A hail identification tracking method and system for the whole life cycle of hail embryos

Abstract

This invention discloses a hail identification and tracking method and system for the entire life cycle of hail embryos, relating to the field of meteorological detection. The method includes: S1 acquiring dual-polarization radar volume scan data; S2 identifying hail embryo particles; S3 determining whether the hail embryo particles are new particles, and if so, creating a life profile; S4 tracking target particles in spatiotemporal correlation and updating the life profile; S5 determining whether the target particles have transformed into hail; S6 acquiring the current life profile of the hail particles; S7 determining whether hail precipitation conditions have been triggered, and if so, proceeding to S8, otherwise returning to S1; S8 analyzing the location of the precipitation area; S9 spatially clustering the precipitation area location and generating a probability map of the hail precipitation area and early warning information. By establishing a "life profile" for each hail embryo particle that includes its trajectory, phase evolution, and environmental thermodynamic conditions, the complete process of hail embryo formation from initial formation to final hail precipitation is tracked. Based on the statistical analysis of the life profile, key transformation characteristics of different hail storm types are adaptively identified, thereby achieving early hail warning.

Description

Technical Field

[0001] This invention relates to the field of meteorological detection, and in particular to a method and system for identifying and tracking hail embryos throughout their entire life cycle. Background Technology

[0002] Hail, a severe convective weather phenomenon, poses a serious threat to agriculture, infrastructure, and the safety of people's lives and property. Hail embryos, as the initial nuclei of hail formation, directly determine the final size and landing area of ​​hailstones through their origin, movement, transformation, and growth processes. Therefore, achieving precise tracking of the entire life cycle of hail embryo particles is of great significance for revealing the mechanism of hail formation, improving the accuracy of hail warnings, and guiding artificial hail suppression operations.

[0003] Current research on hail embryos and hailstones has the following limitations:

[0004] (1) Although simulation studies based on numerical models can simulate the trajectory of hail embryos, they rely on model parameterization schemes, have uncertainties, and are difficult to reflect real cloud physical processes.

[0005] (2) Radar-based hail identification methods mainly focus on macroscopic echo characteristics and forecast index calculation, but fail to achieve fine tracking at the particle level and cannot trace the specific source and evolution path of hail embryos;

[0006] (3) The existing technology has not yet established a complete tracking framework covering the four stages of hail embryo “source-movement-transformation-hailfall”, and also lacks a dynamic threshold determination method that can adapt to different hail storm types.

[0007] With the development of X-band dual-polarization phased array radar technology, its minute-level scanning cycle and hundred-meter-level spatial resolution have made real-time tracking of hail embryo particles possible. How to make full use of high-resolution radar data and establish a real-time, particle-level hail embryo full life cycle tracking method has become an urgent technical problem to be solved in this field. Summary of the Invention

[0008] The purpose of this invention is to design a hail identification and tracking method and system for the entire life cycle of hail embryos in order to solve the above-mentioned problems.

[0009] The present invention achieves the above objectives through the following technical solutions:

[0010] A method for hail identification and tracking throughout the entire life cycle of hail embryos, characterized by comprising:

[0011] S1. Obtain the dual-polarization radar volume scan data of the target area at time t;

[0012] S2. Identify all hail embryo particles in dual-polarization radar volume scan data;

[0013] S3. Determine if there are any new hail embryo particles. If so, use the hail embryo particle as the target particle, create a life file for the target particle, and then proceed to S4; otherwise, proceed directly to S4.

[0014] S4, the previous moment The target particle and the target hail embryo particle at time t are spatiotemporally correlated and tracked, and the life file is updated. This refers to the radar scan time interval;

[0015] S5. Determine whether the target particle has been transformed into hail based on the current transformation rule. If so, treat the target particle as a hail particle, update the life file of the hail particle, and then proceed to S6; otherwise, return directly to S1.

[0016] S6. Obtain the current life record of the hail particles;

[0017] S7. Determine if the hail particles have triggered the hailfall condition. If yes, proceed to S8; otherwise, return to S1.

[0018] S8. Analyze the location of the landing area of ​​hail particles that trigger hailfall conditions;

[0019] S9. Spatial clustering of the landing areas of all hail particles to generate a hail landing area probability map and generate early warning information.

[0020] A hail identification and tracking system covering the entire life cycle of hail embryos, used to implement the aforementioned hail identification and tracking method covering the entire life cycle of hail embryos, comprising:

[0021] Acquisition module; The acquisition module is used to acquire dual-polarization radar volume scan data of the target area over a continuous period of time;

[0022] Identification module; The identification module is used to identify all hail embryo particles in dual-polarization radar volume scan data;

[0023] The first analysis module is used to determine whether each hail embryo particle is a new particle. If so, the hail embryo particle is used as the target particle.

[0024] The creation module is used to create the life profile of the target particle.

[0025] Storage module; The storage module is used to store all life records;

[0026] Tracking module; The tracking module is used to track each target particle and update the life profile;

[0027] The second analysis module is used to determine whether the target particle has reached the extinction condition. If so, the life record update of the extinct target particle is terminated.

[0028] The third analysis module is used to determine whether the transformation law self-learning is triggered after all target particles have completed the current round of tracking and elimination judgment; if so, the transformation law is learned according to the complete life cycle life file sealed in the storage module, and the current transformation law is updated; if not, the current effective transformation law is maintained.

[0029] The fourth analysis module is used to determine whether the target particle has been transformed into hail based on the current transformation pattern. If so, the target particle is treated as a hail particle, and the life file of the hail particle is updated.

[0030] The fifth analysis module is used to determine whether hail particles trigger the conditions for hailfall.

[0031] The positioning module is used to analyze the life records of hail particles that trigger hail conditions to locate the landing area. The warning generation module is used to spatially cluster the landing area locations of all hail particles, generate a hail landing area probability map, and generate warning information. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of a hail identification and tracking method for the entire life cycle of hail embryos according to the present invention;

[0033] Figure 2 The combined reflectance map and reflectance factor provided in the embodiments of the present invention Vertical cross-sectional view;

[0034] Figure 3 This is a vertical cross-sectional view and differential reflectance of the particle phase state recognition HID results provided in this embodiment of the invention. Vertical cross-sectional view;

[0035] Figure 4 The specific differential phase shift rate provided in the embodiments of the present invention Vertical cross-sectional view;

[0036] Figure 5 This is a box plot showing the distribution of polymer AG and ice crystal CR particle numbers at various temperature levels from 08:14 to 08:19 UTC, provided in this embodiment of the invention.

[0037] Figure 6 This is a box plot of hail RH and high-density graupel HDG distribution at various temperature levels from 08:14 to 08:19 UTC provided in this embodiment of the invention;

[0038] Figure 7 This is a box plot showing the distribution of low-density graupel (LDG) and rain (RA) particle numbers at various temperature levels from 08:14 to 08:19 UTC, provided in this embodiment of the invention.

[0039] Figure 8 The curves showing the change in particle number over time for low-density graupel (LDG), high-density graupel (HDG), sleet (RH), and rainwater (RA) provided in the embodiments of the present invention are shown.

[0040] Figure 9 This invention provides the instantaneous growth rate curves of low-density sleet (LDG), high-density sleet (HDG), rain sleet (RH), and rainwater (RA) over time.

[0041] Figure 10 This is the evolution sequence of the maximum vertical velocity and its altitude provided in the embodiments of the present invention;

[0042] Figure 11 It is a time sequence showing the change of the maximum horizontal projected area of ​​the region with a vertical velocity greater than 5 m / s.

[0043] The corresponding figure labels are:

[0044] in, Figure 2 In the diagram, (a1, b1) represent the combined reflectance maps at 08:16 UTC and 08:18 UTC, respectively, and the black arrows indicate the direction of the analysis profile. Figure 2 In the equation (a2, b2), the reflectivity factor is the value along the direction of the arrow at the corresponding time. Vertical section; Figure 3 (a3, b3) in the figure represents the vertical profile of the particle phase identification HID result along the direction of the arrow at the corresponding time. Figure 3 In the equation (a4, b4), the differential reflectance along the direction of the arrow at the corresponding time is... Vertical section; Figure 4 In the equation (a5, b5), the differential phase shift rate along the direction of the arrow at the corresponding time point is... Vertical profile. The contour lines in all profile diagrams represent the vertical velocities obtained through inversion, with reddish-brown representing updrafts (positive values) and blue representing downdrafts (negative values), visually demonstrating the synergistic relationship between the power core and microphysical characteristics. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0046] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0047] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0048] In the description of this invention, it should be understood that the terms "upper," "lower," "inner," "outer," "left," "right," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use, or the orientation or positional relationship commonly understood by those skilled in the art. They are only used to facilitate the description of this invention and to simplify the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0049] Furthermore, the terms "first," "second," etc., are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0050] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, terms such as "set" and "connection" should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0051] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0052] The altitude of a key isotherm at 5°C intervals is calculated by linear interpolation from the radiosonde data at the most recent moment, such as -30°C, -25°C, -20°C, -15°C, -10°C, -5°C, 0°C, etc., thereby dividing the vertical height into several temperature layer intervals of 5°C.

[0053] like Figure 1 As shown, a hail identification and tracking method covering the entire life cycle of hail embryos includes:

[0054] S1. Acquire X-band dual-polarization radar volume scan data of the target area over a continuous time period. The dual-polarization radar volume scan data includes horizontal reflectivity. Differential reflectivity , differential phase Correlation coefficient CC and radial velocity The dual-polarization radar volume scan data undergoes quality control and necessary attenuation correction, and is then interpolated to a unified three-dimensional mesh, which can have a horizontal resolution of 100m and a vertical resolution of 100m, to obtain a preprocessed three-dimensional meshed dataset.

[0055] The continuous time period should cover the entire process of hailstorm from generation and development to dissipation, and the temporal resolution should meet the tracking requirements. It is preferred to use phased array radar data with a scanning cycle of ≤1 minute.

[0056] S2. Identify all hail embryo particles in X-band dual-polarization radar volume scan data.

[0057] Specifically, aquatic particles are identified into seven categories: drizzle (DR), rain (RA), polymeric argillaceous (AG), low-density graupel (LDG), high-density graupel (HDG), hail (RH), and ice crystals (CR). The dual polarization parameter (horizontal reflectance) is also considered. Differential reflectivity , differential phase Five variables—correlation coefficient (CC), temperature (T), and so on—are used as key criteria for identifying aquatic particles. A fuzzy logic algorithm is employed, primarily relying on an asymmetric trapezoidal function for identification. The function formula is as follows:

[0058] ;

[0059] in , , , As a function threshold, for the same polarization parameter of the same type, satisfying η is the variable value, and U is the calculated membership degree of a certain type of aquatic particle. The function threshold is set as shown in Table 1. Based on the radar's ability to identify different polarization parameters and the relationship between each polarization parameter and the particles, the polarization parameters are respectively... , , The weighting coefficients c for C, CC, and temperature are set as shown in Table 2. The weighted summation of U for each of the five parameters at each grid point yields the overall matching score. The calculation formula is: , where c is the weighting coefficient and U is the membership degree of different particle categories. The particle category with the highest numerical value is determined as the dominant particle at that grid point. Low-density graupel (LDG) and high-density graupel (HDG) are identified as hail embryo particles, and RH is identified as hail particles.

[0060] Table 1 shows the threshold settings for the membership function of aquatic particles.

[0061]

[0062] Table 2 shows the weighting coefficients c for each aquatic particle.

[0063]

[0064] S3. Determine if there are new hail embryo particles. If so, use the hail embryo particle as the target particle, create a life file for the target particle, and then proceed to S4; otherwise, proceed directly to S4. Specifically, when the hail embryo particle at time t is compared with the one at time t... The distance between all target particles at time t is greater than the distance threshold. Then the hail embryo particle is a new hail embryo particle; distance threshold The value is 1000m. When creating the life record of the target particle, the initial time is recorded. three-dimensional position polarization parameters And extracting the ambient temperature at that location from the environmental data. and vertical airflow velocity Polarization parameters Including horizontal reflectivity Differential reflectivity , differential phase Correlation coefficient Phase type And extract the ambient temperature at that location from the environmental data. and vertical airflow velocity As the initial record of life;

[0065] S4, the The target particle at time t is spatiotemporally correlated with the hail embryo particle at time t and its life record is updated; specifically including:

[0066] S401, Predicting the first The predicted position of the target particle at time t. Radar scan time interval Specifically, based on the inverted three-dimensional wind field, from the target particle at the... Location at any moment and particle composite velocity vector Predict its expected position at time t. The three-dimensional wind field includes the east-west component u and the north-south component v of the horizontal wind speed, as well as the vertical airflow velocity. It can be obtained through multi-radar three-dimensional variational assimilation, dual-Doppler radar wind field inversion, or single-radar VAD technology; The particle composite velocity vector is determined by the horizontal wind speed { } and the vertical airflow velocity of the environment With the final velocity of the falling particle Synthesis, i.e. ;

[0067] S402, Based on the predicted location Define the search window; specifically: based on the predicted location. Centered on the motion uncertainty, a three-dimensional search window is set up, where uncertainty refers to the uncertainty based on the third... The range of deviation between the spatial position of hail embryo particles predicted by radar observation and wind field data at time t and the actual three-dimensional spatial position of the particle at time t is used to quantify the prediction error of the hail embryo particle trajectory and to set a reasonable search space. The deviation mainly originates from deviations in wind field inversion and particle velocity calculation, as well as position calculation deviations caused by turbulent disturbances within the hailstorm. The window size is dynamically adjusted according to the wind speed and radar scanning interval; in this embodiment, 500m-1000m is selected.

[0068] S403. Within the search window, all hail embryo particles existing at time t are considered as candidate particles.

[0069] S404. Analyze the overall matching degree M between each candidate particle and the target particle; expressed as: Where α, β, and γ are the position matching degrees, respectively. Polarization evolution matching degree Matching degree with physical constraints The weight, , , d is the distance between the actual location and the predicted location. Polarization parameters The maximum possible change per unit time, where e is the natural constant with a value of 2.71. Polarization parameters The weight, The location uncertainty scale is dynamically adjusted based on wind speed and radar scanning interval, with a value ranging from 200m to 500m; horizontal reflectivity. of 5 dBZ / min, differential reflectivity of 0.5 dB / min, compared to differential phase of For 1° / km / minute, the correlation coefficient CC The value is 0.02 / minute, but the specific value can be adjusted according to the radar band and seasonal characteristics; physical constraint matching degree. Used to verify whether the evolution direction of parameters conforms to physical laws, such as differential reflectivity. The correlation coefficient (CC) should decrease monotonically as the target particles transform into hail, and should not rebound without cause; the correlation coefficient (CC) should be higher in high-purity phases; the trajectory of the target particles should be coordinated with the wind field, and there should be no sudden changes that do not conform to dynamics; when the evolution direction conforms to physical laws, Otherwise, reduce;

[0070] satisfy In this embodiment, take It emphasizes the leading role of motion prediction while taking into account the rationality of polarization evolution;

[0071] S405. Determine whether the highest overall matching degree is less than the minimum matching degree threshold. If yes, the tracking is interrupted, the interruption count of the target particle is incremented by 1, and the process proceeds to S406; otherwise, the interruption count of the target particle is reduced to 0, and the process proceeds to S407; the minimum matching threshold is set to 0.6.

[0072] S406. Determine whether the number of interruptions of the target particle has reached the preset number. If so, the target particle has reached the extinction condition, and the life record update of the extinct target particle is terminated; otherwise, proceed to S407.

[0073] S407. Select the candidate particle with the highest overall matching degree as the target particle at time t, and update the life record of the target particle.

[0074] S0. Determine if conversion pattern learning has been triggered. If so, perform conversion pattern learning and update the current conversion pattern. (This is repeated every preset time interval.) (e.g., 5 minutes) or when the number of newly added life records exceeds a preset threshold. (For example, when the quantity is 5), pattern learning is triggered;

[0075] Learning the transformation patterns specifically includes:

[0076] 1) Traverse all life records and count the cumulative dwell time of target particles in each temperature range (i.e., the total number of times or the total number of time steps the target particle appears in that range).

[0077] 2) Count the number of events in each temperature range where phase transition (from low-density graupel (LDG) and high-density graupel (HDG) to hail particles) occurs;

[0078] 3) Calculate the conversion event density for each temperature range = number of conversion events / cumulative residence time. The conversion event density reflects the probability of conversion occurring per unit residence time. Considering the differences in particle sample size across different ranges, the conversion event density is weighted by confidence level, and only ranges with cumulative residence time exceeding the global average are retained for comparison.

[0079] 4) The temperature range with the highest density of transformation events is determined as the optimal growth temperature range. For example, the temperature layer interval is divided into ranges every 10℃, and the corresponding height ranges are obtained (L1:-40~-30℃, L2:-30~-20℃, L3:-20~-10℃, L4:-10~0℃...). The temperature layer interval height range where the highest value of the transformation event density calculated in 3) is located is taken as the optimal growth temperature layer.

[0080] 5) Traverse the life records of all target particles that have undergone phase transitions, and extract the vertical airflow velocity at the moment of transition for each target particle. , differential phase and differential reflectivity ;

[0081] 6) The vertical airflow velocity at the moment of conversion The peak value or 85th percentile of the cumulative distribution is used as the critical conversion rate. ; the distribution ratio differential phase The 95th percentile of the cumulative distribution is used as the phase of the maximum ratio difference. ; Maximum differential reflectivity The 95th percentile of the cumulative absolute value distribution is used as the maximum differential reflectance. ;

[0082] 7) Utilizing the optimal growth temperature layer Critical conversion rate Maximum ratio differential phase and maximum differential reflectivity Update the current learning pattern, represented as .

[0083] This step employs a multi-level progressive association algorithm to associate radar echo signals from different times with the same hail embryo particle. This algorithm fully considers the significant changes in the physical properties (size, shape, density, surface condition) of the hail embryo during its growth process, allowing parameters to evolve reasonably under the constraints of physical laws, rather than requiring strict equality.

[0084] S5. Determine whether the target particle has been transformed into hail based on the current transformation rule. If so, treat the target particle as a hail particle, update the life file of the hail particle, and then proceed to S6; otherwise, return directly to S1.

[0085] S501. Analyze the temperature suitability based on the target particle's state data. Power suitability and physical suitability The status data includes the height h(t) and the temperature range to which h(t) belongs. Vertical airflow velocity , differential phase and differential reflectivity , , ,in, The critical conversion rate, For the maximum ratio differential phase, Maximum differential reflectance; analysis of temperature suitability. Specifically: If ,but ;like and Adjacent, then ;otherwise , The optimal growth temperature layer;

[0086] S502, Based on temperature suitability Power suitability and physical suitability Calculate the comprehensive conversion index , represented as: ,in, , and Temperature suitability Power suitability and physical suitability The weights, in this embodiment, =0.4, =0.4, =0.2;

[0087] S503, Determining the Comprehensive Conversion Index Is it not less than the conversion threshold? If so, the target particle is transformed into hail, becoming a hail particle. The life record of the hail particle is updated, its current phase type is updated to RH, and the transformation time is marked in the record. Then proceed to S6; otherwise, return to S1; the conversion threshold in this embodiment is... The value is 0.7;

[0088] S6. Obtain the current life record of the hail particles.

[0089] S7. Determine if the hail particles trigger the hailfall condition. If yes, proceed to S10; otherwise, return to S8. The hailfall trigger condition is met when at least two of conditions 1, 2, 3, and 4 are satisfied. Condition 1 is: when the proportion of falling hail particles exceeds a certain threshold. At that time, condition 1 is satisfied; condition 2 is: the average vertical velocity of all hail particles. When the value changes from positive to negative, condition 2 is satisfied; condition 3 is: the rate of change of the spatial distribution centroid height of all hail particles is negative for a consecutive preset number of times, which satisfies condition 3; condition 4 is: the spatial distribution range of hail particles expands, indicating that the updraft weakens the binding of the particles. Therefore, when the spatial distribution variance of hail particles is greater than the variance threshold, the structure of hail particles is loose, which satisfies condition 4. 30%;

[0090] S8. Analyze the location of hailstone impact zones that trigger hailfall conditions; specifically including:

[0091] S801, For hail particles exist The position of the moment is The speed of movement is Assuming the horizontal velocity of the particle remains constant during its fall, while its vertical velocity is affected by gravity, an extrapolation model is constructed. , , , where g is the acceleration due to gravity;

[0092] S802, Analysis At time t, the landing time is obtained. and landing coordinates As the location of the landing area.

[0093] S9. Spatial clustering of all hailstone landing areas to generate a hailstone landing area probability map and generate warning information. The existing density-based spatial clustering algorithm (DBSCAN) is used to perform spatial clustering to complete the spatial clustering of hailstone landing areas. Specifically, this includes the following sub-steps:

[0094] S901. Based on the two-dimensional planar spatial distribution characteristics of hail particle landing areas, two core parameters of the DBSCAN algorithm are set: neighborhood search radius ε and minimum number of neighborhood samples MinPts, which can be adjusted automatically according to radar spatial resolution and particle sample size.

[0095] (1) Neighborhood search radius ε: Set based on radar horizontal resolution, with a default value of 2 times radar horizontal resolution (in this embodiment, radar horizontal resolution is 100m, and the default value of ε is 200m).

[0096] (2) Minimum neighborhood sample number MinPts: Represents the minimum number of particles to form an effective cluster. The default value is set to 20. When the sample size is less than MinPts, it is judged as a discrete noise point and is not included in the calculation of the effective hailfall area.

[0097] S902. Obtain the landing area location data. Convert the landing area location coordinates of all hail particles obtained in S8 into a two-dimensional coordinate point set in a Cartesian coordinate system. .

[0098] S903, DBSCAN spatial clustering specific calculation process

[0099] (1) Calculate spatial distance. For the preprocessed point set Q, Euclidean distance is used as the metric for spatial distance between two points. and The formula for calculating the distance between them is: In the formula, B is the spatial Euclidean distance between the two points.

[0100] (2) Identify the core points. Traverse each point in the point set Q. Statistics Center of the circle The number of landing points contained within a neighborhood of radius. .like Then determine As the core point, it is included in the core point set. The core point is the central framework for forming hail accumulation zones, directly determining the effective hailfall area.

[0101] (3) Generate clusters.

[0102] a. Initialize cluster number k=0, set unvisited point set. ;

[0103] b. From the core point set Randomly select an unvisited core point Mark it as visited, from Remove from the middle and create the current cluster. and will join in ;

[0104] c. Establish a seed queue, including the core points. exist Add all points within the neighborhood to the seed queue;

[0105] d. Traverse each point in the seed queue ,like If it has not been accessed, mark it as accessed. Remove from the middle. If As the core point, then exist All unvisited points within the neighborhood are added to the seed queue; if a point does not yet belong to any cluster, it is added to the current cluster. .

[0106] e. Repeat step d until the seed queue is empty, completing the current clustering. The generation of f is then performed, and k = k + 1 is set before entering f.

[0107] f. Repeat steps b to e until the core point set is reached. All core points were accessed.

[0108] (4) After the noise point processing traversal is completed, it is still retained in Points in the data are identified as discrete noise points, are not included in any clusters, and are not involved in subsequent hailfall probability calculations, thus avoiding interference from discrete particles on the hailfall warning range.

[0109] S904. Generation of Hailfall Area Probability Map Based on Clustering Results

[0110] (1) Delineate the spatial boundaries of clusters: For each valid cluster Extract the extreme values ​​of latitude and longitude of all points within the cluster to determine the rectangular latitude and longitude boundaries of the cluster. The latitude and longitude boundaries of all clusters together constitute the total effective spatial range of this hail warning;

[0111] (2) Constructing an aligned grid base map: Based on the original horizontal resolution of the radar, construct a two-dimensional regular grid that is completely aligned with the radar grid within the total latitude and longitude range of all clusters. Each grid cell is an independent probability calculation unit.

[0112] (3) Calculation of hail probability by grid: Taking the grid cluster density of the cluster as the core, the normalized hail probability of each grid cell is directly calculated. If the grid is located in a cluster Within the range of latitude and longitude, ,in, This represents the total number of grid cells within the cluster to which this grid cell belongs. The maximum number of grid cells across all clusters is used to ensure that the hail probability of all grid cells is uniformly mapped to the 0-100% range; if the grid cell is located in a cluster... Outside the range of latitude and longitude, =0.

[0113] (4) Delineation of graded risk zones: Based on the grid hail probability, three levels of risk zones are delineated:

[0114] High-risk core area: probability ≥60% of the grid area;

[0115] Medium-risk affected areas: 30% ≤ <60% of the grid area;

[0116] Low-risk potential zone: 0 < <30% of the grid area.

[0117] (5) Generation of hail landing probability map: Based on the grid probability calculation results and the boundary of the graded risk area, a gridded hail landing probability map is generated, and the latitude and longitude boundaries and corresponding risk levels of each cluster are marked simultaneously.

[0118] This invention achieves fully automated tracking of the entire process, from particle initialization and identification, spatiotemporal correlation, and file updating to transformation determination and hail warning, by establishing a real-time dynamic life record for each hail embryo particle. This method has the following significant characteristics:

[0119] The multi-level progressive association algorithm effectively solves the tracking problem caused by changes in physical properties during the growth of hail embryos, ensuring the continuity and reliability of tracking.

[0120] The adaptive threshold generation mechanism based on life records enables the method to automatically adapt to different hailstorm types and development stages, avoiding the limitations of fixed thresholds.

[0121] The comprehensive conversion index integrates temperature, dynamics, and microphysical information, improving the accuracy and robustness of conversion identification.

[0122] Hail warnings based on swarm particle profiles can be issued 5-10 minutes in advance, buying valuable time for artificial hail suppression operations.

[0123] Based entirely on radar and radiosonde data, it does not rely on numerical models and is suitable for real-time operational applications.

[0124] A hail identification and tracking system covering the entire life cycle of hail embryos, used to implement the aforementioned hail identification and tracking method covering the entire life cycle of hail embryos, comprising:

[0125] Acquisition module; The acquisition module is used to acquire dual-polarization radar volume scan data of the target area over a continuous period of time;

[0126] Identification module; The identification module is used to identify all hail embryo particles in dual-polarization radar volume scan data;

[0127] The first analysis module is used to determine whether each hail embryo particle is a new particle. If so, the hail embryo particle is used as the target particle.

[0128] The creation module is used to create the life profile of the target particle.

[0129] Storage module; The storage module is used to store all life records;

[0130] Tracking module; The tracking module is used to track each target particle and update the life profile;

[0131] The second analysis module is used to determine whether the target particle has reached the extinction condition. If so, the life record update of the extinct target particle is terminated.

[0132] The third analysis module is used to determine whether the transformation law self-learning is triggered after all target particles have completed the current round of tracking and elimination judgment; if so, the transformation law is learned according to the complete life cycle life file sealed in the storage module, and the core parameters of the phase transformation judgment are updated; if not, the current effective transformation law is maintained.

[0133] The fourth analysis module is used to determine whether the target particle has been transformed into hail based on the current transformation pattern. If so, the target particle is treated as a hail particle, and the life file of the hail particle is updated.

[0134] The fifth analysis module is used to determine whether hail particles trigger hailfall conditions; the positioning module is used to analyze the landing area location based on the life records of hail particles that trigger hailfall conditions; the early warning generation module is used to spatially cluster the landing area locations of all hail particles, generate a hailfall area probability map, and generate early warning information.

[0135] This embodiment demonstrates the physical rationale of the invention through a typical case of a single hailstorm. The experimental hailstorm lasted approximately 50 minutes. The strongest echo reached over 60 dBZ, indicating a single hailstorm that had developed after a rearing process. The hailstones formed at 08:08 UTC, and a strong echo zone greater than 60 dBZ appeared within the hailstones at 08:14 UTC. Hailfall began at 08:20 UTC and ended at 08:25 UTC.

[0136] Figure 2 , Figure 3 , Figure 4This paper presents the comprehensive analysis results of the method of this invention in the key stages of hail formation. Through the coordinated display of reflectivity, vertical velocity, dual polarization parameters, and vertical profiles of particle phase states, the spatial overlap between the strong updraft core and the high-reflectivity, graupel, and hail particle regions can be clearly seen. The combined reflectivity at 08:16 and 08:18 UTC, along the direction of the arrow... HID (Hydrometeor Identification) results , Profile observations show that the strongest reflectivity factor of the hailstorm rose above 60 dBZ, and HID identified the formation of a large RH particle swarm within the cloud, indicating that the hail embryos had entered a rapid growth phase. Furthermore, a dynamically strong core region with a velocity w ≥ 15 m / s was formed within each hailstone, with relatively small velocity fluctuations, providing ample time for hail growth. Through the "dynamically critical region" (the area with reddish-brown vertical velocity contours ≥ 5 m / s in the profile), the HID algorithm accurately identified the RH particle swarm, demonstrating the effectiveness of dynamic constraints for accurate hail embryo identification.

[0137] Figure 5 , Figure 6 , Figure 7 The box plots showing the distribution of each particle's quantity across different temperature layers indicate that rain-and-hail (RH) particles are most abundant and most concentrated in the -20 to -15℃ layer (with the smallest box height), followed by the -15 to -10℃ layer. This characteristic not only verifies that the suitable temperature range for hail embryo growth is -40 to -10℃, but also further clarifies that -20 to -10℃ is the core growth region. Moreover, the height of this temperature range matches the core region of the strong updraft (w≥20m / s) in the dynamic analysis, where hail embryos fully grow and transform into hail particles.

[0138] Figure 8 , Figure 9 , Figure 10 and Figure 11 This study demonstrates the multi-parameter temporal evolution during a complete hail embryo growth-transformation-fall process, verifying the effectiveness of the dual-parameter hail determination rule and fusion diagnostic system of this invention. Figure 8 The evolution of particle numbers shows that the period from 08:08 to 08:14 UTC was the rapid formation stage of hail embryos, with a continuous increase in hail numbers. This indicates that supercooled water droplets within the cloud froze extensively on the surface of the hail embryos through frost adhesion, leading to a significant increase in the number of hail embryos (LDG, HDG) and hail. Figure 9 The evolution of the growth rate of various particle types shows that the hail number growth rate peaked at 08:15 UTC, followed by a sharp drop, marking the appearance of the critical point for hail embryo transformation. At this time, the growth rate of high-density graupel (HDG) turned from positive to negative. The simultaneous decrease in the rate of increase of low-density graupel (LDG) indicates that a large number of hail embryo graupel particles have begun to transform into hail (RH). Figure 10 The sequence of maximum vertical velocities corresponding to reflectance greater than 40 dBZ shows that as the cell develops, the maximum vertical velocity and its altitude increase synchronously. The maximum vertical wind speed is reached at 08:20 UTC, which also coincides with the highest altitude at which the maximum vertical velocity occurs during the cell's development, corresponding to an altitude of 5.5 km. Afterward, the vertical wind speed weakens, and the altitude of its maximum value generally shows a downward trend. 08:20 UTC happens to be the start time of hailfall, indicating that after the updraft weakens, large hailstones lose their lifting force and fall. Figure 11 The evolution sequence of the maximum area of ​​the vertical updraft velocity greater than 5 m / s shows that the maximum area of ​​the main updraft occurred at 08:16 UTC. Therefore, it is clear that around 08:15 UTC is the critical time point when hail embryos begin to transform into large quantities of hail. This comprehensively demonstrates the physical principle that dynamic attenuation and microphysical transformation jointly determine the onset of hailfall.

[0139] In summary, this invention, through the specific implementation steps described above, constructs an automated business process from data preprocessing, object identification and tracking, targeted diagnosis, to fusion early warning. All core parameters (such as seed point threshold, clustering weight, scoring threshold, and early warning threshold) can be calibrated based on statistical analysis of local historical hailstorm cases, ensuring the physicality and regional adaptability of the method. The above content is merely an embodiment of this invention, and its scope of protection is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in this invention, under its spirit and principles, should be considered as covered within the scope of protection of this invention.

[0140] The technical solutions of the present invention are not limited to the specific embodiments described above. Any technical modifications made in accordance with the technical solutions of the present invention fall within the protection scope of the present invention.

Claims

1. A method for hail identification and tracking throughout the entire life cycle of hail embryos, characterized in that, include: S1. Acquire X-band dual-polarization radar volume scan data of the target area at time t; S2. Identify all hail embryo particles in the X-band dual-polarization radar volume scan data; S3. Determine if there are any new hail embryo particles. If so, use the hail embryo particle as the target particle, create a life file for the target particle, and then proceed to S4; otherwise, proceed directly to S4. S4, the previous moment The target particle and the target hail embryo particle at time t are spatiotemporally correlated and tracked, and the life file is updated. This refers to the radar scan time interval; S5. Determine whether the target particle has been transformed into hail based on the current transformation rule. If so, treat the target particle as a hail particle, update the life file of the hail particle, and then proceed to S6; otherwise, return directly to S1. S6. Obtain the current life record of the hail particles; S7. Determine if the hail particles have triggered the hailfall condition. If yes, proceed to S8; otherwise, return to S1. S8. Analyze the location of the landing area of ​​hail particles that trigger hailfall conditions; S9. Spatial clustering of the landing areas of all hail particles to generate a hail landing area probability map and generate early warning information.

2. The hail identification and tracking method for the entire life cycle of hail embryos according to claim 1, characterized in that, In S3, when the hail embryo particle at time t is related to the first time... The distance between all target particles at time t is greater than the distance threshold. Then the hail embryo particle is a new hail embryo particle.

3. The hail identification and tracking method for the entire life cycle of hail embryos according to claim 1, characterized in that, S4 includes: S401, Predicting the first The predicted position of the target particle at time t. This refers to the radar scan time interval; S402. Determine the search window based on the predicted location; S403. Within the search window, all hail embryo particles existing at time t are considered as candidate particles. S404. Analyze the overall matching degree M between each candidate particle and the target particle; S405. Determine whether the highest overall matching degree is less than the minimum matching degree threshold. If yes, the tracking is interrupted, the interruption count of the target particle is incremented by 1, and the process proceeds to S406; otherwise, the interruption count of the target particle is reduced to 0, and the process proceeds to S407. S406. Determine whether the target particle's interruption count has reached the preset number. If yes, terminate the target particle's life record update; otherwise, proceed to S407. S407. Select the candidate particle with the highest overall matching degree as the target particle at time t, and update the life record of the target particle.

4. The hail identification and tracking method for the entire life cycle of hail embryos according to claim 3, characterized in that, In S404, the overall matching degree M is represented as: Where α, β, and γ are the position matching degrees, respectively. Polarization evolution matching degree Matching degree with physical constraints The weight, , , d is the distance between the actual location and the predicted location. Polarization parameters The maximum possible change per unit time. Polarization parameters The weight, Let be the scale of positional uncertainty, and e be the natural constant.

5. The hail identification and tracking method for the entire life cycle of hail embryos according to claim 1, characterized in that, S5 includes: S501. Analyze the temperature suitability based on the target particle's state data. Power suitability and physical suitability Status data includes height. , Temperature range Vertical airflow velocity , differential phase and differential reflectivity , , ,in, The critical conversion rate, For the maximum ratio differential phase, Maximum differential reflectance; analysis of temperature suitability. Specifically: If ,but ;like and Adjacent, then ;otherwise , The optimal growth temperature layer; S502, Based on temperature suitability Power suitability and physical suitability Calculate the comprehensive conversion index , is represented as: ,in, , and Temperature suitability Power suitability and physical suitability The weights; S503, Determining the Comprehensive Conversion Index Is it not less than the conversion threshold? If so, the target particle is transformed into hail, and the target particle is made into a hail particle. The life file of the hail particle is updated, and then the process proceeds to S6; otherwise, the process returns to S1.

6. The hail identification and tracking method for the entire life cycle of hail embryos according to claim 1, characterized in that, In S7, the hail trigger condition is met when at least two of conditions 1, 2, 3, and 4 are satisfied; condition 1 is: when the proportion of falling hail particles exceeds a preset threshold. At that time, condition 1 is satisfied; condition 2 is: the average vertical velocity of all hail particles. When the value changes from positive to negative, condition 2 is satisfied; condition 3 is: the rate of change of the centroid height of the spatial distribution of all hail particles is negative for a consecutive preset number of times, which satisfies condition 3; condition 4 is: when the variance of the spatial distribution of hail particles is greater than the variance threshold, which satisfies condition 4.

7. The hail identification and tracking method for the entire life cycle of hail embryos according to claim 1, characterized in that, S10 includes: S801, for hail particles i in The position of the moment is The speed of movement is Assuming the horizontal velocity of the particle remains constant during its fall, while its vertical velocity is affected by gravity, an extrapolation model is constructed. , , Where g is the acceleration due to gravity; S802, Analysis At time t, the landing time is obtained. and landing coordinates As the location of the landing area.

8. The hail identification and tracking method for the entire life cycle of hail embryos according to claim 1, characterized in that, Between S4 and S5, there is also S0, which determines whether conversion pattern learning has been triggered. If so, conversion pattern learning is performed, and the current conversion pattern is updated. Specifically, conversion pattern learning includes: 1) Traverse all life records and calculate the cumulative dwell time of target particles within each temperature range; 2) Count the number of phase transition events occurring within each temperature range; 3) Calculate the conversion event density for each temperature zone = number of conversion events / cumulative dwell time; 4) The temperature range with the highest density of transformation events is determined as the optimal growth temperature range. ; 5) Traverse the life records of all target particles that have undergone phase transitions, and extract the vertical airflow velocity at the moment of transition for each target particle. , differential phase and differential reflectivity ; 6) The vertical airflow velocity at the moment of conversion The peak value or 85th percentile of the cumulative distribution is used as the critical conversion rate. ; the distribution ratio differential phase The 95th percentile of the cumulative distribution is used as the phase of the maximum ratio difference. ; Maximum differential reflectivity The 95th percentile of the cumulative absolute value distribution is used as the maximum differential reflectance. ; 7) Utilize the optimal growth temperature layer Critical conversion rate Maximum ratio differential phase and maximum differential reflectivity Update the current learning pattern, represented as .

9. A hail identification and tracking system for the entire life cycle of hail embryos, used to implement the hail identification and tracking method for the entire life cycle of hail embryos as described in any one of claims 1-8, characterized in that, include: Data acquisition module; The acquisition module is used to acquire dual-polarization radar volume scan data of the target area over a continuous period of time; Recognition module; The identification module is used to identify all hail embryo particles in dual-polarization radar volume scan data; The first analysis module is used to determine whether each hail embryo particle is a new particle. If so, the hail embryo particle is used as the target particle. The creation module is used to create the life profile of the target particle. Storage module; The storage module is used to store all life records; Tracking module; The tracking module is used to track each target particle and update the life profile; Second analysis module; The second analysis module is used to determine whether the target particle has reached the extinction condition. If so, the life record update of the extinct target particle is terminated. The third analysis module is used to determine whether the transformation law self-learning is triggered after all target particles have completed the current round of tracking and elimination judgment; if so, the transformation law is learned according to the complete life cycle life file sealed in the storage module, and the current transformation law is updated; if not, the current effective transformation law is maintained. The fourth analysis module is used to determine whether the target particle has been transformed into hail based on the current transformation pattern. If so, the target particle is treated as a hail particle, and the life file of the hail particle is updated. The fifth analysis module is used to determine whether hail particles trigger the conditions for hailfall. Location module; the location module is used to analyze the life records of hail particles that trigger hailfall conditions to locate the impact area; early warning generation module; The warning generation module is used to locate the landing areas of all hail particles, perform spatial clustering, generate a hail landing area probability map, and generate warning information.

Images