Lightning ground lightning activity track correction method and device, electronic equipment and storage medium

Abstract

The invention discloses a lightning-to-ground lightning activity track correction method and device, electronic equipment and a storage medium, and is used for solving the technical problem that a current lightning-to-ground lightning activity track determination method depends on a single lightning positioning data source and is low in precision. The method comprises the following steps: acquiring meteorological data of a target area and lightning positioning data for lightning ground lightning activities in the target area; wherein the meteorological data comprises satellite data and radar data; determining a lightning initial centroid trajectory according to the lightning positioning data; generating a first centroid correction track corresponding to the satellite data, and generating a second centroid correction track corresponding to the radar data; evaluating the first centroid correction track and the second centroid correction track according to a pre-constructed track evaluation model, and outputting a track correction strategy; and correcting the lightning initial centroid trajectory according to a trajectory correction strategy to obtain a lightning corrected centroid trajectory of the lightning ground-to-ground lightning activity in the target area.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a lightning ground flash activity track correction method and device, electronic equipment and a storage medium. BACKGROUND

[0002] As one of the most destructive meteorological phenomena in nature, lightning disasters pose a serious threat to the safety of power systems. Accurate identification of lightning ground flash activity tracks is not only a key step in improving lightning warning capabilities, but also an important part of building a lightning protection system. By establishing a lightning monitoring network, real-time monitoring and rapid response mechanisms for lightning activity can be achieved, effectively preventing and reducing damage to power transmission and distribution equipment caused by lightning. This ensures the stable operation and reliability of the power system.

[0003] Currently, the determination method of lightning ground flash activity track mainly relies on lightning location data obtained by lightning location systems. Specifically, the track path is extracted by analyzing the spatial distribution of lightning points. However, this method is completely based on a single data source of lightning location data, without introducing any physical observation data to physically constrain or temporally and spatially correct the track identification results. This leads to the phenomenon of track drift, jumping or deviation from the actual thunderstorm main path when facing non-uniform distribution, density mutation or boundary ambiguity of lightning points in strong convective systems.

[0004] The above-mentioned track determination method relying on a single lightning location data source has the problem of low precision, which is not conducive to lightning disaster analysis and subsequent warning. SUMMARY

[0005] The present application provides a lightning ground flash activity track correction method, device, electronic equipment and storage medium, which is used to solve or partially solve the technical problem of low precision of current lightning ground flash activity track determination methods relying on a single lightning location data source.

[0006] The present application provides a lightning ground flash activity track correction method, which comprises:

[0007] Obtaining meteorological data of a target area and lightning location data of lightning ground flash activity in the target area; the meteorological data includes satellite data and radar data;

[0008] Determining a lightning initial centroid track according to the lightning location data;

[0009] Generating a first centroid correction track corresponding to the satellite data and a second centroid correction track corresponding to the radar data;

[0010] According to the pre-constructed trajectory evaluation model, the first mass center correction trajectory and the second mass center correction trajectory are evaluated, and a trajectory correction strategy is output;

[0011] According to the trajectory correction strategy, the lightning initial mass center trajectory is corrected to obtain a lightning correction mass center trajectory of lightning ground flash activity in the target area.

[0012] Optionally, the first mass center correction trajectory corresponding to the satellite data is generated, comprising:

[0013] According to the satellite data, the cloud top physical temperature, the cloud top altitude and the satellite image of the cloud cluster in the target area are determined;

[0014] The satellite image is spatially cropped to obtain a lightning ground flash activity area;

[0015] Based on the cloud top physical temperature and the cloud top altitude, the cloud top cooling rate of the cloud cluster corresponding to the lightning ground flash activity area is determined;

[0016] If the cloud top cooling rate of the cloud cluster corresponding to the lightning ground flash activity area is less than or equal to the cloud top cooling rate of the first cloud cluster adjacent to the lightning ground flash activity area, the initial mass center corresponding to the lightning initial mass center trajectory is offset to the center direction of the second cloud cluster, to obtain a first mass center, and a first mass center correction trajectory is generated;

[0017] The first mass center correction trajectory includes the first mass center; the second cloud cluster is adjacent to the lightning ground flash activity area; the cloud top cooling rate of the second cloud cluster is greater than the cloud top cooling rate of the cloud cluster corresponding to the lightning ground flash activity area.

[0018] Optionally, the method further comprises:

[0019] Based on the cloud top physical temperature and the cloud top altitude, the temperature gradient of the cloud cluster corresponding to the lightning ground flash activity area is determined;

[0020] According to the cloud top altitude and the temperature gradient, the lightning density of a third cloud cluster and the lightning density of a fourth cloud cluster are determined; the third cloud cluster and the fourth cloud cluster are both adjacent to the cloud cluster corresponding to the lightning ground flash activity area;

[0021] If the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area is higher than that of the third cloud cluster, and the lightning density of the cloud cluster corresponding to the lightning flash activity area is greater than that of the third cloud cluster, then the initial centroid corresponding to the initial centroid trajectory of the lightning is shifted towards the center of the fourth cloud cluster to obtain the second centroid, and the first centroid correction trajectory is updated.

[0022] Among them, the updated first centroid correction trajectory; the cloud top physical temperature of the fourth cloud cluster is lower than the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area; the lightning density of the fourth cloud cluster is greater than the lightning density of the cloud cluster corresponding to the lightning flash activity area.

[0023] Optionally, generating the second centroid-corrected trajectory corresponding to the radar data includes:

[0024] Based on the radar data, determine the combined reflectivity, echo top height, and vertical integrated liquid water content;

[0025] Lightning observation points are extracted from cloud clusters formed within each time window and mapped to the radar grid cells where the lightning observation points are located;

[0026] At least one target radar grid cell is selected from multiple radar grid cells when the combined reflectivity, echo top height and vertical integrated liquid water content meet the preset threshold conditions.

[0027] Each of the target radar grid units is taken as a strong convection core observation point. For each strong convection core observation point, a physical weight value is determined based on the combined reflectivity, echo top height and vertical integrated liquid water content corresponding to the strong convection core observation point.

[0028] Based on each of the physical weight values, the latitude and longitude coordinates of the target radar grid cell are weighted and averaged to generate a weighted centroid.

[0029] The weighted centroids under each time window are connected in chronological order to obtain the second centroid correction trajectory.

[0030] Optionally, the step of evaluating the first centroid-corrected trajectory and the second centroid-corrected trajectory according to a pre-built trajectory evaluation model and outputting a trajectory correction strategy includes:

[0031] The first centroid-corrected trajectory and the second centroid-corrected trajectory are preprocessed and then input into a pre-constructed trajectory evaluation model; the trajectory evaluation model includes a trajectory physical index embedding unit and a weight evaluation unit;

[0032] The trajectory physical index embedding unit extracts the persistence index, stability index, and structural complexity index corresponding to the first centroid corrected trajectory and the second centroid corrected trajectory under each time window.

[0033] Based on each of the aforementioned persistence indicators, stability indicators, and structural complexity indicators, the weight evaluation unit performs score aggregation based on a preset trajectory physical indicator weight function to obtain the score result.

[0034] If the scoring result indicates that the score of the second centroid correction trajectory is higher than the preset scoring threshold, and the time window of the second centroid correction trajectory is not missing, then the second centroid correction trajectory is determined as the target correction trajectory, and a trajectory correction strategy is generated based on the target correction trajectory.

[0035] Optionally, the trajectory evaluation model further includes a dynamic switching strategy unit; the method further includes:

[0036] If the scoring result indicates that the score of the second centroid correction trajectory is lower than the preset scoring threshold, or if the time window of the second centroid correction trajectory is missing, then the dynamic switching strategy unit calls the first centroid correction trajectory to compensate the second centroid correction trajectory under the same time window to obtain the target correction trajectory, and generates a trajectory correction strategy based on the target correction trajectory.

[0037] Optionally, the extraction process for the persistence index, stability index, and structural complexity index corresponding to the first centroid correction trajectory and the second centroid correction trajectory under each time window includes:

[0038] The first centroid correction trajectory and the second centroid correction trajectory are sequentially used as the correction trajectories to be evaluated.

[0039] Based on the trajectory to be evaluated and corrected, determine the time span, trajectory duration, lightning event variation, and average lightning event variation of the lightning cluster from the starting point to the ending point.

[0040] Based on the time span, determine the persistence index of the correction trajectory to be evaluated;

[0041] Based on the trajectory duration, the amount of change in the lightning event, and the average amount of change in the lightning event, a stability index for the trajectory to be evaluated and corrected is determined.

[0042] Based on the revised trajectory to be evaluated, determine the average number of thunderstorm splitting and merging events;

[0043] Based on the average number of times, the structural complexity index of the modified trajectory to be evaluated is determined.

[0044] The present invention also provides a lightning flash trajectory correction device, comprising:

[0045] The data acquisition unit is used to acquire meteorological data of the target area, as well as lightning location data for lightning and ground flash activity within the target area; the meteorological data includes satellite data and radar data.

[0046] The initial centroid trajectory determination unit is used to determine the initial centroid trajectory of the lightning based on the lightning location data.

[0047] The trajectory correction generation unit is used to generate a first centroid correction trajectory corresponding to the satellite data and a second centroid correction trajectory corresponding to the radar data.

[0048] The trajectory correction strategy output unit is used to evaluate the first centroid correction trajectory and the second centroid correction trajectory according to the pre-built trajectory evaluation model, and output the trajectory correction strategy.

[0049] The centroid trajectory correction unit is used to correct the initial centroid trajectory of lightning according to the trajectory correction strategy, so as to obtain the corrected centroid trajectory of lightning activity in the target area.

[0050] The present invention also provides an electronic device, the device comprising a processor and a memory:

[0051] The memory is used to store program code and transmit the program code to the processor;

[0052] The processor is used to execute the lightning and ground flash activity trajectory correction method as described above, according to the instructions in the program code.

[0053] The present invention also provides a computer-readable storage medium for storing program code for executing the lightning-to-ground flash activity trajectory correction method as described in any of the preceding claims.

[0054] As can be seen from the above technical solutions, the present invention has the following advantages:

[0055] A method for correcting lightning-to-ground lightning activity trajectories based on meteorological data is provided. Meteorological data of the target area, as well as lightning location data for lightning-to-ground lightning activity within the target area, are acquired. The meteorological data includes satellite data and radar data. Based on the lightning location data, the initial centroid trajectory of the lightning is determined. A first centroid-corrected trajectory corresponding to the satellite data is generated, and a second centroid-corrected trajectory corresponding to the radar data is generated. By introducing multi-source meteorological data, including satellite and radar data, to construct the initial centroid trajectory based on the lightning location data, two types of corrected centroid trajectories with physical constraints are generated, effectively overcoming the problem of trajectory drift or distortion caused by a single lightning location data source under abnormal distribution, unstable accuracy, and systematic errors. Then, the first and second centroid-corrected trajectories are evaluated according to a pre-constructed trajectory evaluation model, and a trajectory correction strategy is output. The initial centroid trajectory is then corrected according to the trajectory correction strategy to obtain the corrected centroid trajectory of lightning-to-ground lightning activity within the target area. This invention utilizes a trajectory evaluation model to quantitatively assess and compare the two constructed corrected trajectories, outputting a trajectory correction strategy that conforms to the laws of physical evolution, thereby achieving dynamic optimization and correction of the initial centroid trajectory of lightning. The technical solution provided by this invention significantly improves the spatial accuracy, evolutionary coherence, and physical rationality of lightning trajectories. It facilitates the correction of trajectories determined from lightning location data, improving the accuracy of lightning flash activity trajectory determination. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 A flowchart illustrating the steps of a method for correcting the trajectory of lightning strikes to the ground.

[0058] Figure 2 A structural block diagram of a lightning flashover trajectory correction device;

[0059] Figure 3 This is a schematic diagram of the structure of an electronic device.

[0060] Explanation of reference numerals in the attached figures: 31, processor; 32, communication bus; 33, user interface; 34, network interface; 35, memory. Detailed Implementation

[0061] This invention provides a method, apparatus, electronic device, and storage medium for correcting lightning flash activity trajectories, which solves or partially solves the technical problem that current methods for determining lightning flash activity trajectories rely on a single lightning location data source and have low accuracy.

[0062] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0063] As an example, lightning strikes pose a significant threat to the safe operation of power systems. Accurate identification of lightning activity trajectories is a core element in improving the efficiency of lightning early warning and response.

[0064] Current lightning trajectory identification methods primarily rely on location data provided by lightning location systems. Specifically, they extract trajectory paths by analyzing the spatial distribution of lightning points. However, these methods heavily depend on lightning location data as a single source of information, failing to incorporate any meteorological or physical observation data to constrain the extraction results physically or correct them on a spatiotemporal scale. Consequently, in complex situations such as uneven lightning point distribution, abrupt density changes, or unclear boundaries in strong convective systems, the obtained trajectories are prone to positional shifts, abrupt jumps, or deviations from the thunderstorm core region.

[0065] It is evident that this type of trajectory recognition method driven by single-source data has significant shortcomings in terms of spatial accuracy, physical rationality, and evolutionary coherence. Its recognition accuracy is low, making it difficult to meet the practical needs of high-precision dynamic monitoring of thunderstorms.

[0066] Therefore, one of the core inventive points of this invention is to propose a method for correcting lightning flash activity trajectories based on meteorological data, addressing the shortcomings of current technologies. By combining meteorological data and correcting the centroid trajectory determined by lightning location data, the accuracy of lightning flash activity trajectory determination is improved.

[0067] Reference Figure 1 The diagram illustrates a flowchart of a lightning flash activity trajectory correction method provided by an embodiment of the present invention, which may specifically include the following steps:

[0068] Step 101: Obtain meteorological data for the target area, as well as lightning location data for lightning and ground flash activity within the target area; the meteorological data includes satellite data and radar data;

[0069] In practical implementation, meteorological data of the target area to be studied, as well as lightning location data for lightning and ground flash activity within that target area, can first be acquired for subsequent calculations. Furthermore, the meteorological data may include satellite data and radar data.

[0070] Step 102: Determine the initial centroid trajectory of the lightning based on the lightning location data;

[0071] In this step, an initial lightning centroid trajectory can be generated based on lightning location data.

[0072] Step 103: Generate the first centroid correction trajectory corresponding to the satellite data, and generate the second centroid correction trajectory corresponding to the radar data;

[0073] In some embodiments, the process of generating the first centroid corrected trajectory corresponding to the satellite data may include the following sub-steps S01 to S04:

[0074] Step S01: Based on satellite data, determine the cloud top physical temperature, cloud top altitude, and satellite imagery of cloud clusters within the target area;

[0075] Step S02: Spatial cropping of satellite images to obtain the lightning flash activity area;

[0076] Step S03: Based on the cloud top physical temperature and cloud top altitude, determine the cloud top cooling rate of the cloud clusters corresponding to the lightning flash activity area;

[0077] Step S04: If the cloud top cooling rate of the cloud cluster corresponding to the lightning flash activity area is less than or equal to the cloud top cooling rate of the first cloud cluster adjacent to the lightning flash activity area, then the initial centroid corresponding to the initial centroid trajectory of the lightning is shifted towards the center of the second cloud cluster to obtain the first centroid, and the first centroid correction trajectory is generated.

[0078] Among them, the first centroid correction trajectory includes the first centroid; the second cloud cluster is adjacent to the lightning and ground flash activity area; the cloud top cooling rate of the second cloud cluster is greater than the cloud top cooling rate of the cloud cluster corresponding to the lightning and ground flash activity area.

[0079] In reality, there may be more than one secondary cloud cluster near the area of ​​lightning flash activity. In this case, a secondary cloud cluster can be selected based on the closest moving distance, the closest or most reasonable cloud top cooling rate, or other preset rules. It is understood that this invention does not impose any limitations on this.

[0080] During movement, the distance traveled will not exceed the actual spatial distance between the current centroid and the center of the second cloud cluster. This distance can be set to 0.6 to 0.8 times to prevent excessive centroid jumps that could cause trajectory discontinuities. The offset amplitude is driven by the difference in cloud top cooling rates. That is, if the cooling rate of the second cloud cluster is much higher than that of the current cloud cluster, the offset amplitude can be appropriately increased, and vice versa.

[0081] Based on steps S01 to S04, and for the first centroid correction trajectory, this invention further proposes a supplementary or enhancing correction strategy. This correction strategy is implemented by performing the following steps S11 to S13:

[0082] Step S11: Based on the cloud top physical temperature and cloud top altitude, determine the temperature gradient of the cloud clusters corresponding to the lightning flash activity area;

[0083] Step S12: Determine the lightning density of the third cloud cluster and the fourth cloud cluster based on the cloud top altitude and temperature gradient; both the third and fourth cloud clusters are adjacent to the cloud clusters corresponding to the lightning flash activity area.

[0084] Step S13: If the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area is higher than that of the third cloud cluster, and the lightning density of the cloud cluster corresponding to the lightning flash activity area is greater than that of the third cloud cluster, then the initial centroid corresponding to the initial centroid trajectory of the lightning is shifted towards the center of the fourth cloud cluster to obtain the second centroid, and the first centroid correction trajectory is updated.

[0085] Among them, the updated first centroid correction trajectory includes the second centroid; the cloud top physical temperature of the fourth cloud cluster is lower than the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area; the lightning density of the fourth cloud cluster is greater than the lightning density of the cloud cluster corresponding to the lightning flash activity area.

[0086] Steps S11 to S13 are the correction strategies for steps S01 to S04, and the first, second, third, and fourth cloud clusters are all cloud clusters adjacent to the cloud clusters corresponding to the lightning flash activity area. Therefore, in reality, the third and fourth cloud clusters may be the first and second cloud clusters.

[0087] It should be noted that the trajectory generation and correction are performed in chronological order, through a series of time windows. The first centroid correction trajectory generation process, the second centroid correction trajectory generation process, and the subsequent trajectory correction process provided by this invention are all executed based on a series of time windows.

[0088] In the aforementioned correction steps, if the joint judgment condition that "the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area is higher than the cloud top physical temperature of the third cloud cluster, and the lightning density of the cloud cluster corresponding to the lightning flash activity area is greater than the lightning density of the third cloud cluster" cannot be met within the current time window, the correction operation will not be forcibly executed. In this case, the centroid shift action will not be performed, and the correction action based on the mechanism described in steps S11 to S13 of the current time window will be skipped directly, and the next time window will be entered to perform the next round of judgment and correction.

[0089] If the fourth cloud cluster cannot simultaneously satisfy both the conditions of "cloud top physical temperature being lower than the target cloud cluster's cloud top physical temperature" and "lightning density being greater than the target cloud cluster's lightning density," then this fourth cloud cluster will not be used as the target cloud cluster for centroid shifting; that is, the operation of shifting the initial centroid towards its center will not be performed. The handling strategy in this case is to terminate the current round of centroid shifting logic, keep the centroid point of the current time window unchanged, and continue to the next time window for correction and judgment.

[0090] In some embodiments, the process of generating the second centroid correction trajectory corresponding to the radar data may include the following sub-steps S21 to S26:

[0091] Step S21: Determine the combined reflectivity, echo top height, and vertical integrated liquid water content based on radar data;

[0092] Step S22: Extract lightning observation points from the cloud clusters formed within each time window and map them to the radar grid cells where the lightning observation points are located;

[0093] Step S23: Select at least one target radar grid cell from multiple radar grid cells whose combined reflectivity, echo top height, and vertical integrated liquid water content meet the preset threshold conditions.

[0094] Step S24: Take each target radar grid cell as a strong convection core observation point. For each strong convection core observation point, determine the physical weight value based on the combined reflectivity, echo top height and vertical integrated liquid water content corresponding to the strong convection core observation point.

[0095] Step S25: Based on each physical weight value, perform a weighted average of the latitude and longitude coordinates of the target radar grid cell to generate a weighted centroid;

[0096] Step S26: Connect the weighted centroids under each time window in chronological order to obtain the second centroid correction trajectory.

[0097] Step 104: Evaluate the first centroid correction trajectory and the second centroid correction trajectory according to the pre-built trajectory evaluation model, and output the trajectory correction strategy;

[0098] In some embodiments, the process of evaluating the first centroid-corrected trajectory and the second centroid-corrected trajectory according to a pre-built trajectory evaluation model and outputting the trajectory correction strategy may include the following sub-steps S31 to S34:

[0099] Step S31: Input the first centroid corrected trajectory and the second centroid corrected trajectory into the pre-constructed trajectory evaluation model after preprocessing; the trajectory evaluation model includes trajectory physical index embedding unit and weight evaluation unit;

[0100] Step S32: Extract the persistence index, stability index, and structural complexity index of the first centroid corrected trajectory and the second centroid corrected trajectory under each time window through the trajectory physical index embedding unit.

[0101] Step S33: Based on each persistence index, each stability index, and each structural complexity index, the scoring is aggregated by the weight evaluation unit based on the preset trajectory physical index weight function to obtain the scoring result;

[0102] Step S34: If the score of the second centroid correction trajectory is higher than the preset score threshold and the time window of the second centroid correction trajectory is not missing, then the second centroid correction trajectory is determined as the target correction trajectory, and a trajectory correction strategy is generated based on the target correction trajectory.

[0103] Furthermore, the trajectory evaluation model also includes a dynamic switching strategy unit. When the score representing the second centroid correction trajectory is lower than a preset score threshold, or when the time window for the second centroid correction trajectory is missing, the dynamic switching strategy unit calls the first centroid correction trajectory to compensate for the second centroid correction trajectory within the same time window, thereby obtaining the target correction trajectory, and generating a trajectory correction strategy based on the target correction trajectory.

[0104] In some embodiments, the extraction process of the persistence index, stability index, and structural complexity index corresponding to the first centroid correction trajectory and the second centroid correction trajectory under each time window includes the following sub-steps S321 to S326:

[0105] Step S321: Sequentially select the first centroid correction trajectory and the second centroid correction trajectory as the correction trajectories to be evaluated;

[0106] Step S322: Based on the trajectory to be evaluated and corrected, determine the time span, trajectory duration, lightning event variation, and average lightning event variation of the lightning cluster from the starting point to the ending point;

[0107] Step S323: Based on the time span, determine the persistence index of the trajectory to be evaluated for correction;

[0108] Step S324: Based on the trajectory duration, the amount of change in lightning events, and the average amount of change in lightning events, determine the stability index of the trajectory to be evaluated and corrected;

[0109] Step S325: Determine the average number of thunderstorm splitting and merging events based on the corrected trajectory to be evaluated;

[0110] Step S326: Determine the structural complexity index of the modified trajectory to be evaluated based on the average number of times.

[0111] Step 105: Correct the initial centroid trajectory of the lightning according to the trajectory correction strategy to obtain the corrected centroid trajectory of lightning activity in the target area.

[0112] Finally, the initial centroid trajectory of lightning can be corrected according to the output trajectory correction strategy to obtain the corrected centroid trajectory of lightning activity in the target area, thus completing the lightning activity trajectory correction.

[0113] In this embodiment of the invention, a method for correcting lightning flash activity trajectories based on meteorological data is proposed. The provided method has at least the following technical effects or advantages:

[0114] (I) By constructing the initial centroid trajectory of lightning based on lightning location data, multi-source meteorological data, including satellite and radar data, are introduced to generate two types of corrected centroid trajectories with physical constraints. A trajectory evaluation model is then used to quantitatively evaluate and compare these two types, thereby outputting a trajectory correction strategy that conforms to the laws of physical evolution. This achieves dynamic optimization and correction of the initial centroid trajectory of lightning. This method significantly improves the spatial accuracy, evolutionary coherence, and physical rationality of lightning trajectories. It effectively overcomes the problem of trajectory drift or distortion caused by a single lightning location data source under abnormal distribution, unstable accuracy, and systematic errors. It facilitates the correction of trajectories determined by lightning location data, improving the accuracy of lightning flash activity trajectory determination.

[0115] (ii) By introducing a dynamic switching strategy unit during the trajectory evaluation process, the first centroid-corrected trajectory is automatically invoked for compensation when the score of the second centroid-corrected trajectory is insufficient or the time window is missing. This achieves the integrity restoration of the trajectory sequence and the utilization of information redundancy, significantly enhancing the robustness of trajectory recognition and the fault tolerance of the system. This strategy effectively avoids trajectory breakpoint problems caused by radar blind spots, data corruption, or timing errors, and is a key mechanism supporting the stable operation of the entire system.

[0116] (III) A mechanism for selecting core observation points for strong convection is constructed using physical constraint factors. A physical weighting function is constructed by multiplying three parameters: combined reflectivity, echo top height, and vertically integrated liquid water content. A physical field intensity criterion is introduced during the centroid calculation process, making the corrected trajectory more physically consistent with strong convection systems. This scheme significantly improves the physical directionality of the trajectory center offset and is a core technical means to enhance the spatial accuracy and practical representativeness of the corrected trajectory.

[0117] (iv) Design a trajectory physical index embedding structure. Integrate persistence, stability, and structural complexity indices for scoring and evaluation. Combine a weighted aggregation mechanism to generate a target trajectory correction strategy, comprehensively quantifying the differences in physical performance among multi-source trajectories, and improving the objectivity and transparency of trajectory correction decisions. This mechanism provides a standardized path for systematically evaluating the comparability of trajectories from different sources, effectively guiding the selection of trajectory sources.

[0118] To enable those skilled in the art to better understand the technical solutions of the present invention, the foregoing embodiments will be described in detail below with reference to specific examples.

[0119] The lightning flash activity trajectory correction method based on meteorological data provided in this embodiment of the invention mainly includes steps S110 to S160. The specific implementation steps are as follows:

[0120] S110. Obtain lightning location data for lightning and ground flash activity within the target area.

[0121] Specifically, step S110 is the foundational step in the entire lightning-to-ground flashover activity trajectory identification and correction process. It aims to provide high temporal resolution and spatial distribution accuracy of basic lightning event information for subsequent trajectory construction, meteorological data association, and physical correction. This step not only serves as data input but also relates to the spatiotemporal accuracy of the trajectory identification starting point and the consistency of physical association with subsequent processing algorithms.

[0122] Furthermore, in compliance with relevant regulations, the server acquires lightning location data within the target geographical area by connecting to a lightning monitoring network, a regional lightning monitoring subsystem, or a third-party high-precision lightning location platform. This lightning location data is typically generated based on the Time Difference of Arrival (TDOA) of electromagnetic waves from multiple ground stations or waveform analysis techniques (such as MDF (Multivariate Discriminant Function) or LPATS (Low-Power Adaptive Threshold System)). It includes parameters such as the timestamp, latitude and longitude coordinates, lightning type (e.g., cloud-to-ground lightning), polarity, and current intensity for each lightning strike event. The server sets spatial boundaries and time windows to filter the target area, ensuring that the collected data has spatiotemporal consistency with the current thunderstorm system activity area.

[0123] For example, in a monitoring task of thunderstorm activity in region A, the server set the time window to 14:00 to 16:00 on June 11, 2024, with spatial boundaries of 116.0° to 118.5° east longitude and 30.5° to 32.5° north latitude. The system automatically retrieved all ground flash events occurring in this region and time period from the connected lightning location network. A total of 12,136 valid lightning location records were obtained. Each record contains the trigger time accurate to the millisecond level and the latitude and longitude positioning with meter-level error, serving as the basic data points for constructing the initial lightning centroid trajectory.

[0124] S120. Determine the initial centroid trajectory of the lightning based on the lightning location data.

[0125] Specifically, step S120 is the core step in this embodiment of the invention, connecting lightning observation data with the trajectory correction model. Its function is to identify and construct an initial centroid trajectory reflecting the spatial movement path of the thunderstorm system based on the location distribution of multiple lightning events in the time series, providing a reference benchmark for the spatial structure and evolutionary trend of subsequent trajectory physical correction. This step not only requires spatiotemporal clustering of lightning points but also the extraction of representative center points of thunderstorm clusters in a reasonable manner to form an initial path line with evolutionary continuity.

[0126] Furthermore, the server first sorts the acquired lightning location data chronologically. It then performs sliding aggregation within time windows (e.g., every 5 or 10 minutes). Within each time window, the server uses a spatial density clustering algorithm (such as DBSCAN (Density-Based Spatial Clustering of Applications with Noise) or an improved adaptive clustering method) to group multiple simultaneously occurring and spatially adjacent lightning location points into a thunderstorm cluster. Subsequently, the geometric centroid of the lightning location points within each cluster is calculated. That is, the latitude and longitude coordinates of all points are averaged to obtain the initial centroid point for that time period. Connecting multiple centroid points chronologically forms the initial centroid trajectory of the lightning.

[0127] For example, during a thunderstorm in region B in July 2024, the server collected a total of 10,042 lightning location data points between 16:00 and 17:00. Six time slices were obtained by setting 10-minute time windows. Within each time slice, the DBSCAN clustering algorithm with a spatial distance threshold of 5 km was used to identify multiple thunderstorm clusters. The main clusters contained 250 to 430 lightning location points in each window. The server calculated the geometric center of each cluster, sequentially obtaining six centroids based on latitude and longitude, and connected them in chronological order to form the initial centroid trajectory of the thunderstorm activity.

[0128] Therefore, by performing this step, the spatiotemporal distribution patterns of lightning and ground flash activity can be extracted in a structured manner, providing initial trajectory values ​​and constraining the trajectory evolution direction for subsequent correction steps. The initial centroid trajectory serves as the input template for the trajectory correction model, providing a unified alignment reference for subsequent satellite and radar corrections and ensuring spatial consistency between data sources. By constructing the initial centroid trajectory, the observational advantages of the lightning positioning system are fully utilized, and an accurate, stable, and traceable foundational path is established for refined trajectory correction.

[0129] S130. Acquire meteorological data for the target area; meteorological data includes satellite data and radar data.

[0130] Specifically, step S130 is a crucial supporting step for realizing the physical correction of lightning activity trajectories. This step extends lightning location data from the electromagnetic characteristics level to the physical evolution level of the atmospheric system by introducing multi-source meteorological observation information. This improves the accuracy and spatiotemporal consistency of trajectory recognition in response to strong convective structures, solving the problem that traditional methods based solely on lightning location data are insufficient to determine the actual core area and physical development trend of thunderstorm systems. It is a prerequisite for constructing a multi-source collaborative correction mechanism.

[0131] Furthermore, the server retrieves satellite remote sensing data and ground-based radar observation data based on the set target area boundaries and thunderstorm activity time windows. Satellite data can select Top of the Atmosphere Brightness Temperature (TBB), Cloud Top Temperature (CTT), and Cloud Top Height (CTH), with a spatial resolution of 4 km and a temporal resolution of 15 minutes. Radar data can select Composite Reflectivity (CR), Echo Top Height (ET), and Vertical Integrated Liquid (VIL), with a spatial resolution of 1 km and a temporal resolution of 5 minutes.

[0132] The server extracts satellite image regions and radar data grid regions covering the initial centroid trajectory of lightning and its vicinity through spatial cropping operations, and performs time alignment processing to ensure that all meteorological data remain synchronized with lightning activity within each time window. During data integration, the server matches satellite image pixels and radar grid points with lightning location points or the initial centroid according to the geographic coordinate system, establishing a one-to-one mapping relationship and providing a spatial registration basis for subsequent trajectory correction.

[0133] For example, in a thunderstorm analysis of region C, the server set the target area to 113.5°E to 116.5°E and 22.0°N to 24.5°N. The time window was from 14:00 to 16:00 on June 12, 2025. The system extracted infrared temperature images of the region at six time points from FY-4B satellite cloud imagery and extracted CR, ET, and VIL data raster data for the corresponding time periods from three Doppler radars in cities a, b, and c. During the data preprocessing stage, the server completed spatial cropping, raster resampling, and temporal matching of satellite pixels and radar grids based on latitude and longitude coordinates, and cached the data into a structured dataset for direct use by subsequent model corrections.

[0134] The technical means provided in this step have the following beneficial effects: First, it enables the linkage modeling between lightning activity and meteorological physical fields, giving trajectory recognition the ability to respond to the real spatial structure of strong convective systems; second, through the collaboration of dual-source data, satellite data provides wide-area coverage and radar data provides vertical structure recognition capabilities, forming a complementary observation system; third, it improves the physical consistency and spatial resolution of trajectory correction, helping to make the judgment of centroid shift more consistent with the evolution of the convective core; and fourth, it provides the necessary input for the calculation of physical parameters of trajectory evaluation indicators (such as temperature gradient, strong convection criteria, radar factor weights, etc.), realizing a closed loop of the indicator-driven evaluation mechanism.

[0135] In summary, by executing step 130, not only is the physical foundation of trajectory correction strengthened, but a cross-modal fusion channel between lightning data and meteorological data is also constructed, promoting the transformation of lightning activity from geometric path recognition to physical field constraint evolution. This is a key information integration link to improve the accuracy and timeliness of thunderstorm warnings.

[0136] S140. Determine the first centroid correction trajectory corresponding to the satellite data, and determine the second centroid correction trajectory corresponding to the radar data.

[0137] Specifically, step S140 is the core step in realizing the physical correction of the initial centroid trajectory of lightning-to-ground flashover activity. By extracting strong convective feature regions from satellite and radar perspectives respectively, the initial trajectory is offset and reconstructed under constraints, thereby obtaining a corrected centroid trajectory that more closely resembles the true core region of the thunderstorm. This scheme elevates the expression of lightning events from a point-like distribution to a strong convective structure, which is a key step in constructing the physical consistency of the trajectory and the rationality of structural evolution.

[0138] Regarding satellite trajectory correction, the server extracts cloud top infrared radiation temperature, cloud top physical temperature, and cloud top altitude from remote sensing images provided by geostationary satellites such as FY-4B, forming a three-dimensional temperature structure field. Within the neighborhood of each centroid point on the initial centroid trajectory, the server performs spatial cropping and brightness-temperature profile extraction based on satellite images, calculating the cloud top cooling rate, temperature gradient, and low-temperature zone distribution characteristics of the corresponding region. If the current centroid falls in a region with gentle temperature changes or high brightness temperature, while its adjacent region contains a cold cloud center with a stronger cooling rate or lower TBB threshold, the server determines that the current centroid is not in the main convection region. Based on this, the centroid is shifted to a direction with a higher cooling rate or lower brightness temperature, forming the first corrected centroid. Each time window (e.g., set to 5 to 15 minutes, with the time window interval determined by personnel or automatically learned) is processed sequentially, and connecting the corrected centroid points constitutes the first centroid correction trajectory.

[0139] For radar trajectory correction, the server extracts combined reflectivity, echo top height, and vertically integrated liquid water content from Doppler weather radar grid data to establish core characteristic factors for strong convection. The server maps lightning observation points or initial centroids to the radar grid area, selecting radar grid cells that meet the criteria of CR > 40 dBZ, ET > 9 km, and VIL > 25 kg / m² as core observation points for strong convection. For each time window, the server constructs a physical weight function using the product of the above three factors, determines the physical weight values, and performs a weighted average of the latitude and longitude coordinates of all core strong convection grid points to generate a weighted centroid. Connecting these weighted centroids in the time series constitutes the second centroid correction trajectory.

[0140] For example, in processing a D-type thunderstorm event in a certain region in 2025, the initial centroid trajectory fell within a region with a cloud top brightness temperature of -38°C in the fourth time window. A cold cloud cluster with a brightness temperature of -47°C and a cooling rate of 6.5°C / h existed nearby. Based on this, the server shifted the centroid at this time point approximately 12km towards the center of this cold cloud cluster, generating the first corrected centroid. Simultaneously, in the radar data within the same time window, the radar grid CR of this shifted region reached 45dBZ, ET reached 10.2km, and VIL reached 28kg / m². The server used this as a high-weight point and calculated the radar-weighted centroid, forming the second corrected centroid. Finally, the server generated the first and second centroid corrected trajectories, each consisting of 6 points, for subsequent evaluation modules to compare physical plausibility and select strategies.

[0141] The technical means provided in this step have the following beneficial effects: First, it breaks away from the traditional geometric centroid modeling method, embedding trajectory recognition into the meteorological physical parameter field to achieve physical constraint reconstruction of the centroid position; second, through independent modeling and complementary coupling of satellite and radar remote sensing methods, it constructs a multi-version expression of the trajectory with cross-resolution characteristics, enhancing the system's adaptability under different weather observation conditions; third, the parameter-driven centroid offset strategy can achieve path optimization and extreme value adjustment, significantly improving the overlap between the trajectory and the main convection; fourth, the first centroid-corrected trajectory is suitable for cold cloud monitoring in wide-area scenarios, while the second centroid-corrected trajectory is suitable for tracking strong convection with a clear vertical structure. The two complement each other, supporting the dynamic strategy selection of the subsequent trajectory evaluation model.

[0142] In summary, step 140, by establishing an independent correction process under multi-source observation, realizes the transition from the initial lightning geometric trajectory to a multi-physical constraint trajectory, which is a key step for accurate identification and dynamic evolution modeling of lightning ground flash activity trajectories.

[0143] In one optional implementation, determining the first centroid correction trajectory corresponding to the satellite data may specifically include: obtaining the cloud top infrared radiation temperature, cloud top physical temperature, cloud top altitude, and satellite image of cloud clusters within the target area based on the satellite data; spatially cropping the satellite image to obtain the lightning flash activity area; calculating the cloud top cooling rate, cloud top infrared radiation temperature threshold, and temperature gradient of the cloud clusters corresponding to the lightning flash activity area based on the cloud top infrared radiation temperature, cloud top physical temperature, and cloud top altitude; if it is determined that the cloud top cooling rate of the cloud clusters corresponding to the lightning flash activity area is less than or equal to the cloud top cooling rate of the first cloud cluster adjacent to the lightning flash activity area, then it is determined that the initial centroid corresponding to the initial centroid trajectory of the lightning will be shifted towards the center direction of the second cloud cluster to obtain the first centroid, and the first centroid correction trajectory will be generated.

[0144] Among them, the first centroid correction trajectory includes the first centroid; the second cloud cluster is adjacent to the lightning and ground flash activity area; the cloud top cooling rate of the second cloud cluster is greater than the cloud top cooling rate of the cloud cluster corresponding to the lightning and ground flash activity area.

[0145] The above steps utilize satellite remote sensing data to physically guide and correct the initial centroid trajectory of lightning, constructing a centroid shift mechanism oriented towards cold cloud structure characteristics. Specifically, based on the brightness temperature variation trend and spatial distribution gradient of cloud clusters, the position of the initial centroid is adjusted to be closer to the true physical center of the strong convection region, thus forming the first centroid-corrected trajectory. Its core lies in fusing the spatial information of lightning location data with the convective structure evolution information from satellite brightness temperature data, achieving a precise spatial representation of lightning activity trajectories.

[0146] In the specific implementation, the server first acquires cloud image products covering the target area from a satellite remote sensing system (such as the FY-4B meteorological satellite) and extracts three key physical parameters: cloud top infrared radiation temperature (TBB), which reflects the intensity of convection and the degree of cold cloud evolution; cloud top physical temperature (CTT), which characterizes the actual thermal state of the cloud layer; and cloud top altitude (CTH), which is used to identify the height and intensity of convection development. The server then spatially crops the acquired raw images, focusing on the lightning flash activity area that includes the initial centroid trajectory of lightning.

[0147] Subsequently, within each time window, the server determines the time-varying rate of change of the cloud top infrared radiation temperature of the cloud cluster containing the centroid point corresponding to the initial centroid trajectory, thus obtaining the cloud top cooling rate. It extracts the lowest TBB value within this region as the cloud top infrared radiation temperature threshold and determines the brightness temperature gradient based on the cross-sectional change trend of TBB from the cluster center to the boundary. If the cloud top cooling rate corresponding to the centroid region is less than or equal to the cloud top cooling rate of the adjacent first cloud cluster, it indicates that the initial centroid is not located in the most active cooling center or convection core region. Based on this, the server identifies a neighboring second cloud cluster with stronger cooling characteristics (higher TBB cooling rate) and determines its spatial relationship. It then shifts the current centroid point towards the center of this second cluster with a certain offset radius, forming a new centroid point, denoted as the first centroid. By performing the above offset operation on each initial centroid within consecutive time windows, the server constructs a complete first centroid correction trajectory.

[0148] For example, during a thunderstorm in July 2025, at a certain moment, the initial centroid was located in an area with a TBB of -39°C, and its cloud top cooling rate was 3.8°C / h. Meanwhile, the adjacent cloud cluster to its northwest had a TBB of -47°C, with a cooling rate of 6.3°C / h. The server determined this area to be the main cold cloud region, corresponding to the dominant center of lightning activity. The initial centroid was shifted approximately 10km in this direction. The recalculated first centroid was located at the core of the cold cloud and updated as the first corrected centroid at the current time point.

[0149] In one alternative implementation, the formula for calculating the temperature gradient is as follows:

[0150]

[0151] in, The brightness temperature of the cloud top corresponding to the center (centroid) of the ground lightning cluster is expressed in Kelvin (K), reflecting the intensity of infrared radiation at that location. This represents the cloud top brightness temperature at the boundary of the cluster, in Kelvin (K). This represents the spherical distance between the cluster center point and the boundary point, in kilometers (km). This represents the spatial gradient of brightness temperature of the cluster, used to determine the rate of change of brightness temperature from the center to the boundary, and is expressed in K / km.

[0152] In one optional implementation, determining the first centroid correction trajectory corresponding to the satellite data further includes: determining the lightning density of the third cloud cluster and the fourth cloud cluster based on the cloud top altitude and temperature gradient, wherein both the third and fourth cloud clusters are adjacent to the cloud clusters corresponding to the lightning flash activity area; if it is determined that the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area is higher than that of the third cloud cluster, and the lightning density of the cloud cluster corresponding to the lightning flash activity area is greater than that of the third cloud cluster, then it is determined that the initial centroid corresponding to the initial lightning centroid trajectory will be shifted towards the center direction of the fourth cloud cluster to obtain the second centroid, and the first centroid correction trajectory will be updated.

[0153] Among them, the updated first centroid correction trajectory includes the second centroid; the cloud top physical temperature of the fourth cloud cluster is lower than the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area, and the lightning density of the fourth cloud cluster is greater than the lightning density of the cloud cluster corresponding to the lightning flash activity area.

[0154] The above steps, based on the satellite-corrected trajectory, further introduce an advanced centroid correction strategy that combines lightning intensity and cloud top physical temperature criteria. Its core objective is to accurately determine whether the initial centroid is located at the optimal thunderstorm activity center by utilizing the spatial intensity characteristics of lightning activity and the physical temperature characteristics of cold cloud structures. Based on this, the centroid is guided to shift towards a region with greater lightning dominance and thermal convection intensity. This improves the accuracy and dynamic response capability of the first centroid-corrected trajectory in terms of spatial consistency and physical representativeness of lightning behavior.

[0155] In the specific implementation, the server, based on the cloud top elevation and temperature gradient data obtained from previous processing, performs lightning density calculations on all adjacent cloud clusters surrounding the lightning flash activity area. Specifically, it counts the number of lightning location points per unit area within each cluster to obtain a numerical lightning density index. Simultaneously, it extracts the cloud top physical temperature of each cluster and marks the cloud cluster corresponding to the current centroid as the target comparison cluster. The server identifies two reference objects from its neighboring clusters: one is the third cloud cluster, which meets the criteria of having a lower cloud top physical temperature than the target cluster and lower lightning density; the other is the fourth cloud cluster, which must simultaneously meet the criteria of having a lower cloud top physical temperature than the target cluster and higher lightning density. If the judgment is correct, it indicates that the current centroid falls in an area with lower thermal intensity but stronger lightning activity, and the fourth cluster represents the optimal area for the superposition of cold cloud structure and lightning density. Based on this, the server shifts the initial centroid spatially towards the center of the fourth cluster, generating a new centroid point. It is designated as the second centroid, serving as one of the nodes in the trajectory corrected from the first centroid.

[0156] For example, in a severe convective event in central China in August 2025, the initial centroid of lightning was located in cluster A. This cluster had a cloud top physical temperature of -52°C and a lightning density of 4.1 lightning events per square kilometer. To its southwest lay cluster B, with a temperature of -56°C and a lightning density of 3.2. To its southeast lay cluster C, with a temperature of -58°C and a lightning density of 6.4. The server determined that cluster A had insufficient thermal structure strength and, although lightning activity was high, was not the optimal region. Therefore, cluster C was chosen as the correction target. The centroid was moved from A to C according to a preset offset radius and direction, forming the second centroid point for this time window, participating in the construction of the first centroid correction trajectory.

[0157] The construction and correction of the first centroid-corrected trajectory can bring the following beneficial technical effects: First, it elevates lightning location data from point density statistics to a standard for judging spatial energy density, introducing the significance of lightning activity as a supplementary criterion for centroid selection, thereby enhancing the physical correlation strength of centroid selection; Second, it combines cloud top physical temperature to construct a two-factor correction strategy, ensuring that the centroid shift not only tends towards areas with high lightning incidence but also aligns with strong cooling convection cores, improving the geometric accuracy of the trajectory's response to the actual thunderstorm center; Third, under the condition of single-source correction of satellite data, by constructing a composite spatial comparison rule of lightning density and thermal differences, it enhances the adaptability of the trajectory correction model in areas with complex cloud structures or insignificant brightness-temperature gradients; Fourth, by dynamically adjusting the centroid shift direction and target cluster selection mechanism, it achieves a combination of spatial continuity and physical centralization of the first centroid-corrected trajectory, constructing a more stable trajectory path with meteorological interpretability.

[0158] In summary, the construction and correction process of the first centroid correction trajectory is driven by two factors: cloud top temperature and lightning density. By judging the relative advantages of neighboring cloud clusters, a dynamic centroid shift mechanism is formed, which provides stronger lightning activity constraints for satellite path correction. This is a key supplementary strategy to improve the sensitivity and accuracy of the trajectory in response to the actual thunderstorm energy field structure.

[0159] In one optional implementation, determining the second centroid correction trajectory corresponding to the radar data may specifically include: determining the combined reflectivity, echo top height, and vertically integrated liquid water content from the radar data; extracting lightning observation points contained in the cloud clusters formed within each time window and mapping them to the corresponding radar grid cells; selecting target radar grid cells from multiple radar grid cells whose combined reflectivity, echo top height, and vertically integrated liquid water content meet preset threshold conditions; using the target radar grid cells as strong convection core observation points, and setting the product between the combined reflectivity, echo top height, and vertically integrated liquid water content corresponding to each strong convection core observation point as a physical weight value; performing a weighted average of the latitude and longitude coordinates of the target radar grid cells based on the physical weight value to generate a weighted centroid; and connecting the weighted centroids under each time window in chronological order to obtain the second centroid correction trajectory.

[0160] The above steps establish a physical criterion model for the core region of strong convection through structured analysis of radar observation data. A physical weighting function is then constructed based on multi-parameter product to calculate the corrected centroid trajectory of lightning activity, thus forming a second corrected centroid trajectory. This step fully utilizes the radar's high-resolution sensing capability of strong convective system structures, combining lightning observations with radar physical field characteristics to ensure that the trajectory correction more closely matches the evolution path of the thunderstorm core, solving the problem of traditional lightning centroid methods neglecting the convective intensity structure.

[0161] In the specific implementation, the server first extracts three key meteorological parameters from the accessed Doppler weather radar dataset: combined reflectivity (CR), echo top height (ET), and vertical integrated liquid water content (VIL). Combined reflectivity reflects thunderstorm intensity, ET represents the height of vertical structure development, and VIL reflects ice phase particles and precipitation potential. All three are significantly positively correlated with the spatial distribution of lightning activity.

[0162] Within each time window, the server identifies cloud clusters corresponding to densely distributed areas of lightning events, spatially maps lightning location points to corresponding radar grid cells, and establishes a grid-level mapping relationship between lightning points and radar data by using fast retrieval algorithms such as KD-tree (K-Dimensional Tree) or raster indexing to achieve coordinate concatenation. The server then selects target grid cells from all mapped radar grids that meet the strong convection characteristic threshold conditions, namely CR≥40dBZ, ET≥9km, and VIL≥25kg / m². These grid cells are considered core observation points for strong convection.

[0163] Subsequently, the server multiplies the CR, ET, and VIL values ​​of each strong convection core observation point, defining this as the physical weight value of that grid cell to characterize its contribution to the spatial centroid of thunderstorm activity. Within this time window, the server performs a weighted average of the latitude and longitude coordinates of all target grid cells based on the physical weight value, obtaining a representative weighted centroid point of the strong convection core. By repeating this operation over multiple consecutive time windows and connecting the weighted centroid points, a second centroid correction trajectory is formed.

[0164] For example, in a thunderstorm event, at a certain moment, 10 radar grid points in the radar data meet the threshold conditions. One of these grid points has a CR of 46 dBZ, an ET of 9.8 km, and a VIL of 28.6 kg / m², with a weight value of 46 × 9.8 × 28.6 ≈ 12918.88. This grid point contributes significantly to the final latitude-longitude weighted centroid. This weighted centroid represents the spatial fusion center of lightning activity and strong convective physical structure within that time window.

[0165] The construction and correction of the second centroid-corrected trajectory can bring the following beneficial technical effects: First, by introducing radar strong convection factors with physical causal relationships, a lightning centroid-corrected path with meteorological constraints is constructed, significantly improving the trajectory's expressive ability at the convective structure level. Second, by adopting a product-type physical weighting function, the dominant role of multiple strong convection factors in centroid correction when they superimpose in the same region is strengthened, effectively reducing the interference of single-factor anomalous disturbances. Third, through grid-level lightning point mapping and weighted methods, the fusion modeling of lightning observation and radar structure is realized, making trajectory correction no longer solely dependent on geometric averages but possessing physical dominance. Fourth, the output second centroid-corrected trajectory has the characteristics of strong continuity, high position reliability, and sensitivity to convective response, making it suitable for high-precision application scenarios such as power grid lightning protection control, strong convection path analysis, and refined thunderstorm early warning.

[0166] In summary, the construction and correction process of the second centroid correction trajectory, by constructing a radar parameter-driven weighted centroid trajectory recognition mechanism, successfully establishes a linkage channel between lightning behavior and meteorological physical structure, realizing the transformation from positioning data to multi-physical feature constrained trajectory. It is the key technical path with the most accurate guarantee and physical consistency in the accurate identification and dynamic physical modeling of lightning ground flash trajectory.

[0167] In one alternative implementation, the formula for calculating the physical weighting function is as follows:

[0168]

[0169] in, This represents the physical weight value of the i-th radar grid point, which is used for subsequent weighted centroid calculation. This represents the combined reflectivity of the i-th radar grid point, in dBZ, and is used to reflect the echo intensity. The echo top height of the i-th radar grid point is expressed in kilometers (km) and is used to represent the vertical development height of convection. The vertical integral liquid water content of the i-th radar grid point is expressed in kg / m², and is used to represent the liquid water column content.

[0170] In one optional implementation, the formula for calculating the weighted centroid latitude and longitude coordinates of the radar data is as follows:

[0171]

[0172] in, , These are the latitude and longitude coordinates of the cluster's weighted centroid, respectively. , Let be the geographic coordinates of the i-th lightning observation point, in decimal degrees; This represents the total number of all valid radar points in the cluster.

[0173] In one optional implementation, to express the activity center of a thunderstorm in the time dimension, the time centroid is defined as the arithmetic mean of the times of all points within the cluster. In this embodiment of the invention, the time centroid is used to quantitatively describe the temporal distribution structure of lightning-to-ground clusters. Its core function is to compress the occurrence times of multiple lightning events within a lightning cluster into a representative average time center point. This allows for precise location and dynamic tracking of the lightning centroid trajectory in the temporal dimension. The time centroid not only plays a fundamental role in time anchoring during trajectory construction but is also a key computational parameter for subsequent trajectory coherence evaluation, event synchronization analysis, and spatiotemporal consistency correction. The specific formula is as follows:

[0174]

[0175] in, The centroid of the cluster in the time dimension represents the average time of occurrence of lightning activity within the cluster, in minutes or seconds; This represents the occurrence time of the i-th lightning strike, with the unit consistent with the final output. This indicates the number of lightning points contained in the cluster.

[0176] S150. The trajectory evaluation model is used to evaluate the first centroid correction trajectory and the second centroid correction trajectory, and the trajectory correction strategy is output.

[0177] Step S150 introduces a trajectory evaluation model on the server side to conduct a systematic and multi-dimensional physical index analysis and quality evaluation of the first and second centroid-corrected trajectories, thereby outputting the optimal trajectory correction strategy. Its core lies in establishing a quantitative trajectory quality assessment mechanism to achieve accurate comparison between satellite-driven and radar-driven trajectories. The final correction path with the highest physical consistency and spatiotemporal continuity is selected in a data-driven manner. This evaluation mechanism not only improves the decision-making intelligence level of the trajectory correction process but also enhances the interpretability and robustness of the result selection.

[0178] In the specific implementation, the server first preprocesses the generated first and second centroid-corrected trajectories, including time series resampling, spatial distribution normalization, and trajectory node alignment, to ensure the comparability of subsequent index calculations. Then, the two types of trajectories are input into the trajectory evaluation model. The trajectory evaluation model comprises two main sub-modules: a trajectory physical index embedding unit and a weight evaluation unit.

[0179] For example, in the analysis of a thunderstorm process in 2025, the duration of the first centroid-corrected trajectory was 70 minutes, the stability score was 0.76, and the complexity was one merging. The duration of the second centroid-corrected trajectory was 85 minutes, the stability score was 0.89, and the complexity was two mergings and one split. The server applied a weighted function (e.g., a weight ratio of 0.3:0.5:0.2) to aggregate the three indicators, resulting in a score of 0.72 for the first centroid-corrected trajectory and 0.87 for the second centroid-corrected trajectory. Since the score of the second centroid-corrected trajectory was higher than the preset threshold of 0.8 and there was no missing time window, the server determined the second centroid-corrected trajectory as the target correction trajectory and output the trajectory correction strategy. The instruction system subsequently adopted the second centroid-corrected trajectory as the final lightning centroid path.

[0180] In one optional implementation, a trajectory evaluation model is used to evaluate the first centroid-corrected trajectory and the second centroid-corrected trajectory, and output a trajectory correction strategy. Specifically, this may include: inputting the first centroid-corrected trajectory and the second centroid-corrected trajectory into the trajectory evaluation model after preprocessing; extracting the persistence index, stability index, and structural complexity index corresponding to the first centroid-corrected trajectory and the second centroid-corrected trajectory under each time window through the trajectory physical index embedding unit in the trajectory evaluation model; performing score aggregation based on a preset trajectory physical index weight function through the weight evaluation unit in the trajectory evaluation model to obtain a score result; if the score of the second centroid-corrected trajectory is determined to be higher than a preset score threshold and the time window of the second centroid-corrected trajectory is not missing, then the second centroid-corrected trajectory is determined to be the target correction trajectory; wherein, the trajectory correction strategy includes the target correction trajectory.

[0181] In the specific implementation, firstly, the server preprocesses the first and second centroid-corrected trajectories. This preprocessing includes time window alignment, trajectory node synchronization, missing data interpolation, and spatial coordinate normalization to ensure that the two trajectories can be compared on the same time and spatial scales. Secondly, the processed trajectories are input into the trajectory evaluation model. The "trajectory physical index embedding unit" in the model extracts three types of core physical indices for each time window. The first type is the duration index, which calculates the total number of time windows spanned by each trajectory from the first centroid to the last centroid. This characterizes whether the trajectory has sufficient coverage of the thunderstorm lifecycle. The longer the time span, the more continuous the thunderstorm process identification. The second type is the coherence index, which measures the degree of change in the number of lightning events within each time window. This is represented by the normalized deviation of the maximum change in lightning events from its mean. The smaller the change, the more coherent and smooth the trajectory. The third type is the complexity index, which counts the number of structural changes based on whether thunderstorm clusters split or merge during trajectory evolution. It is used to measure the ability of a trajectory to respond to the complex behavior of a thunderstorm system.

[0182] Next, the "weight evaluation unit" in the model performs a score fusion of the three physical indicators mentioned above. This unit uses a preset physical indicator weight function to weight and aggregate the indicators. For example, the weights are set as persistence 0.3, stability 0.5, and complexity 0.2, ultimately generating a comprehensive score for each trajectory. If the score of the second centroid-corrected trajectory is higher than the set score threshold (e.g., 0.8), and its time series is complete without missing windows, the server identifies this trajectory as the "target correction trajectory." A trajectory correction strategy is then constructed based on this, which is used by subsequent modules to output the final lightning-to-ground flash activity trajectory.

[0183] For example, during a thunderstorm, the first centroid-corrected trajectory had a duration of 60 minutes, a stability bias of 0.24, a complexity of 1 split, and a comprehensive score of 0.74. The second centroid-corrected trajectory had a duration of 85 minutes, a stability bias of 0.12, a complexity of 2 merges, and a comprehensive score of 0.87. Because the second centroid-corrected trajectory had a higher score and a continuous time window, the server automatically set it as the target correction trajectory and updated the trajectory correction strategy to "use radar-weighted centroid trajectory as the final output".

[0184] The technical means provided in this step have the following beneficial effects: First, it introduces a systematic multi-dimensional physical evaluation structure to achieve quantitative judgment of the quality of the corrected trajectory, eliminate the uncertainty of subjective judgment, and improve the credibility and automation level of trajectory selection; Second, it integrates three dimensions—thunderstorm life cycle characteristics, lightning evolution continuity, and system structural response capability—into a unified model to ensure that the selected trajectory has both temporal integrity and spatial structural rationality; Third, the scoring aggregation mechanism has adjustable weight capabilities, adapting to different meteorological scenarios (such as squall lines that emphasize structural response and localized short-term heavy precipitation that emphasizes temporal stability), and has broad adaptability; Fourth, through an automated strategy output mechanism, it supports the trajectory recognition system to operate in real time in a big data environment, improving the intelligence level and stability of the severe convection monitoring and lightning early warning system.

[0185] In summary, by executing step 150, an intelligent evaluation model for trajectory quality discrimination was constructed, realizing a closed-loop process from physical index extraction to score optimization and then to automatic generation of correction strategies. This is the core technical path for promoting intelligent fusion, dynamic correction and real-time control of multi-source lightning trajectories.

[0186] In one optional implementation, a trajectory evaluation model is used to evaluate the first centroid correction trajectory and the second centroid correction trajectory, and output a trajectory correction strategy. Specifically, it may further include: if the score of the second centroid correction trajectory is determined to be lower than a preset score threshold, or the time window of the second centroid correction trajectory is missing, then the dynamic switching strategy unit in the trajectory evaluation model calls the first centroid correction trajectory corresponding to the time window of the second centroid correction trajectory for compensation, so as to obtain the target correction trajectory.

[0187] To address the potential issues of insufficient scoring in the second centroid-corrected trajectory or missing time window data during trajectory evaluation, the above steps propose a dynamic switching strategy mechanism based on trajectory quality status perception. By introducing some nodes of the first centroid-corrected trajectory through local compensation, a hybrid target-corrected trajectory is constructed. This improves the overall integrity, stability, and robustness of the trajectory results, avoiding the risk of trajectory misjudgment or gaps due to radar data anomalies, interruptions, or structural changes. It is an important means of achieving multi-source trajectory fusion and compensation-tolerant control.

[0188] In the specific implementation, after the server completes the scoring and window integrity detection of the first centroid correction trajectory and the second centroid correction trajectory, it will trigger the dynamic switching mechanism if any of the following conditions are met: the comprehensive score of the second centroid correction trajectory is lower than the preset scoring threshold (for example, the threshold is set to 0.8); the second centroid correction trajectory is missing centroid data or has continuous breakpoints in certain time windows, and cannot form a complete trajectory.

[0189] When the above situation occurs, the server automatically invokes the "Dynamic Switching Strategy Unit" in the trajectory evaluation model. The working logic of this unit includes the following steps: First, it detects all sets of missing time windows or time periods with substandard scores in the second centroid-corrected trajectory and marks them as candidate replacement segments. Then, it finds the corrected centroid node under the corresponding time window in the first centroid-corrected trajectory and fills it as a compensation point into the corresponding time period position of the second centroid-corrected trajectory, forming a new continuous centroid sequence. To maintain the consistency and smoothness of the trajectory structure, the server performs interpolation fitting or transition smoothing algorithms between the two centroids before and after the replacement to avoid abrupt jumps.

[0190] For example, in a severe convective event where radar observations were obstructed, the second centroid correction trajectory had a gap in the time window from 17:45 to 18:00 due to the loss of ET and VIL data, resulting in a score of 0.71, which is lower than the set threshold of 0.8. After the server detected the missing nodes of the second centroid correction trajectory during this time period, it automatically retrieved the satellite correction centroid corresponding to that time period from the first centroid correction trajectory and inserted it into the second trajectory path. At the same time, it performed interpolation smoothing on the trajectory path between the two nodes at 17:30 and 18:15, ultimately constructing a continuous and uninterrupted target correction trajectory.

[0191] In one alternative implementation, the formula for calculating the trajectory stability index (Coherence) involved in the above steps is as follows:

[0192]

[0193] in, This represents the trajectory stability index; the smaller the value, the more stable the lightning activity in the time series. This indicates the duration of the trajectory, in minutes or the number of time windows. This represents the maximum change in lightning events (sudden change in the number or intensity of lightning) within the k-th time window. This represents the average value of the change in lightning events across all time windows.

[0194] In one alternative implementation, the formula for calculating the trajectory structure complexity index (Complexity) involved in the above steps is as follows:

[0195]

[0196] in, Indicators representing the complexity of trajectory structure; This represents the number of trajectories with complex evolutionary characteristics; This indicates the number of times the thunderstorm system splits as recorded in each trajectory; This indicates the number of times thunderstorm systems merged in each trajectory.

[0197] S160. Correct the initial centroid trajectory of lightning according to the trajectory correction strategy to obtain the corrected centroid trajectory of lightning ground flash activity in the target area.

[0198] Step S160 dynamically corrects the initial centroid trajectory of the lightning using a trajectory correction strategy via a server, ultimately generating a corrected centroid trajectory for the target region with multi-source physical constraints. This step is the final stage of the entire lightning-to-ground lightning activity trajectory modeling process. Its core objective is to replace or optimize the initial centroid trajectory with the optimal corrected trajectory structure output by the evaluation model to obtain the final trajectory result that is optimal in terms of physical consistency, structural coherence, and spatiotemporal accuracy. It is a crucial closed loop in completing the process from lightning observation to trajectory output.

[0199] Specifically, the server first receives the trajectory correction strategy output by the trajectory evaluation model. This includes the type identifier of the target corrected trajectory (first centroid corrected trajectory, second centroid corrected trajectory, or fused trajectory) and its spatial node sequence. The server compares the centroid points of the initial centroid trajectory within the time window and performs replacement or interpolation adjustments. If the trajectory correction strategy specifies that the target corrected trajectory is a complete second centroid corrected trajectory, the server directly replaces the centroid nodes of the corresponding window in the original trajectory with the centroid coordinates of all time nodes of this trajectory. If the strategy specifies that the target corrected trajectory is a fused trajectory, i.e., some time windows use first centroid corrected trajectory nodes and the rest use second centroid corrected trajectory nodes (first centroid corrected trajectory + second centroid corrected trajectory), the server determines the source of the current node in each time window and completes the node splicing accordingly. At the same time, it uses smooth connection algorithms such as piecewise linear interpolation and trajectory Bezier reconstruction to eliminate spatial jumps.

[0200] For example, during a thunderstorm activity monitoring process, the initial centroid trajectory experienced a shift in consecutive nodes between 15:00 and 16:00 due to uneven lightning location points. The trajectory evaluation model ultimately output a trajectory correction strategy of "primarily using the second centroid-corrected trajectory, but employing the first centroid-corrected trajectory as a compensation segment between 15:25 and 15:35." The server updates the positions of all nodes on the initial centroid trajectory according to this strategy and performs trajectory smoothing before and after the splicing point of the two data sources to ensure that the final trajectory exhibits no abrupt path changes. After correction, the server outputs a sequence of lightning-corrected centroid trajectories for the target area's lightning-to-ground flash activity within a specified time period, which can be used by subsequent thunderstorm path evolution monitoring, early warning issuance, or 3D thunderstorm reconstruction systems.

[0201] In one optional implementation, the time span of the lightning cluster from the starting point to the ending point is determined based on the first centroid-corrected trajectory; based on the time span, a persistence index of the first centroid-corrected trajectory is determined; based on the first centroid-corrected trajectory, the trajectory duration, lightning event variation, and average lightning event variation of the lightning cluster are determined; based on the trajectory duration, lightning event variation, and average lightning event variation, a stability index of the first centroid-corrected trajectory is determined; based on the first centroid-corrected trajectory, the average number of thunderstorm splitting and merging events is determined; and based on the average number of thunderstorm splitting and merging events, a structural complexity index of the first centroid-corrected trajectory is determined.

[0202] The above steps construct a quantifiable trajectory quality evaluation system oriented towards the physical process, based on the first centroid-corrected trajectory. By extracting three core parameters—persistence, stability, and structural complexity—it comprehensively describes the performance of the corrected lightning trajectory in terms of temporal coverage, evolutionary coherence, and response to changes in system structure. This not only provides data support for subsequent trajectory optimization but also serves as a process variable input in thunderstorm evolution modeling and early warning decision-making, forming a crucial technical foundation for ensuring trajectory reliability and comparability.

[0203] In the specific implementation, the server uses the time series of the first centroid-corrected trajectory as the analysis benchmark and extracts trajectory attribute indicators dimension by dimension: First, the server calculates the total duration of the trajectory, i.e., the persistence index, based on the time span between the first and last centroid time points in the first centroid-corrected trajectory. This index reflects the continuity of lightning cluster activity in the time dimension; the longer the duration, the more completely the trajectory covers the thunderstorm life cycle.

[0204] Then, the server statistically analyzes the changes in lightning events within the time window corresponding to the first centroid-corrected trajectory. Specifically, the server obtains the number of lightning events associated with the spatial neighborhood of the centroid within each time window and calculates the difference in the number of lightning events between adjacent time windows to obtain a lightning event change sequence. The average of this change sequence is then calculated to obtain the mean change in lightning events. The magnitude of lightning event changes in a single time window is normalized to the mean, further summed, and divided by the trajectory duration to obtain a stability index. The smaller the value, the smoother the trajectory is in the time series, and the lower the fluctuation. Next, the server, considering the spatial structure of the first centroid-corrected trajectory, detects whether thunderstorm clusters in the trajectory exhibit spatial breaks, splitting into multiple branches, or multiple trajectory segments merging into one segment. The number of all splitting and merging events occurring throughout the entire trajectory is statistically analyzed, and the trajectory segments are normalized to calculate the frequency of structural changes per unit length of the trajectory, i.e., the structural complexity index, used to quantify whether the trajectory can effectively capture the dynamic evolution of complex convective systems.

[0205] For example, in a thunderstorm event, the first centroid-corrected trajectory spans from 14:00 to 15:20, lasting 80 minutes. The number of lightning events varies across time windows as follows: [22, 24, 25, 23, 27, 26, 25, 24]. The standard deviation of this variation is 1.5, the mean is 2.1, and the normalized stability index is 0.35. Spatially, two splitting events and one merging event are identified. The trajectory is divided into six segments. The structural complexity index is (2+1) / 6≈0.5. Based on this, the server determines that the trajectory possesses good temporal integrity, moderate stability, and adequate structural response capability.

[0206] It should be noted that the persistence, stability and structural complexity indices of the second centroid correction trajectory can be performed by referring to the relevant steps of the first centroid correction trajectory, which will not be elaborated here.

[0207] By calculating various indices for corrected trajectories, the following beneficial technical effects can be achieved: First, a systematic evaluation system for trajectory physical indices is introduced, transforming trajectory analysis from subjective impressions to data-driven judgments, providing a foundation for trajectory optimization, model fusion, and evolutionary analysis. Second, by modeling the changing trends of lightning events, the trajectory model's responsiveness to time series stability and consistency of thunderstorm activity is enhanced, helping to eliminate abrupt trajectories and pseudo-center offset paths. Third, the structural complexity index supplements the dynamic behavior of thunderstorm systems that traditional path models cannot reflect, making it suitable for trajectory modeling needs in complex multi-center or highly vortex-driven convective systems. Fourth, this index system has good versatility and can be applied to the unified evaluation of trajectories from different sources, such as satellite-corrected trajectories, radar-corrected trajectories, and hybrid-corrected trajectories, which is beneficial for the comparative optimization of cross-source trajectory schemes.

[0208] In summary, by executing the index calculation process of the corrected trajectory, a multi-dimensional physical index system based on the trajectory ontology attributes was constructed. This system supports the full-process evaluation of the trajectory from quality perception and quantification to fusion, and is an indispensable key algorithm module for intelligent decision-making on lightning trajectories and modeling of thunderstorm behavior.

[0209] Reference Figure 2 The diagram illustrates a structural block diagram of a lightning flash activity trajectory correction device provided in an embodiment of the present invention, which may specifically include:

[0210] The data acquisition unit 201 is used to acquire meteorological data of the target area, as well as lightning location data for lightning and ground flash activity in the target area; the meteorological data includes satellite data and radar data;

[0211] The initial centroid trajectory determination unit 202 is used to determine the initial centroid trajectory of lightning based on the lightning location data.

[0212] The trajectory correction generation unit 203 is used to generate a first centroid correction trajectory corresponding to the satellite data and a second centroid correction trajectory corresponding to the radar data.

[0213] The trajectory correction strategy output unit 204 is used to evaluate the first centroid correction trajectory and the second centroid correction trajectory according to the pre-built trajectory evaluation model and output the trajectory correction strategy.

[0214] The centroid trajectory correction unit 205 is used to correct the initial centroid trajectory of lightning according to the trajectory correction strategy, so as to obtain the corrected centroid trajectory of lightning ground flash activity in the target area.

[0215] In one optional embodiment, the corrected trajectory generation unit 203 includes:

[0216] The cloud cluster information determination unit is used to determine the cloud top physical temperature, cloud top altitude and satellite image of cloud clusters in the target area based on the satellite data.

[0217] A spatial cropping unit is used to perform spatial cropping on the satellite image to obtain the lightning flash activity area;

[0218] The cloud top cooling rate determination unit is used to determine the cloud top cooling rate of the cloud cluster corresponding to the lightning flash activity area based on the cloud top physical temperature and the cloud top altitude.

[0219] The first centroid offset unit is used to offset the initial centroid corresponding to the initial centroid trajectory of the lightning to the center of the second cloud cluster when the cloud top cooling rate of the cloud cluster corresponding to the lightning flash activity area is less than or equal to the cloud top cooling rate of the first cloud cluster adjacent to the lightning flash activity area, thereby obtaining the first centroid and generating the first centroid correction trajectory.

[0220] Wherein, the first centroid correction trajectory includes the first centroid; the second cloud cluster is adjacent to the lightning flash activity area; the cloud top cooling rate of the second cloud cluster is greater than the cloud top cooling rate of the cloud cluster corresponding to the lightning flash activity area.

[0221] In an optional embodiment, the corrected trajectory generation unit 203 further includes:

[0222] The temperature gradient determination unit is used to determine the temperature gradient of the cloud clusters corresponding to the lightning flash activity area based on the cloud top physical temperature and the cloud top altitude.

[0223] A lightning density determination unit is used to determine the lightning density of the third cloud cluster and the fourth cloud cluster based on the cloud top altitude and the temperature gradient; both the third cloud cluster and the fourth cloud cluster are adjacent to the cloud cluster corresponding to the lightning flash activity area.

[0224] The second centroid offset unit is used to offset the initial centroid corresponding to the initial centroid trajectory of the lightning to the center direction of the fourth cloud cluster when the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area is higher than that of the third cloud cluster, and the lightning density of the cloud cluster corresponding to the lightning flash activity area is greater than that of the third cloud cluster. This results in obtaining a second centroid and updating the first centroid correction trajectory.

[0225] Among them, the updated first centroid correction trajectory; the cloud top physical temperature of the fourth cloud cluster is lower than the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area; the lightning density of the fourth cloud cluster is greater than the lightning density of the cloud cluster corresponding to the lightning flash activity area.

[0226] In one optional embodiment, the corrected trajectory generation unit 203 includes:

[0227] The radar data analysis unit is used to determine the combined reflectivity, echo top height, and vertical integrated liquid water content based on the radar data.

[0228] The lightning observation point extraction and mapping unit is used to extract lightning observation points from cloud clusters formed within each time window and map them to the radar grid unit where the lightning observation points are located.

[0229] The target radar grid cell filtering unit is used to filter from multiple radar grid cells to obtain at least one target radar grid cell whose combined reflectivity, echo top height, and vertical integrated liquid water content meet preset threshold conditions.

[0230] The physical weight value determination unit is used to take each of the target radar grid units as a strong convection core observation point, and for each strong convection core observation point, determine the physical weight value based on the combined reflectivity, echo top height and vertical integrated liquid water content corresponding to the strong convection core observation point.

[0231] The weighted centroid generation unit is used to generate a weighted centroid by performing a weighted average of the latitude and longitude coordinates of the target radar grid cell based on each of the physical weight values.

[0232] The second centroid correction trajectory generation subunit is used to connect the weighted centroids under each time window in chronological order to obtain the second centroid correction trajectory.

[0233] In one optional embodiment, the trajectory correction strategy output unit 204 includes:

[0234] The model data input unit is used to input the first centroid correction trajectory and the second centroid correction trajectory into a pre-constructed trajectory evaluation model after preprocessing; the trajectory evaluation model includes a trajectory physical index embedding unit and a weight evaluation unit;

[0235] The index extraction unit is used to extract the persistence index, stability index and structural complexity index corresponding to the first centroid correction trajectory and the second centroid correction trajectory under each time window through the trajectory physical index embedding unit.

[0236] The scoring aggregation unit is used to aggregate scores based on each of the persistence indicators, stability indicators, and structural complexity indicators through the weight evaluation unit based on a preset trajectory physical indicator weight function to obtain a scoring result.

[0237] The first target correction trajectory determination unit is used to determine the second centroid correction trajectory as the target correction trajectory when the score of the scoring result indicates that the score of the second centroid correction trajectory is higher than a preset score threshold and the time window of the second centroid correction trajectory is not missing, and to generate a trajectory correction strategy based on the target correction trajectory.

[0238] In one optional embodiment, the trajectory evaluation model further includes a dynamic switching strategy unit; the trajectory correction strategy output unit 204 further includes:

[0239] The second target correction trajectory determination unit is used to, when the scoring result indicates that the score of the second centroid correction trajectory is lower than the preset scoring threshold, or when the time window of the second centroid correction trajectory is missing, call the first centroid correction trajectory to compensate the second centroid correction trajectory in the same time window through the dynamic switching strategy unit to obtain the target correction trajectory, and generate a trajectory correction strategy based on the target correction trajectory.

[0240] In one optional embodiment, the extraction process of the persistence index, stability index, and structural complexity index corresponding to the first centroid correction trajectory and the second centroid correction trajectory under each time window in the index extraction unit includes:

[0241] The first centroid correction trajectory and the second centroid correction trajectory are sequentially used as the correction trajectories to be evaluated.

[0242] Based on the trajectory to be evaluated and corrected, determine the time span, trajectory duration, lightning event variation, and average lightning event variation of the lightning cluster from the starting point to the ending point.

[0243] Based on the time span, determine the persistence index of the correction trajectory to be evaluated;

[0244] Based on the trajectory duration, the amount of change in the lightning event, and the average amount of change in the lightning event, a stability index for the trajectory to be evaluated and corrected is determined.

[0245] Based on the revised trajectory to be evaluated, determine the average number of thunderstorm splitting and merging events;

[0246] Based on the average number of times, the structural complexity index of the modified trajectory to be evaluated is determined.

[0247] As the device embodiment is basically similar to the method embodiment, it is described in a relatively simple way. For relevant details, please refer to the description of the method embodiment above.

[0248] It should be noted that, in order to enable those skilled in the art to better distinguish data of the same type but with different actual meanings, the embodiments of the present invention use terms such as "first" and "second" to distinguish and describe some technical features. The terms "first" and "second" are used only for data differentiation and have no other special meanings. It is understood that the present invention does not impose any limitations on them.

[0249] This invention also provides an electronic device, which includes a processor and a memory:

[0250] The memory is used to store program code and transfer the program code to the processor;

[0251] The processor is used to execute the lightning flash activity trajectory correction method of any embodiment of the present invention according to the instructions in the program code.

[0252] More specifically, refer to Figure 3 The diagram illustrates the structure of an electronic device according to an embodiment of the present invention. The electronic device may include: at least one processor 31, at least one network interface 34, a user interface 33, a memory 35, and at least one communication bus 32.

[0253] The communication bus 32 is used to enable communication between these components.

[0254] The user interface 33 may include a display screen and a camera. Optionally, the user interface 33 may also include a standard wired interface and a wireless interface.

[0255] The network interface 34 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0256] The processor 31 may include one or more processing cores. The processor 31 connects to various parts of the server via various interfaces and lines, and performs various server functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in the memory 35, and by calling data stored in the memory 35. Optionally, the processor 31 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 31 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 31 and may be implemented as a separate chip.

[0257] The memory 35 may include random access memory (RAM) or read-only memory (ROM). Optionally, the memory 35 may include a non-transitory computer-readable storage medium. The memory 35 may be used to store instructions, programs, code, code sets, or instruction sets. The memory 35 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the various method embodiments described above, etc.; the data storage area may store data involved in the various method embodiments described above, etc. The memory 35 may optionally also be at least one storage device located remotely from the aforementioned processor 31. Figure 3 As shown, the memory 35, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for correcting the trajectory of lightning strikes.

[0258] exist Figure 3In the electronic device shown, the user interface 33 is mainly used to provide an input interface for the user and to obtain the user input data; while the processor 31 can be used to call an application program stored in the memory 35 for a lightning flash activity trajectory correction method. When executed by one or more processors, the electronic device executes one or more methods as described in the above embodiments.

[0259] This invention also provides a computer-readable storage medium for storing program code for executing the lightning-to-ground flash activity trajectory correction method of any embodiment of this invention.

[0260] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0261] In the embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between devices or units through some interfaces, and may be electrical, mechanical, or other forms.

[0262] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0263] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0264] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0265] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for correcting the trajectory of lightning strikes, characterized in that, include: Acquire meteorological data for the target area, as well as lightning location data for lightning and ground flash activity within the target area; The meteorological data includes satellite data and radar data; Based on the lightning location data, determine the initial centroid trajectory of the lightning; Generate a first centroid-corrected trajectory corresponding to the satellite data, and generate a second centroid-corrected trajectory corresponding to the radar data; The first centroid correction trajectory and the second centroid correction trajectory are evaluated based on the pre-built trajectory evaluation model, and the trajectory correction strategy is output. The initial centroid trajectory of the lightning is corrected according to the trajectory correction strategy to obtain the corrected centroid trajectory of lightning activity in the target area.

2. The lightning flash trajectory correction method according to claim 1, characterized in that, The generation of the first centroid corrected trajectory corresponding to the satellite data includes: Based on the satellite data, determine the cloud top physical temperature, cloud top altitude, and satellite imagery of the cloud clusters within the target area; Spatially crop the satellite image to obtain the lightning flash activity area; Based on the cloud top physical temperature and the cloud top altitude, the cloud top cooling rate of the cloud cluster corresponding to the lightning flash activity area is determined. If the cloud top cooling rate of the cloud cluster corresponding to the lightning flash activity area is less than or equal to the cloud top cooling rate of the first cloud cluster adjacent to the lightning flash activity area, then the initial centroid corresponding to the initial centroid trajectory of the lightning is shifted towards the center of the second cloud cluster to obtain the first centroid, and the first centroid correction trajectory is generated. Wherein, the first centroid correction trajectory includes the first centroid; the second cloud cluster is adjacent to the lightning flash activity area; the cloud top cooling rate of the second cloud cluster is greater than the cloud top cooling rate of the cloud cluster corresponding to the lightning flash activity area.

3. The lightning flash trajectory correction method according to claim 2, characterized in that, The method further includes: Based on the cloud top physical temperature and the cloud top altitude, the temperature gradient of the cloud clusters corresponding to the lightning flash activity area is determined. Based on the cloud top altitude and the temperature gradient, the lightning density of the third cloud cluster and the lightning density of the fourth cloud cluster are determined; both the third cloud cluster and the fourth cloud cluster are adjacent to the cloud clusters corresponding to the lightning flash activity area. If the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area is higher than that of the third cloud cluster, and the lightning density of the cloud cluster corresponding to the lightning flash activity area is greater than that of the third cloud cluster, then the initial centroid corresponding to the initial centroid trajectory of the lightning is shifted towards the center of the fourth cloud cluster to obtain the second centroid, and the first centroid correction trajectory is updated. Among them, the updated first centroid correction trajectory; the cloud top physical temperature of the fourth cloud cluster is lower than the cloud top physical temperature of the cloud cluster corresponding to the lightning flash activity area; the lightning density of the fourth cloud cluster is greater than the lightning density of the cloud cluster corresponding to the lightning flash activity area.

4. The method for correcting the trajectory of lightning strikes according to claim 1, characterized in that, The generation of the second centroid correction trajectory corresponding to the radar data includes: Based on the radar data, determine the combined reflectivity, echo top height, and vertical integrated liquid water content; Lightning observation points are extracted from cloud clusters formed within each time window and mapped to the radar grid cells where the lightning observation points are located; At least one target radar grid cell is selected from multiple radar grid cells when the combined reflectivity, echo top height and vertical integrated liquid water content meet the preset threshold conditions. Each of the target radar grid units is taken as a strong convection core observation point. For each strong convection core observation point, a physical weight value is determined based on the combined reflectivity, echo top height and vertical integrated liquid water content corresponding to the strong convection core observation point. Based on each of the physical weight values, the latitude and longitude coordinates of the target radar grid cell are weighted and averaged to generate a weighted centroid. The weighted centroids under each time window are connected in chronological order to obtain the second centroid correction trajectory.

5. The method for correcting the trajectory of lightning strikes according to claim 1, characterized in that, The evaluation of the first centroid-corrected trajectory and the second centroid-corrected trajectory based on a pre-built trajectory evaluation model, and the output of a trajectory correction strategy, includes: The first centroid-corrected trajectory and the second centroid-corrected trajectory are preprocessed and then input into a pre-constructed trajectory evaluation model; the trajectory evaluation model includes a trajectory physical index embedding unit and a weight evaluation unit; The trajectory physical index embedding unit extracts the persistence index, stability index, and structural complexity index corresponding to the first centroid corrected trajectory and the second centroid corrected trajectory under each time window. Based on each of the aforementioned persistence indicators, stability indicators, and structural complexity indicators, the weight evaluation unit performs score aggregation based on a preset trajectory physical indicator weight function to obtain the score result. If the scoring result indicates that the score of the second centroid correction trajectory is higher than the preset scoring threshold, and the time window of the second centroid correction trajectory is not missing, then the second centroid correction trajectory is determined as the target correction trajectory, and a trajectory correction strategy is generated based on the target correction trajectory.

6. The lightning flash trajectory correction method according to claim 5, characterized in that, The trajectory evaluation model further includes a dynamic switching strategy unit; the method further includes: If the scoring result indicates that the score of the second centroid correction trajectory is lower than the preset scoring threshold, or if the time window of the second centroid correction trajectory is missing, then the dynamic switching strategy unit calls the first centroid correction trajectory to compensate the second centroid correction trajectory under the same time window to obtain the target correction trajectory, and generates a trajectory correction strategy based on the target correction trajectory.

7. The lightning flash trajectory correction method according to claim 5 or 6, characterized in that, The extraction process for the persistence index, stability index, and structural complexity index corresponding to the first and second centroid correction trajectories under each time window includes: The first centroid correction trajectory and the second centroid correction trajectory are sequentially used as the correction trajectories to be evaluated. Based on the trajectory to be evaluated and corrected, determine the time span, trajectory duration, lightning event variation, and average lightning event variation of the lightning cluster from the starting point to the ending point. Based on the time span, determine the persistence index of the correction trajectory to be evaluated; Based on the trajectory duration, the amount of change in the lightning event, and the average amount of change in the lightning event, a stability index for the trajectory to be evaluated and corrected is determined. Based on the revised trajectory to be evaluated, determine the average number of thunderstorm splitting and merging events; Based on the average number of times, the structural complexity index of the modified trajectory to be evaluated is determined.

8. A lightning flash trajectory correction device, characterized in that, include: The data acquisition unit is used to acquire meteorological data of the target area, as well as lightning location data for lightning and ground flash activity within the target area; the meteorological data includes satellite data and radar data. The initial centroid trajectory determination unit is used to determine the initial centroid trajectory of the lightning based on the lightning location data. The trajectory correction generation unit is used to generate a first centroid correction trajectory corresponding to the satellite data and a second centroid correction trajectory corresponding to the radar data. The trajectory correction strategy output unit is used to evaluate the first centroid correction trajectory and the second centroid correction trajectory according to the pre-built trajectory evaluation model, and output the trajectory correction strategy. The centroid trajectory correction unit is used to correct the initial centroid trajectory of lightning according to the trajectory correction strategy, so as to obtain the corrected centroid trajectory of lightning activity in the target area.

9. An electronic device, characterized in that, The device includes a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the lightning and ground flash activity trajectory correction method according to any one of claims 1-7 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program code for executing the lightning-to-ground lightning activity trajectory correction method according to any one of claims 1-7.

Images