A dual-polarization radar hail size classification method, device, medium, and product

Abstract

The application discloses a dual-polarization radar hail size classification method, equipment, medium and product, relates to the technical field of radar meteorological detection and disaster weather monitoring, and comprises the following steps: acquiring hail data sets of S-band radar and C-band radar respectively; the hail data set comprises ground hail reports of different sizes of hail; the radar elevation data and the ground hail report are matched in space and time, and the polarization quantity of each elevation and the radar sampling body height are obtained; the polarization quantity of different sizes of hail is classified into the corresponding height interval according to the radar sampling body height, and the polarization quantity statistical data set is obtained; the membership function corresponding to each polarization quantity is selected based on the radar band and the height interval; the hail size classification result of the distance library is determined according to the observation value of the polarization quantity of each distance library and the membership function corresponding to the polarization quantity by using a fuzzy logic method, and the application can realize the identification of the hail size classification conforming to the domestic hail size classification, and improve the accuracy of the identification result.

Description

Technical Field

[0001] This application relates to the fields of radar meteorological detection and severe weather monitoring technology, and in particular to a dual-polarization radar hail size classification method, equipment, medium and product. Background Technology

[0002] Hail is a severe weather event caused by strong convective storms, characterized by its small spatial scale and sudden onset. In recent years, hail events in my country have become more frequent and increasingly severe, posing a serious threat to infrastructure and socio-economic development. Therefore, effectively improving hail monitoring and early warning capabilities has always been a key focus of severe weather research. Weather radar, which actively emits electromagnetic waves to detect precipitation particles, is an effective tool for real-time monitoring of severe convective weather and has been widely used for hail identification. Early hail identification was mainly based on horizontal reflectivity factors. Z H The single-polarization method, however, relies solely on... Z H Empirical judgments are often insufficient to effectively distinguish between heavy rainfall and hail in practical applications, resulting in high false alarm rates and overestimation of hail size. Dual-polarization radar represents the forefront of current radar meteorological detection technology, providing a new approach to hail identification. Among these, the hail size discrimination algorithm (HSDA) designed for the US S-band dual-polarization radar is flexible, easy to improve, and outperforms traditional single-polarization methods.

[0003] my country has built a weather radar network covering key areas of the country and is gradually carrying out dual polarization upgrades. As of January 2025, 230 radars have been upgraded. How to give full play to the advantages of these dual polarization data in hail monitoring has become an urgent technical problem to be solved. However, the existing HSDA has obvious limitations in my country's operational applications: (1) Insufficient applicability of radar bands: Unlike the US operational radar network, which uniformly adopts the S-band, my country's radar network has multiple models and multiple bands due to complex terrain and generally weak strong convection. The polarization measurement values ​​of the C-band are systematically different from those of the S-band. Therefore, directly applying the HSDA developed for the US S-band will lead to identification errors. (2) Inconsistent hail classification standards: my country's hail classification standards are different from those of foreign countries, and the corresponding membership functions also need to be optimized according to my country's standards. Summary of the Invention

[0004] The purpose of this application is to provide a dual-polarization radar hail size classification method, device, medium, and product, which can achieve identification that conforms to domestic hail size classification and improve the accuracy of identification results.

[0005] To achieve the above objectives, this application provides the following solution.

[0006] Firstly, this application provides a dual-polarization radar hail size classification method, comprising the following steps: acquiring hail datasets from S-band radar and C-band radar respectively; the hail datasets include ground hail reports of different hail sizes; performing spatiotemporal matching between radar elevation data and ground hail reports to obtain the polarization quantity and radar sampling height for each elevation angle; the polarization quantity includes horizontal reflectivity factor, differential reflectivity, and correlation coefficient; based on the radar sampling height for each elevation angle, classifying the polarization quantity of hail of different sizes into corresponding height intervals to obtain a polarization quantity statistical data set divided by radar band, hail size category, and height interval; the polarization quantity statistical data set includes the polarization quantity observation values ​​of all ground hail reports at their matching positions; determining the optimized membership function corresponding to each polarization quantity based on the radar band and height interval; and using fuzzy logic, determining the hail size classification result for each range library based on the observed polarization quantity and the optimized membership function.

[0007] Optionally, the vertical height is divided into the following 6 intervals based on the wet-bulb 0°C and -25°C layers: First height interval: H <H(T w =0℃)-3km; Second altitude range: H(T) w =0℃)-3km≤H <H(T w =0℃)-2km; Third altitude range: H(T) w =0℃)-2km≤H <H(T w =0℃)-1km; Fourth altitude range: H(T) w =0℃)-1km≤H <H(T w =0℃); Fifth altitude range: H (T w =0℃)≤H <H(T w =-25℃); Sixth altitude range: H≥H(T) w =-25℃); where H represents the radar sampling volume height; H(T w =0℃) represents the height at which the wet-bulb temperature is 0℃; H(T) represents the height at which the wet-bulb temperature is 0℃; w =-25℃) indicates the altitude at which the wet-bulb temperature is -25℃.

[0008] Optionally, the optimized membership function corresponding to each polarization quantity is determined based on the radar band and altitude range. Specifically, when the radar band is S-band and the radar sampling object altitude is in the first, second, and third altitude ranges, a two-dimensional membership function is used; the two-dimensional membership function considers two parameters: horizontal reflectivity factor and differential reflectivity; when the radar band is S-band and the radar sampling object altitude is in the fourth, fifth, and sixth altitude ranges, a one-dimensional membership function is used; when the radar band is C-band, a one-dimensional membership function is used.

[0009] Optionally, a fuzzy logic method is used to determine the hail size classification result for each distance library based on the observed values ​​of polarization quantities and the optimized membership functions. Specifically, this includes: using the observed values ​​of polarization quantities in each distance library as input variables, calculating the membership degree of each polarization quantity for small hail, large hail, and giant hail based on the optimized membership functions corresponding to the polarization quantities; performing a weighted average calculation on the membership degrees of each polarization quantity for small hail, large hail, and giant hail to obtain the aggregate value corresponding to each hail category; and determining the hail size classification result for each distance library based on the aggregate values ​​corresponding to all hail categories.

[0010] Optionally, the hail size classification result for each distance library is determined based on the aggregation value corresponding to all hail categories. Specifically, this includes: comparing the aggregation values ​​corresponding to each hail category, and determining the hail category corresponding to the aggregation value with the largest value as the hail size classification result for each distance library; if the hail size classification result for the distance library is large hail or giant hail, and the aggregation value corresponding to the hail size classification result is ≤ a set aggregation threshold or the differential reflectivity is ≥ a set differential reflectivity threshold, then the hail size classification result for the distance library is downgraded to small hail.

[0011] Optionally, the radar elevation data at each elevation angle is spatiotemporally matched with the ground hail report to obtain the polarization amount and radar sampling height for each elevation angle. Specifically, this includes: for each ground hail report, selecting the radar elevation data at each elevation angle before and after the hail time in the ground hail report for spatiotemporal matching; within a set distance radius of the hail location, using a set horizontal reflectivity factor as a threshold to retrieve the location of the maximum horizontal reflectivity factor at each elevation angle, and taking the location of the maximum horizontal reflectivity factor as the matched hail location for each elevation angle; extracting the polarization amount and radar sampling height at the matched hail location for each elevation angle as the polarization amount and radar sampling height for each elevation angle.

[0012] Optionally, hail datasets from S-band radar and C-band radar are acquired separately. Specifically, this includes: acquiring ground hail reports of the target area within a historical time period; classifying ground hail reports into corresponding hail size categories according to the set hail level standards; filtering all ground hail reports using set filtering criteria to obtain valid ground hail reports, and constructing hail datasets for S-band radar and C-band radar respectively; the filtering criteria are: the hail location is within 150 km of the radar station, radar echoes exist within 1 hour before and after the hail time, the echoes are not affected by ground obstruction, and the radar has completed dual polarization upgrade.

[0013] Secondly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to implement the steps of the dual-polarization radar hail size classification method described above.

[0014] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described dual-polarization radar hail size classification method.

[0015] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described dual-polarization radar hail size classification method.

[0016] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a dual-polarization radar hail size classification method, device, medium, and product. By performing spatiotemporal matching of radar volume scan data and ground hail reports, the polarization amount and radar sampling volume height at each elevation angle are obtained. Based on the radar sampling volume height at each elevation angle, the polarization amount of hail of different sizes is classified into the corresponding height interval, resulting in a polarization amount statistical data set divided by radar band, hail size category, and height interval. The radar and hail matching method considering spatiotemporal deviation, as well as the statistical characteristics of radar observation data (polarization amount of hail of different sizes) in different height intervals, can more accurately reflect the relationship between the actual hail size and polarization amount on the ground. Based on the radar band and height interval, the optimized membership function corresponding to each polarization amount is selected, that is, membership function optimization is performed by band. The localized HSDA established for the characteristics of S / C bands takes into account the differences in domestic radar and hail characteristics, and can achieve identification that conforms to domestic hail size classification, thus improving the accuracy of the identification results. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is an application environment diagram of a dual-polarization radar hail size classification method according to an embodiment of this application.

[0019] Figure 2 This is a flowchart illustrating a dual-polarization radar hail size classification method provided in one embodiment of this application.

[0020] Figure 3This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0023] The dual-polarization radar hail size classification method provided in this application can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on another server. Terminal 102 can send the hail dataset to server 104. After receiving the hail dataset, server 104 performs spatiotemporal matching between radar elevation data at each elevation angle and ground hail reports to obtain the polarization amount and radar sampling height at each elevation angle. Based on the radar sampling height at each elevation angle, the polarization amounts of hailstones of different sizes are assigned to the corresponding height intervals, resulting in a polarization data set divided by radar band, hailstone size category, and height interval. This dataset includes the polarization observation values ​​of all hailstone reports at their matching locations, used to analyze the polarization characteristics of hailstones of different sizes and optimize the membership function accordingly. Then, based on the radar band and height interval, the optimized membership function is selected for each polarization amount. Using fuzzy logic, the hailstone size classification result for each distance library is determined based on the observed polarization values ​​and the selected membership function. Server 104 can then feed back the obtained hailstone size classification results to terminal 102. In addition, in some embodiments, the dual-polarization radar hail size classification method can also be implemented by the server 104 or the terminal 102 separately. For example, the terminal 102 can directly perform hail size classification on the hail dataset, or the server 104 can obtain the hail dataset from the data storage system and perform hail size classification on the hail dataset.

[0024] The terminal 102 can be, but is not limited to, various desktop computers and laptops. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers, or it can be a cloud server.

[0025] In one exemplary embodiment, such as Figure 2 As shown, a dual-polarization radar hail size classification method is provided. This method is executed by a computer device, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps 201 to 205.

[0026] Step 201: Obtain hail datasets from S-band radar and C-band radar respectively; the hail datasets include ground hail reports of hailstones of different sizes.

[0027] Step 202: Spatiotemporally match the radar elevation data with the ground hail report to obtain the polarization amount and radar sampling height for each elevation angle; the polarization amount includes the horizontal reflectivity factor, differential reflectivity and correlation coefficient.

[0028] Step 203: Based on the radar sampling height at each elevation angle, the polarization of hailstones of different sizes is assigned to the corresponding height interval, resulting in a polarization statistical data set divided by radar band, hailstone size category, and height interval. The polarization statistical data set includes the polarization observation values ​​of all ground hailstone reports at their matching locations, which are used to analyze the polarization characteristics of hailstones of different sizes and optimize the membership function accordingly.

[0029] Step 204: Determine the optimized membership function corresponding to each polarization quantity based on the radar band and altitude range.

[0030] Step 205: Using fuzzy logic, determine the hail size classification result for each distance library based on the observed values ​​of polarization quantity and the optimized membership function.

[0031] By implementing steps 201 to 205 above, the polarization amount and radar sampling height of each elevation angle are obtained by spatiotemporally matching radar elevation data with ground hail reports. Based on the radar sampling height of each elevation angle, the polarization amounts of hail of different sizes are classified into corresponding height intervals, resulting in a polarization amount statistical data set divided by radar band, hail size category, and height interval. The radar-hail matching method that considers spatiotemporal deviations, as well as the statistical characteristics of radar observation data in different height intervals, can more accurately reflect the relationship between the actual hail size and polarization amount on the ground. By selecting the corresponding membership function for each polarization amount based on the radar band and height interval, i.e., optimizing the membership function by band, and establishing localized HSDAs for the characteristics of S / C bands respectively, the differences in domestic radar and hail characteristics are considered, which can achieve identification that conforms to the domestic hail size classification and improve the accuracy of the identification results. To address the shortcomings of existing hail identification methods in terms of applicability to domestic radar and hail characteristics, and the lack of a systematic and reliable hail size verification dataset in China, this paper proposes a technical solution that includes: constructing a ground hail size dataset; spatiotemporal matching of radar observations and hail reports; statistical analysis of the polarization characteristics of various hail types at different altitudes; and optimization of membership functions. The solution utilizes the horizontal reflectivity factor (…). Z H ), differential reflectance ( Z DR ) and correlation coefficient ( ρ hv The fuzzy logic method is used to more accurately classify hail into three categories: small hail, large hail, and giant hail, providing technical support for hail monitoring and early warning.

[0032] In step 201 above, hail datasets from S-band radar and C-band radar are obtained respectively, specifically including the following steps 301 to 303.

[0033] Step 301: Obtain ground hail reports for the target area within a historical time period; using direct disaster reports from the meteorological disaster management system, disaster reports from the media, and hail photos posted on the Internet by eyewitnesses as evidence, collect ground hail reports for the target area (such as China) within a specific historical time period (such as 2022-2025). Each report includes the time of hail, the location of hail, and the maximum hail diameter.

[0034] Step 302: Based on the set hail rating standards, classify ground hail reports into the corresponding hail size categories. According to the set hail rating standards (small hail: diameter < 2 cm; large hail: 2 cm ≤ diameter < 5 cm; giant hail: diameter ≥ 5 cm), classify individual ground hail reports into the corresponding hail size categories.

[0035] Step 303: All ground hail reports are filtered using predefined criteria to obtain valid reports. Hail datasets for S-band and C-band radars are then constructed. These datasets include polarization observations of all ground hail reports at their matching locations, used to analyze the polarization characteristics of hailstones of different sizes. The filtering criteria are: hail location within 150 km of the radar station, radar echoes present within one hour before and after the hailfall, echoes unaffected by ground obstruction, and the radar having undergone dual-polarization upgrades. Based on these criteria, valid dual-polarization data can be obtained. All valid reports filtered according to these criteria are then divided to construct separate hail datasets for S-band and C-band radars to meet the localization needs of different radar bands.

[0036] In step 202 above, the radar elevation data and ground hail reports are spatiotemporally matched to obtain the polarization amount and radar sampling volume height for each elevation angle. Specifically, this includes: for each ground hail report, selecting radar elevation data (radar volume scan data) from the two times before and the time after hailfall in the ground hail report for spatiotemporal matching; within a set distance radius of the hailfall location, using a set horizontal reflectivity factor as a threshold to retrieve the location of the maximum horizontal reflectivity factor for each elevation angle, and using the location of the maximum horizontal reflectivity factor as the hail location matched for each elevation angle; extracting the polarization amount (including horizontal reflectivity factor) at the hail location matched for each elevation angle. Z H Differential reflectivity Z DR and correlation coefficient ρ hv The elevation angle is determined by the radar sampling height, which is used as the polarization quantity of the radar sampling height. The above matching method was adopted after taking into account factors such as radar volume scan interval, hail reporting lag, and spatiotemporal deviation caused by storm movement.

[0037] In a specific example, the distance can be set to 2 km, and the horizontal reflectivity factor can be set to 45 dBZ.

[0038] After extracting the polarization values ​​and radar sampling heights corresponding to all elevation angles, the radiosonde data closest to the time and location of hailfall were selected. Based on the wet-bulb 0°C and -25°C layers, the vertical height was divided into the following 6 height intervals: First height interval H1: H <H(T w =0℃)-3km; Second altitude range H2: H(T w =0℃)-3km≤H <H(T w =0℃)-2km; Third altitude range H3: H(T w =0℃)-2km≤H <H(T w =0℃)-1km; Fourth altitude range H4: H(Tw =0℃)-1km≤H <H(T w =0℃); Fifth altitude range H5: H (T w =0℃)≤H <H(T w =-25℃); Sixth altitude range H6: H≥H (T w =-25℃); where H represents the radar sampling volume height; H(T w =0℃) represents the height at which the wet-bulb temperature is 0℃; H(T) represents the height at which the wet-bulb temperature is 0℃; w =-25℃) represents the altitude at which the wet-bulb temperature is -25℃. Furthermore, based on the radar sample height, the polarization values ​​are categorized to the corresponding altitude zones, forming a set of polarization statistical data divided by radar band, hail size category, and altitude range.

[0039] Membership functions are used to characterize the degree of matching between each polarization quantity and different hail categories of different sizes. Membership functions that conform to the range and distribution characteristics of polarization quantities are selected for radar bands and altitude ranges. In step 204 above, the optimized membership function corresponding to each polarization quantity is determined based on the radar band and altitude range, specifically including the following steps 401 to 402.

[0040] Step 401: When the radar band is S-band and the radar sampling body height is in the first, second, and third height intervals, a two-dimensional membership function is used; the two-dimensional membership function considers two parameters: horizontal reflectivity factor and differential reflectivity; when the radar band is S-band and the radar sampling body height is in the fourth, fifth, and sixth height intervals, a one-dimensional membership function is used.

[0041] S-band radar: In the lower atmosphere (first altitude range H1 - third altitude range H3), differential reflectivity Z DR Membership function parameters and horizontal reflectivity factor Z H Since the observed values ​​are relevant, the horizontal reflectivity factor is used. Z H With differential reflectivity Z DR Two-dimensional membership function; at high altitudes (fourth altitude interval H4 - sixth altitude interval H6), horizontal reflectivity factor Z H With differential reflectivity Z DR There is no obvious correlation, so a one-dimensional membership function is used (a trapezoidal membership function can be used).

[0042] Step 402: When the radar band is C-band, a one-dimensional membership function is used. C-band radar: Subject to stronger attenuation and resonant scattering, the horizontal reflectivity factor... Z HWith differential reflectivity Z DR There is no significant correlation across all layers (from the first height interval H1 to the sixth height interval H6), so a one-dimensional membership function is used.

[0043] The specific method for optimizing the membership function is as follows: compare the statistical characteristics of the polarization quantity observations with the differences of the default membership function parameters, and use the sensitivity test method to adjust the parameters so that the optimized membership function can match the actual distribution of polarization quantity to the greatest extent and ensure that the classification results based on the membership function have the highest degree of agreement with the ground category.

[0044] In step 205 above, the hail size classification result of each distance library is determined by using the fuzzy logic method based on the observed value of the polarization quantity of each distance library and the optimized membership function, specifically including the following steps 501 to 503.

[0045] Step 501 is the fuzzification step: using the observed value of the polarization quantity of each distance library as the input variable, calculate the membership degree of each polarization quantity for small hail, large hail, and giant hail according to the optimized membership function corresponding to the polarization quantity.

[0046] by Z H , Z DR and ρ hv The observed values ​​are used as input variables. Based on the height range to which each distance library belongs, the corresponding membership function in step 204 is called to calculate the membership degree of each input variable for small hail, large hail, and giant hail.

[0047] Step 502 is the aggregation step: The weighted average of the membership degree of each polarization quantity for small hailstones, large hailstones, and giant hailstones is calculated to obtain the aggregation value corresponding to each hailstone category. The aggregation value corresponding to each hailstone category is calculated by weighting the membership degree of each input variable using formula (1).

[0048] (1).

[0049] In the above formula, Indicates the first Aggregate value corresponding to hail category; Indicates the first Hailstone category number polarization quantity membership function, These represent the horizontal reflectivity factor, differential reflectivity, and correlation coefficient, respectively. Indicates the first The weighting coefficients for each polarization quantity.

[0050] Step 503 is the defuzzification step: determine the hail size classification result for each distance library based on the aggregate value corresponding to all hail categories. Specifically, this includes: comparing the aggregate values ​​corresponding to each hail category and determining the hail category corresponding to the largest aggregate value as the hail size classification result for each distance library.

[0051] To reduce misclassification, a series of size downgrade rules are applied to correct the results. For example, if the hail size classification result in the distance library is large hail or giant hail, and the aggregate value corresponding to the hail size classification result is ≤ a set aggregate threshold or the differential reflectance is ≥ a set differential reflectance threshold, then the hail size classification result in that distance library is downgraded to small hail. Finally, the final hail size category for each distance library is output. In a specific example, the set aggregate threshold can be 0.6, and the set differential reflectance threshold can be 2 dB.

[0052] The lack of verification data for related technologies hinders the development and verification of HSDA: membership functions are determined based on model simulations or real-world statistical data, but my country lacks a systematic and reliable dataset of hail size, which restricts the development and verification of HSDA.

[0053] This application has the following advantages and four beneficial effects.

[0054] Advantage 1: By collecting 586 valid hail reports from China over the past three years, a more complete dataset of ground hail sizes has been constructed than ever before.

[0055] This advantage stems from the multiple ways hail information is collected in step 201. In addition to direct disaster reports from the meteorological disaster management system as the main source, media disaster reports and hail pictures posted on the Internet by eyewitnesses after screening and verification are also a powerful supplement.

[0056] Advantage 2: By analyzing the observation data of S / C band dual polarization radar, compared with previous domestic studies that mostly focused on individual cases of hail, a set of polarization quantity statistical data has been formed for different hail sizes and categories.

[0057] This advantage is based on the hail dataset in step 201, utilizes the radar-hail matching method that takes into account spatiotemporal deviation in step 202, and the statistical characteristics of S / C band radar observation data in different altitude ranges in step 203, which can more accurately reflect the relationship between the actual hail size and polarization on the ground.

[0058] Advantage 3: Improved regional applicability of HSDA: Compared with the previous method of directly applying foreign S-band radar for hail identification, the membership function optimization performed by band in step 204 of this application, and the localized HSDA established for S / C band characteristics in step 205, take into account the differences between domestic radar and hail characteristics.

[0059] Differences in radar deployment: S-band radars are widely deployed in plain areas such as North China, Central China, and South China, while C-band radars are mainly used in plateau and mountainous areas in the central and western regions. When detecting hail, C-band radars are affected by stronger attenuation and resonance effects, exhibiting polarization characteristics different from those of the S-band.

[0060] Differences in hail classification: The intensity of hail in my country is generally weaker than that in the United States, and there are also differences in the classification of hail size. The U.S. National Weather Service defines large hail as having a diameter of ≥2.5 cm, while my country's hail classification standard stipulates that large hail has a diameter of ≥2 cm.

[0061] Advantage 4: Improved hail recognition rate and size classification accuracy of HSDA after optimization: Algorithm evaluation indicators show (see Table 1) that the hail recognition rate of S-band radar at an elevation angle of 1.5° increased from 83.45% to 91.89%; at the same time, the size classification accuracy of S / C-band radar at an elevation angle of 1.5° was 65.07% and 81.78% respectively, an improvement of 5%-6%, and the overestimation rate of size was reduced.

[0062] Table 1. Comparison of evaluation metrics for the original and optimized HSDA at 1.5° and 2.4° elevation angles in the S / C band radar.

[0063] The improvement in algorithm performance is the result of the combined effect of all steps. Step 201 provides a hail size verification dataset for the localization development and verification of subsequent algorithms. Step 204 provides membership function optimization for different radar bands, altitude ranges and hail size categories. Step 205 realizes the localization of HSDA, which can more accurately determine hail size.

[0064] This application also provides an application scenario in which the aforementioned dual-polarization radar hail size classification method is applied. Specifically, the dual-polarization radar hail size classification method provided in this embodiment can be applied in hail warning scenarios. A hail warning scenario includes a data acquisition stage, a hail size classification link, and a warning stage. The hail dataset enters the hail size classification link from the data acquisition stage, obtains the corresponding hail size classification results through human-machine collaboration, and then enters the downstream warning stage. The dual-polarization radar hail size classification method provided in this embodiment belongs to the hail size classification link. Specifically, in the hail size classification process for hail datasets, spatiotemporal matching can be performed between radar elevation data at each elevation angle and ground hail reports to obtain the polarization amount and radar sample height at each elevation angle. Based on the radar sample height at each elevation angle, the polarization amounts of hail of different sizes are assigned to the corresponding height intervals, resulting in a polarization statistical data set divided by radar band, hail size category, and height interval. This dataset includes the polarization amount observations of all hail cases at their matching locations, used to analyze the polarization characteristics of hail of different sizes and optimize the membership function accordingly. Then, based on the radar band and height interval, the optimized membership function is selected for each polarization amount. Using fuzzy logic, the hail size classification result for each distance library is determined based on the observed polarization amount and the selected membership function.

[0065] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 3 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs in the non-volatile storage media to run. The database stores hail size classification data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network. When executed by the processor, the computer program implements a dual-polarization radar hail size classification method.

[0066] Those skilled in the art will understand that Figure 3The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0067] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0068] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0069] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of related data must comply with relevant regulations and be authorized by the owner of the corresponding device.

[0070] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0071] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0072] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0073] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for classifying hail size using dual-polarization radar, characterized in that, The method includes: Hail datasets from S-band radar and C-band radar were acquired separately; the hail datasets included ground hail reports of hailstones of different sizes. The radar elevation data at each elevation angle is spatiotemporally matched with ground hail reports to obtain the polarization quantity and radar sampling height for each elevation angle; the polarization quantity includes horizontal reflectivity factor, differential reflectivity and correlation coefficient; Based on the radar sampling height at each elevation angle, the polarization of hailstones of different sizes is classified into the corresponding height interval, resulting in a polarization statistical data set divided by radar band, hailstone size category, and height interval; the polarization statistical data set includes the polarization observation values ​​of all ground hail reports at their matching locations. The optimized membership function corresponding to each polarization quantity is determined based on the radar band and altitude range. Using fuzzy logic, the hail size classification results for each distance library are determined based on the observed values ​​of polarization quantity and the optimized membership function.

2. The dual-polarization radar hail size classification method according to claim 1, characterized in that, The vertical height is divided into the following 6 intervals based on the wet-bulb 0°C and -25°C layers: First altitude range: H <H(T w =0℃)-3km; Second altitude range: H(T) w =0℃)-3km≤H <H(T w =0℃)-2km; Third altitude range: H(T) w =0℃)-2km≤H <H(T w =0℃)-1km; Fourth altitude range: H(T w =0℃)-1km≤H <H(T w =0℃); Fifth altitude range: H (T w =0℃)≤H <H(T w =-25℃); Sixth altitude range: H≥H(T) w =-25℃); where H represents the radar sampling volume height; H(T w =0℃) represents the height at which the wet-bulb temperature is 0℃; H(T) represents the height at which the wet-bulb temperature is 0℃; w =-25℃) indicates the altitude at which the wet-bulb temperature is -25℃.

3. The dual-polarization radar hail size classification method according to claim 2, characterized in that, The optimized membership function for each polarization quantity is determined based on the radar band and altitude range, specifically including: When the radar band is S-band and the radar sampling body height is in the first, second, and third height intervals, a two-dimensional membership function is used; the two-dimensional membership function considers two parameters: horizontal reflectivity factor and differential reflectivity; When the radar band is S-band and the radar sampling body height is in the fourth, fifth and sixth height intervals, a one-dimensional membership function is used. When the radar band is C-band, a one-dimensional membership function is used.

4. The dual-polarization radar hail size classification method according to claim 1, characterized in that, Using fuzzy logic, the hail size classification results for each distance library are determined based on the observed polarization values ​​and optimized membership functions. Specifically, this includes: Using the observed values ​​of polarization quantities for each distance library as input variables, the membership degree of each polarization quantity for small hail, large hail, and giant hail is calculated based on the optimized membership function corresponding to the polarization quantity. The weighted average of the membership degree of each polarization quantity for small hail, large hail, and giant hail is calculated to obtain the aggregate value corresponding to each hail category; The hail size classification result for each distance library is determined based on the aggregate value corresponding to all hail categories.

5. The dual-polarization radar hail size classification method according to claim 4, characterized in that, The hail size classification results for each distance library are determined based on the aggregated values ​​corresponding to all hail categories, specifically including: Compare the aggregate values ​​corresponding to each hail category, and determine the hail category corresponding to the largest aggregate value as the hail size classification result for each distance library; If the hail size classification result in the distance database is large hail or giant hail, and the aggregate value corresponding to the hail size classification result is ≤ set aggregate threshold or the differential reflectance is ≥ set differential reflectance threshold, then the hail size classification result in the distance database is downgraded to small hail.

6. The dual-polarization radar hail size classification method according to claim 1, characterized in that, The radar elevation data at each elevation angle is spatiotemporally matched with ground hail reports to obtain the polarization value and radar sampling volume height for each elevation angle, specifically including: For each ground hail report, the radar elevation data from the two times before and the last time of hail in the ground hail report are selected for spatiotemporal matching. Within a set distance radius of the hail location, the location of the maximum horizontal reflectivity factor at each elevation angle is retrieved with a set horizontal reflectivity factor value as the threshold. The location of the maximum horizontal reflectivity factor is used as the matched hail location for each elevation angle. The polarization quantity and radar sampling height at the hail location matched at each elevation angle are extracted and used as the polarization quantity and radar sampling height at each elevation angle.

7. The dual-polarization radar hail size classification method according to claim 1, characterized in that, Hail data sets were obtained from S-band radar and C-band radar, respectively, including: Obtain ground hail reports for the target area within a historical time period; Based on the established hail severity standards, ground hail reports are categorized into corresponding hail size categories; All ground hail reports were filtered using set criteria to obtain valid ground hail reports. Hail datasets for S-band radar and C-band radar were constructed respectively. The set criteria were: the hail location was within 150 km of the radar station, there were radar echoes within 1 hour before and after the hail time, the echoes were not affected by ground objects, and the radar had completed dual polarization upgrade.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement the dual-polarization radar hail size classification method according to any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the dual-polarization radar hail size classification method according to any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the dual-polarization radar hail size classification method according to any one of claims 1-7.

Images