the training set are used as input, and the solid precipitation intensity index is used as output. The linear relationship between the solid precipitation feature vectors and the solid precipitation intensity index is learned to train a solid precipitation intensity model. The solid precipitation feature vectors in the test set are input into the solid precipitation intensity model to evaluate the performance of the solid precipitation intensity model, adjust the intercept term and regression coefficient of the solid precipitation intensity model, optimize the solid precipitation intensity model, and obtain the final solid precipitation intensity model; The solid precipitation characteristic vector is input into the solid precipitation intensity model, and the solid precipitation intensity index is output through the solid precipitation intensity model; The process of obtaining the liquid precipitation intensity index includes: Liquid precipitation feature vectors are extracted from the precipitation feature dataset. By combining the liquid precipitation feature vectors in the training set with a multivariate linear regression algorithm, the liquid precipitation feature vectors extracted from the training set are used as input and the liquid precipitation intensity index is used as output. The linear relationship between the liquid precipitation feature vectors and the liquid precipitation intensity index is learned to train a liquid precipitation intensity model. Input the liquid precipitation feature vectors in the test set into the liquid precipitation intensity model, evaluate the performance of the liquid precipitation intensity model, adjust the liquid precipitation intensity model parameters, optimize the liquid precipitation intensity model, and obtain the final liquid precipitation intensity model; Input the liquid precipitation characteristic vector into the liquid precipitation intensity model, and output the liquid precipitation intensity index through the liquid precipitation intensity model; The storage and retrieval module integrates the positioning data and precipitation intensity assessment results of each sub-area, and stores and retrieves them.

2. The system for identifying precipitation phases and storing and retrieving multi-source data in high-altitude and cold regions according to claim 1 is characterized by: The multi-source precipitation data acquisition module collects satellite remote sensing data and ground observation precipitation data in alpine areas and obtains precipitation particle characteristics and precipitation amount characteristic parameters. The process includes: The target detection area in the alpine region is divided into several sub-regions based on topography, climate characteristics, and precipitation distribution characteristics. Different types of acquisition equipment are deployed in each sub-region to simultaneously obtain high-precision ground precipitation data and satellite remote sensing data. The acquisition equipment includes microwave remote sensing equipment, total precipitation sensors, laser raindrop spectrometers, and laser precipitation particle spectrometers to measure the precipitation intensity, precipitation phase, precipitation particle size distribution, and precipitation particle motion characteristics in each sub-region. The satellite remote sensing data includes multispectral, microwave or radar remote sensing precipitation information, which is satellite remote sensing data for each sub-region and is used to invert precipitation type and intensity; the ground observation precipitation data includes precipitation particle data and hourly precipitation data in each sub-region; Data cleaning and normalization are performed on precipitation data, outliers are removed, and missing data are filled. Geometric correction, atmospheric correction, and noise reduction are performed on satellite remote sensing data. Timestamps are added to satellite remote sensing data and precipitation data. Time interpolation and synchronization algorithms are used to synchronize the timestamps of satellite remote sensing data and precipitation data to keep the time series consistent. Based on the hourly precipitation data in each sub-region, the total amount of precipitation in each sub-region within 24 hours and the number of precipitation events in each sub-region within 24 hours are counted, and the precipitation frequency and precipitation volume of each sub-region are further extracted. The precipitation particle data in each sub-region are combined with a laser precipitation particle spectrometer based on the principle of laser scattering to obtain the velocity and particle size of precipitation particles in each sub-region. The precipitation frequency and precipitation volume of each sub-region are used as precipitation characteristics to form a precipitation feature vector. The velocity and particle size of precipitation particles in each sub-region are used as precipitation particle characteristics to form a precipitation particle feature vector. The precipitation feature vector and precipitation particle feature vector are integrated to construct a precipitation feature dataset, and finally the precipitation feature dataset is divided into a training set and a test set.

3. The system for identifying precipitation phases and storing and retrieving multi-source data in high-altitude cold regions according to claim 2 is characterized by: The remote sensing positioning module acquires the positioning data of each sub-area by: Import the pre-processed satellite remote sensing data into GIS software and analyze the satellite remote sensing data of each sub-region through GIS software; Select the geographic coordinate system of satellite remote sensing data and adjust the GIS software coordinate system to make it consistent with the geographic coordinate system of satellite remote sensing data; Number each sub-region and arrange them in numerical order. Then, use the feature extraction tool of GIS software to divide the boundaries of each sub-region and read the location coordinates of each sub-region. The positioning coordinates of each sub-region are integrated and matched with the corresponding labels to obtain a standardized positioning data table of the target detection area.

4. The system for identifying precipitation phases and storing and retrieving multi-source data in alpine regions according to claim 3 is characterized by: The precipitation type identification module classifies precipitation particles into corresponding precipitation types based on the characteristics of the precipitation particles, including: Establish a rectangular coordinate system, use the particle size of precipitation particles in each sub-area as the horizontal coordinate, and the velocity of precipitation particles in each sub-area as the vertical coordinate, and draw a particle size-velocity curve; Select two points on the particle size-velocity curve and calculate the local slope of the curve between the two selected points according to the differential slope calculation formula. Based on the slope characteristics of the particle size-velocity relationship curve, the precipitation type corresponding to the precipitation particles is divided into liquid precipitation and solid precipitation to achieve precipitation type classification; The precipitation type corresponding to the precipitation particles is encoded, the precipitation type code is converted into a precipitation type feature vector, and is incorporated into the precipitation feature dataset to update the precipitation feature dataset.

5. The system for identifying precipitation phases and storing and retrieving multi-source data in high-altitude and cold regions according to claim 4 is characterized by: The precipitation type identification module updates the precipitation type corresponding to the precipitation particles in the following process: Analyze the numerical characteristics of the precipitation phase change index. When the index is lower than the threshold of the precipitation phase change index in high-altitude cold regions, the precipitation type is determined to be solid precipitation. When the index is higher than the threshold of the precipitation phase change index in high-altitude cold regions, the precipitation type is determined to be liquid precipitation. Combined with the precipitation type corresponding to the precipitation particles, when the precipitation type corresponding to the precipitation phase change index is the same as the precipitation type corresponding to the precipitation particles, the precipitation type corresponding to the precipitation particles will not be updated; when the precipitation type corresponding to the precipitation phase change index is different from the precipitation type corresponding to the precipitation particles, the precipitation type corresponding to the precipitation particles will be updated to the precipitation type corresponding to the precipitation phase change index; Based on the precipitation type corresponding to the updated precipitation particles, the precipitation feature dataset is subdivided, and the precipitation feature vector is divided into a solid precipitation feature vector and a liquid precipitation feature vector. The solid precipitation feature vector includes the solid precipitation frequency and solid precipitation volume of each subregion, and the liquid precipitation feature vector includes the liquid precipitation frequency and liquid precipitation volume of each subregion.

6. The precipitation phase identification and multi-source data storage and retrieval system for high-altitude cold regions according to claim 5 is characterized by: The precipitation intensity assessment module includes the following steps to assess the solid precipitation intensity and liquid precipitation intensity: Based on the historical process of the alpine region and in accordance with each historical sequence, the percentile calculation method was used to calculate the percentiles corresponding to the 50th, 80th, 95th and 98th percentiles of the assessment levels of the solid precipitation intensity index and the liquid precipitation intensity index, respectively. These percentiles were used as thresholds to divide the solid precipitation intensity index and the liquid precipitation intensity index into five levels, and the corresponding assessment results were: general, moderate, relatively strong, strong and extremely strong, among which general corresponds to level 5, moderate corresponds to level 4, relatively strong corresponds to level 3, strong corresponds to level 2 and extremely strong corresponds to level 1. Combining the output results of the solid precipitation intensity model and the liquid precipitation intensity model, the solid precipitation intensity and the liquid precipitation intensity are evaluated respectively to obtain the final solid precipitation intensity evaluation results and the liquid precipitation intensity evaluation results.

7. The precipitation phase identification and multi-source data storage and retrieval system for high-altitude cold regions according to claim 6 is characterized by: The storage and retrieval module integrates the positioning data and precipitation intensity assessment results of each sub-region and stores and retrieves them, including: Integrate the liquid precipitation intensity assessment results and the solid precipitation intensity assessment results to generate a multi-source precipitation data retrieval report for high-altitude and cold regions; Based on the MySQL database architecture, a precipitation data storage system is constructed, and the positioning data table of the target detection area and the multi-source precipitation data retrieval report of the high-altitude and cold regions are stored in the precipitation data storage system; The precipitation data storage system includes positioning data, precipitation type and precipitation intensity of the target detection area in the alpine region. Through the precipitation data storage system, accurate retrieval of multi-source precipitation data in the alpine region can be achieved.

Images