Severe convective weather early warning method and system based on monomer identification and lightning jump

Abstract

The invention provides a severe convective weather early warning method and system based on monomer identification and lightning jump, and is applied to the technical field of data processing. The method comprises the following steps: processing high-precision lightning positioning data and meteorological radar data to generate convection monomer characteristic parameter information; processing the ground observation data to generate ground meteorological element associated information; processing the convective monomer characteristic parameter information and the ground meteorological element association information to generate severe convective monomer identification result information; processing the high-precision lightning positioning data based on the strong convection monomer identification result information to generate lightning jump judgment information; and processing the severe convection monomer identification result information and the lightning jump judgment information based on the target severe convection weather early warning model to generate severe convection weather early warning result information.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a severe convective weather early warning method and system based on single-body identification and lightning surge. Background Art

[0002] Severe convective weather, often accompanied by heavy rainfall, hail, and lightning, is characterized by sudden onset, short duration, and small spatial scale. The complexity of this type of weather makes early warning and forecasting extremely challenging. However, accurate early warning is crucial for protecting people's lives and property and reducing disaster losses. It is also indispensable for many areas, including agricultural production, air transportation, and urban operations.

[0003] Current severe convective weather warning technology still has many shortcomings. From the perspective of data acquisition, different observation equipment varies in data accuracy, coverage, and temporal resolution. For example, some lightning location networks lack detailed observations of cloud-to-cloud lightning discharge processes, affecting the comprehensiveness and accuracy of the data. In terms of algorithmic models, existing models struggle to accurately integrate multi-source data, and their simulation and prediction of severe convective weather under complex terrain and climatic conditions are ineffective. This limits the accuracy and lead time of warnings, making them unable to meet actual business needs. Existing models struggle to accurately account for these factors, leading to deviations in warnings.

[0004] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore includes information that does not constitute prior art known to ordinary technicians in this field. Summary of the Invention

[0005] The purpose of this application is to provide a severe convective weather warning method and system based on single-body identification and lightning jump, which at least overcomes the problems existing in the prior art to a certain extent. By integrating high-precision lightning location data, meteorological radar data, ground observation data, historical severe convective weather case data, and topographic and climate characteristic information, a dual-threshold TITAN algorithm is used to process information related to convective single bodies and ground meteorological elements to generate a severe convective single body identification result. Based on this result and the lightning location data, the lightning jump algorithm is used to obtain lightning jump judgment information and determine the lightning jump situation. The target model combines the strong convective single body identification and lightning jump judgment information to generate a risk assessment value, adjust the warning decision parameters, and obtain the severe convective weather warning result after analytical conversion, thereby improving the accuracy and timeliness of severe convective weather warnings, making up for the shortcomings of the prior art, and providing more reliable support for meteorological warning services.

[0006] Other features and advantages of the present application will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0007] According to one aspect of the present application, a severe convective weather warning method based on single-unit identification and lightning surge is provided, comprising: obtaining high-precision lightning location data, weather radar data, ground observation data, historical severe convective weather case data, topographic and geomorphological information of different regions, and climate characteristic information; Based on historical severe convective weather case data and topographic and geomorphological information, as well as climate characteristic information of different regions, high-precision lightning positioning data and meteorological radar data are processed to generate convective cell characteristic parameter information; based on historical severe convective weather case data and topographic and geomorphological information, as well as climate characteristic information of different regions, ground observation data are processed to generate ground meteorological element correlation information; convective cell characteristic parameter information and ground meteorological element correlation information are processed to generate severe convective cell identification result information; based on the strong convective cell identification result information, high-precision lightning positioning data are processed to generate lightning jump judgment information, wherein the lightning jump judgment information is used to characterize whether a lightning jump occurs, the jump time and location; based on the target severe convective weather warning model, the strong convective cell identification result information and the lightning jump judgment information are processed to generate severe convective weather warning result information.

[0008] Another aspect of the present application is a severe convective weather warning device based on monomer identification and lightning jump, which is characterized by comprising: an acquisition module for acquiring high-precision lightning positioning data, meteorological radar data, ground observation data, historical severe convective weather case data, topographic and geomorphic information of different regions, and climate characteristic information; a processing module for processing the high-precision lightning positioning data and meteorological radar data based on the historical severe convective weather case data and the topographic and geomorphic information of different regions, and climate characteristic information to generate convective monomer characteristic parameter .... The ground observation data is processed based on topographic and morphological information and climate characteristic information to generate ground meteorological element correlation information; the convective cell characteristic parameter information and ground meteorological element correlation information are processed to generate strong convective cell identification result information; based on the strong convective cell identification result information, the high-precision lightning positioning data is processed to generate lightning jump judgment information, wherein the lightning jump judgment information is used to characterize whether a lightning jump occurs, the jump time and location; based on the target severe convective weather warning model, the strong convective cell identification result information and the lightning jump judgment information are processed to generate severe convective weather warning result information.

[0009] According to another aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a second processor, the computer program implements the above-mentioned severe convective weather warning method based on single-body identification and lightning surge.

[0010] The present application provides a severe convective weather warning method and system based on monomer identification and lightning jump. The server extracts weather characteristics from historical cases, quantifies topography and climate information, extracts corresponding features from lightning location and radar data respectively, and then fuses them to generate convective monomer characteristic parameters. It generates meteorological element correlation information from ground observation data, laying the foundation for subsequent analysis. In the monomer identification and lightning jump judgment link, the dual-threshold TITAN algorithm is used to process the information related to convective monomers and ground meteorological elements to generate a strong convective monomer identification result. Based on this result and the lightning location data, the lightning jump judgment information is obtained by the lightning jump algorithm to determine the lightning jump situation. The target model combines the strong convective monomer identification and lightning jump judgment information to generate a risk assessment value, adjust the warning decision parameters, and obtain the severe convective weather warning results after analytical conversion, including the occurrence time, affected area, weather type and warning credibility, providing strong support for meteorological warnings.

[0011] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A flowchart of a severe convective weather warning method based on single-cell identification and lightning surge provided by one embodiment of the present application is shown; Figure 2 A schematic structural diagram of a severe convective weather warning device based on single-unit identification and lightning surge provided by an embodiment of the present application is shown; Figure 3 A schematic diagram of an effect based on monomer recognition and lightning jump provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0013] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0014] The following combination Figure 1 The severe convective weather warning method based on monomer identification and lightning surge according to the exemplary embodiment of the present application is described below. Figure 1 As shown, the method is applied to the server and includes: S101, obtain high-precision lightning location data, weather radar data, ground observation data, historical severe convective weather case data, topographic and geomorphological information of different regions, and climate characteristic information.

[0015] In one embodiment, a lightning location system, such as the Beijing EarthNet high-precision lightning location network, is used to obtain high-precision lightning location data. This system monitors lightning occurrence information in real time through multiple detection stations distributed in different locations. For example, during severe convective weather in Beijing from 2:00 PM to 4:00 PM on July 10, 2023, the lightning location system recorded 20 lightning strikes in a certain area of Haidian District (latitude and longitude: 39.95°N, 116.30°E) within 5 minutes from 2:20 PM to 2:25 PM. The lightning intensity ranged from 5 to 10 kA. The lightning frequency was also monitored to suddenly increase at 2:22 PM, from 2 strikes per minute to 5 strikes per minute. These data accurately reflect the changes in lightning location, intensity, and frequency. Meteorological radar data is collected using an S-band radar. During the severe convective weather event in Beijing on July 10, 2023, the S-band radar scanned the area at regular intervals (e.g., 5 minutes). Radar data showed that over Haidian District, the combined reflectivity reached 45dBZ, with strong echoes, a clumpy shape, and a southeasterly velocity of 30 kilometers per hour. These data provide key information for analyzing the cloud structure, precipitation particle concentration, and movement trends of severe convective weather systems.

[0016] Ground observation data was collected from multiple ground-based meteorological stations throughout Beijing. At 2:00 PM on July 10, 2023, a station in Haidian District recorded a temperature of 30°C, an air pressure of 1005 hPa, a humidity of 60%, precipitation of 10 mm in the previous hour, and a wind speed of 5 m / s from the south. These data reflect the ground-based meteorological conditions at the time of the severe convective weather event and are important for understanding the formation and development of severe convective weather. Historical severe convective weather data was obtained from the meteorological database. For example, a query revealed that on August 5, 2018, a severe convective weather event occurred in Chaoyang District, Beijing. The event, characterized by squall line-like winds, occurred at 4:30 PM and lasted approximately one hour. The event affected most of Chaoyang District (roughly between 39.85° and 40.00°N and 116.40° and 116.60°E). The disaster reached moderate intensity, causing some trees to fall and billboards to be damaged. These historical data provide a reference for analyzing the current severe convective weather and help discover the patterns and characteristics of severe convective weather.

[0017] For Beijing, topographic and geomorphological information is obtained using a geographic information system (GIS). Mountainous areas (such as parts of Yanqing District and Huairou District) have a highly undulating terrain, with slopes ranging from 15° to 30° and elevation differences reaching hundreds of meters. Mountainous terrain is predominant. Plain areas (such as parts of Tongzhou District and Daxing District) are relatively flat, with slopes less than 5°, and plains are the primary landform. This topographic and geomorphological information can influence the development and propagation of severe convective weather. For example, the undulating terrain in mountainous areas can enhance upward air movement, promoting the development of severe convection. Long-term meteorological observations and data analysis provide information on the climate characteristics of different regions in Beijing. In summer, due to the urban heat island effect, the average temperature in Beijing's central urban areas (such as Dongcheng District and Xicheng District) is 2-3°C higher than in the suburbs, with relatively low precipitation variability and a dominant southerly wind direction. In contrast, the northern mountainous areas (such as parts of Miyun District) experience relatively low average temperatures, high precipitation variability, and wind speeds reaching extremes of 10-15 m / s. These climate characteristic information are closely related to the occurrence and development of severe convective weather, and play an important role in analyzing the formation mechanism and development trend of severe convective weather.

[0018] S102: Based on historical severe convective weather case data and topographic and geomorphological information and climate characteristic information of different regions, high-precision lightning location data and weather radar data are processed to generate convective cell characteristic parameter information.

[0019] In one implementation, feature extraction is performed on historical severe convective weather data to generate characteristics for the severe convective weather type, occurrence time, duration, impact area, and disaster intensity. Data on severe convective weather events in the region over the past five years were obtained from a meteorological database, using a typical squall line event as an example. Analysis determined that the severe convective weather type was a squall line, a type of weather often accompanied by strong winds, heavy rainfall, and lightning. The occurrence time was 2:30 PM on August 15, 2023, accurately recording the onset of the weather. The duration indicated that the squall line lasted approximately 1.5 hours, from 2:30 PM to approximately 4:00 PM. The impact area, through geographic annotation, revealed that it affected three areas, A, B, and C, in the region, encompassing the city center and parts of the suburbs. According to relevant meteorological disaster assessment standards, the squall line caused some trees to fall and simple buildings to be damaged, resulting in a moderate disaster intensity.

[0020] Quantitative analysis and processing of topographic and geomorphological information and climate characteristics for different regions were performed to generate quantitative factors for topographic relief, landform type, temperature variation, precipitation distribution, and wind speed and vector. The topographic and geomorphological information for the region was quantitatively analyzed using a geographic information system (GIS). Area A in the region is mountainous, with significant topographic relief, an average slope of 20°, and elevation differences ranging from 300 to 500 meters. This yielded a quantitative factor for topographic relief. Area A's topography is primarily mountainous, and the quantitative factor for landform type was determined based on its correlation with severe convective weather. Long-term meteorological observations revealed that the average temperature during the same period of severe convective weather was 2°C higher than normal, resulting in a quantitative factor for temperature variation. Analysis of precipitation data for the region revealed significant precipitation variation over the past 24 hours, resulting in a quantitative factor for precipitation distribution. Combined with wind direction and speed monitoring data, the dominant wind direction was southwesterly, with a maximum wind speed of 12 m / s, resulting in a quantitative factor for wind speed and vector.

[0021] Perform feature extraction and processing on high-precision lightning location data to generate lightning frequency features, lightning intensity features, lightning location features, and lightning jump features. Figure 3 As shown in the figure, a high-precision lightning location system was used to monitor lightning activity between 18:30 and 20:00 on June 4, 2022. The system recorded that in area A, the lightning frequency characteristic was that 15 lightning strikes occurred within the 10 minutes between 18:30 and 18:40. The lightning intensity characteristics showed that most lightning strikes were between 5 and 8 kA. The lightning location characteristics were accurate to the longitude and latitude. For example, at 19:05, a lightning strike occurred at 30.5°N and 110.2°E. In terms of lightning jump characteristics, at 19:02, the lightning frequency suddenly increased from 1 per minute to 4 per minute, showing a significant jump phenomenon. Feature extraction and processing were performed on the meteorological radar data to generate radar reflectivity characteristics, echo intensity characteristics, echo shape characteristics, and echo movement speed characteristics. During the same period, an S-band weather radar scanned the area. Radar reflectivity characteristics showed that the combined reflectivity reached 50 dBZ over Area A, where the squall line passed, indicating a high concentration of precipitation particles in the cloud layer. Echo intensity characteristics showed strong echoes, exceeding 40 dBZ, reflecting the high intensity of precipitation. The echo shape exhibited a typical banded shape, a common feature of squall line weather on radar echoes. Echo velocity characteristics indicated that the squall line echo was moving northeastward at a speed of 35 kilometers per hour.

[0022] Based on the quantification factors of terrain relief, landform type, temperature variation, precipitation distribution, and wind speed and vector, lightning frequency, intensity, location, and surge characteristics, as well as radar reflectivity, echo intensity, echo shape, and echo velocity, were fused to generate comprehensive lightning and radar quantitative characteristics. Considering the mountainous terrain relief and landform type of Region A, combined with temperature variation and precipitation distribution, it was found that lightning in areas with greater terrain relief is more likely to occur in areas of rising airflow, resulting in a certain correlation between lightning frequency and the quantification factors of terrain relief. These quantitative factors were fused with lightning frequency, intensity, location, and surge characteristics to generate a comprehensive lightning quantitative characteristic. Similarly, for radar data, terrain, landform, and climate characteristics affect the propagation and reflection of radar waves. In mountainous areas, terrain relief can alter radar reflectivity and echo shape characteristics. Combined with other quantitative factors, radar reflectivity, echo intensity, shape, and velocity characteristics were fused to generate a comprehensive radar quantitative characteristic.

[0023] Based on the characteristics of severe convective weather types, occurrence time, duration, impact area, and disaster intensity, lightning and radar comprehensive quantitative characteristics were analyzed and processed to generate convective cell characteristic parameters. These parameters characterize the intensity, development trend, and correlation with the surrounding environment of convective cells. The convective cell intensity was found to be relatively strong during this squall line, as evidenced by the high frequency and intensity of lightning comprehensive quantitative characteristics and the high reflectivity and strong echo intensity of radar comprehensive quantitative characteristics. Its development trend was northeastward and gradually weakened over time, consistent with the radar echo velocity characteristics and changes in lightning activity. Regarding the correlation with the surrounding environment, the topography and climate characteristics of Area A, such as the mountainous terrain that promotes upward airflow, provide conditions for the formation and development of convective cells, resulting in a close correlation between convective cells and the surrounding environment. Through these analyses, the convective cell characteristic parameters were generated, accurately characterizing the characteristics of convective cells during this squall line.

[0024] S103: Based on the historical severe convective weather case data and the topographic and geomorphological information and climate characteristic information of different regions, the ground observation data is processed to generate ground meteorological element correlation information.

[0025] In one embodiment, the topographic and geomorphic information and climate characteristic information of different regions are quantitatively analyzed and processed to generate a terrain height quantification factor, a terrain slope quantification factor, a geomorphic type quantification factor, an average temperature quantification factor, a precipitation variability quantification factor, a dominant wind vector quantification factor, and a wind speed extreme value quantification factor. Assuming that when studying severe convective weather in a certain area, areas A and B in the area are used as examples, where area A is a mountainous area and area B is a plain. Geographic Information System (GIS) technology is used to obtain the topographic and geomorphic information of areas A and B. Area A has a large terrain with an average altitude of 800 meters, which is significantly different from the surrounding areas. Based on this, a terrain height quantification factor is generated to measure its impact on ground meteorological elements. The average slope of area A reaches 25°. This large slope affects airflow movement and heat distribution, so a terrain slope quantification factor is generated. Its landform type is mainly mountainous, and the landform type quantification factor is determined. Statistical analysis of long-term meteorological observation data reveals that during the season of severe convective weather, the average temperature in Region A was 20°C, lower than the historical average temperature of 22°C for the same period. This yielded a mean temperature quantification factor. Precipitation variability in this region is high, with significant fluctuations in precipitation over the past month. This yielded a precipitation variability quantification factor. The prevailing wind direction is northwesterly, with extreme wind speeds reaching 15 m / s. This yielded a dominant wind direction quantification factor and a wind speed extreme value quantification factor. Region B has a relatively flat terrain with an average elevation of 50 meters above sea level, a slope of less than 5°, and a plain landform. Corresponding quantification factors for terrain height, slope, and landform type were also generated. Regarding climate characteristics, the average temperature is 23°C, precipitation variability is low, and the prevailing wind direction is southeasterly, with extreme wind speeds reaching 8 m / s. Corresponding quantification factors for mean temperature, precipitation variability, dominant wind direction, and wind speed extreme value were also generated.

[0026] Feature extraction and processing are performed on ground observation data to generate temperature, pressure, humidity, precipitation, and wind speed and direction characteristics. During a severe convective weather event, the ground observation station in Area A recorded the following data: temperature 18°C, pressure 1002hPa, humidity 65%, precipitation 15mm over the past hour, wind speed 8m / s, and northwesterly wind direction. The ground observation station in Area B recorded the following data: temperature 22°C, pressure 1005hPa, humidity 55%, precipitation 5mm over the past hour, wind speed 5m / s, and southeasterly wind direction. These data constitute the temperature, pressure, humidity, precipitation, and wind speed and direction characteristics of Areas A and B, respectively. The temperature and pressure characteristics are fused based on the quantitative factors of terrain height, terrain slope, and landform type to generate quantitative features related to terrain and thermal dynamics. In region A, the temperature (18°C) and air pressure (1002 hPa) characteristics were fused by combining quantitative factors for terrain height (800 meters above sea level), terrain slope (25°), and landform type (mountainous terrain). Due to the high elevation and steep slope of region A, air flow is relatively rapid, leading to frequent heat exchange and resulting in relatively low temperatures. Air pressure also fluctuates due to the topography. By considering the influence of topography on the vertical temperature lapse rate and the effect of topographic obstruction on air pressure distribution, a quantitative topographic-thermal correlation feature was generated. This feature reflects the combined influence of region A's topography on temperature and pressure. In region B, a corresponding quantitative topographic-thermal correlation feature was generated based on quantitative factors for terrain height (50 meters), slope (less than 5°), and landform type (plain terrain), as well as temperature (22°C) and pressure (1005 hPa). Due to the flat terrain of region B, the influence of topography on temperature and pressure is relatively small, resulting in a significantly different feature from that of region A.

[0027] Based on the quantification factors of mean temperature, precipitation variability, dominant wind direction, and wind speed extremes, humidity, precipitation, and wind speed and direction characteristics were fused to generate quantitative characteristics related to climate and dynamics. For region A, humidity (65%), precipitation (15 mm), and wind speed and direction (8 m / s, northwest wind) were fused based on the quantification factors of mean temperature (2°C lower than the historical period), precipitation variability (higher), dominant wind direction (northwest wind), and wind speed extremes (15 m / s). Under these climatic conditions, lower temperatures and greater precipitation variability affect humidity and precipitation, while dominant wind direction and wind speed extremes influence water vapor transport and air movement. By analyzing the interactions between these factors (e.g., using atmospheric dynamics models to consider the influence of wind on water vapor transport and precipitation formation), quantitative characteristics related to climate and dynamics were generated. The corresponding quantitative characteristics of the climate-dynamic relationship were generated for Region B based on its average temperature (23°C), precipitation variability (low), dominant wind direction (southeast wind), and wind speed extremes (8 m / s), as well as humidity (55%), precipitation (5 mm), and wind speed and direction (5 m / s, southeast wind). Region B's relatively stable temperature, low precipitation variability, and weak wind conditions make its quantitative characteristics of the climate-dynamic relationship different from those of Region A.

[0028] Based on the characteristics of severe convective weather types, occurrence time, duration, impact area, and disaster intensity, quantitative characteristics of topography-thermal and climate-dynamic correlations are analyzed and processed to generate surface meteorological element correlation information. This surface meteorological element correlation information characterizes the interrelationships between surface meteorological elements and their degree of correlation with topography, geomorphology, and climate. Assume that during this severe convective weather event, historical severe convective weather case data show that similar severe convective weather events in the region often occur during periods of convergence of cold and warm air, are short-lived, and primarily affect mountainous areas and surrounding plains, with disaster intensity generally ranging from mild to moderate. Combining these characteristics of severe convective weather types, occurrence time, duration, impact area, and disaster intensity, quantitative characteristics of topography-thermal and climate-dynamic correlations are analyzed and processed for regions A and B. In region A, the relationships between surface meteorological elements are more complex due to the topographic and climatic conditions reflected in its quantitative characteristics of topography-thermal and climate-dynamic correlations. For example, higher terrain leads to lower temperatures, which in turn affects the formation and distribution of air humidity and precipitation, while the prevailing wind direction and speed further influence the changes in these factors. Through comprehensive analysis, correlation information of ground meteorological elements in Region A is generated, reflecting the interrelationships between ground meteorological elements in this region and their close connection with topography, landforms, and climate characteristics. Region B also generates information reflecting the correlation of its ground meteorological elements based on its own characteristics. Compared to Region A, the interrelationships between ground meteorological elements in Region B are relatively weak, and their connections with topography, landforms, and climate characteristics are also different. For example, the flat terrain leads to relatively uniform distribution of temperature and air pressure, and the low variability of precipitation also makes the impact of precipitation on other factors more stable.

[0029] S104: Process the convective cell characteristic parameter information and ground meteorological element correlation information to generate strong convective cell identification result information.

[0030] In one embodiment, characteristic parameter information of a convective cell is extracted and classified to generate information on changes in convective cell intensity, echo morphology evolution, and abnormal lightning activity. Within a specific time period, the intensity change information for this convective cell showed a rapid increase in radar reflectivity from 30 dBZ to 50 dBZ within 30 minutes, indicating an increase in convective cell intensity. Regarding echo morphology evolution, the echo shape gradually merged from initially scattered small blocks into a larger block structure, with sharper edges. Anomaly information on lightning activity indicated that the frequency of lightning increased from 2 to 8 per minute within 10 minutes, indicating a significant increase in lightning activity. Correlated information on ground meteorological elements is extracted and classified to generate information on meteorological element mutations and abnormal correlations between elements. The meteorological element mutation information indicates that during the development of the convective cell, the temperature dropped by 5°C within an hour, and the air pressure also fluctuated significantly, dropping rapidly from 1005 hPa to 998 hPa. The abnormal correlation information between elements shows that the correlation between humidity and precipitation is abnormal. Normally, an increase in humidity is accompanied by an increase in precipitation, but in this process, there is no obvious change in precipitation while the humidity increases, which is inconsistent with normal meteorological laws.

[0031] The dual-threshold TITAN algorithm processes information on convective cell intensity changes, echo morphology evolution, lightning activity anomalies, meteorological element mutations, and inter-element correlation anomalies to generate severe convection probability information and severe convection development trend assessment information. The dual-threshold TITAN algorithm first sets two thresholds: a low threshold for preliminary identification of potential convective cell areas, and a high threshold for determining the core severe convection area. In this example, the low threshold is set at 35 dBZ, and the high threshold is set at 65 dBZ. The algorithm calculates the severe convection probability information, indicating an 80% probability that the current convective cell will develop into severe convective weather. Furthermore, analysis of the severe convection development trend assessment information indicates that the convective cell will continue to move northeastward and may further intensify, expected to reach its peak within the next 1-2 hours. Based on the severe convection probability information and severe convection development trend assessment information, target data from the convective cell characteristic parameter information and ground meteorological element correlation information is marked and filtered to generate potential severe convection cell data screening results. For example, areas with radar reflectivity greater than 45dBZ and lightning frequencies exceeding five per minute within convective cells are marked as key areas of concern. In ground-based meteorological information, areas with significant temperature drops and large pressure fluctuations are also flagged. By filtering these marked data, we obtain data screening results for potential severe convective cells, identifying key areas and relevant data that could potentially develop into severe convective cells.

[0032] The results of the data screening of potential strong convective cells are integrated and quantified to generate the strong convective cell identification result information, wherein the strong convective cell identification result information is used to characterize the existence, location, intensity and development trend of the strong convective cell. The characteristic parameter information of the convective cell and the related information of the ground meteorological elements are summarized, and the intensity index of the strong convective cell is calculated through a specific quantification algorithm. This index comprehensively considers multiple factors such as radar reflectivity, lightning frequency, temperature change, and air pressure fluctuation. For example, the analytic hierarchy process (AHP) is used to determine the weight of each factor and construct a judgment matrix. For example, the radar reflectivity weight is 0.4, the lightning frequency weight is 0.25, the temperature change weight is 0.15, and the pressure fluctuation weight is 0.2. These weights reflect the relative importance of each factor to the intensity of the strong convective cell. The data of each factor are standardized to eliminate the dimensional effect. Taking radar reflectivity as an example, assuming that its value range is 20-70dBZ, the radar reflectivity of 30dBZ is standardized using the normalization formula to be The same method is used to process lightning frequency, temperature change and air pressure fluctuation data.

[0033] The calculation formula of the strength index is ,in, is the strength index, is the normalized radar reflectivity, is the normalized lightning frequency, is the normalized temperature change, is the normalized pressure fluctuation. If the normalized data at a certain moment are , then the intensity index . According to the pre-set intensity level standard (such as 0-0.3 is weak, 0.3-0.7 is medium, and 0.7-1 is strong), the current intensity of this single cell is at a medium to high level. The location information is mainly obtained by combining radar echo location and lightning location data. The radar determines the spatial position of the convective cell echo through the principle of electromagnetic wave reflection, and its accuracy can reach the kilometer level. The lightning location system uses the time difference of multiple detection stations to locate the location of lightning, and its accuracy is also relatively high. By fusing and processing these data, the longitude and latitude coordinates can be accurately determined, and it is determined that the strong convective cell exists in the area of 30.5° north latitude and 110.2° east longitude.

[0034] By continuously monitoring the position information of convection cells at different times, such as the position of the cell at time t1 is (30.5°N, 110.2°E), and the position at time t2 is (30.7°N, 110.5°E), using the distance formula (The longitude and latitude here need to be converted to plane coordinates to calculate the distance.) Calculate the distance moved and divide it by the time interval to get a speed of 30 kilometers per hour. Based on the direction of the position change, it is determined to be moving northeast.

[0035] Analyze the changing trend of the intensity index over time. If the intensity index continues to rise at several consecutive monitoring moments, combined with meteorological principles, such as unstable energy accumulation, water vapor supply and other factors, it is judged that the intensity will tend to increase in the next 1-2 hours. Combining the above calculation and analysis results, the strong convective cell identification result information is generated. It shows that there is a strong convective cell in the area of 30.5°N, 110.2°E. The current intensity is at a medium to high level. It is moving toward the northeast at a speed of 30 kilometers per hour. The intensity will tend to increase in the next 1-2 hours. This information comprehensively and accurately characterizes the existence, location, intensity and development trend of strong convective cells, providing key basis for subsequent severe convective weather warnings.

[0036] S105: Process the high-precision lightning location data based on the strong convective cell identification result information to generate lightning surge judgment information.

[0037] In one embodiment, the identification results of a strong convective cell are extracted and classified to generate information on its movement speed, intensity change, and development trend. Analysis of the cell's location at different times revealed that it moved from (30.5°N, 110.2°E) to (30.7°N, 110.5°E) within one hour. The distance traveled was calculated using a distance formula and divided by the one-hour interval, yielding a speed of 30 kilometers per hour. Comparing the cell's intensity index at different times (such as the intensity index previously calculated by integrating radar reflectivity, lightning frequency, and other factors) revealed an increase from 0.5 to 0.6 over the past 30 minutes. This intensity change information is then generated, indicating an increase in cell intensity. Based on a comprehensive assessment of factors such as its movement speed, intensity change, and the surrounding meteorological environment, the cell's development trend is estimated to indicate continued intensification and northeastward movement over the next 1-2 hours. High-precision lightning location data is extracted and classified to generate information on lightning frequency changes, lightning energy distribution, and lightning location migration. Over the past 30 minutes, the frequency of lightning in the area was counted, revealing an increase from 5 to 8 per minute. This information was generated to track lightning frequency changes. The lightning location system captured energy data for each lightning strike. Analysis of this data revealed that lightning energy was primarily concentrated between 5 and 10 kA and was relatively evenly distributed. This information was then generated to track lightning energy distribution. The location of each lightning strike was also recorded, revealing a gradual shift in the lightning location toward the northeast. This information was then generated to track lightning location migration.

[0038] Based on the lightning jump algorithm, the monomer movement speed information, monomer intensity change information, monomer development trend information, lightning frequency change information, lightning energy distribution information, and lightning position migration information are processed to generate lightning jump probability information and lightning jump intensity assessment information. First, a basic lightning frequency threshold is set. (e.g. set to 6 times per minute) and a lightning frequency change rate threshold (For example, if it is set to 0.5, an increase of more than 50% in the lightning frequency per minute is considered a significant change.) Calculate the average rate of change R of lightning frequency in the current time period (such as the past 30 minutes). At the same time, calculate the current average lightning energy E and the energy change trend (for example, by calculating the difference in average energy in different time periods to determine whether the energy is increasing, decreasing or stable). If the current lightning frequency is greater than And R is greater than If the energy trend shows that the energy is increasing (assuming the energy increase is more than 10%), then the probability of lightning surge is high. For example, the current lightning frequency is 8 times per minute, which is greater than the threshold of 6 times. If the energy is greater than the threshold of 0.5 and the energy increase is 15%, the conditions are met and the probability of a lightning surge is 80%, indicating a high probability of a lightning surge. The lightning surge intensity is assessed based on the increase in lightning energy and the change in lightning frequency. If the lightning energy increase is between 10% and 20% and the lightning frequency change rate is between 0.5 and 1, the lightning surge intensity is assessed as moderate. Based on the above example, the lightning surge intensity is assessed as moderate.

[0039] Based on the lightning jump probability information and lightning jump intensity assessment information, the target data in the strong convective cell identification result information and high-precision lightning location data are marked and screened to generate potential lightning jump data screening results. If the lightning jump probability is greater than 70% and the jump intensity is medium or above, the relevant data in that area is marked as potential lightning jump data. For example, in the strong convective cell identification result information, the data within the current location of the cell and a certain range around it (such as within a radius of 5 kilometers) are marked; in the high-precision lightning location data, the lightning occurrence location and related parameter data that meet the above conditions are marked. By screening this marked data, the potential lightning jump data screening results are obtained, and the key areas and related data where lightning jumps may occur are determined.

[0040] The results of the potential lightning surge data screening are integrated and quantified to generate lightning surge judgment information. This information indicates whether a lightning surge has occurred, when it occurred, and where it occurred. The results of the potential lightning surge data screening are integrated and quantified. A comprehensive analysis is performed on information such as lightning frequency changes, energy distribution, and individual movement speed and intensity changes within the marked area. For example, a weight of 0.4 for lightning frequency changes, 0.3 for energy changes, and 0.15 for individual movement speed and intensity changes is used to calculate a comprehensive index. If this comprehensive index exceeds a certain threshold (e.g., 0.6), a lightning surge is determined to have occurred. Assuming the calculated comprehensive index is 0.7, combined with the time of the first occurrence meeting these conditions in the lightning location data, the lightning surge is determined to have occurred at the current time (e.g., 18:30) and at a location (30.6°N, 110.4°E). This generates lightning surge judgment information that clearly indicates the occurrence, time, and location of a lightning surge.

[0041] S106: Process the severe convective weather warning result information and the lightning surge judgment information based on the target severe convective weather warning model to generate severe convective weather warning result information.

[0042] In one embodiment, training data, verification data, and a preset initial model for severe convective weather warning are obtained. Data related to severe convective weather in the area over the past five years are collected as training data, including high-precision lightning location data, weather radar data, and ground observation data. At the same time, information such as the occurrence time, duration, impact range, and disaster intensity of each severe convective weather is recorded. Topographic and geomorphological information and climate characteristic information of different regions in the corresponding time period are obtained from the meteorological database. In addition, severe convective weather data from the past year are selected as verification data. The preset initial model for severe convective weather warning is a multi-layer perceptron (MLP) model, which includes an input layer, two hidden layers, and an output layer. The input layer is responsible for receiving external data, the hidden layer is used to extract features and perform nonlinear transformations on the data, and the output layer outputs the final prediction results. The number of neurons in the input layer is determined by the number of input features. Assuming that after the previous data processing, there are 20 input features, then the number of neurons in the input layer is 20; the first hidden layer is set with 30 neurons, and the ReLU activation function is used to introduce nonlinearity; the second hidden layer is set with 20 neurons, and the ReLU activation function is also used; the output layer has 3 neurons, corresponding to the possibility of severe convective weather (expressed as a probability between 0 and 1), intensity (divided into three levels: weak, moderate, and strong, expressed using one-hot encoding), and affected area (expressed as a longitude and latitude range, such as a two-dimensional vector). The output layer uses the Softmax activation function to handle classification problems.

[0043] The training data is cleaned, feature extracted, and normalized to generate pre-processed feature data. The pre-processed feature data includes meteorological feature data, convective cell feature data, lightning feature data, topographic and geomorphological information of different regions, and climate feature information. The training data is checked for outliers, such as obvious erroneous positioning points in the lightning positioning data. The locations of these points are seriously inconsistent with the surrounding lightning activity and geographical environment, and they are removed. For missing temperature and pressure data in the ground observation data, the average value of the historical data for the same period in the region is used to fill in the gaps. Radar reflectivity features are extracted from meteorological radar data, such as reflectivity values at different altitudes; echo intensity features are used to measure the strength of the echo; echo shape features are used to extract the geometric shape information of the echo through image recognition technology; and echo movement speed features are used to record the movement speed of the echo in different time periods.

[0044] Lightning frequency characteristics are extracted from high-precision lightning location data, counting the number of lightning strikes per unit time. Lightning intensity characteristics are also extracted to obtain the energy intensity of each lightning strike. Lightning location characteristics are accurate to the latitude and longitude. Lightning jump characteristics are calculated using the 2σ lightning jump algorithm. Temperature, pressure, humidity, precipitation, and wind speed and direction characteristics are extracted from ground observation data. Combined with historical severe convective weather data, characteristics of severe convective weather types (such as squall lines and hail) are extracted, along with occurrence time, duration, impact range, and disaster intensity. Topographic and geomorphic information is quantitatively analyzed to generate a quantification factor for topographic relief, such as the standard deviation of altitude within a region; a quantification factor for topographic type, which numerically encodes different landform types, such as mountains and plains; a quantification factor for topographic height, expressed as the average altitude of the region; and a quantification factor for topographic slope, using measured values of topographic slope.

[0045] Quantitative analysis of climate characteristic information is performed to generate temperature change quantitative factors to compare the temperature change amplitudes during the same period in history; precipitation distribution quantitative factors to analyze the differences in precipitation distribution in different regions and times; average temperature quantitative factors to calculate the average temperature in a specific time period; precipitation variability quantitative factors to measure the stability of precipitation changes; dominant wind vector quantification factors to count the frequency of the main wind directions; wind speed extreme value quantitative factors to obtain the maximum wind speed. The extracted characteristic data are normalized to be in the range of 0-1. For radar reflectivity data, assuming that its value range is 10-60dBZ, the radar reflectivity of 40dBZ at a certain moment is normalized to . Similarly, other characteristic data are normalized to generate pre-processed meteorological characteristic data, convective cell characteristic data, lightning characteristic data, topographic information of different regions, and climate characteristic information.

[0046] The preprocessed feature data is processed to generate the early warning model's prediction parameters. The early warning model's prediction parameter vector represents the prediction strategy and related parameter information for the likelihood, intensity, and impact area of severe convective weather. The model first receives this data at the input layer, which is then processed sequentially through two hidden layers. Within the hidden layers, neurons extract features and perform nonlinear transformations on the data using weighted computation and activation functions (ReLU). After multiple iterations of training, the backpropagation algorithm continuously adjusts the connection weights and biases between neurons to better fit the training data. Ultimately, the early warning model's prediction parameter vector is generated. This vector contains the model's prediction strategy and related parameter information for the likelihood, intensity, and impact area of severe convective weather. For example, the prediction strategy determines the probability of severe convective weather based on different feature combinations. Related parameter information includes the connection weights and bias values between neurons in each layer. Based on the early warning model's prediction parameter vector, the pre-set severe convective weather warning initial model is optimized and trained to generate the trained severe convective weather warning model. During training, model parameters, such as the learning rate (assuming the initial learning rate is set to 0.01), are continuously adjusted. Through repeated experiments, an appropriate learning rate is found that results in faster model convergence and better results. Hyperparameters, such as the number of hidden layer neurons and the activation function, are also adjusted to observe the model's performance on the training data. After multiple rounds of training, a trained severe convective weather warning model is obtained.

[0047] The trained severe convective weather warning model is then simulated using validation data to generate validation results. Based on the generated prediction parameter vector of the warning model, the pre-set initial multilayer perceptron model is optimized and trained. During the training process, model parameters, such as the learning rate (assuming the initial learning rate is set to 0.01), are continuously adjusted. Through repeated experiments, an appropriate learning rate is found that results in faster model convergence and better performance. Hyperparameters such as the number of hidden layer neurons and activation function are also adjusted, and the model's performance on the training data is observed. After multiple rounds of training, a trained severe convective weather warning model is obtained. The trained severe convective weather warning model is then simulated using validation data. Various feature data from the validation data are input into the trained model, which then outputs predictions of the likelihood, intensity, and impact area of severe convective weather. For example, the model predicts a 0.8 probability of severe convective weather occurring in a certain area at a certain time, with moderate intensity (represented by one-hot encoding as [0,1,0]), and predicts the longitude and latitude of the affected area to be (30.0°-30.5°N, 110.0°-110.5°E). These predictions are compared with the actual severe convective weather conditions in the validation data to generate a validation result. If severe convective weather does occur in the area, with moderate intensity, and the affected area is (30.1°-30.4°N, 110.1°-110.4°E), validation metrics such as the prediction accuracy and margin of error can be calculated.

[0048] Based on the validation results, the trained severe convective weather warning model is evaluated and adjusted to generate a target severe convective weather warning model. Evaluation metrics such as the model's hit rate (e.g., the proportion of correct predictions to total predictions), false alarm rate (the proportion of incorrect predictions resulting in warnings), and critical success index are calculated. A low hit rate and a high false alarm rate indicate that the model may be overfitting or underfitting. For example, if the model performs well on the training data but has a low hit rate on the validation data, it may be overfitting. Adjustments to the model can be made, such as increasing the amount of training data, reselecting the number of hidden layer neurons and activation functions, or adjusting hyperparameters such as the learning rate. After multiple evaluations and adjustments, a high-performing target severe convective weather warning model is obtained. This model can more accurately predict the time of occurrence, affected area, weather type, and quantitative indicators of warning credibility for severe convective weather, providing reliable support for severe convective weather warnings.

[0049] In another embodiment, the target severe convective weather warning model analyzes and processes the severe convective cell identification information and lightning surge determination information to generate a severe convective weather risk assessment value. A severe convective cell is currently monitored, located at 30.5°N, 110.2°E, with moderate to high intensity and a significant lightning surge. The lightning surge occurred at 18:30. These severe convective cell identification information and lightning surge determination information are input into the target severe convective weather warning model. The model analyzes this information based on its internal preset algorithms and parameters. For example, the model considers factors such as the frequency of past severe convective weather events in the area, the current convective cell development trend (such as the rate of intensity increase, direction, and speed), and the magnitude and duration of the lightning surge. After complex calculations and analysis, the model generates a severe convective weather risk assessment value. Assuming the risk assessment value is expressed on a scale of 0-10, with 0 representing no risk and 10 representing extremely high risk, the calculated risk assessment value is 7, indicating a high risk of severe convective weather in the area.

[0050] Based on the severe convective weather risk assessment value, the warning decision parameter vector within the target severe convective weather warning model is processed to generate a warning deviation correction vector. After obtaining a severe convective weather risk assessment value of 7, the target severe convective weather warning model adjusts the internal warning decision parameter vector accordingly. The warning decision parameter vector contains various parameters used by the model when issuing warnings, such as warning thresholds for different risk levels and adjustment parameters for warning lead times. The model compares the current risk assessment value of 7 with the preset risk threshold. Suppose the preset high risk threshold is 6, and the current risk assessment value exceeds this threshold. The model processes the warning decision parameter vector according to preset rules. For example, if the risk assessment value is high, the model will increase the warning lead time adjustment parameter accordingly to ensure a more timely warning. After processing, a warning deviation correction vector is generated. This vector contains specific information about the warning decision parameter adjustments, such as whether the warning lead time needs to be increased by 30 minutes or the warning intensity level needs to be raised by one level.

[0051] The warning deviation correction vector is parsed and converted to generate severe convective weather warning results. This information characterizes the occurrence time, impact area, weather type, and quantitative indicators of the warning's credibility. Assume that the warning deviation correction vector includes information for increasing the warning lead time by 30 minutes and raising the warning intensity level by one. The final severe convective weather warning results are generated based on the model's previously established correspondence between warning intensity levels and weather types (for example, intensity levels 1-3 correspond to weak convective weather, 4-6 correspond to moderate convective weather, 7-9 correspond to severe convective weather, and 10 corresponds to extremely severe convective weather), as well as the current monitoring time and the convective cell's movement speed and direction. Assume that the current time is 18:30 and the convective cell is moving northeast at a speed of 30 kilometers per hour. After analysis and conversion, the generated warning results are as follows: severe convective weather is expected to occur around 7:00 PM, affecting a 50-kilometer radius northeast of the current location (30.5°N, 110.2°E). The weather type is severe convective weather (including heavy precipitation and lightning), and the warning's credibility is quantified at 80%. This information is displayed to meteorological personnel in a specific manner so they can take appropriate warning measures.

[0052] The server extracts weather characteristics from historical cases, quantifying topographical and climatic information. It then extracts corresponding features from lightning location and radar data, fusing them to generate convective cell characteristic parameters. Ground observation data is used to generate meteorological element correlation information, laying the foundation for subsequent analysis. During cell identification and lightning surge assessment, the dual-threshold TITAN algorithm processes information related to convective cells and ground meteorological elements to generate a strong convective cell identification result. Based on this result and lightning location data, the lightning surge algorithm generates lightning surge assessment information to determine the lightning surge situation.

[0053] In terms of warning model construction and application, training and validation data and an initial model are obtained. After data cleaning, feature extraction, and normalization, the model is trained to generate a warning model prediction parameter optimization model. This model is then verified and adjusted to obtain the target model. Finally, the target model combines information on severe convective cell identification and lightning surge prediction to generate a risk assessment value, adjust warning decision parameters, and analyze and transform the data to produce severe convective weather warning results, including the time of occurrence, affected area, weather type, and warning credibility, providing strong support for meteorological warnings.

[0054] In one embodiment, Figure 2 As shown, the present application also provides a severe convective weather warning device based on single-body identification and lightning surge, comprising: Acquisition module 201 is used to obtain high-precision lightning location data, weather radar data, ground observation data, historical severe convective weather case data, topographic and geomorphological information of different regions, and climate characteristic information; Processing module 202 is used to process high-precision lightning positioning data and meteorological radar data based on historical severe convective weather case data and topographic and geomorphic information and climate characteristic information of different regions to generate convective cell characteristic parameter information; process ground observation data based on historical severe convective weather case data and topographic and geomorphic information and climate characteristic information of different regions to generate ground meteorological element correlation information; process convective cell characteristic parameter information and ground meteorological element correlation information to generate severe convective cell identification result information; process high-precision lightning positioning data based on the strong convective cell identification result information to generate lightning jump judgment information, wherein the lightning jump judgment information is used to characterize whether a lightning jump occurs, the time and location of the jump; process the strong convective cell identification result information and the lightning jump judgment information based on the target severe convective weather warning model to generate severe convective weather warning result information.

[0055] Each embodiment in this application is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the evaluation of the severe convective weather warning method based on single-body identification and lightning jump, the electronic device, electronic device, and readable storage medium embodiment are basically similar to the severe convective weather warning method embodiment based on single-body identification and lightning jump described above, so the description is relatively simple, and the relevant parts can be referred to the partial description of the severe convective weather warning method embodiment based on single-body identification and lightning jump described above.

Claims

1. A severe convective weather warning method based on single-body identification and lightning surge, characterized in that: include: Obtain high-precision lightning location data, weather radar data, ground observation data, historical severe convective weather data, topographic and geomorphological information of different regions, and climate characteristics information; Based on historical severe convective weather data and topographic and climatic information of different regions, high-precision lightning location data and meteorological radar data are processed to generate convective cell characteristic parameter information; Based on historical severe convective weather case data and topographic and geomorphological information and climate characteristics of different regions, ground observation data is processed to generate ground meteorological element correlation information; Process the convective cell characteristic parameter information and ground meteorological element correlation information to generate strong convective cell identification result information; Based on the strong convective cell identification results, high-precision lightning location data is processed to generate lightning surge judgment information, where the lightning surge judgment information is used to indicate whether a lightning surge has occurred, the surge time, and the location. Based on the target severe convective weather warning model, the severe convective cell identification result information and lightning surge judgment information are processed to generate severe convective weather warning result information.

2. The method according to claim 1, wherein Based on historical severe convective weather data and information on topography, geomorphology, and climate characteristics in different regions, high-precision lightning location data and weather radar data are processed to generate convective cell characteristic parameter information, including: Perform feature extraction and processing on historical severe convective weather case data to generate severe convective weather type characteristics, occurrence time characteristics, duration characteristics, impact range characteristics, and disaster intensity characteristics; Quantitative analysis and processing of topographic and geomorphic information and climate characteristic information of different regions are performed to generate quantitative factors of terrain relief, landform type, temperature change, precipitation distribution, and wind speed and vector; Perform feature extraction and processing on high-precision lightning location data to generate lightning frequency features, lightning intensity features, lightning location features, and lightning jump features; Perform feature extraction and processing on weather radar data to generate radar reflectivity features, echo intensity features, echo shape features, and echo moving speed features; Based on the quantitative factors of terrain relief, landform type, temperature change, precipitation distribution, and wind speed and vector, the lightning frequency, intensity, location, and surge characteristics, as well as the radar reflectivity, echo intensity, shape, and speed characteristics, are fused to generate comprehensive lightning and radar quantitative features. Based on the characteristics of severe convective weather types, occurrence time, duration, impact range, and disaster intensity, the comprehensive quantitative characteristics of lightning and radar are analyzed and processed to generate convective cell characteristic parameter information. The convective cell characteristic parameter information is used to characterize the intensity, development trend, and degree of correlation with the surrounding environment of the convective cell.

3. The method according to claim 2, wherein Based on historical severe convective weather case data and topographic and geomorphological information and climate characteristics of different regions, ground observation data is processed to generate ground meteorological element correlation information, including: Quantitative analysis and processing of topographic and geomorphic information and climate characteristic information of different regions are performed to generate quantitative factors of terrain height, terrain slope, landform type, average temperature, precipitation variability, dominant wind vector, and wind speed extreme value; Perform feature extraction and processing on ground observation data to generate temperature features, pressure features, humidity features, precipitation features, and wind speed and direction features; Based on the quantitative factors of terrain height, terrain slope and landform type, the temperature characteristics and pressure characteristics are fused to generate quantitative characteristics of terrain and thermal correlation. Based on the quantification factors of average temperature, precipitation variability, dominant wind direction and wind speed extremes, the humidity characteristics, precipitation characteristics and wind speed and direction characteristics are integrated to generate quantitative characteristics of climate and dynamic correlation. Based on the characteristics of severe convective weather types, occurrence time, duration, impact range, and disaster intensity, the quantitative characteristics of the correlation between terrain and thermal dynamics and the quantitative characteristics of the correlation between climate and dynamics are analyzed and processed to generate ground meteorological element correlation information. The ground meteorological element correlation information is used to characterize the relationship between ground meteorological elements and the degree of correlation with terrain, landform, and climate characteristics.

4. The method according to claim 1, wherein The convective cell characteristic parameter information and ground meteorological element correlation information are processed to generate strong convective cell identification result information, including: Extract and classify the characteristic parameter information of convective cells to generate information on convective cell intensity changes, echo morphology evolution, and lightning activity anomalies; Extract and classify the ground meteorological element correlation information to generate meteorological element mutation information and abnormal correlation information between elements; Based on the dual-threshold TITAN algorithm, the system processes information on changes in convective cell intensity, echo morphology evolution, lightning activity anomalies, meteorological element mutations, and abnormal correlations between elements to generate information on the probability of severe convection and the assessment of its development trend. Based on the severe convection probability information and severe convection development trend assessment information, target data in the convective cell characteristic parameter information and ground meteorological element correlation information are marked and screened to generate potential severe convection cell data screening results; The data screening results of potential strong convective cells are integrated and quantified to generate strong convective cell identification result information, which is used to characterize the existence, location, intensity and development trend of strong convective cells.

5. The method according to claim 4, wherein Based on the strong convective cell identification information, high-precision lightning location data is processed to generate lightning surge judgment information, including: Extract and classify the strong convective cell identification result information to generate cell movement speed information, cell intensity change information, and cell development trend information; Extract and classify high-precision lightning location data to generate information on lightning frequency changes, lightning energy distribution, and lightning location migration; Based on the lightning jump algorithm, the lightning cell movement speed information, lightning cell intensity change information, lightning cell development trend information, lightning frequency change information, lightning energy distribution information, and lightning position migration information are processed to generate lightning jump probability information and lightning jump intensity assessment information; Based on the lightning jump probability information and lightning jump intensity assessment information, the target data in the strong convective cell identification result information and high-precision lightning location data are marked and screened to generate potential lightning jump data screening results; The potential lightning surge data screening results are integrated and quantified to generate lightning surge judgment information, where the lightning surge judgment information is used to characterize whether a lightning surge occurs, the time of the surge, and the location of the surge.

6. The method according to claim 1, wherein Obtain the target severe convective weather warning model, including: Obtain training data, verification data, and preset initial severe convective weather warning models; Perform data cleaning, feature extraction, and normalization on the training data to generate preprocessed feature data, where the preprocessed feature data includes meteorological feature data, convective cell feature data, lightning feature data, topographic and geomorphological information of different regions, and climate feature information; Process the pre-processed feature data to generate early warning model prediction parameters, where the early warning model prediction parameter vector is used to characterize the prediction strategy and related parameter information for the possibility, intensity, and impact area of severe convective weather; Based on the prediction parameter vector of the warning model, the preset severe convective weather warning initial model is optimized and trained to generate a trained severe convective weather warning model; Based on the verification data, the trained severe convective weather warning model is simulated to generate verification results; Based on the verification results, the trained severe convective weather warning model is evaluated and adjusted to generate the target severe convective weather warning model.

7. The method according to claim 1, wherein Based on the target severe convective weather warning model, the severe convective cell identification result information and lightning surge judgment information are processed to generate severe convective weather warning result information, including: Based on the target severe convective weather warning model, the severe convective cell identification result information and lightning surge judgment information are analyzed and processed to generate a severe convective weather risk assessment value; Based on the severe convective weather risk assessment value, the warning decision parameter vector within the target severe convective weather warning model is processed to generate a warning deviation correction vector; The warning deviation correction vector is parsed and converted to generate severe convective weather warning result information, where the severe convective weather warning result information is used to characterize the time of occurrence, affected area, weather type and warning credibility quantitative indicators of severe convective weather.

8. A severe convective weather warning device based on single-unit identification and lightning surge, characterized in that: The device comprises: The acquisition module is used to obtain high-precision lightning location data, weather radar data, ground observation data, historical severe convective weather data, topographic and geomorphological information of different regions, and climate characteristics information; The processing module is used to process high-precision lightning positioning data and meteorological radar data based on historical severe convective weather case data and topographic and geomorphic information and climate characteristic information of different regions to generate convective cell characteristic parameter information; process ground observation data based on historical severe convective weather case data and topographic and geomorphic information and climate characteristic information of different regions to generate ground meteorological element correlation information; process convective cell characteristic parameter information and ground meteorological element correlation information to generate severe convective cell identification result information; process high-precision lightning positioning data based on the strong convective cell identification result information to generate lightning jump judgment information, wherein the lightning jump judgment information is used to characterize whether a lightning jump occurs, the jump time and location; process the strong convective cell identification result information and the lightning jump judgment information based on the target severe convective weather warning model to generate severe convective weather warning result information.

9. An electronic device, characterized in that: include: a first processor; and a memory for storing executable instructions of the first processor; The first processor is configured to execute the severe convective weather warning method based on single-body identification and lightning surge according to any one of claims 1 to 7 by executing the executable instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the second processor, the severe convective weather warning method based on single-body identification and lightning surge as described in any one of claims 1 to 7 is implemented.

Images